0% found this document useful (0 votes)

57 views48 pages

Kurtz Cap 2

This chapter discusses previous theories on state building and their limitations. It aims to address gaps in understanding state building outcomes in Latin America by examining the underlying social and political dynamics, rather than just ecological, institutional or international factors. The chapter reviews common approaches around conflict, resources and institutions, and argues a theory integrating social foundations is needed to explain state building variations in late developing countries.

Uploaded by

Valentina Gómez

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

57 views48 pages

Kurtz Cap 2

Uploaded by

Valentina Gómez

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 48

2

The Social Foundations of State Building

in the Contemporary Era

Chapter 1 posed a broad theoretical question: what accounts for the com-
parative success and failure of state-building efforts on the South American
continent? Empirically, long-run South American state-building outcomes pro-
vide a series of paradoxical results that must be accommodated in any viable
answer to this question. First, the strongest of state institutions emerged in
some of the most unlikely of settings: the impoverished, colonial backwaters
of Chile and Uruguay. At the same time, the wealthier colonial centers of Peru
and Argentina (and Mexico in the north), despite higher levels of human capi-
tal (and thus larger pools of literate, skilled individuals available to staff their
bureaucratic structures) and a legacy of much deeper institutional development
from the era of Spanish colonialism, ultimately produced much less successful
public administrations. While Argentina initially, if belatedly, began the devel-
opment of strong national institutions at the end of the nineteenth century, by
the 1940s, institutional development had devolved into a near-irreconcilable
conflict over the proper scope and role of the state that left its institutions weak
and subject to persistent cycles of populist expansion and antipopulist retrench-
ment. And despite a comparatively much higher level of economic development
than the rest of the continent, and thus in principle more resources with which
to build institutions and manage conflict, a durable institutional settlement
was never achieved. Finally, Peru began its independent life as a weakly institu-
tionalized state, characterized by administrative inefficacy and persistent civil
conflict. And despite repeated efforts to alter the fundamental structure of its
governmental institutions, it remained a continental laggard in this regard. This
range of variation is quite typical of differences among the state administrative
structures found in the late developing nations, and it begs critical questions as
to what brings it about and why the relative differences tend to persist for long
periods of time. It is to that task that this book now turns.

This is far from the first book to wrestle with the causes of state building
and state atrophy.1 Indeed, there is an extensive and compelling literature that
has considered, for example, the ways in which interstate warfare can serve
as a critical impetus to the development of strong, penetrating, and effective
state institutions. The role of conflict has varied in such accounts, ranging
from a direct cause of institutional development to its emergence as a quasi-
functionalist response to an existential external threat to an evolutionary selec-
tion mechanism that removes those states that do not successfully adapt to the
demands of conflict – by rationalizing their bureaucratic structures, creating
modern systems of taxation, or improving their property rights structure, inter
alia – through defeat and dismemberment. Another literature, oddly discon-
nected from the first, has instead considered the causes of state institutional
failure, highlighting the role of large stocks of easily taxed natural resources
as a paradoxical foundation for weakness. In such ecological accounts, large
natural resource endowments, most notably oil, make it possible through the
windfall revenues they are said to provide for states to avoid the difficult tasks of
administrative modernization and domestic taxation. Still others have focused
on a variety of questions of institutional design, suggesting that the incentives
that arise from alternative structures – whether political systems are unitary
or federal, whether local government was directly ruled from the center or
involved the mediation of local elites, or the rules of electoral competition, for
example – induce or impede the strengthening of states over long periods of
time (see, e.g., Soifer 2006; Saylor 2008). And finally, some have emphasized
the role of culture and values in the construction of effective state institutions,
either in the vein of the hoary modernity–tradition dichotomy of moderniza-
tion theory or more Weberian-inspired accounts of the rise of rationality and
bureaucracy in ways linked to particular strands of European Protestantism
(e.g., Gorski 2003).
It is also crucial to differentiate the goals of this book from other work
with which it bears some structural similarities, but also important differences.
While the task here is to examine the dynamics of political development, this
is meant in the specific sense of institution building and administrative reform.
This is not a book that is directly concerned with the origins of distinct political
regimes (even if they do, on occasion, have indirect implications for the claims
made here with respect to state building). And while this book shares some
structural similarity and an emphasis on path dependency in explanation with
the magisterial work of Collier and Collier (2002), including in part a focus
on a labor and popular sector incorporation in the Depression era, both the
specific meaning of incorporation and the causal dynamics that follow from
it differ quite substantially from the usage in that work. This is, of course, to

1 That said, this is a literature that has overwhelmingly (but not entirely) focused on the earliest
examples of effective national state formation and bureaucratic rationalization in central and
northern Europe.

be expected given that the focus here is on administrative institutions that are
very often conserved in strength and reach even as political regimes change
with time, and which are cross-nationally not obviously correlated democratic
or authoritarian regime outcomes.

the state-building debate: what we do and do not know

This book was spurred by the conviction that there are very serious lacunae in
the conventional approaches to understanding national political development.
It was also written in part out of a concern that the passing of world-historical
time has in fundamental ways transformed the challenge facing new states as
they endeavor to build and consolidate administrative institutions. Although –
as will see later – there are important insights to be gained from the tradi-
tional examinations of the state-building process that highlight, for example,
conflict, ecology, or institutions, it is not until these are married to an analysis
of underlying social and political dynamics that a truly compelling account
of contemporary Latin American political development can be constructed, or
one that successfully confronts the new empirical puzzles laid out in the con-
trasting patterns of political development and underdevelopment in the “late
developing” parts of the world. In the section that follows, the insights and
inadequacies of the extant approaches to the understanding of state building
will be examined with an eye to making the case for the centrality of social
dynamics, rather than ecological, institutional, or international-systemic fac-
tors, in laying the long-run foundations of institutional order.
The most well-developed theoretical literatures on state building and decay –
emphasizing conflict as the motor of institution building and resource wealth as
the cause of institutional corrosion – have developed in near-complete isolation
from each other, even as they seek to explain opposite sides of the same theoreti-
cal coin and suffer from similar methodological problems. To begin, some of the
theoretical claims of prominent bellicist and resource curse accounts of polit-
ical development suffer from incomplete specification of their central causal
mechanisms. For example, often missing is a direct explanation of the factors
that determine when and if war or resource wealth is a blessing or a curse.
Instead, institutional development or atrophy is sometimes attributed to the
demands of a strategic or ecological context in a quasi-functionalist sense. This
theoretical strategy, however, must confront the reality that many resource-
wealthy states have sharply improved their governmental institutions –
including some of the wealthiest, most administratively strong states in the
world, such as the United States, Norway, Canada, and Australia. Similarly,
warfare has far more commonly been associated with institutional collapse
than with long-run state building. The approach taken in this book will be
to develop a society-centric account of institutional development that can also
elucidate the conditions under which war and resource wealth do (and do not)
have important consequences for institutional development.

War and Political Development

It is commonplace to view war and/or the threat of war as a force driving
the strengthening of states, or even, for that matter, the emergence of the
nation-state as the dominant form of political organization (e.g., Tilly 1990;
Centeno 2002; Hui 2005). This line of argument traces its roots at least to
the early-twentieth-century writings of Otto Hintze (1975, 183, 199). Mili-
tary competition in this conception drove increases in the tax capacity of the
state and, over the longer term, forced institutional changes that facilitated
economic modernization and industrialization (Hintze 1975, 201). It also, as
a consequence of these efforts, ultimately created state institutions that were
robust enough to place demands even on powerful societal actors.
Tilly (1990, 20) has among the best statements of this perspective, arguing
that “preparation for war . . . involves rulers ineluctably in extraction. War
builds up an infrastructure of taxation, supply, and administration that itself
requires maintenance and often grows faster than the armies and navies that
it serves.” Whether and how this results in the construction of strong national
states depends heavily, however, on the conflicts that emerge between the state
and society over the terms of this extraction, thus implicating social structure
critically in the charting of the course of national destiny.2 International compe-
tition and the diffusion of organizational knowledge eventually made possible
a convergence around what Tilly calls “capitalized coercion” in national–state
administrative form – because of its comparative advantage in military conflict.
Nevertheless, the different roads to this outcome left behind decidedly different
public institutions. In this reading, strategic position is basically prior; political
and social conflict merely shape the specific trajectory of underlying political
development induced by it.
Conflict-centric approaches have formed the dominant perspective on the
formation of strong and effective national states in Europe. For example, Porter
(1994) echoes Tilly in highlighting state development as a competitive response
to military–strategic innovations outside national borders. He shows greater
sensitivity, however, to the possibility that states may simply fail to respond
effectively to external challenges, a fate, for example, that befell Poland during
the Great Northern War (1700–21) and resulted in its dismemberment as an
autonomous political entity. Inevitability is thus removed, but without a clear
specification of the causal processes that determine which states will effectively
respond to external challenges from a conflictual environment. Spruyt (1994)

2 Tilly discusses this in terms of three possibilities derived from differing patterns of state–society
interaction. In coercion-intensive situations, rulers could subdue society (especially the nobility)
and erected massive extractive structures. In capital-intensive systems, rulers were at the mercy
of the voluntary contributions of capitalists for the funds through which to construct mercenary
armies for defense or conquest. Where holders of power and capital were relatively well balanced,
he argues, capitalized coercion emerged, in which domestic elites were incorporated into the
leadership of newly constructed national states that in part reflected their interests.

goes one step further and argues that state formation is not simply a product of
the Darwinian struggle for survival in a world of intense military competition;
other often nonviolent mechanisms, such as international exchange or mimicry,
could be at work as well. Others contend that the absence of substantial inter-
state war makes African or Latin American state building unlikely at best
(Herbst 2000; Centeno 1997, 2002). The converse has also been argued – that
sustained peace will likely at least partially erode the scope and/or cohesiveness
of state institutions (Desch 1996). As, however, the prevalence of interstate
war of a scale sufficient to threaten national sovereignty declines into near-
nonexistence (Mueller 2009), bellicist perspectives tend to produce distressing
and distressingly invariant predictions about the prospects for state building in
the contemporary era.
And even in the anarchic contexts for which they were developed, many
bellicist accounts are missing a clear causal mechanism that would distinguish
success from failure. Why do elites necessarily respond to external threat in an
institution-building fashion? Or, if effective response is not taken as inevitable,
what explains the path taken? Without clear answers to such questions, hypoth-
esis testing is necessarily incomplete, and evidence can be at best only loosely
consistent with any account. Indeed, where careful attention is paid to causal
mechanisms within the bellicist camp, further complexities emerge. For exam-
ple, Cohen et al. (1981) have suggested that strategic conflict may be initiated
by states to build a more powerful foundation for political control over society.
This raises two very interesting issues: (1) conflict’s effects may be asymmetrical
based on state vulnerability – is it external threat or opportunity for conquest
that matters? – and (2) conflict may thus become endogenous; it may be pur-
posefully pursued by modernizing, authority-centralizing elites to make state
building politically viable rather than being a (exogenous) force inducing an
institution-building response.
One excellent response to some of these problems is that of Hui (2004),
who begins by thinking systematically about the different ways in which
states may respond to external threats. She points out that their responses
can include immediate self-defense expedients that are self-strengthening (that
enhance administrative capacity over the longer term) or self-weakening (mobi-
lizing “immediate resource holders” but at a long-term institutional cost).
This approach implies that conflict is of continued theoretical importance,
but without the implicit teleology of other works. A key issue still not fully
answered is to specify the political conditions that shape the trajectories
that states take in response to external threats. Implicitly, this is a question
of where public resources come from, for military–strategic defense is very
costly. The resulting patterns of revenue extraction have been tightly linked
to state-building outcomes. It is here that a connection to the resource curse
literature is manifest, for it is in its purported effects on the revenue system
that natural resource wealth may induce pernicious institutional and political
dynamics.

Natural Resource Rents and Politics

It is not hard to see the prima facie case that underlies the connection made
between natural resource abundance, the creation of “rentier states,” and con-
sequent political and institutional calamity. After all, massive resource wealth
in Zambia and Congo has been associated with moderate to extreme politi-
cal chaos and decades of economic decline. Even greater rents accruing to oil
exporters after the OPEC embargoes of the 1970s have become associated with
institutional chaos in Venezuela and Indonesia; civil insurrection in Algeria;
and instability, underdevelopment, and bloated but ineffective states across the
Middle East. At the same time, Japan, Taiwan, and Korea, among the states
most widely recognized for their exceptional ability to guide their societies and
economies and the unusual efficacy of their bureaucracies, are contexts usually
seen as largely devoid of valuable natural resources.
This conventional wisdom may not be as plausible as it at first appears.
The logic of the rentier state thesis holds that the presence of large mineral
rents induces a series of pernicious outcomes in the state. Public resources are
redirected to inefficient and costly consumer subsidies, to the creation of “white
elephant” industries, and to the construction of patronage and clientelistic
networks to sustain elites in power. These practices are said to induce private
agents to reallocate their activity away from efficient production and toward
the conquest of rents, often resulting in the societal capture of state institutions.
And finally, because of the availability of “easy money” in the natural resource
sector, the very administrative capacity of the state – particularly that involving
the collection of taxes – is undermined, since the generation of “internal”
revenues is no longer necessary to fund ongoing activities. There is a close
affinity of causal mechanisms with the bellicist approach here: both take as
central to state building the creation of an effective and penetrating public
revenue bureaucracy.
The question is, however, where large resource rents are to be had, must
states fail, or do other possibilities exist? Here, again, the problem is one
of causal mechanisms: we need to know why easy wealth necessarily causes
political elites to make choices that spur institutional decay. If there are other
realistic choices to be made, especially if some of them lead eventually to the
construction of higher-quality bureaucracies and stronger states, the resource
curse thesis in its usual formulation is open to serious doubt. Instead of insisting
on the inevitably pernicious effects of natural resource rents, we must then
inquire into the factors that condition the use of such rents: when do they
underwrite stronger states, effective bureaucracies, and national integration,
and when do they induce predation, rent seeking, clientelism, and corruption
(e.g., Kurtz and Brooks 2011; Jones Luong and Weinthal 2010; Dunning 2008;
Karl 1997)? But if this is the case, and resource effects are not inevitable, the
critical causal factor is not natural resource abundance but rather the factors
that condition its use.

But indeed it is the case that quite a few major natural resource–intensive
political economies ended up producing very effective states characterized by
high levels of bureaucratic probity and deep institutional penetration of society.
Mineral wealth, for example, was and is central to the Australian political
economy from the colonial era forward; it was certainly dominant during the
critical period of state building. Natural resource exports – including ores, oil,
and renewable resources – did and do characterize the Canadian economy.
Massive and sudden oil revenues since the 1970s have not, by most accounts,
produced a marked decline in the quality of the British or Norwegian civil
services. And ultimately, even the United States at different points in its history
was heavily dependent on its enormous natural resource wealth. Indeed, with
respect to the last, Schrank (2004) has demonstrated that the effects of resource
abundance in the United States were critically shaped not by the resources
themselves but rather by the social property relationships in which they were
situated in different states.
The linkage in the literature between natural resource wealth and weak
states has been founded in part on conceptual and measurement confusions.
For scholars working in this paradigm, a common strategy has been to identify
resource-dependent countries by the ratio of natural resource exports to gross
domestic product (GDP) in the contemporary era. While at first glance, this
measure seems reasonable enough, it is not consonant with the concept implied
in theory – that access to easily collectible mineral rents promotes bloated,
patronage-producing state structures. It is access to easily extractable rents that
matters and not access specifically to foreign exchange – for citizens typically
can also be bribed, employed, or bought off in domestic currency (Stijns 2005,
2006). What matters, then, is production, not exports.
But the denominator is also problematic – it is not clear why the resource
share of GDP is appropriate as the point of the argument is about the amount
of natural resource wealth available for political distribution and not that there
is quite a lot of it relative to the size of the overall economy. Why would large
amounts of distributable wealth be less politically pernicious in a richer versus a
poorer country – as long as one makes no assumptions as to the inherent quality
of governance in relationship to national income? If both of these measurement
considerations are not taken into account, countries that have potentially very
large natural resource sectors, but that don’t tend to export them (instead they
consume them domestically), or countries that are quite well off would tend to
be coded as resource scarce, despite having enormous natural resource sectors
that produce large politically allocable rents. Indeed, if we consider the top 10
least corrupt governments as defined by Transparency International in 2012,
we find that at least four have enormous natural resource sectors (Canada,
Norway, New Zealand, and Australia), and Denmark is an energy exporter
with notable proven reserves.
This is only the beginning of the problem, however, for a form of what
Schrank (2004) has recently described as “temporal selection bias” is also

a potential cause of serious error. Where resources contribute positively to

economic growth, they are progressively depleted during the course of devel-
opment, leaving such states apparently resource poor later in their develop-
mental trajectory; but wealth did not inhibit growth.3 Indeed, the same logic
applies even if resource wealth is unrelated to economic growth – depletion
over time would still induce a false correlation between resource scarcity and
development. Moreover, if the measure of resource dependence used is a ratio
of resources to GDP, these states would appear to be decreasingly resource
dependent, even if the resource sectors were not depleted. And if economic and
political development are positively correlated, this temporal process will have
a tendency to produce strong false negatives with respect to the relationship
between resource wealth and state efficacy in contemporary-era cross-sectional
(or short–time series) analyses.4 Thus, even a better measure of the size of
the contemporary resource sector is not enough to permit a valid test of the
natural resources–political underdevelopment linkage. It becomes clear that
any such design must encompass a long historical sweep, and indications of
resource intensity of economic life must be considered at the start rather than
the culmination of the process of political development and state building.
The causal dynamics posited in resource curse arguments also warrant
scrutiny.5 If the claim is that natural resource rents induce the creation of “rent-
seeking societies” and corrupt bureaucracies, then this can only have meaning
if this outcome is more likely in their presence than in their absence. But it is
precisely this comparison that is seldom properly addressed – are developing
countries that lack natural resources any less likely to suffer endemic corruption
and clientelism? Should one be surprised that Haiti, Paraguay, and Kyrgyzstan
have ineffective states despite the fact that they are “blessed” with an absence
of massive exported natural resource wealth? Are they institutionally better
off than Ecuador, Bolivia, or Kazakhstan, which have such wealth? For that
matter, should our hopes for improvements in governance and public institu-
tions in Bolivia be raised given the collapse of its once-dominant tin industry
two decades ago, or should they decline as its vast natural gas reserves are
developed?6

3 E.g., the recent closure of the last coal mine in France should not suggest to us that France
developed economically because it is resource poor but rather that development is at least in
part the cause of its contemporary relative resource scarcity.
4 We could note, for example, that the United Kingdom was a very important coal and copper
producer early in its development; Germany had large coal and copper reserves that helped
make it the industrial powerhouse it is today. Neither are major coal producers today. Similarly,
the United States has a deep history of resource-induced development. See Wright and Czelusta
(2003, 2002).
5 Perhaps the best and most-nuanced treatment of the resource curse argument is that of Ross
(2012).
6 To put this in causal inference terms, what is the effect of oil wealth? Perhaps the best way
to approximate this is to examine dynamics over time within a single country, observing the
response of institutions to changes in resource discovery and exploitation patterns. Alternatively,

The literature linking natural resource wealth and political underdevelop-

ment is extensive and nuanced. But at its core, it turns on a simple and com-
pelling idea, summarized by Ross: in such contexts, politicians seize control
over the ability to distribute the natural resource windfall and then “divert state
assets into patronage, corruption, and pork barrel funding . . . [and] once they
hold the ability to reshape resource institutions to their advantage, they may
use the opportunity to create additional, allocable rents to meet their patron-
age and corruption needs” (Ross 2001, 35–37). The implication is that rents,
once available, help consolidate political control, which then makes possible
the further corruption of the state bureaucracy in a downward spiral of mal-
governance, slow growth, cronyism, and patronage politics. Natural resource
wealth is said to have associated effects on public institutions, undermining
the resource extractive (tax) side of the state while permitting the construction
of an oversized but often-captured distributive apparatus. The implication of
both dynamics – though not always explicitly stated – is that where rents are
not available, politicians must use other means to maintain support, includ-
ing modernization of the state and economy or the provision of public goods.
Similarly, the hard work of constructing effective institutions of taxation to
provide necessary revenue both requires strong penetration of society and, as
a consequence, induces citizens to demand accountability if it does not, for
example, provide commensurate improvements in public services (Huntington
1991, 65; Anderson 1987; Ross 2004).7 Both contentions are functionalist,
and the implicit causal mechanisms are underspecified; it is unclear why politi-
cians will steal or misuse rents simply because they exist, nor is it obvious why
the absence of wealth induces improvements in performance (or if it doesn’t
always, what conditions whether it will). Nor are citizen demands for account-
ability inevitable simply because the efficacy of taxation improves. Moreover,
sustained failure would seem all too real an option where resources are scarce.
And effective use of resource wealth (or self-enrichment through mechanisms
that also provide public goods) is a possibility as well.
The most powerful version of this argument, however, sidesteps the question
of elite beliefs or state capture altogether. Instead, the claim is made that because
the natural resource provides the state with a large and readily taxable stream
of wealth, normal institutions of taxation become small and weak relative to
the development of the national economy (Chaudhry 1989, 143; Shafer 1994,
13). Because revenue agencies are understood to be costly (in political and
economic terms) to construct, states will not build such institutions unless

one could consider closely matched countries that differ over the presence of a major resource
sector, though the problem is that appropriate matches are decidedly hard to come by. An
excellent recent treatment utilizing the first strategy in a cross-national econometric frame has
already cast serious doubt on a related conventional wisdom linking natural resource (and oil)
wealth to authoritarian political regime outcomes (Haber and Menaldo 2011).
7 Indeed, such bargains have a long history, as in the promulgation of the Magna Carta, which
ultimately led to consultation of Parliament before the levying of new taxes in England.

they have to. Chaudhry (1994, 5; see also Chaudhry 1989, 107) does add a
qualification that this is likely the case only when “the construction of basic
state institutions coincides with large inflows of external capital [emphasis
added].” Sequencing is thus crucial, and here the claim is that natural resource
wealth undermines the process of state building only at its inception. Already
institutionally well-developed states thus are not likely to suffer from newly
found wealth to anywhere near the same extent. This is consistent with Karl’s
contention that natural resource flows were institutionally corrosive in the cases
she examined, except in Norway, where strong state institutions predated its
oil boom (Karl 1997, 13). A critical implication of both scholars’ work is that
the effect of resource wealth on political development is profound only at a
critical juncture early in the state-building process. Of course, this does not
solve the conundrum of the emergence of very effective states in contexts that
have always had a very important natural resource sector, as in the United
States, Canada, or Australia. Nor does it really address the micrologic behind
this distinction: why wouldn’t politicians in states that come to resource wealth
late have incentives to use it for patronage or populist purposes and to sharply
curtail taxation?8
Where resource-curse scholars do try to fill in the causal mechanisms in their
account, they come to strikingly opposed conclusions over whether an abun-
dance of resource wealth produces an excessively strong society or state. Some
argue that such wealth produces a set of social actors – especially entrepreneurs
in the export sector – that captures public institutions, renders the state inca-
pable of pursuing a coherent “national interest,” and makes it wholly depen-
dent on the success of the mineral sector (Shafer 1994; Karl 1997). Others
suggest that it is typically the state and not society that is too strong, arguing
that the availability of vast sums permits politicians to displace private-sector
elites or make them dependent on public largesse, while the high sunk costs
associated with natural resource (especially mineral) extraction can weaken
business in relationship to the state over time (Moran 1974, chapter 6; Vernon
1977, 171; Chaudhry 1989, 1994). These twin causal mechanisms that under-
lie resource-curse arguments are thus diametrically opposed. Either the access
to rents empowers the incorporating state, or it empowers the rent-seeking
society. The antinomical character of the posited causal mechanisms suggests
that resource-curse theory is closer to shorthand for an observed correlation
than for a causal relationship, and as we have seen, there is also substantial
room to doubt the validity even of the observed correlation.

8 Indeed, such arguments face difficulty explaining why, for example, the discovery of oil in
Norway did little to change the country’s status as one of the highest-taxation economies in the
world, while in Alaska, for example, contemporary oil discovery led to a radical decline in the
tax burden, from the twelfth highest in the nation in 1977 to the lowest in the nation every year
since 1981 (Tax Foundation 2011). Moreover, Alaska is one of two states in the United States
that have neither an income nor a statewide sales tax (Howe and Reeb 1997, 117).

Finally, both the simple and the path-dependent form of the resource curse
argument relies on the assumption that elites will misuse resources in an insti-
tutionally detrimental manner simply because they can. Even where elites are
fundamentally self-interested, this is not necessarily the case. They may well
pursue private gain or the provision of rents for their cronies, but this in
no way precludes them from doing so through the provision of the sorts of
public goods that can modernize the state over time. Politicians may, for exam-
ple, create and distribute rents by engaging in large-scale transportation and
communication infrastructure provision, educational expansion, and military
modernization. Even if the contracts for such work are awarded to cronies
of the government, or access to employment in such projects is distributed as
patronage, the long-term effects are the same: roads and rails are laid, schools
are built, pupils are educated, telegraphs and telephones unite the national
territory, and the coercive face of the state is strengthened.9 These are likely
to reduce domestic transaction costs, promote national integration, strengthen
national defense and/or make possible colonial expansion, raise levels of human
capital, increase the rate of economic development, and broaden the scope of
the central government. This in turn may ultimately create new social groups
with vested interests in the modernization of governance – from middle-sector
groups and public employees to new entrepreneurial elements that operate in
nonresource sectors to reform-minded military officers. Rent seeking thus does
not have to be absent for an influx of resource wealth to have salutary effects
on state building, if the way in which rents are distributed has the effect of
inducing changes favorable to state building as they accumulate over time.
The static and dynamic institutional effects of resource wealth can thus be
opposed.

Institutionalist Accounts
As the preceding makes clear, this book takes issue with some of the contentions
in the regnant literature on the long-run development or underdevelopment of
state institutions. Much more recently, an institutionalist account of state build-
ing in the Latin American region has been developed by Hillel Soifer (2006,
2011). By contrast with the ecological and bellicist approaches, Soifer’s insti-
tutionalist account is complementary to the society-centric theory developed
here.
This institutionalist approach begins with the observation that in the nine-
teenth century, Latin American states took two basic approaches to the estab-
lishment of governing institutions in their more peripheral regions. The imple-
mentation of state policies required the development of local structures of

9 It is crucial, however, that the rents be extracted via the actual provision of the public good
rather than through a wholesale process of theft wherein private actors receive contracts and
payments for infrastructural or public goods provision that is not in fact ever delivered.

administration and raised the question of how they would be staffed. In some
cases, Soifer (2011, 10ff.) points out that central states delegated author-
ity to local elites. In others, they deployed agents directly as bureaucratic
representatives of the central state. This in turn created a set of incentives
that had profound implications for national political development: delegated
authority relied on elites whose income sources were typically “from land-
holdings or other aspects of their status in the local community,” which in
turn made them less dependent on the central authorities and created impor-
tant principal–agent problems for governance. Conversely, deployed agents
not only relied on the center for their incomes but also had an incentive to
induce “cooperation with the effort to expand the reach of the state” (Soifer
2011, 10). This in turn, over time, aggregates into a pattern of local institu-
tional practices that either facilitate or impede state building over the long haul.
Deployed agents, thus, were much more compatible with effective institutional
development.
This is an important point, and these institutional choices likely played
the role assigned them by Soifer. The difference between the society-centric
approach taken here and Soifer’s institutional one has to do with the starting
point. Whereas Soifer treats these institutional developments as structured but
exogenous choices, the approach here would suggest that they are effectively
conditioned by underlying social realities as highlighted in this book. Elites may
alternatively select deployed versus delegated rule, but this is not necessarily a
meaningful choice. Delegated rule, from the perspective of the center, all but
guarantees difficulty in governing, collecting revenues, and mobilizing soldiers,
and as such, it is hard to imagine that any national executive would prefer it
to a deployment strategy. But deployed rule must overcome efforts by local
elites to resist it. Where the institutional account meets the societal account
presented in this book is on precisely this issue: the social relations of local pro-
duction will shape whether local elites will see the imposition of deployed rule
as an existential threat or a burden that may also have potentially quite lucra-
tive implications. That said, further exploration of the societal–institutional
interaction at the local level is very much a productive avenue for future
research.

What Should We Do Next?

Earlier, a series of theoretical and empirical lacunae in the existing scholarship
on state building and atrophy were identified, and theoretical advance will
require that some of these gaps be filled. But there are also additional challenges
that must be considered, particularly when considering the development of
national administrative institutions in the late developing countries, and it is
with this task that the next section begins. Thereafter, a theory of long-term
institutional development and decay will be presented that seeks at once to
supersede or subsume important portions of the existing literature, while at

the same time contending with the distinctive challenges involved in explaining
state-building outcomes in the postcolonial era.10

additional challenges for a theory of contemporary

state building
The problems in the extant literature on state building identified previously do
not exhaust the challenges that face any attempt to construct an alternative
theoretical account applicable to the experiences of state building in the period
after that of the canonical examples in Early Modern Europe. First, any such
account must clearly distinguish itself from explanations of long-run economic
outcomes, for at least in the late developers, political and economic modern-
ization do not necessarily occur in the same times and places. Second, it must
adequately address the issue of time itself, in two senses: political development
unfolds over a long period of time, and the relative hierarchies of institutional
strength that emerge tend to persist. That is, state building is a long-term pro-
cess that likely involves substantial path dependence. As important, attention
must also be paid to the ways in which the emergence and development of
governmental institutions in the “late institutionalizers”11 are quite different
from the conditions that prevailed in the earlier cases of state building. It is
inappropriate to assume – and empirically unlikely to be the case, as we shall
see – that the dynamics that long ago provoked institutional modernization
remain the principal drivers of the substantial variation we observe in the con-
temporary era. Causal heterogeneity by historical era is a distinct possibility,
as, for example, critical conditions in the global economy and the strategic
environment in which states find themselves have changed sharply.12

Economic versus Political Development?

Because so many (though notably, not all) advanced industrial countries are
thought to have comparatively effective bureaucracies and were by and large
located in the European cradle of economic and political development, there

10 This term is meant broadly and is understood to include both the cases of nineteenth-century
national liberation in the South American context and the twentieth-century decolonization and
institution-building experiences of Africa and parts of Asia. There are also obvious differences
among these contexts as well, but they share the experience of attempting to build national
institutions in a world in which powerful and wealthy states already exist.
11 This term is deliberately intended to evoke a parallel with the way in which late developers has
been used. The critical point here is that like economic development, the challenge of creating
effective administrative institutions is different for the countries not in the first wave of political
modernization.
12 Most notably, states are much more economically integrated in the context of late, and especially
late, late development, and the military struggle for national survival characteristic of the
anarchic state system of Early Modern Europe has all but abated in the post–World War II era
(and much earlier in places like Latin America).

has been a tendency to assume that economic and political development are
quite tightly related to each other. This is perhaps most clear in the economics
literature that largely conflates the development of effective governmental insti-
tutions with the emergence of strong or efficient property rights, the latter them-
selves seen as inherently growth producing (e.g., Acemoğlu et al. 2001). But
as long as we are careful to hew to an institutional as opposed to policy-based
notion of what state strength actually is, there is a much weaker empirical foun-
dation for this connection. This is not a claim that strong political institutions
cannot contribute to growth by effectively implementing appropriate economic
policy. They surely can. But the factors that determine the economic strategies
that states select are not the same as those that condition the strength of their
political and administrative institutions, and economic improvement can come
well before institutional development, if the latter comes at all (Glaeser et al.
2004, 298). For example, in terms of wealth and social development, Argentina
was for the most, if not all, of the twentieth century Latin America’s wealth-
iest country and among its best educated, most egalitarian, and least poverty
stricken. At the same time, its political regime was among the most unstable in
the region, and its governing institutions stand out for their comparative inef-
ficacy. Similar points could be made about the United States: very substantial
economic development took place in the Gilded Age after the Civil War, in
the context of a state that did not seek to govern the economy and a judiciary
that was regularly “subverted” because “money and power influenced the path
of justice” (Glaeser and Shleifer 2003, 404). For the first century of its exis-
tence, the American state was characterized by a “radical devolution of power
accompanied by a serviceable but unassuming national government,” and it
was only between 1877 and 1920 that national institutions “first emerged free
from the clutches of party domination, direct court supervision, and localistic
orientations” (Skowronek 1982, 15, 23). Einhorn (2000, 157; 2008) concurs,
pointing out that the signal characteristic of the American national state until
the twentieth century was its weakness.13
Similarly, as we shall see subsequently, Chile and Uruguay comparatively
early on developed effective, central systems of political administration. But
we will also find that the development of these institutions was intimately
connected to the construction of an economic policy regime predicated on
very high levels of public employment, government ownership of industry,
and the maintenance for over a half-century of a broad-based and high-level
tariff protection. The last, in particular, is a decidedly questionable policy
choice in the context of economies of such small scale. And even in Korea,
one of the paradigmatic cases connecting a highly effective state to spectacular
economic development in the postwar era, Cumings (2005, 223) points out
that governmental institutions had been highly centralized and bureaucratized

13 Indeed, Einhorn (2000, 160n25) notes that, outside of emergencies, the US national state “taxed
nothing but imported goods before the Civil War” and that the “tariff dominated until WWI.”

for a very long time – but it was only after 1960 that persistent sharp increases
in economic output were recorded.

The Question of Time

Meaningful state building, at least in most accounts, is something that takes
place over the long term. How long is of course not a matter of consensus,
nor need it be identical across cases. But the factors that condition the ability
of states to centralize and bureaucratize their public administrations do not
produce observable results overnight. Why is this the case? First, there is the
question of the state of existing knowledge, which was a principal issue for the
early modernizers. It is perhaps too easy to forget, but the structure of modern
bureaucratic organization had to be developed and perfected over a long period
of time – and then it had to become adopted into the administrative machinery
of government and ultimately prove its superiority in political, military, and
fiscal terms. Only after this had occurred could it serve as a credible model for
administrative reform. For this reason, Rueschemeyer contends that the emer-
gence of strong, modern bureaucratic states in northern Europe took nearly
a millennium, beginning with the foundation of the University of Bologna
and becoming completed only in the nineteenth century (Rueschemeyer 2005,
144). Indeed, the process of the development of organizational and manage-
rial knowledge is central to Ertman’s (1997, 28) discussion of the influence
of strategic conflict on institutional development – for the earliest moderniz-
ers had to increase the extractive capacity of their states when only feudal or
ecclesiastical models of large-scale organization were available, leading to a
reliance on decidedly suboptimal institutional forms like venal office and tax
farming. Those who faced this challenge later, conversely, had available other
forms of administration as well as a much larger supply of appropriately skilled
labor. And in a world of transaction costs and imperfect subjective models, the
diffusion and adoption even of institutions that provide obvious advantages in
governance is far from automatic or rapid (see, e.g., North 1990, 8).
The question of time becomes even more pressing in arguments about state
building that rely on processes of selection (in the evolutionary sense) – and
many do – to sort out “better” from “worse” institutions. This is of course
most central in the bellicist approach to state building, wherein it is the real-
ity of war and conquest that overwhelms states that do not improve their
administrative structures, thus over time transforming the states system and
inducing convergence around better practices. But any fundamental reorgani-
zation of state and society driven by such an evolutionary14 mechanism requires
a long time frame, for not all rivalries produce conflict, not all defeats imply

14 The term evolutionary is here understood in the natural science sense of change in a population
that is driven by a selection process that induces differential survival – in this case of institutional
forms. It is not meant in the colloquial sense of “gradual” or “incremental.”

dismemberment of states, and not all suboptimal institutional responses result

in defeat.15 But alternative selection mechanisms such as North’s (1981)
emphasis on economic competition or Spruyt’s (1994) focus on institutional
superiority of the sovereign state in making credible commitments and man-
aging interunit political interaction are logically even slower. For they do not
very directly implicate the survival of the national political unit and thus the
lives and fortunes of the elites who govern it.
But it is not enough to make the point that the development of strong insti-
tutions takes time – especially if their institutionalization requires either the
development of new normative structures (as in Rueschemeyer 2005), quasi-
voluntary compliance and cooperation (as in Levi 1989), the creation of pow-
erful supporting constituencies capable of carrying the institutions into the
future, or the creation of settler versus extractive colonial economies and insti-
tutional structures over the course of one or more centuries (as in Acemoğlu
et al. 2001). It not only thus takes time to build durable institutional struc-
tures, but we saw empirically in the opening chapter that the relative lev-
els of institutional development tend to be quite stable over long periods of
time.
This is consistent with a variety of theoretical approaches that have high-
lighted long-term stability of the hierarchy of institutional strength linked to
discrete moments of substantial change as a species of “punctuated equilib-
rium” (Spruyt 1994, 24). Indeed, Acemoğlu and Robinson (2006, 326) point
out that institutional stability is often maintained, even when some formal
political structures are transformed, through the emergence of countervailing
forms of informal de facto power that preserve the core of ex ante practices and
norms.16 Others have pointed out that institutions create vested interests that
tend to perpetuate their patterns of operation even when environmental factors
or changing social or technological structures make them decidedly suboptimal
(e.g., Ertman 2005, 168) or, alternatively, transaction costs and information
unavailability might drive the persistence of poor institutions (North 1990, 52).
While there is much debate as to the underlying causes for the observed pat-
tern of long-run institutional stability, there is a substantial group of scholars
who argue that the state-building process involves path dependencies – dif-
ferences at critical initial moments generate institutional patterns that become

15 This is an important point that is explicit in Hui’s (2004, 184) contention that states can respond
to international strategic rivalry with institution building and taxation or through long-term
self-defeating expedients like excessive indebtedness. But it is not at all difficult to see how
such expedients might help states to overcome an immediate crisis and delay still further the
selection effects involved in international conflict. Indeed, according to her, states utilizing such
expedients can even sometimes defeat competitors who adopt better approaches (though this is
probabilistically less likely).
16 E.g., they point out that the post–Civil War South remained in institutional terms not terribly
different from its antebellum patterns, with the replacement of chattel slavery by a more
informal, semiservile labor system enforced by the Ku Klux Klan and Jim Crow.

self-reinforcing over time. If this is the case – and it is the contention of this book
that it is – the further implication follows that the causal dynamics involved
in state building exhibit hysteresis. That is, the effects of particular indepen-
dent variables depend importantly on what has come before, and thus factors
that might have initiated a virtuous circle of state building at one point in the
development of national–state institutions will not necessarily have that effect
should they emerge later. Windows of opportunity, thus, can close.
The challenge thus is to produce a theory of institutional development and
underdevelopment that can account for long periods of stability punctuated by
periods of change in specific critical moments. It must also develop a theoretical
account that can explain why the factors that were hypothesized to induce a
particular trajectory are not determinative outside of these critical moments.
And finally, empirical evidence must be brought to bear that can evaluate a
path-dependent explanation in its own terms, compare it against explanations
taking a more traditional contemporaneous cause–effect form, and show that
timing is essential to the causes of state building such that subsequent efforts
to change institutional trajectory are ultimately powerfully impeded.

State Building in the Late Developers

At least since Gershenkron (1962), it has been argued that economic devel-
opment in “late developers” involved fundamentally different challenges from
those faced by the first generation to make the transition to modern industrial
capitalism.17 This book contends that a similar dynamic – distinctive challenges
of political development for late developers – characterizes the state-building
process, making it dangerous to rely too heavily on the foundational literature
that focused on the first states in Early Modern Europe. The differences that
are essential have to do with information and the powerful selection pressures
ascribed to interstate war.
The first difference – the global informational environment – represents a
clear “advantage of backwardness.” For the diffusion of administrative forms
allowed late developers to gain the advantages of institutional innovation with-
out paying the costs involved in experimentation, discovery, and repeated
reform. Indeed, by the nineteenth and twentieth centuries, it had become clear
that Weberian, meritocratic bureaucracy was a decided improvement on tra-
ditional feudal, ecclesiastical, or patrimonial alternative models of political
administration. These advantages can be seen in the ability of states to raise
revenues, solve principal–agent problems that impede governability, and raise
the efficiency (cost relative to output) of producing public goods. And by the
second half of the twentieth century, there was no shortage of international
institutions endeavoring to transmit this information, promote administrative

17 Indeed, Gershenkron’s (1962) argument suggested a set of complementary differences in the

development of state institutions that were linked to the challenge of late economic development.

reform, and even subsidize or punitively incentivize the transformation of cor-

rupt or patrimonial political systems through the provision or conditioning of
aid and loans.
The second major difference hinges on the causal status of interstate war.
The leading proponents of conflict-centric approaches to political development
highlight the effect war is said to have in forging elite unity, driving the forma-
tion of powerful revenue bureaucracies, and selecting out of the system those
states that do not adapt and focused their work almost exclusively on the period
of national state development in Europe. But as Hui (2005) has pointed out,
this was an almost unique environment of near-constant, severe conflict over
the course of centuries. Most regional state systems, in fact, have not been char-
acterized by anything like this degree of conflict. Some have argued, however,
that more common and more limited “interstate rivalries” might be sufficient
to drive state building (e.g., Thies 2005, 456; 2004).18 But as such, these the-
ories imply at most very limited actual conflict and minimally little more than
posturing and the display of force. It is hard to see therefore how they are
parallel in theoretical terms to the centuries of war-as-existential struggle that
characterized the European continent.
In the Latin American context, specifically, two features of the strategic
environment are notable. First, while there was some interstate conflict of
a severe character in the nineteenth century, most conflict was civil rather
than international. And by the twentieth century, interstate war had all but
ceased to exist in Latin America. Indeed, only one meaningful war occurred
in South America in the entirety of the twentieth century (the Chaco War
between Bolivia and Paraguay). This is a global trend that Mueller has force-
fully demonstrated, declaring boldly, “War has almost ceased to exist.” In
particular, he finds that interstate conflict that actually alters territorial bor-
ders (i.e., conquest and consequent state death for the losing party) has become
almost extinct (Mueller 2009, 307). And of course, it is precisely this last sort
of war that is hypothesized to drive state building in most bellicist interpreta-
tions.
This leaves us with two possibilities. Either the process of political develop-
ment has been radically undermined by the absence of interstate conflict – and
the pressure this takes off of states to modernize their bureaucracies, increase
their tax revenues, or better organize their administrative systems – or other
factors are principally driving the (large) variations in administrative capacity
to be found in the contemporary developing world. The theory developed next
is designed to do precisely the latter – explain the development of states in a
world in which war to the death is not the state of nature.

18 To qualify as an enduring rivalry for Thies, states need only engage in six militarized interstate
disputes (MIDs) within a 20-year period. MIDs in turn are defined in such a way that the
mere threat or display of military force is sufficient to constitute a militarized dispute. For the
complete operational definition, see Ghosn et al. (2004).

the social foundations of institutional order

For late developers, the initial critical juncture has a comparatively noncontro-
versial starting point. Decolonization and independence required the creation
of an initial set of national political institutions by Creole elites to replace
the administration of the former colonial metropole. This institutional settle-
ment – or a pattern of fragmentation and regionalism if elite agreement on
the basic structure of national government could not be achieved – would
powerfully condition the way subsequent institutional development unfolded.
This initial branching point in the construction of postindependence govern-
mental institutions would be followed by a second decisive moment as these
preexisting structures faced a powerful new challenge: electoral and/or insur-
rectionary mobilization that aimed to initiate the incorporation into political
competition of the representatives of nonelite citizens in the middle and work-
ing classes. This would introduce a further deflection of trajectories of state
building that would decisively condition the relative scope, depth, and efficacy
of South American governmental institutions into the contemporary era. In
the sections that follow, the theoretical argument undergirding this two-stage,
path-dependent explanation will be laid out.

Theory I: Launching a State-Building Trajectory

What we first focus on, then, are the conditions that shape the developmental
trajectories that states face at their birth as independent national-state units.
In the Latin American region, this first critical juncture took place early in
the process of economic modernization of all the countries in the region. As
a consequence, agriculture and mining were typically the dominant economic
sectors, for industrialization was at best small scale and limited in extent. And
it would be the social relations that govern the operation of agriculture – where
the bulk of the population found employment – that were thus critical to the
possibility of effective state building.
But what is the linkage between social relations and institutional develop-
ment? This book argues that where a local elite organizes a labor-repressive
agrarian economy, effective political development, even in the face of war or
wealth, is very unlikely.19 But free labor is not sufficient to produce a positive
institutional trajectory. A second condition is required; that is, the incorpo-
ration of upper-class actors from all major factions into the national political

19 Such an elite could be a settler population, a traditional feudal aristocracy, or simply a locally
dominant upper class. The notion that rural social relations importantly condition long-term
political outcomes is not new, being central to Moore’s (1966) discussion of regime dynamics
and Brenner’s (1977, 1987) explanation of European economic development; see also Aston and
Philpin (1987). The emphasis here on the role that free vs. unfree labor has in the development
of state institutions is inspired directly from the recent pathbreaking work of Schrank (2004)
that considers the role of social structure and resource wealth in US economic development.

system – while maintaining the exclusion of the popular sectors – is crucial

to enabling cooperation in state building and public goods provision activities,
despite whatever cleavages might otherwise divide them. When both conditions
obtain, and this sometimes came about only after a substantial period of con-
flict, elites will come together to centralize power in national institutions that
are able to penetrate and control society to an ever-increasing extent, while
simultaneously building national infrastructure, imposing substantial taxation,
and providing substantial collective and individual benefits to the oligarchy
that created them. An overview of this part of the argument can be found in
Figure 2.1a. But why are these two conditions critical?

Social Relations. Hearkening back to Moore (1966), I contend that the absence
of labor-repressive relations of production is critical to the long-run develop-
ment of, not regimes, but effective political institutions. This distinction is
emphatically not between capitalist and noncapitalist forms of production, for
there are quite a few ways of organizing the economy that do not involve
the exchange of land, labor, and capital on markets but that nevertheless do
not require labor coercion (e.g., labor service tenancy, sharecropping, common
property systems, or cooperative forms of production, inter alia). The emphasis
here is specifically on the mechanism by which labor is recruited and employed
in production – is it an inherently coercive process, or can workers leave the
farms or mines free of any practical or de jure encumbrances? Where rural
elites recruit their workforces through servile or semiservile means (through, for
example, slavery, formal serfdom, indenture, labor corvées, or legally enforced
debt peonage), administrative centralization and state building are decidedly
unlikely.
Why is this the case? Labor-repressive systems leave local elites extremely
vulnerable to the centralization of authority, as they imply the possibility that
national governments may eventually be unable or unwilling to maintain the
strict social control and labor repression on which the agrarian political econ-
omy (and potentially elites’ own physical survival) depends. Even more crit-
ically, in such contexts, agrarian elites will also resist the taxation necessary
to support even military modernization or collective defense against external
aggression, to say nothing of paying the costs for other central state institutions.
Labor-repressive economic systems are of notoriously low productivity (giving
elites little margin for added fiscal contribution) and are inauspicious settings
for most productivity-enhancing investments. Coerced labor forces rarely effi-
ciently adapt and employ improved technologies, and elites are in any event very
unlikely to permit (or pay for) the generalized educational improvement neces-
sary to support such modernization. To do so would empower workers whose
disempowerment is the cornerstone of economic production. And finally, there
is the distributional question: ceding taxation authority and control over coer-
cion to the center would be particularly risky in contexts where urban interests,
competing regions, or other social classes might gain control or influence. They

Downloaded from https:/www.cambridge.org/core. UCL, Institute of Education, on 09 Apr 2017 at 14:38:29, subject to the Cambridge Core
terms of use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139019668.002
(a) Inter-EIite
Substantial Rewards for
National Institution
Institutional Investments;
Cooperation Building Trajectory
Lengthened Time Horizons
Chile
Free Labor Uruguay, after 1876
Economy Argentina, after 1881
Prussia,
after 1850

Inter-EIite Cooperation Possible Local Institution

Conflict Locally; Inter-regional Building
Conflict
Uruguay, before 1876
Post- Argentina, before 1881
Independence
Political
Economy

Servile Economy Limits Constrained

Inter-EIite
Resources for Investment; Central
Cooperation
Local Autonomy Critical State Building

Prussia, before
1810-50
Servile Labor
Economy

Inter-EIite Institution Building Risky; State

Conflict Short Time Horizons Institutional
Atrophy

Peru

Economic and Political Conditions Political Dynamics Initial Institutional Trajectory

(b) Middle- and Working-

Long-Run
Class Party Cooperation;
es e
n

State Building
pr r th
sio

Consensus Exists
De fte

for Statist Development

at /A

Chile
re g
G urin

National Institution Uruguay

Building Trajectory

Chile Emergence of
Uruguay, after 1876 the Social
Argentina, after 1881 Question
B
De efor
pr e t
es he
sio G Middle- and Working- Populist/Anti-
n rea Populist Institutional
t Class Party Conflict;
No Consensus Exists for Cycling
Statist Development
Argentina
es e
n

N/A
pr r th
sio
De fte
at /A
re g
G urin
D

State Emergence of
Institutional the Social
Atrophy Question
B
De efor
pr e t Weak Popular-Sector; Endemic
Peru es he
sio G Endemic Conflict between Institutional
n re at Middle- and Popular-Sector Weakness
Parties; Extended Liberalism
in the Political Economy
Peru

Initial Trajectory Second Critical Juncture Political Dynamics Long-Run Outcome

figure 2.1. The argument in brief: (a) the first critical juncture; (b) the second critical
juncture.

might well, after all, form distributive or redistributive coalitions that exclude
agrarian elites or some subset of them and would thus immediately threaten
the rural social order.
This last threat is quite plausible, for urban-oriented governments have often
moved to incorporate political actors beyond the elite strata as part of a strat-
egy to build the political foundations for industrialization – which generally
would require economic and social reforms anathema to rural elites, includ-
ing, most notably, land reforms and direct resource transfers. Or more likely,
because Latin American governments historically have overwhelmingly rep-
resented the interests of the powerful, agrarian elites might simply have to
contend with efforts designed to promote industrialization at the rural sector’s
expense. Industrialization requires workers, and profitability in nascent indus-
tries requires low wages, an urban labor surplus, and cheap food. In servile
systems with labor effectively bound to the countryside and to low-productivity
agriculture, no such labor force can be recruited, and food prices remain com-
paratively high. Thus even capitalist industrialization threatens such agrarian
systems with radical transformation. And if industrialization requires foreign
exchange that can only be obtained from the export earnings of the agricultural
sector, permitting centralization and institution building would also detonate a
straightforward sectoral clash – as the state seeks to transfer foreign exchange
earnings from agro-exporters to local industrialists.
Where labor is not servile – and this must be considered broadly to include
any form of tenure in which a peasant has the practical legal right to leave
the farm – the development of a modern army and the centralization of power
are not life-or-death threats for local elites. Such situations would naturally
include agrarian social structures characterized by independent family farms
or capitalist agribusinesses employing wage labor but also noncapitalist, tradi-
tional large holdings staffed by peasants sharecropping, working for usufruct,
or working as labor service tenants, as long as they had the legal ability to
depart should they so choose. Some, but certainly not all, such settings have
also been linked to better economic performance and at a minimum greater effi-
ciency in the allocation of labor at least in comparison to servile systems. Elites
in free-labor settings are frequently accustomed to paying something approach-
ing market prices for labor, but that said, they also have the real potential to
invest in labor-saving technology should such costs become onerous; these are
investments that are irrational in servile systems. And, critically, military service
by peasants would not risk the social instability it can bring to servile political
economies when peasants, with newly acquired skills in warfare, return.

National Elite Politics. Although rural politics and rural elites are important,
they do not alone define the amenability of a polity to successful state-building
processes. Equally important are the linkages between the state executive and
the upper classes as a whole. The critical factor here is that elites beyond the
governing faction must have achieved meaningful political incorporation – and

a share in political power. Where central authorities are at odds with pow-
erful regional or local elites, or where the central state is dependent on the
tax-collecting power of provincial strongmen or holders of venal office, polit-
ical centralization will be blocked, and the expanded resource extraction that
drives both state building and human capital formation cannot be imposed; it
founders on the short-term self-interest of uncoordinated elites. But where elites
are incorporated – either through cooperation in the form of an “oligarchic
democracy” or through imposition in more absolutist bureaucratic settings –
their collective interests can be organized within the state, and even difficult
choices that have principally longer-term payoffs can be made. This is partic-
ularly true when they are made in the context of external challenges that, if
unmet, would threaten the collective survival of the elite itself. My argument
does not imply that state building is impossible (or even less likely) in federal
systems, for the existence of subnational political units does not itself imply
that local elites are not simultaneously effectively incorporated in the overar-
ching national political system. The key question is whether regional interests
are meaningfully included in national political institutions in ways that don’t
permanently benefit or marginalize one elite subgroup or region.20
What matters is whether the interests of the central state and those of pow-
erful regional or local elites are harmonized or at least not directly threatening
to each other. An absolutist road to political development would do so by mak-
ing status as a member of a local elite dependent on administrative appoint-
ment, salaried employment, or military service, as defined by the central state
(this could be Soifer’s [2006] “deployed” strategy or militarized authoritarian
dependence as in Vormärz Prussia). In such contexts, the center has discre-
tion over who has access to such employment – positions are neither freely
purchased, nor held as individual property, nor inherited – and the resources
supporting the positions (and thus their status and power) come from the cen-
ter. The alternative path, more typical of Latin America, to elite coordination
involves a much less powerful central executive – one in which local elites effec-
tively hold shared sway. This is the case in “oligarchic democracies” where the
central state becomes the venue in which the collective interests of the upper
classes are defined and articulated, and as such it can credibly be seen as a
provider of collective goods for the elite without simultaneously threatening
its independent authority – by sharing power across factions and/or credibly
enabling the alternation of political control. This is most feasible when con-
tending elite factions and parties are at least somewhat heterogeneous, and thus
cross-cutting, in their sectoral and/or geographic composition. Should elites
manage this, they would then form what McAdam (1982, 38–39), following

20 While one could hypothesize that federal systems might be more prone to intractable conflicts
among competing (regional) elite groups because they reify regional differences, this remains
an interesting empirical question that should be addressed.

Gamson (1968), called a “competitive establishment.”21 However, much as

they may be at odds with each other over some of the issues of the day, they are
united in their desire to avoid the expansion of the political arena to include
subalterns and favor using the state to provide public goods broadly beneficial
to elite interests. Importantly, the provision of public goods can also be used to
generate private benefits – through the allocation, for example, of government
contracts and the distribution of resulting employment through patronage and
clientelistic channels. And in this way, selective incentives can also come to
bear to mitigate the potential free riding of elite beneficiaries of these public
goods. Of course, an often substantial part of the cost of the provision of these
public goods can also be imposed on the middle and working classes (who have
little say in selecting what is provided) through regressive forms of taxation.
Where elite inclusion and popular sector exclusion obtain, the time horizons
of politicians can lengthen, and policy choices that at once enrich the upper
classes but also produce long-run developmental benefits (e.g., expansion of rail
infrastructure, education, economic promotion, or industrialization) become
more likely. Critically, political bargains that create institutions that undergird
elite cooperation are important in this context. They embody an effort to
create an intertemporal agreement that prevents the extreme abuse of power
by incumbents and thereby facilitates longer-run decisions that create policies
that are “more effective, more sustainable, and more flexible” (Spiller et al.
2008, 6) than where such durable bargains cannot be struck.
Of course, this is generally possible only where the political system is effec-
tively and exclusively dominated by upper-class interests, ensuring that newly
empowered public institutions cannot be used in a redistributive way by middle-
class or popular-sector actors. If democratization occurs early and nonelite
social classes achieve real influence, this interelite cooperation around institu-
tional development may quickly break down, and the political and material
resources necessary to underwrite state-building efforts will not be forthcom-
ing. None of this implies an absence of political competition – indeed, the
struggles among contending elite factions are frequently vigorous. But once
real mass suffrage becomes part of the equation, the possibility of launching an
institution-building dynamic becomes more remote, for the now-real threat of
redistributive politics undermines elites’ willingness to support the construction
of institutions strong enough to be able to impose serious costs on them aimed
at promoting nonelite interests.
The vertiginous development of Latin American state institutions – where
it occurred – empirically happened at two important moments. The first was,

21 A similar distinction is made by Higley and Burton (2006, 14–15) between “disunited” and
“consensually united” political elites. The former are locked in zero-sum struggle, whereas the
latter are linked by interlocking social and communications networks that, while they contain
factions that regularly oppose one another, embody the dominance of no particular faction and
are founded on an “underlying consensus about most norms of political behavior.”

as we have seen earlier, in the political settlements that gave birth to viable
national governments (sometimes in the wake of more than a half-century
of postindependence civil conflict). The second, however, occurred as part
of the effort by these states to cope with the emergence of the middle and
working classes as political forces that could no longer simply be excluded
from political participation – at least not at an acceptable cost. This challenge
induced a variety of responses by elites – or the contending actors themselves –
to organize this political incorporation, with a powerful long-term effect on
the character of institutional development. It is to this second critical juncture
that we next turn.

Theory II: Economic Crisis, the Rise of Mass Political

Participation, and the Postwar State
The first critical juncture – the social foundations and political dynamics that
structured the initial formation of national political institutions – launched a
trajectory for Latin American states, either in the direction of improved, more
deeply penetrating governmental institutions or toward persistent weakness,
regionalism, and a central state whose authority was geographically circum-
scribed and often questioned by important societal segments. This initial
branching point does not, however, serve by itself to determine long-run out-
comes. For the institutional structures created in the era of national-state for-
mation all inevitably faced a second challenge, generally sometime between
1900 and the 1940s, of managing the political incorporation of the urban
middle- and working-class masses into politics. This part of the argument is
summarized in Figure 2.1b.
This challenge from below, commonly identified as the “social question,”
would permanently transform the terrain of Latin American politics in author-
itarian and democratic contexts alike. And the manner in which it unfolded
was central to the long-run development of state institutions – even those not
directly related to managing cross-class political conflict. Why is mass political
incorporation a moment in which fundamental institutional trajectories can
be altered? The incorporation of the voices of (at least the urban) popular
sectors fundamentally transforms political contestation by placing the ques-
tion of redistribution squarely in the center of political debate. Middle- and
working-class actors all across the Latin American region mobilized in the first
half of the twentieth century to demand a greater share of the economic pie, a
voice in public policy, and, in many cases, a transformed and expanded role for
the state. But the way in which public institutions were actually restructured
to accommodate and/or control this irruption from below varied widely. But
once created, these institutional patterns, or indeed the inability to construct
institutions with any longevity, persisted over the long haul. Consequently, the
next step is to explore the causal dynamics that account for the trajectories of
political development that were a result of these institutional outcomes.

We cannot, however, abstract away from the first critical juncture, and this
implies that the second critical juncture embodies a form of hysteresis: the same
basic patterns of emergence of the social question produced decidedly different
outcomes depending on the preexisting trajectory on which the state found
itself as a result of the first critical juncture. The second critical juncture thus
deflects the initial state-building trajectory into alternative long-run outcomes
but does not uniquely determine what they might be. This second theoretical
stage would not be necessary, however, if the character of the irruption of
the masses were in fact constant across the cases or reducible to the factors
produced by the resolution of the first critical juncture – but it is not.
The timing and manner in which the masses demanded and gained entry into
national politics varied importantly across the region, and the four cases consid-
ered in detail here, in ways that are exogenous to the institutional and societal
dynamics that gave birth to national institutions in the postindependence era.
The emergence of the social question on the national political agenda was, after
all, typically driven by mobilization from below. Contingent political struggles
were an important part of the determinants of popular-sector mobilization and
the ability to force inclusion. That said, the timing of the political engagement
of middle- and working-class citizens also had a demographic and geographic
foundation – they would first have to come to exist in sufficient numbers (to
be able to effectively press their case). Moreover, patterns of mobilization and
inclusion also reflected how concentrated the population was, or, for example,
the urban population proportion and its immigrant versus citizen composi-
tion. None of these factors follows obviously or necessarily from the resolution
of the first critical juncture. The point is that while the way in which mass
incorporation was initially achieved matters importantly, it is not a product
of the first wave of state building and varies independently across the posi-
tive and negative institutional trajectories that emerged from the first critical
juncture.
The emergence of the social question as an unavoidable political problem –
and the subsequent initial political incorporation of middle-class or working-
class actors – occurred at different times in different contexts. Part of this
difference in timing owes to the characteristics of the national economy in
the early twentieth century, for some agricultural systems – as in grain and
ranching in Argentina – had generated significant associated export processing
industries, while others had few such industrial linkages or were, like mineral
production, typically only able to generate at best modest employment. The
critical dividing line is whether popular-sector emergence came well before
the onset of the Great Depression (and the near-total collapse of the liberal
international economy that it entailed) or during and after this period. For
this moment marked a conceptual sea change that fundamentally shaped the
underlying material interests that these actors brought to the political system.
Where the middle classes, for example, emerged in large numbers before the
Depression, this was a consequence of their linkage to a vibrant, private export

economy. For this was the era of global laissez-faire and free trade, and if large
numbers of such middle-class positions were created, it was because the pri-
vate – principally export – economy was thriving and created associated sectors
employing large numbers of relatively educated middle-sector workers. Where
the middle class emerged later, during and after the Depression, this possi-
bility was precluded. In that context, the bulk of middle-sector occupations
was created within the public sector and linked to the expanding, protected,
domestically oriented industrial sector or tied to administration, education,
health care, or other public goods provision.
Why was this distinction of first-rank political importance? Where the mid-
dle sector was firmly centered in a booming, private export economy of the pre-
Depression era, this would typically produce a political vehicle that favored the
general outlines of that political economy: the gold standard, convertibility, free
trade, and limited public-sector regulation. There were certainly critical reform
demands often made by these new middle-class parties, including the democ-
ratization of the political system and the elimination of corruption, patronage
politics, and vote rigging. But the scope and reach of state institutions were not
a major axis of elite–middle-sector divisions, for their material interests were
comparatively reconcilable with each other. By contrast, when a middle sector
principally grew out of the state bureaucracy, which emerged politically after
the economic tsunami of the Great Depression hit, its political representatives
would be much more amenable to protectionism (itself in any event inevitable
in the short term as trade collapsed), statism, and developmentalist economic
strategies. For at this point in time, not only was the export economy dev-
astated but free-trade economic policies were in near-universal disrepute. The
attractiveness of an alternative rooted in the substantial expansion of the public
sector and an increase in its power and prestige for a middle class already cen-
tered in the public bureaucracy were obvious. And the dominant, nonsocialist
development model available – state-managed, import-substituting capitalism –
would accomplish precisely this. Thus, middle-class parties came to have
common ground with the emergent working-class parties and unions in the
protected industries. Where differences occurred, they were over the level of
redistribution. But this was a bridgeable gap, and strongly redistributive poli-
cies were not typically part of the bargain – nor were they essential features
of statist industrialization. What mattered in this critical juncture for long-run
political development, however, was whether the political representatives of
the middle classes could or could not reach a durable state-building accom-
modation with parties tied to the working class and urban poor. And this
was linked to the timing of their emergence and whether the economic cen-
ter of gravity of the middle class was in the private or public sector; only
where both middle and working classes had much to gain in the expansion
and strengthening of the central state was political accommodation likely. For
as the private sector at the time was largely export oriented and nonindustrial,

if the middle classes were located principally there, they faced a zero-sum con-
flict of interest with a nascent and typically protected industrial sector focused
on the domestic market.
The incorporation of working-class parties into national political competi-
tion often came somewhat later than that of the middle classes (though some-
times they were relatively coterminous). But the politics of the industrial work-
ing class were usually fairly clear: by the time of the Great Depression, the
labor movement and its political representatives typically backed a vigorous
policy of economic reorientation toward the domestic economy, protectionism,
subsidized industrialization, and the creation of myriad social and economic
benefits (i.e., a public role in the provision of pensions, health care, housing,
and education).22
It would be an understatement to say that the pressure that this joint emer-
gence of the popular sectors into national politics put on state institutions –
which were themselves often of comparatively recent vintage – was intense.
For the first time, powerful political voices were demanding major distributive
policies, at the same time as they were advancing a political–economic agenda
that also threatened to reorganize economic winners and losers along sectoral
lines. Traditional elites – typically tied to the agro-export sectors – found
themselves under threat both economically as the Depression devastated trade
and politically as often-radical, working-class actors demanded public indus-
trial investment and protectionism, sometimes in conjunction with emergent
middle-class parties as well. And a large portion of this statist developmental
effort would have to be financed through taxation directly or indirectly levied
on traditional exporters. Such an effort could be rendered politically stable
only where a broad, cross-class popular sector coalition could form to support
it; it necessarily faced very powerful opposition.
The way in which this crisis of representation was resolved involved the
interaction of the preexisting institutional trajectory with the politics of the
emergent middle and working classes. In cases in which the initial path of
political development was favorable to the development of strong institutions
(Argentina, Uruguay, and Chile), the foundation existed for a major expansion
of the reach and power of the central state. But it was far from inevitable,
for the political dynamics that could consolidate a major expansion of state
institutions were quite different from those that might simply bring it about.

22 This was not necessarily a redistributive policy – though by increasing formal sector workers’
cash and social wages, it was perhaps modestly so. The beneficiaries of these reforms in the
formal sector working class, especially its unionized component, were, in Latin America, a
relatively privileged stratum when compared to the majority of the population that languished
in the urban informal sector or comprised the peasantry. Seen only in relation to the traditional
oligarchy, this was a redistributive outcome but not one that favored the unorganized urban
and rural poor.

For countries on an initially favorable path coming out of the first critical
juncture – with cooperation among elites and nonservile social relations induc-
ing the formation of viable national institutions – if the timing of middle-class
political incorporation predated the Depression, in policy terms, middle-sector
workers were typically tied to the oligarchy-dominated export sector econom-
ically and to free trade and laissez-faire politically. By contrast, working-class
interests, not similarly centered in the agro-export economy, would promote
economic redistribution and protectionist industrialization. So if major expan-
sions of the scope and reach of the state were to take place, they were gen-
erally initially carried by the working classes and their political represen-
tatives, acting alone, and only when and where they could achieve power
alone. Moreover, even in the comparatively rare cases in which presidents
came to power with a principally working-class base, and implemented major
reforms to the institutions of the state, a variety of problems would emerge
that would militate against the long-run consolidation of these reforms. First,
the staffing of the rapidly expanding state bureaucracy would of necessity be
dominated by allies of this single political faction. Private sector–based middle-
class parties would not initially cooperate with such an effort, and in turn
their adherents would be excluded from the employment and other benefits
of massive public-sector expansion. Without middle-class party cooperation
at the outset, thus, the large employment gains would tend to exclude par-
tisans of this sector, reducing the interest middle-class parties might have in
subsequent cooperation, for the stabilization of these institutions would not
particularly benefit their constituents. This made such institutional changes a
matter of ongoing political conflict, not a parameter within which politics was
carried out.
Should middle-class parties subsequently return to power, of course, the
situation would be reversed, and they would seek to staff the bureaucracy at
the expense of the incumbents who were largely partisans of their working-
class opponents. And as a consequence, both the structure of public institutions
and the civil servants that staffed them were subject to substantial volatility.
Real, lasting institutionalization of such major institutional changes requires a
broad, cross-class supporting coalition that in this context would be decidedly
difficult to obtain or construct. With the middle classes largely outside the
public sector, lacking a leading role in the design of developmentalist policy, and
with material interests largely tied to the agro-export elite, the formation of an
antioligarchical cross-class coalition behind the strengthening of the state and
public interventionism was too difficult a political lift. Instead, politics would
become a struggle in large measure over the legitimacy of an expanded state role
and over who would control the institutions of the state (and the employment it
represented). Policy would spiral between populist and antipopulist extremes,
with no permanent, widely accepted institutional solution in the offing, for
neither side was likely to win decisively or be permanently defeated. Nor would,
for example, cooptation of a mobilized working class be an option, as the costs

of such a strategy would be immense, and it would require powerful institutions

that by definition did not yet exist to contain and channel their interests.
These dynamics did not take place where middle- and working-class polit-
ical entry came late. In this context, an already-consolidated national state
confronted the inclusion of middle- and working-class political parties, both
of which were pushing for major expansions in public-sector activity. To be
sure, the politics of these two groups were not identical, and middle-class par-
ties were typically not terribly interested in redistributive policies. But on the
core issues of economic management – economic protection, state ownership
of industry, social welfare policies, and state direction of the economy – the
material foundation for a lasting middle-class–working-class accommodation
could be found. A statist development strategy would give middle-class sectors
major expansions in employment opportunities and improvement in the mate-
rial conditions of life as the state administration expanded its economic reach,
enhanced its tax-generating capacity, and began to provide important social
benefits, especially for the white-collar workers in the public sector itself. Mem-
bers of the formal working class as well as domestically oriented industrialists
benefited from the provision of state support as well as extensive economic
protectionism, leading to rapid industrial expansion and increased employ-
ment tied to production for the sheltered domestic economy. This created a
broad-based developmentalist coalition that provided crucial cross-class politi-
cal support for effective and deep state institutions – support that was sufficient
to make possible the consolidation of this major expansion of the institutional
reach and capacity of the state and to be able to enforce its revenue demands in
large measure on the traditional export oligarchy. This coalition in fact could
often also count on some elite support, coming from the expanding industrial
sectors that were reliant on protectionist policies and state-subsidized inputs.
The worst of all possible worlds – from the perspective of strong state insti-
tutions – came in a context where public institutions began on an already
feeble foundation and were suddenly confronted by myriad redistributionist
and statist demands by newly organized middle- and/or working-class actors.
This trajectory began with political elites implacably opposed to the expan-
sion of the state’s infrastructural power, a position that became all the more
determined as demands for heightened taxation, government regulation, and,
most important, redistribution came to be voiced explicitly and powerfully in
national politics. For acquiescence to such an institution-building policy during
the era of the Depression would necessarily imply the imposition of substantial
taxation on agro- and mineral-export elites. By contrast, with limited ex ante
industrialization, an alternative elite coalition based on the industrial sector
was unrealistic, for this group remained as yet underdeveloped and thus of lim-
ited political and economic importance. Similarly, without a history of statism,
or strong central institutions, what middle-class actors that had emerged were
not especially supportive of redistribution or statist developmentalism, tied
as they were to the only important economic pole in the economy – the

agro-export sector. And even were the emergent working classes to come to
power in such a context, there would only be a very limited institutional foun-
dation on which to build, one too weak to constrain elite and middle-class
resistance.
Thus an alliance with middle-sector political actors could not be a viable
path to state building, for the material interests of middle-sector and working-
class groups diverged sharply. Bringing the middle sectors into the coalition
after launching such an institutional trajectory would also be difficult, for
the expansion of the state would have been accomplished by, and staffed by,
partisans of the incumbent government. To make room for a serious ex post
accommodation with middle-class parties, the fruits of public employment and
other patronage distribution would have to be redistributed among erstwhile
political enemies, and questions of longer-term interparty trust would further
bedevil such an arrangement. The only way room for this sharing of resources
could be made would require a curtailment of existing jobs and other resources
for partisans of the working-class–backed incumbent party that launched a
statist trajectory of political development. Such a bargain would thus purchase
an uncertain ally at the cost of weakened political control, the sharing of
decision-making power, and internal division for the incumbent.
Instead, a powerful elite and middle-sector coalition would in these political
contexts likely remain committed to the restriction of the reach of the central
state’s institutions and to the prevention of developmentalist or redistributionist
economic policy (even during the Great Depression). And it would usually
face a comparatively small mobilized working class incapable of instituting or
institutionalizing state-building changes on its own.23 Instead, a more limited
state reach, policy gyrations, and weak institutions would remain the long-term
outcome.

research design and plan of the book

Eric Nordlinger (1968) noted more than 40 years ago that the study of political
development was at least implicitly historically oriented, and most scholarship
made claims that at a minimum required examination of outcomes that emerged
over long periods of time and that the sequencing of key independent variables
was often a critical element in determining them. This is as true today as it
was then – though contemporary scholarship has been more self-conscious
about the implications that the path-dependent character of arguments has
for the testing of theory. For example, as Mahoney (2003, 53) has pointed
out in the context of the path dependencies involved in understanding long-
term economic development in Latin America, “explanations of differences

23 After all, the premise of this path was a relatively limited process of industrialization in the
pre-Depression era, which of needs restricts the size and thus potential economic and political
power of the working classes.

among units that draw on the current attributes of those units will often be
inadequate.” And of course, the matter becomes still more complicated when
theories founded on substantial path dependencies must be evaluated against
alternatives that are contemporaneous in their structure and are premised on
the absence of hysteresis.
The implication is that an adequate theory of political development must of
necessity consider a long historical sweep – if only to meaningfully accommo-
date the testing of alternative theories that are at issue. In fact, some theoretical
claims require even more than the examination of the political development of
contemporary states over the longue durée – most notably, the bellicist con-
tention that conflict and the threat of conflict drive institution building. Here
a long time frame is required even for the specification of the appropriate uni-
verse of cases, for a proper understanding of the effect of war making on state
building cannot be founded on the observation (however accurate) that most
effective contemporary nation-states were forged in the crucible of international
conflict. What is missing here is the fact that state collapse was at the critical
moment in European history a likely outcome of violent international conflict
(Fazal 2004). The consequence is that war or the threat of conflict may be an
enabling condition for initiating a process of state formation and strengthening,
but much scholarship sees this as operating through an evolutionary selection
process rather than a proper causal force. To wit, war may make states, but it
seems to do so only occasionally. Scholars of state formation – by emphasizing
the historical development of actually existing states – have fallen into the trap
of selection bias by virtue of the fact that the states for which war sparked
collapse (or which likely responded to external threats with what Hui calls
“self-weakening expedients”) were often dismembered and thus selected out
of the contemporary sample (Hui 2004). While for those that survived, war
appears to have been an important motivator for institutional improvement,
to correctly assess its causal weight, one must focus on the universe of states
present at the initiation, not denouement, of the state-building process.

An Intertemporal, Two-Stage Research Design

The approach to hypothesis testing taken here involves a series of structured
comparisons. At the first level, the four country cases are divided into paired
comparisons, each of which is designed to emphasize a different critical aspect
of the social–foundational explanation proposed here. In addition, intertempo-
ral comparisons within cases are made to highlight the specific effects of vari-
ables (e.g., social relations, elite politics) that change during the period under
study. The design is two stage insofar as each paired comparison is evaluated
at each critical juncture – the initial formation of viable national political
institutions and the period of the initial political emergence of the middle sec-
tors and working classes. And since the argument is path dependent, implying
that at later stages, departures from the trajectory of political development on

which a case finds itself are exceedingly difficult, the second stage analysis also
highlights the specific moments in which determined efforts at path departure
were undertaken by executives. These moments demonstrate the intense diffi-
culty involved in successful path departure and that variables that at an earlier
period were sufficient to drive state-building outcomes were unable to redirect
trajectories later in the process, once the mechanisms that reproduce outcomes
had been firmly established. Indeed, some of the most powerful authoritarian
regimes on the continent tried, and failed, to institutionalize changes to the
pattern of political development.
Why is hypothesis testing in this book centered on comparative historical
evidence? The justifications for such an approach are both theoretical and
practical. In the first case, arguments about the relationships of war, wealth,
and/or politics to state building are cast here (and by most other scholars) in
path-dependent terms. This precludes at least simple large-N cross-sectional
research designs, insofar as putative causes and effects are hypothesized to
be connected only given the outcomes of prior events, which are of necessity
not recorded in contemporary data.24 In principle, given data sets of long
historical sweep, it might be possible to adequately model theories of this kind
in large-N quantitative data. But even so, it would imply complicated functional
forms and would pose a series of econometric and data challenges. In the
absence of such long-term historical data, efforts to test such theory in cross-
sectional (or short–time series) data would be nonsensical – they would either so
oversimplify the theory in the effort to make it empirically tractable as to vitiate
the hypothesis tests themselves or would be heavily biased with respect to key
hypotheses.25

The Formation of National Institutions

The formation of recognizable, and fairly stable, national political institutions
occurred in South America generally in the second half of the nineteenth cen-
tury. The ultimate outcomes reflect quite substantial institutional divergences
among the former Spanish colonies of the region. The theoretical account devel-
oped in this book to explain these initial differences focuses on two principal
factors: the prevailing social relations of economic production (especially in
the countryside) and the patterns of elite conflict or cooperation. At the same
time, this approach is cast against alternatives emphasizing the causal effects
of war and resource wealth, and it challenges the exogeneity of explanations
that begin with institutional variations to explain long-run outcomes.

24 In contexts where causal dynamics involve critical junctures, cause and ultimate effect are often
widely separate in time, and indeed the “cause” may no longer be present at the time at which
the effect is observable.
25 I would argue that both are critical problems for approaches, such as those of Thies (2004,
2005), that test state-building arguments in cross-national data of quite limited time span.

Figure 2.1 scores the state-building outcomes in this first period – the initial
trajectories were everywhere established by the end of the nineteenth century.
The variations highlighted in this figure were then used to structure the paired
comparisons considered in this book: Peru and Chile, on the one hand, and
Argentina and Uruguay, on the other. In the first period, the analytical focus of
the Chile–Peru comparison emphasizes the polar overall outcomes (state atro-
phy versus substantial state building) that arise from differences on both the key
independent variables. The contrast on the independent variables is very strik-
ing and characterizes essentially the entirety of the postindependence period for
both countries. At the same time, this same comparison is used to shed light on
several of the most plausible alternative explanations. For instance, Peru and
Chile were, in the nineteenth century, in a regional environment characterized
by substantial interstate conflict and rivalry. Indeed, they were among each
other’s principal political and military foes throughout this period, and they
fought two major wars. The last, the War of the Pacific (1879–83), involved
a definitive defeat of Peru by Chile and the temporary occupation of Lima.
Both countries were also – and they remain so to this day – very important
natural resource–exporting economies. In the nineteenth century, these exports
were very heavily concentrated in nitrates; in fact, they were literally the very
same nitrate deposits, for these very valuable mineral territories were annexed
into Chile after the War of the Pacific. Not only does this contrast then permit
us to examine the social–structural theory presented in this book against the
regnant alternatives, it also allows us to consider the question, empirically, as
to when and if war and resource wealth induce or impede national political
development.
If Chile and Peru are very dissimilar cases in terms of the independent
variables in this first period, Argentina and Uruguay are not. Not only were
their basic political economies of the period almost identical (being principally
ranching countries that would later also become major grain exporters), but
both relied almost exclusively on free-wage labor for production in the post-
colonial era. The cases are critical, however, insofar as the comparison (both
within and across) sheds light on the critical role of elite politics. For only very
late in the nineteenth century were their elites able to reach the sorts of compro-
mises that made possible the construction of domestically and internationally
recognized political institutions. To be sure, this was accomplished in different
ways. In Argentina, one faction of the elite effectively militarily defeated its
opponents and was able to impose a more or less hegemonic project – coupled
with sufficient moves to incorporate the defeated side to stave off severe future
disturbances. In Uruguay, two vituperatively opposed political parties – the
Blancos and Colorados – ultimately came to a broad pattern of power sharing
(called coparticipation) that was in a variety of forms able to survive from
the late nineteenth through most of the twentieth centuries. And critically, the
very rapid increases observable in the indicators used to measure the scope and
extent of the institutional reach of each state take place in the immediate wake

figure 2.2. Scoring the cases on the initial trajectory.

of these political settlements – highlighting the crucial role of elite politics even
in a setting not saddled with a large segment of unfree labor.

The Rise of the Popular Classes

The analysis of the second critical juncture seeks to accomplish a somewhat
different task. In theoretical terms, these two critical junctures are sequenced
moments, and thus the outcomes of the second critical juncture depend in
part on the path a state is on when it enters the period of mass electoral
incorporation. As we see in Figure 2.2, three of our cases were on a favorable
trajectory of institutional development (Argentina, Chile, and Uruguay), and
one was not (Peru). From this starting point, the second set of empirical chapters
examine the ways in which these trajectories were deflected or reinforced under
the intense pressures of initial mass political incorporation – which generally
took place in South America between the 1900s and the 1940s.
As Table 2.1 makes clear, a key axis of differentiation is the timing of this
entry into real competition for national political control. In Argentina, the
middle classes entered politics early, and this electoral incorporation produced
a political party that was able to seize executive power well before the Great
Depression. This would, ironically, have decidedly unfortunate implications
for the pattern of national institutional development, as it laid the foundation
for an inability to arrange a lasting accommodation between the middle sec-
tors and a late-entering working-class party. The consequence was a cycling
between populist and antipopulist institutional expansions and retrenchment
that characterized the bulk of the post–World War II era. In Chile and Uruguay,

table 2.1. Scoring the Cases in the Second Critical Juncture

Driving Force
behind
Timing of Initial Effort
Middle- at Major
Class State-
Country Electoral Building
Institutional Entry Program Long-Run Outcome
Initially Chile Late Middle-class– Effective state
strong (during/after working- building
institutions Great class
Depression) accommo-
dation
Uruguay Late Duopoly of Effective state
(during/after multiclass building
Great political
Depression) parties
Argentina Early (after Working-class Populist/antipopulist
First World (Peronist) institutional cycling
War) party alone
Initially weak Peru Early (before Not Persistent
institutions Great successfully institutional
Depression) imposed weakness

by contrast, widespread mass political participation came substantially later.

Indeed, in the former, suffrage was decidedly restricted for much of the first
half of the twentieth century, and well into the 1950s, even basic reforms
like the secret ballot that would enable meaningful political participation had
not yet been instituted. This set the stage for the emergence of middle- and
working-class political forces that were able to collaborate – through quite
different institutional vehicles – in the formation of vastly expanded state insti-
tutions in the context of the developmentalist protectionism of the post–World
War II era.
But showing the different effects of the timing of initial popular-sector entry
into real electoral contention is only part of the goal of the analysis of the
second critical juncture. At the same time, in all these cases, elites initiated
multiple efforts to change the institutional trajectory on which the states found
themselves. I use these efforts to directly explore the constraining effects of
institutional path dependence. Thus, in Chile, one can examine efforts in the
1950s under President Alessandri, and the exhaustive effort by the Pinochet
dictatorship (1973–89), to radically shrink and weaken state institutions. If the
path-dependent specification is correct, however, this task should prove to be
all but impossible for executives even in the context of almost-unchallenged
authoritarian rule. Similarly, this book examines the inability of the Uruguayan

military to reverse the scope and penetration of its national state, even as it
applied a tremendous amount of police and military coercion to effect its state-
weakening goals, which were themselves consonant with advice and pressure
emanating from the international system. And in Peru, we see the most decisive
effects of the sequential character of the state-building process. For, in this case,
we examine a military government that (in the 1967–80 period) actually man-
aged to effectively reverse the structural impediments to institution building
identified for the first critical juncture by using corporatist institutions to force
effective national-level political coordination, while at the same time imposing
serious structural reforms (radical land reform and the empowerment of work-
ers) that effectively put an end to the servility so long characteristic of Peruvian
labor relations. But this effort came very late and well after mass political incor-
poration had taken place. Ultimately, the military was unable to put the genie
back in the bottle – weak institutions, ironically, proved remarkably resilient.
These efforts at path departure help to test the sequential character of the
argument, which is an essential component of any path-dependent theoretical
formulations.

what is state building?

Before proceeding with the empirical analysis in subsequent chapters, how-
ever, it is crucial to be very clear about what is meant by state building. State
building and state strength are not easy conceptual targets, and indeed, it has
taken a variety of different meanings in the political science context. In some
contexts, the emphasis is on military strength, regardless of the underlying
economic and political institutions on which the military relies.26 In other con-
texts, state building is conflated with nation building – the creation of a national
identity, mass political participation, or, for example, cultural unification (see
Tilly 1975, 70). While these sociocultural outcomes are potentially the con-
sequences of a state-building experience, from the perspective of this book,
they do not define it. Finally, some think of state strength in terms of formal
institutional structures, classifying systems that fragment authority or employ
a decentralized federalism with overlapping jurisdictions as inherently weaker
than more unitary, centralized political systems. This also is a conceptualization
not employed in this book.
The approach taken here – and it is widely shared in the contemporary com-
parative politics scholarship on state building – departs from a basically Webe-
rian understanding of the state as the monopolist of the means of legitimate
violence in a defined territorial area (Weber 1978, 54–56). Joining Giddens
(1987, 20), however, I deemphasize the Weberian insistence on the legitimacy

26 E.g., an emphasis on military capabilities would suggest that late Capetian France was quite
strong, whereas an emphasis on the ability of the allegedly absolutist monarch to govern and
tax would suggest quite the opposite (see, e.g., Ertman 1997).

of this monopoly on violence as a definitional aspect of the state; this is a

potential consequence of state building but is not necessarily constitutive of
“state-ness” itself.27 The aspect of the state that is crucial from the perspective
of understanding the state-building process is the character of public institu-
tions. For all states worthy of the name have a public administration of some
sort – for good or ill – and the fundamental challenge of state building is the
construction of institutions that are effective.
But what, then, is efficacy? In part, it implies an efficient solution to the
principal–agent problem involved in governance. Scholars have long recog-
nized that the incentives confronting front-line staff may be at odds with the
policy goals of political executives or their administrative superiors. These
differences could produce anything from active opposition and bureaucratic
sabotage to an indifference that manifests itself in delay, avoidance, or error.
While political leaders (the executive, sometimes the legislature) are charged
with setting public policy, it is the public administration that is tasked with
the actual implementation of political decisions – and depending on the way in
which these institutions function, policy-as-implemented may or may not bear
much resemblance to the choices made by politicians. There is, however, quite
a lot more to it than this. For the coordination of the actions and behavior of
the individuals working within the public administration28 is only the begin-
ning step for the implementation of policy. And indeed, the principal–agent
problems within administrative agencies so frequently highlighted in the litera-
ture are sometimes far from the critical issue: policies must also be imposed on
the societies governed by the state. And important policies almost always have
powerful societal opponents. Moreover, the decisions of political leaders – to
impose taxes, redistribute incomes, provide public goods, impose industrial
or development strategies, inter alia – all imply the creation of very substan-
tial winners and losers. The question from the perspective of state building
is, however, how able (and at what cost) are public institutions to enforce
these decisions, especially on societal actors who have vested interests in resist-
ing them and substantial amounts of economic and/or political power? Thus,
separate from the question of motivating administrative staff to seek policy
implementation is the question of whether public institutions are sufficiently
strong and penetrating to enforce compliance – this requires the ability both
to efficiently detect societal evasion and to impose appropriate sanctions on
malefactors.
Indeed, individuals in civil society frequently struggle to avoid the reach of
the central state. One way this is accomplished is through the effort by elites to

27 As Rueschemeyer (2005, 147) rightly points out, the norms and beliefs that would undergird
legitimacy are in fact typically coercive impositions enforced by state officials.
28 I specifically avoid the term bureaucrats here, for bureaucracy is only one – though generally
the most effective – way in which a public administration can be structured. And as Evans
(1992) has noted, it is generally in rather short supply.

shape public policy toward their private benefit – or to the detriment of their
competitors (e.g., Krueger 1974). Or, alternatively, elites and the masses alike
may engage in efforts to avoid compliance, for example, with the payment of
their juridically defined tax burdens, especially where they are unconvinced
of the benefits or the proportional contribution of others (Levi 1989). This
is particularly the case when it is clear that revenue systems are transparent
in their distributional consequences or where they violate deeply held, and
materially defined but culturally experienced, norms about appropriate types
and levels of extraction (e.g., Bates 1981; Scott 1990, 111; Ibid. 1987, 29, inter
alia). The critical question from the perspective of state building, then, is the
question of the institutional capacity of states to impose and effectively enforce
the revenue and policy initiatives of their leaders. It is what Mann (1993, 59)
in a more general sense calls “infrastructural power,” which
is the institutional capacity of a central state, despotic or not, to penetrate its territories
and logistically implement decisions. This is collective power, “power through” society,
coordinating social life through state infrastructures. It identifies a state as a set of
central and radial institutions penetrating its territories.

State building, then, is the process by which the institutions of public admin-
istration are strengthened in their scope and penetration of society. By impli-
cation, such infrastructural power implies low levels of corruption and ineffi-
ciency, for these would render line civil servants poor agents of central author-
ity. And administrative structures must also have a substantial degree of insu-
lation from particularistic societal pressures not channeled through the formal
political system, ensuring that a true decision-making hierarchy is maintained
and that political choices are not reversed by ex post “lobbying” of adminis-
trative staff charged with implementing them.29 Achieving these goals is not
simply (or even largely) a question of institutional design per se, be it federalism
versus unitary structures or overlapping versus discrete jurisdictions, though
these can potentially have some effects on the infrastructural power of a cen-
tral state. But it is critically about the central state: strong states must be able
to impose policies throughout their national territories. And statelike entities
without a meaningful center (as in, e.g., the United States under the Articles
of Confederation or leagues of city-states in European history) cannot, in the
sense of this book, be “strong.”
Interestingly, much as the term bureaucracy is decried in contemporary dis-
cussions, state building inevitably involves the construction of bureaucratic
institutions. Indeed, the very invention of modern bureaucratic administration
was a breakthrough of world historical proportions, emerging in rudimen-
tary form only around the seventeenth century in Europe (Ertman 2005, 166).
Some have suggested that it provided organizational advantages that could even
be decisive in international warfare, from a superior ability to raise revenues

29 This is roughly what Fukuyama (2004, 8) considers the “strength of institutional capabilities”
in his discussion of the state-building process.

to the organizational and logistical improvements it brought to armies them-

selves. Further highlighting that building a bureaucracy is no mean feat, despite
the contemporary availability of institutional models of efficient bureaucracy
and substantial international promotion of best-practice institutional forms, it
remains in remarkably scarce supply throughout much of the developing world
(Evans 1992, 1995).

What State Building Is Not

Before embarking on a discussion of measurement strategies, it is important to
be clear about what state building is not. Most critically, political development
producing “strong” states in the sense employed here has no implication for
the sort of policy choices (beyond the institutional configuration of the public
administration) made by political elites or the type of regime within which
they are undertaken. Thus, for example, it is widely held that the protection
of private property rights requires a strong, effective state apparatus. But the
former does not entail the latter – and thus measurement strategies that judge
the quality of institutions based on the protection of property rights (or worse
yet, the assumed consequences of particular levels property rights protection)
will go astray.30 Powerful institutions can support any number of property
rights systems, not all of which will involve strong protections of individually
alienable, fee-simple property.31 Some – as in state socialist systems – may deny
private property rights entirely.

30 Acemoğlu and Robinson (1996, fn. 1) are explicit about their elision of policies and institutions,
contending that many of the factors that affect property rights enforcement are policies but will
nevertheless be treated as “economic institutions” by them.
31 Indeed, it is important to note that many contemporary discussions treat private property rights
as simply a continuum from “less protection” to “more protection.” But this construction is
essentially ignorant of the lesson that North (1981, 36) has taught us: that property rights are
alternative bundles of rights and responsibilities attached to the notion of ownership, and that
the key distinction from his perspective is their comparative “efficiency” versus “inefficiency”
in an economic sense. Not all property is individual or individually alienable, and not all strong
property protection is useful from a developmental or governance perspective. Consider that
excessively strong individual property rights can impede the regulation of critical externali-
ties, and the strongest form of property rights – allodial title – would render a market econ-
omy scarcely functional. Individual ownership should not be assumed, either. Earlier forms of
property rights – especially those of the feudal era in Europe – entailed different rights attaching
to distinct individuals for the same piece of property (e.g., lords were entitled to labor from
resident serfs, certain fees and taxes, and a portion of the agricultural output, but peasants
typically had enforceable unsufructory rights over a quantity of land and secondary rights even
over demesne territory, e.g., gleaning and grazing rights [the right of shack, or agistment when
applied to a royal forest], estovers [rights to collect wood], turbary [the right to cut turf for fuel],
prohibitions on lords to fully cut their crops to facilitate grazing and gleaning, or prohibitions
on enclosing the land to prevent the movement of peasant livestock or to restrict access to the
commons). See, e.g., Markoff (1991, 172) on France and Manning (1977, 18 and fn. 2) on
England.

In a similar vein, one cannot assume that effective bureaucratic institutions

will produce good policy choices – be they in the realm of development strat-
egy, fiscal management, public employment, or taxation levels. Effective public
administrations should implement policies efficiently and with a minimum of
graft or penetration by rent-seeking societal interests. But the choice of pol-
icy strategies is ultimately a question for the political system, not the public
administration. And there are myriad examples of strong governments enforc-
ing decidedly suboptimal economic policies (from the tariff escalations that
helped deepen the Great Depression to the intense small-country protectionism
of Uruguay in the 1950s and 1960s to the fiscal and monetary policy errors
that characterize the contemporary deflationary period in Japan).
Ultimately, the conflation of policies with institutions makes analysis
intractable. Acemoğlu et al. (2001, fn. 3) indeed provide the limiting and
illustrative case. For them, the public “institutions” that are critical to long-
run economic performance include “constraints on government expropriation,
independent judiciary, property rights enforcement, and institutions providing
equal access to education and ensuring civil liberties.” Only some of these
can reasonably be thought of as institutions – others are policy choices made
and implemented through institutions (like mass education or political rights).
But the factors that produce effective bureaucratic apparatuses by no means
necessarily imply mass educational expansion, much less broadened civil and
political rights.
We must similarly be careful to avoid the trap of concluding, like
Rueschemeyer (2005, 146), that the emergence of cultural practices that facil-
itate efficient administration need be thought of as part of the state itself or as
a characteristic of “state building” or “full institutionalization.” It is for this
reason that the Weberian notion of authority (legitimate power) was excluded
from the definition of the state employed here. High-quality public institutions
may sometimes propel the development of normative practices that support
compliance with legal and policy enactments, but they are not constitutive of
them. Thus, this book follows in part the perspective of Evans and Rauch
(1999, 749), who emphasize the importance of what they call “Weberian state
structures” – meritocratic recruitment into bureaucracies and long-term career
trajectories that reward performance. While such structures might more effec-
tively deliver public goods – and this over time could induce ideological or cul-
tural patterns of compliance or cooperation like Levi’s (1989) “quasi-voluntary
compliance” or North’s (1990, 47) “informal constraints” on interaction – they
cannot tell us what mix of public goods will be selected for provision, nor do
they ensure the emergence of complementary norms and practices.

Measurement
This focus on the institutional features of the state, while narrowing the con-
ceptual terrain, does not set up an easy measurement task. As we will see,

this is particularly true for this book given its focus on long-run patterns of
political development in Latin America. For not only is a tractable set of cross-
nationally valid measures of the institutional capacity of states required but
these metrics must also exist in the historical record from roughly the early
independence period onward. As we will see subsequently, all the available
options are imperfect, so it is critical to utilize the best available data in a
way that honestly examines the theoretical claims at issue, while being sensi-
tive to the limits that the data imply. Conversely, to permit data availability
to influence – consciously or otherwise – the elaboration of theoretical claims
or conceptual structures undermines the hypothesis-testing process altogether.
Either crucial variables will be omitted – specification error – or they will be
defined in operations that are inappropriate, inducing random and/or system-
atic measurement error.
In thinking about developing metrics to capture the capacities or efficacy of
state institutions, we are essentially confronted with two sets of choices. The
first is whether to rely on “objective” indicators32 of state capabilities or on
perceptions of them. Both approaches are commonplace, and in the contempo-
rary literature, indeed the most commonly used metrics are indeed subjective,
perception-based ones. Notably, especially in the economics literature, scholars
rely on indicators of state efficacy largely derived from perceptions of firms,
international investors, investment consultancies, other “experts,” and occa-
sionally the views of the mass public; they are called on to assess the quality
or probity of public administrations. So, for example, Transparency Interna-
tional’s Corruption Perceptions Index is widely used to rank nations according
to the perceived level of graft in the public administration. Alternatively, Polit-
ical Risk Services, a business consultancy, produces the International Country
Risk Guide, which includes an indicator of what they call “country risk” –
ostensibly a measure of property rights security, which is frequently used as a
proxy for the overall quality of state institutions (as in Acemoğlu et al. 2001).
Perhaps the most comprehensive such metrics are included in the World Bank’s
(2011b) Worldwide Governance Indicators, which combine a broad swath of
different perception-based metrics to construct aggregate measures of gover-
nance across six dimensions (see Kaufmann et al. 2009).
The perception-based approach will not do for both methodological and
practical reasons. It has been widely demonstrated that perception-based
approaches to measuring governance are subject to important selection biases,
conflate policy measures with institutional measures, are limited in their scope,
and are commonly subject to positive and negative halo effects related to recent
economic performance (see, inter alia, for discussions of the problems of these
indicators, Thomas 2010; Kurtz and Schrank 2007a, 2007b; Arndt and Oman

32 The implication here is, of course, not that such indicators are without errors of measure-
ment. Rather, the point is that they rely on concrete, measurable features or outputs of public
administration.

2006). In addition, perception-based indicators have only a narrow temporal

sweep – the earliest dating back only to the 1980s, and there only with very lim-
ited coverage. As we will see, our theoretical contentions will require data over
a much longer historical time frame, rendering these metrics inappropriate.
If we are to rely on objective metrics, this immediately raises the question
of which metrics. One quite valid approach is that of Evans and Rauch (1999;
also Rauch and Evans 2000). Their approach is to utilize a survey instrument
applied to multiple country experts per case – cross-nationally – to assess not
perceptions of governance but rather the objective, “structural” features of
core bureaucratic agencies in 35 countries.33 In principle, a direct measure of
the institutional features of government is optimal. But for our purposes, this
will be problematic, for it is not possible to engage in a similar consistent
characterization of Latin American public bureaucracies from independence
through the present – the historical sweep necessary for the argument devel-
oped earlier in this chapter. While there is some available direct evidence –
documenting the actual on-the-ground operation of public administrations –
its availability is haphazard. In the comparative empirical work that forms the
core of this book, this information will naturally be used to validate the more
consistent but indirect measures that form the core metric used here. It is also
important to note that historical descriptions of the formal rules governing the
operation of public bureaucracies are of at best limited utility given the many
and extensive divergences (many of which continue into the present) from
them that render them poor metrics of actual features of the operation of state
institutions.
The main approach taken here – which is broadly shared in the historical
literature on state building – relies on the measurable and documented inputs
and outputs of public-sector activity. This is what Soifer (2008, 236) has iden-
tified as a “national capabilities approach” to measuring state infrastructural
power. It takes as its starting point that an administrative organization that
is weak (fragmented, inefficient, riddled with patronage and corruption, etc.)
will do a comparatively poor job of accomplishing the tasks assigned to it. To
be a useful measure of public-sector capacity, however, these outputs must be
carefully selected so as to be sure that cross-national differences reflect prin-
cipally variations in the capacity to produce them rather than policy choices
over alternative priorities. As such, the best measures would be of the sorts of
activities that all (or nearly all) states consider to be of primary importance –
for these can be assumed to be pursued vigorously across contexts, and vari-
ations in success will depend principally on variation in the effectiveness of
institutions, not on opportunity costs or political will.

33 Evans and Rauch (1999, 755) are very clear about not asking their informants to characterize
the quality or performance of the bureaucracy. They seek only descriptive information about
objective features of bureaucratic institutions such as the mechanism for entry, the promotion
process, or job tenure.

One such metric – and it is indeed a powerful one – has to do with the ability
of the state to tax the citizens and firms that engage in economic activity within
its borders. At least since Schumpeter (1954, 17), taxation has been seen as
central to the state’s political reach. Indeed, he argued that “[t]axes not only
helped to create the state. They helped to form it. The tax system was the organ
of [state] development of [sic] which entailed the other organs.” And it is widely
used in contemporary scholarship as a measure of state strength (e.g., Cohen
et al. 1981, 905; Gallo 1991, 8; Centeno 2002, 116–26; Thies 2005; see discus-
sion in Soifer and vom Hau 2008, 220).34 There are considerable advantages
to considering the ability to tax as a plausible indicator of the overall level of
state strength.35 First and foremost, it is an activity in which virtually all states
engage of necessity.36 And of course, it is one that is intensely difficult: effective
tax collection requires an effective bureaucratic apparatus that is capable of,
for example, measuring economic activity, recording myriad transactions in the
domestic market, estimating property and land values, ensuring the accurate
reporting of incomes, and then extracting the relevant resources from economic
agents of all types. In the Latin American context – characterized as it is by
generally very high levels of inequality – this further implies that any state
that manages to extract high levels of resources must of needs be deriving an
important portion of them from the wealthier and more powerful members of
society. For that is where the resources are. This is important, for the definition
of state building used here emphasizes the ability to enforce policies not only
on subaltern members of society but also on society’s strongest. Of course, in
wealthy, high-institutional–capacity contexts, taxation levels eventually come
to reflect policy choices as well as the limits of institutional capacity. But for
our cases, the level of extraction is principally defined by the capacity to tax
(indeed, this capacity remains a critical limit on tax receipts even for some
wealthy countries such as contemporary Italy or Greece).
The empirical approach of this book thus begins with an assessment of the
tax capacity of countries in the Latin American region. But it goes well beyond
this as well, for as Campbell (1993, 177) has pointed out, “both the form and
the level of taxation influences [sic] state building.” In the empirical discussions
of our four cases, thus, analysis of over-time changes in the level of taxation
will be complemented by discussions of the types of taxes that are employed.
This variation is quite informative, for it is a much simpler administrative

34 Indeed, Cohen et al. (1981, 905) flatly declare, “We shall use tax revenue to measure state
power.” Centeno uses taxation levels to underscore what he sees as the weakness of Latin
American states, whereas Gallo focuses more on the type of taxation (foreign trade vs. broad-
based internal taxation) to draw inferences about state capacity.
35 It is worth noting that revenues derived from mineral production and/or export (excises, royal-
ties, and income taxes) are included in the broad definition of tax revenues.
36 The only important exceptions would be states that derive virtually all their resources from
international transfers or the proceeds from investments, but this characterizes none of the
states or time periods examined here in the Latin American context.

challenge to collect revenues from, for example, export and import taxes than
it is to impose broad-based taxation on domestic transactions (as in a sales
tax or a value-added tax), and of course, taxing income, profits, and wealth
is administratively and politically difficult to impose. Indeed, it is precisely the
fact that the means of revenue collection also matters, which underscores the
use of tax capacity – ordinary revenue as a share of GDP – in an ordinal fash-
ion as (one) of the set of metrics by which states are compared. To assume
that such data were accurate at the interval level is simply not plausible: it is
far administratively harder to lift tax receipts from 40 to 41 percent of the
economy than from 4 to 5 percent – for the former necessarily entails imposing
difficult-to-collect income, wealth, or broad-based consumption taxes, whereas
the latter does not.37 The reason is easy to see. Easy-to-collect taxes like
tariffs and some excise taxes, were set at high enough levels to theoretically
capture this proportion of GDP, would be high enough to impede imports
or sales altogether, making them self-defeating. A revenue tariff would thus
very quickly become a protectionist tariff and generate little in the way of tax
receipts.
Several caveats are in order here, however. First, it should rightly be noted
that the burden of domestic taxation that is imposed is only partially a question
of administrative capacity; it is also a political and policy question concern-
ing the appropriate level of extraction. Thus, it is possible that tax receipts
as a share of the economy may be low as a result of a political choice to
pursue a limited state rather than the incapacity of governmental institutions
to collect such revenues were tax increases to be enacted. This is, however, a
phenomenon much more likely confined to the advanced industrial countries.
While it certainly accounts for the variation between the comparatively small
state (in taxation terms) of the United States relative to, for example, social
democratic Scandinavia, policy differences do not appear to account for much
of the variation in the Latin American context, particularly in the nineteenth-
and twentieth-century periods that most concern us here. The principal diffi-
culty is actually collecting taxes that are owed, not imposing them in a de jure
sense. As we will see in the next section, the comparative hierarchy (as mea-
sured by taxation) among Latin American states has been very constant for
quite a long period of time, and within countries, extreme swings in the politi-
cal ideology of executives, the regime type,38 or severe economic crises do not
generally appear to have major medium- or longer-term effects on comparative
taxation levels. By way of illustration, the arch-liberal military government of
Augusto Pinochet in Chile taxed at far higher levels than the developmentalist
and redistributive “Revolutionary Government of the Armed Forces” in Peru

37 This has not prevented some scholars from implicitly making the assumption that data are
accurate at the interval level (as measures of state strength). See, e.g., Thies (2005).
38 Slater (2008, 254–55) makes the case more generally that democracy and expanding state
capacity can coincide, even as “authoritarian leviathans” are also a possibility.

of the same time period!39 And of course, that state executives are essentially
revenue maximizers (at least before the very contemporary era) is a founda-
tional assumption of most of the literature on the development of taxation (as
in, e.g., Levi 1989; Olson 1993).40
Taxation is, of course, not the only way to measure the capacity of states.
And where available measures are necessarily indirect, it is appropriate to val-
idate findings using alternative indicators. Usually, this requires analysis based
on public goods production, which tends to be relatively well recorded in the
historical record. The difficulty in using measurable public-sector outputs to
judge the capacity of state institutions is that one must make the accessory
assumption that the states in the sample are all pursuing the output that is
being measured at a comparatively similar level of priority and that the task
is potentially at least a large and complex one (i.e., one that puts substantial
demands on the state and its institutions). But this implies that we must iden-
tify major activities in which all states engage. The two principal candidates
employed here – since they apply well in the Latin American region – are educa-
tion and infrastructure provision. While the precise form that both took varied
across countries in Latin America, it was state support that undergirded the
massification of education and the expansion of roads, rails, and telegraphs.
And as an indicator of capacity, increasing rates of literacy or school enrollment
match the increasing administrative challenge. It is easier to educate students in
urban settings where scale economies, the supply of teachers, and transporta-
tion infrastructure are well developed. As literacy expands, however, it must
necessarily penetrate the countryside, and in a very decentralized fashion –
this is a much costlier task (financially and administratively) and involves the
recruitment and placement of teaching staff in oftentimes undesirable, remote
settings. Thus, improvement on this indicator does not simply represent the
horizontal expansion of the state but also an improvement in its administrative
and penetrative capacities.
For the Latin American states during the era in which state building began
in earnest (from the mid-nineteenth to the early twentieth centuries), the pro-
vision of infrastructure, specifically rail infrastructure, was a critical task. It,
in two very different senses, provides a measure of state capacity. On the one
hand, like education, it was an economically valuable but administratively,
financially, and technically challenging task to undertake, and it was one that
was almost never managed without explicit or implicit state promotion (either
through state ownership or varieties of public subsidy of private infrastructure

39 Of course, the Pinochet period was characterized by far lower taxation than the immediately
preceding socialist government of Salvador Allende. The point, however, is that even as late
as the 1970s, the variation that is produced by shifts in policy is insufficient to reverse a
long-standing hierarchy founded on differences in the capacity to tax.
40 They are, of course, maximizers subject to constraints. But it is these constraints that help make
the measure useful – for the stronger the state, the better “taxation bargain” it can strike with
the societal actors from whom it takes revenue.

provision). It is certainly the case that some rail infrastructure was privately
built and operated (as in Argentina) – but its existence nevertheless required
extensive state participation in the form of loan guarantees, subsidies, land
grants, tax exemptions, or favorable rate schedules, inter alia. Rail infrastruc-
ture measured the sinews of the state in a more literal sense as well, as it also
physically connected the sometimes-remote portions of the national territory
in a single political economy (and market) that could be governed from the
center. For as much as rails unify national markets and improve communi-
cations capacity, they also make the essence of the central state – coercion –
much easier to apply by removing geographic remoteness as a barrier to the
imposition of central authority over often restive regions. They facilitated the
movement of information, troops, and bureaucratic personnel.

What Is to Come?
Any new effort to explain the emergence and divergence of state institutions
and their capacity in the South American region thus faces a daunting set of
theoretical tasks. On the one hand, any treatment must be able to account for
the emergence of strong institutions in unlikely places – after all, places like
Chile and Uruguay inherited virtually no administrative infrastructure from the
colonial period, and those areas that had been colonial centers (most notably
Peru but also Argentina) did not produce the most effective states. But the
account must also be able to explain the persistence in these relative differences
over the subsequent century and a half. Moreover, it must directly engage
the existing, vibrant literature on conflict-induced state building, especially the
pathbreaking work on the formation of national states in Europe.
It is to this empirical and theoretical task that we next turn. In the paired
comparisons that follow, the central hypotheses of this book will be explored
and contrasted in their implications with the regnant alternative accounts.
The empirical discussion proceeds in two steps: the first is concerned with
the emergence (or failure to emerge) of an initial state-strengthening trajec-
tory in the postindependence era, whereas the second examines the ways in
which a second critical juncture sparked by the political incorporation of the
middle and working classes reinforces or deflects the initial trajectory and
reproduces the path of political development through, often severe, subse-
quent political and economic shocks. More specifically, in Chapter 3, a paired
comparison of Peru and Chile – polar outcomes in terms of long-run state
power – during the initial period of national institutional formation is used
to elaborate the origins of their distinct long-run trajectories. In Chapter 4,
the cross-time and cross-country comparisons of Argentina and Uruguay are
used to establish the political side of the foundation of national institutions.
In Chapter 5, the consolidating and deepening effects of the emergence of late
and early middle- and working-class political participation are explored in
Chile and Peru, respectively; in the former, it served to deepen and strengthen

administrative institutions, whereas in the latter, it cemented their inadequacy.

Finally, the politics of mass incorporation will be tied to the divergence in
Argentine and Uruguayan political development from the mid-twentieth cen-
tury onward in Chapter 6. Chapter 7 concludes the book and examines the
potential reach of the theory to a paradigmatic and well-explored case of state
building in the European context: Prussia–Germany.

Kurtz-Latin American State Building in Comparative Perspective PDF
No ratings yet
Kurtz-Latin American State Building in Comparative Perspective PDF
45 pages
Kurtz (2009)
No ratings yet
Kurtz (2009)
42 pages
EXTRA The Hist Form of The State in LA
No ratings yet
EXTRA The Hist Form of The State in LA
23 pages
Ordering Power - Dan Slater
No ratings yet
Ordering Power - Dan Slater
47 pages
Mazzuca (2021), Pp. 1-14
No ratings yet
Mazzuca (2021), Pp. 1-14
18 pages
Book Review Why Nations Fail-Acemoglu An
No ratings yet
Book Review Why Nations Fail-Acemoglu An
5 pages
Review of "Why Nations Fail" by Acemoglu & Robinson
No ratings yet
Review of "Why Nations Fail" by Acemoglu & Robinson
5 pages
Book Review Why Nations Fail-Acemoglu An
No ratings yet
Book Review Why Nations Fail-Acemoglu An
5 pages
Cap1 Kurtz 2013
No ratings yet
Cap1 Kurtz 2013
29 pages
Yale List
No ratings yet
Yale List
5 pages
Political Science Review Analysis
No ratings yet
Political Science Review Analysis
3 pages
Rethinking Weberian Approaches To Statebuilding
No ratings yet
Rethinking Weberian Approaches To Statebuilding
21 pages
1981-Oszlak-The Historical Formation of The State in Latin America - Some Theoretical and Methodological Guidelines For Its Study
No ratings yet
1981-Oszlak-The Historical Formation of The State in Latin America - Some Theoretical and Methodological Guidelines For Its Study
31 pages
The Political Economy of The Rise and Decline of Developmental States
No ratings yet
The Political Economy of The Rise and Decline of Developmental States
14 pages
Charles Tilly The Formation of National States in Western Europe 1975
100% (1)
Charles Tilly The Formation of National States in Western Europe 1975
723 pages
Why Nations Fail: A Student's Guide
No ratings yet
Why Nations Fail: A Student's Guide
11 pages
Tim Knudsen and Bo Rothstein, State Building in Scandinavia, Comparative Politics, Vol. 26, No. 2 (Jan., 1994), Pp. 203-220
No ratings yet
Tim Knudsen and Bo Rothstein, State Building in Scandinavia, Comparative Politics, Vol. 26, No. 2 (Jan., 1994), Pp. 203-220
19 pages
U1 - Doornbos, Woorward, Roque.
No ratings yet
U1 - Doornbos, Woorward, Roque.
38 pages
The Evolution of Modern States
No ratings yet
The Evolution of Modern States
3 pages
Political Science Foundations
No ratings yet
Political Science Foundations
33 pages
Bringing The State Back in - Capítulos 1 A 4
No ratings yet
Bringing The State Back in - Capítulos 1 A 4
168 pages
Failed States
100% (1)
Failed States
54 pages
Toward Consensus in Interpreting Latin American Politics: Developmentalism, Dependency, and "The Latin American Tradition"
No ratings yet
Toward Consensus in Interpreting Latin American Politics: Developmentalism, Dependency, and "The Latin American Tradition"
20 pages
Barkeyparikh Comparativeperspectivesonthestate
No ratings yet
Barkeyparikh Comparativeperspectivesonthestate
28 pages
Note 27 Oct. 2020 À 10 - 20 - 39 AM
No ratings yet
Note 27 Oct. 2020 À 10 - 20 - 39 AM
3 pages
SIERRA, Gerónimo De. Social Sciences in Uruguay. (2005)
No ratings yet
SIERRA, Gerónimo De. Social Sciences in Uruguay. (2005)
48 pages
Political Thought II Short Reading Material
No ratings yet
Political Thought II Short Reading Material
98 pages
State - Building
No ratings yet
State - Building
4 pages
Ds1-Book Review
No ratings yet
Ds1-Book Review
9 pages
Political Dynamics2
No ratings yet
Political Dynamics2
26 pages
Political Development Theories
No ratings yet
Political Development Theories
9 pages
Theil - ANEP - Bird's Eye View of Corruption
No ratings yet
Theil - ANEP - Bird's Eye View of Corruption
56 pages
Oszlak - 1981 - The Historical Formation of The State in Latin America
No ratings yet
Oszlak - 1981 - The Historical Formation of The State in Latin America
31 pages
Sebastian Mazzuca - Corte2023
No ratings yet
Sebastian Mazzuca - Corte2023
27 pages
Why Nations Fail
No ratings yet
Why Nations Fail
9 pages
(Cambridge Studies in Comparative Politics) Valerie Bunce-Subversive Institutions - The Design and The Destruction of Socialism and The State-Cambridge University Press (1999)
No ratings yet
(Cambridge Studies in Comparative Politics) Valerie Bunce-Subversive Institutions - The Design and The Destruction of Socialism and The State-Cambridge University Press (1999)
223 pages
International Relations Assignment# 02 Submitted By: Maryam Malick Submitted To: Ma'am Shagufta Roll No:68 BS ENGLISH (2nd Semester)
No ratings yet
International Relations Assignment# 02 Submitted By: Maryam Malick Submitted To: Ma'am Shagufta Roll No:68 BS ENGLISH (2nd Semester)
14 pages
Book - Authoritarianism, Elite Origins, Democracy
No ratings yet
Book - Authoritarianism, Elite Origins, Democracy
332 pages
Moore 2001 Political Underdevelopment
No ratings yet
Moore 2001 Political Underdevelopment
35 pages
Modern Political Regime Classification
100% (1)
Modern Political Regime Classification
7 pages
(Centeno, Miguel Angel Ferraro, Agustin E.) State PDF
No ratings yet
(Centeno, Miguel Angel Ferraro, Agustin E.) State PDF
451 pages
Panizza ReformwithoutLosers 2004
No ratings yet
Panizza ReformwithoutLosers 2004
29 pages
Susan Strange - The Defective State
No ratings yet
Susan Strange - The Defective State
21 pages
4.2. P. Evans, Embedded Autonomy
No ratings yet
4.2. P. Evans, Embedded Autonomy
23 pages
States, Nations, and Civilizations: Göran Therborn
No ratings yet
States, Nations, and Civilizations: Göran Therborn
18 pages
Reading Summary 9
No ratings yet
Reading Summary 9
5 pages
DFowkes
No ratings yet
DFowkes
3 pages
State Origins in Anthropological Thought
No ratings yet
State Origins in Anthropological Thought
27 pages
YUHUA WANG, State-In-Society 2.0 Toward Fourth Generation Theories of State
No ratings yet
YUHUA WANG, State-In-Society 2.0 Toward Fourth Generation Theories of State
24 pages
Contribution of Anthropology To The Study of State
No ratings yet
Contribution of Anthropology To The Study of State
7 pages
Beyond States and Empires
No ratings yet
Beyond States and Empires
18 pages
Latin America Writings On Architecture Landscape and The Environment 1876 1925 1st Edition Jacques Lambert Instant Download
No ratings yet
Latin America Writings On Architecture Landscape and The Environment 1876 1925 1st Edition Jacques Lambert Instant Download
79 pages
Govermentality
No ratings yet
Govermentality
7 pages
State Collapse: The Dynamics of Somalia'S Decline
No ratings yet
State Collapse: The Dynamics of Somalia'S Decline
17 pages
Dahlstrom y Lapuente Cap 3
No ratings yet
Dahlstrom y Lapuente Cap 3
35 pages
Dahlstrom y Lapuente Cap 2
No ratings yet
Dahlstrom y Lapuente Cap 2
41 pages
Doner
No ratings yet
Doner
47 pages
Moustafa 2003 PDF
No ratings yet
Moustafa 2003 PDF
49 pages
Trochev 2004 PDF
No ratings yet
Trochev 2004 PDF
36 pages
Does Oil Hinder Democracy?: World Politics April 2001
No ratings yet
Does Oil Hinder Democracy?: World Politics April 2001
38 pages
Cambridge University Press World Politics: This Content Downloaded From 200.89.140.130 On Tue, 29 May 2018 13:49:27 UTC
100% (1)
Cambridge University Press World Politics: This Content Downloaded From 200.89.140.130 On Tue, 29 May 2018 13:49:27 UTC
33 pages
2 Approaches Advantages Disadvantages Rostow's Linear Stages Model
No ratings yet
2 Approaches Advantages Disadvantages Rostow's Linear Stages Model
1 page
Ugandan Media Censorship Factors
No ratings yet
Ugandan Media Censorship Factors
57 pages
Modernization Theory
No ratings yet
Modernization Theory
9 pages
Margaret Mead - 'Changing Perspectives On Modernization' - Rethinking Modernization - Anthropological Perspectives
No ratings yet
Margaret Mead - 'Changing Perspectives On Modernization' - Rethinking Modernization - Anthropological Perspectives
9 pages
Features of Modern Democracy: Key Concepts To Understand
No ratings yet
Features of Modern Democracy: Key Concepts To Understand
10 pages
Semi-Colonialism and Informal Empire in Twentieth-Century China: Towards A Framework of Analysis
No ratings yet
Semi-Colonialism and Informal Empire in Twentieth-Century China: Towards A Framework of Analysis
13 pages
World-Systems Theory Overview
No ratings yet
World-Systems Theory Overview
7 pages
Religion vs. Globalization Dynamics
0% (1)
Religion vs. Globalization Dynamics
5 pages
45-Article Text-177-1-10-20200620
No ratings yet
45-Article Text-177-1-10-20200620
15 pages
Afghan Women's Political Challenges
No ratings yet
Afghan Women's Political Challenges
6 pages
Three Paradigm of Governance and Administration: Chinese, Western and Islamic
No ratings yet
Three Paradigm of Governance and Administration: Chinese, Western and Islamic
30 pages
Dependency Theory PDF
No ratings yet
Dependency Theory PDF
12 pages
Impact of Moderization On Caste System in India
No ratings yet
Impact of Moderization On Caste System in India
7 pages
Marshall Berman On Modernity
0% (1)
Marshall Berman On Modernity
10 pages
KITANO Animism Rinri Modernization The Base of Japanese Robots
No ratings yet
KITANO Animism Rinri Modernization The Base of Japanese Robots
4 pages
Civil Military Relations in Venezuela Gisela Sucre
No ratings yet
Civil Military Relations in Venezuela Gisela Sucre
17 pages
Hassan Fathy Revisited
No ratings yet
Hassan Fathy Revisited
12 pages
Modernization in Higher Education
No ratings yet
Modernization in Higher Education
15 pages
Force Driving Globalization
No ratings yet
Force Driving Globalization
12 pages
(A-008) (Sociology of Education) - 2018
No ratings yet
(A-008) (Sociology of Education) - 2018
4 pages
An Urgent Need To Modernize: Public Health Data Modernization Initiative
No ratings yet
An Urgent Need To Modernize: Public Health Data Modernization Initiative
2 pages
The Communication Model Strategy of Kampoeng Heritage Kayutangan To Maintain Sustainable Development in The Tourism Sector
No ratings yet
The Communication Model Strategy of Kampoeng Heritage Kayutangan To Maintain Sustainable Development in The Tourism Sector
3 pages
Radical Functionalism
No ratings yet
Radical Functionalism
224 pages
(Thomas Ricento (Editor) ) Ideology, Politics and Language
67% (3)
(Thomas Ricento (Editor) ) Ideology, Politics and Language
209 pages
Modernization Theory in the Mideast
No ratings yet
Modernization Theory in the Mideast
29 pages
H. Turner, Fascism and Modernization
No ratings yet
H. Turner, Fascism and Modernization
19 pages
Eric A Nordlinger Soldiers in Mufti - The Impact of Military Rule Upon Economic and Social Change
No ratings yet
Eric A Nordlinger Soldiers in Mufti - The Impact of Military Rule Upon Economic and Social Change
19 pages
MAC341 Development Communication Broadcasting
No ratings yet
MAC341 Development Communication Broadcasting
244 pages
Lesson 3 - Theories of Development
No ratings yet
Lesson 3 - Theories of Development
29 pages
The Making of The Chinese Middle Class: Small Comfort and Great Expectations
100% (1)
The Making of The Chinese Middle Class: Small Comfort and Great Expectations
288 pages

Kurtz Cap 2

Uploaded by

Kurtz Cap 2

Uploaded by

2

The Social Foundations of State Building

the state-building debate: what we do and do not know

War and Political Development

Natural Resource Rents and Politics

a potential cause of serious error. Where resources contribute positively to

The literature linking natural resource wealth and political underdevelop-

What Should We Do Next?

additional challenges for a theory of contemporary

Economic versus Political Development?

The Question of Time

dismemberment of states, and not all suboptimal institutional responses result

State Building in the Late Developers

17 Indeed, Gershenkron’s (1962) argument suggested a set of complementary differences in the

reform, and even subsidize or punitively incentivize the transformation of cor-

the social foundations of institutional order

Theory I: Launching a State-Building Trajectory

system – while maintaining the exclusion of the popular sectors – is crucial

Inter-EIite Cooperation Possible Local Institution

Servile Economy Limits Constrained

Inter-EIite Institution Building Risky; State

Economic and Political Conditions Political Dynamics Initial Institutional Trajectory

(b) Middle- and Working-

for Statist Development

National Institution Uruguay

Initial Trajectory Second Critical Juncture Political Dynamics Long-Run Outcome

Gamson (1968), called a “competitive establishment.”21 However, much as

Theory II: Economic Crisis, the Rise of Mass Political

of such a strategy would be immense, and it would require powerful institutions

research design and plan of the book

An Intertemporal, Two-Stage Research Design

The Formation of National Institutions

figure 2.2. Scoring the cases on the initial trajectory.

The Rise of the Popular Classes

table 2.1. Scoring the Cases in the Second Critical Juncture

by contrast, widespread mass political participation came substantially later.

what is state building?

of this monopoly on violence as a definitional aspect of the state; this is a

to the organizational and logistical improvements it brought to armies them-

What State Building Is Not

In a similar vein, one cannot assume that effective bureaucratic institutions

2006). In addition, perception-based indicators have only a narrow temporal

administrative institutions, whereas in the latter, it cemented their inadequacy.

You might also like