Politics after Violence

Soifer_6844-final.indb i 8/17/18 11:53 AM

THIS PAGE INTENTIONALLY LEFT BLANK
 Politics after Violence
Legacies of the Shining
Path Conflict in Peru

Edit ed by Hillel Dav id Soifer

a nd Alberto V erga r a

University of Texas Press Austin

Soifer_6844-final.indb iii 8/17/18 11:53 AM

 Copyright © 2019 by the University of Texas Press
All rights reserved
Printed in the United States of America
First edition, 2019

Requests for permission to reproduce material from this work should be sent to:
Permissions
University of Texas Press
P.O. Box 7819
Austin, TX 78713–7819
utpress.utexas.edu/rp-form

The paper used in this book meets the minimum requirements of

ANSI/NISO Z39.48–1992 (R1997) (Permanence of Paper).

Library of Congress Cataloging-in-Publication Data

Names: Soifer, Hillel David, editor. | Vergara, Alberto, editor.
Title: Politics after violence : legacies of the Shining Path confl ict in Peru /
edited by Hillel David Soifer and Alberto Vergara.
Description: First edition. | Austin : University of Texas Press, 2019. | Includes
bibliographical references and index.
Identifiers: LCCN 2018012485 | ISBN 978-1-4773-1731-0 (cloth : alk. paper) |
ISBN 978-1-4773-1732-7 (library e-book) | ISBN 978-1-4773-1733-4
(nonlibrary e-book)
Subjects: LCSH: Sendero Luminoso (Guerrilla group) | Peru—Politics and
government—1980– | Political violence—Peru—History—20th century. |
Violence—Political aspects—Peru. | Violence—Social aspects—Peru. |
Peru—History—1980–
Classification: LCC HV6433.P4 P65 2019 | DDC 985.06/43—dc23
LC record available at https://lccn.loc.gov/2018012485

doi:10.7560/317310

Soifer_6844-final.indb iv 8/17/18 11:53 AM

 Preface

The idea for this collection emerged from a conversation we had after
attending an especially stimulating iteration of the Tuesday lunch sem-
inar in Latin American politics at Harvard’s David Rockefeller Cen-
ter for Latin American Studies during the academic year 2012–2013,
when both of us were based in Cambridge for the year. In the course
of a series of conversations, we began to realize that scholars of Peru
had not grappled in a systematic way with whether, how, or to what
extent Peruvian politics should be viewed as post-conflict. We were
also struck by the fact that the scholarship in political science on post-
conflict settings emphasized processes like international intervention,
peace building, demobilization, and the creation of new institutions,
none of which characterized the period after Peru’s internal confl ict
came to a close.
Finding ourselves struck by this gap in our understanding of con-
temporary Peruvian politics, we saw answering these questions as a
crucial missing element in the national conversation about the violence
of what we will call in the Introduction the “long eighties.” We there-
fore decided to draw on the community of scholars working on Peru to
begin a conversation around these questions. In so doing, we thought
it very important to bring together scholars, mostly but not exclusively
in the discipline of political science, from the English-speaking aca-
demic community with those based in Peru, in the hope that our con-
tribution to the academic literature on Peruvian politics would not
be divorced from the national conversation about these issues in the
country we all study. We are gratified by the willingness of so many
excellent scholars to join our project; working with our authors has
been not only a real pleasure but also an enlightening experience. We
are not exaggerating in saying that we have learned a great deal in the

Soifer_6844-final.indb vii 8/17/18 11:53 AM

 viii Preface

course of reading drafts of these chapters and engaging in conversa-

tion with their authors.
Our fi rst and most important thanks goes to the David Rockefel-
ler Center for creating a vibrant intellectual community around Latin
American politics. We are also grateful to the center and to Merilee
Grindle, its director at the time, for supporting a conference that gath-
ered most of the chapter authors for an initial conversation about these
issues. Paola Ibarra played an especially important role in providing
logistical support for the conference, and Steve Levitsky, generous in
so many ways, also supported our application for funding.
That initial conference provided us with the opportunity to begin
developing our thinking about how to contextualize legacies of the vi-
olence Peru experienced in the “long eighties.” We did this by bring-
ing together esteemed Peruvianists and experts on other Latin Amer-
ican countries marked by legacies of conflict to discuss early drafts of
many of the chapters, and by bringing the chapters in dialogue with
one another. In addition to the authors of the chapters in this volume,
we thank Jorge Domínguez, Angelica Durán-Martínez, Richard Sny-
der, and Jocelyn Viterna for joining us and contributing so much to
our discussion. We were also very lucky to benefit from the presence
and comments of David Scott Palmer. We were saddened to hear of his
passing as the production process concluded. We also thank Amanda
Milena Alvarez for her careful note-taking, which freed us to actively
participate in the lively conversations over two intellectually stimulat-
ing days.
We are grateful to Kerry Webb for her interest in the project, and to
her and Angelica Lopez-Torres for guiding it through the review pro-
cess so smoothly. Two anonymous reviewers provided insight on each
chapter and how to better tie them together to arrive at a holistic an-
swer to the question with which we began: to what extent and in what
ways are Peruvian politics post-confl ict? We are grateful to them for
their comments.
Finally, we would like to acknowledge the support of our family
members, who kept us motivated when we needed it and distracted us
when we needed a break. Hillel would like to thank Annie Stilz, who has
lived with this project for so many years, compensated only by prom-
ises of more meals in Lima, and Rachel, who arrived partway through
the long process of revisions. Alberto thanks Maria Ines Vasquez, who
provided, as usual, invaluable doses of love and intelligence all along
the way, and Ricardo Vergara and Lula Paniagua, his parents, with
whom he started to discuss these issues back in the eighties.

Soifer_6844-final.indb viii 8/17/18 11:53 AM

 Contents

Preface vii

Introduction: Leaving the Path Behind 1

Hill el Dav id Soife r a n d A lbe rto V e rga r a
1. Shining Path: The Last Peasant War in the Andes 17
José Luis R é n iqu e a n d A dr i á n L e r n e r
2. Civil Wars and Their Consequences: The Peruvian Armed
Conflict in Comparative Perspective 51
Li v i a Isa bell a Sch ubige r a n d Dav id Sul mon t
3. From Oligarchic Domination to Neoliberal Governance:
The Shining Path and the Transformation of Peru’s
Constitutional Order 79
M a x w ell A. C a m e ron
4. The Internal Armed Conflict and State Capacity: Institutional
Reforms and the Effective Exercise of Authority 109
Hill el Dav id Soife r a n d Ev e r et t A. V ieir a III
5. Impact and Legacies of Political Violence in Peru’s
Public Universities 132
Edua r do Da rge n t a n d Noeli a Ch áv ez
6. Peace for Whom? Legacies of Gender-Based Violence
in Peru 157
Jelk e Boest e n
7. Indigenous Activism and Human Rights NGOs in Peru:
The Unexpected Consequences of Armed Conflict 176
M a r i t z a Pa r edes

Soifer_6844-final.indb v 8/17/18 11:53 AM

 vi Contents

8. Political Violence and the Defeat of the Left 202

Paul a Mu ñoz
9. From a Partisan Right to the Conservative Archipelago:
Political Violence and the Transformation of the Right-Wing
Spectrum in Contemporary Peru 226
A lbe rto V e rga r a a n d Da n iel E nci nas
10. Public Opinion, the Specter of Violence, and Democracy
in Contemporary Peru 250
A rt u ro M a ldona do, Je n n ife r L. M e roll a , a n d
Eliz a bet h J. Zech m eist e r
11. Contested Memories of the Peruvian Internal
Armed Conflict 285
Paulo Dr i not
Conclusion 312
St ev e n L ev i tsk y

Works Cited 328

Contributors 371

Index 374

Soifer_6844-final.indb vi 8/17/18 11:53 AM

 INTRODUCTION

Leaving the Path Behind

Hillel David Soifer and Alberto Vergara

This book explores how the Internal Armed Conflict (IAC) of the
long decade of the 1980s has affected Peruvian politics thereafter.1 Al-
though myriad consequences of the conflict are frequently mentioned
in debates on contemporary Peruvian politics, this book constitutes an
initial attempt to provide a unified and systematic assessment of the
extent and nature of its effects.
This is, of course, not the fi rst study of the conflict: as time has
passed, bringing both more information and the critical distance that
analysis requires, our understanding of the Shining Path’s violence and
the Peruvian state’s reaction has grown steadily. Early in the 1980s the
most important research on the confl ict was spurred by surprise about
its outbreak and centered on the rural (and purportedly indigenous)
character of the Shining Path insurrection (see McClintock 1984;
Palmer 1986). The late 1980s saw scholarly debate about whether the
movement was in essence modern or millenarian (Degregori 1991b).
Early in the 1990s the duration of the conflict and the country’s gen-
eral crisis generated a body of work in which scholars contemplated
the abyss (Palmer 1992; Poole and Rénique 1992); in the second half
of the decade researchers strove to respond both to the unexpected
and sudden end of the confl ict and to its link to the emergence of the
authoritarian regime of Alberto Fujimori (Stern 1998c); and in the
2000s the Truth and Reconciliation Commission (Comisión de la Ver-
dad y Reconciliación del Perú, or CVR) elaborated a comprehensive
report (CVR 2003a) while Peru was reestablishing democracy. Since
then, a new body of work has reinterpreted the emergence and devel-
opment of the IAC and has started to ask about its consequences (Wil-
son 2013; Burt 2007; La Serna 2012; Heilman 2010; Theidon 2012;
del Pino 2017).

Soifer_6844-final.indb 1 8/17/18 11:53 AM

 2 Introduction

This volume adds to the previous body of scholarship by systemat-

ically exploring how the IAC has affected contemporary politics and
institutions in Peru. As this introduction will make clear, our analy-
sis diverges from diagnoses claiming that the IAC is the central factor
shaping contemporary Peru, as well as from those that dismiss its im-
pact. Instead, we argue that the IAC left important legacies, but that
as time passes these coexist with, affect, and are affected by a vari-
ety of other variables and processes. Thus legacies from violence—and
from other historical processes—inevitably mesh with new historical
processes to shape contemporary politics. The past of brutal violence
has not left Peruvians, but the country is not defined only by this past.
Our main task in this book is therefore to disentangle the intricacies of
legacies of violent conflict from other historical legacies, and from the
autonomous political and institutional development of contemporary
Peru. We believe we substantiate the claim that the IAC left important
traces in the country, but that it also has had divergent and unequal
consequences, ranging from a mild impact on some dimensions of con-
temporary politics to a fundamental influence on others. By disaggre-
gating the concept of legacies and exploring in detail some distinct di-
mensions of recent politics, we show the extent, and also the limits, of
the conflict’s impact on contemporary Peru.
The following pages of this introduction lay out a theoretical frame-
work that will allow our contributors to investigate the consequences
of the IAC in contemporary Peruvian politics and institutions. We be-
gin by describing basic elements of the conflict that Peru suffered in
what we call the “long eighties.” The second section of the introduc-
tion provides an overview of the legacies that the comparative politics
literature has attributed to the Peruvian conflict and highlights some
points of departure for our reexamination of these issues. Third, we
lay out an original theoretical framework to assess the consequences
of the conflict for contemporary Peruvian politics. We close by briefly
introducing the book’s chapters.

The “Long Eighties”

In 1978, after ten years of military dictatorship, Peruvians elected a

Constituent Assembly that would guide the country into a new dem-
ocratic era. Víctor Raúl Haya de la Torre, the long-standing leader of
the Alianza Popular Revolucionaria Americana (APRA), received the

Soifer_6844-final.indb 2 8/17/18 11:53 AM

 Introduction 3

largest number of votes, and after half a century of persecution he

was elected president of the Constituent Assembly. Communist par-
ties, allowed to participate, received a third of the vote and constituted
a solid new political bloc. This assembly produced the 1979 constitu-
tion, which established universal franchise for the fi rst time in Peru’s
history. Governed by this document, Peru held its fi rst genuinely dem-
ocratic general elections in 1980: for the fi rst time in the nation’s his-
tory, there were no proscribed political parties, and no adult citizen
was legally denied the right to vote (see Sanborn 1991).
Yet even as a tide of democratization swept across Latin America
in the next decade, the period came to be labeled as the “lost decade”
because of the severe economic crisis that swept across the region. In
Peru, however, the 1980s were a twofold loss, as economic disaster un-
folded hand-in-hand with harsh political violence.
The Partido Comunista del Perú-Sendero Luminoso (which we will
call the Shining Path, Sendero Luminoso, or SL throughout the vol-
ume) launched what it termed its “popular war” against the Peruvian
state in 1980. As chapter 1, by José Luis Rénique and Adrián Lerner,
traces, this declaration of hostilities initially went unnoticed in the Pe-
ruvian public sphere. When news of its emergence spread early in the
decade, political leaders and government agencies saw the insurrec-
tion as a bad joke. Thirteen years later, when the Shining Path was
defeated and the confl ict came to an end, the country had suffered,
according to the government-sponsored Truth and Reconciliation Re-
port of 2003, more than sixty thousand conflict-related deaths (CVR
2003a). The report described the conflict as the longest and broadest
in national history, while showing that it inflicted higher costs both in
human and economic terms than did the war against Chile in the nine-
teenth century. In 1993, the same year that the Shining Path leader
Abimael Guzmán asked for a halt to hostilities, Peru approved a new
constitution pushed by the fujimorista majority that ended the short-
lived democratic experiment embodied in the 1979 constitution. This
period of democratic experiment, economic crisis, and insurgency
from 1979 to 1993 is what we call the “long eighties.”
The Shining Path was a Maoist faction initially situated within the
highly fragmented Peruvian Left of the 1970s. Its aim, inspired by the
ideology of Mao Zedong and Peruvian thinker José Carlos Mariáte-
gui, was to trigger a bloody revolution that would tear apart the Pe-
ruvian state. The Shining Path was a peculiar insurgency from a Latin
American perspective both in terms of the messianic cult of its leader,

Soifer_6844-final.indb 3 8/17/18 11:53 AM

 4 Introduction

Abimael Guzmán, also known as “Presidente Gonzalo,” and in terms

of its indiscriminate use of violence (Gorriti 1990; Degregori 1991b;
La Serna 2012; Heilman 2010; Hinojosa 1998; Portocarrero 2012).
Between 1980 and 1982 violence began to escalate, especially in
Ayacucho, where five provinces were placed under a state of emer-
gency in October 1981. 2 In 1982 Shining Path actions increased in
the central Andes, which resulted in the replacement of the ineffec-
tive police by the Peruvian armed forces. As a result of both of these
shifts, the violence escalated at the end of 1982 and the insurrection
achieved national scope. In 1983, the killing of eight journalists from
important Lima media outlets in the highlands of Ayacucho brought
national attention to the confl ict for perhaps the fi rst time, revealing
the barbarous violence that had silently enveloped the central Andes.
Between 1983 and 1986 several massacres took place in the southern
and central highlands departments of Ayacucho, Huancavelica, Junín,
and Apurímac.3 To make things worse, the Movimiento Revoluciona-
rio Túpac Amaru (MRTA), a Cuban-inspired urban group, launched
its own armed uprising against the Peruvian state in 1984.4
The second half of the 1980s saw several important shifts in the
conflict. The Alan García administration elected in 1985 tried to im-
plement a new development-centered strategy against insurgents, but
this approach quickly failed. In 1986 the state massacred hundreds
of Shining Path members in several jails, and the MRTA opened a
whole new front of violence in the Peruvian jungle. In 1988 a paramil-
itary group called Rodrigo Franco burst into the confl ict, perpetrat-
ing a few selected killings. By the end of the 1980s, the economy hit
bottom, which favored a brutal Shining Path offensive especially tar-
geted at the capital city of Lima. Simultaneously, as chapter 4, by Hil-
lel David Soifer and Everett A. Vieira III, details, the army’s counter-
insurgency strategy switched from broadly indiscriminate violence to
more targeted attacks. After the election of Alberto Fujimori in 1990,
Grupo Colina, a new paramilitary group, carried out several massa-
cres and other acts of violence, and by 1991 around half of the Peru-
vian population was living under a state of emergency. The confl ict
took a sharp turn in 1992, when Guzmán was captured and the Shin-
ing Path’s fortunes declined sharply. In 1993 all the main Shining Path
leaders asked for a peace accord, though a marginal contingent formed
a faction called Prosequir that continued the armed conflict in jungle
areas. Heavily associated with drug-dealing, Prosequir’s impact was
marginal. With these events, the decade of the long eighties came to a
close in 1993 and the legacies of the IAC started to be identifiable.

Soifer_6844-final.indb 4 8/17/18 11:53 AM

 Introduction 5

It was a long confl ict with traumatic effects on the country. The
CVR report, using statistical techniques designed to account for the
underreporting of civilian victimization, estimates that approximately
sixty-nine thousand people died. Unlike in every other case of inter-
nal conflict in Latin America or most such confl icts around the world,
in Peru the deadliest actor was not state forces, but the Shining Path,
to which the CVR report attributed responsibility for 54 percent of
all deaths. Adding to the trauma, the violence was widespread across
Peru. Only two of the country’s twenty-five departments, Moquegua
and Madre de Dios, had no reports in the CVR investigation of casu-
alties related to the political violence. The CVR report found that the
central Andean region was particularly affected by the violence, with
Ayacucho alone containing 40 percent of the victims. Almost 40 per-
cent of the casualties were suffered by the poorest quintile of the coun-
try’s population, mainly in rural areas, and 75 percent of all deaths
were among speakers of an indigenous language, compared to some
25 percent of the total population. Hence, the demographic and ter-
ritorial disparities in violence revealed deep and historical divisions
within Peru.
And yet all this destruction was only part of the Peruvian “lost de-
cade.” The devastation of the IAC was accompanied by economic mis-
management that contributed to an unprecedented general crisis. Be-
tween 1980 and 1989, Peru’s gross domestic product (GDP) contracted
by 4.8 percent as the productive capacity of the country plummeted
(Llosa and Paniza 2015). The fiscal deficit reached 12 percent of GDP
in 1989, and in the same year inflation rose by more than 3,000 per-
cent. In 1990 inflation reached 7,500 percent, and the Peruvian GDP
per capita fell to 1960 levels (Banco Central de Reserva del Perú
2005). Between 1985 and 1990 the rate of formal employment shrank
from 53 percent to a mere 5 percent of the employed population (Pa-
rodi 2008). In addition to economic collapse, a cholera epidemic rav-
aged the country in 1991, affecting at least 700,000 citizens (MINSA
2011). In a nutshell, the IAC and other social and economic factors not
only brought on a political and economic crisis but left the country on
the verge of a crisis of stateness (Corrales 2003).
A new phase in Peruvian history began in the mid-nineties with the
1993 constitution, which still rules the country today, as discussed
in chapter 3, by Maxwell A. Cameron. This new institutional frame-
work for Peruvian politics and society was followed in the new mil-
lennium by an exceptional cycle of economic growth that transformed
much of the Peruvian society. This new phase has implied, in both ob-

Soifer_6844-final.indb 5 8/17/18 11:53 AM

 6 Introduction

jective and subjective terms, the onset of a post-conflict period. If we

accept the objective quantitative threshold of one thousand casualties
per year often used to defi ne a civil war (see, for example, Kalyvas
and Balcells 2010), then Peru’s conflict ended in 1994. In more sub-
jective terms, although a few remnants of the Shining Path have sur-
vived in the Peruvian jungle, for most Peruvians “the era of terrorism”
is doubtless a matter of the past, though as Arturo Maldonado, Jenni-
fer Merolla, and Elizabeth Zechmeister show in chapter 10, the fear of
terrorism nonetheless remains highly salient.
Based on this discussion and our reading of the broader literature,
we use the term “post-confl ict” to refer to the aftermath of conflict
within Peru that involved significant violence by multiple armed actors
against both one another and the civilian population. Although we
do not seek to pin down an exact moment when Peru became “post-
conflict,” we defi ne this period as encompassing the years from the
mid-1990s to the present. This volume thus seeks to explore legacies of
the conflict of the long 1980s that have affected and continue to affect
Peruvian politics.

Legacies in Comparative Perspective

The literature on contemporary Peruvian politics is divided between

scholarship attributing a decisive effect to the confl ict and giving it a
central causal role in explaining outcomes of interest, and work that
downplays its effects to focus instead on other features of the coun-
try’s history or more contemporary causal factors. This volume seeks
to assess this surprising disjunction by evaluating which features of
post-confl ict Peruvian politics have been decisively shaped by the IAC,
which have been affected in less fundamental but still important ways,
and which seem to be a product of other causal forces.
A significant body of scholarship on Latin American and Andean
politics has argued that the Peruvian case diverges from its counter-
parts in several important ways. Very often the political violence of
the eighties is said to be the crucial variable that explains these anom-
alies. Arguments along these lines have been especially common in re-
search on the development of civil society. In these accounts, social
organization in Peru is said to be distinctively fragmented and weak
as a result, at least partially, of the IAC. During the fi rst years of the
new millennia, when indigenous movements successfully emerged in

Soifer_6844-final.indb 6 8/17/18 11:53 AM

 Introduction 7

Andean countries, Peru was an “anomaly” with no national indige-

nous movements (Yashar 2005, 240). Mallon (1998) largely attributed
the absence of ethnic identification among Peru’s popular sectors to
class-based conceptions of politics that dominated the Peruvian Left
and Shining Path. Even more directly, Yashar (2005) compares Peru
to Bolivia and Ecuador, and argues that in contrast to these neighbor-
ing countries, violence disrupted indigenous social organizations and
networks in Peru, making it hard to politicize and mobilize ethnic di-
visions. And contrary to other cases in the Andean countries, ethnic
parties also failed to emerge in Peru, which Van Cott (2005) claims
was a consequence of the IAC. In essence, scholars have argued that
the whole chain of ethnic politics, from identity to party representa-
tion passing through social organization, was severely impacted by the
conflict.
Scholars have also attributed the evolution of Peru’s political par-
ties in the last decades to the IAC. The weakness of the Peruvian Left
over recent decades is especially striking in comparative perspective
(Levitsky and Roberts 2011). The IAC is said to have played a prom-
inent role in this weakness: “If the Shining Path exacerbated social
fragmentation, it also contributed to the polarization of the IU” (Rob-
erts 1998, 260). One mechanism driving this outcome was the sense of
fear that arose in Left-aligned civil society leaders during the confl ict,
which had the effect of weakening social movements (Burt 2006). In
the same vein, legacies of violence constituted a hurdle shared by the
Peruvian and Colombian Left, two countries that did not join the so-
called Left Turn (Cameron 2011, 394). On the opposite side of the con-
temporary political spectrum, the enduring presence of a fujimorista
party has been attributed to the enormous popularity Alberto Fuji-
mori gathered after his government defeated the Shining Path (Levit-
sky and Zavaleta 2016). Finally, the weakness of Peru’s regional par-
ties and regional elites has also been traced back to the IAC (Vergara
2015a).
In contrast, other studies of post-confl ict Peru have not granted the
same causal importance to the IAC.5 Kenney (2004) provides an ex-
planation of the breakdown of Peruvian democracy in 1992 that cen-
ters on the lack of cooperation between the executive and legislative
branches. After the 1992 self-coup (autogolpe) that Fujimori carried
out, several researchers focused on the politics and policies of the new
political era. Many accounts of the neoliberal turn in Peru attribute
no significant effect to the IAC in exploring either the introduction

Soifer_6844-final.indb 7 8/17/18 11:53 AM

 8 Introduction

of neoliberal policies or their effects (Wise 1994; Arce 2005). Nor do

scholars explaining the implementation of new poverty alleviation
programs (Schady 2000) or of decentralization (McNulty 2011) attri-
bute a causal role to the violence of the 1980s. More recent research
on democracy and technocracy in Peru has also given only marginal
room to the IAC (Dargent 2015; Vergara and Encinas 2016). Similarly,
work by scholars like Arce (2014) and Meléndez (2012) attributes the
recent expansion of local conflicts and protest in Peru to contempo-
rary political and economic factors and sets aside the causal role of the
IAC, as do recent accounts of Peruvian “democracy without parties”
(Levitsky 2013; Zavaleta 2014).
Hence, as we see, the weight given to the IAC in scholarly explora-
tions of contemporary Peruvian politics varies fundamentally, ranging
from key explanatory piece to ignored process. By digging into the Pe-
ruvian case in detail, this book attempts to provide a systematic analy-
sis of these issues. We seek not to challenge or confi rm any particu-
lar fi ndings cited in the preceding paragraphs, but rather to produce a
synthetic account of the consequences of the IAC for contemporary Pe-
ruvian politics.
To do so, we unpack several dimensions of contemporary Peruvian
politics, society, and institutions to evaluate the extent to which they
have and have not been shaped by the IAC. This systematic evalua-
tion requires a framework that takes history seriously. As we explain
in the next section, we need a careful temporal framework to distin-
guish causal chains that originated in the long eighties from factors
that emerged thereafter, and also to distinguish the confl ict from the
longer historical processes of a fragmented society, polarized politics,
and weak institutions that long predate it, and that also shape contem-
porary Peru in fundamental ways.

Studying Post-Conflict Peru

In laying the groundwork for the chapters that follow, several thorny
methodological challenges in the study of post-confl ict politics must
be confronted. First, we cannot simply draw from the existing schol-
arship on post-confl ict settings, because it sheds only limited light on
our case. Scholars of post-confl ict politics have tended to explore is-
sues like peace negotiations and the implementation of the resulting
accords (Walter 2002; Fortna 2003; Barnett 2006), the effects of in-

Soifer_6844-final.indb 8 8/17/18 11:53 AM

 Introduction 9

ternational intervention in bringing an end to confl ict and securing

peace (Paris 2004; Fortna 2008; Stanley 2013), the political integra-
tion of former combatants (Blattman 2009; Humphreys and Weinstein
2008), and the outcomes and impact of post-confl ict reconstruction
projects (Baranyi 2008; Call and Wyeth 2008). Yet none of these pro-
cesses unfolded in Peru.
In the Latin American context, much has been written about how
violence, defi ned broadly enough to include inequality and social ex-
clusion, has marked societies. To take but one example, the collection
edited by Koonings and Kruijt (1999) investigates the “long-term con-
sequences of violence, repression, and arbitrariness” (p. 2). Tracing the
roots of this violence to social inequality, especially in the rural realm,
and to struggles over political incorporation and social inclusion, they
see it as one of the crucial “recurrent features of the Latin American
political landscape” (p. 2). In studying the “legacies of repressive dicta-
torships and civil wars” (p. 3) they seek to explore the origins and con-
sequences of violence in contemporary Latin America.
Yet in focusing on the legacies of the Sendero conflict, our collec-
tion takes a distinct approach. To understand why we focus on a spe-
cific confl ict rather than the broader phenomenon of violence, one need
only note a striking irony of the Peruvian confl ict: Sendero Luminoso’s
armed insurgency began on the very day that democracy returned to
Peru in 1980, and the end of the conflict coincided not with democra-
tization but with the consolidation of the authoritarian regime of Al-
berto Fujimori. It is reasonable to situate a specific internal confl ict as
nothing more than part of a broader phenomenon of structural vio-
lence in a case like Guatemala, where armed opposition emerged due
to the level of repression and political closure, and thus the civil war
itself can be seen as a consequence of the nature of authoritarian rule
(Wickham-Crowley 1992; Goodwin 2001). But for the Peruvian case,
as chapter 1, by Rénique and Lerner, and chapter 2, by Livia Isabella
Schubiger and David Sulmont, make clear, the internal conflict cannot
be reduced to being only a reflection or a component of broader pro-
cesses of repression and inequality. We do not deny that long structural
patterns of exclusion might affect the conflict—and chapter 6, by Jelke
Boesten, shows these in an especially vivid manner in exploring issues
of gender violence in contemporary Peru. However, beyond these long-
term continuities, the armed confl ict generated its own set of indepen-
dent causal impacts on politics. This book is based on the presumption
that these effects are worthy of exploration in their own right.

Soifer_6844-final.indb 9 8/17/18 11:53 AM

 10 Introduction

Therefore, we deal with the political legacies of confl ict rather than
post-confl ict peace and reconstruction processes (most of which did
not unfold in the Peruvian case) or legacies of violence more generally.
To do so, we develop a new approach to studying the impact of the
conflict in Peru. We hope that our new approach might not only shed
light on the political legacies of the IAC but also aid scholars of other
post-confl ict settings as well, should they seek to shift their attention
from the issues now central to that literature to the political impact
of the conflict itself. Since our focus is the set of legacies of the IAC
that shape post-confl ict politics, we need to begin by delineating the
boundaries of what we mean by “political legacies.”
In taking on this challenge, we seek to complement a rich body of
recent work that has explored the aftermath of the conflict in ways
that are distinct to the approach we seek to develop. Kimberly Theidon
and others taking an anthropological approach have superbly analyzed
relationships within and among local communities after the conflict
(Theidon 2012; del Pino and Yezer 2013). Olga Gonzáles and Cynthia
Milton, among others, have explored the traces the confl ict has left on
artistic and cultural expression (Gonzalez 2011; Milton 2014; Faverón
2006). Other academics have peered into the memories of main actors
(Milton 2018; Asencios 2017).6 Our book does not investigate the cul-
tural dimensions of the confl ict’s aftermath or observe it from a micro
level. Instead, we focus on the macropolitical and institutional conse-
quences of the confl ict.
In focusing on these consequences, we seek to respond to what Blatt-
man and Miguel (2010, 43) describe as “perhaps the most pressing area
for future empirical research” on post-conflict settings: the “social and
institutional legacies” of confl ict. Understanding those legacies is ex-
actly the goal of this book. However, we do not believe it is possible
to generate analytically precise causal claims if we do not disaggregate
the results beyond broad concepts like institutions or “democracy” (see
Kier and Krebs 2010). In seeking causal precision, we explore how the
legacies of the conflict shaped political institutions, civil society, and
political attitudes and participation. That is, we unpack the macro-
political consequences, allowing us to provide a nuanced analysis of
post-confl ict Peru disaggregated across distinct political, social, and
institutional realms. But before turning to a description of these realms
and introducing the chapters to follow, we need to defi ne the concept
of post-confl ict legacies and identify a set of mechanisms by which the
legacies of conflict can be produced.

Soifer_6844-final.indb 10 8/17/18 11:53 AM

 Introduction 11

Disaggregating the Legacies of Conflict

To explore the legacies of Peru’s conflict, we distinguish three causal

channels through which its effects are produced and reproduced,
and by which it has impacted contemporary Peruvian politics: war-
time mechanisms, post-confl ict legacies, and the political struggle over
legacies.

Wartime Mechanisms

By wartime mechanisms we refer to those factors that were produced

by the conflict, emerged during the confl ict period, and have persisted
into the post-confl ict era. These are what Wittenberg (n.d.) calls “con-
tinuation legacies” in his typology. The most prominent set of war-
time mechanisms discussed by our authors relate to the violence it-
self: the chapters by Eduardo Dargent and Noelia Chávez (chapter 5),
Maritza Paredes (chapter 7), Paula Muñoz (chapter 8), and Alberto
Vergara and Daniel Encinas (chapter 9), among others, show that vio-
lence impacted the social and political organization of many sectors of
Peruvian society in ways that endured after the conflict ended. While
in most cases the effects tended to weaken societal and political actors,
Paredes shows that indigenous organization in the Amazon region was
strengthened in response to violence, and Vergara and Encinas show
that violence led to the growth of a more powerful and more conserva-
tive Catholic Church as a central actor.
In chapter 3, Cameron explores how the broader sense of crisis en-
gendered by the confl ict created the opportunity for Fujimori not only
to carry out his autogolpe but also to craft a constitution that solidified
authoritarian and neoliberal control. Like Cameron, Soifer and Vieira
show in chapter 4 that institutional changes made during the confl ict—
in this case, reforms of the security apparatus—persisted long after the
conflict ended.

Post-Conflict Legacies

In contrast to wartime mechanisms, which refer to aspects of the po-

litical arena that were altered during the conflict, post-confl ict legacies
refer to those factors that emerge in the post-confl ict era as the result
of effects of the conflict era. We thus defi ne post-confl ict legacies as
factors that emerge in the post-confl ict period as the result of the con-

Soifer_6844-final.indb 11 8/17/18 11:53 AM

 12 Introduction

flict. Here, too, violence itself is an especially important mechanism.

Maldonado, Merolla, and Zechmeister, for example, show in chap-
ter 10 that worry about terrorism remains strikingly high in Peru by
comparison to other Latin American countries, and that this fear is es-
pecially salient among the cohort of Peruvians who lived through the
conflict. This fear not only persists over time but also has sizable and
systematic effects on democratic attitudes. In a similar vein, Dargent
and Chávez show in chapter 5 that the violence in public universities in
the 1980s contributed to the weakening of student mobilization in the
post-confl ict period, and Muñoz shows in chapter 8 that the Left’s in-
ability to distance itself from Sendero’s actions during the conflict led
it to “become associated in political discourse with terrorism.”
Another set of post-confl ict legacies centers on civil society. Here,
the effects are distinct in different sectors of society: while Vergara
and Encinas show in chapter 9 that the weakening of those sectors of
civil society that supported parties of the Right has empowered tech-
nocrats and therefore reinforced neoliberal continuity, Paredes shows
in chapter 7 that the human rights groups and lawyer-activists that
emerged during the confl ict and were strengthened by international
support after the confl ict ended (and especially in the later years of the
Fujimori regime) were important allies for indigenous actions in the
post-confl ict period.
A third set of legacies is highlighted in chapter 3 by Cameron, who
seeks to explain neoliberal continuity. Cameron highlights the con-
tinuing resonance of neoliberal ideology that emerged during the con-
flict, and the ways it has been sustained by the kind of state built in the
post-confl ict era.

Struggle over Legacies

Whereas both wartime mechanisms and post-conflict legacies are

products of features of the conflict period, a third category of ways in
which post-confl ict politics is shaped by the conflict results from the
choices and struggles of political actors, who frame, shape, and ma-
nipulate features of the confl ict and its legacies for their own purposes
(Krebs 2009, 255). We separate the discussion of legacies that take
this form in order to highlight the agency of political actors in shap-
ing them, and to label these constructed legacies. The most commonly
echoed type of constructed legacy our authors highlight is the stigma-
tization of the Left (Muñoz, chapter 8), human rights issues (Paulo

Soifer_6844-final.indb 12 8/17/18 11:53 AM

 Introduction 13

Drinot, chapter 11), and collective mobilization more generally (Dar-

gent and Chávez, chapter 5) as associated with terrorism. On the other
side, activists have also at times sought to exploit acts committed by
the state, especially ones that occurred under the Fujimori adminis-
tration. Another use of constructed legacies appears in intra-state and
intra-elite power struggles; Soifer and Vieira show in chapter 4 that
memories of the conflict were used by actors seeking to strengthen the
intelligence and surveillance apparatus in the Fujimori era, and Cam-
eron shows more broadly in chapter 3 that pro-neoliberal actors have
sought to frame the conflict in a way that justifies the neoliberal model
as a central component of the defense against a recurrence of the un-
rest of the 1980s.
Incorporating this type of legacy into our analysis allows us to trace
both the direct and objective legacies of the confl ict, and the impact of
the more subjective or constructed struggles on the legacies of the con-
flict. The confl ict left objective legacies, but the political struggle over
the interpretation or construction of those legacies can be as impor-
tant as the objective ones.

Disaggregating Consequences

In all, this volume aims to provide a nuanced account of the conse-

quences of the IAC for politics in contemporary Peru. We do not pre-
sume or argue that the IAC is the most central factor shaping poli-
tics in contemporary Peru. Peru, to put it simply, is “post-too many
things”: contemporary Peru might be described as not just post-
conflict, but also post-hyperinflation, post-party system collapse, and
post-authoritarian. Instead of arguing that the post-conflict lens is
the most useful frame for understanding Peruvian politics, we seek to
evaluate the extent of its utility. The country is still shaped by the IAC
of the long eighties, but as time goes on, it becomes harder to disentan-
gle consequences of the IAC from all the other variables and processes
that have also affected Peru in the two decades since the IAC was fi n-
ished.7 These include a booming economy (Ghezzi and Gallardo 2013),
the emergence of a new middle class (World Bank 2013), the deteriora-
tion of political representation (Levitsky and Zavaleta 2016), and the
ever-increasing presence of informality (W. Mendoza 2017), among
others.
However, we do believe that the IAC unleashed some legacies that

Soifer_6844-final.indb 13 8/17/18 11:53 AM

 14 Introduction

significantly shaped some areas of contemporary politics and institu-

tions in Peru. In order to grasp them, we need to unpack the broad no-
tion of “contemporary politics” into manageable realms and explore
each in turn. The book divides contemporary politics into four realms
in which we observe the consequences of the confl ict: state institu-
tions, civil society, political parties, and public opinion, each of which
is divided into specific dimensions that are the subject of individual
chapters. This analytical approach allows us to make precise claims
about the causal link between the IAC and each of those dimensions.
Our framework distinguishing three types of legacies and the authors’
focus on case studies of the Peruvian experience rather than sustained
and systematic comparisons with other Latin American countries al-
lows us to construct a careful and detailed picture of the consequences
of the IAC on contemporary Peru. The concluding chapter of this vol-
ume by Steven Levitsky synthesizes the earlier chapters’ fi ndings to es-
tablish the extent to which the country is and is not decisively shaped
by the confl ict.
The book’s fi rst section provides historical and theoretical back-
ground. Chapter 1 gives a detailed historical overview of the IAC in
Peru, while chapter 2 places the Peruvian case in the context of the
contemporary political science scholarship on internal confl icts. The
next section deals with state institutions. Chapter 3 highlights how
neoliberal ideology, enshrined in the 1993 constitution, both took hold
in Peru in response to Sendero’s perceived threat to individual free-
dom and was sustained by the discrediting of other ideologies in the
course of the long decade of conflict. Chapter 4 explores how the con-
flict shaped the state’s coercive capacity, emphasizing how reforms of
the military and of policing and surveillance institutions, as well as the
reach of the state over territory, were the product of all three of the
types of mechanisms elaborated in this introduction. Chapter 5 inves-
tigates how the conflict deepened the crisis of Peru’s public university
system in several important ways.
The third section explores civil society and the impact of the IAC.
Chapter 6 makes the case that both ineffectual state policies against
gender violence and recent mass mobilization defending women’s
rights can be traced to the IAC. Chapter 7 focuses on indigenous mo-
bilization in the political context of the contemporary boom in extrac-
tive development in Peru, which provides a new setting for the study
of indigenous movements and the responses they draw from the state.
Both chapters show that the IAC had a decisive impact on contempo-
rary Peruvian civil society, and that its effects continue to be felt.

Soifer_6844-final.indb 14 8/17/18 11:53 AM

 Introduction 15

In the fourth section, on political parties, chapter 8 analyzes the

link between the IAC and the leftist sectors of the political spectrum,
while chapter 9 examines connections to the political right. In chap-
ter 8, Muñoz contends that the end of political violence came with not
only the military defeat of the Shining Path but also the political defeat
of the legal Left. She argues that the prospects of the Left for effec-
tively becoming an institutionalized and electorally appealing politi-
cal party were undermined by the manner in which the confl ict ended.
Vergara and Encinas describe in chapter 9 the transformation of the
Right from organic political parties in the 1980s to a new condition
they label as a “conservative archipelago.” Part of this transformation,
they contend, was related to the IAC.
The fifth section contains two chapters dealing with Peruvian pub-
lic opinion, though from very different perspectives. Chapter 10 pro-
vides a statistical and experimental analysis of the attitudes of Peru-
vian citizens toward terrorism and their implications for democratic
support. In contrast, chapter 11 takes a qualitative approach in explor-
ing the comments of YouTube users on videos related to memorials of
the conflict in order to document the collective struggle to construct
the memory of the IAC in the absence of a collective memory proj-
ect in the country. The volume’s conclusion places all of these in-depth
chapters in a wider Latin American perspective, producing a synthetic
evaluation of the ways in which contemporary Peruvian politics are
decisively shaped by the IAC.
In all, our aim in compiling this volume is to deepen the process of
analytical inquiry into the legacies of violence for contemporary Peru.
Yet we seek to go beyond that: for most of the contributors of the
book, the Shining Path’s brutal actions, and the consequences they un-
leashed, were vivid, tragic, and major events. As Schubiger and Sul-
mont remind us in their chapter, the department of Ayacucho, where
the conflict began and had the deadliest consequences, saw an absolute
loss of population between the censuses of 1981 and 1993. The “long
eighties” were, in other words, a national catastrophe. Regardless of
where we stand politically or how we interpret such a history of vio-
lence, it is clear that the scars and legacies of the confl ict continue to
trouble its heirs in ways that cannot be ignored. Recognizing and an-
alyzing these effects in a systematic manner is an indispensable step
on the path to developing a national conversation about the confl ict.
Such a national dialogue, as Drinot’s chapter shows, has not emerged
in a rational or inclusive way. Instead, a systematic understanding of
the conflict and its legacies seems to remain elusive and opaque for

Soifer_6844-final.indb 15 8/17/18 11:53 AM

 16 Introduction

most Peruvians. Expanding on Dargent and Chavez’s insightful de-

scription of contemporary university students, what seems to prevail
in the country is a “silent majority.” Our aim, then, beyond our aca-
demic purposes, is to help the country to recognize some of the ways
in which it is shaped today by the savage violence of the long eighties,
and also some of the ways in which contemporary politics is rooted in
other historical factors. A clear understanding of these complex pro-
cesses linking past and present will always be a necessary requirement
for fi nding one’s way to a better future.

Notes

1. “Internal Armed Confl ict” is the term used by the Peruvian Truth and
Reconciliation Commission to characterize the violence between the Shining
Path and other non-state armed groups and the Peruvian state. The notion of
“long eighties” is explained below.
2. The following paragraphs rely on the CVR Final Report to constitute a
brief historical reconstruction of the Peruvian Internal Armed Confl ict. For a
detailed account of how the confl ict unfolded, see chapter 1 by José Luis Ré-
nique and Adrián Lerner.
3. The army was responsible for, among others, the Sacos, Pucayacu, Ac-
comarca, and Putis massacres; the Shining Path perpetrated, among others,
the Lucanamarca and Huancasancos massacres.
4. Since the confl ict between the MRTA and the state was less bloody than
that pitting Sendero against the state, and since public perception of the vi-
olence Peru suffered focused more heavily on the latter, we describe the vio-
lence from 1979 to 1993 as the Sendero confl ict throughout the volume.
5. To be fair, it is difficult to fi nd research on contemporary Peru that fully
dismisses the IAC, since most of this work does attribute some role to the gen-
eral turmoil of the eighties. Yet in contrast to the work discussed in the previ-
ous paragraph, the scholarship we discuss here does not grant the IAC a cen-
tral role in its analyses.
6. In addition to the academic literature on cultural legacies of the con-
flict, a nonacademic but vivid and fruitful body of work both on the IAC and
its legacies has boomed in several genres in contemporary Peru. Scholars and
intellectuals personally involved in the confl ict have produced engaging and
eclectic books that have resonated far beyond the academic realm (see, among
others, A. Gálvez 2009; Gavilán 2012; Agüero 2015; Cisneros 2015). Fiction
writers have peered into the period as well (see, among others, Roncagliolo
2006; Colchado 1997; Cueto 2005; Thays 2008; M. Vargas Llosa 1983; see
also the notable graphic novel by Cossio, Rossell, and Villar 2008); and fi lm-
makers have also successfully explored it (among others, see Lombardi 1988;
E. Mendoza 2017; H. Gálvez 2015; del Solar 2015; C. Llosa 2009; Calero
2016).
7. On these methodological questions relating to temporality, see the
analysis of the Mexican Revolution and its consequences in Knight 1985.

Soifer_6844-final.indb 16 8/17/18 11:53 AM

 CHAP T ER 1

Shining Path:
The Last Peasant War in the Andes
José Luis Rénique and Adrián Lerner

The process of modernization begins with peasant revolutions

that fail.
Ba r r i ngton Moor e , Soci a l Or igi ns of Dic tatorsh ip
a n d De mocr ac y

In mid-1988 the hitherto mysterious leader of the Communist Party

of Peru, better known as the Shining Path (Sendero Luminoso), broke
his long silence. In his fi rst interview since the beginning of the insur-
rection, Abimael Guzmán Reynoso boasted of the successful progress
of a rebellion whose central protagonists, he claimed, were “the poor
peasants” of Peru. In Guzmán’s words, the poor played a primary role
in the “People’s Guerrilla Army,” serving “as fighters and command-
ers.” Led by this organization, the “people’s war” spread through the
highlands, “from one border to the other, from Ecuador to Bolivia,”
along a region that was “the historic axis of Peruvian society and its
most backward and poor area” (Guzmán Reynoso 1988). Eight years
earlier, the start of the Shining Path uprising had been received with
disbelief. Even as it continued to grow, leftist leaders had insisted that
it was “infantile.” President Fernando Belaunde, during his second pe-
riod in power (1980–1985), had labeled the insurgents “rustlers.” As
late as 1983, military officers had declared that eliminating the Shin-
ing Path would take only a few weeks (CVR 2003a, 2:60).
Was this, then, effectively a “peasant war”? If so, how was it possi-
ble in a country where one of the most radical agrarian reform efforts
in contemporary Latin America had been recently enforced, and where
more than 60 percent of the population had become urban? Why did
other political and social actors seem to miss it? These questions lie at

Soifer_6844-final.indb 17 8/17/18 11:53 AM

 18 José Luis Rénique and Adrián Lerner

the center of the era of violence experienced by Peru during the 1980s
and early 1990s: “the most intense, widespread, and prolonged period
of violence in all of its republican history” according to the Truth and
Reconciliation Commission (Comisión de la Verdad y Reconciliación
del Perú, or CVR) created by the Peruvian government in 2001 to in-
vestigate it (CVR 2003a, 8:245). This chapter presents the most sa-
lient aspects and the chronology of the Internal Armed Conflict (as
the CVR defi ned it) that ravaged Peru during these years. It pays par-
ticular attention to the roles attributed to the Andean peasantry as a
historical subject and agent in the ideologies and practices of the two
main armed sides of the confl ict, the Shining Path and the coercive
forces of the Peruvian state, amid the shifting dynamics of the con-
flict and the broader societal transformations that characterized the
period. It was a war predominantly fought and endured by peasants.
Our chapter shows that the peasant character of the war was a central
aspect of the revolutionary ideology and praxis of Abimael Guzmán
and the Shining Path, and also of the different counterinsurgent strat-
egies of the Peruvian state, but that the on-the-ground experiences and
effects of “peasant war,” both at the local level and in broader strate-
gic terms, were very different from the preconceptions and theoretical
perspectives of these actors. We begin with an introduction to the rad-
ical tradition of which the Shining Path was a part, but from which it
also departed, and to the counterinsurgent traditions of the Peruvian
armed forces. The second section of the chapter focuses on Ayacucho
and its place in Peruvian society and politics during the fi rst half of
the 1980s, to describe and explain how a fringe Maoist group became
a serious revolutionary challenge in the highlands, and then launched
a war on the national scale. We then shift to an analysis of how both
the Shining Path and the new Peruvian government, led by the Alianza
Popular Revolucionaria Americana (APRA) and president Alan Gar-
cía (1985–1990), switched strategies as violence expanded its reach to
new territories, and new actors emerged along with a sense of deepen-
ing national crisis. Finally, we describe the end of the war during the
fi rst years of the government of Alberto Fujimori (1990–2000) and
touch on some of the political uses and consequences of the confl ict’s
aftermath during the 1990s.
The perception of Sendero Luminoso as an alien infection that sud-
denly attacked Peruvian society with its dogmatic violence still looms
large in the country’s collective memory. And yet, already in 1998 the
historian Steve J. Stern had synthesized a generation’s worth of schol-

Soifer_6844-final.indb 18 8/17/18 11:53 AM

 The Last Peasant War in the Andes 19

arship to argue that Sendero’s origins and development were located

within Peruvian history while also working against it. While the Shin-
ing Path was in many ways a logical episode in a century-old radi-
cal tradition, and part of a broader social and political process, in its
extreme ideological rigidity it also defi ned itself explicitly as against
other projects of change, as seen in its efforts to crush the velasquista
reforms and other governmental efforts as well as all non-senderista
leftist parties and initiatives emanating from civil society. A Marxist
revolutionary party that advocated a peasant revolution provoked a
war whose main victims were peasants and other Marxist activists,
and whose immediate objective was the destruction of the Peruvian
state (Stern 1998a). This paradoxical trajectory mirrored the singu-
larities of postcolonial Peru, a society that the historian Paul Gooten-
berg has recently described as still carrying scars that seem to be a
“perpetual obstacle” to “real horizontal bonds of nationhood, a le-
gitimate state, real dynamic capitalism, even real modernity” (Goo-
tenberg 2014). These were the scars of empires, wars, conquests, and
revolutions past; of colonialism, liberalism, centralism, and peripheral
capitalism; of the divisions and durable inequality that still defi ned the
country (Tilly 1999).
The other key armed actors of the conflict, the coercive forces of the
Peruvian state, were also trapped by these paradoxes. The nation’s po-
lice and armed forces were marked by tense relationships with civilians
in charge of political power, inter- and intra-institutional conflict and
organizational dysfunction, and corruption. Pervasive rifts based on
ethnicity, class, region, and gender status were exacerbated by the ir-
regular warfare that characterized the conflict. Similarly, their partici-
pation was no less driven by ideological stances than those of Guzmán
and his senderista comrades. Their strategies, their conceptions about
what kind of conflict they were fighting, and their ideas about the roles
they and other social actors played in the confl ict were all based on in-
herited diagnoses of Peruvian society, and on the application of dated
Cold War counterinsurgent doctrines to the Andean countryside.

Paths to War: The Radical Tradition, Maoism,

and Counterinsurgency in Peru

On May 17, 1980, the eve of the fi rst presidential elections in seven-
teen years and after more than a decade of military rule, a handful of

Soifer_6844-final.indb 19 8/17/18 11:53 AM

 20 José Luis Rénique and Adrián Lerner

hooded individuals ransacked an electoral center and burned the bal-

lot boxes in the small village of Chuschi, Ayacucho. The event went
largely unnoticed in the Peruvian media’s electoral coverage. In the fol-
lowing months, dogs were found hanging from lampposts in Lima,
with signs that denounced Deng Xiaoping’s “betrayal” of Mao Ze-
dong’s line at the helm of the People’s Republic of China. Mystery
shrouded these actions. For those familiar with the intricacies of the
Peruvian Left, however, there was no doubt about their senderista or-
igins. The Shining Path was the most radical voice in a Maoist chorus
derived from the moscovita–pekines (Sino–Soviet) split of the 1960s,
which produced myriad “Communist Parties of Peru,” usually distin-
guishable by their surnames: Red Star, Red Flag, Pukallacta, Shin-
ing Path. Each claimed to be the legitimate nuclei for the “reconstruc-
tion” of Peru’s Communist Party founded in the 1920s by Jose Carlos
Mariátegui. The Shining Path’s call to launch a “people’s war” was
its distinctive feature. By 1980, however, few believed it was a realis-
tic goal.
The Peruvian Left had emerged in the 1920s in the symbolic space
shaped by a radical tradition fi rst articulated by Manuel González Prada
after the nation’s traumatic defeat in the War of the Pacific against
Chile (1879–1884). A poet and essayist, González Prada penned fi-
ery texts denouncing the Peruvian republican project as a mere con-
tinuation of colonialism, represented by the centralist power of coastal
Lima, which had been founded by conquistadors as the capital of the
Viceroyalty of Peru to rule over the Andes and its peoples. He saw the
conflict between coast and highlands as more than geographic: to craft
a modern nation, radicals would have to end the dominance of coste-
ños and criollos (those who lived on the coast and were of European
descent) and fi nd a “true Peru” in the heart of the Andes, where the In-
dian masses still awaited redemption. The formula of a “long march”
to redress the course taken since the Spanish conquest in 1532 would
become a second skin for twentieth-century Peruvian leftists (Rénique
2015a).
From then on, a range of “isms” (indigenismo, serranismo, in-
caísmo, telurismo) explored and illuminated the coast-highland gap,
advocating for a Peruvian nationhood based on Andean roots. In the
1920s, two of them gained extraordinary traction. To varying degrees,
the new revolutionary visions crafted by Víctor Raúl Haya de la Torre
and José Carlos Mariátegui, as well as the movements they founded,
aprismo and comunismo, carried the imprint of this rhetorical arse-

Soifer_6844-final.indb 20 8/17/18 11:53 AM

 The Last Peasant War in the Andes 21

nal. While the former’s roots lay in the proletarian mixed-race (mes-
tizo) urban and rural populations of the country’s coastal region, the
latter persisted in the search for the Indians of the sierra as the main
revolutionary actors. At least in theory, searching for the mobilizing
myths of the Andean masses against Lima’s postcolonial domination
continued to be the defi ning mark of the Peruvian Left after the defeat
of the upheavals of the 1930s. Hopes for an effective long march from
the countryside to the city then reappeared in the 1960s, as hundreds
of peasant communities invaded haciendas throughout the highlands.
Both the guerrillas of the Revolutionary Left Movement (MIR) and
National Liberation Army (ELN) of 1962–1966 and Juan Velasco’s
Revolutionary Government of the Armed Forces of 1968–1975 fell
within the scope of the radical tradition.
By the 1970s, Maoism was one of the three major branches of a Pe-
ruvian Left that was experiencing explosive growth. The “muscovite”
Communists, together with the “reformist” groups, adopted a position
of “critical support” to Velasco’s “revolution” and its agenda, which
largely focused its efforts on a massive agrarian reform. However,
as historian Jaymie Heilman has recently shown, it was not uncom-
mon for grassroots actors to resent the implementation of the reforms
as an outside intrusion. It could even derail some of the longstand-
ing political struggles of increasingly dynamic peasant organizations
and other activists—the men and women who in their everyday lives
experienced, and indeed often successfully challenged, what genera-
tions of radicals, including the anti-oligarchic reformist generals, de-
scribed as the quasi-static “true Peru” (Heilman 2017). Moreover, and
despite the nationalist military regime’s atavistic, pro-peasant rheto-
ric, intensive demographic and socioeconomic transformations by then
meant that the proverbial “true Peru” had in fact become something
akin to a floating signifier. Following agrarian pressure and disloca-
tion, massive immigration toward urban centers on the coast meant
that the highlands were not the country’s demographic core anymore.
Lima alone accounted for close to 30 percent of the nation’s inhabi-
tants by the 1980s (Matos Mar 2004; Calderón Cockburn 2016). Un-
der those circumstances, emerging groups and social movements could
be the crucial targets for leftist expansion. The immigrants, the new
limeños or informales, became the protagonists of the renewed revo-
lutionary agenda drafted by a “new Left.” Inspired by the Cuban Rev-
olution, the new Left consisted of a series of small organizations in a
chronic state of atomization, all of which aspired to radicalize velas-

Soifer_6844-final.indb 21 8/17/18 11:53 AM

 22 José Luis Rénique and Adrián Lerner

quista reforms by riding the wave of popular mobilization triggered

by the military government’s initiative. For them, José Carlos Mariá-
tegui was still a guiding source, although now largely in conjunction
with the likes of Antonio Gramsci, Alain Touraine, Ernesto Laclau, or
José Aricó as part of an ideological reassessment in which democracy
and the “bourgeois” culture of human rights were fundamental val-
ues. Rural scenarios, and the very idea of a long march from the coun-
tryside to the city, became secondary. For this increasingly sophisti-
cated Left, the agrarian question no longer was at the center of Peru’s
revolutionary transformation.
That was clearly not the case with their Maoist comrades. They
chose, instead, to wholly reject velasquismo, arguing that it was a
“fascist” regime that had to be resisted through radical mobilization.
The “people’s war” was their obvious next step. The senderistas were,
as historian Iván Hinojosa explained, the “poor provincial cousins” of
Lima’s more mainstream Left (Hinojosa 1998). While leftists in Lima
were distancing themselves from the revolutionary paradigm of the
1920s, senderistas saw the reforms of the Velasco regime as an op-
portunity for ideological and organizational reaffi rmation. This con-
viction compelled them to take back the abandoned “shining path” of
José Carlos Mariátegui, particularly his ideas about the peasantry as
the only force capable of ending the semifeudal structures that were
still the pillars of what they termed “bureaucratic capitalism” in Peru
(Rénique 2003, 20–53; Rénique 2015b). In this narrative, Mariátegui
remained the founding figure of a proletarian party whose legacy was
at risk of being tarnished by the new Left’s “revisionism,” defi ned by
Guzmán as the “complete negation of Marxism,” and usually embod-
ied in the “long and putrid experience with frontism (frentismo).” The
United Left was the latest Peruvian version of that experience; it was
nothing less than “a cancer” to be “ruthlessly eliminated” (Guzmán
Reynoso 1988). Mariátegui had introduced the “universal truths” of
Marxism-Leninism to Peru, in a process that concluded with the emer-
gence of the Shining Path. Their ability to initiate “popular war” was
proof that they were Mariátegui’s legitimate children (Guzmán Rey-
noso 1989). The fusion of Mariátegui and Mao made such a fi nal
step possible. In Peru, according to Guzmán, “we fi nd theses similar
to those that Mao has made universal; today Mariátegui would be a
Marxist-Leninist-Maoist” (Guzmán Reynoso 1988).
Maoist practice, however, often contrasted with such combative
rhetoric. According to a study based on the major public university

Soifer_6844-final.indb 22 8/17/18 11:53 AM

 The Last Peasant War in the Andes 23

in Lima, Maoism captivated students of migrant backgrounds, who

found it a useful tool for carving a place for themselves within a sys-
tem that tended to marginalize them. A Peruvian brand of Maoism
was forged largely through imagery—disseminated by Chinese propa-
ganda—of the Great Cultural Revolution, such as Red Guards ques-
tioning the establishment at will, as well as through decontextualized
quotations from Chairman Mao, which provided a veneer of ideologi-
cal empowerment encouraged by the “Little Red Book” and the man-
uals of Marxism that originated in Beijing and Moscow (Lynch 1990).
Under the influence of this Maoist campus culture, which compul-
sively encompassed all levels of the university community, public uni-
versities tended to enter into a self-destructive process of radicalization
(Degregori 1990b).
Whereas Lima’s leftist intellectuals tentatively hinted at the possi-
bilities of formulas such as Cholo-comunismo (an Andean version of
the “Eurocommunism” of those years), or even of Haya-leninismo (as
the basis for a Left-APRA alliance), senderista identity seemed fi rmly
poised to undertake the challenges to come. Its leader and ideologue
Abimael Guzmán charted the route ahead in distinctively terse fash-
ion: “The class gives birth to the Party and the Party rises and begins
to walk, it is the child of the revolution. The Party can never be crushed
or destroyed. The Party will inevitably triumph. This Party forged it-
self, Mariategui is its founder. It is done” (CC-PCP 1979). Closely fol-
lowing the Mao of the 1930s, Guzmán believed in the transformative
power of peasant mobilization, “a hurricane so swift and violent that
no power, however great, will be able to hold it back.” He framed the
peasantry as a force capable of sweeping “all the imperialists, war-
lords, corrupt officials, local tyrants and evil gentry into their graves.”
In the ideological-strategic scheme of the Shining Path’s leader, this
massive, class-based mobilization would force every potential revolu-
tionary to pick one of only three possible alternatives: to trail behind
the peasant movement; to stand in its way and oppose it; or to march
ahead and lead it toward the seizure of power (Mao 1927). By adopt-
ing the latter choice, Guzmán believed, his Communist Party of Peru–
Shining Path would be forged as a party of “iron-clad cadres who dare
to challenge death to snatch the laurels of victory” that was capable of
unleashing a new era (Guzmán Reynoso 1988, CC-PCP 1986a).
Sendero’s “people’s war” was, therefore, a fundamentally ideologi-
cal product. The conviction and willpower that made it possible were
based on the dissemination of a set of beliefs in which revolution was

Soifer_6844-final.indb 23 8/17/18 11:53 AM

 24 José Luis Rénique and Adrián Lerner

a precondition for true Peruvian nationhood (this was “the path” of

Mariátegui and the radical tradition), and on a universal revolution-
ary doctrine (Marxism-Leninism-Maoism) which held that at the cru-
cial moment of revolution, subjective “spiritual factors” could be more
decisive than objective “material” ones (Cook 2014). This meant that
the “Initiation of the Armed Struggle” (Inicio de la Lucha Armada—
ILA), the launching of the war, was the key issue of revolution. Initia-
tion would also solve the problem of leadership. It enthroned Abimael
Guzmán as “President Gonzalo,” the supreme political chief and strat-
egist of revolution. It turned him into the infallible creator of the the-
ory of the people’s war in Peru—the so-called Pensamiento Gonzalo
(“Gonzalo Thought”). In “Gonzalo’s” view, violence was a “supra-
human historical fact” that individuals would assume without any
moral qualms. After the initiation of armed struggle, war itself would
provide the elements for victory (Poole 1994).
The rebellion would then start as a formidable act of will, carefully
planned by its leader. Guzmán himself boasted that he was perfectly
aware that “rivers of blood” would flow (Guzmán Reynoso 1988). The
revolution would coincide with the recurrence of crisis in the country
every ten years. Those relapses, he believed, expressed the inability of
“bureaucratic capitalism” to acquire stability in a society still char-
acterized as semi-feudal, despite the military regime’s much bragged-
about agrarian reform (CC-PCP 1985). From this perspective, the cri-
sis of the late 1970s, with its massive social mobilization that included
the two largest strikes in the country’s history among its social and
demographic transformations, announced itself as a particularly dra-
matic one. Moreover, it was combined with general elections and a re-
gime transition from an increasingly conservative and repressive mili-
tary dictatorship to electoral democracy. The candidate favored to win
the presidential election was Fernando Belaunde. After having been
ousted in 1968 by a coup carried out by armed forces emboldened by
their anti-guerrilla campaigns of the 1960s, Belaunde could hardly
have been expected to decisively deploy the military against a new rev-
olutionary insurrection. That it also would likely take a new govern-
ment some time to become fully operational would make “conditions
riper for revolution.” It was “a very favorable juncture for initiating
the people’s war.” This is why the Shining Path launched its war on
the eve of the presidential elections (Lerner 2017; Lust 2013; Guzmán
Reynoso 1988).
Ideology provided the necessary cohesion. Mao’s quotations consti-

Soifer_6844-final.indb 24 8/17/18 11:53 AM

 The Last Peasant War in the Andes 25

tuted the doctrinaire bedrock. Oral transmission was as important as

the written word in filtering down the basic principles of the people’s
war. Later, as combat progressed, “Gonzalo Thought,” based on the
adaptation of Marxism-Leninism-Maoism to Peruvian reality, and on
Guzmán’s interpretation of the ideas of José Carlos Mariátegui com-
bined with those of Mao, became the blueprint for collective action.
Mariátegui was portrayed as the discoverer of the revolutionary po-
tential of Peru’s indigenous peasantry, who had synthesized a secu-
lar history of oppression into a class-based or “proletarian” view of
what he termed “The Indian Problem” (Mariátegui 2007). Mao had
inscribed peasant rebellion into a broader revolutionary tradition, go-
ing back to the “Paris commune” of 1871, and his was deemed the
most advanced version of Marxism-Leninism because of its poten-
tial for success in peripheral countries. Guzmán made two visits to
China during the 1960s, the second of which coincided with the early
years of the Cultural Revolution and the doctrinaire debates of the
Red Guard. During those trips, he took courses in politics, philoso-
phy, and military strategy. He left a young cadre and obscure philoso-
phy professor, originally schooled in the Stalinist bureaucratic ways of
the local Communist Party, and returned fully converted and imbued
with the prestige that contact with the Revolution provided. From then
on, “Mao thought” was “the spiritual bomb of the poor” (Lin 1966),
and “a single spark could start a prairie fi re” (Mao 1930). Guzmán
thereafter oriented his political work to adapt the script of the Chi-
nese peasant wars of the 1930s and the 1940s to the Peruvian scenario
of the 1980s. His efforts coincided with an expansion of radical mili-
tancy (including Maoism) in Peru, particularly on university campuses
(Rothwell 2013, 55–62).
Some of the social and political fuel for the transformation of an or-
ganization that counted less than twenty members in 1970 into the jug-
gernaut of the 1980s came from recent nationwide processes. The first
was a unique cycle of agrarian mobilization. Peasants had been mobi-
lizing, occupying lands, and generally eroding the traditional agrar-
ian regime of the Peruvian highlands since the 1950s, but the process
was decisively propelled by the state in support of its agrarian reform
in the fi rst half of the 1970s. This led to the creation of a national net-
work of organizations, from district leagues to departmental federa-
tions and a National Agrarian Confederation inaugurated in 1970 in
the unused building of the Parliament. A heavily politicized and mobi-
lized countryside was the result (Cant 2015). Although the turn to the

Soifer_6844-final.indb 25 8/17/18 11:53 AM

 26 José Luis Rénique and Adrián Lerner

right that followed the toppling of Velasco by the “Second Phase of the
Revolution” led by General Francisco Morales Bermúdez (1975–1980)
severely limited the scope of the agrarian reform, in reality its design
and implementation had always proved extremely problematic. What
effectively resulted was a truncated agrarian reform that left a host
of unresolved problems, voids of power, and confl icts; a landscape of
what Enrique Mayer has called “Ugly Stories,” many of which allowed
senderista activists propitious opportunities to establish their presence
locally (Mayer 2009). Finally, at the top, the unique political juncture
was defi ned by a transition to democracy whereby the armed forces
handed power back to the same civilian ruler they had ousted twelve
years earlier.
The Peruvian state was far from a unified entity as the country faced
the insurgent challenge of the 1980s. In terms of the cleavage between
civilian and military leaders, twelve years of a regime led by generals
who constantly stressed the inability of civilians to rule had worsened
lingering civilian-military distrust. As discussed in chapter 4 of this
volume, the other main coercive force of the state, the police, faced nu-
merous internal issues and also had strained relationships with both
the armed forces and the civilian population. Policing was divided into
three separate forces, the Civil Guard (Guardia Civil), the Republican
Guard (Guardia Republicana), and the Investigative Police (Policía
de Investigaciones del Perú), with different and sometimes compet-
ing and overlapping identities and functions. Recurring authoritarian
regimes had used police forces for political surveillance and persecu-
tion, severely damaging their popular image. Chronic lack of funding
further deteriorated police forces, marking them as unpopular, ineffi-
cient, and corrupt institutions. The military, furthermore, had tended
to treat the police as their subalterns. Tensions reached a boiling point
in the “Limazo” of February 5, 1975, during Velasco’s government, as
a police strike in Lima was followed by widespread unrest and looting,
to which the regime responded by unleashing military troops against
the police and civilians. Moreover, the armed forces did not constitute
a cohesive front either. The military regime of 1968–1980 had been
marred by divides between “radicals” and “conservatives,” often in-
tertwined with traditional rivalries between the army, the navy, and
the air force on the one hand, and with ethnic and class differences on
the other hand.
Parallels between the early 1980s, with the guerrillas, and the
1960s, with counterinsurgent military interventions and Belaunde’s

Soifer_6844-final.indb 26 8/17/18 11:53 AM

 The Last Peasant War in the Andes 27

demise, did indeed seem ominous. As Guzmán hoped, this made what
was already a delicate situation even more politically dangerous. The
Belaunde government initially framed the Shining Path as common
criminals and, as a consequence, opted for a police-based response,
even after the fi rst State of Exception was declared in October 1981.
The police intervention carried its own problems. Relations between
police and local populations throughout Peru were already invariably
strained and occasionally violent well before the onset of the confl ict.
Local police officers were often in shameful material conditions and,
particularly in rural settings, physically isolated. They were used to
dealing with petty crime, not well-organized attacks by revolutionary
groups. These conditions, along with broader issues affecting police
forces, led to a mix of ineffectiveness and brutality in the immediate
police response to the Shining Path in Ayacucho. The decision to grant
the armed forces exceptional power over vast regions in January 1983
put them in an impossible position: the demands of the war were dis-
proportionate to their capacities and to their constitutional mandate.
They were expected to control a conflict that had already unfolded
in decisive ways. They were sent to crush a fringe rebellion, only to
fi nd themselves enmeshed in a surprisingly long, irregular, and bloody
war, during which they became the second-greatest perpetrators of hu-
man rights violations and suffered substantial casualties. Significantly,
many of their shortcomings can also be traced to their own ideas and
actions, which were based on strategies unsuited to the challenge
posed by the Shining Path in Ayacucho and beyond. The underlying
ideology included an obsession with traditional geopolitics that led to
massive spending sprees on airplanes and other ineffectual equipment.
The strategy was initially based on anti-guerrilla experiences and field
manuals from the 1960s, and on the leadership of high-ranking offi-
cers trained in the infamous United States Army School of the Ameri-
cas. It initially framed the confl ict in the traditional military terms of
regular counterinsurgency and territorial domination.
The situation for Andean peasants deteriorated when the strategies
and attitudes of state coercive forces encountered the everyday reality
of the Shining Path’s war in Ayacucho. The quest for territorial con-
trol and for clearly identifiable sides clashed with the Maoists’ hit-and-
run tactics, infiltration of the local population, and strategic exploi-
tation of state violence to bolster their long-term claim for hegemony.
Later during the confl ict, the Peruvian armed forces adopted a policy
of military alliances with peasant communities, shifting to more inte-

Soifer_6844-final.indb 27 8/17/18 11:53 AM

 28 José Luis Rénique and Adrián Lerner

gral forms of “political warfare” and to “low intensity” warfare and

targeted attacks. The police, in the meantime, made significant prog-
ress in the realm of intelligence. These changes would eventually prove
extremely successful in securing the defeat of the Shining Path, but
also led to systematic human rights violations.

Spark into Fire: The Onset of War in 1980s Ayacucho

May 17, 1980, became the key date in the Shining Path’s revolutionary
calendar. It was the so-called ILA, during which the party switched
from “labor with unarmed hands” to a display of the “armed word.”
The ILA implied a “red zone” in the north of Ayacucho, on an axis
that linked Huanta and Huamanga (Guzmán’s original strongholds),
based on the gradual penetration of villages and peasant communities.
It also meant that the senderista leadership would rely on a network of
cadres capable of articulating regional plans to be replicated through-
out the country. Ultimately, the ILA would position the organization
as the revolutionary vanguard, above a “revisionist” Left that increas-
ingly embraced “bourgeois democracy.”
The regional context of Ayacucho made the rise of Sendero possi-
ble. Ayacucho offered a microcosm of the contradictions of Andean
modernization that had emerged in the two decades preceding the sen-
derista uprising, in line with a post-oligarchic national transition that
offered no clear break from the past. By the late 1970s, the contra-
dictory effects of the velasquista agrarian reform had been added to
stagnation, marginality, regional dismemberment, and other tradi-
tional negative effects of capitalist penetration in the region’s country-
side. The anti-oligarchic reform had hit large landowners the hard-
est; it was the fi nal blow to the traditional hacendado departmental
elite. But it allowed the survival of small landowners at the district
level: gamonal power, in many ways the “heart of darkness” of the
old agrarian regime in the Andes, remained in place. The reopening
of the Universidad Nacional San Cristobal de Huamanga (UNSCH)
in 1957 completed the unique regional picture. It became a sociopolit-
ical laboratory in which radical professors like Guzmán established a
direct connection with the offspring of peasant communities. The ex-
ceptional opportunity was, as senderistas would put it, “a gift from
the reactionary establishment.” The UNSCH represented the long-
pursued promise of regional development based on scientific knowl-

Soifer_6844-final.indb 28 8/17/18 11:53 AM

 The Last Peasant War in the Andes 29

edge, but also aided the radical transformation proposed by the Mao-
ist texts discussed in student circles and “popular schools.” Students
went looking for the tools of modernization and returned as agents
of revolutionary change. Their exuberance resulted in posters that
showed Guzmán as a blend of Amauta (the traditional image of the
Andean teacher, often associated with Mariátegui) and “Great Helms-
man,” followed by a mass that recalled both the Chinese Red Guard
and the immemorial image of Andean jacqueries or peasant revolts.
Not by chance, Huamanga was described as an “Andean Yenan” (De-
gregori 2011, 23–51; Rénique 2013).
Similar patterns of local politicization had been established in Aya-
cucho since the early twentieth century, but they had always been con-
nected to national currents originating from the political center in
Lima: aprismo, trotskismo, belaundismo, and velasquismo (Heilman
2010; La Serna 2012). The endogenous peripheral origin of the move-
ment was unusual. Never in Peruvian political history had a party
with defi ned revolutionary objectives emerged in the country’s periph-
ery to challenge central power. This peripheral status allowed Sendero
to shape an era of “enigma, exoticism, surprise.” The rise of the insur-
gency was as shocking as the crumbling of the state (Stern 1998a). But,
as Steve Stern has recently asked, again: How could a small group of
revolutionaries from one of Peru’s most backward regions, so out of
step with the rest of the Left’s acceptance of electoral transition from
military rule to democracy, redirect the course of national political
life? (Stern 2013). The backwardness, isolation, and marginality that
were associated with the senderistas became factors that helped the
Shining Path become such a formidable challenge, a threat that was
properly understood only when it was too late to be easily contained.
From its Maoist ayacuchano perspective, the movement exploited the
fissures of Peruvian society with surprising effectiveness. Claiming to
represent the semi-feudal character of the country, they advocated the
historical significance of a peasant-based war for Peru, knowing that
their strategy would sink the country into a bloodbath. Carlos Iván
Degregori witnessed fi rsthand, as a political rival, the transition of the
Shining Path from university party to war machine. He later wrote
that the process evoked a “dwarf star,” in which “matter gets so com-
pressed that it acquires a great specific weight, disproportionate to its
size.” An accumulation of willpower, discipline, and ideological co-
hesion evolved from a position of exceptional marginality, just when
the “revisionist” Left was reaching its zenith, leading to the extreme

Soifer_6844-final.indb 29 8/17/18 11:53 AM

 30 José Luis Rénique and Adrián Lerner

underestimation of the political acumen of senderismo (Degregori

1990a). Ayacucho offered the conditions to take this process of encap-
sulation to unexpected heights. There, Andean Maoism would achieve
its purest form.
Wearied by the vast waves of urban mobilization they had been
forced to confront since 1976, which included two massive national
strikes and several regional ones, the armed forces consistently disre-
garded intelligence reports about radical activities in Ayacucho. Even
the army’s official history of the war cites early reports about Maoist
organizations, and the violent struggles over free education in Huanta
in 1969 had already revealed the leadership (and momentary imprison-
ment) of one Abimael Guzmán (Comisión Permanente de Historia del
Ejército del Perú [hereafter CPHEP] 2010, 17; Degregori 2011). Fur-
thermore, before handing over power to Belaunde, the military report-
edly got rid of archives containing critical information on subversive
activities in the rural districts of the south-central highlands (Gorriti
1990, 83). Blatant myopia about the enemy was an inevitable result. In
April 1984 the minister of war still believed that Maoists in Ayacucho
were getting support from Colombia’s M19 or Chile’s MIR guerrillas
(Pease 1989, 582). In a memoir published in 1989, General Clemente
Noel of the Peruvian army, who ruled Ayacucho with an iron fist as
its fi rst political-military commander in 1983, still traced the roots of
the conflict to the guerrillas of the 1960s, and described the strategy of
the Shining Path in terms of their territorial gains (Noel Moral 1989,
25, 41). As stated in the introduction to this volume, scholarly disori-
entation added to the confusion of the early 1980s. Stuck in Cold War
grand revolutionary equations, scholars of Peru had a hard time figur-
ing out the singularities of provincial insurgents (McClintock 1998, 3;
see also Poole and Rénique 1991; Starn 1995a; Starn 1991; Degregori
2012b, 37–70).
During the fi rst years of the confl ict, the insurgents profited from
their own effective strategy and organization as well as the fragility of
the state in Ayacucho and poor decisions by the central government.
While the attack on Chuschi’s voting center became the symbolic start
of the armed struggle, the Shining Path had hit before, striking a min-
ing camp in Cerro de Pasco in Peru’s central highlands and a police
station in the lower-class suburb of San Martin de Porres in Lima.
Still, the most important initial efforts were carried out by its Main
Regional Committee (Comité Regional Principal) in Ayacucho, Apurí-
mac, and Huancavelica, all poor, heavily indigenous rural regions of

Soifer_6844-final.indb 30 8/17/18 11:53 AM

 The Last Peasant War in the Andes 31

the south-central highlands. This was Sendero’s stronghold. Trained

groups known as “armed groups without arms” attacked police sta-
tions and mining centers in isolated towns. These raids were conducted
to execute authorities and steal weapons, ammunition, and dynamite
for future targets. They also reinforced the central role of ideology in
the Shining Path’s strategy: they did not need weapons to launch the
revolution. Conviction sufficed. By 1982, they had made important
military progress, begun attacking larger towns, and transitioned to-
ward more vicious, symbolically charged acts of violence against po-
lice officers and other local leaders (CVR 2003a, 2:32–33).
Some of the most consequential landmarks of this process were the
attack on the police station in Tambo (La Mar, Ayacucho), on Octo-
ber 11, 1981; the escape of Shining Path militants from Huamanga’s
prison on March 2, 1982; the attack against a police outpost in Vilcas-
huamán near the end of that month; the death of Edith Lagos, one of
the militants who had fled prison, on September 2, 1982, and the fu-
neral that followed it; and the decision by President Belaunde to grant
complete control of the “Emergency Zones” to the armed forces on
the last day of that year. The attack against the police in Tambo led
to the fi rst declaration of the state of exception in Ayacucho. It also
brought the fi rst large contingent of antisubversive police forces, a
paratrooper unit known as “Los Sinchis,” into the confl ict. Their re-
cruits came mainly from Lima and other coastal cities, and they al-
ready were notorious in Ayacucho for their repression of the movement
for free education in Huanta in 1969. Their return in late 1981 was a
turning point in the confl ict. Early on, the Sinchis managed to push
back against Sendero’s momentum after months of initial “easy” vic-
tories against ill-prepared police officers who often were repudiated by
local residents due to patterns of abusive behavior. But the Sinchis also
became a symbol of state brutality. Stationed for long periods under
bad conditions in regions they did not know, and fighting an enemy
that was indistinguishable from the local populations they were sup-
posed to protect, the Sinchis were the beginning of a state-led “dirty
war” that led to numerous episodes of deeply racist and gendered vi-
olence against local peasants, playing directly into the Shining Path’s
long-term strategy of generating a wave of “reactionary violence” that
would eventually feed their peasant army (CVR 2003a, 2:109).
Sendero’s attack on the Huamanga prison in March 1982 created
an aura of invincibility. A Shining Path detachment launched a coor-
dinated attack that saw armed men free prisoners, while strategically

Soifer_6844-final.indb 31 8/17/18 11:53 AM

 32 José Luis Rénique and Adrián Lerner

located sharpshooters made sure they all could leave the city. Sendero
appeared to control the whole city. More than three hundred prison-
ers, including seventy senderistas, escaped. Days later, police officers
kidnapped wounded members of the Shining Path from the city’s hos-
pital and murdered them, in an episode that fed widespread percep-
tions of police ruthlessness. Less than a month later, a minor attack
against the police station in Vilcashuamán resulted in an injury to a
guard. President Belaunde traveled to the site to show support for the
police. This marked a defi nitive step in the insurgency’s transforma-
tion into a national issue.
One of the senderistas who had escaped Huamanga’s prison in
March was nineteen-year-old Edith Lagos. In September, Lagos died
in an armed confrontation in Apurímac. Her funeral in Huamanga
was attended by a crowd calculated by some observers to have reached
tens of thousands. Even the archbishop of Huamanga, a staunch anti-
Communist, performed a special mass for Lagos (Guerrero 2006;
Caro Cárdenas 2006). These events, along with an ever-increasing
number of attacks, seemed to turn the tides of the war, and of part of
the public sphere, in favor of the unknown band of Maoists from Aya-
cucho. A military witness described the period as “the golden years
of senderismo” (Hidalgo Morey 2004, 112). Initial victories attracted
militants who had distanced themselves from the electoral path taken
by their “legal Left” organizations, which they saw as a betrayal of the
principles of the new Left, originally born as a response to the “pac-
ifism” of the “old left” (Pumaruna 1967; Béjar 1990; Flores Galindo
2007; Lust 2013).1
The government’s reluctance to send the armed forces to fight the
insurgents unnerved the military and large segments of the popula-
tion: by mid-1982, a poll revealed that 60 percent of Peruvians wanted
the armed forces to take over in Ayacucho (CPHEP 2010, 42). At the
same time, the electoral Left, a political force at the time, viewed mil-
itary intervention with apprehension. The minister of war, General
Luis Cisneros Vizquerra, warned that if the armed forces were sent to
the Emergency Zones, they would inevitably kill lots of innocent peas-
ants. Cisneros also repeatedly confronted the legal Left, accusing them
of carrying out subversive acts themselves. In the meantime, legislation
passed as early as 1981 helped frame the conflict in terms of “terror-
ism,” eventually allowing the armed forces great leeway in combating
their enemies and helping to create a long-lasting narrative about the
war (CVR 2003a, 2:45, 175).

Soifer_6844-final.indb 32 8/17/18 11:53 AM

 The Last Peasant War in the Andes 33

Amid this ambivalent political climate and constant escalation of

violence, the Shining Path proclaimed the formation of its “People’s
Guerrilla Army” (Ejército Guerrillero Popular) on December 3, 1982,
to celebrate Guzmán’s birthday. Weeks later, Belaunde gave in. In what
was characterized as a true “abdication of democratic authority,” he
opted for a purely military solution, giving the armed forces authority
to control the Zones of Emergency, which expanded in the following
years along with subversive activity to encompass vast territories (Bar-
ton 1984). Close to two thousand personnel from the army and navy
were sent to Ayacucho. Military witnesses described a tremendously
hostile atmosphere as the troops reached Huamanga and other cit-
ies, and, later in 1983, massive electoral absenteeism in local elections
showed the degree to which Sendero swayed regional politics (CPHEP
2010, 48, 54–55).
The armed forces immediately banned journalists and human rights
activists from the Emergency Zones. Only days after their arrival in
early January 1983, it was reported that peasants in Huaychao, a ru-
ral village in Huanta, had killed a column of senderistas. The military
and the president publicly praised them and encouraged others to fol-
low their example; others suspected it was a cover for state repression.
When a group of journalists from Lima traveled to the region, despite
the prohibition, to investigate, tragedy struck: Sinchis had told peasant
communities to attack any newcomers, and the peasants of Uchurac-
cay did just that, murdering the journalists and their guide with stones
and sticks. For many, the Uchuraccay massacre became the ultimate
symbol of the conflict. It put Ayacucho at the center of media atten-
tion and further polarized all sides. It also led to myriad interpreta-
tions, all of which brought Andean peasants under scrutiny. Perhaps
the most prominent interpretation, issued by the state-sponsored in-
vestigative commission led by the famed writer Mario Vargas Llosa,
contained long-standing prejudices about “primitive” Andean peas-
ants and reduced the problem to a clash of civilizations living in isola-
tion from one another: a modern, coastal, urban Peru, and a “deep,”
primitive, rural, Andean hinterland. As a scholarly critic later argued,
these ideas, along with the violence itself, revealed that Peru was “in
deep trouble” (M. Vargas Llosa 1983; Mayer 1991). Only very re-
cently have scholars shed light on the “Uchuraccay tragedy” in all its
historical and local depth, showing that it was marked by the mem-
ory of long struggles over community, land, and citizenship (del Pino
2017). The scars were the product of contact, not of isolation.

Soifer_6844-final.indb 33 8/17/18 11:53 AM

 34 José Luis Rénique and Adrián Lerner

When the armed forces entered the Shining Path’s “red zone,” the
conflict acquired its full, bloody, face. The Andes had not seen such a
degree of violence since the Chilean invasion in the central highlands
in 1881–1884, or perhaps the Túpac Amaru rebellion of the 1780s.
With no clear strategy, severely limited intelligence, and hindered by
the influence of colonialist counterinsurgent formulas ranging from
the French guerre révolutionnaire to the American counterinsurgency
of the Vietnam era and Argentina’s “fight against subversion,” the mil-
itary behaved as an occupation force (Cobas 1982; Masterson 1991;
Rodríguez Beruff 1983; Villanueva 1973; Degregori and Rivera Paz
1993). The militarization of the state was an old idea among the Peru-
vian military, and it fueled both Belaunde’s and the Left’s misgivings
and Guzmán’s boldness (Toche Medrano 2008, 3). From that perspec-
tive, the velasquista and the Southern Cone regimes had been two dif-
ferent modes of counterinsurgent politics: the fi rst was aimed at the
cooptation of the popular sectors, while the second sought the elimi-
nation of the insurgency, expecting to obtain a quick victory regard-
less of social and political costs (Gorriti 2003, 97).
According to a military source, “the nature of the enemy as well
as the circumstances” often made “indiscriminate repression” the
norm (Hidalgo Morey 2004, 211). Very few listened when the mili-
tary claimed that, to achieve a lasting victory, the state would have to
fulfill its role and take care of the structural socioeconomic and polit-
ical problems of the country (Noel Moral 1989, 38). It took only two
months for General Noel to be sued by a public attorney for the al-
leged murder of three Ayacucho peasants, and weeks later Belaunde
told the press that he would throw human rights reports “directly
into the trashcan” (CVR 2003a, 2:182–183). Army barracks like “Los
Cabitos” in Huamanga became killing and torture grounds, and con-
flicts increased between the armed forces and the police, which had ef-
fectively become a subaltern force, trained in counterinsurgent tactics
on the field by the military, frequently in brutal ways (CVR 2003a,
2:114–115). As a historian of the period put it, peasants by then of-
ten feared Sendero, but the police and armed forces were often hated
(Taylor 1998, 94; CPHEP 2010, 50). Noel’s successor in 1984, Gen-
eral Adrián Huamán, had experience fighting the guerrillas of 1965.
But he was also a native of Ayacucho, fluent in Quechua, and advo-
cated a more “social” and developmental approach to the war that was
based on efforts to gain the support of the local peasant communities.
He was known for paternalistic gestures, from the traditional distribu-

Soifer_6844-final.indb 34 8/17/18 11:53 AM

 The Last Peasant War in the Andes 35

tion of bread to the organization of the harvests and control over so-
cial spending in the region. He also supported the notion of peasant
self-defense organizations (comités de autodefensa, or CAD) to com-
bat the insurgents, although he refused to arm or train them. Never-
theless, his approach clashed with civilian fears of a new military take-
over of government functions and political life, and he did not last
long. Crucially, moreover, human rights violations by the armed forces
spiked under his leadership, and the fi rst mass graves were found in
Ayacucho (CVR 2003a, 2:183).
The military onslaught did see the armed forces inflict significant
casualties on the rebels. The military increased its number of patrols
and took back police stations and villages after a long period of ab-
sence. But they were fighting the wrong war, misreading their enemies’
strategies. The insurgents did not budge. They too chased victory re-
gardless of human costs. The territorial approach to the war taken by
the armed forces meant that they were susceptible to the Shining Path’s
hit-and-run attacks, and whenever they took back a village, often bru-
tally, they would then leave it at the mercy of senderista retaliation. It
was the era of massacres and mass graves. Among many examples, the
emblematic case of the Santiago de Lucanamarca massacre illustrates
the level of degradation in those years. In mid-March 1983, a com-
munity vigilante group killed a Shining Path operative in a village in
Huancasancos, Ayacucho, an early symptom of the gradual souring of
the senderista relationship with the local population. In revenge, sixty
rebels captured and killed sixty-nine individuals, including women
and children, on April 3, using axes and machetes or shooting them
in the head at close range. The killing was cold and deliberate. In his
1988 interview, Guzmán confirmed the authorship and its motivation.
In the face of an attempt to “use the masses against the Party,” “[t]he
Central Leadership itself planned the action and gave the instructions.
That’s how it was. [. . .] They understood that they were dealing with
a different kind of people’s fighters, that we weren’t the same as those
they fought before” (Guzmán Reynoso 1988). Lucanamarca was part
of a broader pattern. Numbers confi rm the dramatic turn in the con-
flict that many by then were calling a “war,” although another two de-
cades passed before serious quantitative research was published. With
3,996 fatalities and 2,742 detainees, 1983 and 1984 were the bloodi-
est years. Between 1982 and 1983, the number of ammonium nitrate-
fuel oil (ANFO) explosives and dynamite sticks stolen throughout the
highlands went from 12,366 to 185,473 (CVR 2003a, 6:510). ANFO

Soifer_6844-final.indb 35 8/17/18 11:53 AM

 36 José Luis Rénique and Adrián Lerner

and TNT were logical weapons in a mining country, and symbols of a

war built from scratch, without foreign patrons or supplies. Combat-
ants learned on the go and stole their weapons from their enemies. It
was an army rooted in values of sacrifice and commitment, and on the
militants’ willingness to dissolve their individuality in the name of col-
lective action. They paid their “quota of blood,” to use a phrase from
the semi-religious senderista jargon.
For some years, then, “Gonzalo Thought” proved effective with lo-
cal units in charge of specific jurisdictions. Sendero’s initial success
was based on a combination of cohesion and flexibility: although “the
party controls everything,” decisions on how to apply a general direc-
tive dictated by the leadership was left to local commanders. Led by
the graduates of the First Military School, they would decide where
to attack or who to “annihilate.” Local commanders transmitted the
party’s orders, and local recruits provided local knowledge. The “pop-
ular war” moved forward through a patchy dynamic carefully planned
to take advantage of the countless rifts and confl icts within a rural so-
ciety; depending on site-specific circumstances, it also absorbed and/or
destroyed the “peasant work” of other leftist competitors. Locals re-
sponded with their own strategies of survival to an ever-changing con-
flict that often caught them “between two fi res” (Berg 1992; Aroni
Sulca 2016b). While the paradigm of a peasantry “between two fi res”
has been heavily criticized for comparable contexts, it is worth not-
ing that in this case, Sendero’s strategy of “beating the countryside”
(batir el campo) in those years consisted precisely of unbridling vio-
lence from both sides, in order to force peasant communities to take
sides (Grandin 1995).

Feet of Clay: The Expansion of Violence and the Growing

National Crisis during APRA’s First Government, 1985–1990

In a remarkable turn of events in the history of the Andean region, by

the mid-1980s Ayacucho had produced a “syncretic phenomenon” en-
compassing the key elements that Raj Desai and Harry Eckstein have
identified in modern peasant insurgencies: (1) the ideology and organi-
zation of “modern revolutions,” visible in the Leninist party structure
of professional cadres that harnessed and directed the energy of the
masses toward transformative goals; (2) the operational doctrines of
guerrilla warfare as developed in Asian confl icts, which made it possi-

Soifer_6844-final.indb 36 8/17/18 11:53 AM

 The Last Peasant War in the Andes 37

ble to build strength and coercive capacity for specific political goals;
and (3) the “spirit” of traditional peasant rebellion, an “archaic” ele-
ment that, adequately tapped, could supply the “passionate energy” to
propel a major modern insurgency. This mix, according to Desai and
Eckstein, could display the force of a natural calamity, as Peruvians
were tragically learning (Desai and Eckstein 1990).
By the time the national government deployed the military, the
Shining Path insurgents were already expanding their insurrection into
new territory in the jungle, in other highland regions, and on the coast.
In the Amazon rain forest, their efforts were focused on the Huallaga
Valley, a potentially rich region containing the world’s main center of
coca production that offered an ideal terrain for hosting a “guerrilla
army.” In the highlands, the Shining Path would assert its presence in
the “red zone,” move toward larger cities, expand toward the northern
sierra, and, above all, focus on the central highlands. In the central si-
erra, key national transport and electricity grids could be sabotaged.
The central highlands also included important agricultural and min-
ing areas (crucial to getting dynamite). The region was envisioned as
a “second ring” from which to pressure the central coast, particularly
Lima. The national capital, with its “enormous masses of the poor in
its neighborhoods and shantytowns” (the “fi rst ring”) offered promis-
ing possibilities for the “popular war,” and, as the seat of political and
economic power, it was the war’s ultimate prize (Guzmán Reynoso
1988; Kernaghan 2009).
But although Sendero seemed ready to take the offensive, some re-
gions were harder than others to penetrate. Colonel Hidalgo Morey,
for example, acknowledged “military maturity” in the performance of
the senderistas in two critical areas: the Huallaga and Mantaro val-
leys (Hidalgo Morey 2004, 144). The pattern of expansion seen in
Ayacucho and the neighboring provinces of Apurímac and Huancavel-
ica was not possible in Junin or Puno, however, where peasant orga-
nizations were much stronger. The Shining Path met resistance there,
despite managing to gather support early on to dismantle the agrar-
ian units created by the national government’s agrarian reform, which
were often perceived by local peasants as a continuation of the old ha-
ciendas. Their attempts to impose economic autarky and their aggres-
sive methods of recruitment were rejected in these regions, where peas-
ant communities were more mercantile and less “semi-feudal” than
those in the original “red zone” (Manrique 1998; Rénique 1998). In
Amazonian regions, even murkier situations emerged. As it was con-

Soifer_6844-final.indb 37 8/17/18 11:53 AM

 38 José Luis Rénique and Adrián Lerner

sidered a vital zone of refuge from the armed forces, the Shining Path
attempted to impose extreme forms of regimentation in certain Am-
azonian territories, which eventually incited opposition from the lo-
cals, who were organized in ways that were frequently alien to the
senderistas. In a pattern that has been all too common in the relation-
ships between Amazonian indigenous populations and colonists, local
resistance was often met with extreme violence (del Pino 1998). The
Shining Path carried out some of its most devastating attacks against
native populations in the jungles of central Peru. The Ashaninka peo-
ples were hit particularly hard by actions that have been character-
ized as genocidal. The Peruvian Truth Commission estimated that in
a total population of 55,000, nearly one-fifth (or some 10,000) of the
Ashaninka were displaced, 6,000 were murdered, and 5,000 were held
captive in labor camps (CVR 2003a, 5:161–162).
Furthermore, in all regions, including their Ayacucho stronghold,
the very strategies that had allowed Sendero to make progress were be-
ginning to reach their limits. Their manipulation of preexisting local
conflicts and tremendous use of violence to exploit power vacuums in
rural areas to create “the new power” through “popular committees”
eventually came back to haunt them in many places. This became par-
ticularly acute starting in 1983, when the Shining Path began employ-
ing a new generation of local cadres, who often belligerently ignored
traditional communal norms, igniting all kinds of clashes and even
rebellions within peasant communities. The senderistas also reacted
with extreme violence when the armed forces were deployed to Aya-
cucho. Seeking to dispel the notion that the military offensive would
defeat them by 1985, they furiously attacked the communities that
had been reclaimed by the armed forces. These “counter-restorations”
(contrarreestablecimientos) typically placed severe restraints on the
mobility of peasants, disrupted their economic cycles, used their chil-
dren for armed actions, and were marked by shocking acts of violence
such as those at Lucanamarca (CVR 2003a, 2:46–47; Gavilán 2012).
Therefore, senderista expansion, although sizable, was marked by
a “consistent capacity” both “to win and to squander an initial po-
litical base” (Stern 1998b). Acquiring support through “popular jus-
tice” and inciting resistance to their authoritarian ways were two di-
mensions of the same dynamic. It would be years, however, before the
state’s coercive forces could take full advantage of these contradic-
tions. Greater success was perhaps not possible while Peru remained
under Belaunde’s government, which was paralyzed by the memory of

Soifer_6844-final.indb 38 8/17/18 11:53 AM

 The Last Peasant War in the Andes 39

the mid-1960s drama, but also unwilling to engage in a more ambi-

tious developmental and political agenda. Given the limitations of the
Shining Path’s capacity to “conquer bases throughout the country,”
the destiny of the war hinged on the state’s capacity to fi nally adminis-
ter the countryside efficiently. The change of regime in 1985 seemed to
offer that possibility of rectification. Under those circumstances, Sen-
dero’s move toward Lima was an opportunistic quest to cover its re-
gional weaknesses rather than a decision based on its military prow-
ess (Hidalgo Morey 2004, 124). This explains why, in mid-1986, the
Shining Path announced that the “people’s war” was now “taking the
countryside and the city as a single whole, with the countryside the
main theatre of armed action and the city a complementary but neces-
sary one (CC-PCP 1986a).”
Given its historical background and organizational resources, the
fi rst regime of the long-vetoed Aprista Party created great expecta-
tions about what it could do in the fight against subversion. Unlike his
predecessor, the young and energetic Alan García Peréz (1985–1990)
seemed ready to act in two critical directions: targeting subversion in
a more integral way and enforcing respect for human rights. The fi rst
involved bringing in the state to act upon the socioeconomic factors
that fueled rural unrest by initiating economic development programs
like Plan Sierra and Trapecio Andino, and through new representa-
tive spaces like the Rimanakuy (regional forums in which the presi-
dent and his cabinet met community representatives to discuss matters
of development as well as security). The second was to act decisively to
contain state violations of human rights. On that front, García sent a
promising message just two weeks after his inauguration, in response
to the “massacre of Accomarca,” another of the confl ict’s representa-
tive episodes. On August, 14, 1985, in Accomarca, a small village of
Vilcashuamán, Ayacucho, at least sixty-two people, including twenty-
six children, were executed by two army patrols led by Sub-lieutenant
Telmo Hurtado. The massacre was part of a revenge mission in re-
sponse to a Shining Path ambush suspected to have been performed
with the complicity of the community. Soon after the news broke, and
with some pressure from the Peruvian Congress, President García dis-
missed the commander of the Ayacucho Emergency Zone, Wilfredo
Mori, and the rest of the command of the armed forces (CVR 2003a,
7:102–112).
Less than a year into García’s government, any positive intentions
in this area were cut short by the impact of another infamous mas-

Soifer_6844-final.indb 39 8/17/18 11:53 AM

 40 José Luis Rénique and Adrián Lerner

sacre. Prisoners accused of terrorism were killed as police and mili-

tary agents repressed a coordinated protest launched in June 18–19,
1986, in several jails in Lima. Any semblance of a civilized implemen-
tation of the counterinsurgent war sank in the blood of the victims of
summary executions and bombings in prisons like El Frontón, Canto
Grande, and Lurigancho. With more than 250 fatalities, the episode
drew significant unwanted international attention, as it coincided with
a meeting of the Socialist International hosted by García in Lima. The
“massacre of the prisons” gave the Shining Path another opportunity
to tout its discipline and revolutionary mystique. A few months earlier,
images transmitted by the international press had emphasized the con-
trast between the sordid world of Peruvian jails and the frantic order
prevailing in the “luminous trenches of combat” of the pavilions con-
trolled by the Maoists. After June 1986, senderista prisoners were en-
shrined as the epitome of “heroism” and “sacrificial spirit” (CC-PCP
1986b; Rénique 2003; Infante 2007).
Guzmán claimed that far from discouraging the senderistas, the
“cold-blooded barbarism” displayed against them in the prisons ac-
tually fed their combativeness. As suggested by the following quota-
tion, senderista texts of the late 1980s could be read, paraphrasing
the theoretician of jihad Abu Bakr Naji, as a manual for the “manage-
ment of savagery” (Naji 2006): “Their criminal barbarity has fi lled the
masses. Their new heights of infamy have been registered forever in
the memories of countless masses who will mete out crushing punish-
ment to those who are politically and militarily responsible, no mat-
ter how long it may take” (CC-PCP 1986a). Writing from his hideout
in Lima after a decade in hiding, and physically detached from most
of the bloody realities of war, by 1988 Guzmán seemed to believe that
the “people’s war” was moving toward its next critical stage: “stra-
tegic equilibrium.” This was, in theory, a transition between “strate-
gic defense” and “strategic offense,” which would be possible through
the acquisition of “modern means of war.” In the meantime, “humble
dynamite” would remain the “weapon of the people.” How could a
war fought with “the simplest weapons that everyone from among the
masses can wield” turn into a full-scale war for national power? Al-
though Guzmán thought his party had a “mass character,” he never
purported to turn it into a “mass party” like those embracing “rot-
ten revisionist positions.” Rather, it was “an instrument of war like
the one Lenin himself would demand”—the kind of party that led the
October Revolution with 80,000 members in a country of 150 million

Soifer_6844-final.indb 40 8/17/18 11:53 AM

 The Last Peasant War in the Andes 41

inhabitants. Through the organisms of “the new power,” they would

be able to bring to the center of the system the energy accumulated in
the peripheries. This would provoke a crisis of such magnitude that
it would bring about foreign intervention, a situation in which “the
United States can mobilize our neighboring countries” or “they could
intervene directly, with their own troops.” Then, “the oppressed na-
tion-imperialist contradiction would become the main one, and that
would give us an even broader basis on which to unite our people”
(Guzmán Reynoso 1988).
This scenario was clearly drawn from the experience of China af-
ter the Japanese invasion of 1937. Was it an even remotely realistic sce-
nario in fin-de-siècle Latin America? According to this idea, the pos-
sibility of transforming the “popular war” into a “patriotic war” led
by the Shining Path depended on its capacity to push the country into
further chaos. This alone would have been no small feat, but the rest
of Guzmán’s plan was based on an even shakier premise: a worldwide
revolutionary offensive. By then, however, “capitalist roaders” had
taken over even Mao’s fatherland, and both the Berlin Wall and the
Soviet Union were disintegrating. The world was marching in the op-
posite direction to that imagined by “President Gonzalo.” The bur-
den of fulfi lling this pipe dream fell squarely on the shoulders of his
ragtag “People’s Guerrilla Army” and the increasingly younger and
inexperienced “masses,” who periodically convened to destroy and
kill throughout the country. Guzmán believed that by the end of the
1990s, his party’s historical mission would be to usher in a new mil-
lennium in which communism “will be defi nitively stamped on his-
tory and humanity will take a leap from the realm of necessity into
the realm of liberty.” Unsurprisingly, the new historical chapter began
with the Shining Path’s providential decision to “climb the slopes of
another mountain in order to scale more brilliant summits.” Absolute
victory was the only acceptable outcome, and all compromise was the
“black flag of the reaction” (Guzmán Reynoso 1988). Achieving their
ends would justify spilling copious amounts of blood, as the suicidal
resistance in the prisons, ordered by Guzmán, showed. But not, of
course, the blood of the leader, who guaranteed the party’s continuity.
The twentieth-century era of revolutionary militancy was withering
away everywhere, as the world entered neoliberal times. Somehow, at-
tempts to relaunch it continued in Peru. In 1984 the Movimiento Revo-
lucionario Túpac Amaru (MRTA), a new armed organization, entered
the fray as a renewed expression of the classic guerrilla insurgents of

Soifer_6844-final.indb 41 8/17/18 11:53 AM

 42 José Luis Rénique and Adrián Lerner

the 1960s. A detailed analysis of the MRTA is beyond the scope of this
chapter, but suffice it to say that at numerous points they added to the
atmosphere of violence and confusion in 1980s Peru. They usually fol-
lowed the laws of war more closely than the Shining Path, but they
also kidnapped and murdered civilians. In some cases, they fought the
senderistas over resources and strategic territory. They also made the
waters of the political arena murkier. For example, after reaching an
agreement with President García for a year of truce, they put him un-
der the spotlight again when, days before the end of his presidential
term, more than forty inmates associated with the MRTA (including
their leader and the member of a prominent family linked to APRA,
Víctor Polay Campos) escaped prison. They also complicated Alberto
Fujimori’s triumphalist narrative between December 1997 and April
1998 when they held dozens of high-profile hostages for four months
in the residence of the Japanese ambassador to Peru. Emerretistas, as
militants of the MRTA were known, managed to be a thorn in the side
of everyone for more than a decade (Meza Bazán 2012).
With his proposal to blend socialism and Andean traditions within
the framework of “mariateguismo,” the historian Alberto Flores
Galindo reignited discussions about the political traction of the radi-
cal tradition. “Millenarism and messianism loom in Peru because pol-
itics here is not just a secular activity,” Flores Galindo remarked in
1986. By then, the Partido Unificado Mariateguista, the largest party
of the legal Left, was wracked by an internal debate triggered by a
proposal to create an armed self-defense organization (Rénique 2004,
245–250). Other less politicized Peruvians looked for hope among
the profusion of evangelical groups (some of which featured an un-
deniably nativist slant) that seemed to crop up throughout the mar-
gins of “official Peru” (Curatola 1997; De la Torre 2005; Ossio Acuña
2015). As Stern observed, in the 1980s and the early 1990s “the sense
of enigma and surprise was not really limited to Sendero Luminoso
alone [. . .] Peru as a whole seemed to lurch from surprise to surprise”
(Stern 1998a, 3).
Economic ruin was widespread from 1988 onward. After García’s
demand-driven economic policies succumbed to a cycle of combined
hyperinflation and recession, as well as to intense liberal opposition,
his response, a structural adjustment, unleashed a social crisis of un-
known proportions (Parodi Trece 2002). The discovery of a state-led
paramilitary extermination group with close ties to APRA’s higher-
ups contributed to the feeling of instability and further eroded the gov-

Soifer_6844-final.indb 42 8/17/18 11:53 AM

 The Last Peasant War in the Andes 43

ernment’s legitimacy (Comisión investigadora de grupos paramilitares

1989). The widely shared notion that the country was confronting the
worst crisis in its republican history fueled both despair and radical-
ization. Was Lima ripe for a fi nal assault? Were Sendero Luminoso’s
regional bases strong enough to sustain this leap forward, or was this
ultimately an escape forward? Apparently, although they had extended
their clandestine network throughout most of the country, their ex-
pansion was not accompanied by a net growth of their forces (Hidalgo
Morey 2004, 144). To fill this growing gap they resorted to two strat-
egies, one rural and one urban, that led to a spiral of death matched
only by the period from 1982 to 1984. More than two thousand fatal-
ities occurred each year between 1987 and 1992, with victims now in
almost every department of the country. In the cities, particularly in
Huamanga, Lima, and Huancayo, the Shining Path made growing use
of urban terror by means of car bombs and the so-called “commands
of annihilation.” They also added to the anxiety of the urban popu-
lations with attacks against the power grid that generated long black-
outs, “armed strikes,” and infiltration of local grassroots organiza-
tions, particularly in shantytowns and other popular neighborhoods,
and of public universities (CVR 2003a, 1:55).
In the countryside, Sendero intensified demands upon its rural bases
by using ever more coercive methods of recruitment. Sendero’s actions,
in combination with the extended presence of the armed forces, took
yet another turn for the worse. The intricate configurations of those
years are only starting to be disentangled by researchers. The anthro-
pologist Kimberly Theidon recently described the complexity of the lo-
cal rural milieus:

Without denying the pressures exerted by the senderista cadre as well

as the armed forces, the idea of being “caught between two fi res” does
not help us understand the brutal violence that involved entire pueblos
or the fact that there was a “third fi re,” comprised of peasants them-
selves. In the words of many villagers, “we learned to kill our broth-
ers.” (Theidon 2012, 5)

Amid these dramatic developments, the state and its coercive forces
had begun adjusting their strategies. The Counter-Terrorism Direc-
tion (DIRCOTE), a specialized investigative police unit, was created
in 1983. Despite working for long periods in extremely precarious and
dangerous conditions, DIRCOTE achieved striking results that pro-

Soifer_6844-final.indb 43 8/17/18 11:53 AM

 44 José Luis Rénique and Adrián Lerner

vided better intelligence on the Shining Path and led to important cap-
tures. State forces quickly realized that the hits that would hurt the
Maoists the most were not military, but political. They targeted the
leaders, often with a majority of the funding coming from foreign
sources (including from the US Central Intelligence Agency and the
Israeli Mossad). DIRCOTE eventually managed to work in relative
independence from the political pressures of the time. Since the cap-
tured Shining Path members rarely exposed their comrades, partially
because the party’s compartmentalized structure limited such knowl-
edge, the police unit focused heavily on analyzing documents found in
raids. DIRCOTE eventually underwent several changes of names and
personnel, but kept a core of capable investigators. Its resources were
limited, despite increased support during the years of the APRA gov-
ernment, so by 1987 they decided to focus solely on the Shining Path’s
political and logistical apparatus in Lima. This decision proved fateful
five years later, when one of their raids led to the capture of Abimael
Guzmán in the capital. Given Sendero Luminoso’s vertical organiza-
tional structure, his capture proved to be a crucial moment in the his-
tory of the war (Jiménez 2000; CVR 2003a, 2:110–150).
The armed forces also adapted. In 1989 they changed their geo-
graphic orientation from traditional military regions to “Anti-
Subversive Fronts” based on the Shining Path’s spatial expansion
(CVR 2003a, 2:198). More importantly, a trend emerged that would
help define the course of the confl ict and of Peruvian politics in the de-
cade to come. Influenced by the Taiwanese School of War, the armed
forces adopted a more integral and political conception of war. This
led to broadly approaching the political realm as part of the war ef-
fort, a strategy that Fujimori and his spymaster, Vladimiro Montesi-
nos, would later exploit to rule the country (CVR 2003a, 2:225–234).
But in the countryside, the integration of the conflict’s social and po-
litical aspects with its military facets led to the formation of an armed
alliance between the military and the peasants: “for the fi rst time in
nine years the Armed Forces were securing peasant confidence in the
most disputed areas of the conflict” (Degregori and Rivera Paz 1993,
13). This approach led to the formation of thousands of self-defense
committees armed by the Peruvian state, a critical force in under-
mining the senderista rearguard. By the early 1990s there were some
3,445 rondas campesinas, which would ultimately defeat Sendero at
its “peasant war” (Hidalgo Morey 2004, 158).
Still, during those years, tragic setbacks slowed the process. One

Soifer_6844-final.indb 44 8/17/18 11:53 AM

 The Last Peasant War in the Andes 45

such event was the massacre of Cayara, Ayacucho, in May 1988, dur-
ing which thirty-five adults and children were killed by two military
patrols that accused the cayarinos of collaborating with Shining Path.
The massive cover-up that ensued exhausted what little credibility the
regime still had (CVR 2003a, 7:203–214). Moreover, the armed forces
approved a new Manual of Military Operations in the Emergency
Zones while Peru was still under APRA’s rule. The manual explicitly
established a policy of eliminating non-armed members of the Shin-
ing Path—the members of its “people’s committees.” Its application,
said Ollanta Humala Tasso, then an army captain, encouraged human
rights violations and stirred crises of conscience among young officers
like himself (Humala Tasso 2009, 40–42).
Meanwhile, the Shining Path’s goal of bringing the insurgency to
the country’s capital raised daunting challenges for the senderistas.
From infiltrating cultural groups through organizations such as the
Movement of Popular Artists to preparing “armed strikes” that resem-
bled rehearsals for a fi nal assault on Lima, the party was working to
the limits of its resources. Mistakes, therefore, were costlier than ever
before. Two in particular epitomized the new turn. The fi rst was the
assassination of María Elena Moyano on February 15, 1992. Moyano
was a popular leftist and feminist Afro-Peruvian grassroots leader
from the Villa El Salvador neighborhood; her body was blown into
pieces in front of her family after she had been shot for opposing Sen-
dero’s infiltration of local politics. In the second, on July 16 of that
year, a car bomb placed in the Tarata street of the traditional upper-
middle- class district of Miraflores caused twenty-five fatalities and
wounded as many as two hundred. The viciousness of these two events
captured the imagination of limeños of all classes: not coincidentally,
popular movies have been produced about both. Eventually, Guzmán
and his second in command (and living partner), Elena Iparraguirre,
apologized for them. They argued that the car bomb was supposed to
explode in front of a nearby bank at off-hours without causing civil-
ian casualties, and that the use of dynamite against Moyano had been
an excess committed by those in charge of an otherwise justifiable ex-
ecution of someone alleged to be collaborating with the security forces
(Burt 2011).
At the national level, the battle for the Upper Huallaga Valley ap-
peared by the late 1980s to be the key for propelling the “people’s
war” to a “strategic offensive” given the growth of the illegal coca
market. Control over part of its income would provide the Shining

Soifer_6844-final.indb 45 8/17/18 11:53 AM

 46 José Luis Rénique and Adrián Lerner

Path with the means to acquire the modern weaponry required to take
the offensive. The region was also a key source of manpower for the
assault on Lima. After a complex, multisided struggle that involved
MRTA, the US Drug Enforcement Agency, the Peruvian armed forces,
international drug cartels, and coca planters, the Shining Path man-
aged to gain a degree of control over the region by 1989, establish-
ing itself as a mediator between producers and the Colombian cartels
in a chapter of the Latin American Cold War that seemed to forecast
the subsequent era of the “wars on terror”: an apocalyptic orgy of vio-
lence fueled by the insurgents’ capacity to sustain themselves through
autonomous revenue streams linked to drug trafficking (Dreyfus 1999;
Kay 1999; Gonzáles 1992; Dawson 2011, 226). But Sendero’s role was
challenged by a growing counterintelligence force that had no moral
qualms about making deals with the cartels—the armed forces’ more
“integral” approach to counterinsurgency could show remarkable eth-
ical flexibility, particularly when there was money to be had under the
table. Vladimiro Montesinos, an obscure lawyer with ties to both the
cartels and the CIA as well as a former Peruvian army captain who
had been expelled on espionage charges, began a meteoric ascent that
would turn him into the country’s foremost behind-the-scenes politi-
cal operator by the early 1990s (CVR 2003a, 2:201).
In retrospect, it is understandable that not only the morality but
also the effectiveness of Guzmán’s strategies has been questioned from
all sides. Colonel Hidalgo Morey has pointed to the theoretical incon-
sistency of Guzmán’s “strategic equilibrium,” questioning how an or-
ganization that increasingly depended on forced recruitment could
trust its recruits to operate modern weaponry (Hidalgo Morey 2004,
161–162). Even one of Guzmán’s comrades, Oscar Ramírez Durand,
also known as “Comrade Feliciano,” questioned Guzmán’s integrity.
Ramírez, who would lead a Sendero column that remained at war long
after Guzmán had abandoned the group, doubted Guzmán’s reasons
for perpetually postponing a move to the countryside. Guzmán’s ab-
sence created an insurmountable divide between his clique and those
who carried the burden of the war on the ground. His lack of field ex-
perience—“he never faced a siege, he has never been in an operation,”
“Feliciano” accused—made any serious debate about military issues
with him impossible (CVR 2002a). Unlike the accounts of Hidalgo
Morey and Oscar Ramírez Durand, a report prepared by Gordon Mc-
Cormick for the US Secretary of Defense was written before the Shin-
ing Paths’ collapse. McCormick argued that he could not imagine the

Soifer_6844-final.indb 46 8/17/18 11:53 AM

 The Last Peasant War in the Andes 47

Maoists “taking Lima,” but he believed they could generate “politi-

cal disintegration”; they were able and willing to “make a bad situa-
tion worse.” Whether they would cause such a collapse hinged on their
strength in the countryside, their ability to isolate Lima from the rest
of Peru, and, ultimately, whether the Peruvian government would fi-
nally understand the way Sendero worked (McCormick 1990, 77).

Authoritarian Epilogue:
Fujimorismo and the Narrative of Success

In a 1992 US congressional hearing about the situation in Peru, Ber-

nard W. Aronson, the US assistant secretary of state for Inter-Amer-
ican Affairs, minced no words: “Make no mistake,” he said, if the
Shining Path were to take power, “the result could be genocide,” as its
goal was “to remake Peruvian society just as the Khmer Rouge set out
to reduce Cambodia to the Year Zero.” Reinforcing his somber warn-
ing, Aronson then quoted Abimael Guzmán himself: “The revolution
will triumph after the Peruvian people ‘cross over the river of blood’”
(Crossette 1992). Six months later, Guzmán was in custody. In the
fi rst speech given after his capture, President Alberto Fujimori referred
to him as “Abimael the exterminator,” a “monster” who with inhu-
mane coldness had ordered the assassination of María Elena Moyano,
the slaughter of Tarata, and countless other incursions into neighbor-
hoods and villages in which men and women had been executed in the
most horrific ways (Burt 2011). Guzmán’s capture, on September 12,
1992, was a turning point in the making of a powerful narrative of
what Carlos Iván Degregori labeled “the fi rst media-based dictator-
ship in Latin America” (Degregori 2000a). The coup on April 5 of that
year, when Fujimori ordered the military to close the Congress and im-
prison key critics, was legitimized in the eyes of many by the capture
of Guzmán. The regime then acquired its fi nal shape during the next
months through two decisive events: the failed “constitutional coup”
led by General Jaime Salinas Sedó, and the elections of a Democratic
Constituent Congress on November 13 and 22.
The years from Fujimori’s surprising electoral victory over Mario
Vargas Llosa in June 1990 to the legalization of the “Fujicoup” in May
1992 were, as the title of a propagandistic video suggested, “Three
Years That Changed Peruvian History” (Fujimori 1993). The master
fujimorista narrative was displayed in that piece. It presented Presi-

Soifer_6844-final.indb 47 8/17/18 11:53 AM

 48 José Luis Rénique and Adrián Lerner

dent Fujimori as a man who saved the nation from an evil alliance of
terrorists and drug traffickers attempting to transform Peru into “the
world’s fi rst república cocalera”; “an empire of vice and crime” that
would have destroyed millions of lives. This nefarious objective relied
on the passive complicity of the traditional parties—the partidocra-
cia that held power during most of Peru’s recent democratic history.
Suddenly, through Fujimori, the enigmatic drama of the last few years
had found a plausible explanation. Importantly, now words were sup-
ported by deeds. Fujimori was quick to develop an effective presiden-
tial style. Freed from hyperinflation, and with Guzmán in jail and the
terrorists on the run, Peru’s future looked promising for fi rst time in
a long while. According to polls, 70 percent of the population agreed
with the Fujicoup (Fujimori 1993). The path seemed clear for realiza-
tion of the fujimorista promise: a democracy based not on speeches
but on good administration, “conducted by selfless technocrats instead
of self-serving politicians” (Conaghan 2005, 3). The video also fea-
tured an ad hoc, self-serving historical plot. From the military caudi-
llos of the past to the civilians and their crony political parties of more
recent times, a depressing continuity of selfish and incompetent rulers
was described. It was a history of frustration in which even innovators
like Haya de la Torre ultimately betrayed their ideals. This accumula-
tion of wrongdoings was crowned by the “worst crisis in republican
history,” providentially solved through the rise of Alberto Fujimori, a
man with a different vision and values rooted in his immigrant condi-
tion who was courageous enough to take the reins of a country in ru-
ins (Fujimori 1993; see also Jochamowitz 1993).
To his popular touch, Fujimori quickly added the critical support
of the military command and the economic elites, as well as that of
the international fi nancial community. By late 1992 a consensus had
emerged: a legally and morally unrestrained government was needed
to end the calamities of the 1980s (Vergara 2013, 176). It was not, of
course, a novel proposition: from Odría to Velasco, the call for iron-
fisted rule evoked other moments of Peru’s troubled recent political
history. Fujimorism emerged as a suitable vehicle to bring the coun-
try back from the brink of collapse, because it was capable of creating
consent around neoliberal reform, provided political support to run
an anti-insurgent campaign unencumbered by “excessive” legal con-
straints, and presented itself as a voice of the people in spite of the un-
popular nature of his economic policies.
The “people’s war” came to its end as a result of defeats on two

Soifer_6844-final.indb 48 8/17/18 11:53 AM

 The Last Peasant War in the Andes 49

fronts. The fi rst was effective investigative police work. Anti-terrorist

intelligence forces captured Sendero Luminoso’s leaders without fi r-
ing a single shot, and the organization fell apart as soon as Guzmán
was imprisoned. The second breakthrough resulted from the mass mo-
bilization of the so-called self-defense committees, or anti-subversive
peasant rondas. The peasant war was turned on its head: those ex-
pected to be the central protagonists of the “people’s war” had re-
belled against it. For many of the ronderos, siding with the armed
forces meant emancipation from senderista abuses. Rondero identity,
under these circumstances, became defiant: a rondero’s aggressiveness
could be directed toward a variety of purposes. This ambiguity was re-
flected in the Peruvian Truth Commision’s characterization of ronde-
ros: “In no other actor of the war is the line between perpetrator and
victim, between hero and villain, as thin and porous as in the case of
the CAD [comités de autodefensa]” (CVR 2003a, 2:288; Starn 1993).
The rondas had been several years in the making, but only in late
1991 were they fully embraced by the state. They were then acknowl-
edged as “free and spontaneous” civil society associations, although
many of them had been created by the armed forces. Their numbers
doubled between 1993 and 2000, by which time there were around
eight thousand legally recognized CADs, with some half a million
members. While opponents saw in them the “militarization” of soci-
ety, Fujimori believed they still had a role to play in the broader scheme
of national security (CVR 2003a, 2:300). A new rural armed player,
under governmental control, had been born in the territory in which
autonomous and rebellious peasant organizations had prevailed dur-
ing the 1960s and 1970s. The CADs offered direct local connections
for a regime that loved to emphasize Fujimori’s hyperactive agenda.
Disregarding regional bureaucratic structures, Fujimori toured the An-
dean areas, usually accompanied by the minister of the presidency,
who directed an agency that played a fundamental role in expedit-
ing the process of funding small infrastructural projects and coopt-
ing leaders at the local level. The military acted as Fujimori’s political
platform, organizing his tours, providing security and transportation,
and even organizing “spontaneous” demonstrations of public support.
In this sense, parallels between fujimorismo and velasquismo point to
continuities that deserve deeper analysis (Vergara 2014).
The “people’s war” had, paradoxically, afforded Fujimori the op-
portunity to continue the peculiar kind of counterinsurgent work de-
veloped by Velasco. It was an opportunity to handle the old problem

Soifer_6844-final.indb 49 8/17/18 11:53 AM

 50 José Luis Rénique and Adrián Lerner

of integration, the so-called “Indian problem” of the early twentieth-

century, which had historically frustrated the making of the Peruvian
nation. Velasco conceived a corporatist solution in which, under the
inspiration of Túpac Amaru, an organized peasantry would take its
place in a militarized republic. In contrast, Fujimori offered a style
of doing things, without the nuisance of professional politicians and
palabrería (excessive talk). Instead, his was a pragmatic tone, where
gente sencilla (simple people) like himself would stand on hierarchies
to build a nation tailor-made to meet its population’s greatest needs. It
was a “sui generis democracy” that would open channels for Peruvi-
ans to become the protagonists of modernization (Conaghan 2005, 4,
7). The state would thus meet the “popular overflow” (desborde popu-
lar) that had transformed Peru into a vibrant but chaotic urbanizing
nation since the mid-twentieth century. In the “capitalist revolution”
to come, millions of former peasants would, somehow, fi nally fi nd
their place in the “official Peru” from which they had until then been
excluded (Collier 1976; Degregori, Blondet, and Lynch 1986; Golte
and Adams 1987; de Althaus 2008; Rochabrún 2007).
Few could have predicted that by 1993 the man who had described
dialogue with other political forces as “cheap demagoguery” would
recognize the achievements of the Fujicoup. Fujimorismo had coopted
even Abimael Guzmán (Rénique 2003, 106–108; Guzmán Reynoso
1988). Fujimori’s “intelligence adviser,” Vladimiro Montesinos, used
the opportunity to articulate a reelection campaign and, further
ahead, an authoritarian, murderous regime that has been described as
“the most corrupt in the history of Peru” (Quiroz 2008, 439).

Note

1. For a more detailed discussion of this issue, see chapter 8 in this volume.

Soifer_6844-final.indb 50 8/17/18 11:53 AM

 CHAP T ER 2

Civil Wars and Their Consequences:

The Peruvian Armed Conflict in
Comparative Perspective
Livia Isabella Schubiger and David Sulmont

Throughout the 1980s and 1990s, Peru was ravaged by the longest
and deadliest armed conflict in the Republic’s history (CVR 2003a). It
has been estimated that between 1980 and 2000, about seventy thou-
sand people lost their lives in the conflict (Ball et al. 2003). Most of
the victims have been attributed to the insurgent group Shining Path
(Sendero Luminoso) and the Peruvian government, although other
armed actors were active as well, in particular rural self-defense
forces (known as the comités de autodefensa or rondas campesinas)
and the Movimiento Revolucionario Túpac Amaru (MRTA), a revo-
lutionary armed group much smaller and shorter-lived than the Shin-
ing Path.1 While some small remnants and splinters of Sendero Lu-
minoso are still active in remote pockets of Peru’s periphery as of
2017, the Shining Path has been considered defeated in most regions
of the country since the mid to late 1990s. Nevertheless, the legacies
of political violence still impact Peruvian politics and society to sig-
nificant degrees. This chapter puts the conflict and its repercussions
into theoretical and comparative perspective. We largely focus on
the armed conflict between Sendero Luminoso and the Peruvian gov-
ernment, and less so on the MRTA or the peasant rondas, the lat-
ter of which were central to the eventual weakening of the Shining
Path. We argue that recent scholarship on political violence helps
to illuminate a number of seemingly idiosyncratic aspects of the Pe-
ruvian case regarding both the dynamics of the conflict and post-
conflict politics. We conclude by highlighting several puzzling fea-
tures of the Peruvian confl ict that may yield fruitful avenues for fu-
ture research.

Soifer_6844-final.indb 51 8/17/18 11:53 AM

 52 Livia Isabella Schubiger and David Sulmont

Armed Conflict in Peru

The Peruvian Internal Armed Conflict, as it is called in the Peruvian

Truth and Reconciliation Commission (Comisión de la Verdad y Re-
conciliación del Perú, or CVR), sprang from a very particular Peru-
vian historical context. When Sendero Luminoso decided to initiate its
armed struggle on May 17, 1980, Peru was on the eve of its first dem-
ocratic elections after twelve years of military rule. The Revolutionary
Government of the Armed Forces, which had deposed the democratic
government of Fernando Belaunde in October 1968, undertook some
of the most radical and progressive social reforms in Peruvian history
(McClintock and Lowenthal 1985). The economic and social reforms
of the military government, particularly the agrarian reform, put an
end to the rule of traditional landowners over millions of indigenous
peasants in the Andes as well as the political and economic hegemony
of the oligarchic class of hacendados and landowners over Peruvian
society (Pease 1986).
By the end of the 1970s, a democratic transition took place in the
midst of an economic crisis and popular mobilizations against the mil-
itary government. A Constitutional Assembly was elected in 1978, a
new constitution was approved in 1979, and general elections took
place in 1980 when Belaunde won the presidency for a second time.
The new constitution included many progressive civil, political, and
social rights and universalized political citizenship by allowing illit-
erate adults over the age of eighteen to vote. Under the previous con-
stitution, voting had been restricted to literate people over twenty-one
years of age, excluding millions of citizens—mainly from rural areas
and with indigenous backgrounds.
Contrary to other military dictatorships in Latin America (such as
in Chile from 1973 to 1990 and Argentina from 1976 to 1983), which
targeted left-wing parties and militants and caused massive human
rights violations, left-wing organizations and leaders had suffered rela-
tively low levels of repression under the Peruvian military government.
They were able to build political organizations that played an impor-
tant role in the social and political mobilizations during the second
half of the 1970s and the electoral processes of the 1980s. Indeed, left-
wing political parties (including radicals and Communists) success-
fully participated in the new elections. The Izquierda Unida electoral
alliance received the second-most votes in municipal elections in 1980,

Soifer_6844-final.indb 52 8/17/18 11:53 AM

 Civil Wars and Their Consequences 53

in which people from more than a thousand constituencies in rural

and remote areas were able to elect their local authorities.
Although stark social and economic inequalities remained, in 1980
Peru was a society undergoing major social, political, and democratic
transformations. A large portion of its rural and indigenous popula-
tion was claiming fuller citizenship rights; vast amounts of land were
redistributed among peasant and indigenous peoples; and radical left-
wing parties and social organizations were gaining access to the for-
mal democratic process.
In that context, Sendero Luminoso, a small radical Maoist political
party, decided to initiate its armed struggle for a Communist revolu-
tion in Peru. Sendero’s influence at the time was mainly limited to some
provinces of the department of Ayacucho, one of the least-developed
regions of the country. 2 Sendero’s fi rst “military” action took place in
Chuschi, a small town in Ayacucho’s countryside, where they burnt
the electoral materials the day before the general elections of May 18,
1980. As Carlos Ivan Degregori points out (Degregori 2012a, 38–40),
the insurgency of Sendero Luminoso came as a “triple surprise”: it was
not anticipated by the government and its intelligence services, the po-
litical parties and social organizations, or the social science and aca-
demic community. None of them foresaw that this small and highly
radicalized group would be able to unleash such a wave of violence in
the coming years—at the very moment when the democratic transition
was opening new channels of participation and representation for di-
verse sectors and social groups in Peruvian society.
In sharp contrast to other cases of civil wars in Latin America (e.g.,
El Salvador, Guatemala, or Nicaragua), during the most violent years
of the conflict, democratic institutions (a free press, parliamentary pol-
itics, some judiciary independence, and elections) continued to func-
tion (with many limitations) in Peru between 1980 and 1992.
As chapter 1 discusses in detail, the Internal Armed Conflict
spanned almost two decades. According to the statistical estimates of
the CVR, the conflict’s seventy thousand deaths were attributed to Sen-
dero Luminoso (46 percent); state agents and countersubversive forces,
including autonomous or state-organized self-defense peasant forces
(30 percent); and other agents or circumstances, such as the MRTA,
nonidentified agents, and combat casualties (24 percent) (CVR 2003a,
Annex 3, 34).3
The conflict is generally considered to have begun on May 17, 1980,

Soifer_6844-final.indb 53 8/17/18 11:53 AM

 54 Livia Isabella Schubiger and David Sulmont

Figure 2.1. Peru 1980–2000: fatal casualties reported to the Truth and
Reconciliation Commission by year and perpetrator. Data source:
Comisión de la Verdad y Reconciliación (CVR 2003a).

the day of Sendero’s fi rst armed attack against the state. The end of
the conflict is more difficult to determine. By the close of the 1990s,
Sendero Luminoso was essentially defeated, both militarily and polit-
ically. However, small and sporadic actions carried out by remnants
and splinters of subversive organizations were still being reported in
remote regions of the country at the time of this writing in 2017. The
mandate of the CVR was to investigate events and human rights vio-
lations that occurred between 1980 and 2000, but as can be seen in
figure 2.1, most of the reported victims died between 1982 and 1993
(94 percent), with two clearly identified peaks in 1984 and in 1989–
1990. Figure 2.1 shows different patterns of violence across time and
among actors. A fi rst period of intense violence by both Sendero Lu-
minoso and the state’s agents occurred in 1983 and 1984, and was fo-
cused mainly in Ayacucho. In this time period, Ayacucho’s country-

Soifer_6844-final.indb 54 8/17/18 11:53 AM

 Civil Wars and Their Consequences 55

side witnessed a bloodbath within its peasant population: 18 percent

of the fatal victims reported to the Truth and Reconciliation Commis-
sion correspond to events that took place in 1984, mostly in that re-
gion. During the second peak of violence from 1989 to 1992, Sendero
became by far the main perpetrator of fatal attacks. In this second pe-
riod, the state’s counterinsurgency strategy became less indiscriminate
and relatively less lethal.
The fi nal phase of the confl ict started with the capture of Abimael
Guzmán in September 1992. At about the same time, a “repentance”
law granted immunity to militants of subversive organizations who
surrendered themselves to the authorities and provided information
that could lead to the incrimination of other members of their orga-
nization. In October 1993, television broadcasts showed Guzmán and
most of the higher leaders of Sendero Luminoso who had been cap-
tured by the police; from their prison, they recognized their defeat and
asked for negotiations to “end the war.” After that, violent events in
the country declined sharply.
The conflict’s massive toll of victims destroyed thousands of fami-
lies and hundreds of communities. Despite some high-profile condem-
nations of prominent figures such as Alberto Fujimori on human rights
crimes, most of the crimes committed by state agents remain unpun-
ished. Thousands of victims and their families still wait for justice and
reparation.

Civil Wars and the Peruvian Conflict

During the past decade, scholarship on political violence has moved

away from an understanding of civil wars as a homogeneous category
(Fearon and Laitin 2003; Collier and Hoeffler 2004). Scholars have
instead started to appreciate and theorize the various layers of conse-
quential diversity across and within civil wars, recognizing variables
such as armed actors’ identities and motivations, as well as patterns of
mobilization and violence in armed confl ict (Cederman and Gleditsch
2009; Kalyvas 2003; Weinstein 2007). We focus here on those the-
ories and typologies that have emphasized the implications of war-
fare technology (Kalyvas and Balcells 2010), ideology (Gutiérrez Sanín
and Wood 2014; Thaler 2012), and identity (Cederman, Wimmer, and
Min 2010; Wucherpfennig et al. 2012) for the internal dynamics and
development of civil wars.

Soifer_6844-final.indb 55 8/17/18 11:53 AM

 56 Livia Isabella Schubiger and David Sulmont

Warfare Technology

To begin with, the Peruvian confl ict represents an unambiguous case

of “irregular” (i.e., asymmetric) civil war, often treated synonymously
with “guerrilla war” and also encountered in cases such as El Salva-
dor (FMLN), Afghanistan (Taliban), or Nepal (CPN-M) (E. Wood
2008; Kalyvas and Balcells 2010; Balcells and Kalyvas 2014; Balcells
and Kalyvas 2015).4 Kalyvas and Balcells (2010) distinguish irregular
conflicts from those in which military power is distributed more sym-
metrically, either at high or low levels: in conventional civil wars, state
forces and rebel groups engage in direct confrontations with access to
heavy weaponry (e.g., the Bosnian and Croatian wars in the 1990s),
while in symmetric nonconventional civil wars, both state actors and
rebel groups are militarily weak (e.g., Congo-Brazzaville in the 1990s).
According to Kalyvas and Balcells (2010), 54 percent of all major civil
wars5 between 1944 and 2004 were fought as irregular civil wars, a
proportion that is higher (66 percent) for wars that started before the
end of the Cold War era (1944–1990).
Insurgent groups emerging under these conditions not only oper-
ate covertly to avoid direct confrontations with their militarily supe-
rior opponents, but also face particular challenges of coordination and
internal control if they are to thrive as cohesive organizations. This
makes internal institutions for screening and recruitment, discipline,
and ideological indoctrination particularly important (Gutiérrez Sanín
2012; E. Wood 2010; Hoover Green 2011; Schubiger 2013). More-
over, and likely to a greater extent than in other types of armed con-
flict, civilian collaboration and support is often critical for the survival
of insurgent groups in this type of war (E. Wood 2008; Balcells and
Kalyvas 2014).

Ideology

Many irregular wars have been fought by rebel groups that repre-
sent some flavor of Marxist orientation (the latter broadly defi ned).
Of those 147 armed conflicts studied by Kalyvas and Balcells, 38 were
fought by Marxist armed groups, of which 10 had secessionist aspi-
rations and 28 pursued nonsecessionist revolutionary goals. In other
words, 26 percent of these confl icts saw governments being confronted
by armed groups with a generally Marxist outlook, and 19 percent
by groups that were both Marxist and revolutionary, like the Shining

Soifer_6844-final.indb 56 8/17/18 11:53 AM

 Civil Wars and Their Consequences 57

Path, the FARC in Colombia, the CPI-M in India, or the CPN-M in

Nepal (Balcells and Kalyvas 2015, 7–8).
Balcells and Kalyvas (2015, 6) identify four common organizational
principles shared by Marxist insurgent groups, regardless of their ex-
act ideological orientation: (1) a strong political party, (2) a heavy em-
phasis on centralization and discipline, (3) mass indoctrination of
both combatants and civilians, and (4) well-developed institutions for
state-like governance. The Shining Path indeed adopted all those prin-
ciples. In particular, its own blend of Marxist-Leninist-Maoist ideol-
ogy played a central role. Sendero Luminoso is also a clear case of an
armed group endorsing an ideology that was both externally intru-
sive and implied a fi rm internal institutionalization within the orga-
nization’s ranks (Schubiger and Zelina 2017). The group’s leadership
placed high emphasis on the indoctrination of not only its cadres but
also the “masses” in areas it controlled.6 Consistent with the principles
of guerrilla war, the Shining Path rarely confronted state forces di-
rectly, instead focusing on indirect strategies such as targeted assassi-
nations of state representatives and suspected collaborators, massacres
against communities of alleged civilian defectors, and eventually also
high-profile bombing attacks in urban areas (CVR 2003a). In contrast
to many other Marxist movements during the Cold War, however,
Sendero Luminoso was remarkably independent from external sources
(Kalyvas and Balcells 2010; Balcells and Kalyvas 2015). Moreover,
and even more strikingly, Sendero Luminoso was much more violent
toward the civilian population than comparable leftist Latin American
armed movements of its time (CVR 2003a).

Identity

Despite its origins in the impoverished and predominantly indige-

nous regions of a country rife with horizontal inequalities—the lat-
ter a main driver of protracted ethnic civil wars (Cederman, Gleditsch,
and Buhaug 2013)—the Peruvian armed conflict does not qualify as
an “ethnic confl ict” in the canonical understanding of the term, ac-
cording to which both claim-making and mobilization map neatly
onto ethnic divides (Wucherpfennig et al. 2012; Vogt et al. 2015). In
Peru, the insurgents’ Marxist-Leninist-Maoist mobilization frame was
clearly ideological, revolutionary, and nonethnic. And while the sen-
deristas claimed to fight on behalf of the most marginalized segments
of the population, neither the government nor the insurgents recruited

Soifer_6844-final.indb 57 8/17/18 11:53 AM

 58 Livia Isabella Schubiger and David Sulmont

their fighters exclusively from one particular ethnic group. This mir-
rors broader patterns in Peruvian society: during the second half of
the twentieth century, ethnicity rarely figured as an explicit focal point
of social and political mobilization, despite—or because of—the fact
that it has long been a major source of horizontal inequality and dis-
crimination (e.g., Sulmont 2011; Sulmont and Callirgos 2014; Thorp,
Caumartin, and Gray-Molina 2006; Yashar 2005; Paredes 2010;
Thorp and Paredes 2010). Especially among the most marginalized
indigenous people of Peru, identity-based categories such as ethnicity
have often been perceived as a barrier to social mobility, rather than a
basis for collective action (Sulmont 2011, 25).
Viewed from this perspective, Sendero Luminoso’s mobilization
approach is similar to those leftist revolutionary movements in Latin
America that also clearly adopted a nonethnic discourse, such as the
Fuerzas Armadas Revolucionarias de Colombia (FARC) or the Frente
Farabundo Martí para la Liberación Nacional (FMLN) in El Salva-
dor. At the same time, as we will discuss later, the Peruvian case shares
some similarities with those confl icts in which armed movements have
likewise adopted a class-based ideology, but where ethnicity neverthe-
less remained highly salient throughout the war, as in the confl ict be-
tween the Ejército Guerrillero de los Pobres (EGP) and the government
of Guatemala, for example.7
During the armed conflict, while issues of ethnic identity did not
dominate claim-making and mobilization, ethnicity was an important
correlate of people’s vulnerability to violence (we return to this sub-
ject later in the subsection entitled “Violence”). Moreover, the rele-
vance and complexity of ethnicity for power relationships still visible
in Peruvian society at large was mirrored in the armed groups’ internal
hierarchy: Sendero’s leadership was dominated by light-skinned edu-
cated elites, while lower-ranking members often represented more in-
digenous and more marginalized segments of the society (CVR 2003a,
8:99; Starn 1995b, 551; Starn 1998, 229). Insurgent governance over
communities sometimes at least temporarily reversed such hierarchies,
as “insurgent-allied residents who ruled some towns briefly in the early
1980s were generally poorer and more indigenous than those they sup-
planted” (E. Wood 2008, 551). At the same time, indigenous commu-
nities were also among the fi rst to openly—and violently—resist the
Shining Path (e.g., Coronel 1996; del Pino 1998).
In summary, the Peruvian armed conflict seems a paradigmatic ex-
ample in the universe of armed insurgencies during the Cold War era

Soifer_6844-final.indb 58 8/17/18 11:53 AM

 Civil Wars and Their Consequences 59

in Latin America and elsewhere, classifying as a clear case of irreg-

ular, nonethnic civil war. Moreover, like many other rebel groups of
its time, the Shining Path was an insurgency of Marxist orientation
against a government alleged to represent a capitalist future and a co-
lonial past. At the same time, however, Peru is also a very complex
case that does not neatly map onto established typologies and theoret-
ical divides. Having discussed the Peruvian conflict with reference to
the dimensions of warfare, ethnicity, and ideology, we turn to patterns
of violence, governance, and mobilization next.

Violence, Governance, and Mobilization

As outlined previously, the armed insurgency came as a surprise for

many analysts in Peru. Why did many people join the Shining Path?
Why did the confl ict reach such unexpected levels of intensity, and
how can we explain the patterns of mobilization, violence, and gover-
nance the confl ict saw?

Insurgent Mobilization

Debates on the microfoundations of civil wars, and individual partic-

ipation in war in particular, have centered on the relative importance
of factors such as grievances rooted in horizontal inequalities (Ceder-
man, Gleditsch, and Buhaug 2013), selective material incentives (Col-
lier and Hoeffler 2004), and outright coercion (Humphreys and Wein-
stein 2008). On the reasons for joining the Shining Path rebellion of
the 1980s and 1990s, scholarship on Peru resonates with several ar-
guments advanced in this broader literature by highlighting economic
and political marginalization, the lack of social mobility for young
people (including many educated men and women), and political in-
doctrination, but also coercion in Sendero recruitment (e.g., CVR
2003a; Chávez de Paz 1989; del Pino 1998; Degregori 1998a; Mc-
Clintock 1989; E. Wood 2008). Only rarely do scholars point to the
role of selective economic incentives, although some argue that those
were relevant for certain subgroups of Sendero recruits as well (Wein-
stein 2007), particularly in regions where Sendero extracted taxes
from drug traffickers and protected coca-growing peasants.
One strategy to evaluate the validity of competing arguments on in-
surgent mobilization for the case of Peru is to directly examine who

Soifer_6844-final.indb 59 8/17/18 11:53 AM

 60 Livia Isabella Schubiger and David Sulmont

joined the Shining Path. Unsurprisingly, there are no probabilistic sam-

ple data on the social composition of Sendero Luminoso, as systematic
surveys have not been feasible in this context of clandestine mobiliza-
tion. Moreover, data on theoretically relevant variables are challeng-
ing to collect. For example, ethnic identities are difficult to measure
in Peru due to the fluidity and complexity of how ethnic identity is ex-
perienced and perceived (e.g., Sulmont 2011; Sulmont and Callirgos
2014). However, based on the testimonies of men and women detained
and accused of terrorism in various prisons within the country, the Pe-
ruvian Truth Commission built a database of 1,158 inmates accused of
terrorism. For 821 of these prisoners, it was possible to determine the
alleged affiliation, with the overwhelming majority allegedly belong-
ing to the Shining Path (85.3 percent) and a minority to the MRTA
(14.7 percent) (CVR 2003a, statistical annex, dataset dec_pen_est).8
Even if there are no means of evaluating the statistical representativity
of these data, it is the most comprehensive and systematic database on
insurgent militants in Peru up to this date.
The data set compiled by the Truth Commission shows that a mi-
nority of the 821 inmates had an indigenous mother tongue (29 per-
cent in the Shining Path, 14 percent in the MRTA) (CVR 2003a, statis-
tical annex, dataset dec_pen_est) although the percentage of Sendero’s
inmates with an indigenous fi rst language was higher than in the gen-
eral population (29 percent vs. 19 percent).9 Both the members of the
Shining Path and the MRTA were better educated than the average for
members of their generation and their departments of origin. More-
over, their members were young: Over 50 percent of the Sendero in-
mates were from 20 to 30 years old at the time of their arrest. More
than 75 percent of the arrested senderistas were 40 or younger at the
time of being interviewed by the CVR (CVR 2003a, statistical annex,
dataset dec_pen_est). Using court records to construct a different data
set of 183 prisoners convicted of terrorism in Lima during the period
from 1983 to 1986, Chávez de Paz (1989) comes to similar conclu-
sions. Not distinguishing between members of the Shining Path and
the MRTA, he fi nds that most convicts were young and from the most
economically marginalized provinces of the country, but that a signif-
icant proportion of them were highly educated. Specifically, 35.5 per-
cent of those examined had enjoyed some form of university education,
with or without having attained a degree. The proportion of highly ed-
ucated convicts was even higher among women (Chávez de Paz 1989).
Despite the obvious limitations concerning statistical representa-

Soifer_6844-final.indb 60 8/17/18 11:53 AM

 Civil Wars and Their Consequences 61

tiveness, the available data support the notion that many of Sendero’s
recruits with rural and/or indigenous backgrounds had experienced
some form of social mobility by migrating to the cities and becoming
more educated (CVR 2003a). The Peruvian Truth Commission argues
that, when confronted with prevalent social inequalities, racial dis-
crimination, and economic exclusion, a minority of those young peo-
ple felt attracted to Sendero’s ideology and discourse of radical change
promising the chance of a better and fairer life in a new society (CVR
2013a, 8:247).
This last argument echoes theories of “relative deprivation” (Gurr
1970) that link social grievances to the propensity of people to rebel.
In a study comparing attitudes toward illegal protest among the gen-
eral public and the students of two public universities in Lima in 1987,
Muller, Dietz, and Finkel (1991) found that alienation from the po-
litical system was an important source of discontent among students
(many of them migrants or children of migrants from other parts of
the country).10 At the same time, discontent alone could not explain
the propensity to participate in illegal protest, which was strongly con-
ditioned by the expectancy of success and the perceived importance
of personal participation in illegal protest. As Gonzalo Portocarrero
(1998) suggests in another study of attitudes toward violence among
young people in Lima in the 1990s, those latter factors can be en-
hanced by exposure to an organization which promises radical social
change and an improvement of the social conditions for their partici-
pants and their communities.
A significant proportion of Sendero recruits were women, a fact that
received considerable attention (e.g., Nash 1992). Wood and Thomas
(2017) argue that political ideologies affect both the demand for and
supply of female fighters in rebel groups. Their cross-national study of
rebel organizations active between 1979 and 2009 suggests that leftist
rebel organizations are more likely than groups endorsing other types
of ideologies to feature high proportions of women in combat roles.
Sendero Luminoso is a case in point. However, and again, the exact
numbers are highly debated. Among the subpopulation of those impris-
oned and studied by the Peruvian Truth Commission, 18 percent were
female (CVR 2003a, statistical annex, dataset dec_pen_est); women
constituted 16 percent of the small sample of those convicted of ter-
rorism examined by Chávez de Paz (1989, 27), while other estimates
go up to 40 percent (Barrig 1993, 96–97). Again, though, the limited
quality and representativity of these data and estimates must be kept in

Soifer_6844-final.indb 61 8/17/18 11:53 AM

 62 Livia Isabella Schubiger and David Sulmont

mind. Overall, the proportion of women in the Shining Path has been
assessed to be comparable to that in the insurgencies in El Salvador
(FMLN) or Sri Lanka (LTTE) (E. Wood 2008, 552), around 30 per-
cent, and as unusually high in the universe of armed groups (ibid.).
Despite this high proportion of female members, Sendero Lumi-
noso did not devote special attention to gender equality in its writ-
ings, instead largely subsuming issues of gender under the discourse
of class. Consistent with assessments that have described Sendero’s re-
lationship to women as largely instrumental (Coral 1998, Balbuena
2007), women that held highly influential positions inside the Shin-
ing Path often did so due to their personal relationships with men—
the most prominent examples being Guzmán’s partners Augusta la
Torre and Elena Iparraguirre (CVR 2003a; E. Wood 2008, 552). At
the same time, the very fact that women did have the opportunity to
play significant roles in the Shining Path across all ranks clearly did
not go unnoticed, and was perceived by some women as a hitherto
unavailable avenue to social mobility. Based on field research in Aya-
cucho and Junín, Steven Zech notes that “significant female leader-
ship/participation within Sendero affected the movement’s discourse
when they rounded up the community for ‘chats’ and indoctrination
efforts. One ex-Senderista I’ve spoken with numerous times over the
past few years even went so far as to suggest that one of the key objec-
tives of the revolutionary movement was to confront a society that rel-
egated a woman’s role to procreation. So, in some cases, Sendero may
have given women new ideas about their role within their communi-
ties and Peruvian society more broadly.”11 Overall, for both men and
women, participation in war was shaped not only by their position in
society, but also to significant degrees by their exposure to strategies
of recruitment, governance, and violence, factors emphasized in theo-
ries that stress the endogenous nature of preference and network for-
mation during war (Kalyvas 2006; E. Wood 2003; E. Wood 2008; Ar-
jona and Kalyvas 2012).

Counterinsurgent Mobilization

One very consequential dynamic of the war was the mobilization of

rural residents into counterinsurgent groups. In some cases—particu-
larly in the mid-1980s and the 1990s—these rondas campesinas or co-
mités de autodefensa (CADs) were imposed, organized, or co-opted

Soifer_6844-final.indb 62 8/17/18 11:53 AM

 Civil Wars and Their Consequences 63

by government forces, who actively tried to foster armed resistance

against the Shining Path. In the 1990s, the rondas were officially in-
corporated into the state’s counterinsurgency efforts (Degregori et al.
1996; Fumerton 2002). To a presumably greater extent than in cases
such as Guatemala, however, civilians also rose up autonomously
against the insurgents (Degregori et al. 1996; Starn 1995b; Fumerton
2002; Schubiger 2013).
Many conventional theoretical approaches that privilege the pre-
war social or economic position of individuals as explanatory factors
for wartime mobilization are challenged by the overlapping motives
of insurgent and counterinsurgent recruits (Arjona and Kalyvas 2012;
Humphreys and Weinstein 2008; Gutiérrez Sanín 2008), as well as in-
stances of collective and individual side-switching that have been doc-
umented in cases ranging from Colombia to Sierra Leone (Oppenheim
et al. 2015)—and also Peru (Gavilán 2012; del Pino 1998). Indeed,
and as with militia mobilization more generally (Jentzsch, Kalyvas,
and Schubiger 2015), variation in ronda formation in Peru seems to
be inadequately explained with an exclusive focus on structural vari-
ables. While the quality of preexisting institutions has been shown to
be an important motivating and enabling factor facilitating resistance
to Sendero (La Serna 2012),12 the insurgents’ abusiveness and authori-
tarianism as well as an ideological position at odds with the values and
traditions of local communities likely played an equally important role
(Degregori et al. 1996; Degregori 1998a). In other words, while griev-
ances certainly served as a mobilizing resource for Sendero, the insur-
gents also strongly alienated the local population from early on (De-
gregori 1998a).
Less acknowledged is exposure to state violence as a driver of au-
tonomous counterinsurgent mobilization. Schubiger (2013) shows that
the state-led collective targeting of civilians during the counterinsur-
gency campaign of 1983–1985 had a lasting effect on the incentives
of communities to organize armed resistance against the Shining Path,
if at times only to publicly signal their nonaffiliation with insurgent
groups,13 and in spite of the tendencies of indiscriminate state violence
to also foster insurgent recruitment. The latter effect of state violence
is well documented beyond the Peruvian case (e.g., Goodwin 2001;
E. Wood 2003). These dynamics underscore the endogenous nature
of civilian collaboration during war (Kalyvas 2006; E. Wood 2003;
E. Wood 2008) that was also central for insurgent recruitment: as the

Soifer_6844-final.indb 63 8/17/18 11:53 AM

 64 Livia Isabella Schubiger and David Sulmont

conflict went on, exposure to violence and other wartime experiences

became important sources of motivation to join or support particular
actors, to defect to their opponents, or to stay neutral or flee.

Violence

While ethnicity did not serve as the defi ning source of mobilization
and claim-making during the armed conflict, it was one of the ma-
jor correlates with people’s vulnerability to violence. Indeed, wartime
violence was by no means evenly distributed across ethnicities, with
about 75 percent of the victims of the armed confl ict speaking one of
the country’s native languages (CVR 2003a, 8:108–109).14 While this
partially reflected higher proportions of indigenous people in the areas
most affected by the conflict, both state forces and the Shining Path
displayed certain racist tendencies in their application of violence (De-
gregori 1998a; del Pino 1998; CVR 2003a, vol. 8).
At the same time, patterns of targeting adopted by insurgent and
counterinsurgent actors varied markedly, both in comparison to each
other and over time. As outlined previously, the counterinsurgency ap-
proach of the Peruvian government changed drastically over the course
of the conflict: after an initial period of passivity, counterinsurgent
violence turned intense and became indiscriminate when the armed
forces entered the emergency zones in 1983. The ensuing years bear
some resemblance to the counterinsurgency campaigns in other Latin
American countries, where state violence against civilians was even
more extreme (e.g., Goodwin 2001; E. Wood 2003; E. Wood 2008;
Wickham- Crowley 2015). However, over the course of the whole con-
flict, and in stark contrast to cases such as Guatemala or El Salva-
dor, Sendero Luminoso appears to have surpassed even the govern-
ment in the extent of its lethal violence against civilians (CVR 2003a;
E. Wood 2008). Moreover, while insurgent violence grew more in-
discriminate and intense over the course of the war, state violence
turned increasingly selective over time (Degregori 1998a; CVR 2003a;
E. Wood 2008).
Insurgent and state violence also displayed wide variation in regard
to particular forms of victimization: sexual violence, for example, was
mostly conducted by state agents, who were recorded as responsible
for 83 percent of the reported cases by the Peruvian Truth Commis-
sion, while 11 percent were attributed to Sendero Luminoso and the
MRTA (E. Wood 2008; CVR 2003a, 6:201–202).15 Using a data set

Soifer_6844-final.indb 64 8/17/18 11:53 AM

 Civil Wars and Their Consequences 65

on state-perpetrated sexual violence in Guatemala and Peru compiled

from the Truth Commission documents and nongovernmental human
rights organizations, Leiby (2009) fi nds that in contrast to Guatemala,
where sexual violence was most often perpetrated during sweeps and
massacres, most victims of state sexual violence in Peru were violated
while in detention (Leiby 2009, 454–455). At the same time, she states
that while the Shining Path adherents committed sexual violence much
less often than state personnel (still outstripping the MRTA), many of
Sendero’s actions “were among the most brutal attacks” (Leiby 2009,
466). Chapter 6 in this volume provides a more detailed discussion of
this particular instantiation of violence.
In a bid to explain both the timing of Sendero’s decision to launch
the armed conflict as well as its extensive use of violence against mem-
bers of other leftist groups, Ron (2001) argues that tactical choices
were shaped by ideology in consequential ways. It is indeed difficult to
understand the particular strategies and forms of violence adopted by
Sendero Luminoso and the armed forces without taking their ideolo-
gies and institutions into account. This insight resonates with recent
research on other cases that has started to explore the role of ideology
and armed group institutions in shaping patterns of wartime violence
against civilians (Hoover Green 2011; Thaler 2012; E. Wood 2010;
Gutiérrez Sanín and Wood 2014). These studies have shown that the
type and quality of insurgent institutions—for example, institutions
for regularized political indoctrination—themselves typically shaped
by ideology to significant degrees, influence not only the extent to
which armed group leaders manage to control the intensity of violence
against civilians, but also the consistency with which combatants wield
particular forms of violence while eschewing others (Hoover Green
2011; Thaler 2012; E. Wood 2010; Gutiérrez Sanín and Wood 2014).
Ideology can also “overload” actors’ strategies with meanings that be-
come increasingly disconnected from the social, cultural, and political
frames of the previous life of individuals involved in highly ideologized
organizations, as well as from the experience of those whose lives these
organizations claim to be improving (e.g., Wieviorka 2009, 152).
Sendero Luminoso’s use of openly displayed violence to sanction be-
havioral transgressions of community members in areas where it held
early influence was intrinsically connected to its aspirations to shape
and rule over the everyday lives of civilians according to their ideologi-
cal principles. During the early years of its establishment in areas of en-
demic state weakness, the Shining Path was able to garner considerable

Soifer_6844-final.indb 65 8/17/18 11:53 AM

 66 Livia Isabella Schubiger and David Sulmont

support through acts of “social cleansing” (Degregori 1998a, 136), the

violent punishment of thieves, adulterers, and other community mem-
bers accused of violating the local order (see also, e.g., del Pino 1998,
161). However, this support vanished once both insurgent rule and vi-
olence escalated to extremes (e.g., del Pino 1998; Degregori 1998a).

Governance

Rebel governance—defi ned as the “set of actions insurgents engage in

to regulate the social, political, and economic life of non-combatants
during war” (Arjona, Kasfi r, and Mampilly 2015, 3)—took an intru-
sive form in the case of the Shining Path, with the insurgents striv-
ing to impose their own set of tight social rules in the areas they con-
trolled. That insurgents establish state-like structures is by no means
unique to the Peruvian case. A growing body of literature studies how
armed groups govern civilian life in areas they control (e.g., Weinstein
2007; Mampilly 2011; Arjona 2014; Arjona, Kasfi r, and Mampilly
2015; Arjona 2017). Marxist insurgencies, but also revolutionary in-
surgencies of other ideological orientations (Kalyvas 2015), appear to
have been particularly dedicated in their aspirations to govern terri-
tory through comprehensive political institutions (Balcells and Kaly-
vas 2015; Huang 2016).
As Guillermo O’Donnell (2004) argues, many Latin American
states have not been able to fully enforce and ensure the monopoly of
coercion and administrative means throughout their territories, leav-
ing aside areas where different types of informal, patrimonial, or ma-
fia rule can persist. Those areas often coexist with democratic national
institutions, but the government is unable to fully enforce the rule of
law.16 Wickham-Crowley (2015) shows that, in line with his theory on
the conditions favoring insurgent governance and its collapse, Sendero
Luminoso was able to establish a strong presence in areas largely side-
stepped by the land reform under Belaunde’s fi rst government (1963–
1968) and the following military regime (1968–1980). He further
holds that, similar to certain regions in El Salvador and Guatemala,
Sendero governance was strongest in regions affected by violent gov-
ernment repression (cf. Goodwin 2001; E. Wood 2003), and that it
was weaker where other political parties or social organizations had
been established fi rst. At the same time, the heavy blows infl icted by
targeted, intelligence-driven actions on behalf of the state in addition
to long-term and massive counterinsurgency operations eventually
helped to undermine insurgent military power and territorial control,

Soifer_6844-final.indb 66 8/17/18 11:53 AM

 Civil Wars and Their Consequences 67

hence also destroying the insurgents’ capacity to act like counter-states

(Wickham-Crowley 2015, 66–67).
To be sure, Sendero undermined its own ability to govern by vio-
lating its obligations to their presumed civilian constituency, failing to
protect the population from state violence (McClintock 1989, 90; Isbell
1992, 90; Degregori 1998a, 141; Fumerton 2001, 482, 484; Weinstein
2007, 191–192), escalating its own level of coercion and civilian abuse
(Degregori 1998a), and undermining its constituencies’ livelihoods, for
example by closing off markets (Wickham-Crowley 2015, 65; Degre-
gori 1998a). These fissures turned many citizens against the Shining
Path (e.g., Degregori et al. 1996; Degregori 1998a; del Pino 1998).
Importantly, the insurgents were not the only nonstate actors provid-
ing wartime governance and local order where state presence was weak;
the rondas also frequently fulfilled that role (García- Godos 2006; see
also Fumerton 2001; Schubiger 2013). The rondas transformed their
communities in ways that reached beyond the conflict, for example by
reshaping their relationship to the state (García- Godos 2006).

(Some) Legacies of Political Violence

The legacies of the armed confl ict are multiple and complex. As polit-
ical legacies of the conflict are discussed in-depth in other chapters of
this book, we focus here on three areas that are not the main subject of
those chapters: sociodemographic transformations, economic impact,
and truth and justice.

Sociodemographic Transformations

The conflict in Peru destroyed thousands of lives and families, as well

as hundreds of communities in the areas most affected by violence.
Particularly in Ayacucho, the conflict had deep demographic repercus-
sions. As we have shown in figure 2.1, most of the lethal violence oc-
curred between 1982 and 1993. In about that same time period, Peru
performed two national censuses, the fi rst in 1981 and the second in
1993. A comparison of these censuses reveals that Ayacucho was the
only region in the country that experienced a net loss in population in
those years: in 1993 it had a 2.2 percent smaller population than in
1981. According to the CVR statistical estimates, the number of vic-
tims in Ayacucho between 1981 and 1993 could amount to 24,237, or
roughly 4.8 percent of the total population of Ayacucho in 1981. Con-

Soifer_6844-final.indb 67 8/17/18 11:53 AM

 68 Livia Isabella Schubiger and David Sulmont

sidering that the fatal casualties were mostly young men in a reproduc-
tive age, the economic consequences and the disruption of family lives
were especially severe in that region.
Although the significant number of victims and the indiscriminate
character of violence in some periods and regions (particularly in Aya-
cucho in 1983–1984) might suggest otherwise, Sendero Luminoso
and state agents frequently targeted particular members of the soci-
ety. Casualties were especially high among social and political lead-
ers in the most affected locations, mostly in rural and indigenous re-
gions. The Peruvian Truth Commission has estimated that 23 percent
of the victims killed by the Shining Path were local elected authori-
ties and leaders of social organizations. In fact, according to the CVR
database, the number of civil local authorities (majors, councilmen,
judges, public officials, etc.) and social organization leaders killed by
Sendero Luminoso is higher than the number of military or police ca-
sualties caused by that organization (1,682 vs. 1,671) (CVR 2003a).
Those fatalities represent a loss of social and political capital that se-
verely affected political parties (from both the left and right wings of
the political spectrum) as well as social organizations from peasant
and indigenous communities. This kind of violence also had a deter-
rence effect, preventing people from engaging in political or social mo-
bilization and provoking lasting effects within the Peruvian party sys-
tem and civil society organizations (see chapters 7, 8, and 9).
Forced displacement was probably one of the most important de-
mographic consequences of the civil war in Peru. It has been estimated
that 600,000 people—approximately 2.7 percent of the national pop-
ulation, according to the 1993 census—were forced to leave their res-
idences during the confl ict in Peru (Diez Hurtado 2003). Other con-
flicts in the Latin American region have also forced huge numbers of
displaced people: 6 million in Colombia up to 2014, from a total pop-
ulation of 48.9 million (12 percent of the population) (Internal Dis-
placement Monitoring Centre 2014); between 500,000 and 1.5 million
in Guatemala from a total population of 10 million during the con-
flict years there (Comisión de Esclarecimiento Histórico 1999, 3:211).
In Peru, as in Colombia, forced displacement was mostly internal. In
contrast, for Guatemala it has been estimated that 100,000 people be-
came refugees in neighboring countries (W. Wood 1994, 610).
As in other confl icts, forced displacement or migration is a source
of deep social transformation, destroying economic resources and dis-
rupting traditional cultures, communities, and ways of life (Castles
2003; W. Wood 1994). During the fi rst phase of displacement, peo-

Soifer_6844-final.indb 68 8/17/18 11:53 AM

 Civil Wars and Their Consequences 69

Table 2.1. Peru 1981–1993, change in social development indicators in

Ayacucho and at national level

Ayacucho Peru

Social indicators 1981 1993 Change 1981 1993 Change

Access to running water 17 27 159 43 47 109

(% households)
Access to public sanitation 8 14 175 35 40 114
(% of households)
Access to electricity 10 25 250 46 54 117
(% of households)
Post-secondary education 4 11 289 7 17 232
(% of people)

Data source: INEI (2017a, 2017b).

ple are very vulnerable to serious problems such as mental health is-
sues, domestic violence, and alcohol abuse (Moya Medina 2010). Two
decades after the conflict, mental health problems remain an impor-
tant consequence for affected families and survivors (Velazquez 2014).
These issues might have aggravated the problem of gender violence
against women, which is part of a longer history of violence and in-
equality in Peru, as Jelke Boesten discusses in chapter 6.
Within-region displacement transformed the geographical distribu-
tion of the population. Flight from the countryside to the cities was a
widespread phenomenon in the areas affected by the armed conflict,
particularly in Ayacucho. According to the 1981 census, 36.5 per-
cent of Ayacucho’s population lived in urban areas, whereas in 1993
it was 48.1 percent, the highest relative increase in urban population
of any region in the country (the national urban population rose from
65.2 percent to 70.1 percent between those years).
Adaptation to new living conditions was extremely difficult for peo-
ple fleeing conflict-ravaged areas, particularly when they lost relatives
(mostly young males) and most of their belongings. But resettlement
in urban areas over time gave displaced people access to better liv-
ing conditions and social services that were unavailable in remote ru-
ral areas. Moreover, as we can see in table 2.1, some social develop-
ment indicators (access to running water, sanitation, and electricity;
rate of postsecondary education) actually rose at higher rates in Aya-

Soifer_6844-final.indb 69 8/17/18 11:53 AM

 70 Livia Isabella Schubiger and David Sulmont

cucho than the national average during the confl ict years, even in the
midst of a severe economic crisis at the end of the 1980s.
After a harsh initial phase of adjustment, many displaced people
from rural areas in Ayacucho’s main cities gained access to health care
services (pre- and post-natal control, hospitalized childbirth, pediatric
services) that improved some health indicators like mother and child
mortality rates (Moya Medina 2010).

Economic Impact

It is difficult to assess the economic impact of the conflict. Some stud-

ies estimate that its economic cost between 1980 and 1993 rose to ap-
proximately US$21 billion (CVR 2003a, 8:230), a figure that repre-
sents 48 percent of Peru’s gross national product in 1993. Insurgent
groups targeted public infrastructure, mainly electric pylons, causing
frequent blackouts in major cities, but no strategic infrastructure was
severely damaged during the conflict. Violence in several regions dis-
couraged private and public investment, and it mostly affected regions
that already had a marginal role in the overall economic output of the
country.
The conflict overlapped with a severe economic crisis that affected
Peru—as well as many other countries in Latin America—during the
1980s, although the economic crisis had structural causes different
from the ones related to the armed conflict. Thus the armed conflict
added additional burdens to the economic and political instability of
Peruvian society by the late 1980s and early 1990s.
The conflict also had an indirect and more lasting impact on the
ideological framing of economic policy. The fact that Sendero Lumi-
noso presented itself as a Communist and Marxist group seeking rad-
ical social transformations severely damaged the public perception of
legal left-wing leaders, parties, and movements. This produced a sort
of stigmatization of left-wing politics and policies (see chapter 8), in-
cluding economic policies that, among other things, emphasized the
state’s central role in economic and social development. This might
have facilitated the rise and hegemony of the neoliberal economic
thinking in Peru and the consolidation of the right-wing political spec-
trum that supports these policies (see chapters 3 and 9).

Truth and Justice

As in other Latin American countries where large-scale human rights

violations were committed (e.g., Chile, Argentina, Guatemala, and El

Soifer_6844-final.indb 70 8/17/18 11:53 AM

 Civil Wars and Their Consequences 71

Salvador), the armed confl ict in Peru led to the emergence of a human
rights movement. Several victims’ organizations and human rights
NGOs were created during the conflict, some of them sponsored by
the Catholic Church, the Evangelical churches, and international part-
ner organizations. As Maritza Paredes argues in chapter 7, the human
rights movement played an important, and somewhat unexpected,
role in Peruvian politics and civil society in the decades following the
conflict.
During the democratic transition in 2000, the Peruvian human
rights movement was particularly engaged in the advocacy for and the
formation of the Peruvian Truth and Reconciliation Commission. Fol-
lowing and studying the best practices of previous truth commissions
in Chile and Argentina, but particularly in South Africa and Guate-
mala, the Peruvian CVR set new standards for this kind of project,
which later influenced similar commissions in other parts of the world
(Hayner 2011; Landman 2006).
The CVR presented an alternative narrative of the conflict that
challenged the “victorious” version of the Fujimori regime and the
military as defeating Sendero Luminoso and pacifying the country. It
also presented strong evidence that portrayed Sendero Luminoso as
the main perpetrator of deadly violence during the confl ict, in sharp
contrast with other confl icts in Latin America. The CVR provided the
human rights movements and the victims with important tools (legit-
imacy, data, forensic evidence, recommendations concerning repara-
tions, etc.) to advance their claims for justice and reparations from the
state and the judiciary system. However, as in the cases of other truth
commissions (Hayner 2011; Bakiner 2016; Landman 2006), the role
of the CVR is still a matter of controversy. Even if some public opin-
ion polls have portrayed a relatively favorable image of the CVR, most
people saw it as unable to directly contribute to a reconciliation pro-
cess (Sulmont 2007; Barrantes 2007).
Despite those controversies and their temporary character, as Onur
Bakiner (2016) points out, truth commissions can produce important
changes in human rights accountability, and the short- and long-term
effects of those impacts ultimately depend on the interplay of forces
in civil society, political actors, and the institutions in a particular
country.
Since most of the victims were members of some of the most mar-
ginalized groups of Peruvian society, their political leverage to advance
judiciary processes or reparation policies remains limited. In contrast,
sectors of the political establishment that were in power during the

Soifer_6844-final.indb 71 8/17/18 11:53 AM

 72 Livia Isabella Schubiger and David Sulmont

1980s and 1990s have regained political influence, enhancing their

ability to contest the narrative produced by the CVR.17 As this vol-
ume’s chapter 11 shows, contested memories of the Internal Armed
Conflict are one of the recurrent features of political and historical de-
bates in contemporary Peruvian society.
The consequences of high-profi le judicial trials or cases on human
rights abuses are still shaping or influencing the political careers of
some of the most important political actors in Peru. To give a few ex-
amples, Alberto Fujimori’s sentence in 2009 to twenty-five years in
prison for human rights violations has been a crucial issue for the fu-
jimorismo as a political movement and its identity. In June 2017 the
office of the attorney general decided to reopen a human rights vio-
lation case involving former president Ollanta Humala (2011–2016)
dating to when he was the military commander of a countersubver-
sive base in the Amazon region in 1992 (Redacción EC 2017).18 And in
September 2017 a new trial began concerning the 1986 “Massacre of
El Frontón,” the extrajudicial execution of surrendered inmates when
units of the Peruvian Navy suffocated a mutiny of Sendero inmates on
the prison island of El Frontón. This trial may involve Alan García,
Peru’s president from 1985 to 1990 and 2006 to 2011, and a presiden-
tial candidate in 2016, who ordered the navy to intervene in this oper-
ation (Mejía Huaraca 2017).
The huge number of human rights violations committed during the
armed conflict still represents a heavy burden for the victims’ rela-
tives and, at an institutional level, for the judiciary system, where sev-
eral cases are still pending or are being prosecuted many decades af-
ter their occurrence. In mid-August 2017, after twelve years of trial,
lengthy prison sentences where issued to military personnel found re-
sponsible for human rights violations in the case of Los Cabitos, a mil-
itary base in the city of Ayacucho where many “suspected terrorists”
were tortured, assassinated, or disappeared in the 1980s and 1990s.19
In June 2016 the Peruvian parliament enacted the law “For the Search
of the Disappeared Persons during the Period of Violence 1980–2000,”
which committed the Peruvian state to researching, locating, identify-
ing, and returning to their relatives the remains of an estimated thir-
teen thousand people who disappeared during the confl ict (Ministerio
de Justicia y Derechos Humanos 2017).
Concerning justice and reparations, there is a case to be made for
either a glass half-filled or half-empty, depending on the perspective of
the observer. Justice and reparations are still actively pursued and de-

Soifer_6844-final.indb 72 8/17/18 11:53 AM

 Civil Wars and Their Consequences 73

bated, with some advances and setbacks (Macher 2014). These issues
will continue to challenge the commitment of the Peruvian state to hu-
man rights protection and democratic values for many years or even
decades to come. As long as democratic institutions will allow public
debate on these matters, we can expect continuing discussion regard-
ing the legacies of the Internal Armed Confl ict in Peru.

Remaining Puzzles and Avenues for Future Research

The Peruvian armed conflict illuminates several issues and dynam-

ics that are still poorly understood in the literature on political vio-
lence. In this conclusion, we would like to highlight four: (1) the role
of ideology in armed conflict; (2) the intersection of social inequalities
and the consequences for wartime violence and mobilization; (3) the
social and institutional legacies of armed conflict; and (4) the capac-
ity of transitional justice mechanisms to foster long-term post-conflict
reconciliation.
In their conclusions, the Peruvian Truth and Reconciliation Commis-
sion fi nds that the “immediate cause” of the confl ict was the decision
of the PCP-Sendero Luminoso to initiate an armed struggle against the
Peruvian government, and that this decision was mainly driven by its
“fundamentalist” brand of ideology (CVR 2003a, 8:246). French soci-
ologist Michel Wieviorka (2009, 2015) argues that violence can emerge
when individuals and social groups in a society are not able to engage
in institutionalized confl ict relationships. From this perspective, there
is an important distinction between confl ict and violence. Confl ict is
an inherent feature of societies, implying several actors—in most cases
with unequal assets and social status—that compete against each other
with the aim of transforming their relationship. For some groups, the
unavailability of institutionalized channels to engage in such relation-
ships can lead to what Wieviorka calls “desubjectivation.”20 “Desub-
jectivation” is the impossibility to become a subject, which he argues
gives rise to the appearance of individuals or groups that may resort to
instrumental or expressive violence in order to manifest their existence
or transform social relationships on their own terms, without engag-
ing in negotiation with other actors. 21 When groups reach a high level
of isolation from mainstream social processes and social movements,
they can become “hyper-subjects,” overloading their actions with new
meanings through a dogmatic and fundamentalist ideology, political

Soifer_6844-final.indb 73 8/17/18 11:53 AM

 74 Livia Isabella Schubiger and David Sulmont

or religious (Wieviorka 2004). From this perspective, Sendero can be

described as a “hyper-subject.” Indeed, through its violence and ideo-
logical purism, the Shining Path increasingly distanced itself from the
population it claimed to represent. Why, however, did Sendero adopt
this particular ideology, and why did it choose such an extreme and
counterproductive approach? Why do armed groups adopt specific ide-
ologies, and what are the implications?
Pronounced and often highly counterproductive forms of vio-
lence demand a deeper engagement with armed group ideology than
the reduction to its instrumental use (Ron 2001; Gutiérrez Sanín and
E. Wood 2014). Scholars of political violence are just beginning to un-
derstand the role of “ideas and normative commitments that moti-
vate and coordinate, as the bearers of identities, strategies, and insti-
tutions” (Gutiérrez Sanín and E. Wood 2014, 222) in armed confl ict.
The issues to be explored range from the endorsement of particular
ideas and programs by the leadership to the mechanisms that foster
ideological transformations and the micro-foundations underlying the
normative commitments of rank-and-file combatants (Gutiérrez Sanín
and E. Wood 2014; Hoover Green 2011).
Another question scholars should continue to explore is why spe-
cific social, economic, and political inequalities are sometimes incor-
porated in armed group mobilization frames—and indeed, become the
very core of a conflict’s cleavage—while in other very similar circum-
stances they are not. Similarly, why do patterns of targeting and vio-
lence often overlap with these cleavages, but sometimes do not? Despite
a large body of literature that has illuminated the role of ethnicity in
armed conflict (e.g., Cederman, Gleditsch, and Weidmann 2011), the
complex intersections of identity, mobilization, and violence are still
poorly understood, partially due to an often overly simplifying distinc-
tion between different “types” of conflict. Peru is a case in point. The
fact that ethnic cleavages were not the primary source of insurgent
claim-making and mobilization should not be mistaken as an indicator
for the nonethnic character of violence. As we have seen, the Peruvian
case challenges common conceptualizations of the role of ethnicity in
armed conflict. While clearly not qualifying as an “ethnic civil war,”
and despite the irrelevance of ethnicity for the conflict’s “master cleav-
age” (Kalyvas 2003), wartime patterns of violence revealed a stark de-
gree of ethnic discrimination, although in complex ways. In this re-
gard, the confl ict mirrored dynamics in politics more generally, both
in Peru and other Latin American countries, where the salience and

Soifer_6844-final.indb 74 8/17/18 11:53 AM

 Civil Wars and Their Consequences 75

relevance of ethnicity failed to disappear despite the general tendency

to mobilize citizens based on class-based rather than identity-based is-
sues (Yashar 2015). Future research should look more closely at the
role of ethnicity, even in conflicts where identity-based categories have
seemingly little relevance at fi rst sight.
Another area that needs further exploration are the social and in-
stitutional legacies of political violence. War, through violence, trans-
forms individual and group preferences, behaviors, and identities. Be-
ing a victim or perpetrator of violence can lead to new meanings for
social action, or to actions whose initial meanings or purposes be-
come lost as others arise. Armed conflict also transforms social rela-
tionships in other ways and more indirectly, through the destruction
of social capital via the removal or killing of leaders and the disruption
of whole communities through forced displacement. Another socially
disruptive legacy of civil war that has been documented in Peru is so-
cial and political polarization at not only the national and regional
level but also the very local level, induced through the simultaneous
dynamics of pro- and counterinsurgent mobilization (Theidon 2006;
E. Wood 2008). At the same time, and as documented in various post-
conflict contexts, exposure to wartime violence can also increase the
capacity of communities and individuals for political participation and
local collective action (Bellows and Miguel 2009; Blattman 2009; Gil-
ligan, Pasquale, and Samii 2014).
One of the most notable consequences of the confl ict was the emer-
gence and growth of a very vibrant human rights movement that lev-
eraged its influence to shape the transitional period, for example, by
successfully advocating for the installment of a truth and reconcilia-
tion commission. Yet many questions regarding the legacies of the civil
war still remain to be explored. How did the war affect gender rela-
tions, for example, and how does this effect vary across subgroups—
such as ordinary citizens, former senderistas, and members of civilian
self-defense groups? How did wartime mobilization at the local level
affect postwar collective action, and what are the mechanisms driving
this transformation? What determines the trajectories of former com-
batants as they reintegrate into society, and how are social, economic,
and ethnic inequalities affected by exposure to violence wielded by
state actors and armed groups?
Finally, future research should continue to explore the role of truth
commissions in post-confl ict societies. When political independence
and high scientific and methodological standards are met (Landman

Soifer_6844-final.indb 75 8/17/18 11:53 AM

 76 Livia Isabella Schubiger and David Sulmont

2006), truth commissions can challenge some “common knowledge”

about the confl ict and enhance the cause of human rights movements
and victims. However, high expectations concerning reparations and
“reconciliation” can undermine the legacy and role of such commis-
sions (which are transitory by defi nition). Some actors involved in the
conflict can regain power and influence in post-confl ict societies, par-
ticularly when some of them have always been a part of the “power
elites,” rendering it more difficult to implement mechanisms for tran-
sitional justice. The role of truth commissions in empowering demo-
cratic actors, movements, and institutions should be a matter of future
discussion and research. 22

Notes

We warmly thank Fabian Morgenthaler, Abbey Steele, Manuel Vogt, Steven

Zech, the editors, and two anonymous reviewers for helpful comments and
input.

1. The comités de autodefensa (CADs) were in many cases promoted by

the state and embedded into its counterinsurgency policy, especially during
the 1990s. This is also the reason why the Comisión de la Verdad y Reconci-
liación, when aggregating the victims, often attributed fatalities identified as
victims of the CADs to agents of the state.
2. According to the 1981 national census, Ayacucho had 503,392 inhab-
itants, or 3 percent of the national population; 63.5 percent of them lived in
rural areas, in contrast to 34.8 percent of the national population. The illiter-
acy rate was 45 percent, the third-highest in the country (the national illiter-
acy rate in 1981 was 18.1 percent), and more than 88 percent of Ayacuchanos
spoke an indigenous language (mainly Quechua), in comparison to 25 percent
of the national population (INEI 2017a).
3. It is important to make a distinction between casualties or victims re-
ported to the CVR and CVR’s own statistical estimates. During its fieldwork,
the CVR received nearly 17,000 testimonies, through which it could identify
23,969 fatal casualties (CVR 2003a, 1:119). Fatal casualties were defi ned as
including people who were murdered, disappeared, or died in combat. Among
the reported victims, Sendero Luminoso was identified as the perpetrator in
nearly 54 percent of the cases; 33 percent were attributed to state agents (ex-
cluding self-defense forces); and the rest were ascribed to other or nonidenti-
fied actors or circumstances (CVR 2003a, 1:136–137). To assess the probable
magnitude of the confl ict, estimates were performed using statistical methods
described in Annex 3 of the fi nal report (see also Ball et al. 2003).
4. While the classification of the Peruvian armed confl ict as a “civil war”
is contested in the Peruvian literature, according to the standard defi nition of
a civil war used in political science—an internal armed confl ict between the

Soifer_6844-final.indb 76 8/17/18 11:53 AM

 Civil Wars and Their Consequences 77

government of a state and at least one opposition group reaching certain fatal-
ity threshold—this categorization is straightforward; see, for example, Kaly-
vas and Balcells (2010).
5. Kalyvas and Balcells include internal armed confl icts that resulted in
at least one thousand war-related deaths in total and during at least one sin-
gle year. Many internal armed confl icts do not reach this threshold. The Pe-
ruvian case did.
6. For examples, see the cases of the “Comité Zonal Fundamental Cangallo-
Victor Fajardo, comité principal,” “La Violencia en las comunidades de Luca-
namarca, Sancos y Sacsamarca,” and the studies concerning the presence of
Sendero Luminoso in public universities, described in volume 5 of the Peru-
vian Truth Commission Final Report (CVR 2003a).
7. According to the crossnational ACD2EPR 2014 data set (http://www
.icr.ethz.ch/data/acd2epr), the EGP also engaged in ethnic claim-making (Vogt
et al. 2015), which was not the case in Peru.
8. We refer to the data set published by the CVR, which deviates slightly
from the figures in volume 8 of the report, as the latter was fi nalized before
the data set.
9. Compared with data from the 1993 National Census (INEI 2017c).
10. On the subject of Sendero Luminoso and its effects on Peru’s public
universitities, see chapter 5.
11. Steven T. Zech, personal (written) communication, January 2016.
12. See also Arjona 2017 on the relationship between the FARC and civil-
ians in Colombia.
13. On similar consequences of indiscriminate state violence in other con-
fl icts, see also Lyall 2009, 337; Kalyvas 2006, 167–168. For anecdotal evi-
dence on Peru, see, for example, Fumerton 2002; García-Godos 2006; Wein-
stein 2007.
14. This compares to Guatemala, for example, where the proportion of
indigenous victims was even higher (Thorp, Caumartin, and Gray-Molina
2006, 456).
15. Focusing on rape as one particular form of sexual violence, Cohen
(2013) fi nds that in a data set of eighty-six civil wars between 1980 and 2009,
a minority of isolated rape incidents were reported as perpetrated by armed
actors in 8 percent of confl icts. In those cases with reported rape, both state
and nonstate actors committed this form of sexual violence in 62 percent of
confl icts. In 31 percent of these cases, only state actors reportedly commit-
ted rape, and in 7 percent rape was only perpetrated by insurgents (Cohen
2013, 467).
16. Using some version of Tilly’s (1990) theories on warfare and state
building, other authors, like Centeno (2002), also highlight the “incomplete”
features of state building in Latin American countries and their limited capac-
ity to ensure the monopoly of coercion in their societies.
17. Alan García, president from 1985 to 1990, was reelected with a new
mandate in 2006–2011; Keiko Fujimori, the daughter of Alberto Fujimori,
contested the second round of presidential elections in 2011 and 2016.
18. In August 2017 Ollanta Humala and his wife were issued “preventive”

Soifer_6844-final.indb 77 8/17/18 11:53 AM

 78 Livia Isabella Schubiger and David Sulmont

arrest warrants on corruption charges, but one of the judge’s arguments for
issuing the warrants was the possible implication of the former president in a
human rights violations case (extrajudicial executions) in the town of Madre
Mía in 1992.
19. Evidence discovered at Los Cabitos suggests that crematorium ovens
were used to dispose of victims’ corpses.
20. The “subject” can be defi ned as an autonomous social actor (either an
individual or a group) capable of formulating choices despite dominant so-
cial pressures and engaging in social relationships with other actors. On the
theory of the subject, see the following works by the French sociologist Alain
Touraine: Touraine 1992; Touraine and Khosrokhavar 2000.
21. It is possible to see in those insights the classical sociological theme
and theory of anomie developed in Durkheim 1933; Durkheim 1951; and
Merton 1968.
22. On those issues concerning the role and impact of truth commissions,
we fi nd the work of Onur Bakiner (2016) particularly interesting.

Soifer_6844-final.indb 78 8/17/18 11:53 AM

 CHAP T ER 3

From Oligarchic Domination to Neoliberal

Governance: The Shining Path and the
Transformation of Peru’s Constitutional Order
Maxwell A. Cameron

Out of the chaos created by the Shining Path’s prolonged people’s war,
there arose in Peru a new constitutional order. The constitution of
1993, written in the aftermath of President Alberto Fujimori’s 1992
presidential self-coup, or autogolpe, replaced the constitution of 1979,
which had been drafted as part of the transition from military rule.
Whereas the fi rst constitution preceded Peru’s internal confl ict, the
second occurred during it. By comparing the two texts, and the cir-
cumstances surrounding their adoption, we can begin to assess one el-
ement of the legacy of the internal confl ict: the emergence of the neo-
liberal governance system that persists to this day.
Two lessons are apparent. First, the internal confl ict influenced the
development of Peru’s constitutional order in two ways: it created the
crisis that enabled a new constitution to be written, and it weakened
the Left and popular organizations necessary to contest neoliberalism.1
Analysts agree that the 1993 constitution was more authoritarian and
neoliberal than its 1979 predecessor (Rubio Correa 2012; Teivainen
2002; Planas 1999; García Belaunde 1996). It rolled back social fea-
tures of the 1979 constitution and facilitated the concentration of
power in the hands of the executive branch of government. 2 It also
proved remarkably enduring, in large measure because it aligned Peru
with neoliberal precepts. Peru did not emulate other countries in the
region that undertook constitutional reforms (often called republican
“refounding”) as part of a left turn (Cameron and Hershberg 2010;
Levitsky and Roberts 2011; Ellner 2014). One could argue that the
Shining Path foreclosed the possibility of a left turn by creating an
emergency situation which resulted in a constitution that locked in a
neoliberal economic model and, with it, a correspondingly limited elec-
toral democracy.3

Soifer_6844-final.indb 79 8/17/18 11:53 AM

 80 Maxwell A. Cameron

Second, and more optimistically, the confl ict weakened but did not
destroy Peru’s defective democracy. The constitutions of 1979 and
1993 provided similar protections for fundamental rights and free-
doms. The 1993 constitution was written under Fujimori, but it did
not fully sanction the kinds of abuses of power that occurred under his
rule, and indeed the president almost immediately found himself chaf-
ing against the constraints it imposed on his government. If we place
both constitutions within a longer historical perspective, the 1979
constitution appears to reflect a deeper process of societal democrati-
zation under military rule and thereafter. Furthermore, although the
1993 constitution rolled back certain social democratic features of
the 1979 constitution and rolled out neoliberal ones, it nonetheless re-
tained other constitutional and democratic elements; moreover, some
of the authoritarian features were ultimately overturned. The role of
the Shining Path within this larger process of democratization was
negative but insufficient to destroy Peru’s electoral democracy.
To support these claims, this chapter is organized into six parts.
The fi rst examines the breakdown of oligarchic domination and the
process of social democratization that culminated in the 1978 Constit-
uent Assembly. The second interprets the Shining Path within the con-
text of this transformation. The third section examines how, due to the
emergency situation, the stresses imposed on the newly democratized
political regime led to a rupture of the constitutional order. The fourth
section compares the 1979 and 1993 constitutions. The fi fth section
discusses the emergence of neoliberal governance techniques fostered
by the 1993 constitution. The sixth section places Peru in the context
of left turns elsewhere in Latin America. The fi nal section concludes.

The Breakdown of Oligarchic Domination

The initial crisis of the oligarchic state in the 1920s and 1930s did not
result in the breakdown of oligarchic domination, which remained en-
trenched, especially in the countryside, into the mid-twentieth century.
Conflict between Peru’s military and the leader of the American Pop-
ular Revolutionary Alliance (Alianza Popular Revolucionaria Ameri-
cana, or APRA), Víctor Raúl Haya de la Torre, resulted in a veto on
Haya de la Torre holding public office and the postponement of neces-
sary social, political, and economic reforms. When these reforms were
fi nally adopted, it was, paradoxically, under the tutelage of reformist

Soifer_6844-final.indb 80 8/17/18 11:53 AM

 From Oligarchic Domination to Neoliberal Governance 81

military officers who seized power in a coup in 1968. The rural oligar-
chy was fi nally destroyed by an extensive land reform, the creation of
peasant cooperatives, unionization, industrial communities, and cor-
poratist institutions. The aim of the military was to modernize the na-
tion while limiting class confl ict. Instead of attenuating class confl ict,
however, the reforms exacerbated it, and Peru entered a period of vio-
lence that would ultimately cost tens of thousands of lives and untold
amounts of property damage.
Prior to the sweeping reforms undertaken by the military regime be-
tween 1968 and 1980, Peruvian society was characterized by structural
dualism between the coast and the sierra—a source of cultural hetero-
geneity that dated to the colonial period (Cotler 1976, 35). The coast
was the seat of criollo culture; the highlands, of indigenous cultures.
The urban areas along the coast monopolized technologies of social
communication, including newspapers and television, and were inte-
grated into global economic markets. The sierra, with its preindustrial
economic arrangements, was a sort of archipelago containing “vast
pockets of isolation” in which “traditional social forms of organiza-
tion,” like indigenous communities, coexisted with large landholdings,
called latifundias or haciendas (Cotler 1976, 36). On the haciendas
and latifundias, the relationship between landlords and rural bosses
(or gamonales), on the one hand, and peasants and landless indigenous
workers (or colonos), on the other, was particularly repressive and ex-
ploitative. The gamonal system relied on savage punishment and dis-
cipline to dominate the “Indians.”4 But it also rested on manipulation
that was “made possible, among other causes, by the monopoly exer-
cised by the dominant on knowledge of the Castilian tongue” (Degre-
gori 1989, 10). Power was concentrated in the person of the landowner.
Mestizos dominated the professions: lawyers, judges, governors, po-
lice, merchants, mayors, and tax collectors were recruited overwhelm-
ingly from among them. They monopolized access to written texts and
restricted literacy and education to guarantee their domination. Only
the literate could elect or be elected to public office in this system. The
rural oligarchic system resembled a “triangle without a base.”5 It was a
system of total and despotic power.6
There was less oppression, but life was still precarious for those liv-
ing in indigenous communities. These communities were based on col-
lective ownership of land, and internal cohesion was structured by
kinship and particularistic networks. Indigenous communities relied
on ancestral norms and communal practices to coordinate the activi-

Soifer_6844-final.indb 81 8/17/18 11:53 AM

 82 Maxwell A. Cameron

ties of their members. Small in scale, their social lives were organized
around face-to-face communication and anchored in generational hi-
erarchies; their collective knowledge was stored in biological memory
and transmitted orally; their connection to the larger division of la-
bor was precarious and sporadic. Yet they were cohesive collectivities,
bound together in tightly knit groups based on reciprocity and mutual
aid, and capable of surviving many external threats, both environmen-
tal and human. There were constant tensions, however, as large land-
lords sought to expand their holdings at the expense of indigenous
communities.
The ultimate collapse of oligarchic domination began in the late
1950s. The expansion of literacy and education in the countryside,
combined with the spread of mass communication networks, espe-
cially radio, diminished the power of mestizos and gamonales and en-
abled challenges to oligarchic domination (Handelman 1975, 55–58;
Heilman 2010, 96–119). Peasant unrest in La Convención near Cuzco
in 1958–1962 focused on the inequities of a system of domination
based on massive haciendas. Peasants began to organize into unions
and to undertake land invasions. Inspired by the example of the Cu-
ban revolution, urban intellectuals joined the struggle (Blanco 1972).
In short order land invasions spread throughout the highlands, involv-
ing hundreds of thousands of peasants. Although an incipient guerrilla
movement was quickly put down, it called attention to the ways in
which Peru’s socioeconomic problems stemmed from the backward-
ness of the rural oligarchy, the lack of national integration, and a weak
and dependent state (Béjar 1970).
Military officers who fought against the guerrillas were able to di-
rectly observe the oppression and misery created by the very order they
were expected to defend; some of them decided it was time for change
(Cleaves and Pease García 1983, 216). One such officer was Juan Ve-
lasco Alvarado, a junior officer who seized power in a coup on Octo-
ber 3, 1968. As the leader of the self-styled revolutionary government
of the armed forces, he embarked on sweeping reforms to break the
domination of the rural oligarchy and lay the foundations for a new
development model based on the inclusion of workers and peasants,
appreciation of indigenous culture, and redistribution of land and in-
come (Lowenthal 1975; Chaplin 1976; McClintock and Lowenthal
1983). “One of the major goals of antioligarchic and nationalist rev-
olution,” according to Julio Cotler (1975, 50), was “the homogeniza-
tion of the social structure, which facilitates the expansion of capital-
ist forms of production.”

Soifer_6844-final.indb 82 8/17/18 11:53 AM

 From Oligarchic Domination to Neoliberal Governance 83

The military introduced agrarian reform, peasant cooperatives, job

stability, industrial communities, educational reform, expansion of so-
cial security, and support for unionization (Alberti, Santistevan, and
Pásara 1977; Balbi 1989; Eckstein 1983; Stephens 1980; McClintock
1981). In state corporatist style, the officers attempted to carry out
reforms while minimizing class confl ict and guaranteeing the unity
of the state (Stepan 1978).7 The opposite was achieved: the reforms
unleashed the kind of social forces that would call into question all
constituted power, especially the military. This transformation en-
abled pent-up demands for change to be expressed in diverse forms
of collective protest and unrest (Mayer 2009). The promotion of
unions and industrial communities led to unprecedented worker mo-
bilization. Support for higher wages and better working conditions in-
creased militancy. One union leader said that if he could get the same
increase in wages for his members by striking or bargaining with the
employer, he would always strike because this taught workers the im-
portance of militancy and solidarity.8 Such strategies encouraged cla-
sismo—a “class-oriented and confrontational political mentality” in
which “struggle” was seen as the most appropriate means by which to
achieve collective goals (Stokes 1991, 87, 91, 97).
The struggle fueled an uneven, incomplete, and precarious pro-
cess of “social democratization” (Lynch 1991, 72). The new pressures
for citizenship and participation were no less significant because they
occurred under military rule. Workers and peasants experienced the
sense of dignity that comes from recognition as fully human agents
capable of collective action and the exercise of political power (Pa-
rodi 1986). To a considerable extent, this process of social democra-
tization would continue in urban areas as waves of rural-urban mi-
grants moved to the cities. As peasants and indigenous peoples moved
to the cities and founded new urban communities in the shantytowns
around Lima, they became pioneers of new settlements; as agents of
change, they carried within them the potential for democratization of
the society.9
Massive general strikes in the late 1970s and regional protest move-
ments, often led by teachers, contributed to a process of radicaliza-
tion of the population. Other middle-sector groups that became more
militant were bank and public-sector employees. The union move-
ment, long under the control of the APRA party, was captured by the
Community Party during the military regime, and the General Con-
federation of Peruvian Workers (CGTP) emerged as a unifying voice
for organized labor. Maoist radicals controlled the teachers union;

Soifer_6844-final.indb 83 8/17/18 11:53 AM

 84 Maxwell A. Cameron

Trotskyists were influential within peasant confederations. These

groups formed the nucleus of an emergent “new Left” (Stephens 1983;
Bernales 1987).
The military decided to extricate itself from government—but, or
so the generals insisted, not from power. Between 1977 and 1980 an
elite-controlled process of transition was initiated. Unlike the situa-
tion in some other Latin American nations that would undertake tran-
sitions to democracy in the 1980s, the Peruvian transition involved
drafting a new constitution.10 The military government decided to
convene a constituent assembly in an effort to manage the transition
and ensure the reforms it had implemented during its decade in office
were ratified (Teivainen 2002, 59; Lynch 1991, 135). To the surprise of
nearly all concerned, the Left captured about one-third of the seats in
the Constituent Assembly (Bernales 1980, 70).11
The result was a remarkably progressive, albeit aspirational, consti-
tution that recognized popular sovereignty as the basis of government,
stressed the dignity of all citizens as equals, acknowledged the impor-
tance of the common good and life within community, and called for
a free, just, and educated society without exploitation in which the
economy would be at the service of the people, not vice versa. The
right to vote was given to illiterates, enfranchising an estimated one
million souls (Lynch 1991, 135). The constitution contained an expan-
sive list of human rights, including social and economic rights like the
right to social security, health, work, and to negotiate collective agree-
ments. Multiple forms of property, private and public, social and in-
dividual, were recognized. The new constitution stated that “the state
could have extensive direct intervention in the economy through pub-
lic enterprises and other means,” wrote political scientist Teivo Teivai-
nen. According to its Article 110, the economic regime was primarily
based on “principles of social justice” (cited in Teivainen 2002, 59).
The constitution captured the spirit of the political transformation
that had begun and which represented a significant step toward the de-
mocratization of Peruvian society. It outlined a highly democratic po-
litical regime that enshrined the separation of powers, a bicameral leg-
islature with a senate to represent the regions, a five-year presidential
term without reelection, and a nondeliberative military. New institu-
tions were created, such as the Tribunal of Constitutional Guarantees
and National Magistrature Council. The Public Ministry and Senate
were redesigned (García Belaunde 1996, 37–38). With this progressive
constitution in place, Peru was able to move to elect a democratic gov-

Soifer_6844-final.indb 84 8/17/18 11:53 AM

 From Oligarchic Domination to Neoliberal Governance 85

ernment in May 1980. That election day would also mark the start of
a confl ict that would ultimately take upward of sixty-nine thousand
lives and inflict extensive material and psychological damage on the
nation.

The Shining Path

One interpretation of Peru’s internal conflict is that it was an unin-

tended consequence of the transformation of Peru that began with the
breakdown of oligarchic domination. The Shining Path exploited di-
visions within Peruvian rural society to construct a powerful collec-
tive identity and mobilize deeply felt resentments against an unjust
order; in so doing, however, it reproduced and exacerbated the injus-
tices of that order. Whereas land reforms are often designed to prevent
peasant rebellions, in Peru they triggered one. The land reform broke
the back of the rural oligarchy, but it also displaced mestizos, ending
their monopolies of knowledge, and creating a vacuum of power. This
vacuum was filled by the Shining Path. The violence instigated by the
Shining Path, in turn, further eroded the division between coast and
highlands as waves of migrants abandoned the countryside for the cit-
ies, especially Lima.
The Shining Path reflected the provincial social order from which it
emerged, even as that order passed into extinction.12 Fire was fought
with fi re in an effort to fundamentally remake the state and society.
Unlike other revolutionary movements in Peru and Latin America, the
Shining Path sought to replace what it called a corrupt, bureaucratic,
corporatist, and fascist state with a new Maoist state. The party would
be the nucleus of an entirely new political, social, and economic order.
This meant that the party had to be everything to its members—a to-
tal organization—and its leadership had to wield absolute power.
The Shining Path has also been described as a project by provin-
cial mestizos to reestablish power in a disrupted social system, as well
as a channel for social mobility and political power for ambitious,
newly educated members of the indigenous population. The leaders
of the Shining Path were heavily drawn from among educated mesti-
zos whose social status was undermined by the agrarian reform (De-
gregori 1997, 182). The displaced middle strata sought new routes to
status and domination through the revolutionary party. Their most
successful recruitment efforts targeted young men and women from

Soifer_6844-final.indb 85 8/17/18 11:53 AM

 86 Maxwell A. Cameron

peasant communities who had achieved a level of education sufficient

for them to expect to occupy greater status and power than they could
by remaining within their families’ communities.
The Shining Path offered not only a recipe for change, but also a
guide to objective Truth. According to this vision, “traditional power,
based not only on the monopoly of the means of production but also,
moreover, on the monopoly of knowledge and its deceptive manipula-
tion, is brought down by the dominated who break both monopolies”
(Degregori 1989, 13). The guide to a new objective Truth would be the
leader, Abimael Guzmán. He sought to construct a new monopoly of
knowledge based on quasi-religious devotion to foundational texts and
authors. Hence the insistence that Guzmán, who held degrees in phi-
losophy and law, was a major intellectual force not only in Peru, but
worldwide, and was comparable in stature to Marx, Lenin, and Mao.
Devotion to the leader was modeled on the Old Testament-style con-
struction of a people of the book. Pensamiento Gonzalo, as Guzmán’s
thought was called, was a vehicle for indoctrinating recruits and forg-
ing them into instruments of revolution. Under the guidance of a van-
guard steeped in knowledge of supposedly profound and inaccessible
texts, they were asigned a sacred mission.13 With the grandiloquence
of the parochial intelligentsia, Guzmán proclaimed Peru the epicenter
of global revolution.

Democratization under Stress

Writing in the 1980s, anthropologist José Matos Mar (1985) argued

that “official Peru” had lost its monopoly on power and could no lon-
ger exclude and marginalize the Andean majority (or what he called
the “marginal Peru”). “Popular overflow” (desborde popular) signaled
the end of despotic oligarchic power: Peru’s popular sectors could no
longer be ignored or marginalized by the nation’s social, economic,
and political institutions.14 The revolution toward which the Shining
Path sought to direct the popular overflow was avoided only because
there was a democratic alternative. As precarious as Peru’s democracy
seemed at the time, it was the decisive obstacle to the Shining Path’s
designs, as was implicitly recognized by that organization’s leader-
ship when they initiated the armed struggle by burning ballot boxes
on election day in May 1980. The resiliency of democracy was not
fully appreciated by the Peruvian Left, which was divided between the

Soifer_6844-final.indb 86 8/17/18 11:53 AM

 From Oligarchic Domination to Neoliberal Governance 87

kinds of commitments needed to participate in the democratic process

and the desire to build a mass movement with the potential option of
seizing power by revolutionary means.
The various parties that created the United Left were attracted by
the prospect of gaining political power by electoral means, but they
were not ready to abandon mass politics outside democratic institu-
tions. Their most promising democratic leader, Alfonso Barrantes,
won the mayoral race in Lima in 1984, positioning him to run for
president in 1985 and 1990. For others, mass struggle included the
possibility, eventually, of armed struggle. The suggestion that Peru in
the mid-1980s was in a prerevolutionary phase did not seem wildly
implausible. For some activists, Shining Path militants were showing
that revolutionary struggle was still possible, while the United Left
was caught up in the decidedly unrevolutionary politics of electioneer-
ing. “If the Shining Path had not existed,” speculates Gustavo Gorriti
(1999, 11), “the left’s incorporation into the system would have been
more visible and complete, and today the Marxist left would be a pil-
lar of democratic stability, part of a lively and vibrant political pro-
cess, and an entirely peaceful one.”
The other stress on democracy was economic crisis. Although the
center-right developmentalist Popular Action party won the 1980 elec-
tion under the leadership of Fernando Belaunde Terry (1980–1985),
it quickly became apparent that the new democratic government was
not up to the twin challenges of insurgency and economic crisis. Mon-
etarist policies adopted by industrialized nations in the early 1980s in-
creased Peru’s debt burden, while competition among exporters low-
ered the prices of commodities that were the primary source of hard
currency. Peru was forced to renegotiate its international debt obli-
gations just as the country emerged from military rule. President Be-
launde implemented orthodox economic policies that placed the bur-
den of adjustment on Peru, particularly the urban poor, many of whom
were forced into the informal economy in unprecedented numbers. He
left the counterinsurgency strategy to the military and seemed un-
aware of the extent of the violence and repression in Ayacucho. By the
end of Belaunde’s term there was a pervasive sense of ungovernability
and drift. Drastic measures to restore economic growth and govern-
ability seemed in order.
The 1985 election brought APRA to power, and for about eigh-
teen months there was a renewed sense of optimism among Peruvians.
The youthful president, Alan García Pérez, promised to place a ceiling

Soifer_6844-final.indb 87 8/17/18 11:53 AM

 88 Maxwell A. Cameron

on debt payments and stimulate growth through heterodox or popu-

list economic policies: deficit spending, price controls, and wage in-
creases. For the Right, these policies did not seem too different from
what the Left had to offer, and so the term “Apro-Communism” reen-
tered the political vocabulary after many dormant years, despite the
fact that there was certainly no alliance between APRA and the IU.
After an initial spurt of economic growth, heterodox policies gener-
ated the predictable evils of inflation, capital flight, and shortages, to
which García responded, in a fit of pique, by attempting to national-
ize the banks. Armored vehicles broke down the doors of the Banco de
Crédito and entered in a cloud of tear gas.15 Conservative sectors ral-
lied behind novelist Mario Vargas Llosa and his movement, Libertad.
Vargas Llosa teamed up with the author of The Other Path, Hernando
de Soto, to propose a radical reorientation of Peru’s economy, soci-
ety, and politics. It was the inception of Peruvian neoliberalism (see
M. Vargas Llosa 1991; A. Vargas Llosa 1991; A. Vargas Llosa 1994).
De Soto spent his formative years in Switzerland, Canada, and the
United States. He studied in the Graduate Institute of International
Studies in Geneva, which had recruited Ludwig von Mises and Wil-
helm Röpke, founding members of the Mont Pèlerin Society. Upon
his return to Peru, de Soto invited Friedrich von Hayek to a meeting
in Lima and subsequently founded the Instituto Libertad y Democ-
racia to help create an intellectual climate favorable to neoliberalism
(Mitchell 2009, 396). De Soto’s fi rst contribution to public policy was
working under the García government in a project for property rights
legislation and administrative reform. In 1989 he published The Other
Path, an ambitious effort to interpret the migration of indigenous Pe-
ruvians to the cities and their survival strategies as reactions to a state
that had failed to provide property rights to the poor. De Soto explic-
itly linked governance and markets. Either Peru would carry out a cap-
italist revolution driven by the emerging entrepreneurs of the informal
economy, or there would be a violent revolution (de Soto 1989, 233).16
The fi nal years of the García administration were so chaotic and de-
structive—hyperinflation reached levels comparable to Weimar Ger-
many, and a campaign of car bombs hit Lima as Shining Path activists
expanded operations into the capital city—that a growing number of
Peruvians began to accept that major sacrifices would be necessary to
restore economic growth and political order. Although Vargas Llosa
lost the 1990 election to Alberto Fujimori, his ideas were adopted.
Vargas Llosa even offered his program and his team of neoliberal

Soifer_6844-final.indb 88 8/17/18 11:53 AM

 From Oligarchic Domination to Neoliberal Governance 89

technocrats to the newly elected president. De Soto quickly made the

transition and began to work with Fujimori. He was instrumental in
providing international connections that would consolidate the new
president’s commitment to reforms.
Fujimori may have cast himself as a candidate who would not im-
plement shock therapy, but once in office he was quickly persuaded
that economic austerity and aggressive counterinsurgency measures
were unavoidable. A hastily arranged trip to Washington and Tokyo
convinced the president-elect that he would get little international sup-
port unless he adopted what had come to be known as the “Washing-
ton Consensus”: privatization, liberalization of trade and investment,
deregulation, labor market flexibility, monetary conservatism, fiscal
restraint, and, above all, limiting the role of the state in the economy.
This was tough medicine to swallow for a country already in a virtual
state of economic collapse, but, as is so often the case, the depth of the
crisis created an opportunity to implement thoroughgoing reforms.
The perception—ill-informed, perhaps, but widespread—that the
Shining Path was at the cusp of a strategic equilibrium with the armed
forces created the conditions for an interruption of the democratic or-
der. Important sectors within the armed forces concluded that Peru
needed a period of prolonged political stability that could only be
achieved by hard-line (mano dura) measures. Although they initially
plotted to topple Fujimori, their scheme (the plan verde,17 as it was
known) fell into the hands of Fujimori’s security adviser, Vladimiro
Montesinos, who then adapted the strategy to build a civil-military co-
alition that would support the April 1992 presidential autogolpe, in
which the Congress was closed, the courts purged, and the constitu-
tion suspended.
The autogolpe was a moment of institutional dissolution. It is far
from clear that this draconian measure was necessary to arrest the
deteriorating security situation. The Congress appeared to be quite
prepared to concede extraordinary powers to the president, and the
counterinsurgency strategy was beginning to work. Within a few
months a meticulous police investigation that had started before the
autogolpe would result in the capture of Guzmán in a safehouse in
Lima, but that only seemed to vindicate Fujimori’s mano dura. More-
over, although it was less clear at the time, Fujimori and Montesinos
had implicated themselves and others within the armed forces in hu-
man rights crimes.
Whatever the logic of power that led to the autogolpe decision, the

Soifer_6844-final.indb 89 8/17/18 11:53 AM

 90 Maxwell A. Cameron

public was clearly prepared to support it, as Fujimori intuited from the
way in which his attacks on Congress and the political establishment
played in public opinion. Following the autogolpe, opinion polls showed
Fujimori had most Peruvians behind his authoritarian measures. The
support was neither capricious nor unreasonable; it was rooted in pro-
found and widespread fear (Schulte-Bockholt 2013, 97). The fear was
not limited to the business community and the armed forces. It ex-
tended to virtually the entire middle class as well as working- class and
poor rural communities, which felt the brunt of the confl ict. The per-
vasive sense of fear paralyzed social movements and made it hard to ar-
ticulate opposition to Fujimori (Burt 2009b, 315–349).
Of course, there was opposition from lawmakers, lawyers, judges,
and constitutional law experts, as well as other defenders of the rule of
law, and this opposition would wind up expressing itself in extremely
important ways as legal professionals within the courts and in lawyers’
guilds challenged the erosion of constitutional protections throughout
Fujimori’s tenure in power. But their defense of the rule of law had few
echoes among the public. Privileged groups like the business commu-
nity and technocrats were willing to surrender legal guarantees in re-
turn for order and stability; underprivileged groups were never pro-
tected by the rule of law in the fi rst place.
Tepid opposition came from the international community, particu-
larly the Organization of American States (OAS). Such opposition was
mollified when Fujimori traveled to a General Assembly of the OAS in
the Bahamas and, at the advice of Hernando de Soto, agreed to con-
vene a “pretentiously—and redundantly—named Democratic Constit-
uent Congress” (CCD) to rewrite Peru’s constitution and call new elec-
tions in 1995 (García Belaunde 1996, 39). Elected on November 22,
1992, the CCD did not have any members from the parties Acción
Popular, APRA, and Libertad, all of which abstained. The CCD met
from January to September 1993 and produced a text that was sub-
mitted for a referendum on October 31, 1993. Fifty-two percent of
the public gave it their approbation, but one-third of the electorate ab-
stained and 9 percent cast blank or void ballots.18
Peru was, of course, facing an institutional crisis, but “the consti-
tution had little or nothing to do with this crisis” (García Belaunde
1996, 39). Fujimori did not announce the need for a new constitution
on April 5, 1992. The idea of changing the constitution came about
in response to criticism of the autogolpe—it was not part of Fujimo-
ri’s initial intention, which was to govern by plebiscitarian means and

Soifer_6844-final.indb 90 8/17/18 11:53 AM

 From Oligarchic Domination to Neoliberal Governance 91

for longer than a single term. Rewriting the constitution was a fallback
position to salvage a course of action that faced unexpected obstacles.
But Fujimori must have felt emboldened by the outpouring of popular
support for his hardline measures. He quickly realized that he could
rewrite the constitution to his advantage. He approached the task not
in a spirit of constitutionalism, however, so much as a desire to perpet-
uate himself in power.

Comparing the 1979 and 1993 Constitutions

Despite its more centralist and neoliberal character, the constitution of

1993 was otherwise quite similar to its predecessor. There were those,
like the members of the Andean Commission of Jurists19 who argued
that a change in the constitution was not really necessary. Instead, the
CCD should have simply modified the 1979 constitution. “However,”
as Domingo García Belaunde (1996, 40) observed, “the regime’s ‘ju-
rists’ warned that it would be dangerous to leave the 1979 Consti-
tution in force, even with important reforms. Article 307 drastically
sanctioned all authors of coups. This was the Sword of Damocles that
could be used in the future against the current government.”20
Fundamental questions for any constitution include: who is the au-
thor and with what right does that author claim the power to create
a new order? (Bernal 2017; Cameron 2013). The framers of the 1979
constitution described themselves as “We, representatives to the Con-
stituent Assembly,” who, exercising the “sovereign power of the peo-
ple” (“ejercicio de la potestad soberana del pueblo”) that the people
“have conferred upon us,” promulgate the constitution. The language
is active and inclusive. The appeal to popular sovereignty places the
constitution within the Enlightenment tradition of the French and
American Revolutions.
The 1993 constitution made no reference to popular sovereignty.
Shifting to the third person rather than the fi rst-person plural, it says
that the “Democratic Constituent Congress, obeying the mandate of
the Peruvian people” (“obediciendo el mandato del pueblo peruano”)
has “resolved to give the following constitution” (“ha resuelto dar la
siguiente constitución”). The term “representative” is not used. The
image conjures a delegation of power to a third party which returns to
the people in the form of a gift.
The omission of any reference to popular sovereignty in the 1993

Soifer_6844-final.indb 91 8/17/18 11:53 AM

 92 Maxwell A. Cameron

constitution must have been deliberate. The framers of the 1993 con-
stitution feared popular sovereignty and sought to order society from
above. They also dropped the references to the integration of Latin
American peoples and independence from imperialism, which in the
Latin American context are also important features of popular sov-
ereignty, as well as to historic and revolutionary leaders like Túpac
Amaru and Simón Bolívar. The 1993 constitution created a unicam-
eral legislature of 120 members with a single electoral district, which
can be dissolved by the executive if he or she lacks confidence in two
consecutive cabinets. Participatory innovations were adopted, includ-
ing the use of referenda and recall (García Belaunde 1996, 42).
The neoliberal cast of the 1993 constitution has been widely noted
(Rubio Correa 2012; Teivainen 2002). García Belaunde (1996, 43)
wrote that “the state practically disappears from the economic sphere
taking on a modest subsidiary role.” Whereas the 1979 constitution
described the state in interventionist terms, the constitution of 1993
sought to minimize state involvement in the economy and give space
to private enterprise. Article 60 stated that “the State may subsidiarily
engage in business activities, directly or indirectly, for reasons of high
public interest or manifest national convenience,” only when autho-
rized by law (emphasis mine). This norm, according to Alberto Ver-
gara and Daniel Encinas (2016, 162) “is the cornerstone of the neolib-
eral citizenship regime and its derived policies that have progressively
taken root in Peru.” The constitution also extended national treatment
to foreign enterprises and reinforced private property rights. Article 70
made the right to property inviolable, stating that there can be no ex-
propriation unless it is justified by national security or public neces-
sity, and only then with full compensation. “Social interest” or “public
utility” were no longer sufficient reasons, as in the 1979 constitution.
Article 58 enshrined free enterprise. The role of the state is defined to
prevent monopolies and ensure competition. Public enterprises are not
to be given special treatment relative to the private sector, and the state
should only play an entrepreneurial role in cases of “high public inter-
est or manifest national convenience.” The right to strike is regulated
“so that it will be exercised in harmony with the social interest,” ac-
cording to Teivo Teivainen (2002, 157, 158).
Certain contract laws (involving agreements between the state and
investor) cannot be modified subsequently by legislation. Otherwise,
conflicts over contracts are to be resolved in the courts. Article 79,
which states that the Congress cannot increase spending, deprives the
legislature of one of its biggest levers—the power of the purse. Arti-

Soifer_6844-final.indb 92 8/17/18 11:53 AM

 From Oligarchic Domination to Neoliberal Governance 93

cle 89 deals with the agrarian regime, establishing that native lands
are imprescriptible, meaning they cannot be acquired through posses-
sion due to occupancy or other requisites. However, the constitution
does not say these lands are unembargable or inalienable, as in the
previous constitution, which means they can be bought and sold or
foreclosed. This was a key reform to enable the breakup of peasant co-
operatives in favor of parceleros (small private ownership of land).
Perhaps the most authoritarian features of the 1993 constitution are
its emergency provisions: under chapter VII, constitutionally guaran-
teed rights can be suspended during a regime of exception. In these pe-
riods the executive has extraordinary powers. Marcial Rubio Correa
(2012, 224) refers to this as the constitutionalization of temporary dic-
tatorship. However, after forty-five days the suspension of constitu-
tional guarantees must be approved by Congress. Under Article 200,
no judge can challenge the decision to impose a state of exception. The
constitution of 1993 also recognizes the rondas campesinas (peasant
self-defense organizations).
No sooner was the constitution adopted than Fujimori began to
seek ways around it. His style of leadership—chaotic, delinquent, and
unpredictable—was incompatible with basic principles of constitution-
alism. For example, he insisted that he had the right to run for three
terms in office. 21 Accordingly, shortly after Fujimori’s fi rst reelection,
Congress introduced the so-called Law of Authentic Interpretation of
the Constitution, which stated that the president was eligible for an-
other term since he had only been elected once under the new consti-
tution. 22 Since the law had the clear intent of benefiting a single indi-
vidual, it violated the principle of generality, as well as the hierarchy
of laws, by imposing a particular interpretation of the constitution by
means of ordinary legislation. The Lima Bar Association challenged
the constitutionality of the law in the Constitutional Tribunal that
had been created by the 1993 constitution. The government passed an
“organic law” requiring an extraordinary majority of six out of seven
votes in order to declare a law unconstitutional. This meant only two
votes were necessary to veto any decision, and two members of the
Tribunal had close ties with the intelligence service. They upheld the
law while the rest rejected it. When a majority declared the law inap-
plicable to Fujimori, the majority in Congress fi red the members who
had ruled against reelection, leaving the Constitutional Tribunal inop-
erative for the remainder of Fujimori’s term and opening the way for
his unconstitutional attempt to run for a third term.
And yet, despite Fujimori’s contempt for his own constitution, the

Soifer_6844-final.indb 93 8/17/18 11:53 AM

 94 Maxwell A. Cameron

nondemocratic manner in which it was contrived, the absence of a

compelling necessity to rewrite the entire constitution, and, above all,
the fact that the constitution was conceived and implemented in the
middle of an emergency situation, the 1993 constitution stands as a
remarkably enduring document. It survived Fujimori and the Shining
Path. The process of social democratization in Peru since the 1950s
was to some extent irreversible. But the neoliberal tenets of the con-
stitution have become irreversible as well. The 1993 constitution es-
tablished a clear neoliberal policy order that has remained untouched,
even as reforms have modified many of its important political features.

Neoliberal Governance

The pro-market policies adopted by Fujimori, which remained intact

after he fled the country in 2000, fundamentally reordered Peruvian
politics by linking political and economic stability to the success of
market-led development. Fujimori would later return to face trial and
imprisonment, but his constitution survived intact. Four successive
elected governments upheld the economic model it enshrined. In part,
the continuity reflected a fear of returning to the past. Peru became
more like Chile, which, after the Pinochet dictatorship from 1973 to
1990, elected left-wing governments (under presidents Ricardo Lagos
and Michelle Bachelet) that refrained from tampering with the pro-
market economic model of the military regime, although they modi-
fied political features of that regime and introduced compensatory so-
cial programs. Chile did not experience political violence on the same
scale as the internal confl ict in Peru, but the 1973 coup was a trau-
matic event without precedent in Chile’s history, and the fear it gen-
erated discouraged governments from tampering with the economic
model after the transition from authoritarian rule. The constitutions
of Peru and Chile were adopted by referenda under authoritarian gov-
ernments in circumstances that were neither revolutionary nor demo-
cratic, yet both have endured.
Neoliberalism fundamentally reconstructed the Peruvian body pol-
itic. By neoliberalism I mean not merely a list of policies (for example,
the “Washington Consensus”) designed to promote competitiveness
and growth, but more basically the rules and incentives that encour-
age competition throughout society (Foucault 2004; Brown 2015; Dri-
not 2014, 171). The goal of neoliberalism has never been simply to

Soifer_6844-final.indb 94 8/17/18 11:53 AM

 From Oligarchic Domination to Neoliberal Governance 95

change policies and institutions, but to do so in order to shape the be-

havior of individuals and organizations. Although the purpose of get-
ting incentives (and policies, or institutions) “right” is often framed in
terms of the need for growth and competitiveness, these goals are nor-
matively desirable because they enable public officials to employ tech-
niques of governance grounded in economic rationality. The neoliberal
project in Peru as elsewhere has been about the constitution of a social
order around a specific set of freedoms—above all, those associated
with property, competition, and entrepreneurship. It is not, however, a
constitutional project that involves any notion of popular sovereignty.
Neoliberalism is designed to be a guarantee against collectivist efforts
at social engineering and forms of rationality that transcend the indi-
vidual’s pursuit of his or her own ends. A threat to individual liberty
and rationality was, above all, what the Shining Path represented to
neoliberals. In few countries around the world can the link between
neoliberalism and governance be seen more clearly.
Teivo Teivainen sees what he calls “economism” as a defi ning fea-
ture of the 1993 constitution. “The neutrality and autonomy of the
economic sphere vis-à-vis the political organs of the state was in var-
ious ways enshrined in the constitution of 1993,” Teivanen writes,
“which made it a clear example of the constitutional politics of econ-
omism, one of the central elements of which is the attempt to make an
economistic social order persist through the creation of constitutional
constraints. The new constitution had various such constraints de-
signed to ‘bind the future’ so that neo-liberal policies could not be eas-
ily reversed” (Teivainen 2002, 157). Teivainen draws upon the work
of Gramscian scholar Stephen Gill to argue that in the “new consti-
tutionalism”23 associated with neoliberalism, “the scope of democ-
racy is restricted by defi ning various governance institutions and the
issues they deal with as ‘economic’ and using the doctrine of economic
neutrality to produce a dichotomy between the economic and political
spheres” (Teivainen 2002, 17).
Building on work by Michel Foucault (2004; see also Brown 2015),
I argue that the “economism” of neoliberal governance is accompanied
by a political dimension. 24 And it is not merely destructive but also cre-
ative. Neoliberalism was not just a set of policy prescriptions, or even
prohibitions, but rather a technique of governance based on the use of
rules and incentives to promote maximizing competitive utility in all
spheres of life (Foucault 2004, 118–121). These measures also serve to
discourage collective action and collectivist identities, and they may be

Soifer_6844-final.indb 95 8/17/18 11:53 AM

 96 Maxwell A. Cameron

more important in this respect than any legal or constitutional limita-

tions imposed by the constitution. By fostering competition and entre-
preneurship, they encourage the active pursuit of growth, investment,
and monetary gain. It is important, however, to make a distinction
that may not have been sufficiently obvious to Foucault: neoliberalism,
operating through a micropolitics of competition, is entirely compat-
ible with oligopolies, cartels, price fi xing, predatory practices, lobby-
ing, influence peddling, and other noncompetitive arrangements within
and among major corporations. These arrangements, which are any-
thing but competitive, give Peru’s electoral democracy oligarchic fea-
tures. 25 A small handful of major corporate groups dominate much of
the productive and fi nancial activity in the country, and these groups
exert enormous lobbying and direct influence over policy as well as the
market (F. Durand 2004; Crabtree and Durand 2017). These de facto
corporate powers enjoy what I call negative power: power that rests
on the purposeful destruction of those forms of infrastructural power
that might enable collective action threatening property, market com-
petition, and corporate power.
Negative power is the opposite of the kind of power that mobilizes
society’s resources for collective purposes: it is the power to dissolve,
to obstruct, to discourage, to exclude, to undo or not do things. This is
power that primarily operates through the micro-level effects of com-
petition, but it also comes in the form of direct (but rarely transpar-
ent) interventions in politics in which major corporations exercise their
veto over almost any area of public policy that affects their interests.
It restricts elected officials’ scope for policy choice. The 1993 consti-
tution guarantees that even if an elected official promises to do some-
thing or has a mandate to implement a policy, it does not happen un-
less it is consistent with the logic of the market. It discourages elected
officials who may wish to deviate from neoliberalism, and it guaran-
tees freedom in the marketplace, market-led growth, and macroeco-
nomic stability. This freedom is purchased at the expense of collec-
tive action or public policy initiatives that would achieve ends beyond
what is permissible within a neoliberal model of development.
In short, neoliberal governance presents itself as a solution to both
economic backwardness and collectivist threats to individual and en-
trepreneurial liberty, and it does so by promoting competition and in-
centives to get ahead at the micro level while preserving the macro-
level corporate power of oligopolies. The neoliberal state is not the
night watchman state of classical liberalism—a state that regulates

Soifer_6844-final.indb 96 8/17/18 11:53 AM

 From Oligarchic Domination to Neoliberal Governance 97

competition and ensures that people freely express their natural incli-
nation to barter and trade—but a state that actively establishes com-
petitive economic rationality through demobilization, deregulation,
depoliticization, privatization, surveillance, targeting, and thereby fos-
ters a culture of entrepreneurship and consumerism.
Demobilization, or breaking up nonmarket “guilds and combines”
in de Soto’s (1989, 208) terms, is achieved by promoting flexibility and
the deregulation of certain spheres of economic activity, particularly
in relation to the labor market. The prohibition against guilds and
combines does not apply to major corporations. Deregulation means
“increasing the responsibilities and opportunities of private individ-
uals and reducing those of the state.” The objective is to “depoliti-
cize the economy in order to protect the state from the manipulation
of redistributive combines” (1989, 249). The state should focus on en-
forcing efficient rules rather than managing production or allocating
resources.
Demobilization and the deregulation of collective bargaining were
intended to undermine union activity; to reduce the density of union-
ization; and to discourage the use of strikes and other forms of collec-
tive bargaining and struggle by making such action useless. The Min-
istry of Labor stopped facilitating collective bargaining or upholding
workers’ rights, and began to do just the opposite—it worked to pro-
mote a flexible workforce and a labor market with minimal regula-
tions and safeguards. Private service contracts proliferated at the ex-
pense of stable work. In the rural areas, efforts were made to continue
to promote the parceling of land and the breakup of peasant cooper-
atives. The ongoing power of the teachers’ union was blamed for the
slow progress of educational reform.
Privatization was vigorously pursued in not only the sphere of pro-
duction but also a wide range of critical social services like education,
health care, and pensions. Privatization has been advanced somewhat
by stealth. Rather than denying access to free public education, suc-
cessive governments have instead failed to make improvements and al-
lowed the educational system to fall into such a state of disrepair that
private schools have proliferated and become the norm even among
the poor. The same is true of the health care system, where private
pension plans, following the Chilean model, have been introduced.
Even coercive authority was deregulated (de Soto 1989, 251). Polic-
ing has become reliant on para-police organizations, such as Seranaz-
gos in certain neighborhoods of Lima, as well as the use of informal

Soifer_6844-final.indb 97 8/17/18 11:53 AM

 98 Maxwell A. Cameron

urban police (or “cops for hire”), gangs deployed for protection, re-
servists, and private security forces (Schulte-Bockholt 2013, 60–66).
At the same time, the media whips up a climate of fear through a gro-
tesque focus on violent criminal activity. Spying and monitoring have
become normalized by the surveillance functions of the state. Succes-
sive governments have exhibited intense hostility toward nongovern-
mental organizations and have criminalized the activities of protest
movements.
Public policy making has focused on targeted spending rather than
policies aimed at achieving universality. Government spending has fo-
cused heavily on infrastructure projects, which have helped link rural
communities, facilitate commercialization of agricultural goods, and
promote internal trade. Social programs have been financed through
special funds (with organizations like FONCODES devoted to the dis-
bursements of such funds). As a result, a patchwork quilt of channels
and mechanisms earmarks particular resources for specific communi-
ties and purposes. The use of special funds to channel resources is jus-
tified by the need to avoid redistributive pressures from legislative ap-
propriations (Barrantes 2009).
A cultural shift is apparent in the rise of entrepreneurship and ram-
pant consumerism. De Soto’s glorification of the small entrepreneur
is widely accepted in mainstream media. The clasismo of the 1970–
1980 period was replaced with a new ethic of getting ahead: todo se
consigue por la lucha (“everything is achieved through struggle”) was
replaced with hay que competir para ganar (“one must compete to
win”). A substantial market for self-help books for budding entrepre-
neurs has arisen, and the media celebrates the success stories of pro-
vincial entrepreneurs and merchants. A whole section of El Comercio
is devoted to emprendedores. 26 The pioneering TV program Promo-
viendo, hosted by Guido Pennano, the minister of industry under Fu-
jimori in 1990–1991 before he was imprisoned for fraud, is devoted to
success stories of small-scale entrepreneurship.27 Business gurus like
Nano Guerra García provide models for budding capitalists. 28 Yet de-
spite the spread of emprendedurismo, there are no collective associ-
ations of emprendedores. 29 Consumerism forms part of this culture
shift. Alberto Vergara (2015b) uses the phrase compra y calla (“buy
and be quiet,” a play on the familiar exhortation of parents to children
to come y calla, or “eat and be quiet”) to capture the mix of conspicu-
ous consumption and political paternalism that characterizes contem-
porary Peru.

Soifer_6844-final.indb 98 8/17/18 11:53 AM

 From Oligarchic Domination to Neoliberal Governance 99

Neoliberalism as a technique of governance has advantages: it is

easier to govern a nation of aspiring entrepreneurs than one of class
warriors. Yet neoliberalism has been beset by internal contradictions.
Despite very high rates of growth—due in large measure to favorable
international commodity prices and foreign investment in extractive
industries—Peru continues to be a mediocre performer in interna-
tional rankings of competitiveness. In part, this is because competitive-
ness requires robust institutions and substantial public investments; it
also requires planning and coordination, and it assumes investments
in the well-being and productive capacity of the working population.
And yet Peru has lagged in the implementation of these sorts of institu-
tional reforms. The bad externalities that proliferate—disease, pollu-
tion, congestion, violent crime—affect the poor disproportionately. In
the roulette of market opportunities, the dice are loaded against small
entrepreneurs and workers.
Peru’s state has underperformed in terms of redistribution. This is
not to deny successes in the area of poverty reduction; indeed, it is
consistent with the strictures of neoliberalism to spend money on pov-
erty alleviation programs. De Soto (1989, 251) wrote that “to redis-
tribute to the poorest and least fortunate members of the population”
is an important function of the state, provided it does not discourage
production or distort markets. Yet the various programs devoted to
poverty alleviation in Peru have been modest in impact compared with
similar programs in other countries.30 Part of the problem appears to
be a technocratic orientation in the implementation of the programs,
which inhibits the scale and speed of their delivery (a point to which I
return below).
The Peruvian state has also underperformed in the area of the pre-
distribution (as opposed to the redistribution) of resources. The canon
minero is a program for redistributing royalties from mining opera-
tions that has the unintended effect of generating confl ict. Regional
and local governments compete for resources disbursed by the central
government through the powerful Ministry of Finance. There are hun-
dreds of flashpoints of conflict throughout Peru. Extractive industries
tend to be located precisely in poor rural areas where governance is
most precarious. These industries generate enormous wealth, but they
impose huge social, economic, and environmental costs on impover-
ished regions. The canon minero system exacerbates these tensions.
Public services have been allowed to deteriorate despite macroeco-
nomic growth and stability. Public education, especially in rural areas,

Soifer_6844-final.indb 99 8/17/18 11:53 AM

 100 Maxwell A. Cameron

is of extremely poor quality. In Lima, half of all school-aged children

go to inexpensive and poor-quality private schools because the public
schools are so bad. There have been improvements in health, as well as
public infrastructure, but the contrast between public and private ser-
vices remains like night and day. Public insecurity caused by everyday
crime and violence is another persistent social ill.
Despite the many limitations of neoliberalism in Peru, changing this
policy orientation and model of governance has thus far proved to be
impossible. On the one hand, opposition to neoliberalism has been un-
dermined by the policies of demobilization, deregulation, depolitici-
zation, privatization, surveillance, social targeting, and the culture of
entrepreneurship and consumerism. On the other hand, legacies of vi-
olence and conflict have made it particularly difficult to develop col-
lective capacity for popular mobilization, including the construction of
political parties and movements of the Left. In the absence of collec-
tive pressures for change, the urgency to undertake comprehensive re-
forms to address Peru’s social deficits is lost, and the capacity to hold
governments accountable for their failures is minimized.
This helps us to understand—and perhaps begin to resolve—the de-
bate over neoliberal policy continuity. The debate has been framed as
a disagreement between the state capture thesis, associated most re-
cently with the work of Crabtree and Durand (2017), and the thesis,
articulated by Vergara and Encinas (2016) and Dargent (2015), that
technocrats and bureaucrats enjoy extraordinary power and autonomy
within the Peruvian state. I argue that the highly concentrated eco-
nomic power of the business elite in Peru allows it to exercise inor-
dinate political power over policy-makers operating in a highly con-
strained neoliberal state, but this power is fundamentally negative and
exclusionary; it entails the power to exercise a veto over public policies
that might threaten the core interests of the organized business com-
munity. The neoliberal state operates in accordance with economic ra-
tionality, but the high degree of autonomy of technocrats is possible
because of the weakness of the Left and social movements. Vergara
and Encinas (2016, 163, 170) implicitly acknowledge this when they
suggest that the weakness of the political class in relation to techno-
crats is due to the lack of organized political parties and the precari-
ousness of their connections to civil society, as well as the “fragmented
and uncoordinated” nature of civil society itself. Such weakness, they
note, “undeniably contributes to the general weakness of the political
class and reform attempts.” In other words, neither the weakness of

Soifer_6844-final.indb 100 8/17/18 11:53 AM

 From Oligarchic Domination to Neoliberal Governance 101

the state vis-à-vis the business establishment or the relative autonomy

of technocrats vis-à-vis politicians can be treated in isolation from the
devastating effects of both the internal confl ict and neoliberal restruc-
turing on Peru’s party system and civil society. This condition is neces-
sary to explain the continuity of neoliberal governance and Peru’s ab-
stention from Latin America’s left turns.

Peru and Latin America’s Left Turns

At the dawn of the twenty-fi rst century, many Latin American de-
mocracies made “left turns” (Cameron and Hershberg 2010; Levitsky
and Roberts 2011). This was an unexpected reversal of two decades
of deepening neoliberalism. The end of the Cold War had seemed
to promise an end of history and politics. Latin America’s left turns
showed, however, that politics and history never really ended; they
were part of a larger and more diffuse movement of opinion against
neoliberalism born of globalization, financial and monetary crises,
and rising inequality (Rosanvallon 2006, 147–159). Latin America’s
left turns did not constitute a repudiation of liberalism so much as a
desire to overcome its insufficiencies in the Latin American context—
whether the lack of human development, the neglect of public insti-
tutions, or the absence of effective legal guarantees.31 The alternative
was not illiberalism but post-liberalism.32 Given a history of radical-
ism and political instability, it seemed odd that when the tide turned,
Peru missed the current.33 The puzzle can be explained by comparing
Peru with its neighbors and by analyzing recent elected governments.
The imperviousness to the regional leftist trend seemed particularly
curious because Bolivia and Ecuador, the two Andean nations most
similar to Peru in terms of inequality, long-standing histories of so-
cial exclusion, and persistent political instability, both elected (and re-
elected) left-wing governments. They also achieved high levels of eco-
nomic performance and improvements in social indicators. Bolivia is a
far poorer country than Peru, but it invested more than twice as much
in education and now has rates of educational attainment higher than
Peru. It also invested more in health care. Ecuador also invested sub-
stantially more in health care and education. 34 In both countries, pow-
erful social movements prepared the ground for left turns.
Internal conflict appears to be a powerful deterrent to left turns.
Bolivia today faces many of the same challenges as Peru, but the na-

Soifer_6844-final.indb 101 8/17/18 11:53 AM

 102 Maxwell A. Cameron

tional revolution of 1952 gave it an earlier start on a reformist project

that has many similarities to the current reforms under its president
Evo Morales. The revolution, while not exactly a peaceful affair, was
not a major bloodbath either, and Bolivia did not experience heavy-
handed repression in the subsequent decades on the same scale as Peru.
Historically, the Bolivian military has been weak and corrupt. As a re-
sult, powerful social movements have emerged that were unafraid to
push their case to the point of winning power by electoral means. “In
Peru,” as Eduardo Silva (2009, 231) notes, “significant insurrectionary
movements and a turn to authoritarianism that closed political space
during Fujimori’s presidency inhibited the formation of associational
power and horizontal linkages across social movement organizations.”
Colombia is the only other country in Latin America that has also
been forced to grapple with a comparably brutal insurgent movement:
the Fuerzas Armadas Revolucionarias de Colombia, or FARC. The Cen-
tro Nacional de Memoria Histórica estimates that nearly 220,000 peo-
ple were killed in Colombia between 1958 and 2012.35 So powerful
was the rejection of the Left that the Colombian government, after
painstaking negotiations with the FARC, was unable to secure the fa-
vor of a majority of voters in an October 2016 peace plebiscite. Like
Peru, Colombia did not undergo a left turn. This is not to say that all
internal conflicts militate against the left turns. The Farabundo Marti
National Liberation Front, which led a broadly supported popular in-
surrection that fought the Salvadoran government to a standstill before
negotiating a peace accord, committed relatively few human rights vi-
olations. As a democratic political party, it successfully ran candidates
in municipal elections across El Salvador, fought for a share of legis-
lative seats, and ultimately won presidential elections in 2009. It was
the brutality of the Shining Path and the FARC that damaged the pros-
pects for the legal Left in those cases, as well as the ongoing repression
of leftist activists that these internal confl icts generated.
Turning to the experience of recent governments, of the four presi-
dents elected following the Fujimori decenio, two were neoliberal tech-
nocrats, and two were politicians (one of whom was an outsider). Ale-
jandro Toledo rode to power on the wave of protests that contributed
to the collapse of the second Fujimori government, but he had neither
the inclination nor the social base necessary to alter the economic de-
velopment model enshrined in the 1993 constitution. A rupture with
the neoliberal model was clearly on the political agenda in 2006 with
the campaign by Ollanta Humala and his insurgent ethno-nationalist

Soifer_6844-final.indb 102 8/17/18 11:53 AM

 From Oligarchic Domination to Neoliberal Governance 103

movement. It is extremely significant that Humala was a former mili-

tary officer who fought against the Shining Path in Tingo María, Huá-
nuco. Thus, even though he ran on a program of progressive reforms
called “La gran transformación,” nobody could accuse him of being
tainted by subversion. If anything, there were serious questions about
whether he had participated in human rights crimes.
Humala narrowly lost that election to Alan García, who strategi-
cally positioned himself as a newly converted defender of the status
quo with a promise of “responsible change.” Despite his catastrophic
mismanagement of government in the 1980s, García seemed like the
safer choice. As the leader of Peru’s last remaining organized politi-
cal party, he could have chosen to emphasize social inclusion, but in-
stead he governed according to neoliberal priorities. 36 As Paulo Drinot
observes, notwithstanding his adoption of neoliberal rhetoric, García
was prepared to use the sovereign power of the state to crush opposi-
tion to his neoliberal policies (Drinot 2014, 177). When Amazonian
peoples opposed the expropriation and destruction of their land by
extractivist companies, his repressive response was reminiscent of the
massacre in El Frontón during his fi rst term in office.
When Humala won office after his second bid for the presidency
in 2011, the results were instructive for our analysis. “While Humala
won the second round on a platform of much more moderate social
change than in 2006, on reaching office he switched abruptly to the
right” (Crabtree and Durand 2017, 126). He agreed to focus on pro-
moting economic growth, macroeconomic stability, and the avoid-
ance of any destabilizing measures. This as a perfect example of nega-
tive power.37 Humala found himself constrained by a hostile business
community, critical media, and powerful right-wing adversaries in the
Congress and the courts. The strategy of the Right was to ensure that
Humala would avoid any temptation to follow the populist example of
Hugo Chávez in Venezuela. They expressed outrage when he appeared
to be considering the nationalization of the Spanish multinational Rep-
sol, which would have provided resources for populist redistribution;
they fought hard against efforts to require prior consultation of indig-
enous peoples; and they lobbied to ensure that major mining projects
like Conga would be approved.38 There was no countervailing force
against these pressures. Humala lacked a strong party organization or
social base to hold him accountable. The result was that he was able to
accomplish little of what he promised, leading to disappointment in the
ranks of his electoral base and defections of allies in Congress.

Soifer_6844-final.indb 103 8/17/18 11:53 AM

 104 Maxwell A. Cameron

The election of Pedro Pablo Kuczynski was, in a sense, the ultimate

vindication of technocratic neoliberalism. A wealthy white investment
banker with extensive experience working in the World Bank and the
International Monetary Fund, Kuczynski could hardly have been more
unlike the majority of Peruvians. As the oldest leader to be elected
president in Peru—having worked under the fi rst government of Fer-
nando Belaunde Terry in 1967–1968, and returning in the 1980s to
serve in cabinet under the second Belaunde government and then as
minister of fi nance under Alejandro Toledo—Kuczynski was, however,
well known to Peruvian voters. When matched against Keiko Fujimori
in a runoff election after two potentially stronger contenders were dis-
qualified by the National Election Board, Kuczynski benefited from
a strong anti-fujimorista animus among voters. A respectable third-
place showing for the Left put its leader Verónika Mendoza in the po-
sition of being able to endorse Kuczynski, helping him win the runoff.
In December 2017, President Kuczynski, facing allegations of corrup-
tion, narrowly survived an impeachment vote and then proceeded, to
the dismay of the Left, to pardon Fujimori. When it emerged that al-
lies of the president offered opposition politicians financial rewards if
they voted against impeachment, Kuczynski resigned in March 2018.

Conclusion

I began this chapter by suggesting that two different—but not incom-

patible—lessons could be drawn from an analysis of the impact of the
internal conflict on the transformation of Peru’s constitutional order.
The fi rst was a version of Naomi Klein’s shock doctrine: the internal
conflict had a major impact on Peru because it created the emergency
situation within which it was possible to adopt a more neoliberal and
authoritarian constitution. The second lesson is that the Shining Path
failed to revolutionize Peruvian society and politics. The inability of
criollo republican institutions to contain the pressures of popular mo-
bilization, and the inevitable pressures for redistribution in an ex-
tremely unequal society, led to a process of social democratization that
was arrested but not reversed by neoliberal governance in the 1990s.
There is evidence for both claims.
The perception of an existential threat allowed a pro-market eco-
nomic model that could be implemented as not just a solution to the
effect of populism and economic crisis, but also a technique of gov-

Soifer_6844-final.indb 104 8/17/18 11:53 AM

 From Oligarchic Domination to Neoliberal Governance 105

ernment. The 1993 constitution would not only enshrine a neoliberal

economic model and reconstitute political power around its inviola-
bility, but also lock in an approach to governing that would direct
the power of the state to neutralizing collective movements through a
combination of demobilization, privatization, surveillance, targeting,
and entrepreneurial culture. As a result, two decades later, Peru did
not follow its Andean neighbors Venezuela, Bolivia, and Ecuador in
adopting post-neoliberal constitutional reforms.
But it is also important to recognize that the Shining Path failed in
large measure because there was a democratic alternative to revolu-
tion. Not an incrementalist or reformist movement, the Shining Path
presented itself as a harbinger of wholesale transformation of state and
society. It threatened constituted power by exploiting the greatest vul-
nerability in Peru’s social order: the unequal and oppressive nature of
social and political institutions, and the enduring and ongoing cultural
legacies of colonialism, including political centralism, social exclusion,
and denial of cultural recognition. The challenge for Peru’s democ-
racy is to show that these problems can be addressed within the frame-
work of a democratic regime and state. Those who advocate progres-
sive change could do worse than to return to the path of reform and
social democratization initiated before the Shining Path and the Peru-
vian military scorched the earth, while learning from more recent ex-
periences with constitutional change that have flourished in the region.
The 1979 constitution is likely to remain a point of reference for such
an agenda.39

Notes

An early version of this chapter was presented in a workshop at Harvard Uni-

versity, May 19–20, 2014. I am grateful to Hillel Soifer, Alberto Vergara,
Teivo Teivainen, Francisco Durand, Fabiola Bazo, and two anonymous refer-
ees for advice and guidance.

1. See chapters 7 and 8 in this volume.

2. On the meaning of neoliberalism see Foucault 2004; Brown 2015; Jones
2012. See also the section of this chapter on the emergence of neoliberal gov-
ernance techniques fostered by the 1993 constitution. See Peck (2010, 26) on
roll-back versus roll-out neoliberalism.
3. In effect, Peru is a case of Naomi Klein’s (2009) “shock doctrine.”
4. It involved the torture of the body that Foucault (1979, 9) called
“punishment-as-spectacle.”

Soifer_6844-final.indb 105 8/17/18 11:53 AM

 106 Maxwell A. Cameron

5. This term was fi rst used by Cotler, and subsequently by others including
Handelman (1975, 45) and McClintock (1981, 65).
6. Michael Mann (1986, 169–170) makes a useful distinction between
despotic power, which involves actions that are implemented by rulers with-
out routine, institutionalized negotiation with opposition groups, and infra-
structural power, which is the capacity to penetrate society to implement po-
litical decisions. For a discussion, see Soifer 2008.
7. “We [represented] the Aristotelian Mean,” said one officer. “We wanted
the law to be upheld—as it has to be—but with liberty.” Cited in Pásara
(1983, 329).
8. For an excellent and thorough discussion of clasista unionism, see Balbi
1989, 79–90. See also Tovar 1985.
9. Consider the eloquent words of Carlos Iván Degregori, Cecilia Blon-
det and Nicolás Lynch (1986, 21), writing about the Cruz de Mayo barrio in
Rímac (italics in original): “De ser siervos, waqchas, clientes o plebeyos, a lo
largo de su periplo los fundadores de Cruz de Mayo se convierten en parte del
contigente de pioneros que, al invadir tierras y construir nuevos asentamien-
tos llevan (o traen) el proceso de democratización social al corazón mismo de
dominio oligárquico y burguéz dependiente, a Lima.” (“From being servants,
waqchas [poor], clients or commoners, through their journey, the founders of
Cruz de Mayo became part of the continuum of pioneers who, by invading
lands and constructing new settlements, carry (or bring) social democratiza-
tion into the very heart of the oligarchy and bourgeoisie, to Lima.”)
10. Brazil also adopted a new constitution as part of its transition to
democracy.
11. The major groups were FOCEP, PSR, PCP, UDP.
12. Given inevitable social change, the Shining Path was in a “race against
time” (Degregori 1992, 35).
13. Guzmán professed to be influenced by Thomas Mann’s account of the
story of Moses in The Tables of the Law.
14. A similar process of popular overflow was reflected in the music scene
(see Bazo 2017).
15. As a graduate student at the time, I observed the tank attack and in-
haled the tear gas while walking through downtown Lima to post a grant ap-
plication in the Correo Central.
16. De Soto appears to have concluded that his advocacy of informal en-
trepreneurs actually contributed to the defeat of the Shining Path. Eduardo
Dargent (2014) disagrees.
17. See Schulte-Bockholt 2013, 91–92.
18. By contrast, there was only 16 percent abstention in the Constituent
Assembly elections (Bernales 1980, 44).
19. See the essays in Del Golpe de Estado a la Nueva Constitución (Lima:
Comisión Andina de Juristas, 1993).
20. Similar deliberations occurred within the Chilean military junta after
the 1973 coup. See Barros 2002.
21. Article 112 read: “The presidential term is for five years. The president
can be reelected immediately for one additional term. After another consti-

Soifer_6844-final.indb 106 8/17/18 11:53 AM

 From Oligarchic Domination to Neoliberal Governance 107

tutional period has transpired, the ex-president can run again, subject to the
same conditions.”
22. The law stated: “interpreted in an authentic manner, the reelection to
which Article 112 of the Constitution refers is limited to presidential terms
initiated after the date of promulgation of the text of the Constitution. In con-
sequence, interpreted authentically, in the calculation one does not retroac-
tively include presidential periods initiated prior to the entry into force of the
Constitution.”
23. See Gill and Cutler (2015) for a more recent discussion.
24. Paulo Drinot (2014) draws on Foucault’s discussion of neoliberalism in
his essay on García’s second term.
25. I am grateful to Teivo Teivainen for this point.
26. See http://elcomercio.pe/noticias/emprendedores-516839.
27. See http://www.promoviendo.tv.
28. See http://aeg.pucp.edu.pe/boletinaeg/notaegresados/189_egresados
.htm.
29. I am grateful to Francisco Durand for this observation.
30. Peru’s Ministry of Development and Social Inclusion is involved in
a number of projects aimed at poverty alleviation. These programs include
PRONAA (which prevents malnutrition affecting three million children);
FONCODES (offers temporary job creation for 400,000 people); CUNA
MAS (supports early child development in the fi rst three years of life, bene-
fiting 80,000 children); JUNTOS (a conditional cash transfer program with
700,000 users); and PENSION 65 (supports 170,000 elderly people over
sixty-five years of age in highlands).
31. The sense of disappointment with market reforms could be read in the
subtitles of the two major books written by Hernando de Soto. The Other
Path, written in 1989 (its title is a play on the name of Peru’s revolutionary
movement, the Shining Path), heralded an “invisible revolution in the Third
World” as an alternative to underdevelopment and “Marxist-Leninist funda-
mentalism,” as Mario Vargas Llosa wrote in a preface for the book (p. xx). De
Soto’s second book, The Mystery of Capital, written in 2000, promised to ex-
plain “Why capitalism triumphs in the West and fails everywhere else.”
32. Scholars like Yashar (1999) and Arditi (2010) wrote of “post-liberal”
democracy.
33. Roughly a dozen countries (including Venezuela, Brazil, Chile, Ec-
uador, Bolivia, Paraguay, Argentina, Uruguay, El Salvador, Nicaragua, and
Costa Rica), encompassing most of the population of the region. Latin Ameri-
can nations often move in packs: shifts in policy from export-oriented growth
to substitution industrialization and then to neoliberalism have often been
taken by the region’s nations in tandem. Peru has occasionally been an out-
lier—it was a “late adopter” of populist reforms, and it adopted them under
military rule. Nevertheless, there are broad similarities in the patterns and se-
quences of national development trajectories across the region, to which Peru
is no exception.
34. See http://hdr.undp.org/en/countries/profi les/PER and http://hdr.undp
.org/en/countries/profi les/BOL.

Soifer_6844-final.indb 107 8/17/18 11:53 AM

 108 Maxwell A. Cameron

35. See http://www.centrodememoriahistorica.gov.co/micrositios/informe

General/estadisticas.html.
36. This argument is developed in Cameron 2011. García (2005, 10–16),
in a book published before the election, argued that globalization demanded
modernization. Chile was the most successful case in Latin America, and free
trade agreements could prevent the erosion of competitiveness. His position
was consistent with an emphasis on social inclusion and justice, but in prac-
tice his main concern was investment, growth, and free trade.
37. See http://elcomercio.pe/politica/gobierno/ollanta-humala-presento-hoja
-ruta-que-buscacambio-rumbo-noticia-756721.
38. For an account of the attempt to acquire Repsol, see Vergara and En-
cinas 2016, 174–175.
39. Such an agenda could recognize popular sovereignty and the right to
self-determination of indigenous people within their ancestral territories; seek
a new balance between the rights of workers and the rights of employers, as
well as the elimination of labor services and support for unionization; restore
the right to renegotiate contracts with foreign investors; democratize political
parties; and perhaps include buen vivir (living well, in harmony with nature)
as a principle to guide public policy and an alternative to the neoliberal focus
on growth. See Gudynas 2011.

Soifer_6844-final.indb 108 8/17/18 11:53 AM

 CHAP T ER 4

The Internal Armed Conflict and

State Capacity: Institutional Reforms
and the Effective Exercise of Authority
Hillel David Soifer and Everett A. Vieira III

This chapter draws on existing research, contemporary newspaper ac-

counts, and novel analysis to investigate the effect of Peru’s Internal
Armed Conflict on the capacity of the state, and specifically its coer-
cive capacity. On the one hand, we fi nd that the confl ict played an im-
portant role in shaping the design of state institutions in the coercive
realm, but on the other, the confl ict had a more limited impact on the
ability of those institutions to exercise authority and extend their influ-
ence over territory. We argue that a holistic perspective on state capac-
ity that takes into account both institutional design and institutional
strength allows us to paint a more nuanced picture of how conflict has
and has not shaped the contemporary Peruvian state.
Our chapter explores the three facets of the armed conflict’s poten-
tial effects that frame this book as a whole: wartime mechanisms, con-
flict legacies, and the politicization of the conflict. We trace how each
might have influenced state capacity. The fi rst section evaluates a se-
ries of changes to the state’s coercive capacity that took place during
the armed conflict. We examine trends in the funding and size of the
state’s armed forces, showing that while the conflict brought increases
in this dimension of coercive capacity, these vanished as soon as the
conflict ended, leaving no lasting shift. By contrast, the confl ict did
have some impact on the structure of these institutions; it spurred in-
stitutional changes that shaped their effectiveness and the unity with
which they operated. It also resulted in significant learning in the realm
of counterinsurgency.
The second section investigates the ways in which the memory or
lessons of the confl ict have been manipulated by political actors and
institutions in Peru’s security apparatus. Here we examine how state

Soifer_6844-final.indb 109 8/17/18 11:53 AM

 110 Hillel David Soifer and Everett A. Vieira III

actors seeking to increase the power of the intelligence agencies and

to limit the salience of human rights reforms have leveraged the recent
past of confl ict in the service of their goals. We conclude that a signifi-
cant effect can be found here: the conflict has played an important role
in shaping the strategies used by actors within the state as they strug-
gle for institutional resources, autonomy, and power.
In the third section, we turn to lasting legacies of the confl ict af-
fecting the state’s coercive capacity. Here our findings show once again
that the conflict has shaped institutional design more than institutional
strength. We fi nd that institutional changes made during the conflict
have tended to remain after it came to an end, despite their long-term
suboptimality. We also fi nd a surprising pattern of increased postwar
state territorial reach focused precisely in those areas where the state
was unable to oversee elections in 1989.
After presenting these findings, we consider the broader implica-
tions for the study of state capacity in the conclusion. Broadly speak-
ing, we suggest that our fi ndings pose a paradox to studies of confl ict
and state capacity. On the one hand, they dovetail quite well with the
robust body of scholarship on the role of conflict as a historical cause
of long-term institutional outcomes, and on the striking stability of in-
stitutional design (Ertman 1997; North 1990; Soifer 2016). On the
other hand, we show that even as it has shaped their design in lasting
ways, the confl ict has relatively little lasting effect on the capacity of
those institutions. We also suggest that scholars of war and the state
should incorporate the territorial dimension into their study of state
capacity.

Studying State Capacity

Scholars have long explored whether and how confl ict shapes state ca-
pacity. While focused exclusively on the Peruvian case, our chapter
builds on this literature. Most research fi nds that conflicts like that
seen in Peru, in which (as detailed further below) the state did not
mobilize massive manpower or resources, are unlikely to lead to sig-
nificant and lasting increases in state capacity (Downing 1992; Tilly
1992; Centeno 2002; Thies 2005). Yet because these studies focus on
cross-national evidence rather than within-case analysis and rely on
crude proxies for state capacity, they lack the nuance possible in a deep
investigation of a single case such as the one we conduct.

Soifer_6844-final.indb 110 8/17/18 11:53 AM

 The Internal Armed Conflict and State Capacity 111

Our chapter makes three important choices in terms of how we ap-

proach the study of state capacity. First, we focus on its coercive facet.
In so doing, we move away from the focus of most existing studies
that explore how conflict affects state capacity by focusing on the ex-
tractive efforts of the state (Centeno 2002; Thies 2005; Slater 2010).
We make this analytical move for several reasons. First, because the
most direct consequence of state-making efforts that respond to secu-
rity threats is military mobilization, we argue that the effect of conflict
on state capacity is best evaluated in its coercive facet. Second, because
other elements of state-making such as taxation are shaped (even dur-
ing wartime mobilization efforts) by many factors other than state ca-
pacity, they provide a more indirect assessment of how confl ict makes
the state.
The second distinctive feature of our approach to state capacity is
that we make a multifaceted assessment of coercive capacity. Rather
than choosing one indicator, we examine separately the size and fi nanc-
ing of coercive institutions, various aspects of their institutional design,
their effective extension across the national territory, and the effective-
ness with which they operate. This multifaceted approach, which fits
well with recommendations by Saylor (2013) and Soifer (2008) on the
measurement of state capacity, lets us bring a new perspective to the
debate about how confl ict shapes “stateness” by showing that it affects
some aspects but not others.
Third, in the spirit of the volume as a whole, we shift away from
cross-case comparison to a concerted exercise in within-case analysis
that is intended to tease apart the legacies of the conflict from those of
the broader political and economic crisis of the 1980s. In so doing, we
speak both to the Peruvian experience and to the broader question of
what kinds of conflicts, under what kinds of circumstances, produce
state-building.

Wartime Mechanisms

We begin our analysis by investigating how state coercive capacity

changed during the years of confl ict itself, tracing the evolution of the
resources devoted to security provision and counterinsurgency as well
as the quality of security institutions. Much scholarship on the state ar-
gues that when mobilization, extraction, and other elements of stateness
are increased during wartime, they will remain high thereafter (Tilly

Soifer_6844-final.indb 111 8/17/18 11:53 AM

 112 Hillel David Soifer and Everett A. Vieira III

1975; Centeno 2002). This relationship derives from various mecha-

nisms that underpin the so-called ratchet effects (Campbell 1993) by
which state capacity increases during a confl ict but does not decline
after its end. Based on the broad consensus that war-making shapes
state capacity, we should expect that increases in military spending and
manpower taking place during the confl ict should not be reversed in its
aftermath. We examine whether a step change or ratchet effect can be
observed as military size and manpower respond to the conflict.

Money and Manpower

Perhaps the simplest way to assess the state’s capacity is to examine the
resources at its disposal that can be used to implement its chosen pol-
icies.1 In the coercive realm, those resources are both fiscal and mate-
rial; we can assess both the spending on security and the size of coer-
cive organizations—in other words, money and manpower. We begin
here with the very basic question of how the funding and size of the
Peruvian military evolved during the confl ict with Sendero.
Military spending data from the Correlates of War project, mea-
sured in current-year US dollars, shows no ratchet effect for this opera-
tionalization of state capacity. We observe an upward trend in military
spending during the pre-confl ict era of the 1970s, when the country
was under military rule. This is followed by a highly unstable pattern
during the years of confl ict: spending rose more than 50 percent from
1979 to 1980, and nearly tripled between 1981 and 1982. It then de-
cayed over the remainder of the Belaunde years, falling back to 1979
levels before tripling from 1986 to 1987. That year, however, was an
outlier, and thereafter spending remained relatively constant for the
remainder of the confl ict, with a small uptick in 1996 likely due to mo-
bilization for the border conflict with Ecuador. Thus we see no lasting
effect of the conflict on military spending; levels in the 1990s were no
higher than those of the pre-conflict decade.
We also use data from the Correlates of War project to chart the
size of the armed forces, measured on a per capita basis, over the same
time period. The data reveal a significant increase in military mobi-
lization in the pre-conflict period: from 1970 to 1980 the per capita
size of the military nearly doubled. Given that the country was ruled
by the armed forces during these years, that pattern is not surpris-
ing. Yet the trends in military mobilization after the onset of the Inter-

Soifer_6844-final.indb 112 8/17/18 11:53 AM

 The Internal Armed Conflict and State Capacity 113

nal Armed Conflict are quite striking: after a 10 percent increase be-
tween 1980 and 1983 in per capita military size, the manpower of the
armed forces fell by 33 percent over the remainder of the decade, and
remained at a constant level during the fi nal years of the confl ict. In
terms of military manpower, then, no significant increase in state ca-
pacity happened during the conflict. There was no lasting ratchet ef-
fect in this case, nor even evidence of temporary gains in state capac-
ity during the confl ict.
Trajectories of coercive resources do not fit with the expectations
of the “bellic” school of state-building. Military spending returned to
pre-conflict levels by 1990, and remained there except for the brief
spike in 1996–1997. This suggests the absence of the classic ratchet ef-
fect of war-making on state capacity; rather than a step change in mil-
itary spending that persisted after the confl ict, we see instead a tempo-
rary increase that decayed after the confl ict ended. Similarly, the data
show no ratchet effect for military size either. After a peak in the early
1980s, the per capita size of the armed forces slowly declined over the
remainder of the period of internal confl ict and the subsequent decade.
While this level remained higher than that of the early 1970s, the per
capita size of the armed forces is actually smaller than it was during
the late years of military rule.

Institutional Changes

Overall, in terms of resource-based assessments of state capacity, we

fi nd no lasting impact of the Internal Armed Conflict. Yet state capac-
ity cannot be reduced to the resources at its disposal; one must also ex-
amine the quality of state institutions to get a broader picture of the
state’s capability to impose order. In this area, we fi nd that the Peru-
vian case follows the expectations of the bellic school of state forma-
tion more closely: the threat posed by the Sendero insurgency spurred
a series of fundamental shifts in state institutional design, which we
trace below in the police, the military, and the intelligence services. In
each instance, we see significant institutional reforms during wartime,
which are the focus of this section. Later in this chapter, we show that
these institutional reforms, though suboptimal in the longer term,
tended to remain in place after the confl ict came to a close; thus they
represent a significant legacy of the confl ict, and one that persisted de-
spite longer-term negative consequences in the post-confl ict setting.

Soifer_6844-final.indb 113 8/17/18 11:53 AM

 114 Hillel David Soifer and Everett A. Vieira III

Police
At the onset of the internal confl ict in 1980, Peru had three differ-
ent national police organizations: the Guardia Civil (GC), which had
a loose military structure; the Policía Investigaciones del Perú (PIP),
the most clearly civilian; and the Guardia Republicana (GR), the most
militarized of the three (Costa and Neild 2005, 218). These forces
competed for resources and power within Peru’s security apparatus,
and tensions manifested in armed clashes among police units, includ-
ing a “police rebellion” (Gorriti 1999, 73) on September 6, 1980, that
was defused only by presidential intervention. The consistent pattern
of “fragile armed standoffs” between units of distinct police forces
(ibid., 113) was also a product of corruption, as elements within each
of the three police forces competed for control over illegal activities
including, but not limited to, the narcotics trade. The quality of each
of the police institutions was also hampered by poor leadership re-
sulting from the politicization of the initial appointments made by Be-
launde (ibid., 39). The interior minister, under whose aegis all three
police forces were housed, saw “imposing his authority over his police
chiefs” to be a bigger challenge than the insurgency (ibid., 72).
The organizational problems within each force, and tensions be-
tween them, impeded the state’s efforts to respond to Sendero Lumi-
noso, and to perceive the broad challenge the insurgency posed (Gorriti
1999, 62). In 1981, as the intensity of the confl ict ramped up, the po-
lice continued to identify and arrest a significant share of the perpetra-
tors of individual actions, but did little to address the organization be-
hind these actions. This can be seen in the effects of the fi rst state of
emergency declared in the armed confl ict, which was announced on
October 12, 1981, and applied to five provinces of Ayacucho. Though
the police were able to take control of the emergency zone, their in-
ability to gain civilian support and their poor intelligence hamstrung
their ability to do significant damage to the Shining Path, which sim-
ply went dormant for this brief period. The police played a limited role
in the conflict for the remainder of the Belaunde years, as military in-
tervention took center stage.
Upon taking office in 1985, Alan García and his administration
sought to alter the course of counterinsurgency efforts away from a
military focus that had been ineffective as well as responsible for mas-
sive human rights violations. Doing so, however, required police re-
form. A series of initiatives unfolded over subsequent years, culminat-

Soifer_6844-final.indb 114 8/17/18 11:53 AM

 The Internal Armed Conflict and State Capacity 115

ing in the December 1988 creation of the Policía Nacional del Perú
(PNP), which unified the three forces in the police sector (CVR 2003a,
186). Many experts and the leaders of various police institutions saw
this unification as hasty and potentially unwise, and proposed post-
poning it until it could be planned more carefully (ibid., 186). But uni-
fication went ahead largely because it was seen as a key element of the
state’s response to Sendero’s gains in the central and southern sierra.
Police unification was one example of the conflict’s effects on the de-
sign of state coercive institutions. It is less clear, as we will show, that
these changes had a significant impact on the quality of policing.

T he A r m ed Forces
As their counterinsurgency interventions proved both ineffective and
politically unpalatable, the armed forces also became a site of reorga-
nization. Efforts in this area were intended to bring the armed forces
under the control of the elected government in order to provide bet-
ter oversight of counterinsurgency policy (CVR 2003b, 275). Orga-
nizational changes were highlighted by the April 1987 creation of the
Ministry of Defense, which consolidated the various heretofore sepa-
rate security agencies, including the Ministries of War, Air, and Navy,
the Joint Command of the Armed Forces, and the National Defense
Secretariat.
But another institutional shift that occurred in the course of the
conflict worked in the opposite direction, and also had important im-
plications for how the military would subsequently operate. Work-
ing at cross-purposes to military consolidation under civilian author-
ity was the granting of complete autonomy to zone commanders and
the leaders of individual military units in the zonas de emergencia, be-
ginning with Law 24150, which was introduced at the very end of Be-
launde’s presidency and persisted nearly untouched until 2004 (Jas-
koski 2013b, 42, 87). As we will discuss, the autonomy of coercive
actors has persisted, with important consequences for state capacity.

In t elligence Serv ices

Another crucial change to Peru’s security apparatus over the course of
the conflict was a series of efforts to bolster the intelligence services be-
ginning in the mid-1980s. At the onset of the confl ict, the intelligence
services were of extremely poor quality. The Servicio de Inteligencia
Nacional (SIN) had been weakened sharply under the Morales Bermú-
dez administration in terms of its reach into the rural highlands, as the

Soifer_6844-final.indb 115 8/17/18 11:53 AM

 116 Hillel David Soifer and Everett A. Vieira III

number of paid informants fell sharply (Gorriti 1999, 199). This left
the police “groping around” (ibid., 45) as it confronted the fi rst insur-
gent actions. Military intelligence, in particular, consistently sought a
foreign element underpinning the Shining Path, and “could only con-
ceive of guerrillas in terms of a classic Castro-style movement” (ibid.,
52). Reporting was “not only seriously deficient but intentionally in er-
ror” (ibid., 51) as intelligence officers sought to curry favor with their
superiors.
As strategists became aware that counterinsurgency tactics were
alienating the rural population, the emphasis on intelligence increased
(Weeks 2008, 53). This prompted officials to seek an end to the fierce
rivalries and lack of coordination between the various intelligence
agencies, and to improve their quality. Facing political obstacles that
limited his ability to eliminate existing agencies, García chose instead
to create a new unit to complement those already present. He estab-
lished a special intelligence group within the PIP intelligence unit, the
Grupo Especial de Inteligencia del Perú (GEIN), and tasked it with the
explicit purpose of capturing Abimael Guzmán and striking at Sen-
dero’s political structure (Tapia 1995b, 34; Taylor 1998, 51; CVR
2003a, 198).
The role of intelligence grew even more sharply under Alberto Fuji-
mori (Obando 1994, 119). Though the SIN had been the formal coor-
dinating body for the various intelligence organizations for more than
a decade, the Fujimori-Montesinos years saw its power grow. One im-
portant moment in this process was Law 25635, which took effect on
July 28, 1992, and formally established the SIN as the consolidated
and centralized hub for counterinsurgency (CVR 2003a, 229). A key
consequence of this move was a shift of control over intelligence to an
institution that came under limited government oversight; the indepen-
dent operation of the intelligence service was a key characteristic of
the Fujimori regime. Though the growing sway of the intelligence ap-
paratus was, in large part, a consequence of the president’s close rela-
tionship with the head of the SIN, Vladimiro Montesinos, it was also
very much influenced by the security concerns engendered by the inter-
nal conflict.

Learning

In addition to resources and institutional design, state capacity is also

affected by institutional quality. To explore this, we focus on the ex-

Soifer_6844-final.indb 116 8/17/18 11:53 AM

 The Internal Armed Conflict and State Capacity 117

tent of learning in the realm of counterinsurgency. Here, once again,

we fi nd that significant institutional change occurred. We examine
whether there were changes in the state’s counterinsurgency strategy,
whether these changes increased effectiveness, and whether improve-
ments can be attributed to the lessons obtained in the course of the
conflict. Since the bulk of our evidence supports the role of the con-
flict and learning in the evolution of counterinsurgency strategy, it sup-
ports the fi nding that the Peruvian coercive apparatus was strength-
ened as the result of the confl ict.
At the onset of the Sendero uprising, the training and expertise
of the armed forces had been principally focused on external threats
rather than internal ones (Taylor 2006, 35–37). As a result, the mil-
itary lacked the training necessary for internal, unconventional war-
fare (Koc-Menard 2006, 335–336). As discussed previously, though
the police were reasonably effective in direct response to Shining Path
actions—suspects were arrested with regularity—this mode of re-
sponse was ineffective in terms of addressing the broad sweep of the
insurgency, as manifested in the failure of the police to weaken the
Shining Path during the initial emergency declaration of late 1981.
The fi rst coherent counterinsurgency strategy emerged autono-
mously within the military, without guidance from Belaunde, as it
faced its initial deployment to Ayacucho in late 1982 (Jaskoski 2013b,
40–41). This was nothing short of a scorched-earth campaign intended
to eradicate the insurgency and instill in the minds of the population
that they should support the military over Sendero, combined with an
expectation (unmet due to economic crisis) that other state agencies
would fund economic development to win over peasants (Taylor 1998,
43; Jaskoski 2013b, 42). In the vernacular of Stathis Kalyvas (2006),
this type of indiscriminate violence did not differentiate between sen-
deristas and local residents, but whereas Kalyvas emphasizes the in-
ability of armed actors to differentiate between civilians and insur-
gents, the Peruvian military appears to have been unwilling to make
this distinction. This sentiment is captured by the former minister of
war, General Luis Cisneros: “If to kill two or three senderistas it is
necessary to kill 80 innocents, then it does not matter . . . The peas-
ants have to decide where they wish to die: with Sendero or the armed
forces” (Granados 1987, 27, 33). Unsurprisingly, human rights vio-
lations committed by the military spiked during this period of indis-
criminate violence (Obando 1998a, 196; Koc-Menard 2006, 336).
Toward the end of the 1980s, a retreat from indiscriminate violence

Soifer_6844-final.indb 117 8/17/18 11:53 AM

 118 Hillel David Soifer and Everett A. Vieira III

under García began to cohere into a new counterinsurgency strategy.

This new approach “gave emphasis to the socio-political, economic,
and psychological dimensions of waging unconventional warfare”
(Taylor 2006, 36). Elements of this strategy, several of which are dis-
cussed elsewhere in this chapter, included programs to appeal to ci-
vilians, encouragement of the rondas campesinas, intelligence service
reorganization and unification to improve coordination, giving the mil-
itary political control in zones of emergency, and increasing the pen-
alties for those accused of terrorism (Tapia 1995b, 35–36; Obando
1998a, 198–199; Taylor 1998, 50–52; Taylor 2006, 36–37). This new
strategy was carefully delineated in a new counterinsurgency manual
issued in August 1989 (Ministerio de Defensa 1989). This manual sug-
gests a conscious effort by the military to shift its emphasis to an effort
to win over the hearts and minds of the population in order to defeat
Sendero, centering on the acquisition of local knowledge and the im-
portance of gaining local support. And while local tactics were left up
to individual zone commanders and unit leaders, there was a clear shift
toward this new model of responding to Sendero after the late 1980s.
There is strong evidence that counterinsurgency became strikingly
more effective in the late 1980s. Perhaps the clearest instance of these
fruits of learning can be seen in the direct gains against the Shining
Path achieved by General Alberto Arciniega during his deployment in
the Upper Huallaga Valley. Rather than view the cocaleros as crimi-
nals and eradicate their crops, he sought favorable relations with the
locals. Arciniega also improved the morale of his troops through in-
creased interaction between officers and their subordinates. These im-
provements led to increased cooperation with the local population
and led to victories against Sendero (Taylor 1998, 50). More broadly,
the increased effectiveness that resulted from the shift away from
scorched-earth policies could be seen in various parts of the country
in the beginnings of trust-based relationships between peasant com-
munities and local military units, in alliance against a common enemy
(Tapia 1995a, 24). And this emergence of a faint but crucial degree of
trust facilitated the spread of the rondas campesinas, which were fatal
to Sendero (Starn 1996, 245).
The spread of more effective counterinsurgency efforts was marked
by both horizontal and vertical mechanisms of diffusion. First, the
decentralization of counterinsurgency was an organizational feature
facilitating learning. While, as discussed above, this limited the ex-
tent to which edicts from the center were implemented in a consis-

Soifer_6844-final.indb 118 8/17/18 11:53 AM

 The Internal Armed Conflict and State Capacity 119

tent and rule-bound manner, it also facilitated learning by allowing a

climate of experimentation. Because zone commanders were rotated
regularly and were independent of the units stationed in various loca-
tions, experiences from one region could travel to other parts of the se-
curity apparatus. There was, therefore, space for a horizontal form of
learning in the military organization. But policy change also resulted
from top-down pressures from the commanding heights of the state.
The growing salience of human rights abuses on the part of the mil-
itary, and their negative implications for counterinsurgency, led Gar-
cía to opt for a change in strategy to avoid the policy trap of terror-
ism whereby a heavy-handed response to the violence killed innocents
along the way and thus reinforced support for Sendero (Taylor 1998,
37; Gupta 2008, 205).
This learning on the part of the security apparatus resulted in more
effective counterinsurgency, and represents one way in which the con-
flict spurred a significant increase in state coercive capacity. Whereas
earlier we traced how the confl ict sparked changes in institutional de-
sign but not necessarily in the effectiveness of the state, here we high-
light an important counterexample. Disaggregating state capacity into
multiple indicators allows the observation of this complex pattern of
mixed consequences.

The Conflict as Strategic Frame

Beyond its direct effects, we also fi nd evidence that the confl ict has
shaped how actors in the post-conflict context view their roles and the
means by which they seek to pursue power and resources. We fi nd that
the conflict has in this way shaped both state-society relations and in-
teractions among institutions within the state. Two arenas in which
this type of legacy is particularly clear are the intelligence apparatus
and the military.
As mentioned above, much of the increase in the surveillance ca-
pacity of the intelligence services under Fujimori can be traced not to
necessities of the internal confl ict, but to the logic of the regime; as
subsequent revelations showed, surveillance was a key component of
its strategy of holding power. But the role of intelligence in counter-
insurgency served as justification for its further expansion in the
post- conflict period and generated resistance to attempts to increase
oversight over surveillance through intelligence reform. Once the sur-

Soifer_6844-final.indb 119 8/17/18 11:53 AM

 120 Hillel David Soifer and Everett A. Vieira III

veillance apparatus had grown in power and escaped the oversight of

state leaders, it became very hard to put the genie back in the bottle.
The intelligence apparatus continued to grow in budget and in surveil-
lance capacity, and to operate autonomously from government over-
sight, both under Fujimori and thereafter. This continued even after
the SIN was deactivated and a replacement established under Alejan-
dro Toledo (Weeks 2008). Even after a series of intelligence-related
scandals under Toledo and the CVR’s call for reforming or eliminating
the military’s role in intelligence, little has changed. In consequence,
this aspect of security remains largely outside the oversight of national
government institutions despite a wave of security sector reforms, and
despite its formal separation from the armed forces2 (Obando 2008–
9). Thus references to the conflict were used to justify the increased
power and autonomy of the surveillance apparatus during the Fuji-
mori era.
References to the conflict have also served as justification in the mil-
itary’s challenges of government oversight, especially in episodes of
contention over human rights reforms since Fujimori’s fall. Military
officers continued to see the army’s central mission as counterinsur-
gency, and felt that they could not adequately carry out this mission in
a context where they were restricted by human rights oversight (Jas-
koski 2013b). In addition to the expected efforts of the military hier-
archy to protect its members from prosecution for violations during
conflict, Maiah Jaskoski’s interviews with officers who served in emer-
gency zones during the 1980s and 1990s show that they saw account-
ability measures as hampering their central mission of continued coun-
terinsurgency against the remnants of Sendero Luminoso. Rather than
protesting or rebelling against human rights reforms, the military has
pressed against them by underperforming its missions in the realm of
counterinsurgency, regularly disobeying orders to “conduct more as-
sertive counterinsurgency operations” (Jaskoski 2012a, 81). Civilian
authorities under Toledo and early in the second García administra-
tion were unable to compel the army into full performance of counter-
insurgency. Only when the García government passed Law 29166 in
December 2007, which once again broadened the jurisdiction of mili-
tary courts and modified the rules of engagement to reduce the vulner-
ability of soldiers to human rights charges, did the army resume asser-
tive efforts (ibid., 83). Thus lessons learned during the confl ict led the
armed forces to resist control on the part of the central state, which
undermined coercive capacity in important ways for much of the de-
cade after the confl ict ended.

Soifer_6844-final.indb 120 8/17/18 11:53 AM

 The Internal Armed Conflict and State Capacity 121

Legacies

We have already discussed a series of changes to institutions made in

the course of the conflict. A vast body of scholarship in sociology (Ert-
man 1997), political science (Barnett 1992), and economics (North
1990) has argued that state institutions tend to endure in their initial
form despite their suboptimality. Building on this scholarship, we now
examine whether some of the changes discussed previously have per-
sisted into the post-confl ict period, and show that these have affected
state capacity in lasting and negative ways. We also examine a novel
legacy of conflict for state capacity: the extent to which the state can
reassert effective control over territory that fell outside its grasp dur-
ing the confl ict.

Institutional Continuity in Government-Military Relations

State capacity is defi ned by the extent to which the central state can
generate the implementation of its chosen policies to their fullest capa-
bility. Thus it is shaped by relations between the commanding heights
of the state and its bureaucratic agents (Migdal 1988; Soifer 2008).
In the coercive realm, then, state capacity is affected by the extent to
which authorities can command and control military institutions.
We therefore examine the evolution of government-military rela-
tions in Peru. Against the common tendency to explore these issues
under the rubric of democratization, we show that they cannot be un-
derstood without reference to the confl ict, which shaped government
oversight over the coercive apparatus in important and lasting ways.3
We must concede that factors other than the confl ict did play a role. In
particular, the fact that 1980 marked a transition from military rule
to democracy meant that the military entered the decade, and the con-
flict, under a strained relationship with state leaders (Obando 1998b,
386). These tensions shaped Belaunde’s reticence in deploying the
armed forces to Ayacucho until Christmas of 1982 (DeGregori and Ri-
vera 1993, 9).
We identify a dramatic decay of civilian control over the course of
the 1980s that can be traced directly to the conduct of the confl ict.
A broad scholarly consensus supports this view that the fight against
Sendero Luminoso reduced government control over the military. As
described above, a key element of this reduced control was the auton-
omy of military officers in emergency zones, to whom much discre-
tion was given, whether de facto (DeGregori and Rivera 1993, 7) or

Soifer_6844-final.indb 121 8/17/18 11:53 AM

 122 Hillel David Soifer and Everett A. Vieira III

de jure (Mauceri 1991, 98). These officers were accountable to the cen-
tral state only to a very limited extent, and even then only to military,
rather than civilian, institutions of justice. Except for fragmentary ef-
forts during the fi rst two years of the García administration, the 1980s
saw “few efforts to assert civilian control over the military” (Cameron
and Mauceri 1997, 240). Indeed, even García’s changes to the military
command structure reveal that rather than serving as an agent of state
leaders, the military enjoyed significant autonomy from the state. Gar-
cía had to co-opt high-ranking officers with promotions, political ap-
pointments, and access to resources to guarantee their support for the
regime. While this strategy cemented alliances between civilian and
military leaders, it further undermined the control state leaders exer-
cised over the military leadership (Obando 1998b).
The Fujimori years saw a deepening of patterns that had emerged in
the previous decade rather than significant institutional change. While
the counterinsurgency conflict continued to rage during Fujimori’s
fi rst years in power, the fragmentation and local autonomy that had
marked the late García years escalated (Mauceri 1991, 98–99). Mili-
tary officers’ lack of accountability remained a theme as human rights
abuses escalated and protection from prosecution (and especially civil-
ian prosecution) was extended. Particularly striking was the legal con-
struction of protection of soldiers for delitos de función, which came
to cover any action committed in an emergency zone (Jaskoski 2012a,
76), and the ban on review of sentences handed down by military tri-
bunals in civilian courts. Institutional changes, such as the 1991 de-
cree that empowered the military to confiscate resources and conscript
individuals as it saw fit, increased military autonomy (Mauceri 1991,
99). The military enjoyed influence over not only these reserved do-
mains of counterinsurgency and human rights, but also government
policy more generally. At the same time, however, the internal institu-
tions of the military came to be deeply politicized, generating a debate
about how much autonomy the military really enjoyed during the fu-
jimorato (Obando 1998a, 199; Avilés 2009). Scholarship on this pe-
riod minimizes the impact of the counterinsurgency conflict, focusing
instead on regime dynamics to account for patterns of government-
military relations and arguing that Fujimori’s lack of party support,
tensions between branches of government, and the role of Montesinos
are sufficient to account for these patterns. Yet we highlight that many
of the changes observed during this period represent continuity with
trends that fi rst emerged during the confl ict. Thus, our fi ndings fit well

Soifer_6844-final.indb 122 8/17/18 11:53 AM

 The Internal Armed Conflict and State Capacity 123

within the framework developed by North (1990), Ertman (1997), and

other scholars who show that institutional changes made as states seek
to respond to severe challenges will persist thereafter, even when they
limit state capacity in the long run.

State Territorial Reach

Yet we reach a puzzling fi nding when state capacity is instead assessed

in terms of its ability to effectively exercise control across its national
territory. This final section of the chapter shows that places where the
state could not exercise control at the height of the conflict saw sig-
nificant state-building in the post-conflict period. Existing theories of
the effects of internal conflict on the state overwhelmingly neglect this
territorial aspect of state capacity. This is a striking omission, given
that internal conflicts have a fundamentally territorial nature (Kaly-
vas 2006). By showing how this dimension of state capacity evolved in
the post-confl ict period, we paint a more holistic picture of the rela-
tionship between conflict and stateness. As will be detailed further, we
also fi nd some suggestive evidence of state-building in the post-conflict
period that has not been explored by scholars of contemporary Peru.
This section seeks to measure the extent and limits of the reach of
the Peruvian state by improving on existing measures such as those de-
veloped by Kent (1993), Koc-Menard (2007), and McDougall (2013).
We assemble original data on a new measure of state territorial reach,
assessed at the provincial level.4 This allows us to develop more spa-
tially disaggregated measures than those used by other scholars, and
to measure state reach over the entire national territory rather than
just the conflict zones. We assess the reach of state institutions at the
height of the conflict, using election irregularities in the municipal
elections held on November 12, 1989, which Sendero tried to spoil, as
an indicator. The failures of these municipal elections represent a good
measure of limits on the state’s ability to implement a basic set of pol-
icies, and in particular they capture variation in the security climate
at the local level, since insecurity (as will be discussed) was central to
explaining why candidates withdrew from elections and voters opted
not to risk going to the polls. We then examine the extent to which the
state’s authority has been restored to those regions, or whether state
weakness has remained constant, thus revealing the extent to which
state weakness that appeared5 in the course of the conflict continues to
mark Peru to the present.

Soifer_6844-final.indb 123 8/17/18 11:53 AM

 124 Hillel David Soifer and Everett A. Vieira III

During the November 1989 municipal elections, Sendero Luminoso

attacked municipal officials and candidates, threatened to kill many
others, and seized and burned libretas electorales in an effort to dis-
rupt the vote. Candidates withdrew in many municipalities, and in
others none could even be found who were willing to run.6 We use
data on district-level election returns to assess the extent of four types
of electoral irregularities that occurred in each province and thus cap-
ture state territorial reach.7 A fi rst type of irregularity is No Elections,
which refers to entire provinces for which no election results are avail-
able. A second, related, type is Missing Districts, which refers to prov-
inces in which more than 25 percent of districts have no election re-
sults reported. Because voting was mandatory, we can also assess two
other kinds of election irregularity. Our third type of irregularity is
Spoilage, which refers to provinces where null and blank votes made
up 60 percent or more of votes cast in at least one district.8 The fourth
type is Low Turnout, which refers to provinces that had one or more
districts where fewer than 100 total votes (including null and blank
ballots) were cast.9 We assemble district-level data into scores for each
province on four kinds of election irregularities, which might reflect
the limited ability of the state to implement voting procedures. We use
the province (n = 188) as the unit of analysis rather than the district
because our measures of contemporary state reach are available at this
level of aggregation.
Our portrait of election irregularities in the 1989 municipal elec-
tion sheds light on the limits to the state’s reach at this moment in the
conflict. These irregularities were found in many parts of the coun-
try; only 72 of the country’s 188 provinces were unaffected. Of the
country’s 25 regions, all but two (Moquegua and Tumbes) saw elec-
tion disruptions. Problems were widespread within many regions: in
Ayacucho, Huancavelica, Ucayali, and Junín, every province saw elec-
toral disruptions, as did all but one in Huánuco. Seven of ten prov-
inces in the department of Lima suffered from election irregularities,
as did the capital’s port of Callao, which represents its own province.
All four types of irregularities we examined were widespread. No
elections at all took place in 29 of the country’s 188 provinces, includ-
ing four in Ayacucho, Apurimac, and La Libertad, three in Huánuco,
and two each in the regions of Amazonas, Ancash, Cusco, Junín,
and Ucayali. The complete absence of elections was not limited to re-
mote provinces with few residents; provinces in this category included
Cajatambo on the outskirts of Lima. Returns were missing from at

Soifer_6844-final.indb 124 8/17/18 11:53 AM

 The Internal Armed Conflict and State Capacity 125

least 25 percent of districts in an additional 30 provinces, including

five in Huánuco, four each in Arequipa and Junín, and three each in
Ancash, Puno, and Lima. Thus missing returns affected at least some
share of elections in nearly one-third of the provinces in the country,
impacting 410 districts. Widespread spoilage affected 203 districts in
61 provinces. Finally, low turnout was found in an additional 103 dis-
tricts spread across 43 provinces.
But our fundamental interest is in the question of whether this ter-
ritorial weakness of the state persisted after the confl ict ended. To as-
sess this, we drew on an Index of State Density (Indice del Densidad
del Estado—IDE) constructed by the UNDP (2009) for each province
in Peru and calculated for 1993 and 2007. Scores are composed of five
dimensions: doctors per ten thousand residents, percentage of house-
holds with electricity, percentage of households with water and sew-
age provision, secondary school enrollment rate, and percentage of the
population possessing an official identity card. Though this set of in-
dicators moves our object of examination away from the purely coer-
cive facet of state capacity, it allows the assessment of the subnational
reach of state institutions.
To examine whether state weakness revealed in the 1989 elections
continued to resonate thereafter, we ran a series of regression models
that explore whether election irregularities are associated with subse-
quent differences in scores on the UNDP state density index. To avoid
saturating the models, we added few control variables. We controlled
for the human development indicator level (available from the same
UNDP data set) in the province to account for the direct or indirect ef-
fects of socioeconomic development on the IDE, which has increased
steadily over the course of the post-conflict period. In the models for
2007, we also controlled for a lagged dependent variable: the relevant
measure of state reach in 1993. As our independent variables, we used
each of the four types of election irregularities, as well as an election
disruption index for each province that aggregates all four. This dis-
ruption index can take on values from zero to three, since the absence
of results for the entire province cannot co-occur with any of the other
three types of irregularities.10 The mean value for this province-level in-
dex of electoral irregularity is 0.84, with a standard deviation of 0.77.
Table 4.1 shows the results of our fi rst set of analyses. Model 1,
which explores the effects of the overall election disruption index on
state reach, fi nds that disruptions have a positive and statistically sig-
nificant effect on provincial levels of state presence in 2007, even if we

Soifer_6844-final.indb 125 8/17/18 11:53 AM

 126 Hillel David Soifer and Everett A. Vieira III

Table 4.1. Models 1–3, effects of electoral irregularities on

post-conflict state reach

Model 1 Model 2 Model 3

DV: IDE 2007 DV: IDE 2007 DV: IDE 2007

Constant 0.382*** 0.354*** 0.418

(0.091) (0.090) (0.092)
HDI 1993
Boycott index 0.013***
(0.005)
Spoilage 0.288***
(0.008)
No election results
Missing districts
Low turnout 0.015*
(0.009)
IDE 1993 0.639*** 0.640***
(0.042) (0.043)
HDI 2007 0.008 0.063 −0.429
(0.174) (0.172) (0.177)
# obs 188 188 188
Prob > F 0.0000 0.0000 0.0000
R-squared 0.8460 0.8514 0.8423
Adjusted R-squared 0.8435 0.8490 0.8398

*p < .10
**p < .05
***p < .01

control for state presence in 1993 and the HDI scores. The remaining
models explore different manifestations of election irregularity sepa-
rately: model 2 shows the results for the spoilage indicator, and model
3 for low turnout.11 Most strikingly, the effects of spoilage are not only
statistically significant, but positive: provinces that contained districts
with high ballot spoilage in 1989 had higher levels of state capacity in
2007. This suggests that in the intervening years, the state extended its
reach into those areas where it had lacked control in 1989.
Models 4–8, shown in table 4.2, examine the five separate indica-
tors of the IDE as dependent variables. They show that spoilage is asso-
ciated with an increase in all aspects of the state’s presence, suggesting

Soifer_6844-final.indb 126 8/17/18 11:53 AM

 The Internal Armed Conflict and State Capacity 127

that the state has been able to extend its reach into areas where Sen-
dero was able to spoil the 1989 election. The spoilage of the 1989 elec-
tion is associated with statistically significant increases in all but one of
the indicators of state capacity, and these effects are substantively quite
large. In 2007, the provinces where one or more districts had a spoiled
election in 1989 saw increased presence of state institutions in several
ways. All else equal, in provinces where Sendero was able to spoil elec-
tions in 1989, 2007 saw electricity provided to 4.2 percent more house-
holds (unweighted average of provincial rates = 56.3 percent) and
water and sewage to 3.9 percent more households (average = 39.8 per-
cent). Similarly, where spoilage had occurred, secondary school enroll-
ment was increased by 2.1 percent (average = 67.4 percent) and iden-
tity registration (average = 96.0 percent) had increased by 1.1 percent.
The fact that this relationship remains after controlling for HDI sug-

Table 4.2. Models 4–8, effects of spoilage on individual dimensions of

stateness in 2007

Model 4 Model 5 Model 6 Model 7 Model 8a

DV: DV: Water/ DV: DV: DV: ID

Electricity sewage Doctors Education Cards

Constant 11.106 36.350** 12.010*** 48.874*** 100.193***

(12.303) (14.538) (3.219) (7.466) (5.128)
Spoilage 4.183** 3.870* 0.706 2.077* 1.082**
(1.732) (1.996) (0.450) (1.067) (0.429)
HDI 2007 40.042* −33.582 −15.489*** −8.122 −10.407
(22.237) (26.126) (5.704) (13.922) (9.754)
Lagged DV 0.666*** 0.905*** 1.344*** 0.688*** 5.645**
(1993 score) (0.040) (0.053) (0.062) (0.040) (2.367)
# obs 188 188 188 188 188
Prob > F 0.0000 0.0000 0.0000 0.0000 0.0020
R-squared 0.7018 0.6733 0.7551 0.6856 0.0769
Adjusted 0.6969 0.6679 0.7512 0.6805 0.0618
R-squared

a
Because there is no 1993 value for identity registration, we use the 1993 over-
all IDE here. This likely explains the poor fit of this model.
*p < .10
**p < .05
***p < .01

Soifer_6844-final.indb 127 8/17/18 11:53 AM

 128 Hillel David Soifer and Everett A. Vieira III

gests this is not simply a function of economic development, but a man-

ifestation of an increasingly present Peruvian state.
This pattern of increasing state presence in the post-confl ict era
in zones heavily controlled by Sendero has gone unnoticed in exist-
ing scholarship.12 As the introductory chapter to this volume shows,
much scholarship on contemporary Peruvian politics has been writ-
ten without systematic engagement with the legacies of confl ict. The
issue of social spending and public good provision is no exception: our
fi ndings here, though suggestive, show a clear pattern of investment in
public good provision in confl ict-affected areas, and in particular in
those areas where Sendero Luminoso exercised control at its height,
which cannot be accounted for by existing explanations. Scholarship
on FONCODES and on social welfare provision more generally under
Fujimori has tended to emphasize motives of clientelism and political
mobilization (Schady 2000) as well as compensation for the costs of
the radical economic restructuring his government implemented. Un-
der Toledo, social provision is commonly explored through the lens
of poverty reduction (St. John 2010, 109). Yet beyond our suggestive
fi nding about public good provision outcomes such as school enroll-
ment and electricity provision, the historical record also reveals that
the conflict seems associated with the targeting of public good ef-
forts: for example, the Fondo de Compensación Municipal, founded in
1994, focused on rural areas (Degregori, Coronel, and del Pino 1998,
244), and the Juntos CCT program introduced by Toledo in 2004 was
fi rst rolled out in one hundred districts in Ayacucho, Apurímac, Huan-
cavelica, and Huánuco. Though commentators (St. John 2010, 117)
describe these as being the “poor” regions of the country, it may be no
accident that they are also the most confl ict-affected.
In all, our results show that the lack of state reach was erased rela-
tively quickly after the confl ict ended. Though its effects should not be
overstated, we fi nd that some degree of state-building in terms of state
territorial reach seems to have occurred in the post-confl ict period.
We suggest that this issue is worthy of further exploration in terms of
our understanding of the politics of social welfare in recent decades
in Peru. A logic that begins with conflict legacies suggests that in ad-
dition to poverty alleviation and clientelism as motivations for state
spending in these areas, we might uncover a state-building project—
limited by fiscal conditions and the economic model (see chapter 3)—
that sought to extend the state’s control over areas where Sendero held
sway at the end of the 1980s. Understanding the extent and nature of

Soifer_6844-final.indb 128 8/17/18 11:53 AM

 The Internal Armed Conflict and State Capacity 129

this state-building project might lead scholars to adjust their under-

standings of the post-confl ict governments in important ways.

Conclusion

This chapter has assessed the effects of the Internal Armed Conflict
on several different dimensions of state coercive capacity. On the one
hand, we found few lasting effects in terms of the size and funding
of the armed forces. Though these facets of coercive capacity did see
change to some degree in the course of the confl ict, wartime shifts dis-
appeared after the confl ict came to a close. Yet on the other hand,
we fi nd that the confl ict impacted the design of state coercive institu-
tions in important and lasting ways. The war with Sendero spurred a
series of reforms of the army and police, as well as significant shifts
in counterinsurgency doctrine, which have persisted after the confl ict
ended, though as Jaskoski (2013b) shows, that doctrine is only im-
plemented to a limited extent. Moreover, the experiences of confl ict
served as a central justification used by the military and intelligence
apparatus for expanded autonomy and scope of action during the Fu-
jimori administration and subsequent post-confl ict administrations. In
this way, the Peruvian coercive apparatus and its relationship to the
state’s commanding heights remain shaped by the conflict. Finally, we
show that significant territorial extension of the state occurred in the
post-confl ict period, focused precisely on those regions where the state
had been weakest at the confl ict’s height.
Our fi ndings suggest that conclusions about the effects of confl ict
on state capacity depend on how one defines and assesses the outcome.
Peru’s Internal Armed Conflict seems to have transformed coercive in-
stitutions in important ways, as did the wars in early modern Europe
that have been the focus of the bellic school of theory on state-building
that began with Tilly (1975). This fi nding seems quite surprising given
the claims of Centeno (2002) that confl ict in Latin America generated
“blood and debt” but little institution-building, and the claims of Thies
(2005) that internal conflict weakened rather than built states. Yet in
fact it is consistent with both of these seemingly contradictory claims
in the literature. Like Centeno and Thies, we fi nd limited effects of
the conflict on the effective exercise of authority by the state. Yet like
Downing (1992), Ertman (1997), and other studies of European state
formation, which buttressed their claim that war made states by refer-

Soifer_6844-final.indb 129 8/17/18 11:53 AM

 130 Hillel David Soifer and Everett A. Vieira III

ence to the institutional changes that confl ict spurred, we fi nd that it

led to significant institutional shifts. Whether state capacity should be
studied in terms of institutional design or institutional effectiveness is,
of course, beyond the scope of our analysis. But our findings suggest
that the conclusions we can draw about how it is changed by confl ict
depend on how state capacity is measured.

Notes

For comments on earlier versions of this chapter, we thank Jorge Domínguez,

Angelica Durán Martínez, Tulia Falleti, Gustavo Flores-Macías, Agustina Gi-
raudy, Matthias vom Hau, Diana Kapiszewski, Jennifer Pribble, and Deborah
Yashar as well as participants in the conference organized by the editors at
DRCLAS, Harvard University, members of the Princeton Latin American Pol-
itics Working Group, and the anonymous reviewers. We thank Noelia Chávez
and Madai Urteaga for assiduous research assistance.

1. In broader cross-national studies of state capacity, this is often done

through the use of national income or government revenue as a proxy for
those resources. See, for example, Fearon and Laitin (2003), who use GDP
per capita as a proxy for state capacity. For a discussion of the strengths and
weaknesses of input-based approaches to assessing state capacity, see Soifer
2008; Soifer 2012; and Fukuyama 2013.
2. Indeed, this lack of civilian oversight over intelligence remained a con-
cern under the Humala administration as well. See Angel Páez, “DINI destina
el 70% de su presupuesto para acciones secretas,” La República, June 14, 2013,
http://www.larepublica.pe/14-06-2013/dini-destina-el-70-de-su-presupuesto
-para-acciones-secretas.
3. Because our focus is on government control rather than civilian control
in particular, we use the term government-military relations rather than civil-
military relations throughout.
4. See de la Calle (2017) for a similar, though not identical, use of the
same data as an indicator of territorial control during Peru’s confl ict.
5. The lack of similarly widespread failures to implement elections ear-
lier in the decade supports our claim that this manifestation of state weakness
emerged in the course of the confl ict.
6. Contemporary news reports indicated problems in regions ranging
from Tingo María (see Mario Munive “Uchiza le dice Sí a las elecciones,”
La República, November 5, 1989) to Ayacucho (Carlos Valdez, “Candidato
IU se reinscribe para elecciones en Ayacucho,” La República, November 9,
1989), while scholarly accounts point to similar issues in locations including
the Mantaro Valley (Manrique 1998, 205), Puno (Rénique 1998, 320), and
Huanta (Coronel 1996, 61).
7. We use the province as the unit of analysis rather than the district be-
cause our measure of post-confl ict state reach, as described below, is measured

Soifer_6844-final.indb 130 8/17/18 11:53 AM

 The Internal Armed Conflict and State Capacity 131

at the provincial level. Election data are drawn from http://www.infogob

.com.pe/Eleccion/eleccion.aspx, last accessed May 30, 2016.
8. In addition to Sendero Luminoso’s attempts to prevent the elections
from proceeding, the MRTA called for spoiled ballots (“Uchiza le dice Sí a las
elecciones,” La República, November 5, 1989). While blank and null ballots
might be interpreted as evidence of a protest vote in other contexts, the calls
for such behavior on the part of insurgents combined with threats of violence
against voters in a context of mandatory voting leads us to include spoilage as
an indicator of state weakness.
9. We are reluctant to use the 1981 or 1993 census data to calculate turn-
out as a percentage of population due to the high levels of internal displace-
ment during the confl ict. Our threshold of 100 total votes is a very conserva-
tive measure: the mean of votes per district was 2,755.
10. A binary index of the form “boycott = 1 if no results for province OR
two of the other three conditions hold” produced nearly identical results, as
did various other specifications for the election irregularities variable.
11. Models with the missing results indicators as an outcome produce no
statistically significant results; this suggests that these indicators may be cap-
turing data problems rather than an insecure electoral climate. This hunch is
supported by the fact that electoral results are missing for the entire province
of Cusco, which was not likely a place where Sendero was able to disrupt elec-
tions to this degree.
12. One exception is a brief discussion in Degregori, Coronel, and del
Pino 1998, 245, which points to a “particularly impressive” increase in social
spending in “areas of the country previously hit by the war with Sendero Lu-
minoso” but does not explore the reasons for this systematic pattern.

Soifer_6844-final.indb 131 8/17/18 11:53 AM

 CHAP T ER 5

Impact and Legacies of Political Violence

in Peru’s Public Universities
Eduardo Dargent and Noelia Chávez

No one would deny the severe impact of political violence on Peru’s

public universities from 1980 to 1995. As highlighted by diverse works
about the causes and consequences of political violence, the Shining
Path (PCP-SL) was born in, and its trajectory was closely intertwined
with, public universities.1 A significant number of PCP-SL leaders were
university professors, many of its members were students, and the
group acted violently within universities, including such acts as threats
and assassinations. State repression carried out by military and para-
military groups also struck public universities, with disappearances
and selective killings on campuses (Degregori 1990a; Sandoval 2002;
CVR 2003a; Sandoval and Toche 2007; Reátegui 2008; Jave, Cépeda,
and Uchuypoma 2014, 44–81). Universities were described as inse-
cure, politicized, permanently on strike, and incubators of terrorism.
Thus it is reasonable to see in political violence one of the main causes
of the current crisis of public universities.
Nonetheless, many other problems also affected public universi-
ties before and during the internal confl ict (Lynch 1990; CVR 2003a,
582–594; MINEDU 2005; Degregori and Sandoval 2009, 28–49; Jave,
Cépeda, and Uchuypoma 2014, 35–44; Disi 2017). As discussed in the
next section, the academic degradation of public universities started
at least two decades before political violence, and was driven by a set
of internal factors as well as elements of the broader context. Mass
enrollment coupled with low budgets led to a severe decline in aca-
demic quality in what had heretofore been small and elitist institutions.
Moreover, radicalism was not limited to Sendero: during the 1970s and
1980s radical Marxist and Maoist groups were prevalent in universities
(Lynch 1990; CVR 2003a, 594–598). University authorities’ patrimo-
nial and frequently corrupt management of these institutions, a char-

Soifer_6844-final.indb 132 8/17/18 11:53 AM

 Political Violence in Peru’s Public Universities 133

acteristic that has remained salient for the last forty years, also con-
tributed to the decline of public universities (Lynch 1990; Degregori
and Sandoval 2009, 38–41; N. Chávez 2014), as did the economic cri-
sis that hit Peru in the late 1980s. In sum, although we focus on the im-
pact of violence on public universities, this effect cannot be studied in
a vacuum, separated from other issues affecting them since the 1960s.
Without pretending to fully disentangle the effects of violence from
these preexisting trends and the impact of the economic crisis, we pro-
pose that violence fundamentally changed the image of public uni-
versities and deepened their many problems. Violence, we argue, was
fundamental to making public universities a “non-issue” in the pub-
lic debate and to determining their current precarious state. In specific
terms, we argue that political violence produced three legacies that
continue to have a significant impact to the present day.
First, violence installed an image of public universities as chaotic
and violent in the minds of political elites and society at large. Since
the onset of the internal confl ict, not only have these institutions come
to be perceived as precarious and politicized, but some political and
economic elites have portrayed them as being violent, a radical threat
for society. This “stigma” remains strong in public opinion even if
such trends have significantly reduced (Gamarra 2010; Jave, Cépeda,
and Uchuypoma 2014; Nureña 2016). As discussed in the introduc-
tion to this book by Soifer and Vergara (p. 12), the frequent struggle
of political actors and other elites that “frame, shape, and manipulate
features of the conflict and its legacies for their own purposes” repro-
duces these negative images, keeping them salient still today.
The second legacy refers to the departure from these institutions of
two groups that are crucial for quality and meritocracy: as a result of
the chaos among public universities, both middle- and higher-income
students and high-quality professors migrated to private universities
that bloomed in the 1980s. Violence also had an effect on the qual-
ity of education, as competitive professors shifted to private universi-
ties or just stopped teaching out of safety concerns.2 Public universities
never regained their quality after the confl ict, and this weakness led to
decreased salience in the country’s public and academic debates.
Third, though more tentatively, we also propose that violence and
its memory played a part in weakening students’ capacity for collective
action. Even today students retain an association between the exces-
sive politicization and radicalism of the eighties and students’ politi-
cal organization, leaving university politics in the hands of clientelistic
student representatives and radical political groups. This weak capac-

Soifer_6844-final.indb 133 8/17/18 11:53 AM

 134 Eduardo Dargent and Noelia Chávez

ity for student collective action contributes to the absence of student

mobilization in pursuit of higher education reforms or other political
issues in Peru, a clear contrast with countries such as Chile, Colombia,
or Mexico that have witnessed student mobilization and protests in re-
cent years.
Together, these three legacies constitute severe obstacles to exter-
nal or internal efforts to reform public universities. While neoliberal
higher education reformers in the rest of Latin America debated in the
nineties about what to do with public universities, they were not an
issue for market reformers in Peru, who instead promoted profit in
higher education as a response to the growing demand for university
education and ignored public universities almost entirely. And despite
important efforts in recent years, public universities still remain low
as a political priority and in public salience, despite their relevance for
educational reform in Peru.
To advance these claims, we focus mainly on the case of the Univer-
sidad Nacional Mayor de San Marcos (UNMSM). Although it has bet-
ter quality and infrastructure than other public universities in Peru, it
portrays well the common trajectory of public universities affected by
violence. Also, as Peru’s most important public university, San Marcos
attracted significant attention from the media and became a referent
for the public as a result of the events there. Throughout this chapter
we also present information from other universities, particularly the
Universidad Nacional San Cristobal de Huamanga (UNSCH), the in-
stitution where Sendero started its actions and built its initial support.
The chapter proceeds as follows: in the fi rst section we briefly pres-
ent some antecedent conditions to document the problems of public
universities before violence. We then focus on the impact of violence in
these institutions from 1980 to 1995, showing how these events were
reported to the public. Third, we present the above-mentioned legacies,
provide evidence to justify our claims, and explain how these legacies
impact the possibility of reforming public universities. We conclude
with some general thoughts about the implications of our fi ndings for
the challenge of reforming public universities.

Antecedents: The Public University

before Violence (1950–1980)

Although we cannot properly discuss in detail the causes of the cur-

rent crisis of Peruvian public universities, our aim in this section is to

Soifer_6844-final.indb 134 8/17/18 11:53 AM

 Political Violence in Peru’s Public Universities 135

provide a baseline from which to judge our analysis of the independent

impact of violence and its legacies. We focus on three contributory fac-
tors that predate the conflict: the massive growth of students without
a comparable budget increase, Marxist radicalism that affected plu-
ralism and quality within universities, and clientelism in university
government.
The crisis of public universities was heavily influenced by a process
that started in the 1960s: the massive growth of student recruitment
without a similar growth in budgets (CVR 2003a, 582–587; Degre-
gori and Sandoval 2009, 25–34; Sandoval 2002). Efforts to democra-
tize higher education resulted in a significant increase in the number of
public universities, from five in 1955 to twenty-three in 1979. At the
same time, there was no significant increase in the budget for higher
education. The process of democratization achieved its goal of opening
public academic institutions to lower socioeconomic classes, especially
students from rural areas (CVR 2003a, 582–587). However, the re-
sulting shortfall in funding crippled academic quality in what had tra-
ditionally been relatively small and elitist institutions.
As public universities expanded, considerably less attention was
given to academic quality. There were more than budgetary reasons for
this shortcoming; politics also played a role. Progressive political forces
saw universities as recruitment grounds for progressive or even revolu-
tionary politics (Lynch 2002, 301). As a result, politics within univer-
sities became more about demanding bigger enrollments and defending
students’ stability rather than addressing academic quality. With only
a few exceptions, academic quality strongly decreased in these insti-
tutions in the years to follow, and the poor quality of many academic
programs left the promise of higher education unfulfilled.
Some governments showed more interest than others in reforming
these institutions, but in general political elites were disinterested in
public universities. One of our interviewees explains this lack of inter-
est by the growing distance between radical groups and political elites,
who saw no point in fi nancing an institution controlled by groups that
portrayed the state as an enemy (interview with Burga; all interviews
are listed at the end of the chapter). In sum, low budgets, high enroll-
ment levels, and lack of political interest were all problems affecting
academic quality by 1980 when the internal confl ict began.
Second, by 1980 Marxist radicalism was already strong in the uni-
versities, and it was not limited to the PCP-SL. Throughout the fifties
and sixties, APRA and Acción Popular (AP) were gradually pushed
out of universities by increasingly radical leftist groups. Eventually,

Soifer_6844-final.indb 135 8/17/18 11:53 AM

 136 Eduardo Dargent and Noelia Chávez

these more radical groups achieved control over administrative and ac-
ademic positions, and eroded academic pluralism with a quite simplis-
tic version of Marxism (Degregori 1989; Degregori 1990a; Degregori
1990b; Lynch 1990). Furthermore, the resources provided by univer-
sities allowed these radical groups to detach themselves from national
politics. In open contrast to APRA or Acción Popular, or even to other
leftist groups that started to compete in elections in the democratic
transition from 1978 to 1980, these radical groups saw universities as
an end in themselves. Universities provided them with resources and
jobs, and served as recruitment grounds. These groups therefore re-
jected democratic politics and criticized other leftist parties for taking
part in a “democratic fiasco” (CVR 2003a, 588–594). Political isola-
tion and tight control of universities became the political identity for
these groups. Yet unlike the PCP-SL, these groups believed that revo-
lutionary conditions did not exist in the country, justifying in this way
their decision not to take up arms (CVR 2003a, Conclusion 137).
A quite radical (and simplistic) version of Marxism became domi-
nant within public universities, limiting academic pluralism and meri-
tocracy. This is highlighted by the number of Marxist-oriented courses
taught during these years. A review of courses taught in 1985 within
three academic disciplines at San Marcos (history, anthropology, and
sociology) shows that at least 18 to 30 percent of courses, including
many core courses in each discipline, had significant Marxist content.
This is a modest estimate, as we are only defi ning as “Marxist” those
courses with evident Marxist names, but others with a more neutral
name may also have included strong Marxist content. Although the SP
was also present within universities, it was never one of the dominant
groups: academic radical Marxism and Maoism cannot be reduced to
its presence. Radicalism that included quite extreme versions of Mao-
ism was broadly present in national universities by 1980 even in the
absence of Sendero, with serious consequences for academic pluralism
and university quality.
This radicalism coexisted with our third factor, a patronage-oriented
use of resources by university authorities and student representatives.
By controlling government and administrative positions and develop-
ing clientelistic linkages with student representatives, political groups
achieved tight control over public universities. By Peruvian law, student
representatives have a one-third share in public universities’ representa-
tive bodies, which elect authorities such as deans and presidents (recto-
res). In the 1960s and 1970s, an already clientelistic tradition deepened

Soifer_6844-final.indb 136 8/17/18 11:53 AM

 Political Violence in Peru’s Public Universities 137

and became aligned with ideologies. Positions were obtained on the ba-
sis of loyalty and ideological agreement more than on academic merit.
For the external observer, this mixture of pragmatic clientelism
with bursts of radicalism is counterintuitive. But as suggested by Lynch
(1990), this radical discourse can serve to justify closure to more trans-
parent processes and meritocratic reforms. As we show below, this cli-
entelism existed alongside violence, remained strong and even grew un-
der governmental intervention in the late 1990s, and is prevalent to this
day, although in a more pragmatic form. Clientelism would have been
a problem with or without violence.
In sum, there were already severe problems in public universities be-
fore 1980 that affected their academic quality and public image. It is
quite likely that these problems would have kept them on a negative
trajectory, gradually leading to the growth of private universities as an
option for professors able to garner higher pay and students with the
capacity to pay more for education. The economic crisis that affected
budgets and salaries also would have contributed to this process. Yet
we will argue that within this trajectory, violence constituted a tip-
ping point, deepening these problems and damaging popular percep-
tions of public universities. The image of chaos and violent radicalism
that came to be associated with public universities, the degree of tal-
ent and quality lost during those years, and the accelerated growth of
private universities in the 1980s and 1990s are all phenomena related
to political violence. These trends partly explain why these institutions
were largely ignored in Peru, while in the rest of the region neoliberal
reformers initiated, or tried to initiate, reforms in public universities.

Conflict-Era Mechanisms:
The Impact of Violence on Universities (1980–1995)

Violence by both the PCP-SL and state security forces impacted public
universities. The insurgents targeted universities, aiming to recruit stu-
dents and build support for their armed struggle (CVR 2003a, Con-
clusion 137). The PCP-SL was successful in becoming a long-stand-
ing presence on campuses, though it was considerably less successful
in achieving control over academic and administrative positions. Con-
trol over dorms and dining halls served to create connections with
poor students, many of them migrants from rural areas who became
attracted to Maoism (CVR 2003a; Sandoval and Toche 2007, 56–58).

Soifer_6844-final.indb 137 8/17/18 11:53 AM

 138 Eduardo Dargent and Noelia Chávez

Members of the academic community lived side by side with profes-

sors, students, and administrative workers known or suspected for
their militancy or sympathy with Sendero. Violent actions by the PCP-
SL included death threats against professors and students, surveillance,
and even disappearances. According to the CVR, the PCP-SL was re-
sponsible for killing or disappearing thirty-one students (17.6 percent
of all students killed or disappeared) (CVR 2003a, 610).
Security forces also acted violently within universities. In the fi rst
years of the armed conflict the state carried out few activities in uni-
versities, but gradually these institutions became the center of intelli-
gence surveillance, police intervention, and later military (and para-
military) actions including disappearances and selective killings. The
army and the police carried out repressive actions at several univer-
sities. According to the CVR, 2.3 percent (118) of all the victims at-
tributed to the state were university students (CVR 2003a, Conclusion
140). The direct attacks against universities critically changed these
institutions.
Although the press presented a dramatized view of the actual con-
ditions and overgeneralized the case of universities such as San Mar-
cos, the Universidad Nacional San Cristobal de Huamanga, and La
Cantuta to other institutions barely touched by violence, the situation
was certainly chaotic. Professors of social sciences at San Marcos in-
terviewed for this chapter recall how members of Sendero were con-
stantly interfering with academic activities, which led some of their
colleagues to teach their classes off campus for their own security.
Critical thinking was associated by external actors and the military
with terrorismo. The presence of violent groups within universities,
state repression, and the mounting economic crisis combined to cre-
ate an institution in chaos. As will be discussed later, this degree of
violence had an impact on the decision of high- and middle-income
students and of competitive professors to leave public universities for
private ones, a trend that did not reverse after the violence ended. And
these risks increased the costs of becoming involved in university poli-
tics, weakening student associations.
Our goal in this section is not only to show through the review of
media sources the dramatic impact of violence during those years, but
also to make the reader appreciate the extent of the negative percep-
tions of public universities as presented by the media. In this way, we
give plausibility to our claim that these years increased the stigmatiza-
tion of public universities as violent and radical “focos subversivos.”
A review of three national press sources—a centrist political magazine

Soifer_6844-final.indb 138 8/17/18 11:53 AM

 Political Violence in Peru’s Public Universities 139

(Caretas), a leftist-inclined newspaper (La República), and a conser-

vative paper (Expreso)—show these dramatic events and the way they
were framed by the media. Most of these reports focus on San Marcos
and, to a lesser degree, La Cantuta. Other universities, even if deeply
affected by violence, as were the UNSCH and the Universidad Nacio-
nal del Centro, did not receive similar coverage in a national press that
focused on Lima.
News reports about violence within universities were frequent in
the eighties. In 1984, Caretas reported the presence of pro-PCP-SL
students in the demonstrations against a proposed new higher educa-
tion law. La Cantuta was frequently presented as the main center of
Marxist-Leninist indoctrination in Lima (see, for example, “Cuando
Sendero Sopla,” Caretas no. 786 [February 1984]). But most news sto-
ries refer to state interventions in these institutions. During the gov-
ernment of Alan García (1985–1990), the police conducted raids in
universities at least four times, and images of the raids were pub-
lished in the press. For example, in 1987, approximately 4,400 troops
broke into La Cantuta, San Marcos, and UNI, arresting 793 people
and fi nding weapons and propaganda. Similar news coverage circu-
lated in 1989 when the minister of defense deployed military forces
to San Marcos and La Cantuta (“Disparo al Aire,” Caretas no. 1054
[April 24, 1989]). That same year, Caretas published a report about
Sendero’s control of the student cafeteria at San Marcos and its partic-
ipation in student elections (“Apanado Senderista,” Caretas no. 1066
[July 17, 1989]). During those years we find similar stories and images
displayed on the covers and in the headlines of La República. The im-
age of a university in chaos was thus produced by not only conserva-
tive media, but all media sources.
Consonant with the rise of law-and-order discourse discussed in
other chapters in this volume, various groups in society saw these in-
terventions as justified. For a growing sector of the population, the
PCP-SL was taking advantage of a weak democracy. As will be shown,
political parties like the rightist Partido Popular Cristiano (PPC) de-
manded fi rmer action by the government. University authorities usu-
ally opposed these actions on the principle of university autonomy, a
hard-fought concession achieved by professors and student representa-
tives. But the image presented in the media was one of the PCP-SL us-
ing public universities to their advantage, somehow similar to the news
reports of jails turned into schools of ideology and recruitment.
In the 1990s some of these trends persisted, and in some cases, such
as the Universidad Nacional del Centro, they intensified. Government

Soifer_6844-final.indb 139 8/17/18 11:53 AM

 Figure 5.1. Universities and violence in newspapers, 1985–1988. Source:
La República, PUCP Newspaper Archive.

Figure 5.2. Universities and violence in newspapers, 1985–1989. Source:

La República, PUCP Newspaper Archive.

Soifer_6844-final.indb 140 8/17/18 11:53 AM

 Political Violence in Peru’s Public Universities 141

Figure 5.3. Universities and violence in newspapers, 1992–1995. Source:

Diario La República and Diario Expreso, PUCP Newspaper Archive.

interventions in universities continued on four occasions after Alberto

Fujimori was elected in 1990: once in 1991, twice in 1992, and again
in 1995 in a more permanent manner (discussed later). Two of the in-
terventions were conducted after the PCP-SL’s attacks in Miraflores,
which had great public impact, as the government argued that both
attacks were planned in public universities. In sum, public universi-
ties were closely associated with violence during the eighties and early
nineties: they harbored terrorists and were a threat to society. This im-
age was transmitted through media sources.
Though universities were in chaos and the PCP-SL was very much
present, a closer look shows that the reality within universities did not
fully match these images. An overwhelming majority of students were
not radical, much less PCP-SL sympathizers, and starting in the late
1980s internal resistance to terrorist groups began to grow. Beginning
in 1988, San Marcos students organized to resist the PCP-SL (Caretas
no. 1023 [September 2, 1988]). In 1991, Caretas reported an increas-
ing demand among students to deradicalize universities. After Abimael
Guzmán’s capture in September 1992, violence gradually declined

Soifer_6844-final.indb 141 8/17/18 11:53 AM

 142 Eduardo Dargent and Noelia Chávez

Figure 5.4. Universities and violence in newspapers, 1992–1995. Source:

Diario La República and Diario Expreso, PUCP Newspaper Archive.

within the universities. By 1995 the unrest was considerably amelio-

rated. However, as some press reports show, the predominant image
projected by the media continued to be one of violence and radicalism.
After the violence subsided, universities continued to struggle with
many of the problems of previous decades. Academic and student rad-
icalism is less prevalent, but budget shortages, low academic quality,
and patrimonialism still are very much present in these institutions.
Even if conditions are better than in the 1980s, the quality of pub-
lic universities is still very low (with some important exceptions, such
as Universidad Nacional Agraria, Universidad Nacional de Ingeniería,
and some schools in the University of San Marcos). If many problems
in higher education are similar to the ones before 1980, what are the
distinctive legacies of political violence for these institutions?

Legacies of Political Violence (1995–2015)

We propose that political violence produced three legacies that impact

public universities and the possibility of their reform. First, a percep-

Soifer_6844-final.indb 142 8/17/18 11:53 AM

 Political Violence in Peru’s Public Universities 143

tion that they are violent and chaotic has remained strong among the
public and elites, even after the 2000 democratic transition when vio-
lence was no longer part of the daily life of these institutions. Second,
the migration of students and professors from public to private uni-
versities because of violence not only affected these institutions’ qual-
ity, but also reduced the internal demand for reform. Third, and more
tentatively, we also suggest that violence affected students’ capacity
for collective action, which has reduced the possibility of internal de-
mands for reform and reduced the capacity of student associations to
mobilize and to wield influence in the public sphere.

The Radical Image of Public Universities

Intervention commissions were appointed for six public universities,

including San Marcos, by the Fujimori government in May 1995, just
a month after his reelection.3 The interventions were justified by point-
ing to the risk these institutions were said to still represent for so-
ciety. Congress approved the intervention in a secret session after a
bomb killed four in the María Angola hotel in Miraflores. According
to Caretas (no. 1365 [June 1, 1995]), a leader of the government ma-
jority linked the bomb with public universities to justify the interven-
tion. Although the measure had reorganization of these institutions
as its alleged goal, no significant reforms were carried out. The com-
missions focused on reducing Marxism in academic plans and purg-
ing professors justly or unjustly associated with the PCP-SL and other
radical groups. In San Marcos, around 150 professors who were Pa-
tria Roja activists were dismissed (interview with Germaná). Marxist
classes at San Marcos were drastically reduced. In the 1994 academic
plans of two of the above-mentioned schools, we only fi nd two Marx-
ist courses, and none in a third school. Although most professors re-
mained in their positions, radical Marxism lost considerable theoreti-
cal predominance in academic programs.4
What we see in this episode is a fi rst example of how the image
of chaos on campus can be used politically by elites: no matter what
the internal conditions of public universities were in reality, politicians
and the media were easily able to depict these institutions as centers of
radicalism and a threat to society. This image of public universities re-
mains strong within elite and public opinion, as if the 1980s are still
with us. To this day, when the issue of public universities makes it to
the front pages, it is usually in violence-related news, such as an al-
leged resurgence of the PCP-SL. Political elites frequently depict pub-

Soifer_6844-final.indb 143 8/17/18 11:53 AM

 144 Eduardo Dargent and Noelia Chávez

lic universities as spaces where the “terrorist threat” is alive and well
(Sandoval and Toche 2007, 59).
To be sure, radicalism still is present in these institutions, and we
do not intend to minimize this problem. As we will discuss, there still
are sympathizers of Sendero in public universities, an issue that de-
serves political attention. And radical Marxism of a quite simple and
dogmatic style is still part of everyday campus life. But our main point
is that despite the fact that this radicalism is much weaker than in
the past, and that violence has been drastically reduced within uni-
versities, the depiction of these institutions as centers of radicalism re-
mains strikingly strong. As documented by some authors, this creates
a strong stigma among professors and students, especially those from
UNSCH, that affects their professional careers (Gamarra 2010; Jave,
Cépeda, and Uchuypoma 2014).
A second event that shows the relevance of this legacy in a some-
what less direct manner was the approval of Decreto Ley 882 in No-
vember 1996. Through Decreto Ley 882 the government allowed
profit at all levels of education, allegedly to respond to a rising demand
for quality education (Sandoval 2002). The result was a dramatic in-
crease in for-profit private universities, of which the vast majority were
of very poor quality. We see this development not as an instance of
how government elites used the legacy of violence to advance reform of
higher education, but as an example of how public universities were so
thoroughly discredited and weakened by their recent violent past that
they simply became absent from debates that deeply affected them. In-
terestingly, in striking contrast to other countries in the region where
neoliberal reforms aimed to reform higher education through changes
in public universities, the Peruvian reform simply marginalized public
institutions (Degregori and Sandoval 2009, 34–35).5
The “neoliberal” higher education reform in Peru consisted of al-
lowing private entrepreneurs to open universities rather than reform-
ing public institutions in line with neoliberal prescriptions. The gov-
ernment did not have to justify the reform, nor their abandonment
of public universities, in political terms simply because the Peruvian
public was not interested in the fate of these institutions. Our news
search shows almost no reaction to the reform. As concluded by De-
gregori and Sandoval (2009, 48), rather than a neoliberal reform of
public universities, what prevailed in Peru was “una lógica casi delib-
erada de mantener en precariedad permanente a las universidades” (“a
quasi-deliberate logic to permanently keep universities in this precari-
ous condition”).

Soifer_6844-final.indb 144 8/17/18 11:53 AM

 Political Violence in Peru’s Public Universities 145

Later in the decade, during the democratic protests against Fuji-

mori’s government from 1997 to 2000, higher education’s legacy of
poor public perception was used by the government to attack and
delegitimize students demanding democratization and economic re-
forms. Government leaders frequently disqualified student protesters
as PCP-SL sympathizers or as manipulated by violent groups. One of
the phrases used by students during protests in response to this type of
government propaganda was “somos Estudiantes, no somos terroris-
tas” (“we are students, we are not terrorists”).6 The intervention com-
missions allowed the Fujimori regime to exercise tighter control over
opposition organizations that emerged within the public universities.
The stigma did not recede with the end of fujimorismo. Even in
democratic times in which the discrediting of democratic protestors is
not necessary, the legacy remains quite strong. How does this image of
dangerous universities reproduce itself two decades after the end of vi-
olence? In part, it is carried forward by the memories of those who di-
rectly experienced those dramatic events. But it also remains strong
partly because media sources have continued to present this nega-
tive image of public universities over and over again during the subse-
quent two decades. Media headlines and news reports frequently link
public universities with this latent danger. As shown in figure 5.5, the
stigma against public universities and students remains strong, espe-
cially among students of San Marcos and Huamanga (Jave, Cépeda,
and Uchuypoma 2014, 132).7
Political elites use this image as part of their rhetoric. For exam-
ple, the media have recently reacted with great alarm to the activi-
ties of the Movimiento por la Amnistía y los Derechos Fundamentales
(MOVADEF), which advocates the liberation of PCP-SL members. As
shown in the news headlines, these activities are reported as a revival
of the PCP-SL in San Marcos and other public universities. Even uni-
versity authorities sometimes depict student protests as examples of
radicalism linked to MOVADEF, despite the fact that many of these
protests center on accusations of abuses and corruption by university
authorities.
Yet the reality within public universities is quite different from this
image. Not only are the activities of these radical groups marginal, but
also there is no evidence to suggest that radicalism is growing.8 Despite
the prevalent view of universities as hotbeds of radicalism, the latest
research on students’ political opinions in San Marcos shows that the
image of “young troublemakers” is a stereotype that has become ob-
solete.9 A 2011 poll (SENAJU 2011) shows an overwhelming majority

Soifer_6844-final.indb 145 8/17/18 11:53 AM

 146 Eduardo Dargent and Noelia Chávez

Figure 5.5. Universities and violence in newspapers, 2007–2013. Source:

Diario Correo, El Comercio, and Perú 21, Internet images.

of students within San Marcos rejecting MOVADEF and the PCP-SL.

Eighty percent consider MOVADEF a façade of Sendero; 78 percent of
the students think that their actions and proposals are negative for the
country, while only an insignificant 2 percent think otherwise. Other
questions provide similar evidence. A sizable majority rejects radical
ideologies. Only 16 percent of students surveyed think MOVADEF’s
presence is worrisome, and few see the PCP-SL’s presence as a concern.
All our interviews and our own experience doing research in public
universities confi rm this contrast between the public image of these in-
stitutions and their internal reality.
But perception is sometimes more important than reality. Ongo-
ing perceptions of chaos and violence help us to understand why pub-
lic universities were abandoned for almost three decades in Peru. Un-
til quite recently, reform proposals for public universities have been
absent. Democratic governments after the transition irresponsibly cre-
ated new universities (five by Alejandro Toledo and sixteen by Alan
García), but without proposing and advancing reform plans. We be-
lieve our fi rst legacy—the negative image of public universities as sites

Soifer_6844-final.indb 146 8/17/18 11:53 AM

 Political Violence in Peru’s Public Universities 147

of subversion and violence—strongly determined this outcome by re-

inforcing the idea among popular and elite audiences that investing in
public universities is a waste of public funds.
Not until 2014, fourteen years after the democratic transition and
after more than thirty years had passed since the last serious debate
over the role of public universities, was a new university law govern-
ing both public and private universities debated and approved in Con-
gress. The law launched a broad reform effort in the Peruvian state,
replacing ineffective institutions and promoting higher standards in
universities. As a result, the Ministry of Education now has a special
office in charge of higher education (Dirección General de Educación
Superior, DIGESU) and a new institution, the Superintendencia Na-
cional de Educación Superior (SUNEDU) aims to assure basic quality
in public and private universities. Although the ongoing reform is not
a priority for the current administration, it is still underway. It is too
soon to tell if these changes will fundamentally change the perception
of public universities.

Cohort and Quality Effect

A second legacy of violence was that it led to the exit of two crucial
groups from public universities, deepening a process that had already
started in the previous decades. First, middle- and higher-income stu-
dents that had in the past chosen to study in public institutions due
to their remaining prestige gradually moved to good private univer-
sities, and the composition of student cohorts also declined in qual-
ity. Second, quality professors, especially those with sufficient capac-
ity to fi nd positions in private universities, also left public universities.
After the violence ended, these trends did not reverse, as middle- and
higher-income students now predominantly attend private institutions
and quality professors fi nd better career opportunities in these insti-
tutions. These students and teachers are crucial drivers of internal re-
forms, which are less likely to occur in the absence of pressure from
these groups.
Regarding the cohort effect, in the 1980s, many students with the
capacity to pay for education left public universities for private ones.
This process enhanced the rising private institutions, a trend that deep-
ened with the 1996 reform. Students from more comfortable fi nancial
backgrounds who choose a public university despite being able to af-
ford private university tuition contribute to enhance academic quality

Soifer_6844-final.indb 147 8/17/18 11:53 AM

 148 Eduardo Dargent and Noelia Chávez

at the public institutions, their option to exit allows them to demand

better educational quality and to be less fearful of retaliation. Losing
this group of students is thus costly for public universities.
We lack hard data from the 1980s to document this cohort effect,
especially our claim about the better academic quality of these stu-
dents. Yet some indirect information confi rms this shift in the socio-
economic composition of public university student bodies. First, the
exponential growth of private institutions and the number of students
registered in them shows this migration toward private institutions. In
1980 there were thirty-five universities in Peru, ten private and twenty-
five public. By 1994 the numbers were equal, with twenty-eight pub-
lic and private universities apiece (MINEDU 2005). The recruitment
rate at private institutions kept growing in the 1980s, while the rate
at public ones grew only very moderately (Sandoval 2002). But it is
not just an issue of numbers; our argument centers on the departure
of the middle and higher classes from public institutions. Although we
do not have numbers for the 1980s and 1990s, information from na-
tional polls in 2000 and 2006 combined with actual recruitment num-
bers in private and public institutions show this trend. According to
Diaz (2008), in 2000 the upper socioeconomic quintile had consoli-
dated their presence in private universities: approximately 56.8 percent
of all upper quintile registered students were in private universities. In
2006 this figure increased to 72.4 percent (Diaz 2008, 104, 113). The
fi rst university census in 2010 confi rms this socioeconomic divide be-
tween private and public universities: the higher the income level, the
lower the number of students in public universities. Nowadays, with
the mushrooming growth of low-quality but affordable private uni-
versities, only the bottom income group (earning less than approxi-
mately US$165 per month) are more likely to attend public than pri-
vate universities.
During the 1980s competitive professors also abandoned public
universities, bringing down academic quality. Professors had to teach
in an atmosphere of fear during the armed conflict. Sometimes they
had to teach with candles due to blackouts, or to deal with radical stu-
dents who insulted them during classes. Furthermore, many professors
were threatened, attacked, and even murdered (interviews with Burga,
Germaná, and López). We have only anecdotal evidence of this effect
in San Marcos from our interviews. But it is quite clear in the case
of UNSCH, the university where the Shining Path started its actions.
UNSCH had seen significant gains in the 1970s that resulted in bet-

Soifer_6844-final.indb 148 8/17/18 11:53 AM

 Political Violence in Peru’s Public Universities 149

ter conditions for teaching and doing research. Although still weak in
comparative perspective and very much politicized, this university at-
tracted good professors and built connections with academic commu-
nities in Peru and abroad after it was reopened in the 1960s (Degre-
gori 1990b). International professors taught in UNSCH during those
years, and young academics with high-quality credentials saw working
there as an interesting career option (interviews with Sala and Mén-
dez). Part of this success was built through an alliance between univer-
sity professors and development projects funded by the international
community: foreign funds passed through UNSCH, raising levels of
academic quality and providing resources for research. Paradoxically,
at the same time that Sendero was growing within UNSCH, a more se-
rious academic community was also gaining in relevance.
For the fi rst years of the internal confl ict, the PCP-SL and the mil-
itary kept armed actions away from UNSCH, but violence gradually
penetrated the institution (Gamarra 2010, 66–71). Both the PCP-SL
and the armed forces threatened professors, who left the university.
Foreign aid projects were cut back or canceled during those years, es-
pecially those calling for research in rural areas. As a result, by the
end of the decade Huamanga had lost the faculty that had made it a
better public university (CVR 2003a, 640–648). In this case we can
more safely conclude that violence altered the university’s trajectory
and made competitive professors leave.
Of course, the economic crisis also contributed to both currents of
departure. The diminished budgets caused strikes that paralyzed uni-
versities for many months, reduced salaries, and degraded equipment
and building facilities. With or without violence, students with eco-
nomic means would have likely eventually migrated to private univer-
sities. And a university in crisis may not have been able to respond to a
growing demand for higher education. Nonetheless, violence speeded
the process of migration, and the chaotic condition of the public uni-
versities prevented even a partial adaptation of the public system to
this growing demand.

Diminished Student Collective Action and

Its Effects Within and Outside Universities

Although more research is necessary to document this claim, we ten-

tatively propose a third legacy: we suggest that violence within pub-

Soifer_6844-final.indb 149 8/17/18 11:53 AM

 150 Eduardo Dargent and Noelia Chávez

lic universities affected students’ collective action capacity. The violent

period beginning in the 1980s not only undermined student associa-
tions by increasing the risk for taking part in them, but also strongly
delegitimized student political participation by associating univer-
sity politics with violence and radicalism. Student organizations were
weakened during those years both as a direct result of violence and
by this growing stigma. We propose that a silent majority within uni-
versities now perceives university politics as corrupt or too radical to
take part in, and leaves political activity in the hands of individuals
and groups that do not advance proposals for academic quality. Fur-
thermore, this weakness and detachment depresses student mobiliza-
tion for higher education reform in the public sphere. As we discuss in
the concluding section, only deeper research can clarify the soundness
of this legacy.
There are two arenas in which politics take place within the univer-
sity: elections for student representatives in instances of cogovernment
(official bodies in which students have representatives), and through
student political organizations that compete for student unions (fe-
deraciones, or centros federados). Yet neither arena now sees student
mobilization for improved educational quality. As mentioned previ-
ously, several studies indicate that student representatives on govern-
ing bodies receive clientelistic benefits from authorities in exchange for
their support in internal elections to maintain these groups in power.
And political groups competing for student unions tend to be consid-
erably more radical than the average student rather than representing
widespread student interests. Furthermore, and partly due to this in-
ternal weakness, Peru is a negative case of student mobilization, with
few students in the public arena either demanding better higher educa-
tion or supporting other political agendas (Disi 2017).
How can we relate violence to this outcome? As discussed before,
in the 1970s radicals and antisystemic speeches spread among students
(Sandoval and Toche 2007, 55). University politics were more extreme
than the positions favored by most political parties, more centered in
radical debates than in electoral competition outside campuses. Inter-
nal politics were also a competition between these radical groups, who
won representative positions in governing bodies and controlled stu-
dent unions. But in general, the student body was more involved at
that point both in national politics and internal politics than it is to-
day. And even if it was well known that student representatives re-
ceived patrimonial benefits, many of these benefits were directed to-
ward their organizations, not for their personal gain.

Soifer_6844-final.indb 150 8/17/18 11:53 AM

 Political Violence in Peru’s Public Universities 151

The violence and radicalism of the 1980s further delegitimized pol-

itics within universities. On the one hand, as discussed, violence itself
weakened student organizations. The suspicion over these groups, and
the way political elites brought other radical groups and Sendero Lu-
minoso together, increased the risk of active engagement in student
unions. But a more widespread problem was that student organiza-
tions were frequently linked in the public discourse to this radicalism.
University politics was associated with violent radicalism, even if these
groups had important differences with the PCP-SL and frequently op-
posed it. The abstract radicalism of university politics was difficult to
distinguish from active radicalism of the kind promulgated by Sen-
dero. Student unions lost representativeness among a student body
tired of ideological and violent discourses. As evidence of the lack of
patience with this radicalism, a considerable number of students sup-
ported the government intervention in 1996.
In the 1990s student political organization was disarticulated and
lost relevance. There was a gradual detachment of representatives in
governing bodies from the weakened student representative organi-
zations. By the mid-1990s, just a few student federations remained
(J. Chávez 1999, 56). Their activities were focused on recreation,
sports, and services, leaving behind political activism (Venturo 2001,
99). Since 1997–1998 we have witnessed a growing interest in national
politics through democratic protests against fujimorismo, but no sig-
nificant change in regard to participation in university politics. After
the democratic transition, these public protests declined again (Disi
2017, 172–173). Students, in general, do not trust and are not active
in their representative institutions, nor in student organizations. This
internal weakness has obviously affected students’ capacity to mobi-
lize in the democratic arena in favor of university reform or other po-
litical issues.
In this vacuum created by the legacies of the internal confl ict, a new
type of student representative has emerged. Students who take part in
representative institutions tend to be independent actors that negoti-
ate their support with university authorities. These negotiations are
quite different from those before or during the 1980s, when radical
groups occupied these positions; representatives now demand and re-
ceive benefits in return for supporting authorities (interview with Ger-
maná; N. Chávez 2014). These clientelistic coalitions fear meritocratic
and quality-centered reforms that could weaken their control over re-
sources. As a result, few real reform efforts were conducted or de-
manded within universities and in the public sphere after the demo-

Soifer_6844-final.indb 151 8/17/18 11:53 AM

 152 Eduardo Dargent and Noelia Chávez

cratic transition. In the recent upsurge of university protests for quality

education in Latin America, Peru is (again) a negative case.
Two recent case studies show the strength of this patrimonialism
among student representatives. When documenting the problems faced
by Quechua-speaking students in UNSCH, Villacorta (2012, 179–
180) fi nds that authorities seek the support of representatives by pro-
viding them with benefits such as better access to the students’ res-
taurant and dormitories, and by providing scholarships or material
goods.10 Noelia Chávez (2014) fi nds a similar situation at the Univer-
sidad Nacional de la Amazonía Peruana (UNAP) in Iquitos. The uni-
versity does not have a student federation, and representatives in gov-
erning bodies are closely associated with authorities. The authorities
fi nance these students’ campaigns in exchange for their support once
elected. Their handouts include money, laptops, cell phones, and plane
tickets for academic congresses. In general, students do not trust or en-
gage with political groups acting within universities. Political groups
competing for the control of student unions are now mainly radical or-
ganizations such as Patria Roja and the even more radical MOVADEF.
These groups compete every two years for control over the national
federation, Federación de Estudiantes del Perú. The majority of stu-
dents are little interested in these contests.
Academic works since the nineties underscore the lack of interest in
university politics among public university students. A research proj-
ect about political participation of young people in twenty-five univer-
sities between 1996 and 1997 shows how negatively ideas and political
groups within universities were viewed (J. Chávez 1999). The students
interviewed considered university politics to be corrupt and negative
for public universities. Most of them believed that those involved in
university politics were seeking their own material good or other goals
not related to improving academic quality (J. Chávez 1999, 96). The
prevailing view was that student representatives are corrupt, bought
by authorities or political groups.
The case of San Marcos shows some of these trends. The university
does not have a federation representing the entire student body. Fur-
thermore, recent studies (for example, Nureña et al. 2014) show that
only 13 percent have a strong interest in university politics, 16.4 per-
cent have taken part in student assemblies, and 13.8 percent are part
of a political organization. Similarly, Jave, Cépeda, and Uchuypoma
(2014, 102–107) fi nd a strong contrast between students’ interest in
national politics and other collective activities (study groups, politi-
cal activism, cultural activities) and university political participation.

Soifer_6844-final.indb 152 8/17/18 11:53 AM

 Political Violence in Peru’s Public Universities 153

By weakening student organizations and building a strong stigma

against student politics, political violence of the eighties seems to have
affected students’ political engagement within the universities. And
the weakness of student associations and collective action in general
helps to explain the comparatively low student mobilization in Peru.
This weakness becomes more striking when we compare the Peruvian
case to Chile, Colombia, and Mexico, countries where student associ-
ations frequently take part in protests to demand better education and
conditions in universities (Disi 2017).
Yet we are aware that there are good reasons to consider the re-
lationship between political violence and the current disengagement
of students from university politics as tenuous. The patrimonialism,
radical politics, and corruption at the root of this disenchantment ex-
isted before the eighties. Some scholars have questioned whether there
continues to be a link between violence and lack of political interest
or participation. Focusing on San Marcos, Nureña (2016) argues that
we should not confuse lack of participation in internal politics with
lack of political interest. San Marcos, he argues, is considerably po-
liticized, or at least as politicized as other sectors of Peruvian society.
Political participation there occurs through less traditional organiza-
tions (e.g., academic circles, cultural groups). The lack of political par-
ticipation within universities is better explained by a range of mecha-
nisms used by authorities and current student representatives to close
and limit students’ participation (e.g., clientelism, corruption, formal
requirements), and also by the perception of many students that po-
litical representatives are ineffective, radical, or corrupt, not necessar-
ily because they think university politics is associated with radicalism
(Nureña 2016, 124).
More generally, the current weakness of political parties and civil
society associations in Peru makes it difficult to expect high politi-
cal engagement from university students within or outside of universi-
ties. Universities seem to be a microcosm of larger trends seen in Peru-
vian society. In his comparative study of student mobilization in Latin
America, Disi proposes that links between student movements and op-
position parties are an important factor.11 In Peru, his negative case,
the weakness of leftist parties limits the possibility of student mobi-
lization. Thus there are good reasons to think that even without vio-
lence, student participation in internal politics or in the public sphere
will be low in Peru.
Nonetheless, though we agree that more research is necessary to
confi rm this link and that, as shown by Nureña, there are important

Soifer_6844-final.indb 153 8/17/18 11:53 AM

 154 Eduardo Dargent and Noelia Chávez

caveats to consider, we believe violence is still a relevant factor in com-

parative perspective, and not only due to its impact on student orga-
nization. The perception of university students as radicals or violent
is a legacy that certainly affects and delegitimizes student organiza-
tions, and stigmatizes political activism in a more fundamental way
than in other societies. The weakness of student associations, partly
attributable to violence, also limits students’ participation in the pub-
lic sphere. The counterfactual argument we propose is that without vi-
olence we could expect more internal politics, as well as more relevant
expression of students’ concerns in the public sphere. More research is
needed to properly assess this causal claim.

Conclusion

This chapter has shown that the violence of the Internal Armed Con-
flict had a strong and independent effect on public universities, leaving
strong legacies that persist to this day: an image of violence and radi-
calism; sizable effects on student and faculty composition, with impli-
cations for academic quality; and obstacles to students’ capacity for
collective action. By limiting the possibilities of external and internal
reform efforts, these legacies contribute to sustaining the current crisis
condition of public universities. If our diagnosis about these legacies is
right, then those interested in the reform of public universities should
pay attention to two issues.
First, any serious reform plan must take into consideration the
strength of the negative and violent image of public universities among
the population in order to mobilize popular support and legitimacy.
The image of radicalism is not ingrained only among elites. The mem-
ories of the 1980s and the repeated discourse in which public univer-
sities are perceived as lairs of terrorism have rooted this idea in the
broader population. Thus, reformists must come to terms with this
fact in their efforts to show the population that the deeper problems in
public universities relate to clientelism, corruption, academic medioc-
rity, and lack of funding rather than to radicalism.
Second, there seems to be a silent majority within the public uni-
versities waiting to be politicized. The anger, frustration, and political
disengagement of a majority of students seems a fertile ground to ad-
vance new ideas about the benefits of reforming public universities. But
advancement will require a very different student politics that breaks

Soifer_6844-final.indb 154 8/17/18 11:53 AM

 Political Violence in Peru’s Public Universities 155

from current clientelistic patterns. Reformist academic authorities may

fi nd in students the allies they need to break down resistance to reform.
The current support of some student organizations for the ongoing uni-
versity reform is an optimistic sign in the midst of the many problems
and challenges of Peruvian public universities.

Interviews

Burga, Manuel. Former dean of UNMSM. Lima, January 16, 2014.

Germaná, César. Professor, Social Sciences Department, UNMSM. Lima,
January 20, 2014.
López, Sinesio. Professor, Social Sciences Department, UNMSM. Lima, Jan-
uary 16, 2014.
Méndez, Cecilia. Former professor of history in UNSCH, current professor
in UC Santa Barbara. Santa Barbara, California (interviewed via Skype),
March 11, 2014.
Mori, Jorge. Former student activist at UNMSM and expert in higher educa-
tion. Lima, February 8, 2014.
Sala, Núria. Former professor of history in UNSCH. Lima, March 7, 2014.
Sandoval, Pablo. Profesor, Anthropology Department, UNMSN and re-
searcher in the Instituto de Estudios Peruanos. Lima, January 28, 2014.

Notes

1. The relevance of public universities within the Peruvian conflict is ex-

emplified in the fi nal report of the Truth and Reconciliation Commission
(Comisión de la Verdad y Reconciliación, or CVR), which dedicates a the-
matic chapter and four case studies to the universities.
2. In Peru, it is quite common to have professors that are only part-time
lecturers; they do not make careers within universities but teach there out of
vocation and/or for prestige.
3. These universities were UNMSM, Universidad Nacional de Educación
Enrique Guzmán y Valle “La Cantuta,” Universidad Nacional Federico Vi-
llarreal, Universidad Nacional San Luis Gonzaga de Ica, Universidad Nacio-
nal Faustino Sánchez Carrión de Huacho, and Universidad Nacional Hermi-
lio Valdizán de Huánuco.
4. A second goal was to expel the so-called eternal students, who profited
from lenient rules to remain registered for many years and often were politi-
cal leaders.
5. For a review of neoliberal reforms in higher education in Latin Amer-
ica, see Schwartzman 1996.
6. “Con Alegría Juvenil por Tercer Día Marcharon Universitarios,” La
República, June 5, 1997.
7. Some more examples of these headlines and reports: “Avanza la Infi ltra-

Soifer_6844-final.indb 155 8/17/18 11:53 AM

 156 Eduardo Dargent and Noelia Chávez

ción del Movadef en las Universidades del País,” Perú 21, September 11, 2012;
“Toman a la Ligera Infi ltración del Movadef en las Universidades,” La Razón,
April 4, 2014; “Terroristas del Movadef se Infi ltran en más Universidades,”
La Razón, February 6, 2014; “ANR: El Movadef está en las Universidades,”
El Trome, November 10, 2012; “Movadef y Simpatizantes de Abimael Guz-
mán Regresan a San Marcos,” Diario Correo, August 23, 2013.
8. Jeffrey Gamarra, a professor at UNSCH, argues that students in this
university today are quite different from those of the 1980s; he says they are
much more interested in technical careers than in any sort of radical activities
(Gamarra 2012).
9. Jave, Cépeda, and Uchuypoma (2014, 132) show the extent to which so-
cial science faculties at San Marcos and UNSCH have been stigmatized.
10. Villacorta also documents a series of problems mentioned previously,
including lack of meritocratic faculty recruitment, poor quality, and abuses
against students.
11. Disi explains mass mobilizations in Chile and other Latin American
countries as being due to increased enrollment of lower-income students in
higher education, eased by the availability of bank credits that can be repaid
by students after their graduation. These costs, coupled with the difficulty
graduates of low-quality institutions have in fi nding jobs and increased griev-
ances about academic quality, mobilize students to demand reforms. Disi ar-
gues that Peru did not witness a similar process mainly because lower-income
students accessed higher education through low-cost private institutions and
there were no easily available bank or public credits. Also, in contrast to
Chile, public universities were free of charge. The author argues that “the dif-
ferences in mobilization between Chile and Peru can be attributed to a great
extent to the relative absence of fi nancial grievances among Peruvian college
students” (Disi 2017, 169). Also important, as mentioned, is the lack of con-
nections with political parties.

Soifer_6844-final.indb 156 8/17/18 11:53 AM

 CHAP T ER 6

Peace for Whom? Legacies of

Gender-Based Violence in Peru
Jelke Boesten

In August 2016, a multitude of women, their families, and their friends

took to the streets of Lima to protest the high levels of violence against
women in Peru and the impunity routinely accorded to the perpetra-
tors of this violence. Never before had so many Peruvians protested vi-
olence against women, even if there had been ample reason to do so.
In this chapter, I will explore why this mass mobilization happened at
that particular point in time by examining the extent to which the vio-
lence against women in 2016 might be interpreted as a legacy of the vi-
olence of the Internal Armed Conflict (IAC) or as a result of persistent
historical structures of violence and inequality. I also consider whether
the contemporary response to such violence from both civil society ac-
tivists and the state should be seen in light of the continuous battles
over truth, justice, and reconciliation.
In exploring the hypothesis that the contemporary violence against
women is a legacy of a much longer history of violence and inequality, I
will focus in particular on what aspects might be seen as a sequel to the
Internal Armed Conflict. I will ask if high levels of peacetime violence
might be seen as either a wartime mechanism or a post-conflict legacy.
To examine this, I draw from my research in the archives of the Truth
and Reconciliation Commission and other sources for my book Sex-
ual Violence during War and Peace: Gender, Power, and Post-Conflict
Justice in Peru (2014). But I am also interested in exploring how the
lack of justice and visibility regarding cases of conflict-related violence
against women contrasts with the more recent mobilization of hun-
dreds of thousands of people to protest against continuous high lev-
els of violence against women. I argue that perhaps historic cases are
too politically and socially divisive to work as examples that promote

Soifer_6844-final.indb 157 8/17/18 11:53 AM

 158 Jelke Boesten

broader gender justice; instead, it may be that the struggle against the
everyday violence women and girls experience across lines of class, eth-
nicity, geography, and age has fi nally found its historic momentum,
with capable activists to lead the way and a political opportunity to
rise to the challenge of demanding justice and social change.

High Levels of Violence against Women

as Sequel and Continuum

As is now widely known but not necessarily widely accepted in and be-
yond Peru, gender-based violence was an important dimension of the
political violence that enveloped Peru in the 1980s and 1990s. Gender-
based and sexual violence, understood as violence perpetrated for spe-
cifically gendered reasons (i.e., because of being a woman, or because
of being queer, or employing sexual violence), was used by all the
armed groups involved in the IAC: Shining Path, MRTA, the Peruvian
military, and indeed the peasant self-defense forces. It is slowly com-
ing to light that the MRTA, and possibly the Shining Path as well, tar-
geted queer men and women in the Loreto region (CVR 2003a, 2:432–
433). In addition, the Shining Path stands accused of mass enslavement
of the Ashaninka people of the Amazon, including sexually enslaving
Ashaninka women and girls (CVR 2003a, vol. 5). The use of different
forms of gender-based violence within the Shining Path still requires
research, as there is evidence of forced marriages, rape, and infanticide
(CVR 2003a, vol. 6). These incidences of gender-based violence are
all expressions of male dominance in precarious and contested spaces.
Most notable is the use of sexual violence on the part of the Peru-
vian military against civilians as well as people suspected of terror-
ism. This topic is comparatively well researched (Henriquez and Man-
tilla 2003; Henriquez 2006; Boesten 2014a). Drawing on testimonies
of witnesses, perpetrators, and survivors of sexual violence, the Truth
and Reconciliation Commission (Comisión de la Verdad y Reconci-
liación, or CVR) found that sexual violence was widespread and even
systematically used against populations suspected of terrorism, as
well as against civilians (CVR 2003a, vol. 6). The CVR documented
538 cases of rape in which the victim-survivor could be identified by
name and surname, but estimated that this represented only 7 percent
of actual cases of rape. The victim profile of this violence reflects geo-
graphical, ethnic, class, age, and gender divides: young, rural, and in-

Soifer_6844-final.indb 158 8/17/18 11:53 AM

 Legacies of Gender-Based Violence in Peru 159

digenous women were most vulnerable to sexual violence and abuse.

But this was not the only vulnerable group: according to the CVR,
80 percent of women imprisoned in Lima experienced sexual assault
and abuse from security personnel, and 25 percent were raped.
Based on the data, the CVR concluded that members of the police
and armed forces were the main perpetrators of systematic rape. While
we must recognize the need for further research on the use and prac-
tice of gender-based violence by all armed groups in the conflict, it is
worth noting that state forces were the main perpetrators of such vio-
lence. This is especially important because it is the state’s function to
provide security for its citizens, because it points to the state as a major
agent in reinforcing and reproducing existing inequalities, and fi nally,
as will be discussed later, because none of the perpetrators of such vio-
lence have been held to account. The data confirms that gender-based
violence is highly political, not only in war, but also in peacetime.
The patterns of rape perpetrated by state forces reflect the wide-
spread idea that rape is used as a weapon of war. The military used
sexual torture in its prisons and rural military bases, raped women
and girls during raids in homes and villages as a means to terrorize the
population, and used women and girls as booty and as tools for a pol-
itics of divide-and-rule between communities as well as to establish
and maintain hierarchies between soldiers (Henriquez 2006; Boesten
2014a). Hence rape was used, albeit not formally, as “part of a system-
atic political campaign with military purposes,” that is, as a weapon
of war (Skjelsbæk 2010).
It is difficult to establish whether soldiers were ordered to rape pris-
oners or civilians, and it is unlikely that evidence of such orders will
ever be found. As I have argued (Boesten 2014a), testimonies of vic-
tim-survivors and others suggest that specific military masculinities,
which rely heavily on performances of heteronormative dominance
and the use of extreme violence and sexual violence as social capital
within military hierarchies, facilitate the widespread practice of sex-
ual violence. For example, hierarchies between more and less available
women based on perceptions of race and class—a hierarchy of sexual
availability which draws strongly on existing societal stratifications—
was used to establish and reproduce hierarchies between soldiers. Ac-
cording to testimonies from soldiers, young poor indigenous women
were available to be raped by the troops, even if they were “just” com-
munity members, civilians, and not suspected of any political activ-
ity. But women who were captured on suspicion of terrorism and who

Soifer_6844-final.indb 159 8/17/18 11:53 AM

 160 Jelke Boesten

were seen as whiter or better educated than the cholitas from the vil-
lage would fi rst be raped by the commander in charge. The men high-
est in rank could choose among the prisoners, and could even choose
to keep women and “enamorarlas” (seduce them) for longer periods of
time. Such abusive behavior from military officers clearly encourages
the rank and file to do the same.
There are further indications that sexual violence against local
women and girls was turned into a practice that created social capital
within the armed forces: girls were ordered to come to military bases
and prostitute themselves during parties, girls were forced to negoti-
ate with soldiers about the conditions of their abuse, such as whether
they would be raped by one or two soldiers rather than gang raped, or
the reward for being abused, such as receiving information about loved
ones. In such ways, sexual dominance over local girls may have given
these soldiers certain masculine credibility in the eyes of their peers
and, especially, their superiors. There is evidence of soldiers boasting
to each other about their conquests and their violence; there is evi-
dence of collective pornographic spectacles in which soldiers violently
gang raped girls and women, dead or alive, watching each other and
spurring each other on (Boesten 2014a, 19–42). Such experiences sug-
gest a process of drawing soldiers into a spiral of brutalization (Mit-
ton 2015), using the most “obvious” victims—the most vulnerable to
abuse—as a tool. From such a perspective, the sexual abuse of women
served as a means not only to terrorize, fragment, and dominate the ru-
ral population or the prison population but also to foment a loyal army
of young men willing and able to continue to perpetrate atrocities.
The idea that gender-based violence during the conflict reflects the
patterns of inequality that existed before and after the war is impor-
tant: it suggests that such violence is an exacerbation of existing re-
lationships. The abuse of women during the war may have been ex-
tremely cruel and very widespread, but at the same time, such violence
was imaginable, possible, and scripted along lines of existing violence
and inequalities. This means that some women are more vulnerable
than others to harassment by men in public spaces, or in the work-
place, because of vectors of inequality such as race/ethnicity, class, and
age. But all women are vulnerable to gender-based violence in their
own homes.
There is no data concerning levels of violence against women before
the 1980s, but women reported to the CVR that during the confl ict the
armed forces were not the only perpetrators of violence against them.

Soifer_6844-final.indb 160 8/17/18 11:53 AM

 Legacies of Gender-Based Violence in Peru 161

Henriquez and Mantilla (2003) show how violence against women in

their own homes and communities often escalated in zones with high
levels of political conflict. We do not know whether that is caused by a
breakdown of social order, opportunism, or the frustration and trau-
matic context in which families were forced to live. What we do know
is that most men and women in the highly affected areas were compro-
mised by the violence, either through victimization or recruitment by
the Shining Path, the self-defense forces, or the armed forces (Theidon
2012). If we add that fact to the idea of what armed groups do to the
young men and women who become involved in the confl ict, including
the possibility of brutalization among those forced into spirals of vio-
lence, then perhaps the rise of domestic violence in a context of par-
ticipation in political violence is not surprising.1 Following the analyt-
ical framework set out by Soifer and Vergara in the Introduction to
this volume, the high levels of gender-based and family violence in ar-
eas most affected by the armed conflict of the 1980s and early 1990s
could be called a “wartime mechanism.” However, despite the lack of
verifiable statistics, violence against women, including sexual violence,
also seems to have been high before the armed confl ict.
Violence against women in wartime is not the same as violence
against women during peacetime, but there certainly is a continuum.
The difference lies in the extreme cruelty and frequency of wartime vi-
olence, as well as the drawing in of those who might otherwise not
have become victims or perpetrators of gender-based violence. But
there are also similarities: in wartime and peacetime, women and girls
perceived as being of lesser value because of race and class—often de-
termined by criteria such as citizenship status (does she have papers?),
poverty, geography, language, education, dress and physical aspects
such as height, color of eyes, and type of hair—are more vulnerable
to rape. Young women perceived as being cholas—of indigenous de-
scent—who work in wealthier households are historically perceived as
legitimate targets for men’s sexual satisfaction (Boesten 2014). Like-
wise, young women, or rather, adolescent girls—the main victims of
soldiers’ sexual abuse—are also the main victims of contemporary
peacetime sexual abuse (Mujica 2011).
The scale of peacetime violence should also be taken into consid-
eration: according to data from the Programa Integral de la Lucha
Contra la Violencia Familiar y Sexual in Ayacucho (2005), physical
violence against children is common, and sexual violence against chil-
dren rampant (see also Boesten 2010, 148). Current figures of violence

Soifer_6844-final.indb 161 8/17/18 11:53 AM

 162 Jelke Boesten

against women collected by the National Bureau of Statistics through

a household survey method show that in 2014, 32.3 percent of women
experienced physical violence perpetrated by their intimate partner at
least once in their lifetime, 7.9 percent experienced sexual violence,
and a staggering 72.4 percent experienced psychological and/or verbal
violence, often in addition to other forms of violence (ENDES 2014).
A study that looked at police statistics and data from women’s emer-
gency centers (one-stop multisectoral centers set up in the 1990s to
report abuse and seek support) and the Ministry of Women and So-
cial Development for the years 2000 to 2009 concluded that on aver-
age, about 7,000 cases of sexual violence were reported each year to
such institutions. However, in 2009 the Institute of Forensic Medicine
alone carried out 34,153 exams of “sexual integrity,” examinations
of women and girls who reported sexual abuse but whose claims were
not entered into the statistics of the police, emergency centers, or the
ministry. Seventy-eight percent of the formally reported denunciations
concerned girls eighteen years old or younger. Of these, 10 percent
concerned children aged nine and under; 25 percent involved children
aged ten to thirteen; and 45 percent represented adolescents aged four-
teen to seventeen (Mujica 2011). Wartime mechanism or not, these are
extraordinary figures which raise the question of whether peacetime is
truly peaceful for women and girls.
One conclusion could be that a high rate of post-conflict violence
against women is a consequence of confl ict-related violence. Epidemiol-
ogists agree that a range of factors contribute to high levels of interper-
sonal violence, particularly sexual and gender-based violence, which
tend to be more prevalent in post-conflict societies. These factors in-
clude high levels of trauma among both men and women, alcoholism
and drug use, violence against children in their homes, and the preva-
lence of “toxic” masculinities (Gould and Jewkes 2013; Guedes et al.
2016). In Peru, NGOs, civil society organizations, and state institu-
tions such as Women’s Emergency Centers in Ayacucho cite trauma
and alcohol and drug use as strongly affecting levels of violence against
women and children (Boesten 2014a, 127). Violence against children
as a form of disciplining them, i.e., as communication, was also re-
ported to be high throughout the 1990s and early 2000s (Boesten
2010, 148). According to the National Household Survey from 2014
(ENDES 2014), 25 to 29 percent of parents still use corporal punish-
ment with their children. But the practice of corporal punishment (and
the effects it may cause) is more likely attributable to historical-cultural
factors rather than political violence.

Soifer_6844-final.indb 162 8/17/18 11:53 AM

 Legacies of Gender-Based Violence in Peru 163

The idea of “toxic” masculinities—understandings of manhood as

necessarily violent, controlling, and (hetero)sexually predatory—has
gained traction among feminist researchers and epidemiologists. The
prevalence of toxic masculinities is often associated with histories of
violence and exclusion. For example, South African scholar Pumla Di-
neo Gqola (2015) asserts that a history of violence, dispossession, and
exclusion experienced by large parts of that nation’s population has
created a situation in which asserting dominant masculinities could
become a form of resistance against the violence of the system, as well
as the preferred form in which the dominant group asserted its power
within its own group and against the marginalized population. Other
scholars, such as Chris Dolan (2003), who examines masculinities in
post-confl ict Uganda, or Kimberly Theidon (2009), who looks at men’s
reintegration in Colombia, suggest that the undermining of men’s roles
as breadwinners and heads of households has created a masculine cri-
sis often compensated for through violent and controlling behavior, es-
pecially against more vulnerable groups such as women. These studies
all indicate that the social, political, and economic history of post-
conflict societies shapes social relationships, and the prevalence of
sexual and gender-based violence, in the present. Thus high levels of
gender-based violence could be seen as a post-confl ict legacy.
But in Peru, a complex historical dynamic of violence against and
dispossession of the indigenous population forms the background to
continuing high levels of gender-based violence. In patriarchal socie-
ties, women have long been subordinated to men, with certain male
privileges—including the right to rape one’s wife—enshrined in law
until fairly recently. Furthermore, in postcolonial societies, whiter
men’s dominance over indigenous people continues to shape women’s
vulnerability to racialized and sexualized violence in their homes, com-
munities, and places of work. Peruvian literature is littered with nar-
ratives of the abuse of chola domestic servants, largely perceived as le-
gitimate sexual targets for coming-of-age of adolescent boys and as
legitimate targets for men in venting their sexual frustration. The vul-
nerability of this group of women was recently attested to by Obdu-
lia Guevara Neyra, the general secretary of the Union of Domestic
Workers-Lima (SINTTRAHOL), who claims that 60 percent of do-
mestic workers today experience some form of sexual abuse.2 The ex-
perience of gender-based violence, then, runs along a continuum from
cat-calling to emotional, physical, and sexual violence in homes, com-
munities, and workplaces, and to rape and femicide in war and peace-
time. The continuum also runs through history, from colonial times to

Soifer_6844-final.indb 163 8/17/18 11:53 AM

 164 Jelke Boesten

modern Peru. Perpetrators may be powerful men who feel entitled to

women’s bodies of all colors, or they may be subaltern men who feel
entitled to women and girls still more vulnerable than themselves. An
authoritarian political culture grounded in patriarchal and racist rela-
tionships across society fuels these toxic masculinities and perpetuates
the vulnerability of women and girls.
Thus, it is fair to say that the widespread sexual violence during the
Internal Armed Conflict in Peru showed parallels to sexual violence in
peacetime. Wartime rape worked as a means to reproduce and perpetu-
ate existing historical inequalities based on gender, race, class, age, and
sexuality. Contemporary peacetime sexual violence does so as well,
but in a less obvious and more privatized manner. Post-conflict socie-
ties such as Peru, but also Guatemala or South Africa, show high lev-
els of interpersonal violence, including sexual violence, and this is cer-
tainly related to wartime histories of violence and confl ict. But while
high levels of peacetime violence are not necessarily a direct sequel
to a particular conflict (although they may be in specific communi-
ties or families), we should recognize the broader role and function of
gender-based violence, and particularly of sexual violence, in reproduc-
ing structures of inequality in war and in peace, before, during, and af-
ter armed conflict. In this light, the idea of gender-based violence as a
wartime mechanism or post-conflict legacy is difficult to sustain.
Explicit efforts to reach reconciliation and justice among former en-
emies make post-conflict eras an important opportunity to address
historical hierarchies such as those associated with ongoing sexual vi-
olence. Transitional justice could, and should, transform the social re-
lations that feed into violence, providing the state and society a chance
to address gender inequality and violence associated with it. In that
sense, the way in which the state and society deal with confl ict-related
sexual violence in the post-conflict moment could start a process of
positive change for women and men. The CVR did the best it could
in the limited time it had to uncover the truth about the systematic
perpetration of sexual violence by all armed groups. Likewise, the in-
stitution tasked with documenting victims of violence, the Registro
Único de Víctimas, did an important job in recording cases of sex-
ual and gendered violence. Thus, the information needed to build a
policy of accountability exists, and it is the responsibility of the state
to acknowledge and account for what was done in its name. A for-
mal apology on the part of the state, the military, and the police to all
women and men harmed by the sexual violence would send a power-

Soifer_6844-final.indb 164 8/17/18 11:53 AM

 Legacies of Gender-Based Violence in Peru 165

ful message heralding a new era in which gender-based violence is no

longer acceptable, and paving the way for effective judicial and policy
interventions to address wartime victimization as well as continuing
widespread violence in contemporary Peru. Unfortunately, this has not
happened; perpetrators of sexual violence still evade punishment, and
patriarchy is alive and well.

Patriarchy and Impunity

The protests of August 2016 in Peru were triggered because men con-
tinue, with impunity, to inflict violence on women. Criminal justice
may not solve the root causes of gender-based violence, but the prob-
lem will certainly persist if perpetrators are consistently getting away
with violence and abuse. Two particular cases of unpunished violence
against women entered the public consciousness in July 2016. In both
cases, the courts decided to give the perpetrators suspended sentences
after deciding the forensic evidence suggested the harm done was “mi-
nor.” In the fi rst case, in Ayacucho, a woman named Arlette Contreras
was assaulted by her boyfriend, who beat her up when she tried to flee
the hotel where he had taken her. In a video that was widely circulated
on the Internet, footage from security cameras showed the boyfriend,
Adriano Pozo, naked, dragging Contreras by her hair through the re-
ception area. Pozo was captured and put on trial, but a judge gave him
a suspended sentence and a fi ne, arguing that the nature of Contreras’s
wounds suggested that he had not intended to rape or kill her, and had
inflicted only minor harm.3 The sentencing judge, a woman, was an
acquaintance of Pozo’s father, an Ayacuchano governor.
The second case was that of Lady Guillén, a celebrity in the world
of cumbia, who endured assaults by her boyfriend for a year before re-
porting the violence when she felt her life was in danger. Photos of her
beaten, disfigured face, with stitches around her eyes, were widely cir-
culated in newspapers since 2012. But in 2016, the judge in the case
considered Guillén’s wounds minor and decided that her life had not
been in danger.4 The aggressor, Ronny García, received a suspended
sentence after having spent several years in pretrial imprisonment. Re-
leased, García now stands accused of violence by a new girlfriend.
These two emblematic and well-publicized cases were matched by sim-
ilar cases throughout the country. Women and girls were disfigured,
raped, and killed by their partners or former partners, and their cases

Soifer_6844-final.indb 165 8/17/18 11:53 AM

 166 Jelke Boesten

were reported in small corners of the printed media, but justice was
generally not done. The massive protests of 2016, never seen before
in the history of activism against gender-based violence in Peru, led to
new investigations of the Contreras and Guillén cases, but many other
cases have yet to receive the attention they merit.
As I have detailed elsewhere (Boesten 2012; Boesten 2014a), im-
punity for violence against women is generally high in Peru. Despite
the presence of well-established laws and protocols, there are mul-
tiple problems in the Peruvian state’s response to sexual violence. A
lack of training in gender awareness and recognition of sexual vio-
lence continues to debilitate the police, judiciary, and forensic medi-
cine services designed to support battered women. A lack of adequate
funding for services that work, or should work, such as women’s po-
lice stations and emergency centers, undermines the promise these ser-
vices embody. A lack of political collaboration among the district and
municipal authorities who are in charge of allocating funding to some
of these services further impedes them, while the lack of a sufficiently
independent and effective judiciary constrains gender justice. Overall,
the dissonance between a relatively good legal and policy framework
on the one hand and weak implementation on the other stems from the
unresolved tension between a patriarchal state which puts the male-
headed family at the center of all considerations, and the need for a
policy that treats women’s rights as inalienable.
As in many parts of Latin America, the role of a conservative Cath-
olic Church is particularly harmful in Peru. The progressive Libera-
tion Theology that emerged in the late 1960s largely disappeared
with the demise of the Left in the 1990s, and as Vergara and Enci-
nas show in chapter 9 of this volume, the more conservative sectors
of the Church gained a more central presence. Opus Dei, led by Car-
dinal Luis Cipriani, archbishop of Lima, is now the dominant Chris-
tian voice and holds tremendous power over politics and institutions,
particularly with regard to gender politics. Recently, in response to
campaigns for the decriminalization of abortion in cases of rape, as
well as in response to the demonstrations of August 2016, Cipriani
has spoken out against what he calls the danger of the “gender ideol-
ogy,” wielding his influence to denounce campaigns that could unset-
tle the Church’s patriarchal power. While it is difficult to sustain a po-
sition that endorses violence—indeed, a whole range of agencies, state
institutions, and businesses supported the August 2016 march against

Soifer_6844-final.indb 166 8/17/18 11:53 AM

 Legacies of Gender-Based Violence in Peru 167

gender-based violence—it appeared possible to defend heteronorma-

tivity and the “sanctity of life” by accusing women and girls of pro-
voking sexual harassment and even rape.5 Hence, patriarchy is not the
archaic concept that it should be by now, but rather, continues to build
and rebuild its historic momentum.
The persistence of patriarchal relationships, underpinned by a nor-
mative belief in a natural gender inequality, stands in contrast with
the increasing number of women in higher education, in formal em-
ployment, and in national and local politics. Whereas many scholars
judge women’s representation in national politics during the 1990s
as affi rming gender stereotypes rather than unsettling them (Blondet
2002; Schmidt 2006; Boesten 2010), women’s participation has be-
come consolidated since the start of the twenty-fi rst century, if not
without encountering resistance (Krook and Restrepo Sanín 2016). In
2015–2016 we saw the rise of a new generation of left-of-center female
politicians: Verónika Mendoza stood for president for the Frente Am-
plio, and Marisa Glave and Indira Huilca proved to be strong voices in
Congress in support of human rights. Mendoza, Huilca, and Glave are
unafraid to defend women’s rights and to discuss sexism in politics,
representing a new feminist voice in national politics. Considering the
increased prominence of women in public life and the gap in progres-
sive politics that is now being filled by a new generation of feminist
politicians and activists, perhaps the pushback from conservative sec-
tors in society should be seen as a response to that increasing cry for
gender equality. For example, the newly approved school curriculum
for 2017, which includes gender equality as an objective, was consid-
ered to promote gender “ideology” and homosexuality, undermining
the family as cornerstone of society according to conservative sectors,6
and was given as a reason to force the widely popular and acclaimed
education minister Jaime Saavedra to resign in December 2016.
This tension between increasing gender equality on the one hand
and persistent patriarchal attitudes in some institutions and sectors on
the other hand is what allows for impunity from punishment for sex-
ual violence to persist, but it also allows for protest, in Peru and in-
deed throughout Latin America. In Peru, this tension is also coming to
the fore in relation to historic crimes against humanity, particularly re-
garding criminal justice in relation to sexual violence perpetrated by
the Peruvian military in the 1980s and truth fi nding and political ac-
countability in the case of forced sterilizations in the 1990s.

Soifer_6844-final.indb 167 8/17/18 11:53 AM

 168 Jelke Boesten

Truth, Justice, and Reconciliation; or, Struggles over Legacies

When the CVR concluded that sexual violence in the Internal Armed
Conflict had been systematic, particularly on the part of the armed
forces, it also allowed for several cases to be identified and investi-
gated more thoroughly in the interest of seeking criminal justice. By
2012, human rights organizations had investigated and presented six-
teen cases to the public prosecutor. In Febrary 2018, three military fig-
ures were convicted and given prison sentences for the kidnapping and
rape of a student in 1992. This fi rst positive result is a landmark rul-
ing, and it may help the emblematic case that is currently under way
against eleven more former military personnel.7 This case, commonly
known as Manta y Vilca after the two communities affected, concerns
fourteen complainants.8 The hearings before the National Criminal
Court started in July 2016 and could last for several years.
It has taken thirteen years since the publication of the CVR report
for this case to come to court. There are several reasons for this slow
process, and of course for the continuing impunity in most other cases
of conflict-related sexual violence. First, there is an overall reluctance
to prosecute military officers for violations of human rights. The Peru-
vian military and police force commanded counterinsurgency efforts
against two very destructive insurgent groups, and for obvious rea-
sons they do not like to be criticized for their actions. In addition,
while the transitional government of President Valentín Paniagua in
2000–2001 purged the relevant institutions (military, National In-
telligence Service, judiciary, and electoral committee) of corrupt and
violent officials who had upheld the Fujimori regime (Taylor 2005),
this did not prevent the election of Alan García or Ollanta Humala.
García, in power from 2006 to 2011, was also the president in 1986
when the navy was sent in to suppress a prison uprising in El Frontón,
where at least ninety senderistas were killed in extrajudicial execu-
tions. Ollanta Humala served in the military in the 1990s and was ac-
cused of being involved in human rights violations in Tingo María.
Electoral support for García and Humala, as well as for Keiko Fu-
jimori, the still-popular daughter of Alberto Fujimori, indicates that
perhaps there simply is not much governmental or popular support for
positions which seek to address the excesses of the counterinsurgency
of the 1980s and 1990s, including prosecution of former military fig-
ures for violations of human rights.

Soifer_6844-final.indb 168 8/17/18 11:53 AM

 Legacies of Gender-Based Violence in Peru 169

Of course, the military and the Ministry of Defense also influence

what is possible in terms of who can be prosecuted and who cannot,
and what evidence is available and what is not. As such, as soon as hu-
man rights organizations started to investigate specific cases of mili-
tary rape based on CVR testimonies, the Ministry of Defense refused
to release documents which could reveal the identity of specific sol-
diers in specific times and places. It claimed these particular archives
had burned down.
Furthermore, a lack of resources and investigative capacity in hu-
man rights law impedes the adequate prosecution of cases of sexual vi-
olence. Cases of sexual violence are notoriously complex in any con-
text; hence the relatively low conviction rates in most parts of the
world. But the complexity of such cases is highly influenced by nor-
mative understandings of what rape is, who can be victimized, and
who can be a perpetrator. Historically, rape in marriage has seldom
been recognized, as husbands were accorded an unquestioned right to
their wives’ bodies, and women were expected to acquiesce. Likewise,
rape was often denied both by society as well as in the courts because
of how women behaved, what they wore, where they went, or what
they had done. Traditionally, the only prosecutable forms of rape have
been those involving a clearly innocent victim and a perpetrator who
is a stranger to that victim (data shows, however, that most sexual vio-
lence is perpetrated by people in relationships of trust with the victim).
So gendered prejudices impede adequate prosecution. In Peru, the his-
torical divisions along lines of race, class, and gender further constrain
a more objective legal perspective upon cases of rape. This is wide-
spread in “everyday” peacetime cases such as that of Lady Guillén and
Arlette Contreras, as well as in cases of conflict-related rape. For ex-
ample, while international law says that in a context of war there can-
not be a context of consent, in practice, the lines between consent and
coercion seem just as porous as during peacetime: CVR interviews
show how interviewers sometimes dismissed women’s claims of having
been raped based on assumptions about the nature of consent, even in
the overall violent context of war. Likewise, conceptions of what con-
stitutes an injury, or of who speaks the “truth,” have been questioned
by judges presiding over cases of state-perpetrated rape—even when
the evidence is a child born in captivity. Prosecuting cases of confl ict-
related rape are difficult for the same reasons that human rights viola-
tions perpetrated by the military are difficult to prosecute, but in ad-

Soifer_6844-final.indb 169 8/17/18 11:53 AM

 170 Jelke Boesten

dition, the specific coordinates of sex crimes and the institutionally

embedded prejudices based on race, class, and particularly gender fur-
ther impede successful prosecution in Peru.
The successful prosecution in 2016 of two former military officers
in Guatemala for crimes against humanity, sexual violence, and do-
mestic slavery perpetrated against indigenous women in the commu-
nity of Sepur Zarco in the 1980s is groundbreaking because of this
complexity. This success is now followed by the Manta y Vilca case
considering crimes against humanity and sexual violence in 1980s
highland Peru. Although Manta y Vilca might take years to resolve in
court, that it has reached court in the fi rst place is already an achieve-
ment. But however revolutionary the case might be, there is little pub-
lic interest in Manta y Vilca, for all the reasons described previously.
In addition, the victim-survivors, poor indigenous women of mature
age, are not interesting to the wider public, nor are the perpetrators of
much public interest as mestizo and cholo former soldiers living at the
margins of contemporary society.
At the same time, there has been increasing pressure on the Peru-
vian government in recent years to investigate and account for forced
sterilizations carried out in the mid-1990s by the Fujimori administra-
tion under the banner of the Programa Nacional de Planificación Fa-
miliar. The goal of this National Family Planning Program was to re-
duce population growth to 2 percent by decreasing average fertility
from 3.6 to 2.5 children per woman by 2000. One program objec-
tive was to improve maternal and child health, and the effort on the
whole emphasized freedom to choose, reproductive rights, and gen-
der equality. In practice, however, rural physicians were given quotas
to sterilize a certain number of women each month, and were prom-
ised improved working conditions and resources for their cooperation.
Doctors who were uncooperative were threatened with negative conse-
quences. According to various sources, many poor, largely indigenous
women were sterilized, including between 600 and 10,000 forcibly or
against their will. Many more—up to 300,000—were treated in unhy-
gienic and unprofessional circumstances (CLADEM 1999; Congreso
de la Republica 2002; MINSA 2002; DEMUS 2008).
The developmental rationale for the program—to reduce population
growth for the sake of economic stability—was based on the idea that
one can reduce poverty by reducing the poor population. Misconcep-
tions founded on sexism (only women are responsible for reproduction
and hence for poverty) and racism (indigenous women and men are

Soifer_6844-final.indb 170 8/17/18 11:53 AM

 Legacies of Gender-Based Violence in Peru 171

too ignorant to control their own fertility even when given the right re-
sources) among both those in command as well as those implementing
the policy made coercion in the sterilization program possible (Boesten
2010). Several investigations have taken place since the coercive prac-
tices behind the sterilization program were uncovered (Congreso de la
Republica 2002; MINSA 2002; DEMUS 2008) including one by the
US Congress in 1998 to account for the role of USAID funding for the
program.9 However, to this day, no in-depth research has documented
the actual number of women affected, nor has any criminal investiga-
tion accounted for the actors involved. Research shows that local physi-
cians and nurses had an important role to play in the program’s imple-
mentation (Boesten 2010; Gianella 2014). Although it seems necessary
to hold them accountable, it would be too easy to blame only some in-
dividual doctors. Rather, it is essential that those who designed and en-
forced the program be held to account for the harm they have done.
Governmental responsibility for the violence perpetrated in its name is
crucial if the objective is to break through the gendered and racialized
structures of inequality.
After the fi rst uncovering of the program by human rights activist
Giulia Tamayo in 1997,10 the forced sterilization program at fi rst drew
the attention of only a relatively small group of academics and activ-
ists. The case revealed a series of tensions and contradictions in Pe-
ruvian society and its desire for change that are difficult to reconcile:
progressive feminist ideas about equality and inclusivity clash with
conservative and patriarchal ideas about women’s roles and their re-
sponsibility for reproduction.
Since 2012, a new generation of activists has worked to document
and publicize forced sterilizations carried out by the Family Planning
Program, and to pressure the government into creating a register of
victims that can be used in administering reparations, conducting an
investigation into political accountability, and exploring criminal ac-
countability (Ballón 2014). Despite this heightened activism, the judi-
ciary permanently closed the case in November 2016.
There are several ways of understanding the state’s lack of enthusi-
asm for addressing these historic abuses. First, and most importantly,
both the sexual violence of the armed conflict as well as the forced
sterilizations are representative of persistent inequalities grounded in
ideas about race, class, and gender. Clearly, the racist connotations of
the sexual violence meted out on the rural and prison population in
the 1980s and 1990s, and of the forced sterilizations of the late 1990s,

Soifer_6844-final.indb 171 8/17/18 11:53 AM

 172 Jelke Boesten

have helped reproduce and further entrench those inequalities. Any

transitional justice policy would have to address this in order to chal-
lenge these inequalities. While the report of the CVR certainly focused
on these structural inequalities and the violence that it produced, it
was not able to unsettle the narrative by drawing in new voices or
significantly challenging existing hierarchies. Challenging structures
of inequality grounded in entrenched prejudices perhaps requires new
thinking and new actors on the political stage. The entrenchment of
inequality means that indigenous women (and, arguably, most indige-
nous men) simply do not have a voice in Peruvian society. Most claims
that indigenous women want to make on the state, or on broader so-
ciety, have to be channeled via urban activist networks and NGOs,
which draws them into a whole different set of political struggles.
Another way of explaining the lack of enthusiasm for addressing the
historic abuses is by placing them in the broader context of contempo-
rary post-IAC battles over truth and justice. Transitional Peru is divided
between those who view the confl ict through a military- conservative
perspective, and those who are grounded in a human rights perspective
as laid out by the CVR’s fi nal report (Drinot 2009). In regard to gen-
der issues, the conservative-military perspective employs a patriarchal
orientation akin to contemporary opposition to “gender ideology” as
discussed previously. Human rights groups, in contrast, support gender
equality and women’s rights. Human rights organizations also actively
seek to provide a voice to those marginalized by the persistent inequal-
ities in Peruvian society. For the victim-survivors of sexual violence at
the military bases of Manta and Vilca, Huancavelica, as well as for
those sterilized under the Fujimori regime, this means that their cases
are channeled into the public sphere via vocal human rights groups
such as APRODEH, IDL, COMISEDH, and DEMUS. This places vic-
tim-survivors and their legal cases in the middle of the battles over
truth and memory, between those who support a military-conservative
narrative and those who favor a human rights perspective. But it is vic-
tim-survivors’ human rights that are at stake here. Their past, present,
and future are largely lost in the struggles among urban elites over IAC
legacies that Paulo Drinot describes in chapter 11.

Conclusion

In 2016, two emblematic historical cases of institutional sexual vio-

lence against women, one involving the systematic rape of women in

Soifer_6844-final.indb 172 8/17/18 11:53 AM

 Legacies of Gender-Based Violence in Peru 173

rural Huancavelica, and the other the forced sterilization of poor rural
women, could not get civil society to mobilize as effectively as could
the two “everyday” cases of Lady Guillén and Cindy Contreras. While
there are reasons to believe that high levels of contemporary violence
against women are related to the long and complex history of insti-
tutionalized racism, sexism, and political violence, it is apparent that
it will be difficult to use the post-confl ict moment to unsettle the en-
trenched and intersecting inequalities that are the root of this violence.
There is no clear causal relationship between conflict-related gender-
based violence and peacetime levels of violence that can be quantified
and made concrete.
Considering these constraints, how was it possible for three women
in July 2016 to mobilize fifty independent organizers in twenty-four
hours, forty-five thousand collaborators and protesters in five days,
and about half a million women, men, and children in time to march
against gender-based violence on August 13? First, I believe that the
three Ni Una Menos instigators managed to mobilize a cross-class al-
liance with women who did not have a history of activism or politics.
This is unique, and was largely accomplished through social media,
which does not discriminate (as much) as does word-of-mouth mobi-
lization. The effective use of different Facebook pages as well as in-
stant messaging had a democratizing effect upon the often-unequal
relationship between civil society organizations or NGOs and grass-
roots groups. The Facebook page set up to serve as an organizational
platform quickly turned into a platform for sharing painful mem-
ories, many of which had never been revealed before. The organiz-
ers respected this rain of testimonies and set up alternative social
media tools to continue organizing the protest march. An open con-
fessional space used by fifty thousand people is a conscience-raising
forum beyond most feminists’ dreams, and it allowed for people to
speak and participate who otherwise would have stayed invisible. The
number and severity of the experiences shared on this page, as well as
the speed at which they circulated, also drew in allies who are usu-
ally more difficult to mobilize for cases of gender-based violence: men,
private-sector participants such as business sponsors, and indeed even
representatives of the state (the police, the judiciary, and the newly ap-
pointed president all made public statements and appearances as part
of the protests).
The political moment was important: Keiko Fujimori, a symbol of a
guilty and violent past, and arguably representing (or at least defend-
ing) the conservative-military historical perspective, lost the elections

Soifer_6844-final.indb 173 8/17/18 11:53 AM

 174 Jelke Boesten

held in April 2016. Verónika Mendoza, a representative of a hopeful

new left-wing alliance arguably championing a human rights histori-
cal perspective, also lost. The winner, Pedro Pablo Kuczynski, if not a
political outsider or a representative of a new generation, was chosen
for his technical perspective, or, one could argue, for his seeming neu-
trality in long-standing political feuds. The Ni Una Menos campaign
stepped into a political vacuum in which progressive politics, as well
as opposition to them, was in disarray. The fi rst weeks of this new gov-
ernment was the perfect moment to mobilize the Peruvian population
against something that often falls in the gaps between two highly di-
vided political camps.
At the same, the Ni Una Menos campaign focused on ongoing every-
day violence that women and girls experience regardless of their polit-
ical, ethnic, geographical, or class origins. While this campaign was
certainly supportive of emblematic historical cases of sexual violence,
these cases were not central to the Ni Una Menos narrative. Much of
the transitional justice literature advocates for the post-confl ict mo-
ment to be crucial in effecting transformational change. We hope that
the shock of truth will lead to justice, and that criminal justice and rep-
aration will lead to conciliation and transformation. But perhaps it is
the other way around: fi rst things need to change, and then perhaps
there will be justice.

Notes

1. We also need to consider the possibility that the CVR asked women
questions about the violence they experienced that were not often asked be-
fore more systematic monitoring of rates of violence against women began.
Holly Porter found that a majority of raped women during the Ugandan in-
ternal confl ict claimed to have been raped by intimate partners, not by active
combatants (2015).
2. Obdulia Guevara Neyra, interview with Gabriela Wiener in La Repu-
blica, http://larepublica.pe/impresa/domingo/835623-en-lima-mas-del-30-de
-trabajadoras-del-hogar-todavia-esta-en-una-situacion-de-semiesclavitud.
3. See http://larepublica.pe/impresa/sociedad/785994-siento-miedo-ahora
-yo-soy-la-prisionera-dice-cindy-arlette-contreras.
4. See http://www.peruthisweek.com/news-lady-guillen-110009.
5. “[They tell us] there are many abortions among young girls, but nobody
has abused these girls. Often it is women who put themselves on display, pro-
voking men.” Cardinal Juan Luis Cipriani, archbishop of Lima, Peru, on na-
tional radio in response to campaigns against sexual violence and in favor of
the legalization of abortion in case of rape, July 30, 2016, RPP Radio.

Soifer_6844-final.indb 174 8/17/18 11:53 AM

 Legacies of Gender-Based Violence in Peru 175

6. See http://larepublica.pe/politica/829690-el-85-aprueba-que-curriculo
-escolar-promueva-igualdad-de-genero-segun-ipsos.
7. See Jacqueline Fowks, “Perú condena por primera vez a militares por
violaciones sexuales cometidas en los años de confl icto interno,” El País, Feb-
ruary 9, 2018, https://elpais.com/internacional/2018/02/09/america/1518201
594_889441.html.
8. One more case is under investigation by the Interamerican Commis-
sion for Human Rights: http://idehpucp.pucp.edu.pe/comunicaciones/notas
-informativas/cidh-admite-caso-emblematico-de-violacion-sexual-ocurrido
-durante-periodo-de-violencia/.
9. The Peruvian Population Control Program Hearing Before the Subcom-
mittee on International Operations and Human Rights, 105th Cong., Febru-
ary 25, 1998, http://commdocs.house.gov/committees/intlrel/hfa48459.000
/hfa48459_0f.htm.
10. Tamayo had to flee the country as a consequence of her work.

Soifer_6844-final.indb 175 8/17/18 11:53 AM

 CHAP T ER 7

Indigenous Activism and Human

Rights NGOs in Peru: The Unexpected
Consequences of Armed Conflict
Maritza Paredes

In the following pages, I address the armed confl ict’s unintended ef-
fects on indigenous activism. In particular, I focus on the increas-
ing cooperation between indigenous organizations and human rights
NGOs arising from the context of internal conflict. The existing liter-
ature has thoroughly explored the damage inflicted on indigenous mo-
bilization by the war initiated by Sendero in the central Andes at the
end of the last century (Yashar 2005; Van Cott 2005; Burt 2009b).
War records showing the impact of the internal conflict on indigenous
fatalities reveal that 75 percent of those killed in the conflict were of
indigenous origin, mainly from the departments of Ayacucho, Huan-
cavelica, Huánuco, and San Martín (CVR 2004).1
Moreover, a well-known direct negative consequence of the internal
conflict on indigenous mobilization at the turn of the twenty-fi rst cen-
tury has been the lack of convergence between integrating indigenous
organizations in the highlands and emerging and dynamic indigenous
organizations in the Amazon (Yashar 2005; Van Cott 2005; R. Smith
1996). Unlike in Bolivia and Ecuador, indigenous peasant organiza-
tions in Peru’s highlands did not transform their “class-based” frames
into “ethnic-based” ones. 2 The declining reputation of class ideologies,
which were embraced by indigenous peasant organizations in the high-
lands, left these groups without political resources for legitimate mobi-
lization (Paredes 2010). Grassroots indigenous organizations from the
Amazon moved away from their peers in the highlands and converged
with one another in a process of indigenous internationalization, in-
tersecting with the demands of the global environmental movement
(R. Smith 1996). As noted by a vast literature, at the turn of the cen-
tury, Peru did not see the emergence of a pan-ethnic indigenous move-

Soifer_6844-final.indb 176 8/17/18 11:53 AM

 Indigenous Activism and Human Rights NGOs 177

ment in reaction to neoliberal policies that undermined indigenous col-

lective rights.
Nonetheless, this study departs from these previous analyses, as
it focuses on the unexpected effects of armed conflict on indigenous
activism in contemporary Peru. Indigenous activism in this new cen-
tury has taken place in a context characterized by the end of the in-
ternal conflict and democracy, economic growth based on booms in
the markets for metals and hydrocarbons mainly extracted from indig-
enous territories, and the internationalization of legalized structures
for the protection of indigenous rights. By examining the development
of these conditions, this chapter concludes that the effects of internal
conflict on contemporary indigenous mobilization are complex, with
multidirectional configurations. In particular, this chapter focuses on
a second and less-studied indirect consequence of the internal confl ict
on indigenous mobilization: the increasing impact of human rights
NGOs on indigenous activism.
In the following, I focus on the building of new alliances with hu-
man rights NGOs and how these alliances reshape both the opportu-
nities and the challenges for indigenous organizations to mobilize in
contemporary Peru. These NGOs became an important actor, with ca-
pacities to provide legal advice to indigenous organizations and com-
munities in a context of internationalization and expanding legal-
ization of indigenous rights (Morgan 2004). In Latin America, legal
activism has become increasingly prominent over the last two decades
(Rodríguez-Garavito and Rodríguez-Franco 2015), particularly on in-
digenous and environmental issues (Sieder and Barrera 2017). Indige-
nous mobilization in the region has been portrayed as a form of “legal
subaltern cosmopolitanism” (Rodríguez-Garavito and Sousa Santos
2007, 19–20) connected to transnational networks of activists, which
act and exercise pressure on highly legalized structures (Keck and Sik-
kink 1998; Brysk 2000; Sieder 2011). The most important of these
structures has been International Labor Organization convention 196
acknowledged by most Latin American countries and ratified by the
Peruvian state in 1994.
Indigenous mobilization in Peru and Latin America is occurring
today in this new context that includes both threats from extractive
industries acting in indigenous territories and opportunities from le-
galized structures providing new forms of activism. Human rights
organizations in Peru emerged from the internal confl ict with signif-
icant legal training on constitutional and human rights, local and in-

Soifer_6844-final.indb 177 8/17/18 11:53 AM

 178 Maritza Paredes

ternational advocacy expertise, and grassroots connections in rural

areas where indigenous communities live. On the basis of these ca-
pacities, human rights NGOs have been crucial to the revitalization
of indigenous activism in Peru, in spite of the challenges of grassroots
organizations.
The role of human rights NGOs has become crystallized in the signifi-
cant events of the so-called Bagua protests and the subsequent approval
and implementation of the Law on Prior Consultation (no. 29785, 2011)
and its regulating decree (Supreme Decree 001-2012-MC, 2012). This
law is the fi rst of its kind in the region, and its implementation is still
highly controversial in the country (Schilling–Vacaflor and Flemmer
2013; Flemmer and Schilling–Vacaflor 2016; Sanborn, Hurtado, and
Ramírez 2016). Human rights NGOs supported the legalization and
formalization of consultation rights. In turn, the promulgation of the
Law on Prior Consultation (no. 29785) has expanded the possibilities
of domestic litigation around these rights and their implementation.
Relying on the analysis of events and interviews around the case of
the so-called Bagua protest, I will trace the history of alliances between
indigenous organizations and human rights NGOs as an indirect con-
sequence of the internal war. The chapter is organized in six sections.
The next section presents the old and new setting for indigenous mo-
bilization in Peru. The third section explains how armed conflict in
Peru stimulated the formation of significant human rights organiza-
tions, closely connected to broad international networks that sought to
defend citizens from the injustices of internal war. The fourth section
shows how as political violence diminished, and subsequently post-
conflict efforts (such as the Truth and Reconciliation Commission, or
Comisión de la Verdad y Reconciliación [CVR]) faded, these human
rights organizations with crucial legal capacities became important al-
lies of indigenous organizations and activists on indigenous rights. The
fifth section suggests that the role of these organizations, however, can-
not compensate for the still significant Andean and Amazonian divi-
sions in the movement. In the sixth section, I offer my conclusions.

Old and New Settings for Indigenous Activism in Peru

Pioneering studies argued that restricted political association during

the armed conflict (Yashar 2005) and violence against leaders and
their communities (Burt 2009b) weakened indigenous networks in

Soifer_6844-final.indb 178 8/17/18 11:53 AM

 Indigenous Activism and Human Rights NGOs 179

Peru (Yashar 2005, Van Cott 2005). At the end of the twentieth cen-
tury, Bolivian and Ecuadorian indigenous peasants and lowland orga-
nizations produced broader and more powerful pan-ethnic coalitions
that eventually became indigenous political parties, such as the Movi-
miento al Socialismo (MAS), which brought Evo Morales to power in
Bolivia.3 In Peru, these pan-ethnic coalitions were not formed, and the
weakness of its indigenous movement is often seen as exceptional in
regional perspective.4
Other authors have also noted the distinctive influence of the Peru-
vian radical Marxist Left on the class identity of indigenous peasant
organizations at the communal level (Van Cott 2005; Degregori 1989;
Paredes 2010; Paredes 2011). For Paredes (2011), who reported an in-
crease in indigenous association activity by peasants in the highlands
at the end of the 1980s, 5 the problem was not the lack of opportunity
for association as reported by Yashar (2005), but that the radical dis-
course of “class confl ict” propagated by the Shining Path made mo-
bilization on class terms virtually impossible in Peru throughout the
1980s and 1990s. As explored in detail in chapter 8, within a context
of brutal internal war and economic and political crises, it was easy in
the 1990s for the government of Fujimori to blame the chaos on polit-
ical parties, and particularly on the Marxist Left and their message of
“class conflict.”
Peasant organizations in Peru that traditionally had been built along
class lines openly associated with the Left, such as the fi rst national in-
digenous peasant union (the Confederación Campesina del Perú, or
CCP), were hurt by the campaign against the Left. This harm was due
to not only the increasing persecution of their leaders and direct attacks
on their organizations and their lives by Sendero and Fujimori’s govern-
ment, but also the confusion created among the public concerning the
difference between these peasant organizations and Sendero. The per-
secution and stigmatization of class-based ideologies in Peru seriously
dampened the potential for an indigenous movement to emerge from
previously active indigenous peasant organizations (Paredes 2010). At
the end of the 1980s, the CCP incorporated about 250,000 members
of 500 organizations—federations, unions, community groups, and
others—from seventeen of the twenty-five departments in the coun-
try, mostly from the highlands (Matos Mar and Mejía 1980). Yet in the
1990s this incorporation failed to transform itself into a vigorous in-
digenous organization, as did its peers in Bolivia and Ecuador.
In contrast to the experience of highland indigenous peasant organi-

Soifer_6844-final.indb 179 8/17/18 11:53 AM

 180 Maritza Paredes

zations, Amazonian indigenous organizations converged into a move-

ment that was able to mobilize and grow, even during the 1990s and
the ongoing internal confl ict. They organized during this decade on
the basis of a multicultural rights framework and drew support from a
growing number of international allies from the environmental move-
ment (Greene 2009; R. Smith 1996). The divergence between highlands
and lowlands organizations in Peru, with almost no bridges connect-
ing them, is the most direct and persistent consequence of the internal
conflict on indigenous mobilization in the late twentieth century. As
we will see, this direct consequence of the internal conflict would sub-
sequently interact with other indirect and unintended consequences to
shape the complex dynamics of indigenous activism in Peru today.
In the new century, indigenous activism has taken place in a changed
domestic and international setting. The political, economic, and insti-
tutional scenario in which Yashar, Van Cott, and others studied indig-
enous activism and politics has changed. The new century is charac-
terized by new political regimes and a super cycle of commodity boom
in the region. Since 2004, Peru and the Andean countries have expe-
rienced a natural resources boom with increasing investments in the
extractive sector, focused on mining and hydrocarbons (Thorp et al.
2012). Based on this region-wide boom across the Andean countries,
Peru has experienced a cycle of sustained growth, with average annual
growth rates of 6 percent for the period 2002–2013 (Dargent et al.
2017).
The other side of economic growth has been social confl ict (De
Echave et al. 2009; Paredes and de la Puente 2017; Defensoría del Pue-
blo del Perú 2013). Communities have accused extractive industries of
land and resource grabbing (Haarstad and Fløysand 2007; Bebbington
and Bury 2013), contamination (Muñoz et al. 2007; Orta-Martínez
and Finer 2010; Orihuela and Paredes 2017), and unfair or ineffec-
tive redistribution of economic and social benefits (Arellano-Yanguas
2011; Arce 2014; Ponce and McClintock 2014).
Indigenous people have been crucial actors in these confl icts as min-
ing and hydrocarbons concessions and indigenous land rights have in-
creasingly overlapped in the last decade. In the Amazon, at the end of
2008, 72 percent of the territory belonging to indigenous communities
was under concession to petroleum and gas exploration and exploi-
tation (Benavides 2010; Orta-Martinez and Finer 2010). In the An-
des, by 2017, 40 percent of indigenous community lands overlap with
mining concessions (Martínez 2018). In many cases, there is uncer-

Soifer_6844-final.indb 180 8/17/18 11:53 AM

 Indigenous Activism and Human Rights NGOs 181

tainty over property rights due to poor land titling, and indigenous
communities are ready to defend the lands they have been rightfully
occupying.6
The commodity boom cycle in indigenous territories is thus provok-
ing a rapidly emerging set of challenges in the region that has revealed
the weakness of institutionalized relations between indigenous peoples
and the state. However, the internationalization of indigenous move-
ments has set forth an agenda centered on environmental justice, ter-
ritorial autonomy, and the political implementation of collective and
international rights (Brysk 2000; Keck and Sikkink 1998) and has in-
creased the salience of the international and domestic setting for indig-
enous legal action (Anthias 2014; Sawyer 2013a; Sawyer 2012; Saw-
yer 2004; Perreault 2013; Bebbington 2011; Bebbington 2013; Sawyer
and Gomez 2008; Sawyer and Gomez 2012; Rodríguez-Garavito
2011). Indigenous people now have an internationally recognized le-
gal right to consultation when a development is going to affect their
territory and livelihoods. They can also appeal to the Inter-American
Commission on Human Rights (ICHR) when their states fail to com-
ply with established conventions.7 The effectiveness of international
pressure varies from country to country. However, as we argue here,
in Peru, the opportunity to work together with human rights NGOs in
this new context of legal complaints is adding new characteristics and
dynamism to a previously weak setting for indigenous activism.

The “Dog in the Manger” of Garcia’s Government

and the “Bagua Protests”

Indigenous mobilization in Peru experienced a tipping point under the

government of Alan García from 2005 to 2010. In 2007, President
Garcia received special legislative powers from the Congress to imple-
ment the Free Trade Agreement signed between the United States and
Peru. With this power, he approved a package of decrees known as the
“Law of the Jungle” in 2008. The new legislation sought to tailor the
national legal system to global markets by facilitating increased for-
eign direct investment in Peru and greater integration of the nation’s
economy into the global economy. The García administration sought
to adapt the national legal system to the requirements of international
capital by facilitating procedures for foreign investment in indigenous
territories (Hughes 2010; Schmall 2011).8

Soifer_6844-final.indb 181 8/17/18 11:53 AM

 182 Maritza Paredes

The aggressive defense of these legal measures by President Gar-

cía showed that his goal was not only to achieve economic growth but
also to drive cultural transformation through the “de-indigenization”
of the country. In a public campaign led by García himself, the gov-
ernment portrayed communal rights (especially in the Amazon) as an
“impediment” to economic development in Peru (García Perez 2007b).
In a series of articles called El Perro del Hortelano (the dog in the man-
ger),9 García described indigenous people as “incapable” of economic
production on their “abandoned” lands. In García’s “dog in the man-
ger” philosophy, indigenous communal ownership was an anachro-
nism preventing development and causing poverty in the Amazon (Me-
rino 2015; Drinot 2011; Hughes 2010). As Drinot argues, by invoking
“communists, protectionists and environmentalists” as enemies of the
country, García was referring to “a recalcitrant anti-capitalist Other,
which is Peru’s indigenous population.” He claimed the indigenous
population represented “backwardness” (Drinot 2011, 183).
Indigenous peoples opposed the government’s decrees. Drawing
particularly vehement opposition were Decrees 1090 and 1064, which
transferred 45 million hectares of protected Peruvian Amazon forest
to private companies (Merino 2015). Additionally, Decree 1015 jeop-
ardized the collective territorial rights of indigenous communities by
weakening protections guaranteed under Peruvian law. For example,
it changed the condition that in order to reallocate a communities’
land to outsiders, two-thirds of the general assembly of the commu-
nity must approve reducing the threshold for acceptance to a simple
majority (Merino 2015). The Inter-Ethnic Development Association of
the Peruvian Amazon (Asociación Interétnica de Desarrollo de la Selva
Peruana, or AIDESEP), the major organization of Amazonian indige-
nous peoples, called for demonstrations in 2008 that lasted for fifteen
months during 2008 and 2009. This conflict between indigenous com-
munities and the central government has been the longest that has Peru
has faced in the new twenty-fi rst century, and the most serious in terms
of its consequences.
When the conflict emerged, indigenous organizations from both the
highlands and the Amazon had begun sustained coordination for the
fi rst time in Peruvian history. Organizations from the Amazon and
the Andes both saw danger in legislation that reduced indigenous ter-
ritorial rights in both regions. The situation created an opportunity
for building pan-ethnic alliances. During this crisis, indigenous orga-
nizations strengthened their alliances with an agreement10 known as
the Pacto Unidad, which includes eight organizations from the two re-

Soifer_6844-final.indb 182 8/17/18 11:53 AM

 Indigenous Activism and Human Rights NGOs 183

gions.11 This alliance became part of a dynamic network of activism,

in which human rights NGOs and the Peruvian government’s Om-
budsman’s Office also participated, campaigning in the Congress and
the media for the institutionalization of prior consultation for indige-
nous communities.
In the new century, organizations like the CCP embraced interna-
tional indigenous rights and framed their demands around commu-
nal rights and the identity of “original peoples.” After years of em-
phasizing class discourse and alliance with leftist parties, new leaders
managed to rearticulate alliances and demands along the lines of in-
digenous international legal rights for indigenous peoples.12 Another
organization, the Confederación de Comunidades del Perú Afectadas
por la Minería (CONACAMI), also emerged from communities af-
fected by mining, mainly in the highlands (Paredes 2006). Still, the
strongest organization was the AIDESEP. Established in the 1980s, its
strength was based on the affiliations of 1,350 communities, approxi-
mately 80 federations and regional organizations, and 65 different in-
digenous peoples of the Amazon, as well as its multiple national and
transnational alliances, mainly with the international environmental
movement (R. Smith 1996). The coordination among all of these orga-
nizations represented a new step forward in the building of pan-ethnic
alliances that had not previously been possible.
La Coordinadora Nacional de Derechos Humanos (CNDH) hosted
many of these meetings of coordination among indigenous organiza-
tions from the highlands and lowlands, which were aimed at creating
an integrated national indigenous platform.13 In these meetings and
through open letters, indigenous leaders invoked the implementation
of indigenous rights provided for in the 169 ILO Convention for In-
digenous and Tribal People.14 After the government ratified this con-
vention in 1994, the state of Peru was criticized by both national and
international human rights organizations for the absence of adequate
legislation on prior consultation with indigenous peoples, and the fail-
ure to implement such legislation. The problem became more serious
when the high prices of minerals and hydrocarbons triggered a mas-
sive expansion by extractive industries on indigenous lands and re-
sources, and when the García government threatened legislation pro-
tecting communities’ collective land rights in 2008.
The indigenous organizations decided on a joint response to the de-
crees issued by the García government. However, most of the politi-
cal action occurred in the Amazon, a reflection of the strength of the
movement in this region. On August 9, 2008, activists blocked the op-

Soifer_6844-final.indb 183 8/17/18 11:53 AM

 184 Maritza Paredes

erations of a number of oil facilities in various areas of the northern

Amazon and forced the government to initiate negotiations. Though
these negotiations between indigenous organizations and the govern-
ment continued for months, they ultimately failed, leading to a stron-
ger round of protests in April 2009. Fifteen months of protest and mo-
bilization culminated in tragedy: on June 5, 2009, thirty-three people
died in Bagua, an area in the Amazonas region also called the “Devil’s
Curve,” when police officers tried to evict protesters who had occupied
a highway (Anaya 2009; Cavero 2011; A. Durand 2010; Sosa 2017).
Days before these events, President García had condemned the ac-
tions of indigenous peoples, saying, “These people do not have a crown,
they are not fi rst-class citizens, 400,000 natives cannot tell 28 million
Peruvians: ‘You have no right to come here.’ No way, that’s a very seri-
ous mistake, and whoever thinks that way wants to take us to irratio-
nality and the primitive mindset of the past.” For Santiago Manuin, an
Awajun leader who suffered eight bullet wounds during the protest,15
the mobilization was not only peaceful but rightful: “The government
of that time confronted us, police followed the order of their high com-
mand and the native people defended its right, its identity, its culture,
its own development, its forest, its rivers, its cosmos and its territory.
The strike was peaceful; until that fateful day of June 5, our protest
was to reject the decrees that affected us, and the government did not
consult with us as it is required by the ILO Convention 169.”16
The tragic events in La Curva del Diablo had multiple consequences
for indigenous organizations. On one side, it had a negative impact on
the AIDESEP. The group’s main leader, Alberto Pizango, had to leave
the country after being charged with sedition, conspiracy, and rebel-
lion. Another fi fty-three leaders were also accused of sedition and of
shooting at police, which rapidly undermined the most important re-
source of the Amazonian indigenous movement, its leadership (Greene
2009). The next section explains how, during a significant period
of mobilization and potential demobilization after the events of the
“Curva del Diablo,” alliances with human rights NGOs created new
opportunities for indigenous activism in Peru.

Human Rights NGOs and the Indigenous Movement in Peru

Human Rights, NGOs, and the Indigenous Issue

During the armed conflict, only a small number of human rights or-
ganizations focused on indigenous issues. Though many of these or-

Soifer_6844-final.indb 184 8/17/18 11:53 AM

 Indigenous Activism and Human Rights NGOs 185

ganizations worked mainly with indigenous communities in the An-

des that saw especially dramatic abuses of human rights, there were
also a small number of NGOs centered in the Amazon in the years be-
fore the beginning of political violence in Peru (CAAAP 1992; Chirif
2012; Espinosa de Rivero 2009). Among the best examples of organi-
zations with this profi le are the Episcopal Conference (Comisión Epis-
copal de Acción Social, or CEAS), the NGO Evangelical Peace and
Hope, and the Amazon Center of Anthropology and Practical Appli-
cation (Centro Amazónico de Antropología y Aplicación Práctica, or
CAAAP), founded in 1974.17 Though in the Amazon there were many
other national NGOs and international affiliates, these were the main
ones dedicated to Amazonian politics and environmental issues.18
During the 1980s and 1990s, political violence became the central
axis of these human rights organizations’ work, as groups came to ful-
fill two primary functions during this time period. First, they played
a key role in recording and circulating evidence of crimes against hu-
manity for the CNDH, especially in the 1990s when violence in the
Amazon was at its most severe (author’s interview with Coronado
2015; interview with Torres 2015; interview with Jugo 2015; all inter-
views are listed at the end of the chapter). For example, the CAAAP
and its network of allied organizations (other small NGOs and Cath-
olic Church parishes and prelatures) were the main actors responsi-
ble for collecting information on human rights abuses in the Amazon
for the CVR. Second, organizations such as the CAAAP and its allies
served as the base for other human rights organizations—the Instituto
de Defensa Legal (IDL) and Ombudsman’s Office, among others—to
initiate their human rights work in the Amazon region (interview with
Coronado 2015; interview with Torres 2015).
Toward the end of the armed conflict, community self-defense com-
mittees became a challenge for governance because the state did not
have control over these local organizations, many of which were sit-
uated in indigenous communities. This problem created an opportu-
nity for human rights organizations, which had no previous experi-
ence with indigenous populations, to offer legal advice on community
justice, which has since gained international legitimacy as a right of in-
digenous people.
Self-defense committees spread across the Peruvian territory dur-
ing the armed conflict, but they were fi rst created in the northern high-
lands to solve problems of cattle-rustling (Starn 1999; Paredes 2011;
Degregori et al. 1996). Because of its success, the model of community
self-defense was replicated in other regions, especially in the north-

Soifer_6844-final.indb 185 8/17/18 11:53 AM

 186 Maritza Paredes

eastern area of the country (CVR 2003a). Although military authori-

ties adapted the model to other communities in the central highlands
to combat the Shining Path, outside of this area self-defense commit-
tees were autonomous and worked with other social and local orga-
nizations (Starn 1998; Starn 1999; Degregori et al. 1996).19 By 1993,
there were 4,205 self-defense committees across the country, with
235,465 members that had 16,196 weapons (del Pino 1996). 20 Once the
Shining Path’s retreat was accomplished, these organizations became
a challenge for the state, as it intended to resume authority in areas
where historically its presence had been weak. Human rights NGOs
like the IDL, which had no previous experience with these populations,
as well as agencies such as the Ombudsman’s Office began to work in
the midst of this confl ict between the state and the communities on in-
digenous justice issues. Community justice became a recognized com-
ponent of the multicultural framework. (Sieder 1998; Yrigoyen 2004).
Toward the end of the 1990s, and especially during the democratic
transition, the human rights network’s agenda thus became more di-
verse (interview with Coronado 2015). During this process, indige-
nous issues gained greater institutional presence in NGOs within the
human rights network. In particular, the CNDH, the IDL, and the
Asociación Pro Derechos Humanos (APRODEH) became involved
in indigenous issues (interview with Coronado 2015). Gradually, the
Ombudsman’s Office took advantage of decentralization to work with
peasant and native Amazonian groups (interview with Abad 2014; in-
terview with Luque 2014; interview with Abanto 2013). The CNDH
founded the Working Group of Indigenous Peoples, composed initially
of the NGOs that traditionally had worked with indigenous groups
such as the CAAAP, the CEAS, and Paz y Esperanza. The Working
Group brought together indigenous organizations to develop con-
crete proposals that were responsive to indigenous people’s necessities.
Moreover, the Ombudsman’s Office, which was a pioneer institution
in this realm when it founded the Program for Indigenous Persons to-
ward the end of the 1990s, was convened to participate. 21
As of the early 2000s, indigenous activism in human rights had been
limited in terms of growth (interview with Coronado 2015). Within
the context of Peru’s democratic transition and especially under the
Paniagua government, a window of opportunity appeared to open for
addressing indigenous issues in the country. A series of formalized di-
alogues known as the Mesa de Diálogo were convened, which resulted
in the Plan of Priority Action for Peasant and Native Communities.

Soifer_6844-final.indb 186 8/17/18 11:53 AM

 Indigenous Activism and Human Rights NGOs 187

This plan was the fi rst of its type, and it was developed with the par-
ticipation of both Andean and Amazonian organizations. This process
was widely recognized by indigenous organizations and a variety of
NGOs. However, the plan never materialized into concrete policy ini-
tiatives in any sector, and during the Toledo administration it was in-
stead transformed into the National Commission of Andean and Am-
azon Peoples (Compañía Nacional de Peritos Agrícolas, or CONAPA),
under the direction of the fi rst lady, Eliane Karp. During Toledo’s ad-
ministration, CONAPA did not have the autonomy to promote the
plan, and indigenous demands were deemed as mere declarations in-
stead of something that would actually be met (Pajuelo 2007).
The Legislative Decree package of 2008 that reduced indigenous ter-
ritorial rights, and the protests that began that same year, generated a
sense of emergency which brought together diverse indigenous orga-
nizations to initiate a response that was also coordinated with NGOs
in the human rights network (interview with Coronado 2013; inter-
view with Coronado 2015; interview with Torres 2015; interview with
Lanegra 2014). At this time, the CNDH Indigenous Working Group
added more members, incorporating a range of organizations not tra-
ditionally associated with indigenous issues as well as their sources
of funding. 22 These organizations included those traditionally con-
centrated on campesinos (farmers), like Servicios Educativos Rurales
(SER), and environmental NGOs such as Sociedad Peruana de Dere-
cho Ambiental (SPDA) and Derecho, Ambiente y Recursos Naturales
(DAR) with extensive experience in indigenous and environmental is-
sues in the Amazonian region which had yet to develop a human rights
agenda.23 The Ombudsman’s Office had a more autonomous role, but
began to play a part in strengthening interactions between indigenous
organizations and members of the human rights network (interview
with Coronado 2015).
The shift of these organizations toward indigenous issues resulted
in the development of indigenous repertoires of mobilization with a le-
gal approach, which was not previously a prevalent feature of indige-
nous mobilization. Given that the legislative decrees were the origin
of protests, indigenous organizations began to fi le lawsuits to chan-
nel their demands. The fi rst example of this strategy appeared in reac-
tion to the approval of DL 1015, when the Ombudsman’s Office filed
a claim before the Consitutional Court (Tribunal Constitutional, or
TC) arguing that the government had failed to meet the prior con-
sultation requirement, and thus that the DL was unconstitutional. 24

Soifer_6844-final.indb 187 8/17/18 11:53 AM

 188 Maritza Paredes

With the support of the human rights network, indigenous organiza-

tions began to prepare legislative proposals that would transform ex-
isting processes into national policy innovations regarding indigenous
peoples. After the events in Bagua, the Congress opened a space for
negotiation that resulted in the legal project that later became the Law
of Previous Consultation of 2011 (Alva 2010; Brito 2012; Merino and
Lanegra 2013).
The abovementioned unconstitutionality process, led by the Om-
budsman’s Office, served as a basis for discussions in Congress (Me-
rino and Lanegra 2013; Villenas, Pautrat, and Samaniego 2010; Gam-
boa and Snoeck 2012). In the Bagua events, the Ombudsman’s Office
found an opportunity to apply an advocacy plan they had been pre-
paring for a long time. In 2006 they initiated a training process with
the ILO in preparation for a process of prior consultation in Peru (in-
terview with Abanto 2013; interview with Lanegra 2013). When the
controversial DLs were approved, the Ombudsman’s Office was well
prepared to challenge them on constitutional grounds, arguing that
pursuant to the Peruvian legal framework, if prior consultation was not
enforced, then the DLs were unconstitutional. This claim was repro-
duced by indigenous organizations and supported by institutions such
as the IDL and the CNDH (Gamboa and Snoeck 2012, Ruiz 2009).
Once the concept of unconstitutionality was proposed regarding
the conflict of 2008–2009, the debate shifted to constitutional rights.
The political structure opened opportunities for indigenous peoples
to achieve their objectives, including the possibility of making claims
based on the state’s compliance with its international obligations con-
cerning indigenous rights. In this regard, the CNDH wrote:

We call upon the Congress of the Republic, to revise, debate, and pro-
nounce in a positive and timely way regarding the request to repeal the
following laws . . . , that affect rights of Indigenous Peoples, as rec-
ognized by Convention 169: Concerning Indigenous and Tribal Peo-
ples in Independent Countries, of the International Labor Organiza-
tion, ratified by Peru, through the Legislative Resolution No. 26253,
published in the Official Newspaper “El Peruano” on December 2,
1993. 25

In this debate, organizations such as the CAAAP and the CEAS gave
way to the IDL and the Ombudsman’s Office, which had the tools to fi le
claims and initiate legal actions. 26 Given that organizations from civil

Soifer_6844-final.indb 188 8/17/18 11:53 AM

 Indigenous Activism and Human Rights NGOs 189

society participated as observers in congressional table discussions, co-

ordination with indigenous organizations did not stop (interview with
Coronado 2013; interview with Coronado 2015; interview with Bur-
neo 2013; interview with Torres 2013). Instead, the CNDH’s Work-
ing Group was one of the many spaces where indigenous organizations
were nurtured to participate in congressional legislative debates. Thus,
thanks to the mobilization of indigenous organizations, a direct negoti-
ation space with the executive and legislative branches was established.
This was also achieved because of the support received from human
rights activists, NGOs, and the Ombudsman’s Office, which raised vis-
ibility and facilitated the achievement of negotiation.
Indigenous organizations found in human rights NGOs the sup-
port and resources to mobilize for the implementation of prior con-
sultation, and after the events of Bagua and the approval of the Prior
Consultation Law, they increasingly counted on those NGOs for le-
gal support. This alliance has brought dynamism to the movement, in
a context where indigenous mobilization largely occurs within trans-
national networks of activism and is shaped by the international le-
galization of indigenous rights (Sieder 2011; Rodríguez-Garavito and
Sousa Santos 2007).27 The next section explains how these human
rights NGOs that have come to be a resource for contemporary indig-
enous mobilization emerged as social actors in the context of the inter-
nal war during the 1980s and 1990s.

The Development of Professional Human Rights

Organizations during the Internal Conflict

The human rights movement emerged in Peru before the spread of po-
litical violence in the country. These organizations arose during the
military authoritarian regime commanded by Velasco Alvarado (1968–
1980). The movement was nonpartisan: although these organizations
began with an affi nity with activists on the Left, no human rights or-
ganizations had a party affiliation (CVR 2003a; Youngers and Pea-
cock 2002).
Initially, these organizations were part of a secular movement in
the Catholic Church, a network of vibrant neighborhood and rural
organizations that emerged from the links formed between parishes
and the liberation theology movement that began in the 1970s (Levine
1990; Levine 2006). 28 In Peru, the Second Vatican Council, the epis-
copal conferences of Medellin and Puebla, and the expansion of lib-

Soifer_6844-final.indb 189 8/17/18 11:53 AM

 190 Maritza Paredes

eration theology guided the Church’s role into an activism “from be-
low.” Bishops and priests who went to the countryside to build a more
inclusive church found in the course of their work the growing po-
litical dynamism of indigenous communities, which were seeking to
strengthen their peasant unions, advocate for land titles, and (later)
protect themselves from the violence of Sendero. A discourse of human
rights became central to the Church’s effort to build its pastoral net-
work, which tied together a significant set of organizations among the
most excluded populations in society, including in particular rural, in-
digenous peasant communities.
Young members of these secular organizations, motivated by their
concern about civil detentions during the military regime, pushed for
the creation of a Human Rights Office in 1977 as part of the CEAS.29
Secular organizations had significant independence from the Church’s
hierarchy at the time: although initial support from bishops and
priests was crucial for the establishment of human rights organiza-
tions, these organizations were not subordinated to the ecclesiastical
hierarchy (CVR 2003a; IDL 2003). Indeed, during the armed conflict,
the Church took a highly conservative stance in areas where the worst
of the confl ict was occurring, such as in Ayacucho. 30
Human rights regional committees multiplied during the early 1980s
in the form of social organizations linked to the Church.31 The CNDH
appeared in the mid-1980s, and over time it became the mouthpiece
for a growing network of organizations concerned with human rights,
many of them with an important presence in rural areas and among
indigenous peasant communities. In the early 1980s, when political
violence seemed to have spread throughout the countryside, human
rights organizations across the country were overwhelmed by the mag-
nitude of human rights abuses. At that time their aim was to make vis-
ible the increasing number of forced disappearances, deaths, and tor-
tures at the hands of both terrorist groups and military forces, which
were ignored by the government in Lima and the international com-
munity (Youngers and Peacock 2002).32 The CNDH sought to unify
human rights appeals and to connect regional organizations with one
another, providing the movement with “one voice” to make interna-
tional demands. However, in places where the Church was not present,
it was almost impossible to obtain information about human rights vi-
olations. In the central Andes, where violence was most severe, it was
especially difficult. While organizations such as the mothers of the Na-
tional Association of Kidnapped, Detained, and Disappeared Relatives

Soifer_6844-final.indb 190 8/17/18 11:53 AM

 Indigenous Activism and Human Rights NGOs 191

in Emergency Zones (Asociación Nacional de Familiares de Secuestra-

dos, Detenidos y Desaparecido del Perú, or ANFASEP) demanded jus-
tice for their disappeared relatives, these organizations faced perma-
nent harassment from law enforcement. 33
The CNDH was not an attempt to unify groups from Lima and
other regions within one organization, given that they were very di-
verse, but rather to create connections among them and thus generate
knowledge and a national register of violations (interview with Jugo
2015). The result was a large and flexible network, capable of por-
traying the armed conflict’s consequences. In 1989 these organizations
staged the impressive “March for Peace” in the country, as well as the
fi rst visit of the Inter-American Commission on Human Rights to Peru
in 1990 (CNDH 2015).34
Over time, the increased association with international human rights
mechanisms in a difficult and repressive internal context strengthened
Peru’s human rights movement. In the 1990s the Fujimori government
deployed an aggressive antisubversive campaign that included illegal
tactics such as selective forced disappearance, arbitrary imprisonment,
and the annihilation of people and groups by state forces (CVR 2003a;
IDL 2003). Two groups created in the 1980s gained prominence in
these years by bringing together a significant number of young lawyers
interested in defending human rights in the context of Peru’s violence
and impunity from prosecution. These two groups were the APRODEH
and the IDL, formed in 1983 and 1984, respectively. Both institutions
centered their efforts on defending detained civilians and fi ling interna-
tional complaints regarding crimes against humanity. Both gained vis-
ibility in the public debate and in international circles upon taking ju-
dicial action.
Fujimori’s search for international legitimacy and support from in-
ternational fi nancial institutions presented an opportunity for a fur-
ther strengthening of the movement. Though it had gained legitimacy
due to its extreme economic and security policies (Morón and San-
born 2007), 35 Fujimori government faced consistent national and in-
ternational opposition to its methods due to human rights concerns
(Youngers and Peacock 2006).36 After the 1992 autogolpe, Fujimori
closed the Congress and caused the collapse of the national justice sys-
tem. Furthermore, alleged terrorists faced judgment in the so-called
faceless military courts (Roberts and Penecy 1997). In general, the jus-
tice system was full of abuses, and law enforcement representatives
who had committed human rights violations enjoyed a remarkable de-

Soifer_6844-final.indb 191 8/17/18 11:53 AM

 192 Maritza Paredes

gree of impunity (Burt 2009b; Degregori et al. 2003). The OAS, the
United States, and the international community began to take actions
that forced Fujimori to call for new congressional elections and a Con-
stituent Assembly. It was especially favorable for the human rights
movements that Fujimori’s government needed international legiti-
macy for fi nancial reinsertion just at the time that Bill Clinton took of-
fice as the US president in 1993. Clinton’s government “appointed pro-
Peruvian human rights sympathizers for key positions associated with
Latin America” (Youngers and Peacock 2006, 172) and placed pro-
human rights conditions on US support for Peru’s reinsertion in the in-
ternational fi nancial community.37
This same set of international pressures also compelled Fujimori to
call a referendum in October 1993 that resulted in the approval of a
new constitution, pursuant to which an Ombudsman’s Office was cre-
ated. Although the constitution had many questionable aspects, in-
cluding its treatment of the reelection issue and the placement of com-
munal property in jeopardy (see chapter 3), articles 161 and 162 of the
1993 constitution mandated the creation of the Ombudsman’s Office
as “an autonomous constitutional body, with the function of promot-
ing the protection of basic human rights and monitoring compliance
with state administrative duties and the provision of public services
to the general population.” The human rights movement in Peru, and
eventually indigenous organizations as well, gained a valuable ally
in the state with the creation of the Ombudsman’s Office, which not
only increased the movement’s popular legitimacy but also provided
the rare democratic credentials that the US government and the inter-
national community had sought (Pegram 2008; Roberts and Penecy
1997). As the free media, the legislative branch, and the judiciary be-
came limited, the Ombudsman’s Office rapidly transformed into a key
state agency, acting as a guarantor of democracy and human rights
vis-à-vis the government (interview with Luque 2014; interview with
Abad 2014). The ombudsman also took on an important role in re-
porting the corruption and fraud cases that resulted in the collapse of
the Fujimori regime.
After the defeat of the SL and the collapse of Fujimori’s govern-
ment, a democratic transition ensued, which included the creation of
a Truth and Reconciliation Commission (CVR). The CVR served as
a valuable site of professionalization for many organizations and ac-
tivists of the human rights movement—it generated “shared values,
a body of scientific knowledge, and procedures and systems to apply

Soifer_6844-final.indb 192 8/17/18 11:53 AM

 Indigenous Activism and Human Rights NGOs 193

that knowledge” (O’Flaherty and Ulrich 2009). This was an initiative

that emerged from both Valentín Paniagua’s transitional government
and the international human rights movement: the end of confl ict in
Guatemala, apartheid in South Africa, and dictatorships in Chile, Ar-
gentina, and Uruguay set off an international proliferation of Truth
and Reconciliation Commissions.38 The objective of these commis-
sions was not only to produce a cathartic and restorative process, but
also to prevent the emergence of new deadly conflicts (Gonzáles and
Varney 2013; Hayner 2011; Goodman and Pegram 2012).39 The CVR
in Peru brought together efforts of activists, intellectuals, and human
rights NGOs, who received an official mandate and US$19 million in
funding to work on the identification and registration of human rights
abuses and to train people committed to the cause.40 The CVR not
only increased visibility for human rights organizations through its ac-
tions, but also strengthened them through professionalization and pro-
motion of links with international cooperation.
When new conflicts between the state and indigenous organizations
arose in 2008, human rights organizations had accumulated signifi-
cant experience and professionalism in the legal defense of constitu-
tional rights, and represented an unprecedented resource for indige-
nous people. Although the history of the human rights movement in
Peru has not always been well connected to the indigenous movement,
human rights lawyers had encountered indigenous people in their ru-
ral work during the internal confl ict. Lawyers found these commu-
nities to be those most deprived of their fundamental human rights
during these years, and gained significant experience in working with
their organizations.

Legal Activism under Persistent Regional Fragmentation

Legal activism, however, does not compensate for the range of difficul-
ties that the indigenous movement has inherited from its regional frag-
mentation, which was a direct consequence of the internal confl ict.
Today, the Amazonian indigenous movement still rests on a dense net-
work of regional organizations, which counterbalance the complexity
and geographic breadth of the Amazon. Unlike in the Amazon, indige-
nous organizations in the highlands are characterized by a weaker net-
work of regional organizations. Some regional branches of the CCP
in places like Cusco and Puno are strong actors, but regional organi-

Soifer_6844-final.indb 193 8/17/18 11:53 AM

 194 Maritza Paredes

zations generally maintain low profi les and are weakly coordinated.
Some organizations of indigenous women have been created in an ef-
fort to diversify the agenda of the predominantly male indigenous or-
ganizations (Rousseau and Hudson 2016).
Table 7.1 shows a list of the national and regional organizations
that participated in the process after the Prior Consultation Law was
approved to negotiate the protocol for application of this norm. The
results of this process have been highly controversial, and this chap-
ter sets aside outcomes including limitations on the technical, legal,
and political specifications that can be made to the Prior Consulta-
tion Law and its approved regulations (see Benavides 2012; Benavides
2010; Sanborn, Hurtado, and Ramírez 2016; Schilling-Vacaflor and
Flemmer 2013; Sosa Villagarcia et al. 2012). I simply highlight instead
the more active participation of regional branches of the Amazonian
Movement, which is reflected in the table. These regional organiza-
tions, moreover, made decisions independently from their national or-
ganizations while negotiating and agreeing upon regulations of the
Prior Consultation Law. They allowed the process to continue moving
forward and to be implemented in spite of the opposition of national
leaders (Sosa 2017).
Regional differences between Amazonian and Andean indigenous
organizations still influence how legalized structures are applied and
operate. During the fi rst years of the law’s implementation, it was ap-
plied primarily to hydrocarbon projects and other projects in the Am-
azon. Mining projects that mostly operate in the Andes continued to
be approved in indigenous communities without implementing the
Law on Previous Consultation for several more years.41 It was only five
years after the Prior Consultation Law was approved and began to be
applied, and after a persistent campaign by NGOs and the Ombuds-
man’s Office, that the government announced the norm also would be
applied to mining in the highlands.
Another example of the persistent weakness of Andean indigenous
organizations relative to the Amazonian organizations was the publi-
cation of the Official Database of Indigenous Peoples compiled by the
Vice-Ministry of Interculturality. This is one of the most important
state instruments regarding which communities are officially entitled
to exercise indigenous rights. According to an interview with Paulo
Vilca, a former vice minister of interculturality, the database con-
tained indigenous peoples from both regions, the Amazon and the An-
des, when it was fi rst organized in January 2012, but it was not pub-

Soifer_6844-final.indb 194 8/17/18 11:53 AM

 Indigenous Activism and Human Rights NGOs 195

Table 7.1. Indigenous organizations participating in the deliberations

over the elaboration of the Prior Consultation Law

Organization Acronym Affiliation

Asociación de Nacionalidades Ashaninka del ANAP AIDESEP

valle Pichis
Asociación Regional de Pueblos Indígenas de ARPISC AIDESEP
Selva Central
Indígenas de Selva Central Arpisc Central CARE AIDESEP
Ashaninka del Río Ene
Central de Comunidades Nativas de la Selva Ceconsec AIDESEP
Central
Comisión Especial Permanente de los CEPPAW –
Pueblos lndígenas Awajún Wampís
Comité de Gestión del Bajo Urubamba CGBU –
Consejo de Nacionalidades Indígenas del Conaip AIDESEP
Perú
Consejo Machiguenga del Río Urubamba Comaru AIDESEP
Coordinadora Regional de los Pueblos Corpi-SL AIDESEP
Indígenas de San Lorenzo
Federación de Comunidades Ashaninka del Fecomabap AIDESEP
Bajo Perené
Federación Indígena Regional y del Alto Mayo Feriaam AIDESEP
La Asociación Indígena de Estudiantes Aaupi –
Universitarios de la Amazonia Peruana
Organización de Pueblos Indígenas del Orpio AIDESEP
Oriente
Organización Regional de los Pueblos Orpian-P AIDESEP
Indígenas de la Amazonia Peruana del Norte
del Perú
Unión de Comunidades Aymara Unca –

Source: Compiled by Politai (Sosa Villagarcia et al. 2012).

lished, largely as a result of the mining industry sector’s opposition to

the inclusion of Andean indigenous communities. This statement is in
line with the declarations of another vice minister of interculturality,
Iván Lanegra (Lanegra 2015), who stated that the mining industry’s
opposition to publishing the Database of Indigenous Peoples derived
from the fact that it showed the extent of overlap of indigenous com-
munity territories with major mining projects.42 Despite Lanegra’s res-

Soifer_6844-final.indb 195 8/17/18 11:53 AM

 196 Maritza Paredes

ignation under pressure from the government, there were no regional

organizations that supported or pushed the publication of this data-
base, even though communities in conflict with mining projects in the
Andes had frequently claimed their right to prior consultation. These
communities, however, are local and not linked to a regional organiza-
tion. The database was fi nally published at the end of 2013, and many
Andean communities are now learning to draw on the new rights it en-
tails (Málaga and Ulfe 2017).
According to Vilca, there were few links of solidarity between Am-
azonian and Andean organizations during these debates. For Amazo-
nian peoples, the Andean people’s fight to be recognized in the data-
base constituted a problem because it delayed many prior consultation
processes in the Amazon region. In fact, there is still no defi nitive list
of communities recognized as native peoples, but rather only a “pre-
liminary” list. It is significant that the consultation process to be im-
plemented in the mining sector within the Andes appears to be the
result of pressure from the Ombudsman’s Office and human rights or-
ganizations rather than Andean indigenous movement efforts.
In summary, the major accomplishment of Amazonian organiza-
tions has been to establish consultation as a legal space of negotiation
between the state and indigenous peoples. In this legal space, human
rights organizations working on legal cases involving regional indig-
enous organizations are the main facilitators, and they are working
with communities in both regions. However, it is hard to foresee how
these legal resources will interact with the legacy of regional fragmen-
tation. Legal resources open new opportunities for indigenous activ-
ism. This activism may strengthen indigenous organizations, but it re-
mains to be seen whether this will be enough to transcend the legacy
of fragmentation and build a fi rmer national movement.

Conclusions

In this chapter, I have examined the unexpected and multidirectional

legacies of the internal war on indigenous mobilization today by ex-
plaining the politics of extractive development in Peru that provide a
new contemporary context for indigenous mobilization in the twenty-
fi rst century. I have aimed to analyze the legacies of political violence
within this new context that produced the violent clash between in-
digenous protesters and the police in Bagua. Some institutional trans-
formations, such as the Prior Consultation Law, illustrate a different

Soifer_6844-final.indb 196 8/17/18 11:53 AM

 Indigenous Activism and Human Rights NGOs 197

form of activism where legal allies such as the Ombudsman’s Office,

NGOs, and other human rights stakeholders play a crucial role, and
this change is a legacy of the armed conflict. I have shown how indig-
enous communities in Peru have found the support to use the influ-
ence of the human rights and judicial systems for mobilization. I con-
clude that the Peruvian human rights actors working on indigenous
issues and supporting mobilization and legal claims are a legacy of
the internal war. Last but not least, I argue that the post-confl ict pe-
riod extended the participation and enhanced the professionalization
of many of these actors through the Truth and Reconciliation Com-
mission and, more recently, the Ministry of Interculturality. However,
these new legal resources and capacities coexist with the still deep di-
vision between Andean and Amazonian indigenous organizations,
thus reducing the opportunities for encompassing indigenous activism
at a national level.

Interviews

Abad, Samuel. Former member of Ombudsman’s Office. Lima, April 2014.

Abanto, Alicia. Former member of the Department for Indigenous Peoples,
Ombudsman’s Office. Lima, October 2013.
Burneo, Zulema. Former member of Vice Ministry of Intercultural Affairs;
member of International Cooperation. Lima, September 2013.
Coronado, Hernan. Former member of the CAAAP, the CNDH, Vice Minis-
try of Intercultural Affairs. Lima, September 2013 and July 2015.
Jugo, Miguel. Member of the CNDH. Lima, July 2015.
Lanegra, Ivan. Former vice minister of Intercultural Affairs, Ombudsman’s
Office. Lima, September 2013.
Luque, Rolando. Former CVR commissioner, Ombudsman’s Office. Lima,
April 2014.
Quinn, Albano. Former archbishop. Puno, October 28, 2008.
Sanchez, Diego. Current Indigenous People Office commissioner, Ombuds-
man’s Office. Lima, July 2015.
Soberón, Francisco. Member of the APRODEH. Lima, July 2015.
Torres, Javier. Member of SER. Lima, July 2015.
Vilca, Paulo. Former member of Vice Ministry of Intercultural Affairs. Lima,
October 2013.
Zambrano, Gustavo. Former director of INDEPA. Lima, September 2013.

Notes

1. “Indigenous origin” refers to those who have an aboriginal language as

their mother tongue.

Soifer_6844-final.indb 197 8/17/18 11:53 AM

 198 Maritza Paredes

2. Although indigenous communities in Ecuador and Bolivia also created

“class-based” organizations influenced by leftist parties, communities in the
countryside were free of radical class-based Marxist ideologies. Rather, par-
ties from the Left such as the Kataristas ended up introducing elements of an
ethnic, pro-indigenous discourse into their ideology. See Paredes 2011.
3. Other such parties included Ecuarunari (Confederación Kichwa del
Ecuador), CONAMAQ (Consejo Nacional de Ayllus y Markas de Bolivia),
and CSUTCB (Confederación Sindical Única de Trabajadores Campesinos de
Bolivia.)
4. In her influential book, Deborah J. Yashar argues that unlike in Bolivia
and Ecuador, rural zones in Peru (and Guatemala) became a complicated set-
ting for political and social organization because of the outbreak of the inter-
nal war (Yashar 2005, 249). According to Yashar, the Shining Path destroyed
potential frameworks for legal organization along ethnic lines (247–249).
5. Confederación Campesina del Perú (CCP) bases were more dynamic in
the southern highlands (where the concentration of indigenous populations is
higher) at the end of the 1980s during the worst period of violence in the high-
lands. By 1987, the total number of delegates attending the National Congress
of 1987 was twice the number of the delegations in 1978, and seven times the
number in 1974.
6. Land titling in Peru has been poorly enforced. The official public count
is 1,647 and 1,823 Andean and Amazonian communities, respectively. Nev-
ertheless, other organizations, NGOs, and academic institutions estimate a
larger number. The IBC (Instituto del Bien Común) calculates that there are
666 native communities and 3,303 peasant communities waiting for land ti-
tling, but the Vice Ministry of Intercultural Affairs argues that there are only
500 communities that lack land titles.
7. The case of the Uwa Indigenous Community in Colombia is an example
of indigenous struggle for land titling, sovereignty, and the importance of in-
ternational legal mechanisms and transnational activism in establishing col-
lective rights. See Rodríguez-Garavito and Sousa Santos 2007.
8. This was done through the Law Project 840, the so-called Ley de la
Selva.
9. The fi rst article appeared in October 2007, the second in November
2007, and the third in March 2008.
10. See Perú: Informe alternativo 2013 sobre el cumplimiento del Con-
venio 169 de la OIT (Derecho, Ambiente y Recursos Naturales [DAR], 2013).
11. The eight organizations were the AIDESEP, —CCP, the Confeder-
ación Nacional Agraria (CNA), the Confederación de Comunidades del Perú
Afectadas por la Minería (CONACAMI), the Organización Nacional de Mu-
jeres Indígenas del Perú (ONAMIAP), the Federación de Mujeres Campesina,
Rurales, Indígenas, Nativas, Asalariadas del Perú (FEMUCARINAP), the
Central Única Nacional de Rondas Campesinas del Perú (CUNARC), and the
Unión Nacional de Comunidades Aymaras (UNCA).
12. See https://www.servindi.org/actualidad/4799.
13. See https://www.servindi.org/actualidad/4086.
14. See https://www.servindi.org/actualidad/3258.

Soifer_6844-final.indb 198 8/17/18 11:53 AM

 Indigenous Activism and Human Rights NGOs 199

15. See http://larepublica.pe/politica/844884-santiago-manuin-un-guerrero

-por-la-paz.
16. See http://www.noticiasser.pe/26/04/2014/nacional/reconocido-dirigente
-awajun-y-ambientalista-santiago-manuin-denuncia-persecucion.
17. Historically, the CAAAP worked closely with communities and indig-
enous organizations, supporting diverse projects with the main objective of
promoting development in Amazonian indigenous zones.
18. Other organizations that began developing during the 1990s were
more focused on conservation and environmental issues, such as SPDA, Pro-
Naturaleza, DAR, Instituto del Bien Común (IBC), and national affi liates of
WWF, TNC, and CI.
19. The peasant patrols and self-defense committees (CAD) were organi-
zations that played a crucial role in the war against the Shining Path. Upon
becoming trapped in the violence between the military and the terrorists, the
population understood that it had to organize to survive. In the northern and
southern highlands, there were already peasant patrols before the Shining
Path arrived (Starn 1999). In the central southern highlands, where violence
was more intense, CADs arose as a product of the alliance between the mili-
tary and peasants (Starn 1999; Degregori et al. 1996).
20. The original source is drawn from the Joint Command of the Armed
Forces 1993. See del Pino 1996.
21. It is important to mention that from the late 1990s on, the Ombuds-
man’s Office contained a unit specifically formed to address indigenous issues
(Defensoría del Pueblo 2015).
22. International cooperation after Bagua increased for many of these or-
ganizations in support of the effective application of rights such as prior con-
sultation and the right to communal territory.
23. These organizations included the Asociación Paz y Esperanza,
APRODEH, CAAAP, Centro de Derechos y Desarrollo (CEDAL), Cooper-
Acción, CEAS, DAR, Forum Solidaridad Perú (FSP), Fundación Ecuménica
para el Desarrollo y la Paz (FEDEPAZ), IDL, IBC, SER, and Servicios en Co-
municación Intercultural (SERVINDI).
24. File no. 00014–2008-PI/TC, Records of the Tribunal Constitucional.
25. See http://derechoshumanos.pe/2009/04/pronunciamiento-sobre-el-paro
-indigena/.
26. It is worth noting that organizations concerned with rural issues, such
as Centro Peruano de Estudios Sociales (CEPES), or environmental issues,
such as DAR and SPDA, also produced observations, thanks specifically to
their group of legal specialists.
27. Latin American indigenous mobilization in the twenty-fi rst century is
characterized by a struggle against nation-states and transnational corpora-
tions through networks of transnational activism and the use of legal strate-
gies (Rodríguez-Garavito and Sousa Santos 2007).
28. Liberation theology had a profound influence not only in the religious
and ecclesiastical world, but also on the political world and civil society, espe-
cially in grassroots organizations (Levine 1990; Levine 2006).
29. The CEAS was created on March 11, 1965, under the framework of

Soifer_6844-final.indb 199 8/17/18 11:53 AM

 200 Maritza Paredes

the Second Vatican Council and is a service organ of the Peruvian Episcopal
Conference. See http://www.ceas.org.pe/nosotros.php?n=1.
30. Indeed, the CVR criticized the Catholic Church’s failure to take a
stand against the abuses in Ayacucho. The archbishop of Ayacucho, Juan Luis
Cipriani, today cardinal of Peru, publicly criticized human rights organiza-
tions. “La Coordinadora de Derechos Humanos es una cojudez” (“The Hu-
man Rights Coordinating Committee is bullshit”) he said once to the press
(Revista Caretas, April 14, 1994).
31. In the southern highlands, for instance, the Vicaría de la Solidaridad
was created by Monsignor Albano Quinn “to defend human rights, and the
dignity of the person.” Interview with Monsignor Albano Quinn 2008. In
Cajamarca’s northern highlands, Monsignor Dammert built a church with
Poncho and Sombrero to serve people and satisfy their fundamental rights
(Knecht 2005).
32. In 1983, a group of mothers from Ayacucho in search of official an-
swers regarding the disappearance of their relatives founded ANFASEP.
ANFASEP is the fi rst organization of its kind and an example of local dynam-
ics in emergency zones that found a voice at a later time than the rest of hu-
man rights organizations.
33. Its president, Angélica Mendoza, known in Peru as “Mama Angélica,”
tirelessly searched for disappeared relatives in Peru for almost three decades,
despite receiving a death threat after the fi rst day that she denounced her son’s
disappearance. The original citation is from Amnesty International 1989.
34. In 1990 the Pro-Human Rights Organization of Spain gave its first in-
ternational recognition to the CNDH, giving it “Special Advisory Status” be-
fore the United Nations Social and Economic Council. The CNDH is also ac-
credited to participate in OAS activities.
35. Among these policies, there was a structural adjustment package, the
privatization of many state companies, continuing the payment of the exter-
nal debt, which restored international credit, the especially controversial sign-
ing of anti-drug agreements with the United States; and fi nally, the intensify-
ing of the Auto-Defense Committees (CAD) (Morón and Sanborn 2007, 27).
36. What began as an antisubversive strategy eventually became state ter-
rorism. According to Burt, both the Shining Path and the state used fear and
intimidation to “destroy the moral and material base of civil society organiza-
tions” (Burt 2009b, 28).
37. During their fi rst official visit to Peru, Clinton’s representatives met
with the staff of the National Coordinator of Human Rights before seeing Pe-
ruvian government officials (Youngers 1994, 45). Fujimori’s economic policy
applied structural adjustment policies and relied on the country’s reinstate-
ment into the international fi nancial community. In this context, Clinton was
in a favorable position to pressure Fujimori’s government to apply at least four
human rights measures: a guarantee of access to detention centers for the In-
ternational Red Cross Committee, the acceptance of United Nations Human
Rights Commission assessment and a visit of the International Commission
of Human Rights of the OAS, the initiation of official dialogue with the hu-
man rights coordinator, and public recognition of the importance of working

Soifer_6844-final.indb 200 8/17/18 11:53 AM

 Indigenous Activism and Human Rights NGOs 201

with international and national human rights groups (Youngers and Peacock
2006, 172). Given that Fujimori’s government needed the credit, it had to ac-
cept these requested measures.
38. On the international level, between 1974 and 2009, forty Truth and
Reconciliation Commissions were established in all types of regimes and so-
cioeconomic contexts (Hayner 2011).
39. Despite the unique features of each Truth Commission, most of them
share basic characteristics: they complement the criminal judicial system of
each country, focus primarily on severe human rights violations, establish re-
search time periods, produce a great amount of information, and have a vic-
tim-focused approach (Gonzáles and Varney 2013, 14).
40. The CVR was funded by a cooperative agreement with the United Na-
tions Development Program (PER 01/023). Other funds came from the state,
but also from international agencies such as USAID, German Cooperation
GIZ, Belgian Development Cooperation, and the European Union. See http://
www.cverdad.org.pe/lacomision/ifi nanciera/fi nanciamiento.php.
41. The fi rst prior consultation for indigenous peoples related to mining
projects will take place in the community of Parobamba (Calca, Cusco). This
Quechua peasant community will be consulted about the exploration project
of Minera Aurora Company.
42. “Los secretos mineros de trás de la lista de comunidades indígenas del
Perú.” Ojo Público, July 22, 2015, accessed August 6, 2015, http://ojo-publico
.com/77/los-secretos-detras-de-la-lista-de-comunidades-indigenas-del-peru.

Soifer_6844-final.indb 201 8/17/18 11:53 AM

 CHAP T ER 8

Political Violence and the Defeat of the Left

Paula Muñoz

We do not agree with the use of terrorist methods, because now they
will only contribute to inciting repression, isolating the Left from
the people and giving arguments to the right and the government to
reduce our margins of action.
Rol a n do Br e ña , l e a de r of PCdel P –Pat r i a Roja
(au t hor’s t r a nsl at ion)

This poor provincial cousin, previously scorned by its Communist

relatives, arrived in Lima without asking anyone’s permission,
burned down the house, and ended up with the family name.
I vá n H i nojosa 1

For the new generations of Peruvians it comes as a surprise to learn

that during the 1980s there was a political front called Izquierda Unida
(IU) that governed Lima, was represented in the parliament, and was
an active participant in public debates. This is shocking considering the
widespread attention that the IU once received within Latin America.
To be certain, after the transition to civilian rule, the Peruvian Left was
exceptionally strong in the region, both electorally and socially, and it
continued to be so for most of the 1980s (Roberts 1996, 69).
The political “ignorance” of the new generations is partly explained
by the tremendous discredit that the whole traditional party system
(and politics itself) has suffered since its collapse in the 1990s in Peru.
However, new generations are at least aware of the existence of other
traditional parties, such as the Alianza Popular Revolucionaria Ame-
ricana (APRA), the Partido Popular Cristiano (PPC), and even Acción
Popular (AP). These parties were able to “resuscitate” during the early

Soifer_6844-final.indb 202 8/17/18 11:53 AM

 Political Violence and the Defeat of the Left 203

2000s (Kenney 2003), although they faced many problems (Levitsky

2013; Levitsky and Cameron 2003; Tanaka 2005). In contrast, Peru
so far has not seen the revival of an electorally viable and enduring
leftist political party.
In this chapter, I attempt to account for the limited recent achieve-
ments of the Left by examining what happened during the 1980s when
a competitive leftist group did exist. I will appraise the ways in which
political violence in that era affected the Left in Peru. Were the Left’s
fortunes shaped by the conflict? In particular, how did the Partido
Comunista del Perú–Sendero Luminoso (SL)’s decision to initiate an
armed struggle against the Peruvian state and the democratic political
system impact the fate of the rest of the Left? Moreover, can we iden-
tify post-confl ict legacies that at least partly explain the difficulties
faced today by those actors who want to organize and represent this
side of the political spectrum?
To answer these questions, I distinguish between the political vio-
lence’s immediate effects and the war’s long-term effects (or political
legacies) on the leftist camp. While the wartime mechanisms that un-
folded during the confl ict help to explain the weakening of the Left
and its practical disappearance during the 1990s, the legacies (and
their political manipulation) help us to better understand the diffi-
culties faced by leftists in reorganizing today. I contend that the le-
gal Left’s initial fear, well expressed by Rolando Breña, was fulfilled.
With the end of political violence came not only the military defeat of
SL but also the political downfall of the legal Left.
Two wartime mechanisms directly weakened the Left. First, the SL’s
armed uprising posed a serious ideological challenge for the IU front.
Not being able to provide a clear and unified stance about the use of
violence as a political means to an end, the IU ultimately split. Thus an
important consequence of the SL’s rise in arms was the ideological di-
vision of the IU. Second, through other wartime mechanisms, violence
also weakened and dispersed the Left. The IU ranks and militants be-
came a target of direct physical violence and repression from both SL
and the state. Moreover, the IU’s ranks defected to the two groups
that had risen in arms. The escalation of violence and fear also made
it more difficult for the IU to recruit new members. And after the IU’s
split, the Fujimori government harassed and/or coopted former leftist
cadres as well, making it more difficult for the Left to reorganize.
A third factor not related to the conflict, the organizational lega-
cies of the Left at the time of its inception, provides a complementary

Soifer_6844-final.indb 203 8/17/18 11:53 AM

 204 Paula Muñoz

explanation of why the IU split. There was simply no party organiza-

tion capable of making collective decisions and reacting strategically
toward moderation. In addition, and particularly important, this non-
war mechanism allows us to understand the emergence of the fi rst leg-
acy of the confl ict for the Left. The IU’s incapacity to provide a clear
and unified position against the use of revolutionary violence helped
the right-wing and state actors to normalize the association of the Left
with terrorism in the political discourse.
Finally, the association of the Left with terrorism and Fujimori’s ac-
tive manipulation of this legacy buried the political chances the Left
had for effectively becoming an institutionalized and electorally ap-
pealing political party after the confl ict was over. Moreover, the per-
sistence of this legacy and its manipulation still poses a great challenge
for leftist actors and may lessen their chances of improving electorally.
I organize this chapter in the following way. First, I discuss the ori-
gins of the Left in Peru and the creation of the IU. Second, I show how
the rise of the SL posed an ideological challenge for the Left that even-
tually led to the split of the IU. Third, I present and discuss the orga-
nizational approach to the division of the Left. Fourth, I distinguish
other wartime mechanisms that weakened the Left. In these sections, I
discuss the causal processes that destroyed the Left’s political chances
for becoming an institutionalized political party. Subsequently, I con-
tend that the main legacy for the Left is stigmatization: its association
with terrorism. This legacy was successfully manipulated by Fujimori,
who simultaneously attacked the parties on the left side of the political
spectrum while gaining the popular sector’s trust and coopting several
local cadres of an already shaky legal Left. I conclude by discussing
how these processes shaped the political defeat of the Left and made
the chances of reconstructing and consolidating an electorally appeal-
ing political party a daunting project.

The Origins of the Peruvian Left

Ideologically, the development of the Left in Peru has been circum-

scribed to the Marxist-Leninist camp. Peru did not experience the
emergence of a difference between socialism and Communism or be-
tween a social democratic Left and a Communist radical one, as hap-
pened in other Latin American countries (Adrianzén 2011, 45). More-
over, also in contrast with other countries in the region, the Peruvian

Soifer_6844-final.indb 204 8/17/18 11:53 AM

 Political Violence and the Defeat of the Left 205

Left that eventually joined the IU did not have a long partisan tradi-
tion (interview with Hinojosa 2015; see also Letts 1981). The Peruvian
Left developed a mass political base principally in the 1970s (Rob-
erts 1996, 71). In that decade the leftist cadres and ranks grew signifi-
cantly, despite the fragmentation of its bases (Rochabrún Silva and Ya-
ñez 1988).
The Peruvian Communist Party (PCP), originally founded as the
Socialist Party by Peruvian Marxist José Carlos Mariátegui in 1929,
remained small after its inception (Adrianzén 2011). This party had
to compete for popular support with Víctor Raúl Haya de la Torre’s
party, APRA, which was founded in 1930. Both parties were pro-
scribed from the political system, but while APRA managed to expand
its ranks and become a mass party despite political repression (Vega-
Centeno 1991; Collier and Collier 1991), the PCP remained relatively
small and marginal. APRA’s triumph reinforced the PCP’s political iso-
lation and dependence on the Third Communist International, not al-
lowing room for the development of Democratic Socialism (Adrianzén
2011, 46). Indeed, during the 1930s and early 1940s the PCP showed
not only uncritical submission to the Soviet party line but also accom-
modating behavior that made it suspicious of a “rightward” transfor-
mation (Guadalupe 1988).
Despite APRA’s preeminence, the Marxist Left continued its devel-
opment in the following decades. Peruvian Communism grew to be
principally inspired by the Chinese and Cuban revolutions. In particu-
lar, Mao’s revolution offered a very persuasive model for Peruvian left-
ists disillusioned with the PCP. They found many parallels between
Mariátegui’s indigenism and the Maoist focus on the peasantry (Na-
varro 2010, 157; Rénique 2003, 41). This ideological proximity was
further cultivated by the Chinese government through funding and
training of Peruvian cadres. Increasing disagreement within the PCP
ranks provoked a Sino-Soviet split in 1964 between the PCP-Unidad
and the PCP-Bandera Roja. Through the years to come, Maoism would
take hold intensively in Peru, acquiring unprecedented levels of influ-
ence by Latin American standards (Hinojosa 1998).
For its part, the Cuban revolution inspired the rise of armed guerrilla
movements in the early 1960s. A group of Communist APRA dissi-
dent youngsters (the Movimiento de Izquierda Revolucionaria, MIR)
advanced into the jungle to wage its warfare, while another group, the
Ejército de Liberación Nacional (ELN), did the same in southern Peru.
The guerrillas were quickly and brutally crushed by the military in

Soifer_6844-final.indb 205 8/17/18 11:53 AM

 206 Paula Muñoz

1965. While their existence was brief and disastrous, the dead leaders
became martyrs who inspired a new generation of leftists to follow the
insurrectional path for political change.
Thus, in the years following the guerrillas’ defeat, Peru saw the
rapid expansion of the so called new Left. This was an insurrectional
Left, critical of the PCP and the Soviet approach to Communism (Ré-
nique 2003, 40–41). While the new Left included a myriad of ideo-
logical strands—Trotskyists, Christian Left, “campesinistas” (pro-
peasants), and Maoists (Rénique 2003, 41)—Maoism gradually gained
widespread acceptance during the 1970s (Rochabrún Silva and Yañez
1988; Hinojosa 1998). 2
Amid increasing political polarization, in 1968 the Peruvian military
seized power and launched a progressive military regime that enacted
bold reforms long demanded and fought for, including the agrarian re-
form. These reforms were carried out through an intense campaign
that called for the elimination of exploitation but claimed to be nei-
ther capitalist nor Communist (Rochabrún Silva and Yañez 1988, 78).
The military government was very permissive with the Left. This al-
lowed for its considerable growth. With policies promoting industrial-
ization and the expansion of public education, unionization expanded
significantly, and the Left took advantage of it. At the same time, left-
ists were barely repressed by the military government, particularly
when compared with the harsh-line authoritarian regimes most Latin
American countries experienced (Nogueira-Budny 2013). Indeed, the
military worked instead to weaken APRA’s grip on the labor and the
student movement (Sanborn 1991; Hinojosa 1998), 3 thus benefiting
the Left.4 Furthermore, the military promoted the mobilization and
organization of the popular sector through the Sistema Nacional de
Movilización Social (SINAMOS) (Dietz 1980). However, the regime
did not build a party to politically channel these masses, leaving them
available for mobilization from the Left (McClintock and Lowenthal
1983; Stokes 1995).
More important, the Peruvian military experiment not only al-
lowed the Left’s expansion but also unintentionally radicalized it (Ro-
chabrún Silva and Yañez 1988; Stokes 1995; Sanborn 1991; Hinojosa
1998). As Rochabrún Silva and Yañez point out,

With the exception of the Partido Comunista Peruano (Peruvian Com-

munist Party, or PCP), the Marxist left criticized and combated this
government and these reforms, mostly in ideological terms, trying to

Soifer_6844-final.indb 206 8/17/18 11:53 AM

 Political Violence and the Defeat of the Left 207

prevent the popular classes from being attracted by the rhetoric of the
generals. The countless parties of the Left that appeared then sought
to demonstrate that such policies were reactionary, or a hoax, or not
sufficiently revolutionary. In any case, the greatest efforts on the part
of the Left were exerted trying to differentiate themselves from the
military government (Rochabrún Silva and Yañez 1988, 79).

During the second (conservative) phase of military rule, nascent left-

ist groups connected social discontent with rising economic problems,
organizing a national strike in 1977 that toppled the military govern-
ment. Legal restrictions on participation in the upcoming elections
were not imposed on the Left during either the second phase of the
military regime or the initial phase of the democratic regime (Sanborn
1991); even the most radical militants who explicitly eschewed democ-
racy and publicly embraced revolution were not prohibited. Thus, the
country was surprised when a number of the parties on the Left not
only competed in elections for the fi rst time in Peru in 1978, but won
(together) almost one-third of the seats in the Constituent Assembly.5
The Left had emerged as a powerful political actor, although a frag-
mented one.

No Ideological Change

In assessing the legacies of political violence, it is essential to think of

1980 as a critical turning point for the Left. Two important changes
occurred during this year. On the one hand, the SL declared its war
against the political system and rose in arms on the same day the tran-
sitional election was being held. On the other hand, after a poor show-
ing in the 1980 congressional elections, several left-wing political par-
ties decided to join forces and create an electoral front (the IU) to
improve their chances in the upcoming municipal elections.6 These di-
vergent choices and the tensions each created would mark the Left for
the rest of the decade.
An early consequence of the SL’s rise and of the war itself was the
ideological challenge that the confl ict implied for the rest of the Left.
Surprised by the SL’s decision to take up arms, the IU was not able to
provide a clear, unambiguous, and coherent position on the SL, the use
of violence as a political means, or the SL’s participation in the demo-
cratic system.

Soifer_6844-final.indb 207 8/17/18 11:53 AM

 208 Paula Muñoz

Despite ratifying their decision to participate in the legal political

system in 1980, most of the political parties that had formed the IU
did not reject revolution as their ultimate political goal (Pásara 1990;
Tanaka 1998; Hinojosa 1998; CVR 2003a). On the contrary, as the
Truth and Reconciliation Commission (CVR) has documented, within
the IU, participation in the so called bourgeois democracy was con-
ceived as a transitory stage that would allow the Left to accumulate
strength before conducting a future “revolutionary transformation” of
society (CVR 2003a, 177–178). As Diez Canseco7 says, within the IU,

the electoral, the democratic realm, was subordinated to the accu-

mulation of revolutionary forces. This goal prioritized mass strug-
gle and included the vision that all forms of struggle were pertinent
to achieving revolution. At the same time, this shows the vision that
we, as members of this political period, shared: it was a propitious
time for revolution. Our foundational declaration says “We fight for
the destruction of the Bourgeois state and the conquest of a Govern-
ment that arises from the masses’ revolutionary action.” (Diez Can-
seco 2011, 107)

The SL’s armed insurgency questioned the legal Left’s discursive radi-
calism. During the initial phase of their armed struggle, however, the
IU minimized the SL’s insurrectional potential, ignoring and despising
their actions. The SL was very radical, but still a part of the leftist Pe-
ruvian tradition (Hinojosa 1998) and more generally of the Peruvian
radical tradition of thought (Rénique 2003). Within the Left, the SL
was always perceived as the “provincial and poor relative” within the
“family” of Maoist organizations (Hinojosa 1998). While most Mao-
ist parties were composed predominantly of young Andean migrants,
the SL distinguished itself by virtue of its local, parochial, and dog-
matic character (Hinojosa 1998, 76–77). Consequently, leftist mili-
tants despised this dogmatic, provincial group, underestimating its in-
surrectional capability. And as the SL’s initial actions were attacks in
rural areas, information about them was scarce and imprecise (inter-
view with Zapata 2015).
As the SL’s actions increased and became more aggressive, most
leaders on the Left questioned not the method but the form (terror-
ist acts) and timing of the Shining Path’s struggle (CVR 2003a, 180).
Certainly, differences existed within the IU’s party members regarding
what exactly revolution meant and how it should be achieved. For in-

Soifer_6844-final.indb 208 8/17/18 11:53 AM

 Political Violence and the Defeat of the Left 209

stance, Maoists understood revolution as warfare, but the “izquierda

clasista” had a revolutionary discourse that understood revolution
from a Bolshevik perspective—as demonstrations and street fighting—
rather than equating it to war (interview with Zapata 2015). How-
ever, while these nuances existed internally and some groups openly
criticized the SL from the beginning, the image they continued to pro-
vide externally was one of radicalism. It was (and still is) impossible
for uninitiated people to understand and follow all the intricacies of
these ideological discussions within the Left. For an outsider, all left-
ists were the same, especially since the IU indeed had a “double dis-
course” that accepted “all the forms of struggle,” including the elec-
toral struggle as well as the armed one (interview with Mejía 2015. See
also Diez Canseco 2011, 100; Guerra García 2011). As the IU militant
Susana Villarán recognizes, “I believe that the motto ‘power is born
from the rifle’ continued even after hearing about the frightening mas-
sacres Sendero committed against the poorest peasant people. That
is in the testimonies the Truth and Reconciliation Commission gath-
ered. I believe this is something that must be acknowledged[;] we must
say that we were wrong” (interview with Villarán in Adrianzén 2011,
501). The IU’s position appeared at best as ambiguous and did not al-
low their members to provide a convincing position to the public. As
the CVR stressed,

By sharing a similar ideological matrix, these parties could not take a

clear stance regarding the “revolutionary violence” issue, nor did they
mark boundaries with PCP-SL and MRTA’s actions and thought. This
ambiguity and lack of defi nition towards democracy were the seeds of
future tensions within IU that ultimately led to its rupture when the
parties that comprised it ventured to face them. (CVR 2003a, 181)

Undeniably, the IU’s ambiguity toward revolutionary violence was

even more striking when judging the actions of the Movimiento Revo-
lucionario Túpac Amaru (MRTA). The MRTA was ideologically and
socially closer to other non-Maoist groups within the IU than to the
SL (CVR 2003a; interview with Zapata). Moreover, in contrast to the
SL, the MRTA did not see its political-military project as alternative
or exclusive to the participation of other political parties on the Left
(CVR 2003a, 388). Indeed, the MRTA’s political diagnosis was sim-
ilar to the one developed in other leftist parties within the IU. The
MRTA believed the conflict’s militarization would lead to an armed

Soifer_6844-final.indb 209 8/17/18 11:53 AM

 210 Paula Muñoz

confrontation that would end with a “fascist” military coup which

would wipe out all the legal Left as well. As Hinojosa explains, there
was a difference between believing that you were initiating an armed
struggle, as the SL did, and thinking that the situation was inevita-
bly leading to a general armed confrontation (interview with Hino-
josa 2015), that is, to a revolutionary situation. Thus, while many mil-
itants and leaders within the IU easily settled their differences with the
SL, even after the beginning of their armed struggle (interview with
Hinojosa 2015; interview with Zapata 2015),8 it was more difficult
for them to distance themselves from the MRTA’s actions and to es-
chew revolution as a legitimate political path. This had to do with left-
ist parties’ origins and their insurrectional horizon. The IU was am-
bivalent mostly about the use of violence for making the revolution: “it
was a verbal resource, an aspect of emotional identity” (interview with
Ronaldo Ames in Adrianzén 2011, 211).
In this context, elections and democratic institutions were, indeed,
an enormous additional challenge for the legal Left. Most of these par-
ties “were to apply the Leninist sentence of using the elections and
the parliaments as tribunes of agitation and propaganda, but ended up
trapped by the system they intended to transform (or destroy)” (Hino-
josa 1998, 86). Within the IU, “we lived in a schizophrenic manner:
there was an ardent discussion about how power would be taken by vi-
olence, while we already governed many district and provincial munic-
ipalities, and in Lima and elsewhere in the country the leaders fought
tooth and nail with each other to access positions in Parliament” (in-
terview with Santiago Pedraglio, former member of the IU, in Adrian-
zén 2011, 470). Moreover, distracted by the violence and the question
about their position on the increasing violence, the IU did not take full
advantage of its experience of local government to produce a program-
matic alternative (Gil 2013), as other former revolutionary parties in
Latin America did (Holland 2016).
With time, the electoral logic and democratic experience forced the
Left to reconsider their discourse and actions. This led to the develop-
ment of two incompatible positions within the IU (CVR 2003a, 185–
189). On the one hand, a group of radicals greatly distrusted demo-
cratic institutions, practiced a more confrontational political style, and
maintained revolution as their maxim. This position was represented
by most Maoist and “new Left” parties within the front.9 On the other
hand, organizations and leaders such as Alfonso Barrantes gradually

Soifer_6844-final.indb 210 8/17/18 11:53 AM

 Political Violence and the Defeat of the Left 211

developed a reformist position that favored a democratic regime and

political moderation.
The tension between these two positions grew as the SL and the
MRTA expanded their military presence in the territory and the re-
pression and violence increased. The insurrectional radicalism grew
as many militants, particularly the youngsters, demanded a more con-
sequential position from their leaders. In fact, several parties split and
formed new factions, some of which even joined the MRTA and SL
ranks (Pásara 1990; CVR 2003a, 193–197).
As the chances of defection to the subversive groups grew, some
parties tried to prepare military forces to fight from a “tercera vía”
when the time came (Rénique 1998; Rénique 2004). While Peruvian
society moved away from radical positions and strongly rejected the
SL, the IU was internally divided about how to act. This lack of pro-
grammatic unity eventually led to the IU’s division (Guerra García
2011; interview with Tapia in Adrianzén 2011, 497). In 1989, dur-
ing the infamous National Congress held in Huampaní, the moderates
abandoned the front.
Finally, ideologically, the legal Left also confronted an additional
challenge that they could not surmount: their programmatic proxim-
ity to Alan García’s reformist government (interview with Hinojosa
2015). As soon as this government began, tensions grew within the IU
regarding how to position themselves in relation to it, with moderates
who proposed to collaborate with the government opposing those who
favored a clear stand against APRA (Roberts 1996, 85; Tanaka 1998,
132–133). Moreover, in 1990, after the policies implemented by APRA
resulted in an unprecedented hyperinflationary crisis, the leftist parties
that competed for office could not express an alternative and credible
solution to the crisis. To this we must add the fall of the Berlin Wall in
1989, which was devastating programmatically for the Marxist Left.
As will be discussed, what remains of the Peruvian Left has been un-
able to provide a credible alternative to the neoliberal paradigm imple-
mented and solidified since the early 1990s.
In a nutshell, violence left the Left without its most powerful fan-
tasy: revolution. This was one of the most important direct effects of
political violence on the Left. As Caro contends, the Left ultimately
died because the vision that inspired them was exhausted. Because
the Left was not able develop an alternative discourse (interview with
Caro 2015), the legal Left ended up without anything new to say.

Soifer_6844-final.indb 211 8/17/18 11:53 AM

 212 Paula Muñoz

A Self-Fulfilling Prophecy?

The prediction made by Rolando Breña in the chapter’s opening ep-

igraph came true. As time passed, political violence led to increased
repression, diminished the Left’s options, and distanced it from the
electorate. After the IU’s split, the Left presented two candidates in
the 1990 election—Henry Pease for the IU and Alfonso Barrantes for
Acuerdo Socialista—who between them received only 13 percent of
the vote. Henry Pease, the IU’s candidate, ended up 8.2 percent be-
hind APRA’s candidate, who fi nished third in the race despite its disas-
trous record of governance. The prognosis was fulfi lled, to a great ex-
tent, because the legal Left could not solve its internal contradictions
in time to provide a united and unambiguous position about political
violence (CVR 2003a).
But is there something else that could explain the IU’s inability to
change in this context of extreme violence? Some authors have pointed
out to how the organizational legacies inherited from the Left’s pe-
riod of inception made it more challenging for the IU to adapt to the
democratic context and moderate its positions, even when it was po-
litically imperative to do so (Nogueira-Budny 2013; Gil 2013). As sev-
eral authors have stressed, the actions of the reformist military govern-
ment unintentionally marked the characteristics of the nascent Left as
radicalized and highly fragmented (Rochabrún Silva and Yañez 1988;
Stokes 1995; Sanborn 1991; Hinojosa 1998).
To begin with, the IU could never have a unified position during
the 1980s because it was not one Left (interview with Hinojosa 2015).
As explained previously, during the reformist military government,
a myriad of leftist parties flourished, with distinct organizations and
partisan identifications. In contrast to other Latin American cases in
which leftist parties were forced to adapt to adverse conditions dur-
ing their formative years (Nogueira-Budny 2013; Van Dyck 2016), in
Peru the Left developed during the 1970s within a more permissive
context and, thus, had little incentive or need to change. These or-
ganizational legacies made adaptation difficult, even when it became
evident in the 1980s that change was needed (Nogueira-Budny 2013;
Gil 2013). Consequently, until its division in 1989, “the IU remained
a loose electoral coalition of small parties that continued to compete
among themselves” (Roberts 1996, 83).
The IU was not created with the aim of becoming a political party
with long-term interests of institutionalization in democratic arena.

Soifer_6844-final.indb 212 8/17/18 11:53 AM

 Political Violence and the Defeat of the Left 213

It was formed as the coalition of five political parties and two fronts
that, as has already been stressed, were created not to compete elector-
ally but to pursue revolution. Thus, the IU was born specifically as an
electoral alliance, with the electoral space understood instrumentally
(Diez Canseco 2011, 102, 107). These parties allied to compete in the
municipal elections of 1980, after a very poor showing in that year’s
general elections when the Left entered the race divided into five candi-
dacies (CVR 2003a). Another common goal they shared was to coun-
terbalance the alliance on the Right that was in power, the coalition
between Acción Popular and the Partido Popular Cristiano (Diez Can-
seco 2011, 109).
The IU was not only an electoral front but also a heterogeneous one
that carried the organizational legacies of the Left’s origins. Given that
the parties were highly ideological, disputes abounded and provoked
factionalism. At the same time, partisan identities were very strong,
with markers easily recognized and including such things as the way
militants dressed, how they talked, and even their skin color (inter-
view with Mejía 2015). Moreover, each party maintained its popu-
lar base and fought to make its worldview and interests prevail; since
they usually antagonized each other, it was difficult to reach a consen-
sus when decisions were made (interview with Zapata 2015). In the
PCdelP-Pr leader Rolando Breña’s words,

There were 7 or 8 groups that sought to conquer political spaces for

future actions, but they were rivals not allies. We united because we
had to unite, to avoid being weak independently. But the zeal, the dis-
trust, the sectarianism, and the caudillismo of the big and the small
persisted.10

Second, and more important, the IU remained an electoral front,

and it did not build a common partisan structure in the years to come.
The IU persisted as an inchoate party without a professional structure
and could never develop an effective, centralized leadership (Nogueira-
Budny 2013, 141). They did not invest in party-building.
A Comité Directivo Nacional (CDN), composed by a representative
of each party and two representatives of each of the two fronts (the
Unidad Democrático Popular, or UDP, and the Unión de Izquierda Re-
volucionaria, or UNIR) that composed the alliance, oversaw the IU.
The CDN was a collegial body that elected a coordinator every two
months. This meant that, in practice, every two months the IU had a

Soifer_6844-final.indb 213 8/17/18 11:53 AM

 214 Paula Muñoz

different leader, and sometimes the leaders even held different politi-
cal positions (Tanaka 1998, 134). The only “independent,” nonparti-
san member who could participate in the CDN was Alfonso Barran-
tes, due to his position as coordinator of the IU (Diez Canseco 2011,
102). Moreover, the CDN governed the IU by consensus, not by ma-
jority, “And, since getting eight dogmatic, sectarian, leftist caudillos
to agree on anything is considerably difficult, little could be agreed
upon. [. . .] Consensus-based decision-making stifled much needed re-
forms and critical measures got derailed by ideological disagreements,
tactical differences, and even personal vendettas” (Nogueira-Budny
2013, 141).
The IU avoided incorporating a vast social sector of nonparti-
san sympathizers into its ranks because the parties managed to im-
pose and maintain the partisan quota mechanism in the CDN (Diez
Canseco 2011, 114). Indeed, the IU never created local committees of
its own as did leftist fronts in other countries, such as Unidad Popu-
lar in Chile (interview with Zapata 2015).11 Antonio Zapata contends
that the parties did not want the IU committees. Each party had its
own slogans, chants, and candidates, and they contested with each
other—particularly the dominant ones in the majority of districts such
as the Partido Unificado Mariateguista (PUM), Patria Roja (PR), and
PC-Unidad (interview with Zapata 2015). The meetings of the par-
ties (or “cells”) were very different from those of the IU: while the for-
mer were embedded in defi ning identity and developed in the midst of
important ideological discourses and discussions, the latter consisted
mostly of interparty fights for positions or to impose their ideas (in-
terview with Mejía 2015). As Barrantes once demanded, “[W]e have
to overcome this form of political feudalism where everyone wants to
have their own feudal castle, their coat of arms, and insignia. It does
not contribute to unity and the people do not want it; the people want
a big thing” (interview with Barrantes in Adrianzén 2011, 226).
In general, nondemocratically elected party leaders were averse to
building up democratic institutions for decision making because such
a process would necessarily come at the expense of their own parties
and privileges (Nogueira-Budny 2013, 138). Thus, when a moderniz-
ing sector within the IU pushed to provide a uniform system of iden-
tification (“carnetización”) and implement universal elections to elect
leaders (“one militant, one vote”), the IU was dissolved (interview with
Zapata 2015).
These tensions within the IU grew as elections revealed the front’s
contradictions. For one, the Left had more than enough candidates for

Soifer_6844-final.indb 214 8/17/18 11:53 AM

 Political Violence and the Defeat of the Left 215

the position but lacked undisputed leaders (Hinojosa 1998). When the
IU was formed, the parties within it accepted Barrantes as its coordi-
nator because he was an independent and an attractive electoral leader.
But they struggled all along to accept his leadership and questioned
him almost from the beginning. According to Zapata, Barrantes was
a very charismatic and electorally attractive figure, but he maintained
a tense and complicated relationship with the parties’ leadership (in-
terview with Zapata 2015). In Zapata’s view, Barrantes’s leadership
within the front hindered the IU’s consolidation as a party. However,
other leaders thought he was crucial in maintaining the front’s unity
for as long as it lasted (interview with Rolando Ames in Adrianzén
2011; interview with Henry Pease in Adrianzén 2011). Either way,
the questions around Barrantes’s leadership indicate a deeper problem
with the IU: its inability to reconcile the electoral and the partisan are-
nas, or of articulating them and investing capital accumulated in one
into the other (Tanaka 1998, 139).
Finally, it is important to mention that the IU’s low degree of in-
stitutionalization and professionalization did not allow the Left to
exploit (local) management to develop a brand and gain legitimacy
within a population that still questioned their ability to take on the
government (Gil 2013). As the former PUM militant Carlos Paredes
comments while accounting for the Left’s relatively poor electoral per-
formance in Cusco in 1980, “they saw us on the left as good defend-
ers for direct struggle but not as rulers” (interview with Paredes 2009).
As years passed, an ineffective, fragmented, and undisciplined or-
ganizational structure that made decisions by consensus became en-
trenched. The IU’s “adaptation was thwarted because undemocrati-
cally elected veto-players could derail the institutional and ideological
change required by external changes” (Nogueira-Budny 2013, 142).
The moderate wing, led by Barrantes, was unable to effect top-down
change and fi nally left the front. But soon it became clear that their
political calculus had proved wrong.
The divided Left participated in the 1990 presidential elections. As
mentioned previously, both leftist candidates were very far from ob-
taining the votes needed to attain even the third place.12 Thus, an im-
portant percentage of voters who supported the IU in 1985 were no
longer supporting the Left in 1990.13 But what is particularly surpris-
ing from this election is that the radicals, who retain the IU label, ob-
tained 229,851 more votes than the moderates (IS). In other words,
more votes were taken by Alberto Fujimori from the moderate Left
than from the more radical electorate. This means that Barrantes’s

Soifer_6844-final.indb 215 8/17/18 11:53 AM

 216 Paula Muñoz

electorate was substantially different: those who voted for him were
more pragmatic and moderate, and did not feel adequately represented
by existing social-movement-based parties (Tanaka 1998).14
In sum, the key weakness of the IU was based on its internal char-
acteristics: never able to come together as a party, “they could not
merge” (interview with Zapata 2015). The organizational legacies that
the IU inherited from the Left’s period of inception made it more dif-
ficult to respond to the external circumstances and challenges, includ-
ing the violence process. The IU was not able to adequately respond
to this looming threat during the 1980s (at least not in a timely way).
The SL’s strategic moves militarized the country and questioned the le-
gal Left. Pushed by the SL’s irrational insurgency, the IU was not able
to moderate itself, discipline its heterogeneous cadres, and unambigu-
ously break with its radical rhetoric to support democracy. Because of
this, the legal Left ended the decade divided and electorally defeated.

The Crossfire: Weakening and Dispersing the Left

A second, obvious way in which political violence impacted the legal

Left was by physically targeting it. The Left was attacked by both the
SL and state forces. Moreover, the escalation in violence that came
with militarization disrupted social and political organization in the
popular sector, the most affected by the confl ict (CVR 2003a). The
conjunction of violence and economic crisis further disorganized and
demobilized the Left’s social bases. In the midst of these crises, the re-
mains of the Left fi nally disintegrated (or were co-opted) during Fuji-
mori’s tenure.
From early in the conflict, a considerable number of IU militants
and local cadres were harassed, attacked, and/or killed by both the
SL and security forces. On the one hand, the security forces were sur-
prised by the emergence of the SL, which did not resemble at all the
foquista-style guerrillas they were expecting (CVR 2003a). Security
forces knew almost nothing about the SL and its peculiarities within
the leftist camp. Moreover, the National Security Doctrine in which
they were trained included leftist militants as part of the enemy profi le
they had to target (CVR 2003a, 130). Therefore, it is not surprising to
suspect that the armed forces targeted many legal Left militants and
sympathizers, and that these were subject to indiscriminate repression,
at least during the fi rst years of the confl ict. There is no precise way of
knowing how many of the victims of state repression were leftists.

Soifer_6844-final.indb 216 8/17/18 11:53 AM

 Political Violence and the Defeat of the Left 217

In addition, leftist militants were accused by the Acción Popular’s

government of being senderistas and consequently imprisoned and
tortured. The antiterrorist law (DL 046), enacted in March 1981, lay
the foundation for easily characterizing public protests as terrorist
acts. Consequently, the government imprisoned many leftist militants
who were not associated with the SL and union leaders accused of ter-
rorist charges (CVR 2003a). Thus, of the 123 prisoners held at Luri-
gancho prison on charges of terrorism, almost 36 percent were leaders
of social organizations (Comité de Presos Políticos de Izquierda Unida
e Independientes 1985).
The IU was hit directly and indirectly by subversive groups as well.
Throughout the decade, the SL managed to attract to its ranks other
radical leftist militants who were dissatisfied with democracy and the
IU’s inconsistency, or who imagined the SL victorious (Pásara 1990;
Hinojosa 1998, 91–92), particularly Maoists (interview with Zapata).
The following testimony of a former Puka Llacta militant illustrates
this point:

When SL began the armed struggle, in ’80, ’81, ’82, a division oc-
curred in the university and the magisterio. In Patria Roja, Puka
Llacta was the fi rst faction that split. Then a new faction, Renova-
ción, split and then other one, Viraje, and then another, the Bol-
chevique faction. . . . At least half of all these factions became mem-
bers of Sendero. And a part of the other half joined the MRTA. And
they began their actions in Cusco, which were basically propaganda,
fundamentally making paints and distributing flyers. (interview with
Mamani 2010)

In effect, the IU also suffered defections from militants who joined the
MRTA, which was socially and ideologically closer to other groups on
the Left such as the UDP, the PUM, and the Christians (interview with
Zapata 2015).
While the MRTA competed mostly indirectly with the IU, the SL
confronted it explicitly. The SL never approached or tried to ally with
other Left groups, but instead harassed and killed them. From the be-
ginning of the confl ict, the SL used violent and coercive methods to
gain control of social organizations. These actions included the “selec-
tive annihilation” of their leftist competitors, paralyzing the IU’s polit-
ical organizations (CVR 2003a). The violence against leftist militants
and leaders increased over time.
Trapped in the crossfi re, particularly in rural areas, the IU gradu-

Soifer_6844-final.indb 217 8/17/18 11:53 AM

 218 Paula Muñoz

ally reduced itself to an urban force (interview with Zapata 2015). Be-
tween 1989 and 1992 the crisis accelerated (CVR 2003a). The double
crisis (political and economic) “eroded the structural basis for class-
based collective action by creating a more heterogeneous and informal
work force; it diminished the centrality and strength of organized la-
bor, while fragmenting civil society” (Roberts 1996, 70). Terrorized
people no longer trusted anyone who was not well known to them
(such as family members) and stopped participating in unions and so-
cial organizations. Thus, for instance, after the union leader Pedro
Huilca’s death, no one attended the CGTP anymore (interview with
Mejía 2015). In Lima, a crucial event that fragmented the IU was the
savage assassination of the social leader María Elena Moyano in Feb-
ruary 1992: with that action, the SL succeeded in scattering the Left;
“after that the [IU’s] diaspora began” (interview with Zapata 2015).
After the assault on Moyano and the coup, the relationship between
the Left and the “popular movement” was irreconcilable.
In this context of profound socioeconomic and political crisis, Fu-
jimori built relationships with frightened popular sectors. He won the
political battle for the Left’s “popular base” (interview with Hino-
josa 2015). While skillfully attacking the partidocracia, Fujimori re-
cruited several local cadres of an already shaky legal Left to work for
his reelection. In some cases, the government blackmailed radical left-
ist cadres: it made them choose between being charged with terrorism
or working for the political state that Fujimori was building. In other
cases, the government blatantly co-opted political operators to build
popular support for its reelection efforts (Muñoz forthcoming).15 The
change was symbolically sealed when Martha Moyano, María Elena’s
sister, and the Women’s Federation, previously an IU organization,
joined Fujimori.16 It is clear that Fujimori occupied the electoral and
social space vacated by the Left.

Legacies and Their Manipulation:

The Stigmatization of the Left

Fujimori finished killing the corpse.

Au t hor’s i n t e rv i e w w i t h R ic a r do C a ro, 2015

For the Left, the main legacy of the political violence was that it be-
came associated in the political discourse with terrorism. This associ-

Soifer_6844-final.indb 218 8/17/18 11:53 AM

 Political Violence and the Defeat of the Left 219

ation began well before Fujimori entered the political stage. To begin
with, the SL, which was part of the radical Left, unleashed a national
trauma through its excessive violence, sweeping away popular support
for everything connected to leftist ideologies.17 As the Left came to be
associated with irrational terrorists, by 1988, 60 percent of respon-
dents to public opinion polls in Lima believed the SL had infi ltrated
the unions, particularly the Mining Federation and the CGTP. In time,
even the popular sectors stopped seeing the Left as its defender, turn-
ing instead to Fujimori.18
The SL dragged the Left down with it and provided others with ele-
ments to corner the legal Left.19 Indeed, from the beginning of the con-
flict, each succeeding government discursively equated the Left with
terrorism (interview with Hinojosa 2015). Even other congressmen
called the IU representatives “senderistas” (interview with Rolando
Breña). 20 The Right found it politically advantageous to criminalize
the Left and thus weaken or eliminate it as a political competitor. Of
course, the association between the Left and terrorism was even stron-
ger among police and military personnel, who saw the legal Left as al-
lies of the SL (CVR 2003a, 268). But even some of the leading news-
papers, such as El Comercio, Expreso, La Prensa, and Oiga, held the
Left responsible for terrorist attacks and other escalations in violence
(CVR 2003a, 492).
Certainly, the IU’s ambivalent position on the use of violence as a
legitimate political tactic made it easy to blame the Left as a whole
for the SL’s insanity. By December 1989, 46 percent of respondents to
public opinion polls considered it probable that sectors within the IU
would engage in subversion. In many ways, the IU remained trapped
in highly ideological and unintelligible political discussions that were
disconnected from the general public. By June 1988, polls in Lima in-
dicated that 79 percent of the city’s residents believed such subver-
sion was unjustifiable. Moreover, at least in Lima, public opinion did
not differentiate much between the SL and the MRTA, as leftists did,
and thus the two groups were regarded with nearly equal disfavor. A
strong association with terrorism was the Left’s main war-related leg-
acy. That this association remains pervasive, however, is also partly
the result of the political manipulation of the legacy carried on by Fu-
jimori during the 1990s.
The Left suffered from Fujimori’s attacks, as did parties from the
rest of the political spectrum (Tanaka 1998; Degregori 2000a). How-
ever, the Left had to confront an additional challenge: the stigma of

Soifer_6844-final.indb 219 8/17/18 11:53 AM

 220 Paula Muñoz

the military defeat of the SL and MRTA, which occurred within an in-
ternational context in which Communism elsewhere was crumbling.
Particularly after the coup, antisubversive policies became harsher and
delivered “results,” although at very high humanitarian costs such as
selective disappearances and killings, torture, failure to uphold due le-
gal process, and widespread imprisonment of innocent people (CVR
2003a). Violence had created a “culture of fear” in which “citizens
willingly surrendered rights in exchange for the promise of order and
stability” (Burt 2006, 35). The Fujimori regime manipulated this fear
to bolster support for its increasingly authoritarian rule (Burt 2006;
CVR 2003a, 120–134). Thus, for example, in his rhetoric, Fujimori
put trade unionists in the same category as the SL and the MRTA
(Burt 2006, 48). In this way, the government made sure that “oppo-
nents dared not voice their criticism publicly for fear of being labeled
a ‘terrorist’ and receiving the same treatment they receive—death, im-
prisonment, torture, silencing” (Burt 2006, 51).
Fujimori effectively capitalized on the antisubversive policies, par-
ticularly after the capture of the top SL and MRTA leaders beginning
in 1992. After many years of despair, by June 1994, public opinion
polls in Lima showed that 68 percent of respondents believed terrorism
would decrease, and 45 percent thought that the SL would be defeated
(APOYO, Poll June 1994). Between January 1993 and November
1995, polls of Limeños consistently showed around 70 to 80 percent
approval for the government’s antisubversive policies (APOYO, Poll
November 1995). In 1995 Fujimori was reelected with 64 percent of
the vote. Meanwhile, the IU lost its registration when it failed to cap-
ture 5 percent of the vote (Tanaka 1998, 53).
For the rest of the decade, the Left and other opposition groups
continued to be associated in the public political discourse with “ter-
rorism,” even after terrorist actions diminished (CVR 2003a). The re-
gime was successful in its efforts to reinforce the association between
social protest—a traditional tactic of the Left in Peru—with legacies
of “terrorism” and disorder (CVR 2003a, 122). It did so through con-
tinued deployment of psychosocial operatives directed by Vladimiro
Montesinos from the National Intelligence Service and by using an
increasing number of bought media organizations to spread its mes-
sages. Accusing members of government of being “terrorist” or “red”
was one of the tactics most frequently used (CVR 2003a,122; Fowks
2000). Thus during the 2000 presidential race the attacks through the
media against opposition candidates or activists linked with the Left

Soifer_6844-final.indb 220 8/17/18 11:53 AM

 Political Violence and the Defeat of the Left 221

included association with terrorism or accusations of being tolerant to-

ward terrorist extremism as a constant theme (Fowks 2000, 192). De-
spite these attacks, some former IU leaders, such as Diez Canseco and
Pease, played an important role as opposition congressmen, gaining
recognition within certain niches. However, as we will see, this was
not enough for them to succeed as political candidates in the new dem-
ocratic context. While the Left started the new decade electorally de-
feated, it ended it politically stigmatized.
The force and endurance of the Left’s stigmatization due to its asso-
ciation with terrorism can be observed in multiple occasions in the sub-
sequent post-transition period. But the political manipulation of this
legacy was certainly more blatant during the 2016 elections, in which
a younger leftist leader with more popular appeal, Verónika Mendoza,
came close to qualifying for a run-off. As soon as it became clear that
Mendoza might capture second place in the presidential election, ref-
erences linking her with the SL, the MRTA, and “terrorism” multi-
plied. Among other things, Mendoza was accused of being linked to a
social leader charged with terrorism, 21 with joining a meeting for the
fiftieth anniversary of VR (Vanguardia Revolucionaria) that discussed
former VR cadres joining the SL, 22 including former terrorists and vio-
lent radicals in her congressional list, 23 and even of being herself a ter-
rorist. On this last point, some months before the campaign started,
Carlos Tubino, a fujimorista congressman, also declared that if Men-
doza had been a politician in the 1980s, she would have been a ter-
rorist. 24 After the campaign was over and some weeks before the new
leftist parliamentary group led by Mendoza joined Congress, another
fujimorista congressman, Héctor Becerrill, showed a falsified photo
on Twitter as part of an accusation that Mendoza sympathized with
Sendero Luminoso.25 A child when the confl ict took place, Mendoza
was nonetheless unfairly linked to totalitarian extremism. Although
it is not possible to assess what influence this negative campaign had
over voting, we can certainly see the reproduction and political manip-
ulation of the internal confl ict and its legacy.

The Current Limited Political Relevance

of the Left: Final Thoughts

The SL’s strategic moves militarized the country and cast doubt on the
legal Left and its ideological positioning. Pushed by the SL’s extrem-

Soifer_6844-final.indb 221 8/17/18 11:53 AM

 222 Paula Muñoz

ist and uncompromising insurgency, the Left divided in two irreconcil-

able ideological positions. The IU was not able to moderate itself, or to
unambiguously break with its radical rhetoric to support democracy. It
was not able to adequately respond to the crises of the 1980s (at least
not in a timely way) partly due to its organizational characteristics.
The lack of a truly partisan structure with a centralized command
posed insurmountable practical challenges when the Left had to take a
position on the armed struggle, because it was too difficult for the dif-
ferent leftist parties to agree on anything. Because of this joint process,
the legal Left ended the decade divided and electorally defeated.
What would have happened to the IU if there had been no political
violence? Was it condemned to split anyway due to its organizational-
historical legacies? It is certainly difficult to answer this counterfactual
question. Even without armed violence, the IU may have contained the
seeds of its own destruction due to the organizational shortcomings
within the Left.
That said, Peru was unique in the region due to the extreme radi-
calism of the SL and the bloody character of the struggle it launched
against the state. Even if the IU had split in the 1990s for other reasons
than the armed conflict and its legacy, it seems more likely that a new
leftist party could have formed and that its chances of success might
have been greater. What in other countries was an abstract, theoreti-
cal discussion—armed struggle and the revolutionary road to power—
became all too concrete in Peru; in failing to respond effectively to the
crisis, the Left became divided and stigmatized. As seen in the 2016
elections, this legacy remains a powerful lever that is still manipulated
actively by the Left’s rivals.
Ultimately, the political defeat of most former leftist forces was
sealed in Peru when the radical version of neoliberal reforms imple-
mented by the government brought hyperinflation under control and
resumed economic growth, legitimizing the emerging right-wing and
mano dura order that now prevails. The impressive economic record
achieved by neoliberal governments in the context of the mineral
boom hindered the reemergence of the Left in the early 2000s. If the
main legacy of political violence for the Left has been stigmatization,
for the political Right it has been unorganized success. 26
Although more than a decade has passed since Fujimori’s regime
collapsed, it is clear that the main effects of political violence on the
Left have been, fi rst, its political defeat and, later, a legacy of obsta-
cles to overcome in advancing any new leftist alternative. Post-confl ict

Soifer_6844-final.indb 222 8/17/18 11:53 AM

 Political Violence and the Defeat of the Left 223

legacies also explain the long odds of constructing a party that is both
democratic and electorally viable. Although there are voters who are
dissatisfied with the current state of affairs and willing to support pro-
gressive and redistributive policies, it is difficult for leftist coalitions
to become a credible option for these citizens. And even when some
groups effectively confront this adverse situation, as happened with
the ticket led by Mendoza in 2016, they still face the stigma of past vi-
olence and show a surprising inability to behave more strategically to
increase their chances for electoral success and political survival.

Interviews

Caro, Ricardo. Sociologist, historian, and expert on the history of the Peru-
vian Left. Lima, February 23, 2015.
Hinojosa, Iván. Historian and expert on the history of the Peruvian Left.
Lima, March 26, 2015.
Mamani, Adolfo. Former Puka Llacta militant. Cusco, December 16, 2010.
Mejía, Carlos. PCP-Unidad militant. Lima, March 17, 2015.
Paredes, Carlos. Former member of PUM. Lima, November 7, 2009.
Zapata, Antonio. Historian and former member of PUM. Lima, March 3,
2015.

Notes

I want to thank Madai Urteaga for her diligent work as research assistant to
this project. I also thank Melina Galdós, who initially collaborated with me
on it, and Viviana Baraybar, who helped me fi nish revising it.

1. Breña is quoted in Equis (October 1980): 214; the quotation also ap-
pears in CVR 2003a, 179. For Hinojosa’s remark, see Hinojosa 1998, 78.
2. For many, after the guerrilla failure, Maoism seemed to be the only vi-
able approach to revolution. According to Navarro, “BR’s [Bandera Rojas’s]
central criticism of the MIR’s attempt at revolution was its lack of support
from the campesinos who inhabited the regions in which most of the fighting
took place. To the Maoists, this was a clear sign that the MIR lacked sufficient
understanding of Mariátegui’s teachings about the role of indigenous people
in improving Peruvian society, not to mention its total disregard for Mao’s
view of the role of the peasantry in the revolution” (Navarro 2010, 161).
3. APRA was the traditional rival of the Peruvian military.
4. The government even recognized the Central General de Trabajadores
del Perú (CGTP)—controlled by the Communist Party (PCP-Unidad) as the
official representative of labor (Hinojosa 1998, 83).
5. Some important political groups, such as VR–Político Militar, VR–

Soifer_6844-final.indb 223 8/17/18 11:53 AM

 224 Paula Muñoz

Proletario Comunista, PCP–Patria Roja, as well as the Shining Path, refused

to participate on that occasion (CVR 2003a, 173).
6. The IU was initially formed by the following political parties: Unidad
Democrático Popular (UDP), Unión de Izquierda Revolucionaria (UNIR),
Partido Comunista Peruano (PCP), Partido Socialista Revolucionario (PSR),
Partido Comunista Revolucionario (PCR), Frente Obrero Campesino Es-
tudiantil y Popular (FOCEP), and Partido Comunista del Perú–Patria Roja
(PCdelP-PR).
7. Javier Diez Canseco (1948–2013) was an important leader of the Par-
tido Unificado Mariateguista (PUM) during the 1980s.
8. Zapata stresses how the IU published a release condemning the SL’s ter-
rorist actions as early as 1982. The existence of this release in July 1982 was
effectively recorded by the CVR’s Political Chronology 1978–2000, which
was based on newspaper and archival research.
9. For an overview of the development of the radical Left in Peru, see Hi-
nojosa 1998; Rénique 2003; Navarro 2010.
10. Breña was the leader of the Patria Roja. “Rolando Breña Pantoja: ‘El
Partido Comunista también es conservador,’” interview in La República, Au-
gust 12, 2012, http://larepublica.pe/19–08–2012/rolando-brena-pantoja-el
-partido-comunista-tambien-es-conservador.
11. According to Zapata, the IU did not have local committees even in re-
nowned IU strongholds such as Villa El Salvador, where he worked as munic-
ipal director.
12. Together they obtained just 13 percent of the valid votes. See Fernando
Tuesta’s blog at http://blog.pucp.edu.pe/fernandotuesta/fi les/1990%20Elec
ciones%20Generales%201ra%20NAC%20PRES.pdf.
13. In 1985 Barrantes obtained 1,605,139 votes, which represented
24.7 percent of all valid votes. In 1990 Pease obtained 8.2 percent of the valid
votes and Barrantes 4.8 percent. See electoral results at Tuesta’s blog: http://
blog.pucp.edu.pe/fernandotuesta/node/919.
14. I thank Alberto Vergara for suggesting this interpretation.
15. Many of these former leftist militants were left on their own to survive
the economic crisis and the aftermath of the party system collapse. In talking
with them, one could perceive the deep estrangement they felt from their par-
ties. Many militants felt betrayed and mistreated by their limeño and middle-
class leftist caudillos. One common accusation was that while these leaders
took refuge in their NGOs during the crisis, the provincial “little brothers”
were left to fend for themselves (interviews with former leftist militants in
Piura and Cusco, held between 2010 and 2011).
16. Ibid.
17. Interview with Zapata. See also DESCO’s chronology of political vio-
lence (DESCO 1989).
18. Ibid.
19. Ibid.
20. See “Rolando Breña Pantoja: ‘El Partido Comunista también es conser-
vador,’ ” La República, August 12, 2012, http://larepublica.pe/19–08–2012
/rolando-brena-pantoja-el-partido-comunista-tambien-es-conservador.

Soifer_6844-final.indb 224 8/17/18 11:53 AM

 Political Violence and the Defeat of the Left 225

21. See “Verónika Mendoza rechaza estar vinculada con procesado por
terrorismo,” La República, April 2, 2016, at http://larepublica.pe/politica
/758033-veronika-mendoza-rechaza-estar-vinculada-con-procesado-por
-terrorismo.
22. See “Verónika Mendoza: ‘Me llaman terrorista por rechazar a Sen-
dero Luminoso,’ ” RPP, March 30, 2016, http://rpp.pe/politica/elecciones
/veronika-mendoza-me-llaman-terrorista-por-rechazar-a-sendero-luminoso
-noticia-949791.
23. See “¿Hay terroristas en la lista congresal de Verónika Mendoza? Res-
puesta al artículo difamatorio de Carlos García Tapia,” La Mula, April 9,
2016, https://estadocritico.lamula.pe/2016/04/09/hay-terroristas-en-la-lista
-congresal-de-veronika-mendoza/ricardomilla/.
24. See “Carlos Tubino: si Verónika Mendoza fuese una política en los 80s
habría sido terrorista,” La República, September 1, 2015, http://larepublica
.pe/politica/700327-carlos-tubino-si-veronika-mendoza-fuese-una-politica
-en-los-80s-habria-sido-terrorista.
25. See “Héctor Becerril ‘patina’ con foto trucada de Verónika Mendoza,”
El Comercio, July 19, 2016, http://elcomercio.pe/politica/congreso/hector
-becerril-patina-foto-trucada-veronika-mendoza-238143.
26. See chapter 9 in this volume.

Soifer_6844-final.indb 225 8/17/18 11:53 AM

 CHAP T ER 9

From a Partisan Right to the Conservative

Archipelago: Political Violence and
the Transformation of the Right-Wing
Spectrum in Contemporary Peru
Alberto Vergara and Daniel Encinas

We have two objectives in this chapter. The fi rst is to describe the mu-
tations of the Peruvian Right during the last three decades; that is, the
transformation of a Right organized around traditional political par-
ties but without ideological cohesiveness into a new Right that we call
the “conservative archipelago.” This new Peruvian Right did not de-
velop organic organizations, but it is strongly cohesive in ideological
terms. Its main agenda is to defend the neoliberal economic model that
emerged with the 1993 constitution. Despite party and electoral weak-
ness, the Peruvian Right has effectively defended its interests during
periods in which national circumstances could have propelled changes
in the country’s handling of the economy. This could have occurred,
for example, during the geopolitical period when several Latin Ameri-
can countries shifted to the Left, but the conservative archipelago has
been and still is the main political force in contemporary Peru. The
second objective is to assess the extent to which this transformation
and consolidation is related to Peru’s Internal Armed Conflict (IAC)
from 1980 to 1995.1 To meet these two objectives, we will proceed as
follows. After showing the transformation of the Peruvian Right over
the last three decades, we will establish a theoretical framework and
time frame to assess the potential impact of the IAC on this transfor-
mation. This is followed by an empirical assessment of the alleged re-
lationship. Finally, we conclude by framing the transformation of the
Peruvian Right in the Latin American context.

Soifer_6844-final.indb 226 8/17/18 11:53 AM

 From a Partisan Right to the Conservative Archipelago 227

From the Partisan Right to the Conservative Archipelago

When Fernando Belaunde’s second government tried to introduce a

few initial neoliberal reforms in the 1980s, it faced resistance from
not only “popular” forces, such as the left-wing spectrum and the Pe-
ruvian Alianza Popular Revolucionaria Americana (APRA), but also
groups within its own party, Acción Popular (AP), and its conservative
ally, the Partido Popular Cristiano (PPC). While it is frequently ob-
served that the Peruvian party system of the 1980s captured the left-
right ideological spectrum quite well (Tanaka 1998), rather less noted
is that the right-wing parties lacked ideological cohesion. The Right
had organizational prominence since the political parties were rooted
in society, but they attracted sectors with very different ideological
views. Both the AP and the PPC contained strong interests from the in-
dustrial sector, which distrusted the opening up of trade and blocked
any attempts at liberalization. Likewise, free trade was resisted, espe-
cially within the AP, by several sectors, and especially those linked to
Christian democracy, which dismissed any possibility that such eco-
nomic proposals could be positive.2
In the second half of the 1980s—when Mario Vargas Llosa entered
politics, aware of the new economic ideas that were thriving, politi-
cally and academically, in Europe and the United States—an initial
link was established between neoliberal ideas and the right-wing par-
ties. It became an occasion in which the AP, the PPC, and the new
Movimiento Libertad, led by Vargas Llosa, converged (Requena 2010).
Additionally, it was through Vargas Llosa that the ideas of the Peru-
vian economist Hernando De Soto came to shape the Peruvian version
of global neoliberalism (Adrianzén 2014). Thus, for this brief moment,
the old right-wing parties became connected to the new neoliberal
ideas to create a comprehensive liberal Right that was pro-free market
policies and politically liberal. Mario Vargas Llosa and his cohort in-
troduced new neoliberal economic ideas, but the traditional vehicles of
liberal representation—political parties—made the alliance eminently
democratic and institutionalist. While open markets were defended in
the economy, democratic institutions received political backing, even
when it came to the delicate antisubversive situation that moved Var-
gas Llosa to explicitly defend the notion of “civilizing” the Internal
Armed Conflict (M. Vargas Llosa 1993a, 177). Thus, for a brief time,
El Frente Democrático (FREDEMO, the alliance of right-wing parties

Soifer_6844-final.indb 227 8/17/18 11:53 AM

 228 Alberto Vergara and Daniel Encinas

and the new Movimiento Libertad) embodied a liberal and neoliberal

right-wing party project.
Still, this model of a liberal Right proved fleeting, since the elec-
tion of Alberto Fujimori in 1990 gave rise to a new type of Right: one
that was economically neoliberal but politically illiberal. The fi rst two
years of the 1990s were a critical juncture for the Peruvian Right. Fuji-
mori’s autogolpe in 1992, and the success of his administration in sta-
bilizing Peru’s economy and quelling insurrection, served to legitimize
the combination of neoliberal economic policies and the illiberal ap-
proach to politics and institutions. Not surprisingly, Vargas Llosa un-
leashed a fierce critique of it: “What a dictatorship can lead you to is,
provisionally, a type of growth that is more or less biological, statisti-
cal: rather than comprehensive development, which includes the field
of education, of culture, of the democratization of society and its val-
ues” (Hildebrandt 2008, 353). Yet the model Vargas Llosa criticized
was about to endure, in both spirit and actions. It entailed the devel-
opment of a Right based on personalistic politics, without parties,
but with great ideological cohesion regarding the neoliberal economic
model established by the 1993 constitution. Whereas in the 1980s in-
dustrial sectors linked to right-wing parties could boycott neoliberal
measures, in the 1990s and thereafter they would be denounced and
stigmatized as “mercantilistic” actors and agendas.
To illustrate the differences on economic matters between the con-
servative archipelago, which has prevailed in Peru for the past two de-
cades, and the old partisan Right of the 1980s, it is worth highlighting
two cases. In 2013, President Ollanta Humala proposed that the state
purchase some shares in the Spanish energy company Repsol, which
has considerable operations in Peru, in a weak attempt to change the
course of the country’s neoliberal economy. The right-wing opposition
was unanimous, with a striking level of cohesion and reach, in its re-
jection of this proposal. First, most of the print and television media
constantly attacked the initiative. Second, the most important factor
in dissuading the government from following the aforementioned path
came from within its own ranks. Humala’s minister of the economy,
an official with several years of service in the Ministry of the Economy
and Finance, was against the measure and convinced the president not
to pursue it (Uceda and Rivera 2013). In conclusion, the Peruvian Right
succeeded in neutralizing an antineoliberal endeavor with considerable
ideological cohesion despite being weak in organizational terms. This
Right, effective in defending the economic status quo through an ide-

Soifer_6844-final.indb 228 8/17/18 11:53 AM

 From a Partisan Right to the Conservative Archipelago 229

ological unity that belies its organizational fragmentation, is what we

call the “conservative archipelago.” Ideologically, the conservative ar-
chipelago advocates maintaining the economic order that emerged and
was institutionalized with the 1993 constitution. Organizationally, it
is composed of six types of actors (the “islands”): three of them are po-
litical and institutional, and the other three are social.
The fi rst group of actors consists of several traditional political par-
ties (especially the PPC and APRA) that no longer play a prominent
role as parties with societal roots. Instead, they have become what
we call partidos-bancadas3 whose existence and activity is limited to
a discredited legislature with little relevance in Peruvian politics and
policy (Valladares 2012). Second, fujimorismo, which also shares the
partido-bancada characteristics, has become an effective party brand
capable of competing in national elections and projecting a presence,
albeit a weak one, in the subnational arena. The third important group
of political and institutional actors to defend the conservative agenda is
the technocrats and bureaucrats. Having controlled the state for many
years, they have managed to institutionalize and defend neoliberal
practices without being challenged by the increasingly weak politicians
(Vergara and Encinas 2016).
In addition, there are three “social” islands in the conservative ar-
chipelago. First, there are the business associations, which have great
influence in the state and on public opinion, especially through the
Confederation of Private Businesses and Institutions (La Confeder-
ación Nacional de Instituciones Empresariales Privadas, or CONFIEP).
Second, over the last two decades the Catholic Church—especially
through the very public and vocal Cardinal Juan Luis Cipriani—has
played a decisive role in defending the status quo. Finally, most of the
Lima-based national media has exerted great influence since becoming
ardent supporters of the Peruvian economic model.
In sum, there is an archipelago consisting of six actors who, in an
ideologically cohesive but organically fragmented manner, uphold the
economic system that emerged in early 1993. As seen in the example
of the state’s failed attempt to purchase shares in Repsol (as well as
in several other more everyday dynamics), the conservative archipel-
ago is sufficiently effective to defend and institutionalize the prevailing
economic order in contemporary Peru. Thus, a neoliberal consensus
that is not politically liberal is what unifies this Peruvian Right. The
origin of the conservative archipelago—the authoritarian government
of Alberto Fujimori—was therefore a critical juncture that ushered in

Soifer_6844-final.indb 229 8/17/18 11:53 AM

 230 Alberto Vergara and Daniel Encinas

a cohesive neoliberal Right, while the defeat of the FREDEMO Proj-

ect in 1990 closed off the path for a comprehensive liberal Right. And
the path that the juncture opened is directly linked to the IAC of the
1980s. But how? That is the subject of the next section.

The Internal Conflict and the Transformation

of the Peruvian Right

In a country with a tradition of institutional weakness where the rules

change rapidly and frequently (Levitsky and Murillo 2009), the emer-
gence, survival, and entrenchment of the neoliberal economic model,
as well as the overwhelming consensus around it, is a peculiar occur-
rence in Peruvian history. Since the early twentieth century, the im-
plementation of national projects was constantly interrupted and/or
thwarted. Thus, the neoliberal project, constitutionalized in 1993, has
acquired remarkable stability.
Before continuing with the analysis of the transformation men-
tioned in the previous section, two things should be made clear. First,
the neoliberal economic model was highly successful in transforming
Peru, and the plan largely fulfilled its promises (Vergara 2013). This
success explains why politicians and large segments of the population
have united to maintain it. But (and secondly), the implementation of
the neoliberal model came after not only the economic disaster of the
1980s but also the IAC of that same decade. These crises were strongly
entangled together. Therefore, to what extent did the violence of the
1980s influence the transformation of the Peruvian Right? As with
all the outcomes analyzed in this book, it is difficult to isolate the ef-
fects that arose out of the economic crisis from those that arose out of
the IAC.
To analyze the links between the IAC and the transformation of the
Peruvian Right, this study is organized around the framework sug-
gested in the book’s introduction. First, we look at the open IAC pe-
riod, between 1980 and 1990. Second, we focus on the resolution of
the IAC between 1990 and 1995. Finally, we look at the post-conflict
period from 1995 to the present. In each stage, we show the mecha-
nisms that allowed the transformation of the Peruvian Right. In par-
ticular, we emphasize the relationship between the IAC and the de-
struction of the traditional partisan Right, along with the gradual
construction of the conservative archipelago.

Soifer_6844-final.indb 230 8/17/18 11:53 AM

 From a Partisan Right to the Conservative Archipelago 231

The IAC and Partisan Destruction (1980–1990)

Political parties have two types of assets: material assets and ideolog-
ical assets (Hale 2006). This categorization allows us to observe the
fi rst part of the transformation of the Peruvian Right. During the IAC,
the main assets of the Peruvian right-wing parties were severely dam-
aged, which laid the foundations for the formation of the “conserva-
tive archipelago.” We begin by showing the damage to material assets,
followed by the damage to ideological assets. This fi rst stage is signifi-
cant in the destruction of the traditional Right, but not for the forma-
tion of the new Right.
As explained in detail in the opening chapters of this book, the vi-
olence experienced during the 1980s was brutal. The confl ict that
started with the Shining Path’s (SP) declaration of war against the Pe-
ruvian state caused more destruction than any other Latin American
armed movement (Degregori 2010). The SP’s offensive plan involved
“the murder of local authorities: mayors, governors, lieutenant gov-
ernors, and justices of the peace, and national authorities: ministers,
parliamentarians and other representatives of the state” (CVR 2004).
The goal of the SP was to “create a power vacuum so that they could
establish control over the population more easily” (CVR 2003a, 6:16)
Clearly, this had consequences for the entire country, including its
multiple sectors and actors. Chapters 7 and 8 in this volume show the
considerable consequences of the IAC for the Left and for civil society.
But what was the cost for the right-wing parties?
Let us begin with the erosion of material assets. First, according to
the Final Report of the Truth and Reconciliation Committee (Comi-
sión de la Verdad y Reconciliación, or CVR), 12 percent of those killed
or missing during the IAC caused by the SP4 were authorities.5 When
local leaders are included in the category, the CVR shows that author-
ities represented 17 percent of the total dead or missing6 and 21 per-
cent of the SP’s victims.7 This means that after peasants, authorities
were the group most affected by the violence. Most of these authorities
“were members of political parties that supported the democratic re-
gime inaugurated in 1980,” so the fatalities represented “a heavy blow
to the capacity for political mediation in areas affected by the inter-
nal armed conflict” (CVR 2003a, 1:169). The AP, the PPC, and APRA
bore the brunt of the attacks, but it is even more difficult to quantify
the SP’s direct threats and damage caused in areas under military con-
trol. For example, according to a former APRA secretary, around a

Soifer_6844-final.indb 231 8/17/18 11:53 AM

 232 Alberto Vergara and Daniel Encinas

Table 9.1. Main targets of terrorist activity during the 1980s

May 1980 to July 1985 to

Target July 1985 June 1988 Total

Political party offices 104 202 306

Police stations 412 464 876
High-voltage towers 390 349 739
Railways 23 50 73
Bridges 69 96 165
Banks 135 329 464
Special projects 7 42 49
CORDES 5 59 64
SAIS 8 38 46
Municipalities 96 114 210
Electric power plants 19 48 67
Homes 1 411 412
Registers of electors 17 2 19
Settlements a - 213 213
Total 1,286 2,417 3,703

Source: Ministry of Defense; compiled by the authors based on DESCO 1989.

a
“Centro poblado” in Spanish.

thousand party members were assassinated (Vergara 2015a, 294). In

the case of other human rights violations such as torture, state authori-
ties and local leaders were also the second-most targeted group.
Attacks on party venues and offices also precipitated the erosion of
the parties’ material assets, while increasing the fear of party mem-
bers. Among the many examples of the SP’s targeting of municipalities
and party offices cited in the Final Report of the CVR is an account of
the attack on the AP’s main offices in Lima on July 11, 1983. But be-
sides this crucial incident experienced by a right-wing party, the Min-
istry of Defense’s database shows that party offices and municipalities
were among the main targets of the Shining Path during the 1980s.
It is quite difficult to accurately measure the damage caused exclu-
sively by the Shining Path to right-wing parties because the informa-
tion from the CVR does not separate the parties ideologically. How-
ever, a more qualitative approach such as the DESCO (1989) database8
shows that the right-wing parties were indeed victims of attacks on
party offices, party members’ homes,9 and other types of private prop-
erties as well as victims of assassinations and assassination attempts.

Soifer_6844-final.indb 232 8/17/18 11:53 AM

 From a Partisan Right to the Conservative Archipelago 233

Table 9.2. Attacks on right-wing political partiesa

Number of
representa-
tive cases,
1980–1988

Attacks on party offices and property 22

Attacks on homes and private property 8
of authorities, leaders, and members
Assassinations and assassination 19
attempts against party leaders
Assassinations and assassination 10
attempts against authorities
Assassinations and assassination 9
attempts against party members
Total 68

Source: DESCO 1989; compiled by the author.

a
Classifications made by the authors. Party property includes vehi-
cles; private property includes local businesses; party leaders in-
cludes previous authorities; and party members include candidates.

In short, the material assets of political parties were weakened se-

verely during the course of the IAC, in which party offices, leaders,
and members were constantly targeted by the SP. Thus, the violence
weakened the standing of these political parties so that they struggled
to compete electorally, especially in local and rural areas. Addition-
ally, the SP “was able to establish its presence and operate in large re-
gions of the country, particularly where the state was absent or largely
ineffective, thus building alternative structures of authority [. . .] that
sometimes challenged and sometimes replaced the state” (Burt 2004,
249). Furthermore, toward the end of the 1980s, 32 percent of Peru-
vian territory and 49 percent of the population was under military
control (Degregori 2010). According to Javier Diaz Orihuela, an AP
leader, “terrorism had a strong impact on our ability to get elected.”10
In addition to damaging the material assets of political parties, the
IAC also affected their legitimacy. The deterioration of the country’s
political and economic climate during the 1980s gradually led to a
general perception that these parties, through their elected officials,
were incapable of governing (Tanaka 1998, 54). This was particularly

Soifer_6844-final.indb 233 8/17/18 11:53 AM

 234 Alberto Vergara and Daniel Encinas

evident during the APRA administration from 1985 to 1990 when the
government was unable to solve the crisis, and in fact aggravated it
in every possible way. Beginning in 1988, the economic crisis esca-
lated, and the country’s collapse created desperate conditions in which
opportunities abounded for SL to recruit new fighters and supporters
(M. Smith 1992). As the institutional weakness of the state worsened
and inflation soared, the myth of “Sendero ganador”11 grew stronger
(Chávez 2012). Thus, APRA’s failure was instrumental in promoting
widespread rejection of political parties and further delegitimizing an
entire institutional structure unable to fulfill even the most basic du-
ties of the constitutional state.
The political parties’ difficulties led to poor electoral outcomes
at the national and subnational levels. For example, the AP’s perfor-
mance at the municipal level illustrates its growing disconnect with
the public: in 1980, it won 100 provinces (67.1 percent) and 814 dis-
tricts (57.1 percent); in 1983, this number dropped to 36 provinces
(23.2 percent) and 465 districts (32.1 percent); and in 1986, there were
no AP candidates running in any part of the country. All of this fore-
shadowed its disastrous results of the 1985 general election (G. Ruiz
et al. 2013). According to Kenney (2004), in the 1986 municipal elec-
tions, the four traditional parties together lost only 8 percent of the
national vote. However, in the following elections of 1989, this num-
ber increased to 29 percent. During that vote, the last of the 1980s,
the country’s capital experienced a shock when an outsider, the broad-
caster Ricardo Belmont, won the coveted Lima mayoral race against
the FREDEMO candidate, the AP’s Juan Incháustegui. Thus the par-
tisan liberal-right was defeated by a new, personalistic politics based
on nonorganizational ties with the electorate. It would not be the last
time this would occur.
In 1990, Alberto Fujimori used social unrest as a weapon against
the political parties that had pushed the country toward economic fail-
ure. At the same time, the unsuccessful fight against the SP led vot-
ers to seek stronger government (Crabtree 2010, 364). Thus, the re-
newed economic ideas that Vargas Llosa introduced to the Peruvian
Right were not enough to restore the strength of the parties, which
slowly crumbled throughout the 1980s as voters looked on. The IAC
was conducive to the formation of the contemporary conservative ar-
chipelago, since the confl ict eroded the material and ideological foun-
dations of the previous representative order. This entailed the destruc-
tion of party infrastructure, intimidation and assassination of party

Soifer_6844-final.indb 234 8/17/18 11:53 AM

 From a Partisan Right to the Conservative Archipelago 235

leaders and competitive candidates, replacement of democratic rules

in certain areas of the country, and punishment by voters of deficient
governmental handling of the crisis (and, often, aggravation of that
crisis). Ultimately, this total collapse prevented the economically and
politically liberal Right, represented by Vargas Llosa and backed by
parties with democratic credentials, from obtaining power.

Conflict Resolution and the Emergence

of the Archipelago (1990–1995)

The resolution of the IAC, a crucial moment for the formation of the
conservative archipelago, occurred in parallel with the establishment
of Alberto Fujimori’s government, the April 1992 self-coup, and the
approval of the 1993 constitution. The weakened parties were defeated
during this period, while the political and institutional foundations
were laid for the new conservative ideology and its islands of support.
The resolution of the IAC came at the hands of a neopopulist and
neoliberal outsider, and led to the collapse of the right-wing parties.
Here, an important distinction must be made between the Peruvian
Right and the Left vis-à-vis the armed confl ict. The Left was greatly
affected by the unfolding of the confl ict in the 1980s. However, by
1989, before Fujimori appeared on the national political scene, the Pe-
ruvian Left was already weakening “autonomously.” Initially, the Left
was divided by an insurmountable rift between those who denounced
the SL and those who sought to appease it (M. Smith 1992; Rénique
2004). Once it had been divided in two, the Left won only 7 of the
42 district municipalities of Lima in the 1989 elections—after having
held 21 of them in the middle of the decade. Then, in the 1990 gen-
eral elections, the Left managed to place only 9 senators and 20 dep-
uties in power (out of a total of 60 and 180, respectively), well below
the numbers of the AP, APRA, PPC, and Cambio 90 (Alberto Fujimo-
ri’s fi rst party brand). Thus, the Left’s breakdown was already under-
way well before Fujimori’s authoritarianism took root. Indeed, the res-
olution of the IAC by Fujimori did not do much to weaken the partisan
or electoral Left, but it did weaken the societal Left. This allowed the
SL to easily absorb the Left’s weakened foundations on the peripheries
of Lima and Peru, instilling a sense of fear in unions and other sectors
of organized civil society (Burt 1997).
The Right experienced the opposite. Until 1992, right-wing parties

Soifer_6844-final.indb 235 8/17/18 11:53 AM

 236 Alberto Vergara and Daniel Encinas

were electorally successful, having had 20 senators and 62 deputies in

power in addition to holding 82 provincial municipalities (nearly twice
the number of the left-wing platform Izquierda Unida holdings). But
Fujimori’s consolidation of power destroyed the right-wing parties.
They went from being the most important coalition within national
and subnational representative institutions to overall collapse follow-
ing the autogolpe of 1992. Still, unlike the Left, the societal Right was
strengthened during this period. Thus, the resolution of the IAC and
the emergence of Fujimori’s authoritarianism led to inverse outcomes
for the Right and Left: on the one hand, Fujimori did not eliminate the
partisan Left (it had already started to break down by itself) but did
erode the societal Left; on the other hand, he destroyed the partisan
Right but revived the societal Right. With that, the brief Vargasllosian
liberal-rightist experiment came to an end, and the formation of the
conservative archipelago began.
The resolution of the IAC explains the difference in outcomes. Be-
tween July 1992 and July 1993, the SL’s leaders were killed or cap-
tured. Abimael Guzmán was caught in September 1992 and petitioned
the government to grant his surrender a year later. The conflict, which
had seemed destined to endure, ended more quickly than anyone had
foreseen. And once the myth of “Sendero ganador” had been shat-
tered, popularity soared for the figure responsible for the demise of
the SP and his antirepublican form of government. This had direct and
indirect consequences for the Right, ultimately hastening the dissipa-
tion of the partisan Right and paving the way for the new conserva-
tive archipelago.
First, the defeat of the SL (along with the stabilization of inflation
and Peru’s reintegration into the international fi nancial community)
allowed Fujimori to style himself as the personification of an effec-
tive Right. Two decades earlier, General Juan Velasco had accom-
plished something similar with the reformist parties. Velasco imple-
mented the land reform by authoritarian means and borrowed other
measures from the agenda of the reformist parties (APRA, AP, De-
mocracia Cristiana), which had never implemented them, thus strip-
ping them of their reformist legitimacy and eroding their ideological
bases. Similarly, Fujimori successfully adopted the right-wing agenda,
stressing security and economic order. Like Velasco, Fujimori added
an anti-party rhetoric to his authoritarian efficacy, which was instru-
mental in the collapse of the right-wing parties.12 According to the AP
leader Víctor Andrés García Belaunde, who was a congressman un-

Soifer_6844-final.indb 236 8/17/18 11:53 AM

 From a Partisan Right to the Conservative Archipelago 237

til the 1992 autogolpe, “when Fujimori captured Guzmán, it was all
over, we could no longer compete.”13
At the same time as the disintegration of the partisan Right, the
seeds of the conservative archipelago, the new Right, were planted.
First, technocrats emerged in areas connected to the economy, but
without any links to political parties. As Dargent (2015) showed, the
degree of isolation that Peruvian technocrats achieved between 1992
and 1993 was linked to the authoritarian context. This isolation was
vital for the further development of a new cohort of technocrats and
bureaucrats in the Peruvian state. These technical leaders quickly re-
alized that the right-wing parties had been defeated, and that it was
no longer necessary to deal with them in their attempts to implement
market reforms. It made little sense for them to become members of a
dilapidated right-wing party when they could influence the state with-
out that failed intermediary.
Upon closer examination, it can be seen that the business sec-
tor expanded its presence in Peruvian politics during these develop-
ments. Several business leaders entered politics in a shift that would
be sustained over time. Influence administered through the CONFIEP,
given its proximity to the Fujimori government, became more impor-
tant than personal influence (Arce 2005). In 1994, with CONFIEP
funding, the Peruvian Institute of Economics (Instituto Peruano de
Economía, or IPE) was created as a think tank to supply the govern-
ment with recommendations and policies. This cementing of close ties
between the business sector and political power left an important leg-
acy for the future. Furthermore, the influence of these two factors on
the public sphere only increased during the 1990s and into the new
century.
In turn, the IAC played a major role in fostering a more conserva-
tive and politically active Catholic Church. The church had been mod-
erately critical of the government performance on human rights until
the early 1990s, but this began to change in 1992. An important re-
lationship emerged between Fujimori and Archbishop Monsignor Ci-
priani of Ayacucho. Cipriani was the president’s man in Ayacucho, the
region most affected by the violence. His work went far beyond his re-
ligious duties, and he played a key political role during the IAC. In
fact, Fujimori appointed him as the president of the Social Develop-
ment and Compensation Fund (Fondo Nacional de Compensación y
Desarrollo Social, or FONCODES), the fujimorista social policy (and
clientelistic) entity in Ayacucho. Cipriani held this position for most of

Soifer_6844-final.indb 237 8/17/18 11:53 AM

 238 Alberto Vergara and Daniel Encinas

the decade. When the regional governments dissolved after the 1992
coup and the Transitory Councils of Regional Administration (Con-
sejos Transitorios de Administración Regional, or CTAR) were set up
in their place, each council president was appointed on Cipriani’s rec-
ommendation: “Monsignor Cipriani was Fujimori’s personal represen-
tative in Ayacucho” (Pásara 2014, 16). In short, the IAC era forged a
Church with fujimorista tendencies and experience defending conser-
vative political positions in the public sphere.
Finally, the IAC and the undemocratic manner of its resolution cre-
ated favorable conditions for the endurance of the new economic re-
gime that accompanied the 1993 constitution. The remarkable con-
tinuity and stability of the Peruvian economic model would seem to
support the idea that institutional arrangements tend to endure when
their supporters completely defeat the opponents of the new institu-
tions (Przeworski 1991). In fact, if the stability of the neoliberal insti-
tutional arrangements were compared with the mechanisms and ac-
tors responsible for the frustration of similar reforms in the 1980s,
one would note the substantial differences that occurred in Peru over
a short period of time. One would observe that the right-wing par-
ties disappeared; the organizations and politicians of the protection-
ist Right were swept away; a body of neoliberal technocrats embedded
themselves in a state which constrained itself; the partisan Left com-
mitted suicide; the unions were demobilized; the SP requested a peace
deal from Fujimori’s government; and lastly, the fujimorista brand
made its entry into Peruvian politics. These, and especially the disso-
lution of the liberal Right and the birth of a conservative archipelago,
are the lasting political legacies of the resolution of the IAC.

The Post-Conflict Period: Legacies and

the Struggle over Legacies (1995–2015)

In a text published in 2000, Catherine Conaghan accurately foresaw

the future of the post-Fujimori Peruvian Right. She predicted that with
the foundations for a right-wing style of government established, the
fall of Fujimori would not result in the restoration of the old parti-
san Right in Peru. Indeed, after Fujimori fell, the conservative archi-
pelago became autonomous from the government that had created it
and shaped a new, nonpartisan way of defending the 1993 economic
model. In this section, we examine the consolidation of the conserva-

Soifer_6844-final.indb 238 8/17/18 11:53 AM

 From a Partisan Right to the Conservative Archipelago 239

tive archipelago and how it relates to the IAC. Then, we assess the dis-
pute of political discursion surrounding the IAC. In other words, we
look at the struggles over legacies, or the way actors used and shaped
the IAC based on their political preferences.
When Alberto Fujimori was reelected in 1995 with more than
60 percent of the vote, the international community accepted that the
government had been democratically purified of the original authori-
tarian sin of 1992. After then, the conservative archipelago was con-
solidated through a community of ideologically cohesive actors who
defended the new economic model that came with the 1993 constitu-
tion, but who had no organizational umbrella or party.
During the second half of Fujimori’s government, the weakness of
the political parties and the relationship between the executive and
the legislature, which survived the fall of the Fujimori regime, were
consolidated. In other words, “rampant presidentialism” (Morón and
Sanborn 2007) and the subordinate role of the legislative branch (De-
gregori and Meléndez 2007) were established. After the collapse of
Fujimori’s government, this subordination continued not because of
the actions of a competitive authoritarian regime, but because of the
weakness of political parties, which, without societal foundations, be-
came mere “partidos-bancadas” that support or oppose the wishes of
the executive without major consequences (Valladares 2012). Thus,
the right-wing parties, weakened by the IAC and its resolution as they
are, became the unwavering legacy of the post-Fujimori Peruvian de-
mocracy, and they now play a limited role in defending the conserva-
tive agenda from within the legislature.
Within the general legacy of party weakness, the political co-
alition that represents fujimorismo is a unique case. As recalled by
Roberts (2006), fujimorismo was not interested in forming a parti-
san vehicle, and this disinterest carried a high cost when Fujimori fell
from grace. However, a decade later—in large part because Fujimo-
ri’s daughter, Keiko, was old enough to be a presidential candidate—
fujimorismo took more of an interest in building a party organization
(Urrutia 2011; Navarro 2010). Rather than an institutionalized party,
the movement resembles a weak organization supported by a strong
memory and a strong and enduring leader in Keiko Fujimori. Urrutia
(2011) calls fujimorismo no more than a “partisan vehicle.” Follow-
ing Lupu (2014), we might suggest that fujimorismo is limited to be-
ing a party brand that enjoys popular support when a member of the
Fujimori family runs for office. Keiko Fujimori has reached the sec-

Soifer_6844-final.indb 239 8/17/18 11:53 AM

 240 Alberto Vergara and Daniel Encinas

ond round of two presidential elections, and she and her brother Kenji
have been the Congress members with the most votes in the 2006 and
2011 elections, respectively. However, whenever someone outside the
Fujimori family has led the movement, the number of votes won has
decreased significantly.14 As Max Weber (1995) pointed out in his ex-
planation of charisma, only a relative of the providential individual
or leader can inherit that leader’s charisma. In this case, the inherited
charisma of the chosen leader stemmed from the resolution of the IAC
by Alberto Fujimori’s government.15 It is defeating terrorism, rather
than ending hyperinflation or stabilizing the country’s economy, that
is fujimorismo’s biggest electoral asset, as well as the main rallying
point for fujimorista members and supporters (Urrutia 2011; Me-
léndez 2014). A 2009 nationwide survey by Ipsos-Apoyo found that
66 percent of respondents considered the “defeat of terrorism” to be
the greatest success of Alberto Fujimori’s government, far ahead of the
33 percent who cited “economic stability.” In addition, a nationwide
survey conducted after the fi rst round of the 2011 presidential elec-
tions revealed that 43 percent of those who voted for Keiko Fujimori
did so because of “the good government of her father, Alberto Fuji-
mori.” Ultimately, Meléndez (2014) refers to these elements of recogni-
tion as the birth of political identification or “partisanship.”
The enemies of fujimorismo—the “other” that helped give it an
identity during the 2000s—are organizations and individuals linked
to human rights (Urrutia 2011). In other words, this “other” was com-
posed of actors who would have objected to the resolution of the IAC
through the effective “fi rm hand” approach taken by fujimorismo.
Still, neither the institutionalization of fujimorismo nor its material
assets are the most important aspects of the movement; they are out-
weighed by the cohesive force provided by the memory of Alberto Fuji-
mori’s government, Keiko Fujimori’s repeated strong showings in pres-
idential elections, and the electorate’s conception of fujimorismo as a
propagated brand of order through a fi rm approach. In this way, fu-
jimorismo is a dual legacy of the IAC and its resolution. This means
that, like all Peruvian political parties which must depend on a civil
society weakened by the IAC, fujimorismo is organizationally weak.
However, the main appeal of fujimorismo, which unifies members and
supporters alike, is as the destroyer of the SL. In the national elections
of 2011 and 2016, these assets were linked to the conditions of the
early 1990s and recalled through the key issues of security and eco-
nomic slowdown. Thus, the construction of this partisan brand, with

Soifer_6844-final.indb 240 8/17/18 11:53 AM

 From a Partisan Right to the Conservative Archipelago 241

its ability to compete in national elections and play a role alongside

other parties in the legislature, became a central island of the Peruvian
conservative archipelago.16
Furthermore, the technocratic nucleus, which emerged and gained
independence after the 1992 coup and the implementation of the neo-
liberal reforms, endured into the 2000s. Like many legacies, the con-
solidation of the technocratic core was based on not only the inertia of
the 1990s but also decisions made during the post-Fujimori period. It
is important to note that Valentín Paniagua’s transitional government
retained many of the technical plans of the Ministry of the Economy
and Finance and of other entities linked to the economic sphere, insti-
tutions which were dominated by the experts who delivered the neo-
liberal reforms years earlier (Dargent 2015). During the government of
Alejandro Toledo (2001–2006), who had no party-political backing,
the technocratic sector gained power and became a powerful actor, es-
pecially against weak politicians without the leverage needed to influ-
ence other strong actors in the Peruvian state (Vergara and Encinas
2016). While significant state technocratization also occurred in other
countries, such as Chile and Mexico, in those countries the presence of
institutionalized political parties forced technocrats to consult or ne-
gotiate with the parties to advance their proposals (P. Silva 2009; Cen-
teno 1997). However, in Peru, the legacy of technocratic strength and
autonomy combined with the legacy of party weakness, which allowed
the technocrats to become the main actors in the conservative archi-
pelago and in the defense of the 1993 economic model.
Something similar happened in the Catholic Church. If the relation-
ship between Fujimori and the archbishop of Ayacucho was reinforced
by the IAC, then the resulting legacy deepened during the post- conflict
period, transforming the Church into a central actor in the conser-
vative archipelago. In 1995, Archbishop Cipriani openly defended
the laws that absolved military and police officers involved in human
rights violations. Then, when the Túpac Amaru Revolutionary Move-
ment (Movimiento Revolucionario Túpac Amaru, or MRTA) stormed
the residence of the Japanese ambassador in 1997 and took hundreds
hostage, the Peruvian government appointed Cipriani to negotiate
with the subversive group.17 Shortly after, Cipriani became the arch-
bishop of Lima and thus the head of the country’s Catholic Church.
Following his formal appointment, Cipriani developed ties to politi-
cal power, both at a personal level and as part of the influence enjoyed
by many politicians with links to the Opus Dei (Cipriani’s denomina-

Soifer_6844-final.indb 241 8/17/18 11:53 AM

 242 Alberto Vergara and Daniel Encinas

tion), especially during the government of Fujimori and Alan García.

Additionally, the cardinal also took on a role in the media, thus add-
ing to his mechanisms of influence. Since 2000, Cipriani has broadcast
a two-hour radio show every Saturday morning on Radio Programas
del Perú (RPP), the country’s most influential radio station. On his
show, under the guise of religious preaching, Cipriani often couches
unmistakably conservative views in his commentary regarding issues
on the national public agenda, whereby the defense of the 1993 eco-
nomic model is essential. For example, through his show, he joined the
petition to pardon former President Alberto Fujimori and regularly ad-
vocates a Church whose role “is not to change society, but bring order
to it” (Pásara 2014, 94). This influence in the public sphere amounts
to a sizable capacity for political mobilization within Peruvian soci-
ety, which is significant for a country that otherwise has shortcomings
in this respect.18 In short, the activity of the church under the leader-
ship of the cardinal, and his role in the conservative archipelago, are
inseparable from the IAC. While it is a legacy that was strengthened
and consolidated in the post-confl ict era, its origin is ultimately tied to
the IAC.
Finally, the business associations, whose influence on the state took
hold during the period of the authoritarian resolution of the IAC,
maintained and strengthened their links to power and their influence
on the public sphere. When it came to the creation of public policies,
the Peruvian Institute of Economics (IPE) gained importance through
its proposals and leaders. No right-wing party had more influence than
this think tank. The IPE worked through an effective “revolving door”
mechanism between government institutions and the business sector.
For example, Roberto Abusada and Jorge Baca moved from the IPE to
the government, while Leoni Roca and Fritz Du Bois went in the op-
posite direction.19 Overall, the influence of the business sector during
the 1990s was such that according to an expert, “it became virtually
impossible to discern which sets of policy proposals were drawn from
the business sector as opposed to the government itself” (Arce 2005,
136). In the 2000s, the direct influence over public policies diminished
but did not disappear. Several businessmen held important positions in
the executive on a constant basis. Of the 262 ministerial incumbents
from the beginning of the democratic transition to 2013, the percent-
age of ministers who were also businesspersons was 12 percent dur-
ing Toledo’s government, 14 percent during García’s government, and
19 percent during the Humala government. 20 During the fi rst year of

Soifer_6844-final.indb 242 8/17/18 11:53 AM

 From a Partisan Right to the Conservative Archipelago 243

Kuczynski’s administration the trend deepened. These figures, then,

are relatively stable: the number of ministers who are professional pol-
iticians is steadily decreasing, and that of technocrats and/or business-
men is steadily increasing (Vergara and Encinas 2016). Likewise, the
role of private organizations in the public sphere remained active and
influential. It is undeniable that in several sectors of the state, the busi-
ness sector became an important veto player (Vergara 2012). In this
way, the business sector has become a crucial island of the conserva-
tive archipelago and of the defense of the 1993 economic model.

The Struggle over Legacies

Thus far, we have seen that the gradual consolidation and defense of
Peru’s neoliberal economic regime is associated with the success of the
conservative archipelago, whose formation was the legacy of the IAC
and its resolution. A strong conservative Right was consolidated that
prioritized the maintenance of market freedom but placed less impor-
tance on other freedoms. It was also a Right whose working mecha-
nism within traditional liberal institutions was based more on influ-
ence than on public deliberation. In this regard, the formation and
consolidation of the conservative archipelago stemmed from the fail-
ure of the liberal and institutionalist Right that was active during the
1988–1992 period, in which the convergence of traditional right-wing
parties and the new economically liberal Right was proposed. Thus,
the defeat of one Right and the consolidation of another is a legacy
rooted in Peruvian society and linked to the IAC.
The structuring of the conservative archipelago is the sum of inter-
related legacies (ideological and organizational). Still, the different ac-
tors of the archipelago utilize the memory the IAC and their interpre-
tation of the conflict to defend the economic order. Of course, this
approach is not unique to the conservative archipelago. All actors and
sectors of Peruvian politics seek to build their own narrative of the
IAC. While the Left was the most successful at building academic and
scholarly concern for the violence, the Right has succeeded in vulgar-
izing and disseminating an interpretation of the conflict in which the
Left is linked to the pre-Fujimori economic failure of the country and
to the SP chaos. 21 This narrative then allows for a discourse where the
defense of the neoliberal model becomes the defense of a country free
of terrorism. Thus, in addition to the political and institutional lega-

Soifer_6844-final.indb 243 8/17/18 11:53 AM

 244 Alberto Vergara and Daniel Encinas

cies, there is an open and ongoing race to continue building discursive

and narrative legacies.
Two recent events in the national political arena facilitate an analy-
sis of how this works. The fi rst example is the 2010 mayoral electoral
campaign in Lima. Faced with the possible victory of the leftist candi-
date Susana Villarán, many islands of the conservative archipelago at-
tacked her with arguments tied to the IAC. The leaders of the partidos-
bancadas quickly charged that extremist movements surrounded
Villarán. She was accused of having a pact with the MRTA and being
surrounded by convicted terrorists, 22 and the front pages of several na-
tional newspapers alluded to a (nonexistent) association between Vi-
llarán and Abimael Guzmán. 23 The technocrats also joined in to claim
that Villarán’s candidacy posed danger. Pedro Pablo Kuczynski, for
example, declared that the arrival of extremists to the municipality
of Lima would produce “a shock” in the international markets.24 The
Church likewise had its say. Cardinal Cipriani stated with reference
to Villaran’s candidacy that it was not the right time for ideologies,
and that he sensed “a fear, like the one during the period of the Shin-
ing Path.”25 Consequently, Villarán’s performance in the opinion polls
during the last few days before the election, when these attacks esca-
lated and the electorate took fright, decreased sharply (Tanaka 2010).
This type of discourse, which links terrorism with a challenge to
the economic model, recurs whenever social movements emerge in op-
position to certain large mining investments. The confl ict over the
Conga mining project in Cajamarca is a case in point. 26 In 2011, vari-
ous local and regional actors mobilized against the project. Faced with
the danger of losing one of the largest investments projected for the
upcoming years in Peru, Ollanta Humala’s government, politicians,
the media, the Church, and technocrats raised the specter of terror-
ism and anti-investment. Among the technocrats, a former minister of
the environment stated, “these groups, Santos’ Patria Roja, Saavedra’s
MRTA, and Marco Arana’s Grufides group, are extremists. [The pro-
tests] will result in a terrorist movement.”27 Another minister claimed
that “Santos and Abimael [Guzmán] have caused the same damage to
Cajamarca, Santos even more [. . .] Cajamarca is in a terrible reces-
sion.”28 The fujimorista congressman Octavio Salazar said, “Gregorio
Santos acts like a terrorist when he calls for a revolution.”29 Finally, as
always, Cipriani voiced the Church’s position, stating that the protest-
ers contained “few groups, but nevertheless terrorists.”
These two examples, coupled with the Repsol case mentioned at the
start of this chapter, show how the conservative archipelago acts to de-

Soifer_6844-final.indb 244 8/17/18 11:53 AM

 From a Partisan Right to the Conservative Archipelago 245

fend the 1993 economic order in Peru. There is considerable ideolog-

ical cohesion surrounding the economic model, even though there is
no organic structure that allows different islands to coordinate their
statements or actions. Thus, its defenders do not respond to party
lines, and they are at once inside and outside the government, and in-
side and outside the institutions.

Conclusion

In this chapter, we have examined how the IAC and its resolution af-
fected the Peruvian Right. Furthermore, we have shown the ways in
which these legacies have facilitated the persistence of the neoliberal
economic model that emerged with the 1993 constitution. While much
research has been done about the IAC’s impact on the Left, its impact
on the Right is less studied but just as important. Indeed, we have seen
how the Right mutated from a traditional partisan Right with low ide-
ological cohesiveness to a new Right, which lacks organic partisanship
but has a group of actors with ideological cohesiveness based on the
defense of the neoliberal regime. This is an important transformation
borne of the IAC’s various legacies. Therefore, one way of providing
context to our observation is to offer a comparative perspective: How
significant is this transformation within the Latin American region?
Neoliberal reforms were a crucial moment for Latin America that
significantly altered the social and political landscapes in many coun-
tries. According to Kenneth Roberts (2014), these reforms were a criti-
cal juncture that reshaped Latin American political systems. Thus, the
Peruvian political system was not the only one fundamentally trans-
formed by the economic crisis of the 1980s and the subsequent in-
troduction of neoliberal reforms. Roberts also states that “neoliberal
critical junctures were especially destabilizing where they eroded
party-society linkages and blurred the distinctions between major par-
ties” (Roberts 2014, 20). Both phenomena he describes are remarkably
close to what happened in Peru. Thus, it is tempting to claim that the
neoliberal reforms alone explain the Peruvian case and the transfor-
mation of the Right. However, it is important to highlight that Peru is
not only a post-Washington consensus country but also a post-confl ict
country. This certainly does not invalidate Roberts’s general thesis re-
garding Latin America, but it does suggest that the IAC introduced a
dose of specificity to the Peruvian case.
The fact that party systems were reshaped (strengthened or weak-

Soifer_6844-final.indb 245 8/17/18 11:53 AM

 246 Alberto Vergara and Daniel Encinas

ened) by the neoliberal reforms does not explain the extent of the
changes that Peru experienced compared to other countries in simi-
lar situations. First, the defeat of the Left and of organized civil so-
ciety was more resounding in Peru. If neoliberal reforms alone were
enough to weaken the Left and civil society everywhere else, the ad-
dition of an IAC meant more devastating consequences for Peru. As
we have explored in this chapter, the transformation and hegemony
of the Right cannot be understood without considering the brutal de-
feat of the societal Left during the IAC. Moreover, neoliberal reforms
led to the technocratization of Latin American states (Grindle 2012),
but in few cases other than Peru did technocrats achieve such high lev-
els of isolation and direct influence in the government (Dargent 2015).
This technocratic strength is directly linked to the IAC’s resolution
and the defeat of right-wing partisan political forces. Perhaps Chile is
another case where, like Peru, the technocratic sector achieved auton-
omy and influence in the wake of an authoritarian government that
had emerged following widespread social unrest. Still, even though
neoliberal reforms placed more importance on the business sector and
weakened political parties everywhere else, both outcomes were more
profound in Peru. Furthermore, when the right-wing parties weak-
ened in other countries, it led to the long-term weakening of the right-
wing agenda and not to the strengthening that occurred in Peru. Thus,
the fragmented but ideologically cohesive islands that form the Peru-
vian conservative archipelago are significant for their singularity in the
Latin America context, beyond the common results that neoliberal re-
forms tend to produce.
However, the peculiarity of the Peruvian case stems not only from
the significant changes caused by the IAC but, above all, from the per-
sistence of these changes over time. The fact that neoliberal reforms
fundamentally reshaped party systems did not mean their political sys-
tems were frozen. For example, Bolivia and Argentina are like the Pe-
ruvian case in that they share the two characteristics, mentioned pre-
viously, that Roberts describes as most harmful to party systems.
However, in Bolivia and Argentina, the breakdown of the party sys-
tem subsequently led to a post-neoliberal political life, even though the
party system was not fully restored. In Peru the party system was like-
wise not restored, but in this case a post-neoliberal political life did
not emerge. In other words, the Peruvian case is a significant varia-
tion within a common pattern for the dismantling of systems of rep-
resentation. One viewpoint regarding Latin America’s shift to the Left

Soifer_6844-final.indb 246 8/17/18 11:53 AM

 From a Partisan Right to the Conservative Archipelago 247

claims that Latin America adopted populism in places where party sys-
tems collapsed and outsiders emerged (Kaufman 2011; Flores-Macías
2012). Peru meets both general prerequisites, but they were not crucial
there. This is because in this case, the IAC’s unique and dreadful tra-
jectory was added to the well-traveled path of Latin American neolib-
eral reforms. In Peru, the sticky neoliberal reforms adhered with twice
as much intensity. Therefore, the Peruvian conservative archipelago
defended neoliberal reforms more successfully than any other South
American country because the reforms that emerged were linked to
the IAC and its consequences.

Notes

1. Internal Armed Confl ict (Confl icto Armado Interno) is the term used by
the Truth and Reconciliation Committee (Comisión de la Verdad y Reconci-
liación, or CVR).
2. For more about the confl icts within the Peruvian Right during the 1980s,
see Conaghan and Malloy 1994.
3. In the term “partido-bancadas,” partido refers to a political party
and bancada refers to a number of congresspersons acting as a group in the
legislature.
4. This percentage was calculated by the Comisión de la Verdad y Recon-
ciliación. CVR 2003a, Conclusion 27.
5. According to the CVR, this label includes both local authorities (such as
mayors, governors, lieutenants, and justices of the peace) and national author-
ities (such as ministers, congressmen, and congresswomen). In other words, it
includes state authorities.
6. This percentage was calculated by the authors based on the CVR’s Sta-
tistical Appendix (2,359 of 13,470 victims with information for the occupa-
tion variable). CVR 2003a, Statistical Appendix, p. 86.
7. This percentage was calculated by the CVR. See CVR 2003a, 1:169.
8. This is a record of the SL’s activity based on reports in national news-
papers. In this sense, it should be considered a database of representative cases
rather than a database from which exact frequencies can be obtained.
9. We could infer that at least some of the attacks on homes showed in ta-
ble 9.1 correspond to right-wing authorities, leaders, and party members.
10. Personal interview by the authors of this chapter, May 2014.
11. “Winner Shining Path” refers to the idea that the Shining Path was un-
beatable and would win in the war against the state.
12. For more about the way in which Velasco and Fujimori represent two
similar movements against civil society and political parties, see chapter 3 of
Vergara 2015a.
13. Personal interview, May 2014.
14. As a candidate for the fujimoristas, Martha Chávez obtained only
7.4 percent of the vote in 2006.

Soifer_6844-final.indb 247 8/17/18 11:53 AM

 248 Alberto Vergara and Daniel Encinas

15. A September 2012 national survey by IPSOS-Apoyo showed that

50 percent of respondents named Alberto Fujimori as one of the main individ-
uals to have contributed to the demise of the SL.
16. When Pedro Pablo Kuczynski was elected president in 2016, fujimo-
rismo acquired a more prominent role in Congress, having won 53 percent of
the congressional vote share. This triggered frequent spats and horse-trading
with the executive, but despite all the political noise, little has changed in
terms of the cohesiveness of right-wing actors regarding the neoliberal regime.
17. The MRTA launched its struggle against the Peruvian state in 1984,
and is responsible for 1.5 percent of the fatalities during the IAC (CVR 2003a).
18. For example, the show was useful in organizing the 2013 “March for
Life” in which thousands of people mobilized against all types of abortion,
even therapeutic abortion (for medical necessities such as saving the life of the
pregnant woman).
19. Roberto Abusada founded the IPE and then became an economic ad-
viser for the government. Jorge Baca was the IPE’s director and then became
minister of the economy and fi nance. Leoni Roca was adviser to the prime
minister and later director of the IPE. Fritz Du Bois left his post as an eco-
nomic adviser for the government to take the position of IPE director (Arce
2005, 45).
20. We examined all the ministers in each ministry from the Toledo gov-
ernment (2001) to Humala’s third cabinet (June 2013); there was a total of
262 incumbents, who were classified as “businessmen,” “politicians,” “tech-
nical,” and other such categories.
21. See chapter 11 in this volume.
22. “Unidad Nacional exigió a Susana Villarán que revele si lleva en
su lista a miembros de Patria Roja y el MRTA,” Diario El Comercio, Au-
gust 6, 2010, http://elcomercio.pe/politica/gobierno/unidad-nacional-exigio
-susana-villaran-que-revele-si-lleva-su-lista-miembros-patria-roja-mrta
-noticia-619057.
23. See García Llorens 2010. For more on a previous time period (1990–
2006), see Fowks 2006.
24. “Kuczynski advierte ‘un sacudón’ en los mercados internacionales
por preferencias electorales hacia la izquierda,” Diario El Comercio, Sep-
tember 21, 2010, http://elcomercio.pe/economia/peru/kuczynski-advierte
-sacudon-mercados-internacionales-preferencias-electorales-hacia-izquierda
-noticia-642490.
25. “Cipriani sobre elecciones municipales: ‘No es momento de ideolo-
gías,’” Diario La República, September 25, 2010, http://www.larepublica
.pe/25–09–2010/cipriani-sobre-elecciones-municipales-no-es-momento-de
-ideologias.
26. The mining project was approved during Alan García’s government
(2006–2011), and with almost US$5 million invested, it should be one Peru’s
largest investment projects. The investors are the Newmont Mining Cor-
poration, from the United States (51.35 percent); Buenaventura, from Peru
(43.65 percent); and the World Bank as a minority partner through the Inter-
national Finance Corporation (5 percent).

Soifer_6844-final.indb 248 8/17/18 11:53 AM

 From a Partisan Right to the Conservative Archipelago 249

27. “Brack: ‘Protestas en Cajamarca terminarán en atentados terroris-

tas,’ ” Diario Perú 21, July 7, 2012, http://peru21.pe/2012/07/07/actualidad/
brack-protestas-cajamarca-terminaran-atentados-terroristas-2032087.
28. “Silva: ‘Gregorio Santos le hizo más daño a Cajamarca que Abimael
Guzmán,’” Diario Perú 21, May 7, 2013, http://peru21.pe/politica/silva
-martinot-gregorio-santos-le-hizo-mas-dano-cajamarca-que-abimael
-guzman-2129918.
29. “Octavio Salazar: Gregorio Santos actúa como terrorista,” Diario Co-
rreo, December 31, 2011, http://diariocorreo.pe/ultimas/noticias/EPENSA
-060024/octavio-salazar-gregorio-santos-actua-como-terrorista.

Soifer_6844-final.indb 249 8/17/18 11:53 AM

 CHAP T ER 10

Public Opinion, the Specter of Violence,

and Democracy in Contemporary Peru
Arturo Maldonado, Jennifer L. Merolla,
and Elizabeth J. Zechmeister

Democracy and violence have an intricate relationship in Peru. At the

moment of Peru’s transition to democracy in 1980, Sendero Luminoso
(Shining Path) launched a guerrilla war. The government’s program
against Sendero and, after 1987, against the Movimiento Revolucio-
nario Túpac Amaru (MRTA) employed a range of counterinsurgency
tactics, many of which involved incursions on civil and human rights.
As the rebel groups’ efforts became more lethal and visible, the gov-
ernment’s commitment to democratic rules of the game diminished,
reaching a nadir in 1992 when President Alberto Fujimori executed an
autogolpe under which he temporarily closed the Peruvian Congress.
We are now over three decades removed from Peru’s 1980 democratic
election and the public emergence of Sendero Luminoso, and roughly
two decades past a period in which the violence was at its peak and de-
mocracy suffered near-fatal blows. Today Peru’s democracy is mostly
recovered and terroristic threats of violence have mostly receded. In
this context, we ask about the public opinion legacies of the confl ict:
to what extent does worry about violent attacks occupy the contem-
porary Peruvian public, and with what implications for democratic
values?
We address this question by examining data generated from pub-
lic opinion polls and from an original experiment fielded in the sum-
mer of 2012. We fi rst provide an overview of relationships among the
threat of violence, democracy, and public opinion in Peru. Second, we
assess responses to the AmericasBarometer 2010 study in which re-
spondents across the Latin American region were asked the extent to
which they are worried about violent terrorist attacks; the question

Soifer_6844-final.indb 250 8/17/18 11:53 AM

 Public Opinion, Violence, and Democracy in Peru 251

was intentionally vague so as not to presume a particular perpetra-

tor. We document that mean levels of worry are higher in Peru than in
most other Latin American countries. We then examine predictors of
concerns about terroristic violence in Peru. Here we are able to assess
the extent to which, and how, experiences have translated into what,
for some, are memories of conflict. We fi nd that those who were young
during the peak of terrorist incidents tend to be somewhat more wor-
ried about the possibility of new violent attacks. Yet concern about the
threat of violence is not geographically concentrated in regions that
experienced more fatalities during Sendero’s campaign in the 1980s
and 1990s; instead, worry has diffused and taken root to a greater de-
gree in parts of the country that were comparatively less affected by
the violence. One possible explanation is that while these parts of the
country lack the same level of direct experience with the confl ict, they
also have less information about the degree to which high confl ict ar-
eas have returned to a safer everyday norm.
Third, we assess the relationship between worry about terroris-
tic violence and support for democratic values. We show that those
who are more worried are more likely to prefer iron-fisted rule and less
likely to be politically tolerant, but are more likely to prefer that police
respect the rule of law. Fourth, we introduce an experiment designed
to assess the causal effects of raising the salience (via media reports) of
domestic terrorism in contemporary Peru. We find some evidence, al-
though it is weak, that exposure to news about terrorist threats cor-
rodes democratic public opinion in general. This may reflect a reality
in which news about terrorist plots is not uncommon and the public
is already tuned into terrorism, so that there is little marginal conse-
quence to new information. Nonetheless we also fi nd strong evidence
of important heterogeneity in reactions to terroristic threats: those
who express support for fujimorismo become comparatively more sup-
portive of iron-fisted rule and less politically tolerant when exposed to
the threat of violence.
We conclude that Peru’s violent past has left indelible marks on pub-
lic opinion in the post-confl ict era. The confl icts that ravaged Peruvian
society and democracy in the 1980s and 1990s are reflected today in
a unique public opinion landscape, one which stands apart from that
found in other Latin American countries and contains important fis-
sures in orientations toward the threat of terroristic violence and dem-
ocratic values within the country.

Soifer_6844-final.indb 251 8/17/18 11:53 AM

 252 A. Maldonado, J. L. Merolla, and E. J. Zechmeister

Violence, Democracy, and the Peruvian Public

Sendero Luminoso surfaced at the very moment of Peru’s return to

democratic electoral politics in 1980 when the group claimed responsi-
bility for burning polling stations in Ayacucho. With intellectual roots
grounded in a Maoist tradition, Sendero embarked on a campaign of
terroristic acts. The Peruvian government and armed forces responded
in an arguably tit-for-tat fashion to Sendero’s attacks, using a range
of confrontational counterterrorist tactics and ultimately conspiring
with paramilitary units in the confl ict. The confl ict reached a zenith in
terms of fatalities in the mid-1980s, around the time that the govern-
ment found itself on the defensive against both Sendero and the newly
emerged MRTA.
Economic and political problems, seriously aggravated by the do-
mestic strife, fueled the presidential candidacy of political outsider Al-
berto Fujimori in 1989. Fujimori’s fi rst years in office witnessed an
increase in the severity of the conflict, as paramilitary and guerrilla
groups increased the brutality of their actions. Under the specter of ter-
rorist, economic, and governance crises, in 1992 Fujimori took the ex-
treme measure of executing an autogolpe, in which he shut down Con-
gress, neutered the courts, suspended the constitution, and ruled by
decree until, under international pressure, new elections were held for
a constituent congress. Despite the government’s capture of Sendero
Luminoso leader Abimael Guzmán that same year, the armed conflict
with militant insurgents continued, peaking in 1997 with the MRTA’s
attack on the Japanese Embassy, where they took hundreds hostage
and kept seventy-two individuals captive for more than four months.
The standoff ended when the armed forces stormed the embassy and
killed all fourteen rebels. In 1999 the government won another vic-
tory against the guerrillas when Sendero’s second-in-command, Oscar
Ramírez Durand, was captured. In 2000 Alberto Fujimori’s govern-
ment came to an abrupt end under accusations of corruption and hu-
man rights abuses.
Conflict with armed guerrilla movements in the 1980s and 1990s
took a heavy toll on both the Peruvian population and its democracy.
According to the Truth Commission report (CVR 2004), nearly sev-
enty thousand individuals were killed as a result of the insurgency
and counterinsurgency. Forty percent of dead and disappeared victims
were from the department of Ayacucho, with another 45 percent spread
across the departments of Junín, Huánuco, Huancavelica, Apurímac,

Soifer_6844-final.indb 252 8/17/18 11:53 AM

 Public Opinion, Violence, and Democracy in Peru 253

and San Martín.1 Most of those killed were poor, less educated, ru-
ral, indigenous, young, and male: 38 percent of the victims were from
the poorest quintile; 68 percent had less than a secondary education;
79 percent were from rural areas; 75 percent spoke Quechua or an-
other indigenous language as their fi rst language; and most of the vic-
tims were young males.2
The government’s war against the guerrillas left the state respon-
sible, directly or by association with paramilitary groups, for many
deaths. The Truth Commission report indicates that agents affiliated
with Sendero were responsible for most of the victims (54 percent), but
state actors and tactics contributed to the death toll. The government’s
willingness to side-step the rule of law and its apparent tolerance for
collateral damage in the war against rebel groups led to accusations of
human rights violations. Ultimately, Alberto Fujimori was sentenced
to twenty-five years in prison for his responsibility in two killings in
Lima in the early 1990s, and Vladimiro Montesinos, Fujimori’s for-
mer main adviser, and several military officials were also condemned
to prison for human rights abuses.
Briefly put, both the functioning and quality of democracy decayed
in serious ways during the 1980s and especially in the 1990s. The Pe-
ruvian state ultimately emerged victorious in its war against the guer-
rillas, as evidenced in the fact that the insurgents’ strength and efforts
decreased substantially by the year 2000. But this victory came at a
significant cost to Peru’s democracy.
Following his autogolpe, Fujimori was reelected to the presidency
in 1995 with 64 percent of the vote and maintained a presidential ap-
proval rating between 45 percent and 80 percent until the fi nal period
of his presidency.3 In 1993, one-third of Peruvians identified terror-
ism as one of the most important problems in the country. Arce (2003)
mentions that “the defeat of terrorism” was the second most impor-
tant reason for reelecting Fujimori in 1995 and, even in 1999, 70 per-
cent of Peruvians declared this was a reason for approving of Fuji-
mori’s performance.4 After Fujimori’s resignation and exile in 2000,
57 percent expressed the opinion that he would be remembered in
twenty years for defeating terrorism, but just 4 percent mentioned ter-
rorism as an important problem for the country. In a systematic analy-
sis, Arce (2003) fi nds that political violence (the number of “subver-
sive actions”) can be a significant predictor of presidential approval
in Peru, but it is negative for García (1985–1990) and positive for Fu-
jimori (1990–1997). Thus, while guerrilla activity hurt García’s pop-

Soifer_6844-final.indb 253 8/17/18 11:53 AM

 254 A. Maldonado, J. L. Merolla, and E. J. Zechmeister

ularity, Fujimori’s approval increased with insurgent actions. Arce ar-

gues this difference is due to public perception of state actors as either
soft-liners (García) or hard-liners (Fujimori).
Though the toll that terroristic violence took on Peruvian society
and politics was extensive, the willingness of a population to trade
civil liberties for security is not unique to Peru. As we have discussed
in previous research (see Merolla et al. 2014), scholars have found
that the mass public tends to accept greater restrictions on basic rights
when exposed to terrorist threat (e.g., Davis and Silver 2004; Davis
2007; Huddy et al. 2005; Merolla and Zechmeister 2009). Some work
also shows an elevated willingness among a threatened public to ac-
cept extralegal activities by security officers (Fernandez and Kuenzi
2010; Pérez 2003; Pérez 2009; Malone 2010; Malone 2013; Merolla,
Mezini, and Zechmeister 2013; but also see Ceobano, Wood, and Ri-
beiro 2010). And in previous research we have shown a tendency for
individuals in some countries to prefer a centralization of power in the
executive under conditions of terrorist threat (Merolla and Zechmeis-
ter 2009).
Today in Peru the topic of violent conflict remains salient, though
it no longer is identified by more than a small portion of the public
as the most important problem. Peruvians mentioned the fight against
terrorism and narco-trafficking (a related issue) as reasons for their
approval of García’s government in 2010, but the numbers are small
(9 to 11 percent)5 in comparison with the percentage that mentioned
these reasons for approving Fujimori’s government. While Peruvians
may not be as concerned about terroristic violence, they do not think
that the threat has completely receded and have been dissatisfied with
the way the government is handling the issue. For example, in 2012,
66 percent of the public expressed the opinion that Ollanta Huma-
la’s administration would not defeat terrorism prior to the end of his
term in 2016,6 and 54 percent disapproved of his efforts on the issue.7
These concerns were likely fueled by sporadic incidents attributed to
remnants of Sendero. Reflecting terrorism’s continued salience, knowl-
edge of groups affiliated with terrorism remains high. In a September
2012 poll, 83 percent of individuals could correctly name the founder
of Sendero Luminoso and 62 percent could identify the department
where domestic terrorism emerged in the early 1980s.8
In this chapter we examine public opinion in Peru in the contempo-
rary, post-confl ict period. As a fi rst objective, we seek to understand
the legacies that the conflict has left in terms of worry about violence

Soifer_6844-final.indb 254 8/17/18 11:53 AM

 Public Opinion, Violence, and Democracy in Peru 255

in the country. In Peru, as in many post-conflict societies, there is a

debate over how memories of the conflict have persisted, and among
whom (see Hamman et al. 2003; chapter 11 in this volume). We want
to know whether concerns about terrorist attacks are higher in Peru
than they are in other countries and, if so, then among which popula-
tions do we fi nd the highest levels of worry about violence? A key dis-
tinction we consider is between those with more or less connection to
conflict-era violence. It is the case that in some areas of the country the
conflict was more exacerbated, and certain demographic groups were
more affected by violence during the conflict. Yet levels of worry and
awareness about the confl ict may not necessarily be circumscribed to
those who directly experience confl ict. For example, there is evidence
that the memory of the Peruvian confl ict has spread to younger gener-
ations (but see chapter 11 of this volume). In Ayacucho, older individu-
als are a primary source of information about the political conflict for
younger students (Muñoz-Nájar and del Pino 2012). In short, one po-
tential mechanism for diffusion is conversations with older relatives
who experienced the confl ict and communicate their knowledge, gen-
erating awareness and possibly concern in younger generations.
As indicated above, extant evidence suggests that levels of aware-
ness and worry may have decreased over time, but still remain salient.9
It is possible that over time concerns about terrorism have diffused
across and taken root in locations and subgroups that were not partic-
ularly affected by conflict-era violence. Generally speaking, concerns
among the public are likely heightened when particular incidents call
up memories of the past, such as when remnants of Sendero Lumi-
noso, acting through their affi liate MOVADEF (Movimiento por Am-
nistía y los Derechos Fundamentales), try to participate in political af-
fairs, or when narco-traffickers in alliance with former guerrilleros
attack police and military patrols or stations.
The fact that current incidents can serve to heighten concerns about
terroristic violence leads us to our second core objective, which is to
assess the extent to which the heightened threat of violence in the post-
conflict era correlates with diminished democratic attitudes among the
Peruvian public. Theoretically, we could expect one of at least two re-
lationships: fi rst, it is possible that the legacy of the country’s experi-
ence with democratic decay in response to terrorist threats may leave
the public numb to or unwilling to make such trade-offs in the modern
time period; second, it is possible that the tendency for democratic at-
titudes to erode under conditions of threat persists as part of a univer-

Soifer_6844-final.indb 255 8/17/18 11:53 AM

 256 A. Maldonado, J. L. Merolla, and E. J. Zechmeister

sal tendency. In addition to assessing the general reaction of the public,

we look at subgroups within the public with respect to those who live
in regions that were most affected, and with respect to those who con-
tinue to support the fujimorista political camp.

Worry about Violence in Peru

We begin by assessing levels of worry about the threat of terroristic

acts in Peru. Our data come from the 2010 AmericasBarometer by the
Latin American Public Opinion Project (LAPOP).10 In those national
surveys, citizens across the Americas were asked to report on their
concerns about the possibility of violent terrorist attacks in their coun-
try and, as well, their worry about the possibility that they or a fam-
ily member will be a victim of a terrorist attack. The specific wording
of these two questions is as follows (with their original survey code in
parentheses):

1. (WT1) How worried are you that there will be a violent attack by
terrorists in [Country] in the next 12 months? Are you very, some-
what, a little, or not at all worried, or would you say that you have
not thought much about this?
2. (WT2) How worried are you that you or someone in your family
will become a victim of a violent attack by terrorists? Are you very,
somewhat, a little, or not at all worried, or would you say that you
have not thought much about this?

We recoded responses on a four-point scale, where a higher value

indicates more worry. For the analyses here, we omit those who re-
sponded that they had not thought about it or otherwise did not an-
swer the question. Notably, however, the mean proportion of those
who had not thought about it for the eighteen Latin American coun-
tries as a whole (0.23 for WT1 and 0.22 for WT2) is higher than those
found in Peru (0.15 for both WT1 and WT2). This is consistent with
the notion that concerns about terrorism are more salient in Peru than
in other countries in the region on average.
Figure 10.1 presents average levels, for each country, for the two
worry measures. Examining mean public opinion across the eighteen
countries that are commonly considered to constitute Latin America,
we fi nd that mean concerns about terroristic acts in general and with

Soifer_6844-final.indb 256 8/17/18 11:53 AM

 Public Opinion, Violence, and Democracy in Peru 257

Figure 10.1. Mean worry about terroristic acts across countries. Source:
AmericasBarometer 2010; © AmericasBarometer, by LAPOP.

respect to personal victimization are 2.5, in each case, for the region.11
Thus, across Latin America as a whole, concerns about terrorism are
moderate: the average response is between “a little” and “somewhat”
worried. In Peru, however, these mean concerns each register at 2.9. In
short, Peruvians are more worried about violent attacks than the av-
erage Latin American; the average response to each question in Peru
is equivalent to “somewhat” worried on the question’s response scale.
The differences between Peru’s average responses to these questions
and the regional averages are statistically significant (p < 0.001).
In terms of general worry about terrorist attacks, Peru ranks fourth
among the Latin American countries; it trails just slightly behind Co-
lombia, Ecuador, and Paraguay. In terms of concern about personal
victimization, Peru ranks third in Latin America, just behind Ecuador
and Colombia.12 While recent incidents, even if sparser than during
the conflict era, likely play a role in the comparatively elevated levels of
concern in Peru, it is also probable that current levels of worry in Peru
are linked to experiences with political violence that took place de-
cades ago, during the peak of the confl ict era. To the degree that is the
case, then the evidence presented here provides some support for those
who argue against the notion that Peru is a community without mem-

Soifer_6844-final.indb 257 8/17/18 11:53 AM

 258 A. Maldonado, J. L. Merolla, and E. J. Zechmeister

ory of the past confl ict; that is, the results in figure 10.1 suggest that
the past casts a shadow over public opinion in post-conflict era Peru,
in the form of elevated levels of concern about the possibility of new
violence perpetrated by terrorist groups in the country.
What predicts concerns about violent attacks by terrorists in Peru?
We consider a number of socioeconomic and demographic character-
istics of individuals, paying particular attention to factors that distin-
guish those who were more affected by violence and those who were
less affected during the conflict era. As noted above, victims of the ter-
rorist conflict tended to be young, male, poor, indigenous, rural, and
less educated. We consider age by creating a set of age cohort indica-
tors that capture whether an individual was a young adult in 2010;
was a young adult during the peak of the confl ict (we call this belong-
ing to the “affected generation”); or belongs to the older cohort.13 In
considering gender, it is important to take into account that there is a
significant and wide-ranging body of evidence that women tend to ex-
press more concern about terrorist threats than men (e.g., Brück and
Müller 2009; Friedland and Merari 1985; Huddy et al. 2005; Klar,
Sharvit, and Zakay 2002; Lemyre et al. 2006; Nacos, Bloch-Elkon,
and Shapiro 2011; Nellis 2009). Thus, while men were more affected
in Peru, we expect women to be more concerned. We also include a
measure of wealth,14 ideology,15 indigenous identity (the baseline is
white identity, and we also include dummy variables for mestizo and
other), rural residence, and education.16 In addition, we include a mea-
sure intended to capture the extent of the toll that past violence took
in terms of fatalities in the province in which the individual lives.17
This measure is extracted from the records of the Truth Commission
in Peru. We include a region measure that is coded as intervals of fatal-
ities in the province in which the individual lives, where a higher value
means a province with more deaths and disappearances. We use this
measure in our models as a series of three dummy variables for regions
(intervals) 2, 3, and 4; the comparison category is region (interval) 1.18
The number of fatalities increases with each region measure, with the
fewest in region 1.19
The results of two OLS analyses, one for each dependent variable,
are presented in figures 10.2 and 10.3. 20 For the analyses, all inde-
pendent and dependent variables have been rescaled (via linear trans-
formations) from 0 to 1 for ease of comparison and interpretation.
The figures present the nonstandardized regression coefficients and a
90 percent confidence interval for each independent variable.

Soifer_6844-final.indb 258 8/17/18 11:53 AM

 Figure 10.2. Predicting general worry about violent attacks. Source:
AmericasBarometer 2010 v14; © AmericasBarometer, by LAPOP.

Figure 10.3. Predicting personal worry about violent attacks. Source:

AmericasBarometer 2010 v14; © AmericasBarometer, by LAPOP.

Soifer_6844-final.indb 259 8/17/18 11:53 AM

 260 A. Maldonado, J. L. Merolla, and E. J. Zechmeister

Considering age fi rst, we see that those in the affected age cohort
are more worried in general and with respect to personal victimization,
in comparison to the younger cohort (and to some extent in compar-
ison to the older cohort, which is indistinguishable from the younger
cohort). Thus, there is evidence that the generation of individuals that
was most intensely affected by the violent confl ict expresses greater
levels of concern about the threat of terroristic acts in comparison to
the rest of the Peruvian public. As anticipated, we fi nd that women are
more worried about terrorist threats, a result that is consistent across
both measures. We do not find significant results for wealth, rural res-
idence, or ethnic identity, though, interestingly, the coefficients on the
ethnicity dummy variables are all negative, suggesting that there may
be some negligible tendency for those who identify as white (13 per-
cent of the sample) to be more concerned about the threat of violent at-
tacks by terrorists. The coefficient on education in both models is also
negative; it falls outside of a standard boundary for statistical signifi-
cance in the fi rst model (at p = 0.213) and is statistically significant in
the second model (p = 0.017). This set of results supports the conclu-
sion that those who are less educated are in some ways more worried
about terrorist threat in Peru.
Finally, the results by region are significant but perhaps somewhat
counterintuitive: those who live in regions that were most affected
with respect to fatalities are less worried than those who live in the
least-affected region (the baseline category for the analysis). It seems
that contemporary fear is greater where there is less immediately ac-
cessible information about past terrorist threats and experiences. This
fi nding is in line with the fact that indigenous people, the most af-
fected group by ethnicity, report (marginally) lower levels of worry.
Yet the result seems to run counter to the fi nding that those who are
less educated and those of the affected generation by age cohort are
more worried. Collectively, then, the results speak to the complexity of
the dynamics by which the public opinion legacy of the confl ict era has
diffused across the population, taking root to a greater degree among
some subgroups than in others. With respect to the fi nding on region
of the country, it may be that those who live in regions that were pre-
viously affected are influenced by their direct observation of these re-
gions in the contemporary time period, an experience that may leave
them more certain of their current security than those who live at a
distance from these places from which the past conflict sprang.

Soifer_6844-final.indb 260 8/17/18 11:53 AM

 Public Opinion, Violence, and Democracy in Peru 261

The Relationship between the Threat

of Violence and Democratic Values

Worry about terroristic acts is not a static property: it ebbs and flows
within and across a society over time. One factor that influences this
dynamic is the content of media reports. As noted above, informa-
tion periodically appears in the Peruvian news that raises the salience
of terrorism. In addition to the news of actual violent incidents, dis-
cussion of organizations that are legacies of the confl ict era might
heighten public awareness and concern about terrorism, as might news
on contemporary crime and violence associated with narco-traffickers
or other groups.21 Given that media stories can and do raise the sa-
lience of terrorism among the Peruvian public, we now turn attention
to this question: what consequences does the threat of violence have
for democratic public opinion in the post-confl ict period?
Our basic contention is simple: fear of terrorist attacks and raising
the salience of the threat of violence have consequences for the ways
people think about others and about government. A chief objective of
terrorism is to induce anxiety and fear, and extant research focused
primarily on the United States shows clearly that the threat of terror
significantly affects political attitudes, evaluations of leaders, and be-
haviors in ways that can place stress on democratic values, processes,
and even institutions (Brooks and Manza 2013; Davis 2007; Huddy
et al. 2005; Malhotra and Popp 2012; Merolla and Zechmeister 2009).
While much of this research has been limited to a focus on specific
values and preferences, in other work we argue and fi nd support for
the contention that on average, worry about terroristic violence un-
dermines individuals’ support for democracy and democratic practices
across the Latin American region (Merolla et al. 2014). This line of re-
search supports an expectation that worry about terrorist attacks will
be associated with lower democratic values in Peru. Yet context may
matter, and in a country that experienced a prolonged period of inter-
nal conflict, and in which the government was culpable for many ca-
sualties, and in which concerns about terrorist threats are quite ele-
vated in comparison to other countries (thus potentially at a saturation
point), we may fi nd that individuals either are not as perturbed by news
of the threat of terroristic acts or even rally around democratic institu-
tions. In this section we further develop these two rival expectations.
As we have stated elsewhere, the threat of terrorism highlights indi-

Soifer_6844-final.indb 261 8/17/18 11:53 AM

 262 A. Maldonado, J. L. Merolla, and E. J. Zechmeister

viduals’ inability to control external circumstances that threaten their

fi nancial, psychological, or physical well-being (e.g., Merolla and Zech-
meister 2009; see also Huddy et al. 2005). In attempting to reestablish
feelings of control and safety, individuals may adopt one or more cop-
ing strategies that carry political consequences (Merolla and Zechmeis-
ter 2009). In particular, those in a condition of threat may become less
trusting of others, more intolerant of out-groups, and more supportive
of authority. Moreover, individuals might select to turn over control to
a political actor, such as the chief executive, whom one deems capable
of solving or handling the crisis. In the face of terrorist threats, individ-
uals may come to prefer that the balance of power tilt decidedly in fa-
vor of a stronger executive, one who can then pursue a quick, unhin-
dered resolution to the crisis. 22 These tendencies may also extend to the
practice of the rule of law more generally, as individuals prioritize han-
dling the threat, even if actors go outside the law.
Existing scholarship demonstrates support for a negative relation-
ship between the threat of terrorist attacks and support for democratic
values. In the US context, those with high levels of worry about future
terroristic acts are more willing to increase surveillance on Arabs and
Arab Americans, increase security checks on Arab visitors, and de-
crease visas to Arab countries (see Huddy et al. 2005; Das et al. 2009;
but see also Kalkan, Layman, and Uslaner 2009). They are also more
willing to trade civil liberties for more security on a range of counter-
terrorism measures (e.g., Brooks and Manza 2013; Davis and Silver
2004; Davis 2007; Huddy et al. 2005; Malhotra and Popp 2012; Me-
rolla and Zechmeister 2009). With respect to general measures of po-
litical tolerance outside of the United States, Merolla and Zechmeister
(2009) fi nd that subjects exposed to a terror threat condition in Mex-
ico are less willing to provide various rights to their least-liked group.
Peffley, Hutchinson, and Shamir (2015) demonstrate a similar rela-
tionship between terrorist incidents and decreased political tolerance
using the same type of measure over time in Israel. Finally, Merolla
et al. (2014) show that individuals worried about terrorism in Latin
America tend on average to be less politically tolerant.
Empirical evidence connecting terror threats to support for a stron-
ger executive and weaker support for democratic rules is more lim-
ited, but suggestive. In the US context, in the aftermath of 9/11, the US
public became highly supportive of the administration in general and
with respect to particular policies such as a unilateral foreign policy
approach (Huddy et al. 2005) and “terrorist surveillance” programs,

Soifer_6844-final.indb 262 8/17/18 11:53 AM

 Public Opinion, Violence, and Democracy in Peru 263

which have entailed wiretapping without approval from courts fol-

lowing the Foreign Intelligence Surveillance Act (Brooks and Manza
2013; Malhotra and Popp 2012). In a systematic analysis of survey
data from LAPOP, Merolla et al. (2014) show that individuals wor-
ried about terroristic acts are more supportive of strong leaders and
iron-fisted leadership, and are more supportive of coups under cer-
tain circumstances. Some of these relationships have also been sup-
ported in experimental work. For example, Merolla and Zechmeister
(2009) show that research participants in Mexico who read a news
story about the threat of terrorist attacks expressed preferences for a
stronger executive, compared to those who read an article highlighting
relative well-being in that country.23
Taking stock of all of this scholarship, one could expect to fi nd sim-
ilar linkages between worry about terrorist attacks and support for
democracy in Peru. More specifically, individuals worried about vio-
lence may be less supportive of basic democratic values, such as politi-
cal tolerance; less supportive of the democratic rules of the game; and
more supportive of strong, unencumbered leaders. Raising the salience
of the threat of terroristic acts may also lead to these outcomes. How-
ever, Peru’s particular historical context may condition how individu-
als react to the threat of violence.
Two aspects of the Peruvian case make it unique. First, as shown
in figure 10.1, mean scores of concerns about terroristic acts are com-
paratively high in Peru relative to other Latin American countries.
One likely cause for these elevated concerns is that the remnants of
Sendero Luminoso continue on occasion to carry out violent attacks,
which may leave the public generally more anxious and essentially sat-
urated in its exposure to the threat of violence. In fact, as evidenced in
our earlier discussion of public opinion polls, we know that informa-
tion about domestic terrorism is high in Peru. We may find in particu-
lar that when considering exposure to new information about terror-
ist threats, effects are muted because this information is introduced in
a saturated context. 24 Second, Peru has had a unique experience—in
comparison to other countries in the region—with respect to terror-
ist threats and democracy, particularly with President Fujimori’s tem-
porary shutdown of the Congress and other governmental institutions
in the context of terrorist and economic threats. While he was subse-
quently reelected, the tide of public opinion eventually turned against
his administration, and he was sentenced for his complicit role in hu-
man rights abuses by Peruvian security forces. This history may have

Soifer_6844-final.indb 263 8/17/18 11:53 AM

 264 A. Maldonado, J. L. Merolla, and E. J. Zechmeister

a sobering effect on Peruvian public opinion, inducing a resistance to

nondemocratic options, and even possibly, at least for some individu-
als, a rally around some democratic values in the face of the threat of
violent incidents.

The Relationship between Worry about

Violence and Democratic Values: Survey Data

To test the relationship between worry about terroristic violence and

democratic values using the AmericasBarometer 2010 survey data for
Peru, we fi rst combine the two measures of concern about terrorist at-
tacks into a single additive index, Worry about Terrorism.25 We as-
sess three democratic values: Anti–Iron Fist, Abide by Rule of Law,
and Political Tolerance.26 These measures are based on the follow-
ing questions and are rescaled to run from 0 to 1 so that higher val-
ues correspond to more democratic responses (original survey codes in
parentheses):

Anti–Iron Fist (Dem11): Do you think that our country needs a gov-
ernment with an iron fist (coded 0), or that problems can be resolved
with everyone’s participation (coded 1)?
Abide by Rule of Law (Aoj8): In order to catch criminals, do you be-
lieve that the authorities should always abide by the law (coded 1) or
that occasionally they can cross the line (coded 0)?
Political Tolerance Factor:27 1. (D1) There are people who only say bad
things about the Peruvian form of government, not just the incumbent
government but the system of government. How strongly do you ap-
prove or disapprove of such people’s right to vote? Please read me the
number from the [original 1–10] scale. 2. (D2) How strongly do you
approve or disapprove that such people be allowed to conduct peace-
ful demonstrations in order to express their views? 3. (D3) Still think-
ing of those who only say bad things about the Peruvian form of gov-
ernment, how strongly do you approve or disapprove of such people
being permitted to run for public office? 4. (D4) How strongly do you
approve or disapprove of such people appearing on television to make
speeches?

Figure 10.4 shows the results for the Anti–Iron Fist analysis.28 In ad-
dition to including the Worry about Terrorism measure, we include as

Soifer_6844-final.indb 264 8/17/18 11:53 AM

 Public Opinion, Violence, and Democracy in Peru 265

Figure 10.4. Predicting anti–iron fist attitudes. Source: AmericasBarometer

2010 v14; © AmericasBarometer, by LAPOP.

controls all variables that were accounted for in the previous analyses
in this chapter. Because the dependent variable is dichotomous, we run
a logistic regression analysis.
The results show that those who are more worried about terroristic
acts are less likely to select the anti–iron fist option, which prioritizes
the notion of broad participation to solve problems. That is, the nega-
tive coefficient shows that those who are more worried are more likely
to select the iron fist option. This fi nding is consistent with the frame-
work suggesting terrorist threats increase support for a strong execu-
tive. We also see that, even controlling for worry about terrorism, the
affected age cohort has a negative coefficient. This may suggest that
this group is more likely to select the iron fist option in comparison to
the younger age cohort, but the coefficient does not reach conventional
levels of statistical significance (at p = 0.209). The oldest age cohort
is more likely to choose the iron fist option than the younger age co-
hort. In this case, the coefficient is statistically significant (p = 0.003).
The fact that the younger age cohort is comparatively more likely to
select the democratic option is interesting, given other scholarship and
fi ndings suggesting that young age groups in Latin America are more
likely to express populist and related orientations that run counter to

Soifer_6844-final.indb 265 8/17/18 11:53 AM

 266 A. Maldonado, J. L. Merolla, and E. J. Zechmeister

liberal democracy (e.g., Seligson 2007). 29 The results further show that
(all else equal) those who are wealthy are more likely to select the anti–
iron fist option. Interestingly, those who live in regions that experi-
enced the highest degree of violence are also more likely to select the
anti–iron fist option. The positive sign on the coefficient for region 4
(which included provinces with more than 250 deaths and disappear-
ances, such as Huamanga, Huanta, and La Mar in Ayacucho) is robust
to a model that includes only the region dummy variables.
Figure 10.5 presents the results of a logistic regression analysis for
the Abide by Rule of Law measure. Interestingly, we see here that
worry about violence is significant but associated in a positive direc-
tion with this dependent variable. Those more worried about terroristic
acts are more supportive of police abiding by the rule of law. Though
this is speculation, we posit that it may suggest a concern among wor-
ried individuals that terrorism will be addressed with policies that en-
croach on civil and human rights in a manner consistent with the Fu-
jimori era. It is noteworthy, in this regard, that the police represent
one of the least-trusted institutions in Peru, scoring 38.5 points on a
0–100 trust scale in the 2014 round of LAPOP’s AmericasBarometer.
In addition to the significant result for worry about terrorism, we fi nd
that those who are more educated are less likely to respond that police
should abide by the rule of law. This is also an interesting fi nding, in
that it runs counter to perspectives that stem from a working-class au-
thoritarianism framework (Lipset 1959). It may be that those who are
lower on the socioeconomic scale are more likely to believe that their
own rights will be negatively affected by police crossing the line.30
Finally, we turn to Political Tolerance. In this case, we run an OLS
regression analysis. The results, presented in figure 10.6, show that
worry about terrorism is negatively associated with political tolerance.
The results mirror those we fi nd for the Anti–Iron Fist measure: those
more worried about terroristic acts are less politically tolerant, while
those who live in the most affected region are more politically tolerant.
Again, we see that the younger age cohort (the baseline) tends toward
being more democratic in its public opinion compared to the older co-
horts (though the differences here are not statistically significant).
In sum, in our analyses of survey data from the 2010 Americas-
Barometer survey of Peru, we fi nd that worry about violent attacks
by terrorists is significantly related to democratic public opinion on
the three measures we consider here. The direction of this relationship

Soifer_6844-final.indb 266 8/17/18 11:53 AM

 Figure 10.5. Predicting abide by rule of law attitudes. Source:
AmericasBarometer 2010 v14; © AmericasBarometer, by LAPOP.

Figure 10.6. Predicting political tolerance. Source: AmericasBarometer

2010 v14; © AmericasBarometer, by LAPOP.

Soifer_6844-final.indb 267 8/17/18 11:53 AM

 268 A. Maldonado, J. L. Merolla, and E. J. Zechmeister

varies, such that those more worried about terrorism are more sup-
portive of iron-fisted approaches to politics and less politically toler-
ant; yet, at the same time, those more worried about terrorism are less
likely to support the police crossing the line to catch criminals.

The Relationship between the Threat of Terrorism

and Democratic Values: An Experiment

The survey data analyses provide insight into the ways in which con-
cerns about terrorist threats are related to democratic values in the Pe-
ruvian public, but we are also interested in the causal effect that the
specter of the threat of terroristic attacks has on public opinion. To as-
sess the relationship between exposure to information about terrorism
and democratic values in contemporary Peru, we need an experimen-
tal design. Therefore, and following our previous work (Merolla and
Zechmeister 2009), we designed a survey experiment in which partici-
pants were randomly assigned to a control group, a condition in which
they read either about good times in the country, or a condition that
discussed domestic terrorism. They then answered a variety of ques-
tions on support for democracy.
The sample was designed to be representative of the adult popula-
tion in the metropolitan area of Lima.31 Table 10.1 presents a small
set of descriptive statistics for the sample. The average respondent was
forty years old, with twelve years of education. Half of the sample was
female and 37 percent reported working full-time.
Participants were fi rst consented into the research project. They
then responded to a pretreatment survey that asked about their demo-
graphic and socioeconomic characteristics and their political predis-
positions. Subjects were then randomly assigned to one of the experi-
mental conditions.32 Participants in the treated conditions were asked
to read a news story and then to respond to two close-ended questions
that asked them to recall two facts from that article (with the option to
consult the news story as needed). All participants then responded to
a set of questions asking about their emotional state (using questions
recommended by Marcus et al. 2006). Finally, all subjects responded
to a survey asking about their preferences regarding democratic poli-
tics, including institutions and values.
The core feature of the experimental design is the set of short (around
400–500 words) news stories that were randomly assigned to treated

Soifer_6844-final.indb 268 8/17/18 11:53 AM

 Public Opinion, Violence, and Democracy in Peru 269

Table 10.1. Selected experiment sample descriptives

Mean Working
Percent education full-time
Obs. Mean age female (years) (%)

Lima, Peru 760 40 50 12 37

Note: Values in the table are based on the entire study, which included two
threat conditions that are not analyzed in this chapter.

subjects. These take two basic forms: a “good times” story or a domes-
tic terror threat story. The intention of the good times story is to induce
individuals to reflect on some positive indicators around the world and
within the country rather than to think about current threats. It there-
fore serves as one potential baseline to examine: threat conditions com-
pared to a context of relative well-being. Our other baseline for com-
parison is the control group. The information presented was drawn
from actual sources but edited together by the authors.
The good times treatment began with a statement that the country
is “headed toward a time of increased well-being.” It referred to pos-
itive trends in areas such as education, the environment, and health
in the country, as well in the world. The first paragraph ended with a
note that, according to a recent survey, a “majority” of citizens of that
country report “moderate to high levels of life satisfaction.” The next
four paragraphs focused on positive information about education, the
environment, advances in science (e.g., energy use), and general health
and welfare. The fi nal paragraph read as follows: “In global surveys
of happiness, most countries surveyed have experienced an increase in
happiness over time. High levels of life satisfaction in Peru and around
the globe make sense when viewed from the perspective of these and
other indicators of well-being.”33
The domestic terror threat news story was modeled after those ap-
plied in our prior research (see Merolla and Zechmeister 2009). The
fi rst paragraph referenced warnings that the country is “on the brink
of experiencing a major terrorist attack” by domestic terrorist groups.
The last sentence of that fi rst paragraph indicated that a “majority”
of citizens of that country “are somewhat or very worried about the
possibility of a violent terrorist attack.” The next paragraph talked
about the increase in violence across regions in the country, and spe-

Soifer_6844-final.indb 269 8/17/18 11:53 AM

 270 A. Maldonado, J. L. Merolla, and E. J. Zechmeister

cifically referenced an attack in 2010 by Sendero Luminoso in Auca-

yacu. The third paragraph referenced the intention of domestic groups
to continue to mount coordinated, lethal attacks on citizens in vari-
ous public areas. The fourth paragraph referenced the risk of biolog-
ical and chemical weapons, and a report of a terrorist group working
with the materials required to make ricin. The fifth paragraph refer-
enced a statement by a public official about the lethal intentions of ter-
rorists: “Terrorists still have innocent people in their sights and the
will to murder them. They are always working on the next attack, re-
fi ning their methods.”34

Main Experiment Results

We make use of two of the same measures that we analyzed in the sur-
vey analysis section: Anti–Iron Fist and Abide by Rule of Law. In the
latter case we asked specifically about catching terrorists rather than
criminals. The measures are coded such that a score of one is anti–iron
fist and thinking that police should abide by the law, respectively, and
zero otherwise.
Figure 10.7 presents the proportion of individuals giving the demo-
cratic response across experimental conditions. The proportion of in-
dividuals giving the anti–iron fist response is 56 percent in the con-
trol group, 49 percent in the good times condition, and 58 percent in
the domestic terror condition. We therefore see essentially no differ-
ence between the control and domestic terror group. If anything, we
see a rally around democracy when comparing those in the domestic
terror condition to the good times condition, though the difference be-
tween these two conditions is outside of conventional significance lev-
els (p = 0.15). On the second measure, we do not observe any signifi-
cant differences across conditions. In each case, roughly two-thirds of
respondents think that police should abide by the law when catching
terrorists.
We also included several other indicators in the experiment on sup-
port for democratic values. In a factor analysis of these measures,
we identified a factor that captures support for Democratic Security
Values. The questions that loaded highly on this factor include (fac-
tor loadings in parentheses): if the police suspect that drugs, guns, or
other criminal evidence is hidden in someone’s house, they should be
allowed to enter without fi rst obtaining a search warrant (0.77); if a
person is suspected of a serious crime, the police should hold him in

Soifer_6844-final.indb 270 8/17/18 11:53 AM

 Public Opinion, Violence, and Democracy in Peru 271

Figure 10.7. Proportion preferring anti–iron fist and police abide by law, by
condition.

Figure 10.8. Mean support for democratic security values by condition.

jail until they can get enough evidence to charge him (0.81); and, it is
a good idea for the government to keep a list of people who take part
in demonstrations (0.60). Higher values on the factor indicate more
disagreement and thus greater expression of liberal democratic values.
In figure 10.8, we show the mean on the factor by experimental con-
dition. Mean support for Democratic Security Values is higher in the
control group (0.138) than in the domestic terror condition(-0.095),
and this difference is statistically significant (p = 0.058). Somewhat
surprisingly, mean support for democratic security values is also low

Soifer_6844-final.indb 271 8/17/18 11:53 AM

 272 A. Maldonado, J. L. Merolla, and E. J. Zechmeister

in the good times condition (–0.056), and there are no significant dif-
ferences between this condition and the domestic terror condition.
As with the survey analyses, we also assess attitudes related to po-
litical tolerance. In this case, to measure Political Tolerance, we rely
on the method developed by Sullivan, Pierson, and Marcus (1982), in
which respondents are fi rst shown a list of groups and asked to indi-
cate which group they like the least (they can also fi ll in their own re-
sponse). They are then asked the extent to which they agree or disagree
on a seven-point scale that the group should: be allowed to protest
against the government; be banned from running for political office;
have their phones tapped by government. We recoded the measures
such that higher values reflect more political tolerance, and figure 10.9
shows mean tolerance levels by experimental condition. We see that
mean support for letting one’s most disliked group protest is higher
in the domestic terror condition (3.59) compared to both the control
group (3.36) and good times condition (3.38), though these differences
are not statistically significant. With respect to letting the group run
for office, we again fi nd no significant differences across experimen-
tal conditions, though on this measure mean support is lowest among
those in the domestic terror condition (3.85). Finally, with respect to
phone tapping, opposition is highest in the control group (4.89) and
lowest in the domestic terror group (4.50), though this difference is
outside of conventional significance levels (p = 0.195). Considering all
three measures, in two of them there is a tendency for political toler-
ance to be lower in the domestic terrorism condition, though the ef-
fects do not reach statistical significance.
Overall, when considering the effect of exposure to news about
terrorist threats on democratic public opinion in Peru, the pattern of
results differs from what we observed in the survey data. In the ex-
periment, some results trend in a negative direction (that is, terrorist
threats appear to lower democratic values in comparison to the con-
trol, good times condition, or both), but the only significant negative
effect is with respect to democratic security values and its difference
from the control condition. Weak main effects in the experimental
data could be driven by one of several factors. First, public opinion in
Peru may in general be relatively inoculated to information about ter-
rorist threats, perhaps because such information has already saturated
the population (recall the comparatively high levels of worry about ter-
rorist attacks we fi nd in Peru relative to other countries). Second, in-
consistency in the pattern of results we fi nd in the survey and experi-

Soifer_6844-final.indb 272 8/17/18 11:53 AM

 Public Opinion, Violence, and Democracy in Peru 273

Figure 10.9. Mean political tolerance by condition.

ment data analyses could be driven by the different samples across the
two studies as well as sample size. Since the experiment is based on a
sample in a major metropolitan area, the sample is likely higher in so-
cioeconomic status compared to the national sample used by LAPOP.
Among such a group, we may fi nd in particular that the negative ef-
fects of threat register only among certain subgroups in the popula-
tion. Thus, in the next section, we explore whether the domestic ter-
ror treatment has differential effects among those whose sympathies
lie with Fujimori.

The Moderating Effects of Fujimori Support

Earlier, we identified two potential ways in which individuals in Peru

may react to the threat of domestic terrorism. On the one hand, given
historical experiences with violence in the country, they may resist
democratic decay (or they could rally around democracy) when domes-
tic terrorism is made salient. On the other hand, we may fi nd a neg-
ative effect of conditions of domestic threat on democratic attitudes.
Which pattern of results we fi nd may depend on one’s political dispo-
sitions. We expect that the negative effects of domestic terrorism on
support for democratic attitudes may be more likely among those with
sympathy toward Fujimori, while a rally around democracy would be
likely among those with antipathy toward Fujimori.

Soifer_6844-final.indb 273 8/17/18 11:53 AM

Soifer_6844-final.indb 274

Table 10.2. Effect of treatments and Fujimori proxy on support for democracy measures

Security Support
Anti–iron fist Abide by law values protest Run for office No phone tap

Coef. Coef. Coef. Coef. Coef. Coef.

(S.E.) (S.E.) (S.E.) (S.E.) (S.E.) (S.E.)
Good times −0.014 −0.213 −0.182 0.288 0.399 −0.314
(0.193) (0.203) (0.157) (0.391) (0.422) (0.387)
Domestic terror 0.386** −0.219 −0.330** 0.597* 0.358 −0.049
(0.190) (0.197) (0.153) (0.364) (0.395) (0.363)
Fujimori supporter 0.184 −0.443** −0.233 0.899** 0.847* 0.509
(0.216) (0.218) (0.169) (0.455) (0.488) (0.447)
GT*Fuji −0.214 0.432 −0.002 −0.778 −0.700 −0.115
(0.297) (0.306) (0.238) (0.629) (0.672) (0.614)
DT*Fuji −0.729** 0.481 0.276 −1.108* −1.466** −1.042*
(0.302) (0.310) (0.246) (0.653) (0.697) (0.637)
Constant 0.044 0.611** 0.232** 3.045** 3.667** 4.708**
(0.136) (0.146) (0.111) (0.270) (0.293) (0.269)
N 445 441 423 317 323 321
R-squared 0.015 0.007 0.018 0.015 0.017 0.017

Note: Given that Anti–Iron Fist and Abide by the Law are dichotomous, we use probit analysis. We use OLS for the other measures.
*p ≤ .10 (two-tailed)
**p ≤ .05
8/17/18 11:53 AM
 Public Opinion, Violence, and Democracy in Peru 275

We were not able to test for this relationship in the survey data
given the lack of available measures. While we did not ask respon-
dents in the experiment to evaluate Alberto Fujimori, we do have a few
proxy measures with which to explore these expectations. First, we
asked respondents to report their party identification, and fujimorista
(Fuerza 2011/Partido Fujimorista) was one of the options. Second, we
asked participants to rate Keiko Fujimori, the daughter of Alberto Fu-
jimori and leader of the right-wing Fuerza 2011 Party on a ten-point
feeling scale.35 We combine these two measures to create a dichoto-
mous variable in which a one indicates that the respondent identifies as
a fujimorista and/or has warm feelings toward Keiko Fujimori (6–10
on the scale). About 41.38 percent of the sample either identified as a
fujimorista and/or registered warm feelings toward Keiko Fujimori.36
In the discussion of the results, we refer to respondents as either Fuji-
mori supporters or nonsupporters.
To test whether our Fujimori proxy variable moderates reactions
to the domestic terror condition, we regress each of our dependent
variables on dummy variables for the treatment conditions (with the
control group as the baseline), the Fujimori proxy measure, and in-
teractions between the treatment and the Fujimori proxy measure. A
significant p-value on the interaction term indicates that there is sup-
port for a moderating relationship (Kam and Franzese Jr. 2007). The
results across all of the dependent variables are depicted in table 10.2.
We fi nd that the Fujimori support proxy has a significant moderat-
ing effect on the domestic terror condition for four out of the six mea-
sures: Anti–Iron Fist and the three Political Tolerance measures. We
do not fi nd evidence of a moderating relationship for Abide by Rule
of Law or for Democratic Security Values. For the latter, the domestic
terror condition has a significant main effect only of reducing support
for democratic security values relative to the control group.
Since interaction terms are not directly interpretable, we calcu-
late the effect of the domestic terror condition (relative to the control
group) for Fujimori supporters and nonsupporters. With respect to
anti–iron fist attitudes, we fi nd that those who do not support Fujimori
in the domestic terrorism condition are more likely to select the anti–
iron fist option relative to their counterparts in the control group, and
this effect is statistically significant (p = 0.042). This is evidence of a
rally around democracy for this group. In contrast, Fujimori support-
ers in the domestic terror condition are more likely to fall into the iron
fist category relative to their counterparts in the control group, though

Soifer_6844-final.indb 275 8/17/18 11:53 AM

 276 A. Maldonado, J. L. Merolla, and E. J. Zechmeister

Figure 10.10. Effect of the domestic terrorism condition relative to the

control group, by Fujimori support proxy.

this effect is just outside of conventional significance levels (p = 0.14).

The general pattern of these results supports our expectations for the
two groups.
In figure 10.10, we graph the coefficient on the domestic terror con-
dition (relative to the control group) for supporters and nonsupport-
ers of Fujimori on the tolerance measures. With respect to the protest
measure, we again see a rally around democracy among those who do
not support Fujimori when exposed to the domestic terror condition.
They become just over a half unit, 0.597, more supportive of allow-
ing their disliked group to protest relative to their counterparts in the
control group, and the effect is statistically significant (p = 0.10). For
Fujimori supporters, we instead see a negative effect of the domestic
terror condition relative to the control group (–0.511), though the ef-
fect is not statistically significant (p = 0.347). For the running for of-
fice measure, we again see a positive effect of the domestic terror con-
dition among those who do not support Fujimori, though the effect is
not significant (p = 0.366). We fi nd a negative effect of the domestic
terror condition among Fujimori supporters, who become more than a
full unit less tolerant (–1.107), and this effect is statistically significant
(p = 0.055). We fi nd the same basic pattern for the question on oppo-
sition to tapping the phones of one’s disliked group. There is no signif-

Soifer_6844-final.indb 276 8/17/18 11:53 AM

 Public Opinion, Violence, and Democracy in Peru 277

icant effect for the domestic terror condition among those who do not
support Fujimori, but those who do support Fujimori become over a
full unit (–1.091) less tolerant compared to their counterparts in the
control group, and this effect is significant (p = 0.038).
In sum, on a subset of the democratic values indicators assessed
here, we fi nd evidence of the negative effects of increasing the salience
of domestic terrorism among those who identify as fujimoristas and/or
have warm feelings toward Keiko Fujimori. Meanwhile, those who do
not identify as fujimoristas and/or have cool feelings toward Keiko Fu-
jimori display a tendency to rally around democracy when exposed to
the domestic terror condition, at least with respect to becoming more
opposed to iron-fisted leadership and more supportive of the rights of
their disliked group to protest.

Conclusion

What is the state of public opinion with respect to the threat of vio-
lence and support for democracy in contemporary, post-confl ict Peru?
We fi nd, fi rst, that concerns about the potential for terroristic acts
are more elevated in Peru than in the average Latin American coun-
try. This comports with survey data that documents that Peruvians
tend to be very aware of the country’s recent history with Sendero Lu-
minoso and other groups associated with domestic terrorism cam-
paigns. While the issue is no longer as relevant to assessing the job
performance of the incumbent president as it was during the Fujimori
era, Peruvians remain aware of and concerned about terrorist groups.
While we suspect that the relatively higher concern about terrorism
in Peru is driven in part by continued violent incidents in recent years
(though they have been limited in scope and number), we also suspect
that some of this fi nding may be driven by what anthropologists and
others would refer to as “collective memory” of the confl ict era. One
way or another (either because of continued, if infrequent, incidents
and/or a diffusion of memory), one legacy of the confl ict era is height-
ened concern about the possibility of new terrorist attacks in contem-
porary Peru.
Perhaps in part because of the relatively high mean (and low varia-
tion) in worry about violent attacks by terrorists, we do not find many
significant predictors of these concerns. Yet we do fi nd that those who
were young during the peak of the Sendero period tend to be more

Soifer_6844-final.indb 277 8/17/18 11:53 AM

 278 A. Maldonado, J. L. Merolla, and E. J. Zechmeister

worried than the current young cohort and, at the same time, that
those in regions that were more affected tend to be less worried than
those from regions that saw little violence. This suggests important
heterogeneity in considerations of threat by individuals both within
and across regions in Peru. Contrary to those who think that citizens
in Peru are involved in a “pact of silence” about tragic events in those
years (see discussions in del Pino 2008 and Rénique 2012), we suggest
that elevated levels of worry about violence indicate the presence of a
memory of the confl ict in public opinion in Peru, but one that has dif-
fused in complex ways across regions and subgroups.
Second, we fi nd that there is a tendency in the mass public for con-
cerns about terrorist attacks to be associated with greater preference
for iron-fi rst rule and lower levels of political tolerance. This associ-
ation fits with the general theoretical framework that we have pre-
sented in other work on the threat of terrorism and public opinion,
which suggests that individuals tend to favor centralization of power
in a strong executive and to become less tolerant in the face of terror-
ism (Merolla and Zechmeister 2009). Interestingly, we fi nd that atti-
tudes about rule of law (specifically, the police crossing the line) are
not connected to terrorist threats in the same way; rather, in this case,
those concerned about terrorism are more inclined to prefer that the
police respect the rule of law. It could be that the association between
the war on terrorism and human rights abuses by Peru’s security forces
has left those worried about violence resistant to a solution that sets
the police loose in confronting the threat.
Third, we fi nd that exposure to a news story about domestic ter-
rorist threats in Peru has only minor direct effects on public opinion
in the capital, only decreasing support for democratic values related to
security. It may be that the relative salience of the topic, spurred on by
occasional news stories in Peru, creates a situation in which the public
is relatively inoculated against exposure to information about terror-
ist threats. That is, reactions may have preceded our treatment in this
saturated environment, and thus little direct effect is found (Druck-
man and Leeper 2012). The survey data evidence is in line with this in-
terpretation. Results indicate that less-affected regions, such as Lima,
are where citizens express elevated levels of worry in comparison with
more-affected regions. Of course, it may also be that few significant
effects are found because of the particular sample and its relatively
small size. When we drill down another level in the experiment data,
we do fi nd that those who fall in line with Fujimori react differently

Soifer_6844-final.indb 278 8/17/18 11:53 AM

 Public Opinion, Violence, and Democracy in Peru 279

to news about terrorist threats in comparison to those who are not

supportive of the Fujimori political camp. That is, we fi nd that expo-
sure to news about terrorist threats tends to lead those who are anti-
Fujimori to express more democratic attitudes on several measures
(anti–iron fist and one political tolerance measure), while the reverse is
the case for fujimorismo proponents (across almost all of the tolerance
measures). This provides strong evidence of an important public opin-
ion divide in contemporary Peru, with some who favor addressing ter-
rorist threats through more authoritarian measures and others advo-
cating the opposite.
We conclude that Peru’s experience with Sendero Luminoso and
other violent groups, coupled with the democratic decay that came to
characterize the government’s war against terrorism, has left impor-
tant and unique imprints on public opinion. Public concerns are com-
paratively heightened, though more so among particular groups and
in particular regions than others. Those who are more worried display
relatively greater tendencies toward favoring iron-fisted rule and less
political tolerance, but also more respect for the rule of law. And ex-
posure to information about terrorist threats leads to distinct reactions
among those in the pro- and anti-Fujimori camps.

Appendix: Compliance and Manipulation Checks

We included two types of questions that allowed for compliance and

manipulation checks on the experimental design: a set of questions fo-
cused on information about the news stories and another set focused
on emotions.
We fi rst wanted to measure the extent to which individuals were
paying attention to the newspaper articles. We therefore included ques-
tions about facts from the news stories that were presented. Following
presentation of the news stories, respondents in the treated conditions
were asked two close-ended questions about the articles. During this
time, they were offered the chance to return to the article as needed.
The fi rst question asked whether, according to the news story, more or
less than half of citizens are moderately to highly satisfied with their
lives (for the good times condition) or are worried about the possibility
of a terrorist attack (the threat condition). The correct answer is more
than half. The second question asked whether, according to the news
story, global air quality has improved or deteriorated in the past de-

Soifer_6844-final.indb 279 8/17/18 11:53 AM

 280 A. Maldonado, J. L. Merolla, and E. J. Zechmeister

cade (for the good times condition) or about the location of a terrorist
attack referenced in the news story (for the threat condition). The cor-
rect answer for the former is improved; for the latter, respondents were
asked to select between two choices, one correct and one incorrect. We
then created an Information Acquisition score (0, 1, or 2) for each in-
dividual by tallying the number of correct responses they gave to the
two questions we asked following exposure to the treatment. A mean
of 2 would indicate perfect compliance with the treatment. We fi nd
much higher compliance among those exposed to the domestic terror
condition (mean = 1.47) compared to those in the good times condi-
tion (mean = 0.82).37 That recall was higher for the threat conditions
is consistent with research suggesting that negative events and threats
elicit heightened attention and information-seeking (see discussion in
Merolla and Zechmeister forthcoming); however, it could also be that
respondents resisted the information in the good times condition.
To assess the extent to which the treatments successfully induced
the expected emotional reactions (more negative for those in the threat
condition versus more positive for those in the good times and con-
trol conditions), we examine a set of emotions questions. Following
the treatments (or not, in the case of the control condition), respon-
dents were presented with a series of emotions and asked, for each
one, to “indicate to what extent you are feeling this way right now”
on a five-point scale. Since we needed to use a reduced battery given
space constraints, we followed the advice of Marcus et al. (2006) and
used the markers Watson identified as most reliable: Afraid, Anxious,
Worried, Enthusiastic, Hopeful, Proud, Hatred, Contempt, Bitterness,
and Resentful. We performed a principal components factor analysis
on the ten questions and ran differences in means tests between the
terror threat condition and the control and good times condition for
each factor. We fi nd three factors with an eigenvalue over 1, one on
which positive emotions load highly, one on which emotions related to
fear load highly (Afraid, Anxious, Worried), and one on which emo-
tions related to anger load highly (Hatred, Contempt, Bitterness, and
Resentful).
As expected, we fi nd that those in the domestic terrorism condi-
tion are angrier (mean = 0.105) and more fearful (mean = 0.082) than
those in the control group (mean for anger = 0.009; mean for fear =
–0.059) and good times condition (mean for anger = −0.116; mean for
fear = –0.025).38 Furthermore, they are less positive (mean = −0.076)
than individuals in the control group (mean = 0.074; p = 0.10, one-

Soifer_6844-final.indb 280 8/17/18 11:53 AM

 Public Opinion, Violence, and Democracy in Peru 281

tailed) and good times condition (mean = 0.003; p = 0.26, one-tailed).

These results affi rm the effectiveness of the treatments in inducing neg-
ative emotions in the threat condition relative to the control group or
good times conditions.

Notes

An earlier version of this chapter was presented as a paper at the Legacies of

Political Violence in Contemporary Peru Conference, May 19–20, 2014, Da-
vid Rockefeller Center for Latin American Studies (DRCLA), Harvard Uni-
versity. We are grateful for feedback from the participants, the volume editors,
and the anonymous reviewers. Support for some of this project was granted
from the National Science Foundation (Award IDs 0850824 and 0851136).

1. These departments are in the Andean region, with the exception of San
Martín, which is in the Amazon area.
2. Fifty-six percent worked in agricultural activities. Comparing these sta-
tistics to the broader population, the 1993 national census registered that just
29 percent of Peruvians lived in rural areas and 28 percent worked in the agri-
cultural sector. The 1993 national census registered that just 16 percent of the
population spoke Quechua or another indigenous language.
3. All public opinion data in this section come from Ipsos-Perú databases.
We used the “Opinión Data Plus” search engine to look for “terrorismo” and
“violencia política” descriptors and identified information from polls from
January 1999 to August 2013. Some of these polls are based on representative
samples of Lima and some others represent all of Peru. We present in this sec-
tion what we consider the most relevant fi ndings from our review of available
public opinion data.
4. Ipsos Perú, data from a poll representative of Lima of June 1999, ex-
tracted from Opinión Data Plus. Percentage is of those who expressed ap-
proval of Fujimori’s government.
5. Ipsos Perú, data from polls representative of Lima from August 2010 to
April 2011, extracted from Opinión Data Plus. Percentages are of those who
expressed approval of García’s government.
6. Ipsos Perú, data from a national poll of February 2012, extracted from
Opinión Data Plus.
7. Ipsos Perú, data from a national poll in October 2012, extracted from
Opinión Data Plus.
8. Ipsos Perú, data from national poll in September 2012, extracted from
Opinión Data Plus.
9. This topic has been recently salient because of two publications: Los
Rendidos (Agüero 2015) and Memorias de un Soldado Desconocido (Gavi-
lán 2012). These books have received broad media attention, and the aca-

Soifer_6844-final.indb 281 8/17/18 11:53 AM

 282 A. Maldonado, J. L. Merolla, and E. J. Zechmeister

demic community has intensely discussed them (see Saucier 2015; Torres
2015; González Cueva 2015; and Pajuelo et al. 2013).
10. We are grateful to LAPOP and its major supporters (the US Agency for
International Development, the United Nations Development Program, the
Inter-American Development Bank, and Vanderbilt University) for making
the data available. The AmericasBarometer surveys are conducted via face-to-
face surveys in respondents’ homes on the basis of a complex national sample;
our analyses in this chapter account for the complex sample design by use of
Stata’s svy command and weights provided by LAPOP.
11. In calculating regional averages, we weight all countries to an equal
size, a decision that reflects our belief that countries are an appropriate unit of
analysis. An alternative is to weight countries by population, but then mean
values reflect the public opinion of the most populous countries.
12. While the means vary just at the margins, a series of difference of
means tests affi rms that the mean values in the countries that rank higher
than Peru are statistically distinct from that registered in Peru (at p < 0.001).
13. In our analyses individuals aged 18 to 37 are considered part of the
contemporary young cohort and serve as the baseline, those aged 38 to 67 are
part of the affected cohort, and those aged 68 to 87 are part of the older age
cohort.
14. The wealth measure is constructed based on responses to questions
about ownership of household items. It reflects quintiles of wealth. For more
information on this measure, see Córdova 2009.
15. Ideological self-placement is measured on a 1 to 10 scale, where 1
means “left” and 10 means “right.”
16. The education measure captures four cohorts: none, primary, second-
ary, or higher.
17. The data on deaths and disappearances capture individual fatalities.
In each case, the province and district where the criminal acts occurred are
reported. We aggregate data at the province level and input it into the Amer-
icasBarometer data set. As such, individuals in the AmericasBarometer data
set are tagged to the reported number of dead and disappeared people in their
provinces. We do not aggregate and input data at the district level because it
increases missing cases; for individuals in the AmericasBarometer data set for
whom there is no information at the district level on the numbers of dead and
disappeared people, there is information at the province level.
18. We have classified regions according to the number of dead and dis-
appeared people that occurred in a province. We have divided the data into
four groups according to quartiles: provinces with fewer than 10 victims
(21 provinces, 30.6 percent of observations), provinces with 11 to 70 victims
(17 provinces, 19.3 percent of observations), provinces with 71 to 250 vic-
tims (9 provinces, 8.7 percent of observations), and provinces with more than
251 victims (8 provinces, 41.5 percent of observations). Region 1 is the group
with provinces with the least number of victims and serves as the baseline.
In the LAPOP 2010 data, 414 individuals reside in region 1; 261 in region 2;
118 in region 3; and 562 in region 4. Individuals residing in regions not cov-

Soifer_6844-final.indb 282 8/17/18 11:53 AM

 Public Opinion, Violence, and Democracy in Peru 283

ered by the statistics in the Truth Commission are dropped from the analysis
(n = 145).
19. As a check, we logged the variable that gauges number of victims be-
cause it is highly skewed. We ran models using the logged variable, and results
are robust to the fi ndings presented here (that is, they are significant and have
negative signs).
20. We present results from OLS analyses for the sake of easy interpreta-
tion of the results; the results are robust to ordered probit analyses.
21. Often a direct connection is made between terrorism and narco-traf-
ficking and related activities. For example, in the summer of 2015, the pres-
ident’s party introduced a bill in the Peruvian Congress that connected the
contemporary climate of crime victimization, narco-trafficking, and insecu-
rity to a proposal to make “sicariato” (murder by hire) a crime that can be
prosecuted under antiterrorism laws.
22. This discussion draws from our previous work, including Merolla
et al. 2014.
23. At the same time, they did not fi nd a similar effect in their US 2007
study; thus, it is not always the case that terror threat will compel citizens to
be willing to cede extra institutional power to the executive. An important
and growing body of scholarship supporting a link between security issues
and support for democracy focuses on common crime; a number of scholars
have found links between citizens’ feelings of safety regarding neighborhood-
related crime and their opinions about democratic values, processes, and in-
stitutions (e.g., see Fernandez and Kuenzi 2010; Pérez 2003; Malone 2010;
Malone 2013; Merolla, Mezini, and Zechmeister 2013; but also see Ceobano,
Wood, and Ribeiro 2010).
24. Yet note that Peffley, Hutchinson, and Shamir (2015) fi nd negative ef-
fects of terrorism on tolerance in Israel, a context characterized by persistent
terrorist threats.
25. The measures are correlated at Pearson’s 0.71.
26. We select these dependent variables because we have the same or simi-
lar indicators in the experiment data that we present later in this chapter.
27. In the Peru data set, the Cronbach’s alpha score for these four vari-
ables is 0.87.
28. The confidence intervals are set at p < 0.10, two-tailed; the variables
are all scaled 0 to 1. The constants are not depicted in the graphs for the logis-
tic regressions, but they are available from the authors along with all raw out-
put from regressions and other analyses conducted for this chapter.
29. The results are robust to a simple model that regresses the anti–iron fist
measure on the age cohort dummy variables.
30. The results for Worry about Terrorism and Education are robust to
models that include only these individual independent variables, respectively.
31. The Instituto de Opinión Pública at Pontificia Universidad Católica del
Perú carried out the survey from August 18 to August 31, 2012. They inter-
viewed 760 people in Lima in a face-to-face survey using a complex represen-
tative sample of voting-age adults, with selection at the household based on
gender and age quotas that mirror census statistics. The survey has a +/− 5

Soifer_6844-final.indb 283 8/17/18 11:53 AM

 284 A. Maldonado, J. L. Merolla, and E. J. Zechmeister

percent margin of error. The experiment was designed by two of the chapter’s
authors (Merolla and Zechmeister) and its implementation was overseen in
the field by the other author (Maldonado).
32. Random assignment was achieved by numbering the paper question-
naires, randomly drawing those numbers to generate a random order, and
then reordering the questionnaires according to that order prior to going into
the field. We also had conditions related to international terrorism, but we fo-
cus here on the treatments most relevant for this chapter.
33. Contact the authors for the full texts to the news stories.
34. The experiment instrument contained several checks for compliance
and manipulation. We report on these in the appendix.
35. This question was asked post-treatment, but the experimental condi-
tions had no effect on evaluations of Keiko Fujimori.
36. 38.21 percent of the sample registered warm feelings toward Keiko,
while 11.1 percent identified with the fujimorista party option. If we run the
analyses with these measures separated, the pattern of results is the same.
37. The difference between these two conditions is statistically significant
at p < 0.001.
38. The differences between the domestic terror condition and the control
group are significant for fear (p = 0.10, one-tailed) but not significant for an-
ger (p = 0.22, one-tailed). The differences between the domestic terror con-
dition and the good times condition are significant for anger (p = 0.03, one-
tailed) but not significant for fear (p = 0.18, one-tailed).

Soifer_6844-final.indb 284 8/17/18 11:53 AM

 CHAP T ER 11

Contested Memories of the

Peruvian Internal Armed Conflict
Paulo Drinot

As the 2016 presidential elections showed, the Internal Armed Conflict

(IAC) and the contested memories that reflect, and shape, its legacies,
remain at the heart of political contestation in Peru. During the fi rst
round of elections the fujimorista campaign mobilized a familiar nar-
rative of the confl ict and of the role of Alberto Fujimori in the defeat
of the Shining Path (Sendero Luminoso, or SL). The anti-fujimorista
movement, which, as in previous elections, played a key role in the elec-
toral process, countered with its own narrative of the confl ict and chal-
lenged the notion that Fujimori alone was responsible for the defeat of
SL. Instead, it stressed the authoritarian and corrupt nature of the Fu-
jimori regime, warning that a victory for Fujimori’s daughter, Keiko,
would return Peru to its darkest days. The fujimorista campaign re-
plied by accusing its critics, and particularly the members of the left-
wing coalition the Frente Amplio, of being “terrucos” (see Aguirre
2011). As this suggests, contested memories of the IAC are evident in
Peruvian political life today, twenty-five years after the SL leader, Abi-
mael Guzmán, was arrested and the conflict began to unravel. In this
chapter, I argue that these contested memories are one of the key strug-
gles over the legacies of Peru’s Internal Armed Confl ict.
These contested memories are evident in ongoing debates over El
ojo que llora (The Eye That Cries), a monument in a park in central
Lima that memorializes the victims of Peru’s IAC, one of the most em-
blematic post-confl ict legacies. In this chapter, I examine the ways in
which the monument serves as a point of departure for online debate
on Peru’s “time of fear” by studying several cyberfora, particularly
YouTube videos, which operate as websites of memory. I show that El
ojo que llora monument has come to function as a synecdoche (a part

Soifer_6844-final.indb 285 8/17/18 11:53 AM

 286 Paulo Drinot

that stands for the whole) of the Final Report of the 2001 Truth and
Reconciliation Commission (CVR), but also, arguably, as its simula-
crum, a site where the CVR’s report is commemorated but also where
adherence to its principles and recommendations can be manifested
and performed, as well as challenged. Because of this double function
as synecdoche and simulacrum, El ojo que llora has become a privi-
leged site in which ongoing contestation over the IAC, and particu-
larly over how, and indeed if, the IAC should be remembered, takes
place. While some of this contestation occurs at the physical site of El
ojo que llora, much more occurs in other media and fora, not least in
cyberspace.
Both as synecdoche and as simulacrum, El ojo que llora expresses
remarkably well the post-history of the CVR Final Report. Although
it is only one of many monuments and sites of memory in Peru, El ojo
que llora is the most emblematic of all.1 But it is not an “official” site of
memory. It is used by human rights groups and victim-survivor asso-
ciations, but not by government authorities. To the best of my knowl-
edge, no serving or past president has visited the memorial, let alone
used it as the site of an official act. In this sense, the memorial’s fate
since its construction reflects government attitudes toward the CVR
Final Report which express, at best, neglect, and at worst, hostility. 2
In this sense Peru is different from Chile and Argentina, where un-
der Lagos and Bachelet and the Kirchners respectively, the state took
an active, indeed official, if far from uncontroversial or unproblem-
atic, position in relation to human rights abuses of the past and the
politics of memory (see, for example, Collins 2011; Collins, Hite, and
Joignant 2013; Lessa 2013; Allier Montaño and Crenzel 2015). As
such, the memory battles that are fought in, and over, El ojo que llora
take place against the backdrop of state silence regarding the violent
past. This is not a position of neutrality. It is a position that not only
enables but arguably strengthens memory projects which are explic-
itly anti-CVR. But it is a position made possible because there is lim-
ited demand among the Peruvian population for the CVR’s repertoires
of memory, that is to say, for the range of narratives about, and inter-
pretations of, Peru’s IAC put forward in the Final Report of the CVR.3
In the context of the 2016 elections, the contested memories of the
IAC were thrust into the political arena, as noted. Yet most of the time,
these memories are not fully visible and only occasionally bubble to
the surface. As I will discuss in this chapter, and as I have discussed in
greater detail elsewhere, these contested memories reflect incommen-

Soifer_6844-final.indb 286 8/17/18 11:53 AM

 Contested Memories of the Internal Armed Conflict 287

surable views about the causes of Peru’s internal confl ict, the respon-
sibilities and culpabilities of different actors for the violence, and how
Peruvian society must manage the post-conflict phase (Drinot 2009).
But they are not part of an open or official reckoning with the past. In
part this reflects the fact, as the CVR report showed, that the victims
of the conflict were marginal to Peruvian society: they were predom-
inantly poor, rural, and Quechua-speaking. Despite the efforts of hu-
man rights and victim-survivor groups, these victims had or have to-
day little political leverage, which is not to say that they exist outside
of politics. The weakness of the Peruvian Left, arguably the natural
political channel for these victims’ claims for recognition and repara-
tion, has lowered the visibility of the victims and, more generally, the
visibility of the conflict itself in post-confl ict Peru (see chapter 8 of this
volume).
For this reason, though presidential elections can generate renewed
attention from the general population regarding how the violent past
is remembered, in the absence of either state interest in addressing the
violent past, or political movements or civil society groups with suffi-
cient political leverage to place the unresolved legacies of the violent
past and its victims on the political agenda, the battles for memory in
Peru, with few exceptions, do not translate into concerted state poli-
cies. Nor do they generate sustained interest among the population at
large. And yet, as I will show, these contested memories and the an-
tagonism that they generate in those who engage in Peru’s politics of
memory are clearly essential to understand the broader post-confl ict
legacies of the IAC. The memory struggles that I discuss in relation to
El ojo que llora express not only incommensurable views about the vi-
olent past but also incommensurable views of the present and the fu-
ture that are in turn reflected in the politics of Peru today, not least in
the ways in which contested memories of the IAC are operationalized
in the politics of fujimorismo and anti-fujimorismo.

El ojo que llora and the Memorialization of Peru’s IAC

Although the Cold War shaped the experience of violent confl ict
throughout Latin America, in Peru, unlike in the Southern Cone in the
1960s and 1970s and Central America in the 1980s, it occurred in a
context, the 1980s and 1990s, in which the country was exiting from
a period of dictatorship.4 As such, the Peruvian IAC is relatively sui ge-

Soifer_6844-final.indb 287 8/17/18 11:53 AM

 288 Paulo Drinot

neris within Latin America. The IAC included several armed actors,
including the insurgent groups SL and MRTA, the Peruvian armed
forces, and peasant self-defense committees, which played a decisive
role in the outcome of the war. It was broadly national in scope, al-
though concentrated primarily in the southern Andean region and in
the shanty towns of Lima. However, though it had a nationwide im-
pact, scholars of the IAC have shown that at the local level, and par-
ticularly in Ayacucho, which bore the brunt of the violence, the IAC
was shaped by conflicts both between and within Andean indigenous
communities. These confl icts, between pro-SL and anti-SL commu-
nities, or between pro-SL and anti-SL individuals within communi-
ties, refracted and reproduced in a new, vastly more violent register the
conflicts over land, power, and influence that often antedated the SL
insurgency. Elsewhere, and particularly in Lima, the confl ict took on
other, equally complex, dynamics which scholars are only beginning
to uncover.5
The CVR was set up in 2001 to investigate the causes and nature
of the IAC and to report on human rights violations that took place
in the period from 1980 to 2000. It published its multivolume Final
Report in 2003. The CVR concluded that the Internal Armed Con-
flict had produced a far greater number of victims, some sixty-nine
thousand in total, than had previously been thought. It blamed SL for
the highest percentage of the violence, above that committed by the
armed forces or the self-defense committees, or indeed by the MRTA.6
It also was highly critical of the actions and inactions of the govern-
ments of Fernando Belaunde, Alan García, and Fujimori, of political
parties of the Left and the Right, and of several institutions, including
the Catholic Church. But it also accounted for the violence by framing
it in a sociohistorical analysis that stressed the ways in which Peru’s
deep and intersecting inequalities based on ethnicity, class, and gen-
der were reflected in and reproduced through the violence. The CVR
made several recommendations, including personal and collective rep-
arations and prosecution of cases of human rights abuses, which have
largely been ignored by successive governments (Macher 2014; Huber
and del Pino 2015). However, the arrest, trial, and imprisonment of
former President Alberto Fujimori and of his intelligence chief, Vladi-
miro Montesinos, as a well as of some high-ranking military officers,
on charges of human rights abuses (as well as corruption) in the late
2000s was a significant victory for Peru’s human rights community
(see Burt 2009a and R. Gamarra 2009).

Soifer_6844-final.indb 288 8/17/18 11:53 AM

 Contested Memories of the Internal Armed Conflict 289

The CVR’s report became the object of intense criticism and re-
mains a matter of contention. Some of the criticism, largely wrong-
headed, focused on the methodology used for calculating the number
of victims. Some, mostly from the Left, criticized the CVR’s report
for attributing the largest number of victims to SL and not the armed
forces. However, the bulk of the criticism focused on the CVR’s sup-
posed inherent bias. Even though the composition of the commis-
sioners reflected diverse political opinions, right-wing commentators,
from members of Acción Popular (Belaunde’s party) to former pres-
ident García to newspaper columnists such as Aldo Mariátegui have
accused the CVR of being “caviar,” by which they imply that it merely
reflected left-wing opinion, and even terruco sympathies. Similarly,
even though the armed forces were represented on the commission, the
military, both institutionally and in the personal capacity of several
high-ranking officers, has questioned the legitimacy of the CVR and
produced its own accounts of the confl ict (Milton 2018). The Catholic
hierarchy too, and particularly Cardinal Juan Luis Cipriani, was hos-
tile to the CVR and its report. However, it is within the ranks of fuji-
morismo that the hostility to the CVR is greatest. Fujimoristas have
accused the CVR of being inherently biased against Fujimori and of
whitewashing the crimes of SL and the MRTA.7
Far from providing an account of the IAC around which a consen-
sual memory of the violence and a politics of reconciliation could be
built, as originally intended, the CVR and its Final Report have been
thrust regularly into the very center of the politics of memory and, in-
deed, into the center of politics itself, serving as an issue that is regu-
larly mobilized and operationalized politically by those who support
its fi ndings, primarily the Left and human rights organizations, and
those who oppose it, primarily fujimoristas but also other right-wing
members of what Vergara and Encinas refer to in chapter 9 as a “con-
servative archipelago.” It is, itself, a contested site of memory, as is, in-
deed, the memorial El ojo que llora, which, as I have suggested, has
acted more than any other memory site as both synecdoche and simu-
lacrum of the CVR’s Final Report.8 Designed by the recently deceased
Dutch-born Lima resident artist Lika Mutal, the memorial consists
of a large central rock from which water spurts, surrounded by small
stones arranged in a way that forms paths around the central rock, on
which are engraved the names of over thirty thousand victims of the
violence. The memorial, part of the symbolic reparations envisaged by
the CVR, has been the site of regular commemorations of the publica-

Soifer_6844-final.indb 289 8/17/18 11:53 AM

 290 Paulo Drinot

tion of the Final Report of the CVR and other activities organized by
human rights and victim-survivor organizations.
In late 2006, Peru’s right-wing press criticized the memorial for
including the names of Shining Path members killed during a police
raid on Castro Castro prison, where they were being held (see Aguirre
2013; Feinstein 2014). The controversy was sparked by a ruling issued
by the Inter-American Court of Human Rights, which ordered the Pe-
ruvian state to include the names of these senderistas, now considered
“victims,” in the memorial. However, it soon was discovered that the
names of the senderistas already were featured in the memorial, since
they had been added to a list of names given to Lika Mutal by hu-
man rights organizations. In late September 2007, El ojo que llora me-
morial was again the scene of controversy when it was attacked dur-
ing the night. The attackers threw orange paint on the central rock
and on the stones and battered the monument with a sledgehammer.
The attack followed the decision of the Chilean authorities to extra-
dite Fujimori to Peru to face human rights and corruption charges. In
2000, facing allegations of corruption and malfeasance, Fujimori had
fled the country and taken refuge in Japan. Why he decided to travel to
Chile remains unclear, but his arrest, extradition to Peru, and eventual
trial and sentencing was a landmark process. Fujimoristas continue to
claim that Fujimori’s conviction was a sham. Typically, they blame hu-
man rights organizations and what they call the “caviar” class for his
wrongful imprisonment. Fujimori, they insist, saved Peru from SL, and
that he is a hero, not a criminal.
As this suggests, and as I have argued elsewhere (Drinot 2009), the
attack on and defacement of El ojo que llora, and the debates that sur-
rounded the memorial in 2006 and 2007, reflected the tensions be-
tween two antagonistic and mutually exclusive narratives of the IAC:
a fujimorista “memory of salvation,” and a “human rights memory”
initially mobilized by human rights organizations and later largely re-
produced in the CVR report. These narratives, in turn, are expressive
of different ontologies of violence, i.e., of different interpretations of
(1) the causes of the violence, (2) the responsibilities and culpabilities
of different actors in the confl ict, and (3) how Peruvian society must
manage the post-conflict phase. These narratives of the IAC are not
the only ways Peruvians remember the confl ict, as many scholars have
shown. The range of engagements with the violent past discernible in
Peruvian society cannot be neatly reduced to a human rights reper-
toire of memory and a fujimorista repertoire of memory. Nevertheless,

Soifer_6844-final.indb 290 8/17/18 11:53 AM

 Contested Memories of the Internal Armed Conflict 291

these conflicting memories or narratives of the violent past are rou-

tinely mobilized and operationalized. As such, they shape both how
the violent past is remembered and, in particular, how that past is po-
liticized, as the recent 2016 elections demonstrated. I will discuss this
in greater detail later in this chapter with regard to El ojo que llora’s
cyber projection.
As Cynthia Milton has noted, “the socio-political opening that gave
rise to the CVR also created other spaces for public discussion that
were previously unavailable, thus allowing for alternative media of
‘truth-telling’: for instance, visual and performance art, memory sites,
cinema, stories, humor, rumor and song” (2009, 64–65). The study of
such media as a means to explore how Peruvians make sense of their
recent violent past has produced an expanding and sophisticated lit-
erature. Scholars working in several disciplines have turned their at-
tention to cultural artifacts such as novels, films, photography, plays,
comics, song and other music forms, art, and testimony in order to ac-
cess and analyze the construction of personal and collective memo-
ries of the violent past.9 Such studies reveal the broad range of mem-
ory practices that Peruvians engage in, often at the local level and in
highly specific contexts, as well as the distinctive memory politics that
they participate in through such practices. I stress, however, that these
practices are not reducible to either the memory of salvation or the hu-
man rights memory, but are nevertheless informed and in some ways
framed by these repertoires of memory with hegemonic pretensions.
Cyberspace offers another medium through which to explore how
such memory practices are engendered and deployed.
Scholars increasingly recognize the potential of the Internet as a
medium through which to study what Elizabeth Jelin (2003) has called
the labors of memory. Students of memory are well aware that what
Wulf Kansteiner has called “media of memory” (2002, 195) not only
transmit memory but actually construct it: “all media of memory, es-
pecially electronic media, neither simply reflect nor determine col-
lective memory but are inextricably involved in its construction and
evolution.” Increasingly, therefore, scholars interested in how mem-
ory operates on the Internet have turned to examining cyberfora such
as personal and institutional websites as well as digital artifacts such
as Facebook, Flickr, and YouTube (see Bhattacharya 2010; van Dijck
2011; Knudsen and Stage 2013; Ferron and Massa 2014). In an arti-
cle published in 2011, I examined this potential in relation to Peruvian
and Chilean memories of the War of the Pacific (1879–1884) by study-

Soifer_6844-final.indb 291 8/17/18 11:53 AM

 292 Paulo Drinot

ing the comments attached to YouTube uploads of sections of a Chil-

ean documentary on the war (Drinot 2011). However, so far, there
have been few attempts to draw on the Internet within the broader,
and largely productive, effort to explore the circulation and constitu-
tion of collective memories of the IAC.10

CVR Memories

In this section and the next, I analyze a dozen or so videos on You-

tube and one on Vimeo. These videos vary in several ways. Most are
short, around two to three minutes, although one is over six minutes
long. They fall into three broad categories: (1) professional or semi-
professional productions, (2) amateur productions, and (3) reproduc-
tions of news items. Among the amateur videos, two stand out as at-
tempts to present an artistic proposal. Of those that are not simply
reproductions of news items recorded from television, several include
specially selected music; some “Muzak,” often poignant violins or pi-
ano; some “Andean” music (huaynos); in two cases Silvio Rodríguez
tunes; and in one case the music from the film Schindler’s List, scored
by John Williams. One video includes a live performance. A few vid-
eos are, in effect, a succession of still images overlaid with text, music,
or a voiceover. In most cases, however, video dominates. Some videos
have generated no comments at all, while others have as many as sixty
or ninety comments attached to them. These videos were uploaded at
different times. The earliest dates from May 2007, and the most re-
cent from September 2012 (at the time the research was undertaken).
Several, however, were uploaded shortly after the attack on El ojo que
llora in September 2007.
The dominant narrative in these videos is broadly consonant with
the CVR’s narrative of the confl ict and post-confl ict. This is clear-
est, perhaps not surprisingly, in the video uploaded by “Programa lo
justo,” an online program produced by the Coordinadora Nacional
de Derechos Humanos, a human rights NGO, fronted at the time by
Rocío Silva Santisteban, a human rights campaigner.11 The video fo-
cuses on the events organized at El ojo que llora in commemoration of
the eighth anniversary of the publication of the CVR’s Final Report. It
starts with the camera flying over the stones with the names of the vic-
tims of the violence. The music is poignant (ethereal strings) and we
hear the voice of a woman (she is almost crying as she speaks): “the

Soifer_6844-final.indb 292 8/17/18 11:53 AM

 Contested Memories of the Internal Armed Conflict 293

names of more than 15,000 disappeared [. . .] who do not have the

right to fi nd [. . .] our loved ones.” It becomes clear that the speaker is
a relative of one or more of the disappeared. The camera then shows
children at the monument, looking at the stones and laying hands
made of paper which carry written messages (“we must forget what
happened and must continue building collective memory”) next to a
floral arrangement in the shape of Peru. The camera fi nally cuts to the
speaker, who is seen addressing schoolchildren. She is telling them:
“You must have another future, you must build a nation, a dignified
citizenship, with rights, with opportunities, with equality, and fight in-
equality and corruption, the discrimination that occurs in every cor-
ner of our country.”
The camera then cuts to a middle-aged woman who says she is
moved to see young people engaging with human rights issues. She
suggests that human rights should be included in the school curricu-
lum. Then, after showing children standing around and further images
that suggest the didactic activities that have been undertaken on that
day (sheets of paper with texts such as “What does the state do?”), the
camera cuts to interviews with several schoolchildren. The message, it
soon becomes clear, has gotten through to them: “It gives us a lot to
think about,” says one; “it is moving, and it motivates us so that in the
future we will be able to answer for all of them (podamos responder
por todos ellos) and improve the situation.” Another says: “It is right
that they honour them [. . .] also so that the young can learn what hap-
pened during the twenty years of terrorism.” Yet another: “It encour-
ages us to avoid this happening again and to not forget our past.” Yet
another: “We are today the future of the nation, and therefore we are
the new actors who must make sure that this does not happen again.”
In other words, remembering is an act that allows us to avoid repe-
tition of the past and to construct a better tomorrow. This video ex-
presses perfectly how memory work services the CVR project: mem-
ory is mobilized to instruct through didactic devices but also to move
through a series of affective devices. El ojo que llora is instrumental to
this didactic and affective mobilization of memory.
The Advocacy Project, an NGO based in Washington, DC, pro-
duced and uploaded a video which similarly emphasizes the nature of
El ojo que llora as an active site of memory.12 This video combines
moving music (Silvio Rodríguez but also Andean huaynos), superim-
posed text, and an interview with Renzo Aroni, a historian who works
with the Equipo Peruano de Antropología Forense (EPAF). Aroni is

Soifer_6844-final.indb 293 8/17/18 11:53 AM

 294 Paulo Drinot

seen in the fi nal part of the video playing guitar and singing a song
in Quechua. The video was shot on August 23, 2010, the seventh an-
niversary of the publication of the CVR report. On this day, relatives
of victims added new stones with the names of their loved ones to the
monument. We see Gisela Ortiz, “a relative of the Cantuta case,” ac-
cording to a caption, who tells those who have congregated at El ojo
que llora: “Today we are again present in this space of memory in
homage to each one of our relatives, in homage to the thousands of
disappeared Peruvians, cruelly assassinated during the years of polit-
ical violence and today we do not only remember them [. . .] we also
reaffi rm the hope that the day will arrive when injustice is defeated
by justice in our country.” Another speaker, identified by a caption
as Rocío Paz Ruiz of APRODEH, a Peruvian human rights organiza-
tion, stresses the hope that “fi nally it will be possible in our country to
achieve justice, to preserve the memory of what happened so that it is
not repeated.” The video then shifts to the act of commemoration it-
self, which involves placing additional stones engraved with the names
of victims on the monument.
Another video by the Advocacy Project similarly stresses the func-
tion of El ojo que llora as an active site of memory in the context of
All Saints Day or the Day of the Dead commemorations in November
2011.13 Titled “Remembering the Peruvian Disappeared on the Day
of the Dead,” the video, which at over six minutes long is one of the
longest studied, features several interviews fi lmed at El ojo que llora
with relatives of the disappeared. Some of the interviews are in Span-
ish, others in Quechua. The video also includes white text on a black
background to inform viewers that the IAC resulted in around sev-
enty thousand dead, of which some fifteen thousand were disappeared
and are still unaccounted for. Through the text and the interviews, the
video emphasizes the importance of El ojo que llora as a site where
the memory of the disappeared can be kept alive as part of a proj-
ect of redress, collective and public reparations, and a quest for jus-
tice. It is a site from which civil society, with the help of advocacy
groups, can pressure the government to act. Enrique Pólido Espinoza
from Colcabamba, Ayacucho, one of the interviewed, tells the camera
that his relative is buried in Putaccasa and that he is requesting that
the body be exhumed: “my relative has been abandoned on a moun-
tain as if he were an animal or as if he was worthless.”14 The cam-
era then turns to Gisela Ortiz, now identified as “Representative of
La Cantuta Relatives,” who speaks to the dead: “Today we are here in

Soifer_6844-final.indb 294 8/17/18 11:53 AM

 Contested Memories of the Internal Armed Conflict 295

our Ojo que llora, in this place of memory, to remember each one of
the tears that we have shed for you, the pain of not having you here,
the anxiety of not knowing, the long absence, the compromised si-
lences, the impunity, and today, here, we tell you that there will be no
forgetting, no impunity, no one will impose forgiveness without jus-
tice” (my emphasis).15
The video emphasizes the importance of El ojo que llora as a place
in which the remembrance of the disappeared can be performed, and
therefore a place in which their status as absent or missing can be con-
fi rmed or made real (this is also, more specifically, the function of the
pebbles inscribed with the names of victims). But it also emphasizes
the monument’s role as a place from which demands for justice can be
made: Luis Arones, from Raccana, Ayacucho, says into the camera:
“It is a special moment to remember our relatives, our friends, who
are victims of political violence [. . .] It is nostalgia, remembrance, it
hurts in our souls, and I hope that soon this will end and that the gov-
ernment will consider the petitions we are making.” Similarly, Emilia
Auccasi Julian from Sacsamarca, Ayacucho, speaking in Quechua, tells
viewers: “I’ve come to Lima in search of justice, to claim my rights all
the way from a forgotten village. I want to thank EPAF, which has
come to our village and now has brought us here to demand our rights
and make sure they are respected.” The video concludes with a se-
quence of white text on a black background set to a score of Andean
panpipe music: “There are over 4,500 known mass graves dotting the
Peruvian countryside, most of them in the region of Ayacucho / Peru
still does not have a state policy for the search and identification of its
more than 15,000 disappeared / Neither has it signed the UN Interna-
tional Convention on Enforced Disappearances.”
These videos present El ojo que llora as a site at which the report
of the CVR can be not only commemorated but also restated and in-
deed activated as both synecdoche and simulacrum by what Elizabeth
Jelin has called memory entrepreneurs—a particular kind, perhaps, of
professional memory entrepreneur, as I suggest in the next section.16
These memory entrepreneurs’ statements and performances, through
a combination of didactic and affective strategies, mobilize particular
memories of the IAC that are consonant with the analysis put forward
by the report of the CVR. As such, these videos express the success of
the CVR in establishing a repertoire of memory that human rights and
advocacy groups, as well as victim-survivor groups such as COFADER
(Comité de Familiares de Detenidos Desaparecidos) can, and do, draw

Soifer_6844-final.indb 295 8/17/18 11:53 AM

 296 Paulo Drinot

from in their campaigns.17 However, they also express the failure of

the CVR to achieve much beyond establishing this repertoire. After
all, these yearly acts of remembrance and claims-making on the part
of human rights groups and victim-survivor associations also demon-
strate that the Peruvian government has largely ignored and failed to
enact the recommendations of the Final Report of the CVR, at least
until very recently, in regard to individual reparations.18

Amauteur Memories

In addition to these professional or semi-professional videos put to-

gether by human rights and advocacy groups, several amateur, or per-
haps amauteur, videos on El ojo que llora can be found on YouTube
and Vimeo. These videos present their produsers’ understandings of,
and interventions in, the memory politics of the memorial (see, among
others, Grinnell 2009). As such, they constitute a particularly interest-
ing aspect of the battles for memory over El ojo que llora. They reflect
the participation of nonprofessional memory entrepreneurs who oper-
ate independently of, indeed at the margins of, the human rights and
advocacy groups whose memory entrepreneurship is part of a broader
human rights and social activist agenda. For the amauteur memory
entrepreneurs, such videos are personal contributions to a broader de-
bate. Though some have artistic ambitions, these memory entrepre-
neurs are not, properly speaking, activist-artists like Lika Mutal, the
artist who designed El ojo que llora; the theater group Yuyachkani,
which performs plays that evoke and reflect on the IAC; the retablistas
of Ayacucho who, through their art, intervene in the memory politics
of Peru; or comics artists like Jesús Cossio.19 But their videos illustrate
a demotic, bottom-up participation in the politics of memory in Peru
in cyberspace. They also reveal the ways in which the CVR project has
provided a usable repertoire of memory that extends beyond the hu-
man rights community.
Two of these amauteur videos present explicitly “artistic” treat-
ments of the memorial. Juan Javier Cuadro’s video is a black-and-
white time-lapse which lasts less than a minute. It is accompanied by a
soundscape which includes birdsong, unclear voices, and gunshots fol-
lowed by silence and then the sound of the wind blowing. The video
ends with a quotation in white text on a black background, which
makes reference to the September 2007 attacks on the monument:

Soifer_6844-final.indb 296 8/17/18 11:53 AM

 Contested Memories of the Internal Armed Conflict 297

“‘Controversy is the word that defi nes the monument which is a com-
plaint against violence and that is why it is violated, it is there to re-
mind us of what no one wants to relive but which relives that which
everyone wants to forget’—Carlos García.”20 Similarly, “El ojo que
llora en una Noche en Blanco,” by VictorE50, is a short fi lm that con-
sists of a handheld traveling shot of the stones engraved with names as
the filmmaker walks around the monument. The filmmaker’s shadow
occasionally appears in the shot, as do other visitors to the monument.
These sequences are occasionally cut by stills of what appear to be
people attending a “white night,” or nighttime celebration. The video
maker has uploaded a short description of the video: “The video aims
to introduce the ojo que llora in a white night. As such, we start with-
out sound, taking the dynamism of the white night, in the sections of
the video where the stones do not have engraved names. This way, the
sound is so silent that we can hardly hear the steps, as a way to keep
a moment of silence for the fallen who were not recognized or whose
identities were not known.”21
Beto Serquen’s video, uploaded in May 2007, reflects artistic inten-
tions of a somewhat different kind.22 It starts with about thirty sec-
onds of white text on a black background. The text, in small print, ex-
plains the history of the memorial. The rest of the video consists of a
sequence of photographs (the names of the photographers are listed in
the video credits at the end). The sequence begins with images from
the Qoyllur riti festival, images of the Andes, and images of a group
of indigenous people chewing coca leaves; these images appear to have
been chosen to establish the Andeanness of what is being represented.
The video then switches to images of El ojo que llora memorial, start-
ing with a close-up of the crying stone, and then a series of shots of
different parts of the memorial, followed by images of people at the
memorial, including a shot of Salomón Lerner, one of the CVR com-
missioners. The video’s poignancy is marked primarily by the choice
of music that accompanies the images. The polyphonic choral compo-
sition Hanaq Pacha, sung in Quechua, and believed to date from the
seventeenth century, further stresses the Andeanness of the video. In
this way, it is fair to assume, the video maker has attempted to reflect
through both images and sound one of the key fi ndings of the CVR re-
port; that is, that the Internal Armed Conflict had an overwhelming
impact on Andean Peru and on Peru’s indigenous population. It was a
conflict that impacted the whole of Peru, but it disproportionately af-
fected the indigenous peasantry in the Andes in a way, the CVR report

Soifer_6844-final.indb 297 8/17/18 11:53 AM

 298 Paulo Drinot

suggested, that reflected and reproduced the racialized exclusions that

have characterized Peruvian society since the colonial period.
A video by Ricardo Cuya Vera, titled “Crónica visual,” takes a
far more direct, didactic approach. Uploaded in November 2008, the
video, shot in handheld mode, shows a virtually empty memorial. At
6.42 minutes, the video is one of the longest studied here and includes
a real-time voiceover, which describes El ojo que llora in some detail.
The video maker adopts a portentous form of enunciation in parts of
the video; it is quite likely an attempt to appear to speak from a po-
sition of authority. Although largely a description of the memorial it-
self and how it was built, the video is not mere neutral reportage. The
video maker clearly takes sides in the debates that surround the memo-
rial: “This is not a monument to terrorism; it is a cry for peace,” he de-
clares. 23 Cuya Vera’s voiceover focuses on the September 2007 attack
on the memorial: he explains that it “was attacked, painted and partly
destroyed on 3 September 2007. Orange paint was thrown on it [. . .]
funnily enough this is the color of the party of the Japanese dictator
Alberto Fujimori.” As he walks around the memorial, Cuya Vera reads
out several of the names engraved on the stones, including a whole
group who share the same surname, Baldeón: “There were more than
two Baldéons, there were more than half a dozen, it could have been a
whole a street, a whole village, only history knows this.”
Like Cuya Vera’s video, other videos similarly focus on the attack
on the memorial. In November 2007 a Youtube user identified as
nuovavita2 uploaded two videos which are simply recordings of TV
news reports on the attack. Although the videos are not produced by
nuovavita2, they are clearly used to express a point of view on the at-
tack. This is made clearer still by the title given to one of the uploads,
“Profanación a El Ojo que LLora,” and by a short description of the
video which they have added: “Profanation of El ojo que llora. With
paint, sledgehammers and pickaxes, a group of violent hooligans en-
tered the park where this monument to the dead of the political vio-
lence in our country is located.” Nuovavita2 also adds a comment:
“This issue should be everywhere, not just to talk about it, but to de-
mand immediate and well-thought-out action so that it will no longer
be just ‘a matter for the cholos’ [. . .] only when people in Lima start
dying [. . .] often that was the view of things.”24 Through both the act
of uploading videos captured from the television news reports on the
attack on the memorial and the description and comments, this You-
tube user employs the medium to position him or herself in relation to

Soifer_6844-final.indb 298 8/17/18 11:53 AM

 Contested Memories of the Internal Armed Conflict 299

the debates on El ojo que llora, and, by extension and implication, in

relation to the debates on the CVR.
These amauteur videos are cyber-projections of El ojo que llora:
they restate the memorial’s intended message and function as a synec-
doche and simulacrum of the Final Report of the CVR. But they are
also means through which these produsers of videos position them-
selves in relation to the memorial and its message and function. The
videos therefore are memorials in their own right. Like El ojo que llora
they are a means, a technology perhaps, to mobilize memories conso-
nant with the interpretation of the IAC and the recommendations put
forward in the Final Report of the CVR. They mobilize what I am call-
ing the CVR’s repertoire of memory by employing a combination of
music, video, still photography, and text, all contributing to a narra-
tive and an atmospherics that invoke, and evoke, key arguments in the
CVR’s Final Report, such as the disproportionate impact of the IAC on
the Andean population of the country or the fact that many victims of
the conflict remain unacknowledged; a lack of acknowledgment, more-
over, expressive of the broader conditions that helped bring about the
conflict. More generally, like the professional memory entrepreneurs
examined in the previous section, these amauteur memory entrepre-
neurs mobilize, through their videos, the CVR’s repertoire of mem-
ory of the IAC as a didactic and affective device in their explicit rejec-
tion of the counter-memories, or the fujimorista memory of salvation,
expressed in the attack on the memorial in 2007. In this way, the vid-
eos bear witness to the work of largely unknown, yet clearly commit-
ted, memory entrepreneurs deeply invested in the project of the CVR.

Counter-Memories

The repertoire of memory of the CVR, reflected in El ojo que llora and
projected by the videos, is contested in online debates enabled by the
YouTube comment function. These online debates reflect closely, if in
the particularly abrasive register of online interaction, the broader de-
bates in Peruvian society over how, or indeed whether, the IAC should
be remembered. More specifically, in relation to El ojo que llora me-
morial, the debate focuses on the question of which deaths should be
grieved or are deserving of commemoration.25 For many of those who
contest the narrative of the IAC put forward by the CVR and the ways
in which the memorial understands the category of victim, which in-

Soifer_6844-final.indb 299 8/17/18 11:53 AM

 300 Paulo Drinot

cludes those who were the victims of state violence, El ojo que llora is
a travesty. As Roberto García Armas writes in the comment section
to one video: “I challenge anyone to show me if in the famous ‘monu-
ment’ are the names of the policemen who were ambushed in a cow-
ardly manner in Quebrada Honda Sayapullo, also the names of the
authorities and villagers of the district of Sayapullo who were mur-
dered in a cowardly manner by Sendero Luminoso [. . .] but I’m sure
that we will fi nd the names of the ‘students’ of La Cantuta [Univer-
sity].”26 This hostility toward a memorial that is believed to overlook
and therefore deny the sacrifice of those who fought against and were
the victims of the Shining Path while commemorating those who were
killed by the armed forces as part of a legitimate strategy to defend the
nation from an internal foe is a key leitmotiv of the contest over the
CVR narrative in other video comment sections.
In the comments attached to Ricardo Cuya Vera’s video, for exam-
ple, all four comments are hostile to the memorial: Beatrizz2011 says,
“That would be a good place to take my dog to piss and shit”; David
Panebra says, “THAT MONUMENT IS ONLY VENERATED BY TERRORISTS
[the term used is terrucos]. WHAT INJUSTICE.”27 Several posters criti-
cize the location of the monument and its intended purpose, or, as I
have suggested, its function as both synecdoche and simulacrum of
the Final Report of the CVR. El blanco78 writes: “If that ‘monument’
(?) was a homage to the victims of terrorism, they should have built it
somewhere in Ayacucho where THE MOST AFFECTED lived, not in Je-
sus María [a middle-class district of Lima] WHERE NO ONE WAS AF-
FECTED, the creators of that monument are crazy and the authorities
who gave the authorization [to build the monument] are stupid.”28 Ac-
cording to enrike molinares: “This stupid sculpture which venerates
the terrorist assassins should be moved to La Cantuta [. . .] or some
hill far from the city [. . .] that city which for many years was a tar-
get, a victim of car bombs, blackouts, assassinations, etc, etc.”29 These
somewhat contradictory comments resignify the memorial to the vic-
tims of violence as a monument to terrorism, a claim put forward in
much of the right-wing press and by prominent fujimoristas such as
congresswoman Martha Chávez.30
In the comments attached to a video produced by NAPA 18,
that is targeted at children, we fi nd similar comments. One poster,
XXjUdAs85xX, challenges the idea that the children whose names are
engraved on the stones were innocent as claimed by the narrator in the
video: “Innocent??? I can tell that that pseudo-journalist who writes

Soifer_6844-final.indb 300 8/17/18 11:53 AM

 Contested Memories of the Internal Armed Conflict 301

blogs is younger than 20 years of age.” Another, neko-chan joyjoy, la-

ments the fact that “WE ARE THE ONLY COUNTRY IN THE WORLD THAT
RAISES MONUMENTS TO TERRORISTS.”31 In the comments to another
video, the status of those commemorated by the memorial is called
into question by Troyano2011: “VICTIMS[?] THEY WERE TERRORISTS
FOR FUCK’S SAKE.” Another poster attacks both the relatives of the vic-
tims and the human rights organizations which support them: “Where
were all these crybabies when the terrorists were killing left and right.
Now that they are given money for ‘reparations’ they go on marches
and lie about. Thank God the real Peruvians, the armed forces and
some politicians manned up and eliminated the terrorist fi lth. It’s a
shame that these terrorist sympathizers, these NGOs who help them
and the champagne socialists still proliferate, scoundrels.”32
These comments represent counter-memories of the IAC, that is,
memories counter to those memories elicited by, and in some ways
made possible because of, the CVR Final Report. Like the memories
reflected in the videos I have analyzed, this alternative repertoire of
memory is also mobilized by El ojo que llora memorial. The videos
are therefore perhaps best understood as an “interactive commemo-
rative space” (Knudsen and Stage 2013) where different and opposed
memories of the IAC are mobilized and brought into confrontation
(of course, the memorial itself is an interactive commemorative space).
This is especially evident in the discussions that unfold in the com-
ments sections of some videos, particularly those in which one or more
posters engage in exchanges over several years. Take for example, the
ninety-one comments attached to the video uploaded by Beto Serquen
in May 29, 2007.33 The fi rst comment dates from that year. When this
research was conducted, the most recent dated from 2013. In this ex-
change, although several posters participate, much of the discussion is
led by two posters, ellesar19 and nuphi. These exchanges reflect quite
closely the ways in which the memories of the IAC overlap, and in
many ways come into conflict, with memories of Fujimori’s regime.
The context is the attack on the memorial of September 2007, which,
as we have seen, was blamed on fujimorista supporters, who, it was
claimed, acted out of frustration and in retaliation against the decision
of the Chilean authorities to extradite Fujimori to Peru to face trial. As
is well known, the trial of Fujimori led to his conviction and imprison-
ment (Burt 2009a).
Nuphi’s rhetorical strategy focuses on discrediting Peru’s human
rights organizations and the CVR, which he or she claims are mere

Soifer_6844-final.indb 301 8/17/18 11:53 AM

 302 Paulo Drinot

fronts for Sendero Luminoso. He or she, moreover, dismisses these

organizations as being “caviar” and claims that both the report of
the CVR and the memorial cost vast sums, $10 million in the case
of the former and $300 million in the case of the latter: “They are
left-wing pitucos [well-off elites racialized as white], they are impov-
erished red whiteys with a complex who have recycled themselves into
human rights NGOs funded by Sendero Luminoso money and they
don’t want to be brought to account. [Enrique] Bernales, [Javier] Diez
Canseco, Sofia Macher [human rights activists and left-wing politi-
cians] before were pro Sendero.” Nuphi also condemns the fact that
the names of the forty-one senderistas murdered in the Castro Castro
prison are included in the memorial and reproduces the argument, fi rst
made by right-wing journalist Aldo Mariátegui, that whereas the CVR
demands justice for the Castro Castro dead, it says nothing about Car-
los Hidrogo, a police officer murdered by SL: “How stupid to put the
names of the murderers next to those of their victims. For example,
Carlos Idrogo [sic] the policeman, those from Castro Castro who also
appear there, he was murdered fi rst they gouged out his eyes with a
spoon and then they killed him, this reveals the sick minds of those
terrorists who are now ‘victims’ according to the CVR, how much did
their 10 million dollars ‘Report’ cost the country[?]. The ojo que llora
a rock with some stones how much did it cost[?] 300 million dollars.
Do your research and you will fi nd the truth behind the lies.” Finally,
Nuphi also criticizes the methodology used by the CVR to extrapo-
late the number of victims, a criticism that has been leveled time and
again: “There is a caviar theory to infer that there were not 25 thou-
sand but rather 70 thousand dead, on the basis of a small sample they
infer that if in the little village XXX they killed 100 (obviously not by
SL) according to the terrorist human rights NGOs in the other villages
the same thing happened [. . .] this theory is used to CALCULATE FISH
STOCKS and not in order to fi nd out the number of people killed by
Sendero Luminoso and the MRTA.”
In contrast, Nuphi stresses the role played by the armed forces and
by Fujimori in defeating the SL insurgency. He or she dismisses the
claim that the armed forces and Fujimori were responsible for human
rights violations: “How many poor people [gente del pueblo] did Fuji-
mori kill? La Cantuta, Barrios Altos, this is the only thing they repeat
over and over like a broken record, I can tell you are a terrorist try-
ing to influence opinion but you have no arguments, the MURDERERS
WERE THE TERRORISTS, the Armed Forces manned up in a war people

Soifer_6844-final.indb 302 8/17/18 11:53 AM

 Contested Memories of the Internal Armed Conflict 303

die or do you think that they only take prisoners. Think fi rst and then
talk.” As this suggests, Fujimori may have ordered some assassina-
tions, but such acts were justified. For Nuphi, those who criticize Fu-
jimori are siding with the insurgency: “THE MURDERERS OF SENDERO
LUMINOSO AND THE MRTA can stick with the caviares, the liberator
FUJIMORI is for the people, thank you Fujimori for peace, for stability
and progress [. . .] thank you for your public works [. . .] by the fruit of
their labor shall you know them.” For Nuphi, the CVR not only was a
waste of money but also “opened wounds.” It was heavily biased and
notoriously anti-fujimorista: “The famous CVR is a commission that
was named personally by [former president] Toledo to throw mud on
Fujimori, and the predictable conclusion was that ‘el Chino’ [Fujimori]
was the same as Hitler, Pol Pot, Mussolini, etc. The only thing miss-
ing was for them to say that the terrorists were held in concentration
camps.” He or she dismisses the attack on the memorial as a “psico-
social,” a staged political action aimed at discrediting Fujimori, and fi-
nally suggests that the decision by Chile to extradite Fujimori to Peru
was part of a broader anti-Peruvian strategy devised by that country
in alliance with the caviar class: “They ignore the fact that Chile is be-
hind the protests against the mining companies because in 20 years’
time they will have run out resources, they give Fujimori back in order
to wash their hand of the assassin and thief Pinochet [. . .] CAVIAR TER-
RORISTS UNITED WITH THE CHILEANS AGAINST FUJIMORI.”
These views are echoed by other posters. Namer Letnemip calls
on Peruvians to recognize and celebrate the sacrifice of Peru’s armed
forces in the war against the Shining Path: “I hope that justice will be
served to so many Peruvian soldiers who died defending us from the
terrorist criminals [. . .] hopefully one day [. . .] we Peruvians will stand
up to applaud those who freed us from the terrorist scourge.” Oscar
Palomino calls for the destruction of the memorial on similar grounds
to those expressed in the debates that followed the discovery that the
names of forty-one senderistas killed at the Castro Castro prison had
been included among the engraved stones: “THAT PIECE OF SHIT MUST
BE DESTROYED, or turned into a public urinal, because it is not accept-
able to have the name of a terrorist next to that of an innocent victim.”
Mauritopr interpellates the sculptor Lika Mutal directly and points to
what he or she perceives as the absurdity of the memorial by establish-
ing an equivalence with Nazi atrocities during the Second World War:
“Lika Mutal, why don’t you make a sculpture in Holland and put the
name of the Dutch people who died of hunger (during the Nazi oc-

Soifer_6844-final.indb 303 8/17/18 11:53 AM

 304 Paulo Drinot

cupation) and of their executioners the Nazi leadership, all together

[. . .] Let’s see if they let you make an ojo que llora in your country, be-
cause in Peru you can add the name of 41 Senderista murderers but not
that of the courageous policeman Hidrogo whose eyes were gouged
out with a spoon by these monsters. Lika MUTAL FUCK YOU
[in English in the original]!!.” Finally, Namer Letnemip expresses suc-
cinctly what many posters doubtless feel about the memorial (as con-
fi rmed by the fact that his comment receives a total of five “likes”): “I
would have more respect for the memorial ‘THE PENIS THAT PISSES.’”
These counter-memories are, in turn, countered by restatements of
the repertoire of memory of the IAC established in the CVR report
and reflected in El ojo que llora memorial. A poster with the online
tag ellesar19 takes it upon himself or herself to refute the arguments
of Nuphi and others. In so doing, he or she restates, in broad strokes,
several of the key arguments that the CVR put forward in presenting
its interpretation of the causes of Peru’s descent into violence. For ex-
ample, ellesar19 argues that the causes of the violence were structural,
linked to poverty and marginalization: “Terrorism has not been de-
feated like you believe[.] As long as marginality racism injustice exist,
as long as our brothers from the highlands are forgotten there will al-
ways be a breeding ground for terrorism.” Moreover, he or she sug-
gests, the senderistas were also victims. Many were young and were
manipulated into becoming foot soldiers of the insurgency: “Many of
those terrorists were young people who were brainwashed and were
used as foot soldiers and they also have a mother like you or me who
has suffered because her son was lost (in all senses of the word).” For
ellesar19, there were no victors in the IAC: “Those were 20 years of
war during which we all lost, even the terrorists[.] The only ones who
did not lose were the leaders on both sides Abimael Guzmán and Al-
berto Fujimori both traitors to the Fatherland and Fujimori will not be
executed—as happened when he was in power [a reference to extraju-
dicial killings conducted by paramilitary groups during the Fujimori
government]—but he will receive due process with lawyers and every-
thing.” In refuting the arguments put forward by Nuphi and others,
ellesar19 quotes directly from the CVR report: “I am going to quote
what the CVR says in its preface: ‘But there is a basis to argue that
these two decades of destruction and death would not have been pos-
sible without the deep contempt toward the poorest in society, as ev-
idenced in equal measure by the members of the PCP-Sendero Lumi-
noso and by agents of the state, the contempt that is interwoven into
each moment of the daily life of Peruvians.’”

Soifer_6844-final.indb 304 8/17/18 11:53 AM

 Contested Memories of the Internal Armed Conflict 305

Several of these posters engage each other in debate in the com-

ment section of more than one video. In the sixty-odd comments that
appear below the video uploaded by Carlos Quispe Geronimo (which
is almost identical to the one uploaded by Beto Serquen), for exam-
ple, we fi nd again ellesar19, Oscar Palomino, Namer Letnemip, nu-
phi, and others who feature in one or more additional video com-
ment sections.34 The nature of the debate is broadly similar, if at times
even more verbally violent. An exchange between Oscar Palomino,
ellesar19, and a poster with the tag aravelciers illustrates the antag-
onistic and mutually exclusive repertoires of memory that each camp
mobilizes. For Oscar Palomino, who claims on several occasions that
El ojo que llora is a monument to terrorism, the human rights abuses
committed by agents of the state, such as in the cases of Barrios Altos
and La Cantuta, which the CVR presented as case studies in its Final
Report, were fully justified: “TERRORISTS, TERRORISTS, TERRORISTS,
I’m glad that those from Barrios Altos and La Cantuta are dead, HE-
ROIC GRUPO COLINA THANK YOU!!!! I hope they burn in hell for ever.”
Confronted with such an argument, ellesar19 reproduces an argument
used in the debate with nuphi and which reflects the idea that the vio-
lence ultimately had structural causes: “because of people like you ter-
rorism will return because you have become an animal worse than the
terrorists, the fight against terrorism cannot depend on rifles but fight-
ing against poverty, discrimination, abuse and marginalization.” For
aravelciers, Oscar Palomino’s position is a perfect reflection of the fu-
jimorista narrative: “you are so ignorant that you don’t really know
the history but you dare to comment. Poor you, you are the typical lost
generation created by fujimorismo, what a shame.”
In the same way that the Final Report of the CVR produced a rep-
ertoire of memory that both professional and amateur memory entre-
preneurs can mobilize for a series of objectives, a repertoire of mem-
ory aligned with the fujimorista “memory of salvation” has emerged
which memory entrepreneurs of an opposed political and ideological
persuasion can mobilize to refute the CVR. In this repertoire, as we
have seen, the senderistas were terrorists who deserved to be killed,
extrajudicially if necessary. The armed forces were unquestioned he-
roes who saved the nation from a certain apocalypse. And the CVR
and its supporters are at best naïve fools and at worst terrorist sym-
pathizers or indeed terrucos tout court. The genealogy of this reper-
toire of memory is not difficult to trace, although unlike the repertoire
of memory of the CVR, it does not have a foundational text such as
the Final Report.35 Yet, it is clearly discernible in the ontologies of vi-

Soifer_6844-final.indb 305 8/17/18 11:53 AM

 306 Paulo Drinot

olence that certain sectors of the media (most notably journalists such
as Aldo Mariátegui) and certain political actors, particularly in the fu-
jimorista and APRA parties, the armed forces, the Catholic Church
hierarchy, and sectors of the business community, have privileged. El
ojo que llora, as both synecdoche and simulacrum of the Final Report
of the CVR, serves as an emblematic post-confl ict legacy over which a
seemingly unresolvable struggle is waged between these opposing rep-
ertoires of memory.

Conclusion

In August 2017, the director of the Lugar de la Memoria, Tolerancia e

Inclusión Social (LUM), Guillermo Nugent, was forced to resign when
the minister of culture, Salvador del Solar, under pressure from fuji-
morista parliamentarians, blamed him for having agreed to host an
exhibition titled “Resistencia Visual 1992” that was accused of be-
ing overly critical of Fujimori. The minister claimed that the exhibition
was “biased.” Surprisingly, the exhibition was not canceled, doubtless
to avoid accusations that the minister had, in effect, censored it. How-
ever, this did little to allay widespread suspicion that the minister and,
by extension, the government was willing to sacrifice basic principles,
in this case freedom of expression, for political survival in a context
where the fujimoristas, with full control of the legislature, were us-
ing their leverage to attack the government. The incident demonstrates
vividly the extent to which, under the current Kuczynski administra-
tion as well as during the Humala administration, the fujimoristas’
power to police how the past is remembered can, and is, operational-
ized politically. It also raises questions about what memories are per-
missible in Peru, and why, ultimately, some memories fail to gain trac-
tion while others can develop hegemonic pretensions.36
The fujimorista power to police how the past is remembered, and
therefore the politics of the present, was confirmed when in Decem-
ber 2017 President Kuczynski, who faced an impeachment proceed-
ing called by the fujimorista majority in Congress, decided to pardon
Alberto Fujimori. Kuczynski justified his decision by invoking the fu-
jimorista “memory of salvation”: Fujimori, he argued, had committed
errors and transgressions (not crimes) during a “violent chaotic crisis”
but had set Peru on the path to national progress. It was time, he urged

Soifer_6844-final.indb 306 8/17/18 11:53 AM

 Contested Memories of the Internal Armed Conflict 307

Peru’s young, to set old hates aside, reconcile, and “turn the page.”37
This call for Peru to embrace an amnesic amnesty, however, was met
with mass protests. As this suggests, Peru’s IAC remains an open his-
torical wound, one kept open by the seemingly unending, and unre-
solvable, struggles for memory over Peru’s violent past.
El ojo que llora memorial, as a physical site and a cyber-projection,
as both synecdoche and simulacrum of the Final Report of the CVR,
and as an emblematic post-conflict legacy, provides professional and
amauteur memory entrepreneurs with a physical and virtual emplace-
ment from which to mobilize the repertoire of memory established by
the Final Report of the CVR. They mobilize in order to insist on the
fulfillment of the CVR’s recommendations and to advocate on behalf
of victim-survivors. But El ojo que llora’s actual and virtual presence
also enables those who oppose the CVR to mobilize their own coun-
ter-memories, to perform their own repertoire of memory based on
the fujimorista “memory of salvation.”38 As I have shown, while the
extreme verbal violence of the debate may be specific to the online
format, the repertoires of memory that inform the debate are part of
a much broader memory politics that, in turn, informs and refracts
a mainstream politics characterized to a significant extent by a fuji-
morista/anti-fujimorista divide.
Yet while it is important to recognize this active “memory market,”
to use the term given it by Bilbija and Payne, we must also consider
its scope. For some sectors of the Peruvian population, particularly
for those most directly affected by the IAC such as victim-survivor
groups, these debates are obviously crucial. But for the majority of the
Peruvian population, they appear largely marginal outside of specific
political conjunctures that evoke them. In contrast to countries like
Argentina or Chile, where an official engagement with the violent past
has been a central policy of recent governments, the opposite is true
in Peru, where the governments of Alan García and Ollanta Humala,
both of whom have been directly implicated in human rights abuses
during the IAC, and now of Pedro Pablo Kuczynski have at best side-
lined the issue; directly opposed any official engagement; intervened
to effectively dampen open debate on the violent past; or, as happened
in the context of Fujimori’s pardon, embraced the fujimorista “mem-
ory of salvation.” But equally important is the fact that whereas in Ar-
gentina and Chile numerically and politically important sectors of the
population have mobilized around such issues, in Peru mobilization
on a similar scale either has not materialized or has tended to be re-

Soifer_6844-final.indb 307 8/17/18 11:53 AM

 308 Paulo Drinot

stricted to particular political conjunctures such as the recent electoral

process, Fujimori’s pardon, or localized contexts.
To some extent this is a product of the fact, as argued earlier, that
those most invested in the issues that these contested memories ex-
press, the victims, have limited political leverage. It is also a product of
the fact that in Peru most political forces, as well as much of the me-
dia, the armed forces, the hierarchy of the Catholic Church, and many
business groups—Peru’s conservative archipelago—are largely hostile
to, or, at best, disinterested in, human rights agendas and fi nd no po-
litical capital in engaging with the recommendations of the CVR. It is
a situation made worse by the weakness of the Left, which in other cir-
cumstances would be the natural channel for such initiatives. But the
weakness of the Left, a result of a complex set of circumstances not
unrelated to the IAC, is arguably itself expressive of the Peruvian pop-
ulation’s limited interest in the interpretations of Peru’s IAC and the
recommendations put forward in the Final Report of the CVR, and
of the lack of “appetite” for the type of politics that a full engagement
with the Final Report would entail. The combination of an official dis-
interest in, and at times hostility toward, the human rights memory
and the Left’s weakness creates space for the fujimorista narrative to
thrive.
However, this should not be seen as evidence that the contested mem-
ories I have discussed in this chapter are merely shared or produced by
either a small group of activist NGOs and human rights groups or
a few memory entrepreneur hotheads who express them online. Nei-
ther should it be seen as evidence that the contested memories cannot
fi nd broader audiences, or that they are irrelevant to broader questions
about Peruvian politics and society. As should be clear, these contested
memories, and their mobilization and operationalization in the context
of fujimorista and anti-fujimorista political jousting, are a key dimen-
sion of Peru’s struggles over the legacies of the IAC. Their contested na-
ture is an expression of how Peruvians engage with the past, but also a
reflection of how Peruvians understand the present and envisage the fu-
ture, even when they do not engage actively in the politics of memory.
As such, these memories must not be approached merely as an expres-
sion of the ongoing debates over how to interpret the IAC. They reflect
how, for many Peruvians, as Fujimori’s pardon demonstrates, the vio-
lent past continues to inform the present, and to alert us to the contin-
ued importance of its legacies within current and future politics.

Soifer_6844-final.indb 308 8/17/18 11:53 AM

 Contested Memories of the Internal Armed Conflict 309

Notes

1. On other sites of memory in Peru, see, for example, Feldman 2012 and
Delacroix 2014. This chapter draws on an ever-expanding scholarship on the
contested memories of the Internal Armed Confl ict, including del Pino and
Yezer 2013; Degregori et al. 2015; Milton 2014; Milton 2018.
2. For example, the hostility of the Alan García regime (2005–2010) to
the CVR is well known and came to the fore in the context of the plans to
build the Museo de la Memoria. Although García backtracked on his initial
rejection of German funds offered for the building after Mario Vargas Llosa
intervened, this episode illustrated the more general policy of his regime to-
ward human rights organizations and their attempts to foster a productive en-
gagement with the IAC among the Peruvian population. On Peru’s Museo de
la Memoria, or as it is now known, Lugar de la Memoria, Tolerancia e Inclu-
sión Social (LUM), see Milton and Ulfe 2011.
3. On the idea of demand for memory and the associated notion of a mem-
ory market, see Bilbija and Payne 2011.
4. For recent interpretations of the Revolutionary Government of the
Armed Forces, see Mayer 2009 and Aguirre and Drinot 2017.
5. On the rise of the Shining Path, the best account is arguably the Final
Report of the CVR. See also Palmer 1992; Poole and Rénique 1992; Stern
1998c; Portocarrero 1998; McClintock 1998; Degregori 2012a; Portocarrero
2012. See also Flores Galindo 2010. On the origins of the insurgency and
how it played out at the local and regional level, see Rénique 2004; Taylor
2006; Heilman 2010; González 2011; La Serna 2012; Wilson 2013; Meza
Salcedo 2016; del Pino 2017. See also Theidon 2004; Theidon 2012. Recently,
scholars have started to explore the conflict in greater detail in urban con-
texts, particularly Lima, and among urban youth. See Greene 2016 and Asen-
cios 2016.
6. For further elaboration on this point, see chapter 2 in this volume.
7. For reactions to the CVR and its report, see the online database Cen-
tro de Documentación e Investigación at LUM. See, for example, http://lum
.cultura.pe/cdi/video/congresista-martha-chávez-indicó-que-el-informe-fi nal
-de-la-comisión-de-la-verdad-y-la. Accessed September 17, 2017.
8. On El ojo que llora see Hite 2007; Drinot 2009; Milton 2011; and
Moraña 2012. On memorials and memory in other Latin American contexts,
see Jelin and Langland 2003.
9. For general accounts, see Milton 2014; and Saona 2014. On photogra-
phy, see Poole and Rojas Perez 2010; Murphy 2015; Ulfe and Sabogal 2016.
On novels and literature, see Vich, Hibbett, and Ubilluz 2009. On art, see
Vich 2015. On music, see Ritter 2012 and Aroni Sulca 2016a. On comics, see
Drinot 2017 and Milton 2017. See also Denegri and Hibbett 2016 for an im-
portant discussion of testimony and memory.
10. Saona 2012 is a short article that explores the Facebook page “Un día
en la memoria,” which deals directly with memories of the Internal Armed
Confl ict in Peru. But this study remains an exception.

Soifer_6844-final.indb 309 8/17/18 11:53 AM

 310 Paulo Drinot

11. http://www.youtube.com/watch?v=0JRvnhtzetY. Accessed May 11,

2014.
12. https://www.youtube.com/watch?v=GudouYSRn_o. Accessed May 11,
2014.
13. https://www.youtube.com/watch?v=nc3uHtCYqD4. Accessed May 11,
2014.
14. On the politics of exhumation, see Rojas-Perez 2017.
15. On the case of “La Cantuta,” see chapter 5 in this volume.
16. Jelin (2003, 33–34) defi nes memory entrepreneurs as those who “seek
social recognition and political legitimacy of one (their own) interpretation or
narrative of the past. We also fi nd [memory entrepreneurs] engaged and con-
cerned with maintaining and promoting active and visible social and political
attention on their enterprise.”
17. Two videos uploaded by a certain “Chaskky” in September 2006 illus-
trate how victim-survivor associations use El ojo que llora in their campaigns.
The videos show processions of relatives of victims in El Ojo que llora. In
one of them, the relatives are identified as members of COFADER. They are
shown carrying a large cross on which have been stuck photographs of the dis-
appeared. See http://www.youtube.com/watch?v=nc3uHtCYqD4 and http://
www.youtube.com/watch?v=ZTygU41o62s. Accessed May 11, 2014. On
victim-survivor groups, see de Waardt 2013.
18. On reparations, see, among others, LaPlante and Theidon 2007 and
de Waardt 2013.
19. On Yuyachkani and the retablistas, see Ulfe 2014; Garza 2014; see
also Ulfe 2011. On Cossio, see Milton 2017.
20. http://vimeo.com/47354511. Accessed May 11, 2014.
21. http://www.youtube.com/watch?v=O5ltX5Opz5A. Accessed May 11,
2014.
22. https://www.youtube.com/watch?v=kHN9P7xaptQ. Accessed May 11,
2014.
23. http://www.youtube.com/watch?v=Ioc5HkPfxz4. Accessed May 11,
2014.
24. http://www.youtube.com/watch?v=H8JGm2Ok0rY. Accessed May 11,
2014.
25. On the notion of “grievable” life and death, see Butler 2010; also But-
ler 2006 and Boesten 2014.
26. This is a reference to the students and their professor murdered by
a government death squad called Grupo Colina. Fujimori has been accused
of responsibility for the actions of this group. See Burt 2009a. http://www
.youtube.com/watch?v=NQ_9vNKfagk. Accessed May 11, 2014.
27. http://www.youtube.com/watch?v=H8JGm2Ok0rY. Accessed May 11,
2014.
28. http://www.youtube.com/watch?v=3otNKiFM2nE. Accessed May 11,
2014.
29. http://www.youtube.com/watch?v=3otNKiFM2nE. Accessed May 11,
2014.
30. http://panamericanatv.pe/politica/48412.

Soifer_6844-final.indb 310 8/17/18 11:53 AM

 Contested Memories of the Internal Armed Conflict 311

31. https://www.youtube.com/watch?v=olB5dl95kpM. Accessed May 11,

2014.
32. http://www.youtube.com/watch?v=AMS9ZfHtPl8. Accessed May 11,
2014.
33. https://www.youtube.com/watch?v=kHN9P7xaptQ. Accessed May 11,
2014.
34. http://www.youtube.com/watch?v=3otNKiFM2nE. Accessed May 11,
2014.
35. Of course, the Final Report built on years of human rights activism, so
its foundational text was not an ab initio interpretation of the IAC.
36. The virtual absence of senderista and MRTA memories of the confl ict
in the public sphere is a case in point. Though some accounts are beginning
to circulate, they do so only within a register that is broadly consonant with
the repertoire of memory of the CVR. See the testimonial literature of Agüero
2015, Gavilán 2012, and Gálvez 2015. This is in part a product of a contro-
versial legal framework that criminalizes not only senderista “apología” (con-
doning the Shining Path or advocating terrorism) but also, more broadly, a
moral, intellectual, and political climate that denies space to such memories.
37. See http://larepublica.pe/politica/1162485-ppk-habla-de-excesos-y
-errores-de-fujimori-pero-no-absuelve-dudas. Accessed February 6, 2018.
38. Again, what is lost in this hostile debate are the subtler and more com-
plex stories of the IAC that scholars such as Theidon, Heilman, La Serna, del
Pino, and others are beginning to uncover. These stories show that the strict
lines of separation between victim and perpetrator which both the fujimorista
memory of salvation and the CVR’s human rights memory privileged were
politically expedient, but not always historically accurate or compatible with
attempts by communities to come to terms with the violent past.

Soifer_6844-final.indb 311 8/17/18 11:53 AM

 Conclusion
Steven Levitsky

Contemporary Peruvian politics often seems to diverge in impor-

tant ways from that of its neighbors. As Latin America’s third wave
of democratization reached its zenith in the early 1990s, for exam-
ple, Peru suffered the region’s only full-scale democratic breakdown.
Whereas Bolivia and Ecuador experienced the rise of powerful na-
tional indigenous movements in the 1980s and 1990s, no comparable
ethnic mobilization occurred in Peru (Yashar 2005). And whereas the
Left ascended to power in most of Latin America during the 2000s,
Peru remained a beacon of neoliberal continuity (Vergara and Enci-
nas 2016).
When one asks scholars of Peru why the country’s recent political
trajectory has been so distinctive, responses almost invariably include
mention of its bloody Internal Armed Confl ict (IAC). Yet the causal
processes underlying these claims remain poorly understood. Indeed,
research on the political and institutional legacies of the Shining Path
war has, until recently, been limited.
As Soifer and Vergara note in the Introduction to this volume, isolat-
ing the political consequences of the Shining Path is a challenging task.
Several other developments that occurred just prior to, at the same time
as, or in the immediate aftermath of the IAC also powerfully shaped
Peru’s late twentieth- and early twenty-fi rst-century politics (often in-
teracting in complex ways with the IAC). Among these potential con-
founders are the collapse of the rural oligarchic order in the 1960s and
1970s, the economic crisis of the 1980s, the expansion of the urban in-
formal sector, the radical neoliberal reforms in the 1990s, and the post-
2002 commodities boom.

Soifer_6844-final.indb 312 8/17/18 11:53 AM

 Conclusion 313

Evaluating the consequences of the IAC is further complicated by

the fact that its effects were both direct and indirect. Direct effects in-
clude the weakening of party and civic organizations via the killing,
imprisonment, and widespread intimidation of grassroots activists (see
chapters 5, 8, and 9 in this volume), the redesign of internal security
institutions in the face of insurgency (chapter 4), and the reshaping of
public opinion in a more conservative, law-and-order direction (chap-
ter 10). Other consequences of the IAC are plausible but indirect. For
example, the crisis generated by the IAC contributed to both the rise
of Alberto Fujimori and the success of his 1992 autogolpe, which, in
turn, contributed to a host of outcomes, including the 1993 constitu-
tion, the acceleration of party system collapse, and the consolidation of
neoliberalism (chapter 3 in this volume). It was Fujimori’s authoritari-
anism, not the Shining Path, that “locked in” neoliberalism, constitu-
tionally and sociopolitically, but the success of fujimorismo was clearly
rooted in the IAC. As José Luis Rénique and Adrián Lerner (chapter 1
in this volume) put it, the capture of Abimael Guzmán provided Fuji-
mori’s nascent authoritarian regime with a “fundamental credibility”
that both emboldened and empowered it.
This highlights a broader point made by Soifer and Vergara in their
Introduction: the political legacies of conflict are constructed, con-
tested, and constantly reshaped by political actors. The IAC did not au-
tomatically give rise to any particular institutions, policies, discourse,
or public attitudes. They were created—by state officials, politicians,
media, and other important “memory entrepreneurs” (chapter 11 in
this volume). Thus, any evaluation of the political legacies of the IAC
must pay close attention to how those legacies are mediated by poli-
tics—to how they are constructed, manipulated, opposed, and often
weakened by social and political actors. To be sure, the IAC produced
certain “raw materials” that favored some outcomes over others. For
example, it created an arsenal of new political weapons for potential
use by conservatives while leaving behind a set of difficult organiza-
tional, ideological, and discursive challenges for the Left. Neverthe-
less, it is social and political actors who turn these raw materials into
political reality.
With these important caveats in mind, this concluding chapter at-
tempts to synthesize some of the ideas generated by the chapters in
this volume. At the same time, it seeks to place the Peruvian case in a
broader comparative perspective.

Soifer_6844-final.indb 313 8/17/18 11:53 AM

 314 Steven Levitsky

A Conservative Democracy

The IAC appears to have produced a fundamental shift in the distribu-

tion of political power in Peru. In Argentina, Bolivia, Chile, Brazil, Ec-
uador, El Salvador, Uruguay, and elsewhere in Latin America, two or
more decades of democracy empowered progressive and popular sec-
tor forces, shifting politics and policy leftward during the 2000s. In
Peru, by contrast, the distribution of sociopolitical power continued to
favor conservatives, even after seventeen years of democracy. As sev-
eral chapters in this volume show, this outcome can be plausibly traced
to the IAC. On the one hand, the Left and allied popular movements
were weakened dramatically. On the other hand, an array of forces on
the Right emerged from the confl ict both empowered and unusually
cohesive. As a result, Peru’s post-Fujimori democracy was marked by a
striking degree of conservative policy continuity.

A Weakened Left

The IAC nearly destroyed the Peruvian Left. The United Left (Iz-
quierda Unida, or IU), which had been a serious contender for the
presidency as late as 1989, disintegrated in the 1990s and fell into “ir-
relevance” in the 2000s (chapter 8 in this volume). Although the Left’s
demise may be attributed to numerous factors (including the 1980s
economic crisis, the growth of the informal sector, ties to the disas-
trous García government, the crisis of Communism, and the fact that,
unlike many of its South American counterparts, the Peruvian Left un-
derwent little ideological renovation in the 1970s and 1980s),1 Paula
Muñoz compellingly argues that the IAC played a central role.
The armed conflict weakened the partisan Left in at least two ways.
First, it destroyed much of the Left’s human infrastructure. As Mu-
ñoz shows, leftist politicians and activists were hit hard from both
sides during the war. On the one hand, they fell victim to the Shin-
ing Path’s strategy of “selective annihilation” of its left-wing rivals, the
most notorious case of which was the brutal assassination of María
Elena Moyano. On the other hand, numerous leftists were either killed
or imprisoned during the indiscriminate repression unleashed by the
state-led counterinsurgency (CVR 2003a). Amid the climate of fear
created by the confl ict, countless left-wing activists abandoned poli-
tics, particularly in the interior. As a result, the Left, which had devel-

Soifer_6844-final.indb 314 8/17/18 11:53 AM

 Conclusion 315

oped an impressive grassroots presence in the 1970s and 1980s, saw

its organizational bases disintegrate.
Even more consequential, however, was the IAC’s impact on the
Left’s partisan brand. As Muñoz’s chapter shows, Peru’s legal Left
failed to differentiate itself from armed guerrilla movements during
the 1980s—and this failure proved politically costly. Much of the Pe-
ruvian Left remained openly Marxist and revolutionary in the 1980s
(Roberts 1998). Thus, although IU leaders took great pains to distin-
guish themselves from the Shining Path, their continued commitment
to revolutionary struggle and ambiguous position on violence rendered
these distinctions meaningless in the eyes of many ordinary Peruvi-
ans (Muñoz, chapter 8 in this volume). The Left’s ambiguity with re-
spect to violence was exacerbated by the rise of the MRTA, a more tra-
ditional leftist guerrilla group with ties to elements of the legal Left.
Much of the Peruvian public did not draw sharp distinctions between
the Shining Path and the MRTA, or between the MRTA and the legal
Left. All of these organizations, for example, flew red flags, employed
Marxist discourse, and embraced revolutionary struggle.
The IAC thus engendered a “strong association between Left and
terrorism” that endures—and is actively perpetuated by right-wing
politicians—to this day (see chapter 8 of this volume). Because Marx-
ism was so strongly linked in the public mind to violence and chaos,
leftists who employed any traditional leftist discourse and symbols—
the color red, the hammer and sickle, class-based discourse, the
clenched fist in the air—were easily stigmatized as “terrorists.” Left-
ists of virtually all stripes—including moderates with impeccable dem-
ocratic credentials, such as Lima’s former mayor, Susana Villarán—
were vulnerable to red-baiting. Indeed, there emerged a virtual cottage
industry of red-baiting in the Lima media, as influential figures such
as Aldo Mariátegui dedicated themselves almost exclusively to attack-
ing the Left.
The enduring association between leftism and terrorism undermined
the Left’s electoral performance throughout the post-Fujimori period.
Indeed, at a time when leftist candidates were winning national office
throughout South America, including in neighboring Bolivia, Chile,
and Ecuador, no leftist party in Peru won even 1 percent of the vote in
the 2001, 2006, or 2011 presidential elections.2 Although leftist can-
didate Verónika Mendoza won nearly 19 percent of the vote in 2016,
she, too, was bombarded with accusations of terrorism (even though

Soifer_6844-final.indb 315 8/17/18 11:53 AM

 316 Steven Levitsky

she was only a child during the IAC), and anti-leftism was a major fac-
tor in her failure to qualify for the second-round runoff.
The Left’s weakness extended beyond the electoral arena, however.
The IAC also undermined what Vergara and Encinas call the “social
left,” or popular and civic organizations that often align with left-wing
parties, providing them with grassroots linkages and muscle for mobi-
lization (see chapter 9 of this volume). Peru’s social Left—particularly
the General Peruvian Workers Confederation (Central General de Tra-
bajadores del Perú, or CGTP)—peaked in strength during the massive
popular mobilizations of the late 1970s, providing the IU with a pow-
erful grassroots base (Roberts 1998).
The IAC clearly weakened these popular movements. Senderista at-
tacks and state repression thinned the ranks of many popular organi-
zations. Thousands of social activists were killed or imprisoned during
the 1980s (CVR 2003a). Fear pushed many Peruvians out of the public
sphere and deterred many others from joining it, leaving behind a far
more demobilized social movement landscape than had existed in the
1970s and early 1980s (Burt 2007). The IAC also weakened the social
Left indirectly by facilitating Alberto Fujimori’s 1992 coup and subse-
quent authoritarian rule. Freed of legislative and judicial constraints,
the Fujimori government took numerous steps to weaken the legal and
organizational bases of the labor, student, and other progressive asso-
ciations (see chapter 3 in this volume; also Burt 2007).
The destruction of the social Left was perhaps most evident in the
case of the labor movement. The CGTP, which had been the backbone
of leftist mobilization in the late 1970s (Roberts 1998), was decimated
in the late 1980s and 1990s. Violence against union leaders and activ-
ists—most notably, CGTP leader Pedro Huilca—had a direct demo-
bilizing effect by deterring union activism (see chapter 9 in this vol-
ume). At the same time, as Cameron notes in chapter 3, the Fujimori
government undertook a series of measures to weaken unions and in-
hibit class-based collective action, including measures to expand labor
market flexibility that were far more extensive than those carried out
in most Latin American democracies (see Roberts 1998). Union mem-
bership declined from a peak of 25 percent of the workforce in the late
1970s to 5.7 percent in the 1990s—one of the steepest declines in the
region (Roberts 2014, 100). Although this decline was rooted in vari-
ous factors, including severe economic crisis, industrial restructuring,
and the informalization of the economy, Fujimori’s anti-union mea-
sures clearly exacerbated it.

Soifer_6844-final.indb 316 8/17/18 11:53 AM

 Conclusion 317

The IAC also devastated the student movement. As Dargent and

Chávez show in chapter 5, both the Shining Path war and the Fuji-
mori regime undermined university-based student activism. For one,
Shining Path and MRTA penetration of public universities helped to
create an “image of chaos and violent radicalism” that deterred stu-
dent participation (see chapter 5 in this this volume, p. 137).3 At the
same time, the state’s intervention and takeover of public universities
during the initial authoritarian period (1992–1995), followed by far-
reaching neoliberal reforms that encouraged the growth of for-profit
private universities (while public universities were badly neglected), re-
sulted in the flight of many of the “best and brightest” from public to
private universities and a long-term depoliticization of university stu-
dents. Thus, as Dargent and Chávez point out, student political orga-
nizations were “disarticulated and lost relevance” in the 1990s and
did not recover in the 2000s. A major legacy of Shining Path insur-
gency and Fujimori’s authoritarian counterinsurgency, then, has been
the long-term demobilization of Peruvian university students. This
outcome contrasts sharply with other countries in the region, such as
Brazil, Chile, and Mexico, where students played a central role in pro-
test movements during the 2000s.
The IAC may also have inhibited the growth of a national indige-
nous movement in Peru, as Paredes suggests in chapter 7 (see also Ya-
shar 2005). In much of the highlands, where indigenous organizations
might have mobilized in the 1980s and 1990s, Shining Path violence
and state repression dampened or inhibited popular mobilization (Ya-
shar 2005). Thus, whereas indigenous organizations emerged as pow-
erful national-level actors in Bolivia and Ecuador, in Peru such orga-
nizations remained small, largely confi ned to the Amazon region, and
largely unconnected to national-level parties or social movements (see
chapter 7).
Finally, the stigmatization of protest—another clear legacy of the
IAC—reinforced the weakness of the societal Left. Lima’s conser-
vative establishment and much of the media routinely associate stu-
dent, labor, and now environmental protest with radicalism, violence,
and terrorism (see chapters 3, 5, 7, 8, and 9). Anti-mining protests
are frequently characterized by the media, businesspeople, and right-
wing politicians and technocrats as violent and radical, and they al-
most invariably trigger charges that they are orchestrated by terror-
ists. Indeed, the association of protest with violence and terrorism has
at times been used to justify police repression of protest and the ar-

Soifer_6844-final.indb 317 8/17/18 11:53 AM

 318 Steven Levitsky

rest of protest leaders (such as in Espinar, Cusco, in 2012). This “crim-

inalization” of protest is clearly facilitated by collective memories of
the chaos and violent radicalism of the 1980s. Indeed, as Maldonado,
Merolla, and Zechmeister show in chapter 10 of this volume, fear of
terrorism in Peru is among the highest in Latin America. The stigma-
tization of protest has important consequences for the Left. As schol-
ars such as Eduardo Silva (2009) and Kenneth Roberts (2014) have
shown, the most robust left turns in contemporary Latin America—
for example, in Bolivia, Ecuador, and Venezuela—have taken place in
wake of sustained popular mobilization.
In sum, the IAC and Fujimori’s authoritarianism weakened the par-
tisan Left and decimated (or inhibited the growth of) the kinds of social
organizations—unions, student organizations, indigenous groups—
that played a central role in reviving the Left in Bolivia, Chile, Ecua-
dor, Mexico, and elsewhere during the 2000s. Thus not only was the
Left’s partisan brand badly tarnished, but its social and organizational
bases were eviscerated.

An Empowered (and Cohesive) Right

In contrast, right-wing forces emerged strengthened in the aftermath

of the IAC (see chapters 3 and 9). As Vergara and Encinas remind us,
the established partisan Right suffered considerably during the IAC.
Local leaders and activists from the PPC, AP, and APRA were ma-
jor targets of Shining Path violence.4 All three parties saw their hu-
man infrastructure depleted, especially in the interior. Moreover, the
dramatic failure of the AP (1980–1985) and APRA (1985–1990) gov-
ernments to prevent Peru’s plunge into bloody internal confl ict led to a
steep decline in public support for the established parties (chapter 9 in
this volume; also Cameron 1994).
Yet if traditional conservative parties fared poorly during the 1990s,
the “social Right” strengthened considerably (chapter 9 in this vol-
ume). The strength, self-confidence, and political influence of the busi-
ness elite expanded considerably during the Fujimori period (F. Durand
1999; F. Durand 2003). Not only did the private sector thrive, but it
established a greater political presence via business organizations such
as La Confederación Nacional de Instituciones Empresariales Prividas
(CONFIEP), pro-business think tanks such as the Peruvian Institute of
Economics (IPE), and a range of influential right-wing media.
The technocratic elite also emerged strengthened from the Fujimori

Soifer_6844-final.indb 318 8/17/18 11:53 AM

 Conclusion 319

regime (Dargent 2015; Vergara and Encinas 2016; chapter 9 in this

volume). As Vergara and Encinas show, neoliberal technocrats and bu-
reaucrats thoroughly penetrated the state during the 1990s. And in
the absence of viable partisan or political counterweights, this tech-
nocratic predominance persisted throughout the post-Fujimori period.
Finally, the Church, which shifted markedly to the Right during
the 1990s, was also strengthened during the IAC, as President Fuji-
mori awarded Archbishop Juan Luis Cipriani a high-profile political
role during the counterinsurgency (chapter 9). Appointed cardinal in
2001, Cipriani remained an active and influential political actor, ad-
vocating on behalf of numerous right-wing causes during the 2000s
(Pásara 2014).
Vergara and Encinas thus describe the emergence of a loose but
powerful coalition of right-wing business, media, political, and tech-
nocratic elites in the wake of the Fujimori regime. Although the em-
powerment of business, neoliberal technocrats, and conservative
Church figures was not a direct product of the IAC, it was clearly fa-
cilitated by Fujimori’s authoritarian rule—and as Cameron observes
in chapter 3 of this volume, it was likely reinforced by Fujimori’s 1993
constitution. Fujimori’s authoritarianism and the constitutional order,
in turn, are clearly rooted in the confl ict.
Equally important as the strength of individual members of Peru’s
right-wing archipelago, however, is the striking cohesion among them.
Indeed, notwithstanding the fragmentation described by Vergara and
Encinas, the Right consistently and forcefully closed ranks in the face
of perceived threats during the 2000s. Even minor threats to eco-
nomic orthodoxy such as center-left politician Susana Villarán’s elec-
tion as mayor of Lima, anti-mining protest in Cajamarca, and Presi-
dent Ollanta Humala’s floating of a proposal for the state to purchase
shares of the Spanish energy company Repsol triggered a strong and
unified response from across the conservative archipelago. This elite
cohesion, which contrasts starkly with previous periods in Peruvian
history (Cotler 1978), enabled right-wing actors to exercise a virtual
veto power over macroeconomic policy in the 2000s.
Arguably, then, the most impressive characteristic of the post-
Fujimori Right is not its fragmentation but its cohesion in defense of
the economic status quo. This cohesion was likely rooted in the pro-
longed existential threat posed by the IAC (see chapter 3 in this vol-
ume). Much like the Chilean economic elite in the 1970s, Peru’s eco-
nomic elite faced a powerful threat to its core class interests in the late

Soifer_6844-final.indb 319 8/17/18 11:53 AM

 320 Steven Levitsky

1980s and early 1990s, which, as Jeffry Frieden (1991) argues in the
Chilean case, tends to generate upper-class cohesion.
In sum, the IAC and its authoritarian aftermath had important con-
sequences for the distribution of sociopolitical power in Peru. The
weakening of the electoral Left and the erosion of its social bases, to-
gether with the emergence of a stronger and more cohesive “societal”
Right, shifted politics in a conservative direction (see Cameron 2011
and chapter 3 in this volume). Even under democracy, social protest
was stigmatized, leftist politicians faced constant red-baiting, and ef-
forts to shift policy away from the neoliberal status quo faced resis-
tance from a strikingly unified conservative elite.
This new balance of forces had significant consequences for policy.
Peruvians elected presidents in 2001 (Toledo), 2006 (García), and es-
pecially 2011 (Humala) who might have embraced the regional “left
turn” and broken at least modestly with the Washington Consensus
by expanding the role of the state in the economy or substantially in-
creasing social spending. Yet, facing intense pressure from inside (tech-
nocrats) and outside (business, media) the state on the one hand, and
limited popular mobilization on the other hand, all three presidents
governed on the right (Cameron 2011; Vergara and Encinas 2016).
Even politicians with clear left-of-center platforms, such as Ollanta
Humala, quickly concluded that the neoliberal status quo was the path
of least resistance.
Some scholars point to this striking neoliberal continuity as evi-
dence that Peru did not fully democratize in the 2000s (Lynch 2009).
For these scholars, the persistence of Fujimori’s 1993 constitution is
an indicator of Peru’s “incomplete” transition. Yet by standard defi-
nitions, the post-Fujimori regime is clearly democratic. And although
the origins of the 1993 constitution were authoritarian, the document
itself is broadly compatible with liberal democracy. As Cameron notes
in chapter 3 of this volume, Fujimori’s authoritarianism after 1993
was based on gross violation of his own constitution. Rather than a
product of authoritarian persistence, then, the conservatism of post-
Fujimori democracy is better understood as a product of a fundamen-
tal reshaping of the balance of political power produced by the IAC.

Limited State-Building

If the IAC clearly reshaped the distribution of sociopolitical power in

Peru, its impact on the state’s coercive structures appears to have been

Soifer_6844-final.indb 320 8/17/18 11:53 AM

 Conclusion 321

more modest. Given that recent scholarship has identified counter-

insurgency as a potential impetus for longer-term state-building (Slater
2010), this outcome merits some discussion.
The armed forces are strikingly absent from Vergara and Encinas’s
discussion of Peru’s “conservative archipelago.” This absence is not an
oversight on the authors’ part; rather, it reflects the fact that the con-
servative forces that predominated in post-Fujimori Peru were almost
exclusively civilian. Given Peru’s history of military intervention, and
particularly given the centrality of the security apparatus to Fujimori’s
authoritarian regime (Obando 1993; Obando 1998b; Obando 1999;
Rospigliosi 2000; Soifer and Vieira, chapter 4 in this volume), such an
outcome may appear surprising.
Counterinsurgency tends to bring militarization and an expan-
sion of the domestic security apparatus, often with authoritarian con-
sequences (Burt 2007; Slater 2010). Indeed, in Peru, the IAC eroded
civilian control over the military, contributed to the 1992 coup, and
brought the military into power as part of a civil-military regime
(chapter 4 in this volume; Rospigliosi 2000). As Soifer and Vieira
show, Fujimori’s counterinsurgency brought a vast expansion of the
state intelligence apparatus, which began to operate outside the con-
trol of civilian authorities. Even after the notorious National Intelli-
gence Service (SIN) was deactivated in the early 2000s, a series of in-
telligence scandals—some related to illegal surveillance of politicians,
journalists, and other civilians—continued to mar Peruvian democ-
racy (Weeks 2008).
Nevertheless, military power eroded quite dramatically in the
2000s. Unlike other cases in which military regimes confronted pow-
erful left-wing threats, such as in Chile, Colombia, El Salvador, and
Guatemala, Peru’s security forces largely vanished from the political
scene after 2000. The military command was restructured under in-
terim President Valentín Paniagua (2000–2001), and top military offi-
cials—including armed forces chief Nicolás Hermoza Rios—were tried
and convicted of corruption or human rights abuses. Military influence
eroded considerably. As Soifer and Vieira show, both the size and the
budget of the armed forces were reduced to 1970s levels.
Unlike Colombia, moreover, Peru did not experience the emergence
of powerful paramilitary forces in the aftermath of the IAC. Although
the Peruvian IAC triggered the emergence of a range of paramilitary
groups, including death squads such as the Rodrigo Franco Command
and the Colina Group and armed peasant militias, or rondas campesi-
nas, these organizations were either dismantled or (in the case of many

Soifer_6844-final.indb 321 8/17/18 11:53 AM

 322 Steven Levitsky

rondas) redirected into civilian activities that did not involve the kind
of extralegal repression seen in Colombia.
Thus, whereas other Latin American counterinsurgencies left un-
democratic legacies such as powerful militaries (Guatemala) or exten-
sive paramilitary organization (Colombia), in Peru such legacies were
weaker. This outcome may have been a product of the 2000 transition.
The security forces were corrupted and politicized under Fujimori
(Rospigliosi 2000; Soifer and Vieira, chapter 4 in this volume), which
left them discredited, politically weakened, and vulnerable to prosecu-
tion after Fujimori’s fall. Neither Colombia (a stable civilian regime)
nor Guatemala (which underwent a controlled, military-led transition)
experienced such a rupture with the past. Nevertheless, the question of
why militarism and paramilitarism were so limited during the 2000s
remains a puzzle that merits further exploration.

The Resurgence of Fujimorismo

The persistent strength of fujimorismo—a phenomenon that receives

only modest attention in this volume—is one of the most significant
(and unexpected) legacies of the IAC. At the time of Fujimori’s fall
in 2000, few observers believed that he or his followers had a politi-
cal future. Leaked videotapes had revealed to the world that Fujimo-
ri’s government not only had been authoritarian but also engaged in
scandalous levels of corruption and abuse of power (Cameron 2006).
Moreover, Fujimori had been a notoriously personalistic ruler, creat-
ing and discarding four different political parties during his decade
in power. After Fujimori’s fall, his latest political vehicle, Peru 2000,
disappeared. The fujimorista movement fragmented and was widely
expected to fade away, much like Odrismo—the political movement
founded by former dictator Manuel Odría—did in the 1960s. Yet to-
day, fujimorismo is Peru’s largest and best-organized party. While Al-
berto Fujimori languished in prison, his daughter, Keiko, rebuilt the
fujimorista organization (renaming it Popular Force, or FP). Keiko Fu-
jimori nearly won the presidency in 2011 and 2016, and the FP cap-
tured an absolute majority of seats in Congress in 2016.
Fujimorismo is an example of what James Loxton calls authori-
tarian successor parties, or political parties born of authoritarianism
(and founded by former regime elites) which survive and compete un-
der democratic regimes. Authoritarian successor parties are surpris-

Soifer_6844-final.indb 322 8/17/18 11:53 AM

 Conclusion 323

ingly common. According to Loxton, they emerged in forty-seven of

sixty-five new democracies between 1978 and 2010—eventually re-
turning to power in thirty-five of them (Loxton 2016, 12).
Loxton (2016) argues that authoritarian successors enjoy several
advantages in party-building, including established brands, leftover
patronage and clientelist networks to build upon, and ties to economic
elites that often serve as a key source of funding. However, authoritar-
ian successor parties are likely to thrive under a democracy only where
autocrats are perceived by much of the electorate as having achieved
important policy successes (e.g., Pinochet in Chile, the PRI in Mexico).
Fujimorismo clearly met these conditions. Not only could it draw upon
networks of former state officials, old clientelist brokers, and friendly
business elites to build what became Peru’s best-organized party (see
Levitsky and Zavaleta 2016), but it also maintained a clear—and for
many Peruvians, appealing—brand. As observed in chapter 10 of this
volume, Fujimori is widely credited by Peruvians for defeating terror-
ism. Indeed, this success has produced a relatively favorable view of
his presidency—despite its authoritarianism and corruption (Carrión
2006a). In a 2006 survey, for example, 48 percent of respondents ex-
pressed a positive view of Fujimori’s presidency.5 In 2011, 30 percent
of respondents in a poll carried out by IPSOS ranked the Fujimori gov-
ernment as the most effective in the last fifty years, and in a 2013 sur-
vey, 42 percent described Fujimori’s performance as “good” or “very
good.”6
Fujimorismo’s resurgence was thus fueled, in part, by the persistent
strength of what Drinot in chapter 11 calls the fujimorista “memory
of salvation” (see also Urrutia 2011 and Deming 2013). Public percep-
tions of Alberto Fujimori’s success in combating terrorism enabled the
FP to credibly brand itself as a “law and order” party. This branding
proved especially useful during the 2000s, as rising violent crime rates
placed security issues at the top of the public agenda.7
The resurgence of fujimorismo reshaped patterns of partisan com-
petition, establishing fujimorismo versus anti-fujimorismo as Peru’s
primary electoral fault line. This cleavage was the source of intense po-
larization during the 2011 and 2016 presidential election campaigns.
The fujimorismo/anti-fujimorismo divide cross-cuts the Left-Right di-
vide. Whereas illiberal sectors of the Right tend to sympathize with fu-
jimorismo, many center-right liberals—most notably, Nobel Laureate
Mario Vargas Llosa—staunchly oppose it. In 2011, for example, Var-
gas Llosa and other liberals backed left-of-center candidate Ollanta

Soifer_6844-final.indb 323 8/17/18 11:53 AM

 324 Steven Levitsky

Humala over Keiko Fujimori. This support was arguably decisive in

Keiko’s defeat.
Fujimorismo thus had a twofold effect on the Peruvian Right. On
the one hand, it strengthened the Right, infusing it with a popular
sector base that traditional right-wing parties had lacked (Meléndez
2014). On the other hand, fujimorismo divided the Right. This divi-
sion becomes particularly manifest when perceived leftist or populist
threats are weak, as was the case, for example, in the 2016 presidential
runoff, when Keiko Fujimori faced the right-of-center candidate Pedro
Pablo Kuczynski. Fujimorismo’s resurgence thus threatens to under-
mine the extraordinary cohesion that has marked the right-wing archi-
pelago since 2001. In this sense, it is somewhat similar to Uribismo,
another child of an internal armed conflict, which both strengthened
and divided the Right in Colombia.

The Resilience of Post-Fujimori Democracy

In light of the many challenging legacies of the IAC, Peru’s democracy

has performed surprisingly well since 2001. To be sure, the regime suf-
fers from a range of shortcomings (Vergara and Watanabe 2016). Lev-
els of public dissatisfaction with democratic institutions are among the
highest in Latin America, and in 2016, fujimorismo came within a
hair’s breadth of winning the presidency. Nevertheless, Peru has now
maintained a democratic regime longer (seventeen years) than in any
other period in Peruvian history. Indeed, never before has the mili-
tary been so marginal, human rights and civil liberties so broadly pro-
tected, or presidential power so limited.
Peru’s democratic performance since 2001 stands in contrast to sev-
eral of its Andean neighbors. Peru has experienced no constitutional
ruptures or slides in its competitive authoritarianism, as occurred in
Venezuela, Ecuador, and Bolivia. Nor have there been the kind of hu-
man rights and civil liberties violations seen in Colombia. Indeed, if
one examines comparative democracy indices such as Freedom House,
Polity IV, and Variety of Democracies, Peru’s democracy score since
2001 has remained consistently superior to those of Bolivia, Colom-
bia, Ecuador, and Venezuela.
Peru’s contemporary democratic stability has been achieved despite
several challenging legacies of the IAC, including the militarization of
the 1990s (chapter 4 of this volume), the collapse of political parties

Soifer_6844-final.indb 324 8/17/18 11:53 AM

 Conclusion 325

(chapters 3, 8, and 9), a weakened civil society (chapters 5 and 7), il-
liberal attitudes associated with fear of terrorism (chapter 10), and the
persistence of deeply contested memories of the past (chapter 11). It
was also achieved despite various structural conditions—such as vast
social inequality and pervasive state and institutional weakness—that
have been associated with regime instability elsewhere in the region
(e.g., Bolivia, Ecuador, and Venezuela).
What explains Peru’s relative democratic success since 2001? One
could argue that the regime survived largely by default, in a context
of favorable exogenous conditions. Due to the post-2002 commodi-
ties boom, Peru’s economy grew vigorously in the post-Fujimori pe-
riod. High growth rates are associated with democratic survival, even
in countries with relatively unfavorable structural conditions (see Prze-
worski et al. 2000).
Yet the chapters in this volume highlight some legacies of the IAC
that may have contributed to democratic stability (though not neces-
sarily democratic quality). One of these is the absence of a significant
threat to elite economic interests. In contrast to the 1980–1992 pe-
riod, the post-Fujimori era has been characterized by a weak Left and
low levels of popular mobilization. Although popular sector weakness
hardly contributes to the quality of democracy, the absence of serious
or sustained threats to elite economic interests likely contributed to re-
gime stability.8 Business and right-wing elites had little to fear under
post-Fujimori democracy (Cameron 2011; see also chapter 3 of this
volume). Indeed, Peru became a technocrat’s paradise—in which tech-
nocrats’ ability to make and sustain economic policies was largely un-
affected by democratic politics (Vergara and Encinas 2016). Voters
may elect left-of-center candidates, as they did in 2011, but in the ab-
sence of sustained popular mobilization or effective leftist or populist
parties, state bureaucrats have had little difficulty maintaining the or-
thodox economic model established during the 1990s. Such an out-
come may be problematic from the standpoint of democratic represen-
tation (Vergara and Watanabe 2016), but in the 2001–2016 period,
the security it provided to economic elites likely enhanced democratic
stability.
The IAC and the Fujimori experience may also have produced a leg-
acy of greater societal resistance to authoritarianism and rights abuse.
As Maritza Paredes argues in chapter 7, one “unexpected legacy” of
the IAC was the emergence of a vibrant human rights movement with
“wide international networks of support to defend citizens from inter-

Soifer_6844-final.indb 325 8/17/18 11:53 AM

 326 Steven Levitsky

nal war injustices” (see also chapter 2). These organizations not only
survived Fujimori’s fall but expanded their activities, embracing, for
example, the cause of indigenous and environmental rights.
There is also some evidence that the violence and authoritarianism
of the 1990s may have strengthened liberal democratic values within
an important sector of Peruvian society. As chapters 10 and 11 make
clear, Peruvians remain deeply divided over how to interpret the vio-
lence of the past and how to respond to contemporary and future vi-
olence. As Drinot discusses in chapter 11, many Peruvians continue
to embrace the authoritarian “salvation” memory associated with fu-
jimorismo. As Maldonado, Merolla, and Zechmeister point out in
chapter 10, these citizens are more likely to fear terrorism—and to be
willing to cede civil liberties to combat it. At the same time, however,
another sector of society emerged from the 1990s with a strengthened
commitment to human rights and liberal democracy. In their survey
experiment examining Peruvians’ responses to exposure to domestic
terrorism, Maldonado et al. found an “important public opinion di-
vide.” Whereas fujimorista supporters were more likely to respond to
terrorist threats by adopting intolerant, pro-“iron-fist” views, non-
fujimoristas displayed a “tendency to rally around democracy” when
exposed to domestic terror. In other words, non-fujimoristas grew
more liberal when confronted with IAC-like conditions.
Arguably, then, the “contested memory” that emerged out of the
IAC had a mixed effect on democracy and human rights. Due to a
combination of conservative resistance and state indifference, the “of-
ficial” memory associated with the Truth and Reconciliation Commis-
sion never achieved predominance in Peruvian society (see chapter 11).
Nevertheless, it took hold in an important sector of society. Many
contemporary Peruvian citizens place considerable importance on is-
sues of human and civil rights. As a result, government actions that
threaten those rights, such as efforts to reverse human rights prosecu-
tions and instances of state repression (such as in Bagua in 2008 and
Espinar in 2012) have often generated strong public opposition. And
the electoral ascent of fujimorismo has triggered the emergence of a vi-
brant anti-fujimorista coalition—a heterogeneous front, united by po-
litical liberalism, that mobilized to defeat Keiko Fujimori in 2011 and
2016. Whether this latent coalition can sustain Peru’s fragile liberal
democracy in the face of its many structural challenges, however, re-
mains uncertain.

Soifer_6844-final.indb 326 8/17/18 11:53 AM

 Conclusion 327

Notes

1. On the crisis of the Peruvian Left in the 1980s and 1990s, see Roberts
1998.
2. Many leftists backed successful presidential candidate Ollanta Humala
in 2011, but Humala ran under his own Nationalist Party label and broke
with the Left soon after taking office. In 2016 leftists were a minor part of his
coalition and broke with him soon after he assumed office.
3. As Dargent and Chávez note in chapter 5, the Shining Path reportedly
killed or disappeared 31 students, while 118 university student deaths during
the IAC have been attributed to state security forces.
4. Although APRA was a center-left party during the 1980s, it grew more
conservative in the post-Fujimori period, such that Vergara and Encinas in-
clude it as part of their “conservative archipelago.”
5. Ipsos Apoyo survey, January 2006.
6. GFK survey, June 18–19, 2013.
7. Indeed, the FP’s greatest electoral success in 2011 and 2016 came along
Peru’s north coast, where perceptions of crime were highest.
8. For a similar argument applied to the Chilean case, see Kurtz 2004.

Soifer_6844-final.indb 327 8/17/18 11:53 AM