Punishment and Crime in Scandinavia, 1750–2008

Author(s): Hanns von Hofer

Reviewed work(s):
Source: Crime and Justice, Vol. 40, No. 1 (August 2011), pp. 33-107
Published by: The University of Chicago Press
Stable URL: http://www.jstor.org/stable/10.1086/658891 .
Accessed: 03/02/2012 15:39

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to Crime
and Justice.

http://www.jstor.org
Hanns von Hofer

Punishment and Crime in

Scandinavia, 1750–2008

ABSTRACT
Denmark, Finland, Norway, and Sweden have collected criminal justice
data since at least the first half of the nineteenth century. These data can
be used to describe basic trends in criminal justice interventions as regards
violent and property offenses. They can also be used to screen the poten-
tial effects of criminal justice interventions on crime compared with more
basic structural factors such as urbanization, industrialization, migration,
and control of substance use. The marginal effects of changes in criminal
sanctions appear to be negligible for the development of recorded crime.
These changes include use and abolition of the death penalty, the gradual
replacement of imprisonment with alternative sanctions, and the shifting
use of fines. The influence of urbanization and industrialization appears to
be insignificant, but measures to control alcohol use have had a great ef-
fect on violence, as have changes in the opportunity structure on theft.
Viewed over the long term, criminal justice interventions during the twen-
tieth century were inefficient in controlling the development of crime and
criminals in Scandinavia.

There is always an easy solution to every human prob-

lem—neat, plausible, and wrong. (H. L. Mencken)

Statistical data on crime and punishment in the Nordic countries are

unusually good. Both Sweden and Finland have continuous time-series

Hanns von Hofer is emeritus professor of criminology, Stockholm University, Swe-

den. He is grateful for the help and support received from an anonymous referee,
Amber Beckley, Felipe Estrada, Magnus Hörnqvist, Martin Killias, Chris Powell, David
Shannon, Henrik Tham, Michael Tonry, and Per-Ole Träskman. The usual disclaimer
applies.

2011 by The University of Chicago. All rights reserved.
0192-3234/2011/0040-0004$10.00

33
34 Hanns von Hofer

data relating to murder and nonnegligent manslaughter dating back to

the middle of the eighteenth century. Prison data are available in Nor-
way from the mid-1810s (Christie 1975, p. 300). Publication of court
statistics began in 1828 in Norway, 1830 in Sweden, 1832 in Denmark,
and in 1839–42 in Finland (Verkko 1951, p. 41; Hurwitz and Chris-
tiansen 1983, p. 30).
Official statistics became increasingly important in Europe during
the first half of the nineteenth century.1 When in a letter dated May
21, 1830, the King of Sweden told the courts to collect data regularly
on penal and civil matters, he expressed a common belief of the time
that statistical information could be used to arrive at a deeper under-
standing of the causes of crime and to assist in the reform of the society
(Statistiska Kommitténs Betänkande 1910, pp. 288–89). Only a few
years later, in 1835, Adolphe Quetelet expressed very similar ideas,
observing that

the greater the number of individuals, the more does the influence
of individual will disappear, leaving predominance to a series of
general facts, dependant on causes by which society exists and is
preserved. . . . Since the crimes which are annually committed
seem to be a necessary result of our social organization, and since
the number cannot diminish without the causes which induce
them undergoing previous modification, it is the province of the
legislator to ascertain these causes, and to remove them as far as
possible. (Quetelet 1842, pp. 96, 108)

Over time, confidence in criminal justice statistics waned and the

general focus shifted toward the study of the “criminal man.” The
production of criminal justice statistics, however, never stopped, which
leaves us today in the fortunate possession of a rich collection of con-
tinuous data describing the activities of the criminal justice authorities
in Scandinavia.
The main purpose of this essay is to make these data better known
outside Scandinavia and to show that they can be used not only by
historians but also by criminologists. They make it possible to for-
mulate well-grounded critiques of traditional policies based on the be-
lief that unwanted social behavior can be successfully controlled by

1
Compare Ian Hacking (1990, p. viii), who writes about “the avalanche of printed
numbers from 1820 to 1840.”
 Punishment and Crime in Scandinavia 35

punishment. Using secular time series on homicide, rape, robbery, and

theft, we can show that changing forms and patterns of punishments
have had little or no effect on the long-term development of crime.
The major changes relate to alternative explanations: violence seems
to covary with the consumption of alcohol and theft with changes in
opportunity structures.
The inevitable question is whether criminal justice data can be used
in the way I do here. Three aspects of this contested question are
addressed: technical problems, what criminal justice data actually show,
and fundamental problems associated with interpretation of the data.
A review of the debate is not intended. I present my own position in
these matters.
Those who work with time-series data based on criminal justice sta-
tistics that cover brief periods of a few years or decades often note that
such analyses are disturbed by technical factors, legislative changes, the
restructuring of practices in the police or prosecution services or the
court system, and changes in the routines used to collect and present
statistics. Such changes make the use of criminal justice statistics more
difficult, and sometimes impossible. This fact has in turn led to fears
that using criminal justice statistics over an extended period may pre-
sent so many problems that it is best not to use them at all.
In practice, however, the experience is often quite the reverse. In a
long-term perspective, problems of this kind often have only a mar-
ginal effect on the time series, and longer-term patterns are clearly
discernible without being much affected by technicalities. As long as
one restricts oneself to more general analyses, the available Scandina-
vian data appear to be highly usable.
But what do criminal justice statistics actually reflect? Flemming Bal-
vig has summarized a view shared by many Scandinavian criminolo-
gists:

Criminality may be viewed as a result of the interplay between

three parties. These three parties each constitute an essential ele-
ment in the very definition of what constitutes crime, but analyti-
cally we may choose to place more weight on one in relation to
the others, or at least to take only one of them as our point of
departure.
One of the parties . . . is the victim, the person or persons who
are exposed to the act. If we place ourselves on the side of this
party and evaluate the trend in crime, what we see is a curve of
36 Hanns von Hofer

worry or a curve of anxiety. From this perspective, the curve re-

flects what society has chosen to worry about and how this varies
over time.
The second party . . . is the control party. This is the party that
defines (1) the act as a crime, (2) the person or persons who (are
assumed to have) committed the act as the perpetrator, and finally
(3) the person or persons who are exposed to the act as victims. If
we place ourselves on the side of this party and evaluate the trend
in crime, what we see is a control curve. It is a curve that shows on
what and on whom society has chosen to focus its control re-
sources . . . and how this varies over time.
The third party involved in criminality is the act party, that is,
the criminal act itself or the person/group/organization that com-
mits the act. If we place ourselves on the side of this party and
evaluate the trend in crime, what we see is an action curve. This is
a curve that shows the extent of certain types of acts and how
these vary. (1987, pp. 15–16; my translation from Danish)

Although criminal justice statistics in practice are rarely used as vic-

tim statistics in Balvig’s sense, there is little disagreement that they
may describe the penal control activities of the justice system in a quite
reliable fashion; there is, however, considerable disagreement whether
court statistics may validly be used as an indicator of offenders’ behav-
ior. Court statistics are compiled at the end of a long and complicated
selection process, the mechanisms of which are difficult to grasp, es-
pecially in a historical perspective. We lack systematic knowledge about
people’s reporting behavior over time and we have only incomplete
knowledge about police practice and the ways public prosecutors act.
Furthermore, we do not know how court practice has changed over
the decades under study. As a result, many analysts feel compelled to
warn against the use of court statistics. If they are to be used at all,
they should be interpreted as indicators of the exercise of state control
by means of criminal law—more or less independent from individual
offender behavior. Howard Taylor observed of crime statistics in En-
gland and Wales (1914–60): “The immense value of the statistics to
the historian is not that they allow conclusions to be drawn about the
quality or quantity of crime in the real world, but rather that an his-
torically contextualized reading of the statistics greatly assists in the
reconstruction of the supply-side quotas, policies, priorities and politics
that underlay criminal justice” (1998, p. 25).
An opposing point of view can also be maintained, however, partic-
 Punishment and Crime in Scandinavia 37

ularly in relation to longer-term historical developments of court sta-

tistics. This view freely admits that court statistics are useless as indi-
cators of the true volume of offenses and offenders. Too many factors
distort the data. Not every offense is detected and reported and not
every apprehended offender is sentenced. Even so, the possibility can-
not be ruled out that court statistics describe basic trends in offending,
even if one cannot be sure whether the slopes of the observed increases
or decreases are correctly depicted. This view is primarily founded on
the idea that in states with a more legalistic approach, such as the
Nordic countries (see Reichel 2005, p. 118 for details), court sentences
do reflect the guilt of the accused.2 The state deploys a highly bureau-
cratic process to ascribe guilt, and the parties finally recorded in court
statistics as convicted offenders do not find their way there on a totally
arbitrary basis. In most cases, there is a valid connection between the
offense and the sentence. The legalistic approach also limits the powers
of politicians and other stakeholders to influence the production of
these statistics. Thus, to some extent, court statistics do mirror of-
fending behavior.
My own view is that court and other criminal justice statistics are
“artefacts of contemporary social processes” (Morris 2001, p. 125), that
we are not in a position to make a definitive judgment about their
general usefulness one way or the other, and that the dilemma will not
be resolved by means of theoretical analyses and statements. The prob-
lem is empirical in nature; that is, each intended use of criminal justice
statistics must involve a critical determination of whether they are suit-
able or not as the basis for analysis in the specific instance in question.
Therefore, criminal justice statistics are treated below both as indica-
tors of social control through the criminal justice system, and as in-
dicators of (behavioral) crime trends—if there is sufficient reason for
doing so.
The problems of interpretation do not end here, however. The
Swedish historian Hans Andersson observed:

2
Police and prosecution systems are commonly characterized as being based on
“legality” or “expediency” principle. In systems of the first type, the notion of equal
treatment is interpreted to mean that practitioners have no discretion how to proceed.
They must arrest or prosecute whenever they believe an offense has occurred. In
expediency systems, practitioners may exercise discretion in each case to determine
how the public interest will best be served and may dismiss cases even if they believe
guilt can be proven. The Scandinavian countries are generally viewed as legality-
principle jurisdictions.
38 Hanns von Hofer

Indeed, the link between history and criminology is crucial—the

combination may help both a criminologist and a historian to real-
ize that his subject is a social construct. What constitutes a crime
is determined in the minds of men, and in the written laws that
define which acts are legal and which are not. When a historical
perspective is applied to the comparative study of legal culture,
you can see that some acts are legal at one time in a given society,
and others are not. But changes are most commonly undramatic;
shifts take place in degrees rather than in categories. You can also
see how legal change is linked to social, political and cultural
transformations. Here one should be wary of the risks associated
with a determinist[ic] outlook. Legal historians seem prone to ex-
plaining legal change by means of a simplified sociocultural model.
The concept of legal culture, on the other hand, provides a tool
for investigating why people choose to act in certain ways and how
mentalities and discursive fields are constructed in relation to the
legal system. (2008, p. 14)

If this is true, it is reasonable to ask whether it is meaningful to

study how crime has developed over longer periods of time. Balvig
raises precisely this question:

Every period of time has its special characteristics, a specific spirit

of the time. But this “spirit” also has its own time, its characteris-
tic way of relating to time. The important point in this reasoning
is that you have to be cautious about applying the specific “spirit”
characteristic of the present time to earlier periods of time. To be
very concrete: We are living in a period in which time is perceived
and analyzed in terms of development and trends. We are very
concerned about whether things in our surroundings are increasing
or decreasing. Because of this it feels natural and in a way it is
doubly in line with the spirit of the time to analyze crime trends
and to analyze how trends in crime compare to other things.
However, it is problematic to extend this concern and view to
other periods where this spirit did not prevail. If you do so, you
engage in a kind of epoch colonialism.
For this reason you can discuss whether it is at all meaningful to
ask whether crime has increased over the last three centuries. Per-
haps we can compare the crime situations and identify similarities
and differences between the present and the past. The problematic
thing is analyzing and conceptualizing possible differences as
something gradually and continuously changing. . . .
 Punishment and Crime in Scandinavia 39

The first and most fundamental requirement we make of an

analysis of the way crime changes over long periods of time, then,
is that we divide the analysis up in a meaningful way, so that the
main question will not be to look at crime trends over long peri-
ods but 1) to look at the variation in short relatively homogeneous
periods (periods characterized by the same cultural spirit of the
time) and 2) to look for characteristic differences and similarities
between these homogeneous periods.
The crux of the matter is of course how to create this division.
This is completely dependent on which specific qualities—essential
characteristics—we want to illuminate. And this is a question of
theory, of frames of reference and schemes of interpretation. It re-
quires a thesis—a controlling idea. [For example,] in what way do
we want to understand the act of theft—and in what context?
(1988, pp. 31–32)

Balvig makes an important point when the data are primarily viewed
as indicators of criminal behavior and changes in this behavior (the
“act curve”). It may be the case, for example, that theft offenses in
1850 and theft offenses in 1950 are not comparable phenomena in
terms of their content and significance. Or, as Robert M. Morris put
it: “They [crime statistics] are of their time, and are appealed to, and
employed for, the uses of that time” (2001, p. 125). The basic question
that follows is, Does time matter—always? The answer is not as
straightforward as one might believe. See, for example, McDowall and
Loftin (2005), who claim that U.S. homicide rates between 1925 and
2000 show little evidence of historical contingency.
Thus, because criminal justice statistics may perhaps be interpreted
as indicators of other essential social conditions and of changes in such
conditions, one could also view the study of whether the phenomenon
itself has varied over time—that is, irrespective of content, meaning,
or significance—as a fruitful exercise. This view has been advocated by
Nils Christie: “In my opinion, criminology is of very little interest in
itself. It has somewhat greater interest when it comes to helping society
to cope with crime. But even more important is the enormous potential
criminology has for helping society to understand itself. . . . Crimi-
nological data are central indicators of social conditions. We have to
use criminological data as a mirror of society” (1976, p. 144).
Christie argues that criminal justice statistics should be viewed as
data that describe trends in and the effects of the system of formal
40 Hanns von Hofer

control. But these data could also express, for example, the degree of
social inequality (e.g., as regards gender, age, and poverty) and they
could be used as indicators both of power and influence and of the
lack of power and influence.
This essay is inspired by this point of departure but has somewhat
less high ambitions. It is a fairly traditional analysis of criminal justice
statistics and the associations between them and a few relevant crimi-
nological correlates. My approach is based on the idea that the treach-
erous criminological undergrowth must first be cleared away before
the field is opened up for more “grandiose” interpretations.
My general frame of reference is derived not from historiography3
but from macrocriminology. Long-term criminal justice time-series
data provide unique opportunities to handle theories that postulate a
relationship between punishment, crime, and the structure of society
(Killias 2002, p. 25; Liu 2005, pp. 620–21). The focus is on Swedish
data, since they have been subjected to the most extensive analyses (von
Hofer 2008). Data from Denmark, Finland, and Norway are discussed
in less detail, and only to the extent that they exist in published form,
since I have not conducted any further data collection of my own. In
comparing Scandinavian countries, the main emphasis is placed on sim-
ilarities, not dissimilarities. It is tempting to widen the comparisons to
include other European countries, because many observations seem to
have strong international parallels. Such comparisons could also be
used to falsify idiosyncratic national explanations. Nevertheless, I de-
cided not to attempt systematic comparisons, because the similarities
and dissimilarities in a pan-European perspective deserve comprehen-
sive studies of their own.4 Only a few references to countries outside
Scandinavia appear in the text.
I mostly describe traditional crime categories—that is, crimes of vi-
olence (homicide, criminal assault, rape) and property (theft and rob-
bery)—and their punishments as reflected in statistics relating to court
sentences (such as imprisonment, fines, suspended sentences) and other

3
For a short bibliographical note on current research into the European history of
crime and criminal justice since 1750, see Emsley (2007, pp. 275–78). See also Emsley
(2002, 2005), Rousseaux (2006), and Reinke (2009) for roundups of recent research in
Britain, France, and Germany.
4
See, e.g., Manuel Eisner’s work on violence (Eisner 2003, 2008) and Howard,
Newman, and Pridemore (2000) and Killias (2002, p. 102) for valuable general over-
views.
 Punishment and Crime in Scandinavia 41

findings of guilt (such as summary fines and waivers of prosecution),

here referred to as conviction statistics.
The data consist of national aggregates. It is possible that local and
other diversities have been obscured. However, because in the bureau-
cratic nation state punishment and crime are defined by national penal
codes encompassing the whole country, an aggregate level analysis is
justifiable and appropriate.
The main focus is on description of long-term developments; short-
term fluctuations and their explanations are generally not my focus.5
Yet, if explanations are offered, they follow the idea of “strong” and
“weak” explanations (Törnudd 1996, pp. 120–30). The eventual influ-
ence of “weak” causal agents can be studied only by holding constant
“strong” factors such as the opportunity structure or demographic and
legal changes, which usually account for most of the variation observed.
Thus, the focus is on “strong” factors.
Likewise, in-depth numerical analyses have been omitted. Instead,
preference is given to graphical methods of presentation that benefit
from the striking simplicity of long-term uninterrupted time-series
data.6 The terms “development” and “trend” are used interchangeably,
with trend—at first sight—being void of any deterministic connota-
tion.7 The reader is also cautioned against rash level comparisons be-
tween countries (cf. Falck, von Hofer, and Storgaard 2003, pp. 24–25;
Aebi et al. 2006, p. 23).
I draw substantially on the following works: Verkko (1951), Christie
(1975, 1982), Hurwitz and Christiansen (1983), Balvig (1987), Greve
(1996), von Hofer (2003a, 2008), and Lenke (2009). With a few ex-
ceptions, only books and articles written in English are mentioned be-
low, which may leave readers with a somewhat stinted picture of Scan-
dinavian research in the field.
I draw a number of main conclusions concerning the Swedish data:

• Long-term trends in homicide, criminal assault, rape, robbery and

theft seem to describe the course of traditional crime reliably.

5
Compare the concept of the longue durée in historiography (Braudel 1980, p. viii).
For a broad outline of some “short-lived” changes, see Killias (2006).
6
Admittedly, it may contain potential subjective biases—“the beauty of the data is
in the eye of the beholder” (Ball and Wood 1996, p. 144).
7
In fact, the term “drift” (from the stochastic time-series literature) may have been
more appropriate. It was not used in order to keep the terminology simple.
42 Hanns von Hofer

• Industrialization and urbanization have had a minor effect on the

development of crime, but measures to control alcohol use have been
strongly correlated with reduction in violence as have changes in the
opportunity structure with changes in theft.
• After World War II, traditional crime policy measures—reliance on
police and different forms of punishment—have not produced the
desired effects of protecting ordinary citizens and reforming offend-
ers.
• An empirically informed criminal justice policy should look first for
means of prevention other than those offered by the criminal justice
system.

The remainder of this essay is organized as follows. Section I con-

cerns the development of murder and nonnegligent manslaughter, as-
sault, and rape in Sweden and the other Scandinavian countries. This
is followed in Section II by theft offenses and repeated thefts and rob-
bery. Section III examines trends in convictions and sanctions in the
light of a number of specific variables such as gender, age, nationality,
and prior involvement in crime. Section IV summarizes major findings
and Section V discusses their policy implications.

I. Violence
That this essay starts with a discussion of violence should be under-
stood as a reference to modern thinking, in which individual violence
occupies one of the top rankings among what are deemed social evils.
It has not always been so. Although the current Swedish Penal Code
of 1962 begins its description of specific crimes with crimes of violence,
the previous Penal Code of 1864 started with crimes against religion.
The definitions of violence that I use are purely ad hoc and have been
determined by what is available from the statistical sources: homicide
(infanticide excluded), criminal assault, and rape. Robbery has been
treated as a property crime.

A. Homicide
Two historical series are available in Sweden for murder and non-
negligent manslaughter. Data are available from population and vital
statistics from the 1750s onward, and from conviction statistics from
the 1830s onward (see also von Hofer 1990). Although these two sta-
tistical series are based on different units of measurement (i.e., “killed”
 Punishment and Crime in Scandinavia 43

FIG. 1.—Sweden: Homicide. Number of persons killed according to vital statistics, 1754–
2008, per 100,000 population. Source: von Hofer (2008), table 1.1; updated.

and “killers”), there is a high degree of correlation between them—

suggesting that homicide data may be seen as rather reliable measures
of deadly violence, as an “action curve” in Balvig’s sense.
As figure 1 shows, the most striking feature of the long-term trend
from the 1750s to today is that the rate of homicide victims per capita
has scarcely changed. This pattern of stability remains, even if there is
evidence that the level of homicide may be underreported in the late
eighteenth and early nineteenth centuries (by factors of 2 and 1.5, re-
spectively; Kivivuori and Lehti, in this volume). Whether improved
medical treatment has had an effect on homicide statistics is a contested
question. Næshagen (2005) is a strong supporter of this hypothesis as
regards the historical development of homicide in Norway, yet presents
little convincing evidence. Granath (2008) rejects this idea where
deadly youth violence in modern Sweden is concerned, since the great
majority of homicide victims die immediately at the crime scene.
In relative terms, the path followed by the homicide curve fluctuates,
as can be seen by an upswing in the first half of the nineteenth century,
a low level during the period between the world wars, and an increase
44 Hanns von Hofer

after World War II. The last two trends have also been demonstrated
in other European countries (Eisner 2003, 2008).
National trends in alcohol consumption8 in Sweden closely resemble
those for homicide (Lenke 1990; Norström 1998). Alcohol consump-
tion peaked during the middle years of the nineteenth century (Willner
2001, p. 68), at which time homicides were at their highest. Both series
display their lowest levels during the period between the world wars,
when levels of alcohol consumption were held in check by an alcohol
rationing system. Despite this close covariation, there is consensus
among Scandinavian researchers that alcohol consumption should be
considered as a contributing factor only—dependent on specific cir-
cumstances—to lead to deadly violence (Lenke 1990, p. 135; Rossow
2001).
As figure 2 shows, only the years around the 1870s and 1980s need
additional explanations.9 As regards the later period discussions have
focused on, among other things, the significance of increased social
marginalization and immigration (von Hofer 1990; Wikström 1992).
For the most part, though, homicide has fluctuated in a more or less
erratic fashion, once changes in alcohol consumption have been taken
into account. This is inconsistent with global interpretations such as
the relevance of the large-scale social transformation of Swedish society
that occurred during the nineteenth and twentieth centuries. The tran-
sitions of Sweden from an agricultural to an industrial and then a post-
industrial society have not been reflected in homicide trends to any
great extent.10
Even the choice of sanction has undergone tremendous changes
without leaving significant imprints on the historical development of
homicide. In the middle of the eighteenth century, the death penalty
was widely used (and not only for murder). In 1779, during the reign
of Gustav III, the use of the death penalty was reduced. During the
first half of the nineteenth century, the use of pardons was extended.
Capital punishment was reduced still further with the enactment of the

8
The alcohol factor has always played an important role in the Scandinavian crim-
inological literature (Hurwitz and Christiansen 1983, p. 239; Lenke 1990; Willner
2001); cf. also Edman and Stenius (2007); Sulkunen et al. (2000); and, for a non-
Scandinavian perspective, Leigh and Room (2002).
9
Figure 2 displays the residuals of a simple regression of yearly per capita alcohol
consumption on yearly homicide rates.
10
See Eisner (2008, p. 311) for a similar conclusion in a broader European per-
spective.
 Punishment and Crime in Scandinavia 45

FIG. 2.—Sweden: Homicide. Number of persons killed according to vital statistics, 1861–
2005, corrected to account for per capita (15 years and older) alcohol consumption. No annual
alcohol consumption data are available before 1861. Source: von Hofer (2008), tables 1.1 and
3 (app.); updated.

new Penal Code in 1865, after which the number of executions never
exceeded more than one or two per 5-year period. In 1877, public
executions were abolished. The last execution was performed in 1910,
when the guillotine was used for the first and last time. Abolition was
finalized in 1921 for crimes committed during peacetime and in 1973
for crimes during wartime. The prohibition of the death penalty is now
embodied in the Swedish constitution.
Today, there are basically two choices of sanctions for homicide:
approximately 60 percent of convicted killers are sentenced to impris-
onment and 40 percent are sentenced to compulsory residential psy-
chiatric care. The extent to which convicted killers have been labeled
insane has varied in the statistics maintained since the 1910s—from 28
percent in the 1910s to 70 percent in the 1940s and again in the mid-
1970s. These high rates coincide with low homicide rates and a period
when the treatment ideology flourished in Sweden.
The use of life imprisonment decreased sharply during the second
half of the nineteenth century (see fig. 3). In the middle of the nine-
46 Hanns von Hofer

FIG. 3.—Sweden: Number of persons sentenced to life imprisonment, 1859–2008. Source:

von Hofer (2008, p. 250).

teenth century, even crimes other than murder (e.g., rape, robbery,
repeated theft) could carry a life sentence. Toward the very end of the
period, however, the use of imprisonment for life began to increase
again.
When Swedish trends in the application of the death penalty and in
the number of homicide victims are compared, it is not possible to
discern deterrent effects of the use of capital punishment. Homicide
rates dropped to their lowest level in the years following 1921, when
the death penalty was abolished and alcohol restrictions were intro-
duced. They were highest during the 1840s, when at least 10 persons
were executed each year. Today, homicide rates are no higher than in
the 1870s, the 1830s, and the 1750s—all periods when many executions
were performed. A nil effect of the use of the death penalty can also
be demonstrated by means of time-series analyses on a year-by-year
basis (von Hofer 1990).
Figures 4a and 4b visualize the nil effect between executions and
homicide from 1754 to 1921 with the help of a “Phillips curve” (Phil-
lips 1958). If the number of homicides followed the claims of deter-
rence (or of incapacitation) theories, an overall negative temporal slope
 FIG. 4.—Sweden: Executions versus homicides, per 100,000 population, Phillips-type curve.
a, 1754–79; b, 1779–1921. Source: von Hofer (2008), tables 1.1 and 6.1.
48 Hanns von Hofer

FIG. 5.—Norway: Homicide. Number of persons killed according to vital statistics, 1876–
2008 (5-year average), excluding the years 1941–45, per 100,000 population. Source: Statistisk
sentralbyrå (1995), table 4.11; updated.

should be observable between executions and homicide, since more

numerous executions should yield fewer incidents of homicide and vice
versa. However, such a temporal pattern is not discernible.11
Figure 4a clearly shows that both executions and homicides de-
creased between 1754 and 1779 (a positive temporal relationship is
displayed). There is no clear temporal pattern after 1779 until the ab-
olition of the death penalty in 1921 (fig. 4b). No different picture
emerges when the number of executions is lagged 1, 2, or 3 years.
Denmark and Norway. Since the second half of the nineteenth cen-
tury, the homicide levels and trends in Denmark (Lenke 1990, p. 119)
and Norway (see fig. 5) have been largely similar to those found in
Sweden.12 The levels were particularly low during the interwar years

11
Instruction: start at year 1754 in fig. 4a and follow the curve year by year. See
the appendix for a more detailed explanation of the Phillips curve.
12
See also Eisner (2008, p. 297). In 1814, Norway was forced into a union with
Sweden. Norway kept control of its own independent institutions, except for the foreign
service. The countries separated peacefully in 1905.
 Punishment and Crime in Scandinavia 49

(there were various measures of alcohol restrictions, e.g., prohibition

in Norway and high taxes on alcohol in Denmark).
As regards sanctions, the Danish straffeloven from 1866 retained the
death penalty for murder, among other things, but only four men were
executed between 1866 and 1892, when the last (peacetime) execution
was carried out. No women were executed during this period. The
death penalty was abolished during peacetime in 1930 and during war-
time in 1978, and in 1993 the death penalty disappeared from all areas
of Danish legislation. Following World War II, however, 46 individuals
were executed for crimes committed during the German occupation
(Dahl 1989, p. 35; Greve 1996, p. 50).
In Norway the final civil execution took place in 1876 and the final
military execution in 1948. The death penalty was removed from civil
penal law on January 1, 1905 (the decision had been made as early as
1902), when Norway gained her independence from Sweden. Follow-
ing World War II, however, 37 individuals were executed for crimes
committed during the German occupation (Dahl 1989, p. 35). The
death penalty for crimes committed in wartime was finally abolished
in 1979.
In Finland, the purge of collaborators was limited, owing to the
complicated circumstances of the Finnish war history.13 Pardons were
granted to a great extent and no death sentence was carried out after
World War II ended (Dahl 1987, p. 194).
Life imprisonment was abolished in Norway in 1981 (the maximum
sentence is instead set at a prison term of 21 years). The number of
lifers in Denmark is very small (19 as of September 1, 2008), while the
corresponding numbers were 155 individuals in Sweden14 and 148 in
Finland (Aebi and Delgrande 2010, table 8).15 These differences show
very clearly the significance of national crime policies as regards the
use of the most severe sanction allowed by law—despite the substantial

13
Approximately one-third of Soviet POWs in Finland—22,000 out of 67,000—
died in captivity between 1941 and 1944, mainly of hunger and disease; 1,200 Soviet
prisoners were shot—most of them illegally (Kujala 2009). Altogether 443 Soviet cit-
izens and 85 Finnish citizens were executed after military or civil court proceedings
(Lindstedt 2000).
14
A recent law reform in Sweden is expected to decrease the number of lifers. In
2009, the penalty for murder was increased from 10 years or life to 18 years or life.
15
In 2008, Denmark counted approximately 5.5 million inhabitants, Finland 5.3,
Norway 4.8 and Sweden 9.2 million.
50 Hanns von Hofer

commonality of values that is generally assumed to exist among the

Scandinavian countries (cf. Andersson and Hilson 2009).
Finland. The last death sentence (in peacetime) was carried out as
early as 1826,16 but both the level of and trends in murder and non-
negligent manslaughter differ from those in the rest of Scandinavia.
The reader may wonder why the “spikes” in figure 6a are displayed,
which may make it more difficult to read the figure. It was a deliberate
choice because the uncorrected series demonstrates in a very graphic
way both the stability and the variability of homicide rates contingent
on the societal context and political events. The spike in 1808 coincides
with the Finnish War between Sweden and Russia in 1808–9 and the
spike in 1918–19 with the Finnish civil war in 1918.
As early as the second half of the eighteenth century, the level of
homicide was higher in Finland than in Sweden (Verkko 1951).
Whereas in Western Europe and Scandinavia homicide rates reached
their all-time low during the period between World War I and World
War II, in Finland they climbed to new heights in the aftermath of the
civil war. Notwithstanding these fluctuations, it is possible to discern,
even in Finland, a strikingly stable annual “basic level,” at least from
the 1820s to the 1990s, of approximately three homicides per 100,000
inhabitants a year. Janne Kivivuori and Matti Lehti present a detailed
account of the development of homicide in Finland and Sweden else-
where in this volume.

B. Assault
It is much more difficult to assess the quality of Swedish assault
statistics (see fig. 7). There is no parallel control data series for assault
convictions in Sweden equivalent to vital statistics data for homicide,
although the correlation between the homicide and assault series is
considerable. It may be assumed to be unlikely that people’s readiness
to report assault crimes has diminished historically. On the contrary,
it is probable that levels of tolerance for individual violence have de-
clined over time and have led to an increase in the likelihood that
assaults are reported. It is thus highly likely that assaults were more
common in the nineteenth century than in the twentieth century.17
16
Landtman (1932). The punishment became deportation to Siberia (Anttila and
Heinonen 1975, p. 132).
17
With regard to collective violence, however, the twentieth century has been a
particularly dark period in history considering World War I, World War II, and other
mass atrocities.
 FIG. 6.—Finland: Homicide. Number of persons killed according to vital statistics, 1754–
2008, per 100,000 population. Source: Verkko (1948, pp. 9–10); updated. In panel a, 1807,
1814, and 1965 are missing. In panel b, 1808, 1918, and 1919 are removed and 1965 is missing.
52 Hanns von Hofer

FIG. 7.—Sweden: Assault. Number of convicted persons, 1841–2008, per 100,000 population.
Source: von Hofer (2008), table 2.1; updated.

However, it is reasonable to assume that the meaning of assault as a

legal concept has changed substantially since the 1800s. In the nine-
teenth century, convictions for assault rarely related to the maltreat-
ment of children or women. According to prevailing social and legal
notions, these acts were not defined as crimes—other than in serious
cases. Today, about one-third of assaults brought before the Swedish
courts relate to violence against women (primarily in close relations).
Thus there have been important internal shifts in the types of assault
offenses that are reflected in the conviction statistics. Crimes of assault
against men have declined, while those against women and children
have increased. Likewise, the control of young men by means of the
penal law has also increased, as can be seen from figure 8. In 1924, 15
young men aged 15–17 years were convicted of assault. By 1995, the
corresponding number had risen to 1,826.
The trends in assault convictions to a large extent covary with trends
in alcohol consumption (particularly with the consumption of beer and
spirits in public places; Lenke 1990; Norström 1998).18 Much of the

18
Figure 9 displays the residuals of a simple regression of yearly per capita alcohol
consumption on yearly assault conviction rates.
 Punishment and Crime in Scandinavia 53

FIG. 8.—Sweden: Assault. Number of convicted persons, aged 15–17, 1913–2008, per 1,000
population aged 15–17. 1913–89: age when offense committed; 1990–2008: age at adjudication.
The years 1948, 1951–54, 1957, and 1972 are missing. Source: von Hofer (2008), table 2.5;
updated.

variation in the assault rate disappears when changes in alcohol con-

sumption are taken into account. Once again, it seems that the major
transformation of Swedish society witnessed over the past 160 years
has not been accompanied by an overall increase in the extent of crim-
inal assault. The opposite is closer to the truth.
The increase after the mid-1960s (see figs. 7 and 9) seems to be both
real and conceptual. It is real in the sense that factual violence has
increased (as, e.g., homicide data from vital statistics clearly show; cf.
also the increases in rape and robbery shown below). The increase is,
however, also conceptual because of application and enforcement of
wider social and legal definitions of what is considered to be violence.
The upswing since 1980 has, for example, to do with changes in leg-
islation relating to violence against women, and an increase in the fre-
quency of prosecution of violence among young men (see fig. 8). Von
Hofer (2004; with reference to Soman [1980] and Gurr [1981]) has
speculated that underlying this widened conception of violence are
three basic characteristics of modernization: fewer deviations from ex-
54 Hanns von Hofer

FIG. 9.—Sweden: Assault. Number of convicted persons, 1861–2005, corrected to account

for per capita (15 years and older) alcohol consumption. No annual data on alcohol consumption
are available prior to 1861. Source: von Hofer (2008), tables 2.1 and 3 (app.); updated.

pected normality, increased standardization of life processes, and

greater intolerance of hardship and pain. These traits have decreased
people’s tolerance of what is considered acceptable use of force (as a
source of pain) and thus widened sensitivity toward and conceptions
of what is experienced as violence. That an increased level of sensitivity
to violence can then lead to a broadened definition of criminal violence
may in turn be linked to various interest groups that increasingly define
people’s personal integrity in terms of human rights (Tham 2001).
This definition makes the step to interpreting violations of these rights
as criminal acts a short one.19
The increase in assault convictions in Sweden after 1965 coincided
with an increase in the use of imprisonment as a penal sanction in
connection with assault offenses (Lenke 1990, pp. 143–44). In 1965,
the new Penal Code stiffened the penalty for assault. Before 1965,

19
A case in point is the preamble and art. 4(c) of the Declaration on the Elimination
of Violence against Women (United Nations 1993). An opposite policy approach, which
does not advocate the use of punishment, is the public health approach favored by the
World Health Organization (WHO 2002).
 Punishment and Crime in Scandinavia 55

FIG. 10.—Police-reported assault in Denmark, Finland, Norway, and Sweden from 1950–56
to 2008 per 100,000 population. Source: Falck, von Hofer, and Storgaard (2003), table 2;
updated.

prison sentences accounted for approximately 10 percent of the sanc-

tions imposed in connection with assaults. Thereafter, that proportion
quickly increased to 25–30 percent.
Denmark, Norway, and Finland. Post–World War II assault trends
in Denmark, Norway, and Finland have paralleled that found in Swe-
den. Reported assault offenses increased markedly in the crime statis-
tics of all three countries (Sirén 2002, p. 23; Falck, von Hofer, and
Storgaard 2003, pp. 8–9). By contrast, data from hospital statistics and
self-report and victim surveys indicate quite stable levels throughout
Scandinavia (Estrada 2006; Takala and Obstbaum 2009).
The police reports on assault in Denmark and Norway start from
very low levels at the beginning of the 1950s (see fig. 10). The first
self-report surveys conducted in Scandinavia in the mid-1960s did not
include any questions on participation in or exposure to violence. This
fact may be interpreted as indicating that less serious forms of violence
had not been defined as a social and crime policy problem of impor-
tance at that time.
Hurwitz and Christiansen have summarized the development of vi-
olence in Denmark from a longer perspective in the following way:
56 Hanns von Hofer

FIG. 11.—Norway: Assault. Number of convicted persons (misdemeanors excluded), 1923–

2008, per 100,000 population. The year 1996 is missing. Source: Statistisk sentralbyrå (1995),
table 8.9; updated.

“From 1867–70 to 1905 the rate of violent crimes increased from 0.25
to 0.75 per thousand. Thereafter the level of violent crime decreased
markedly, and despite a considerable increase in the 1930s, the level of
violent crime was the same in 1972 as it has been one hundred years
earlier” (1983, p. 36). According to Lenke, Denmark has to be counted
among the low-violence countries (1990, p. 113). The decline at the
beginning of the twentieth century can be linked to the reduction in
the consumption of spirits, which was produced by huge increases in
the price of alcohol during this period (Willner 2001, p. 46). This
pricing policy was an alternative alcohol policy strategy paralleling the
prohibition policies employed in Norway (1916–27) and Finland
(1919–32) and the rationing policy employed in Sweden (1915–55).
The development of assault convictions in Norway for a large part
of the twentieth century is shown in figure 11, and is quite similar to
the trend found in Sweden. Comparing the trends in figure 11 and
figure 10, the reader may get the impression that conviction rates in
Norway have increased at a steeper rate than police-reported offenses.
However, the discrepancy is a visual artifact due to alternate scaling of
 Punishment and Crime in Scandinavia 57

FIG. 12.—Sweden: Rape. Number of convicted persons, 1842–2008. Source: von Hofer
(2008), table 2.3; updated.

the x and y axes.20 Between 1960 and 2007, both indicators increased
about seven times.

C. Rape
In contrast to the official data series on homicide and to some degree
also assault, it is unlikely that figure 12 represents an accurate descrip-
tion of the “real” trend in rape offenses in Sweden. The frequency of
adjudicated rape cases was extremely low during the second part of the
nineteenth century: on average, fewer than 10 cases per year entered
into the court statistics.
During the nineteenth century, the judicial response to rape bore
little real relation to the actual frequency of this crime. Historians have
shown that convictions for rape were very rare in prestatistical times
(Bergenlöv and Lindstedt Cronberg 2002, p. 95). The low rates may
reflect the weak social and legal position of women in Swedish society,
which meant that in most cases of rape women were not afforded the
opportunity to bring their violators to justice. They were given as little

20
For further details, see Falck, von Hofer, and Storgaard (2003, pp. 24–25).
58 Hanns von Hofer

opportunity to protect themselves through the courts in relation to

rape as they were in relation to common assault.21
The modern increases in the level of registered rapes and of con-
victions for rape might thus partly indicate a somewhat paradoxical
state of affairs: the improved legal protection leads to an increase in
the level of reported offenses, and thereafter also in the number of
convictions, with no need for the level of actual offenses to have in-
creased to a corresponding degree.
This interpretation, however, does not mean that one can “explain
away” the entire increase after World War II as an unveiling of the
“dark figure”: that is, the same amount of rapes are being committed,
but more are coming to light and being prosecuted. It is highly prob-
able that the risk of rape increased, considering that the direct control
of women (through “watchful eyes”) has diminished in post–World
War II Sweden and the amount of time which women spend alone
with men outside their families has increased—facts that some men
take advantage of.
Likewise, it is most unlikely that the upswing after World War II
can be explained by changes in the choice of sanctions for rape. Then
as now, rape typically carried a custodial sanction, to be served either
in a prison or, less commonly, in a mental hospital.
Denmark and Norway. Nordic crime statistics data for the 1950s
indicate a very low level of reported rape offenses (see fig. 13). Since
then, there have been increases in all countries, at the same time the
reach of the criminal law has expanded. The increases are greatest in
Norway and Sweden (Westfelt 2008, p. 473). These two countries also
seem to be those where the political significance of the women’s move-
ment has been greatest (Tham, Rönneling, and Rytterbro, in this vol-
ume). Rape appears today as a crime with major gender-political im-
plications (“men’s violence against women”). Thus the sharp increase
in rape reports and rape convictions in Sweden at the end of the period
is connected to a major change in legislation in 2005 that further wid-
ened the legal concept of rape.22

21
Compare Georges Vigarello’s (2001) “A History of Rape” for a similar interpre-
tation as regards France, and Ekström (2003), who adds the important aspect that only
certain men were regarded as suitable perpetrators: “The question of whether the
alleged crime [rape] was considered possible to prove was dependent upon who reported
whom for what” (p. 204).
22
In statistical terms, the number of police-recorded rape crimes in Sweden has
reached the same level as the number of registered homicides during the civil war in
 Punishment and Crime in Scandinavia 59

FIG. 13.—Police-reported rapes in Denmark, Finland, Norway, and Sweden, 1950–56 to

2008, per 100,000 population. Source: Falck, von Hofer, and Storgaard (2003), table 3; updated.

In a long-term perspective, the increase in Norway is very similar

to that found in Sweden, although it started somewhat later in Norway
(see fig. 14). Because of the low frequency of rape convictions in Nor-
way, the series displays considerable random variation as indicated by
the slump at the end of the 1990s.23

II. Property Offenses

It can be assumed that many theft crimes—especially those of a minor
nature—are not reported and that many reported thefts do not result
in criminal proceedings (cf. von Hofer and Tham 1989, 2000). Nev-
ertheless, it is highly likely that Swedish conviction statistics provide
an approximate description of the trend of nontrivial theft in Sweden.
This view is based on the following arguments: First, until World War

Finland (see fig. 6a above). The technical problems of Swedish police statistics con-
cerning rape have been analyzed in detail by von Hofer (2000), and they explain a
considerable part of the startling discrepancy between levels of rape reports and rape
convictions. For a contesting view, see Lovett and Kelly (2009, p. 97).
23
To keep fig. 14 simple, no rates per 100,000 of the population have been calculated.
Between 1925 and 2005, the Norwegian population increased by about 70 percent.
60 Hanns von Hofer

FIG. 14.—Norway: Rape. Number of convicted persons, 1923–2008. Source: Statistisk sen-
tralbyrå (1995), table 8.9; updated.

II many theft offenses carried a relatively high punishment. Second,

serious theft offenses (such as burglaries and car theft) are still char-
acterized by a comparatively high reporting rate, as is revealed by both
national victim surveys and insurance statistics. Third, for over 100
years now, the historical trends in police-reported theft offenses in
Stockholm and Gothenburg have displayed a pattern similar to the
trend in conviction statistics for the country as a whole. Fourth, there
appear to be straightforward interpretations (see below) for marked
deviations from these trends.

A. Theft Trends
Theft convictions began at a relatively high level in the early nine-
teenth century and then dropped during the second half—as also hap-
pened in other European countries (von Hofer and Tham 2000, p.
21)—reaching a low after the turn of the century, only to increase
sharply again between 1925 and 1972 (see fig. 15).
Over the period from the 1800s to the 2000s, there are a number
of marked deviations from the main trend. These can be correlated
 Punishment and Crime in Scandinavia 61

FIG. 15.—Sweden: Theft. Number of convicted persons, 1841–2008, per 100,000 population.
Source: von Hofer (2008), table 3.1; updated.

with economic hardship (1842, 1846, 1852), famine (1867–68 and

1917–18), alcohol restrictions (the low level in the early 1920s follow-
ing the introduction of alcohol rationing in the 1910s), and war (1914–
18 and 1940–45). These sudden interruptions show that the number
of convictions for theft is affected by broader social events—even
though each specific crime is a personal and individual act (cf. Quetelet
1842, p. 96).
The downswing between 1850 and 1925—two extreme points in the
historical trend—is largely due to diminishing conviction rates among
women, older men, and persons with prior criminal records. Convic-
tion rates among young men started to increase in the second half of
the nineteenth century, with this increase then becoming particularly
steep between 1925 and 1959, as is shown in figure 16.
The upswing in theft convictions shown in figure 15 after 1925 is
largely attributable to young men, persons with previous convictions
(since World War II), foreign citizens (after World War II), and women
(since the beginning of the 1960s).
The leveling off in the rate of increase during the 1960s and the
62 Hanns von Hofer

FIG. 16.—Sweden: Theft. Number of convicted persons, aged 15–17, 1879–2008, per 1,000
population aged 15–17. 1879–1989: age when offense committed; 1990–2008: age at adjudi-
cation. The years 1948, 1951–54, 1957, and 1972 are missing. Data on petty theft are not
available for 1879–1912. Source: von Hofer (2008), table 3.3; updated.

stabilization seen during the 1970s mainly reflect trends relating to

young men, whereas convictions among women and men with prior
convictions tend to rise. Since young men account for a large propor-
tion of convictions for theft, this group determines the shape of the
main curve. Theft convictions among women stabilized about 10 years
later at the beginning of the 1980s, and in the 1990s also among both
male and female recidivists.
The leveling off in theft convictions appears to reflect real stability
when compared with reported criminality, insurance statistics, and vic-
tim surveys. Burglaries—crimes that have a relatively small dark fig-
ure—have remained more or less stable since the early 1970s, com-
pared with the substantial increases of the previous decades.
The marked drop in theft convictions at the very end of the period
(which has no parallels in national victim surveys and self-report stud-
ies) should be seen in the light of major reorganizations of the police
and the prosecution service in Sweden, which led to a temporary loss
of efficiency and reduced output.
 Punishment and Crime in Scandinavia 63

On closer scrutiny, the increase witnessed during the twentieth cen-

tury can be seen as an S-curve. One not very far-fetched interpretation
is that when a certain basic material standard is guaranteed (much of
the theft criminality in the middle of the nineteenth century has been
explained by the extreme poverty prevalent at the time; Sundin 1976),
convictions for theft follow patterns of access to consumer goods (cf.
Stack 1982; Norström 1988). With more to consume, there is always
more to steal. At the same time, a bountiful supply of goods leads to
a reduction in their relative value and to lower levels of physical and
social control exercised in relation to them.
In 2005, for example, the Swedish police registered 620,000 larceny
crimes. More than half were thefts of or from motor vehicles, thefts
of bicycles, or thefts from shops or department stores. At the beginning
of the twentieth century, neither cars nor bicycles existed in abundance,
nor were there any self-service stores. The opportunity structure, that
is, the physical opportunities available for stealing, has changed a great
deal since then.
If the S-curve model is a valid representation of Swedish theft trends,
it would also imply that we are observing an inflection process that
started long before the 1990s—probably as early as in the 1960s and
1970s (von Hofer and Tham 1989, p. 34). The logic of the model
proposes that it would also be futile to look for a few single causal
factors that could explain the observed inflection process. The change
in a single variable will rarely dominate, since the series lie at the end
of a long causal chain, with their values determined by various variables
that often work in different directions (McDowall 2002, pp. 728–29).
The series will move with the general constellation of prevailing forces
and the S-curve should be seen as the product of incremental changes
in the social fabric of Swedish society.
There is also much evidence that the shifting application of penal
sanctions did not play a decisive role in this process. The choices of
sanctions for theft offenses have undergone major changes during the
twentieth century. In 1905, the risk of being sentenced to prison for
theft was 100 percent, whereas today it is close to zero for first-time
offenders. Nevertheless, in statistical terms it is very difficult to show
that the change in the selection of sanctions for theft has contributed
to the development of theft over the course of the twentieth century
to any significant degree (von Hofer and Tham 1989).
Nor should we overestimate the significance of industrialization and
64 Hanns von Hofer

urbanization:24 convictions for theft were no more numerous in the

1950s than 100 years earlier. The best explanatory model is changes
in the opportunity structure (cf. Kick and Lafree 1985).
Figures 17a and 17b depict the temporal relationship between theft
convictions, the risk of being sentenced to imprisonment, and the
changes in the opportunity structure after World War I. If the number
of theft convictions followed the claims of deterrence (or incapacita-
tion) theories, an overall temporal negative slope should be observable,
since high risks should yield low rates and vice versa. However, there
is no such relationship.
The development in theft convictions is characterized by a remark-
able stability (or even a downward slope; fig. 17b) once it has been
corrected to take account of changes in the opportunity structure.25
The periods of crisis experienced around 1868 (starvation), World War
I, and World War II are clearly visible, and the S-curve formerly seen
during the twentieth century has disappeared. Long-term deterrent or
incapacitation effects are not discernible. Thus the theft trend in Swe-
den becomes an illustrative example of Felson’s argument “that a crime
wave does not depend on increased motivation to engage in crime,
providing that there is greater opportunity to act upon motivations”
(1983, p. 668; cf. Cohen and Felson 1979, p. 589).
Denmark and Norway. The number of theft crimes reported to the
police has risen in all Scandinavian countries since at least the begin-
ning of the 1960s (see fig. 18). In Sweden, the increase started as early
as the mid-1920s. Those trends are on the whole much the same as
those found in other Western European countries (Westfelt 2001, p.
68). Police-recorded theft trends in the 1990s may also have been in
the process of changing direction (cf. Falck, von Hofer, and Storgaard
2003, p. 9; van Dijk 2008, pp. 123–42). The available data from na-
tional victim surveys corroborate the leveling off, showing more or less
stable levels in Denmark, Finland, Norway, and Sweden (Westfelt
2001, p. 76; 2008, p. 462).
In the longer term—from the middle of the nineteenth century—

24
Compare Sundin (1990, p. 189, with further references) and Knafla (1996, p. 38).
A major proponent of the popular industrialization/urbanization hypothesis has been
Shelley (1981; but also 1986). See Neuman and Berger (1988) and Barak (2001) for
discussions of different developmental models of crime and crime control.
25
Approximated by reference to the existing number of cars 1923–2005 in fig. 17b
(cf. Wilkins 1964, p. 54).
 FIG. 17.—Sweden: a, Theft convictions (including petty theft) per 100,000 population versus
risk of being sentenced to imprisonment, 1866–1922. b, Theft convictions (including petty
theft) per 1,000 motor cars versus risk of being sentenced to imprisonment, 1923–2005. Phillips-
type curve. Source: von Hofer (2008), tables 3.1 and 5 (app.); updated.
66 Hanns von Hofer

FIG. 18.—Police-reported thefts in Denmark, Finland, Norway, and Sweden, 1950–56 to

2008, per 100,000 population. Source: Falck, von Hofer, and Storgaard (2003), table 5; updated.

there are both similarities and differences between the trends in Den-
mark and Norway, and Sweden (no information is available regarding
Finland). In the absence of specific theft data in Denmark and Norway,
all convictions for penal code offenses have been used as a proxy mea-
sure (see figs. 19a and 19b).26
The similarities between the countries are found in the relative sta-
bility from the mid-nineteenth century to the outbreak of World War
II. The differences are that the trend is slightly downward in Norway
until the end of the 1950s,27 noticeably stable in Denmark during the
same period, but upward in Sweden after the mid-1920s. World War
I and World War II leave an impression on the statistics in all of the
countries, but to a varying extent. Following World War II, convictions
for theft offenses increased in all countries (including Finland), with a
clear change toward the end of the period (also evident in data from
victim surveys).
Because the curves for Denmark and Norway do not include con-

26
For Norway, this approximation is problematic at the end of the period, when
figs. 18 and 19b are compared: fig. 18 shows a decrease even in Norway, while fig. 19b
displays a continuous increase.
27
See also diagram 3.3-1 in Christie (1975, p. 241).
 FIG. 19.—a, Denmark: Number of convicted male offenders, 1867–1972, per 100,000 of
male population above age of criminal responsibility, and number of persons convicted for
offenses against the Penal Code, 1950–2008, per 100,000 population. Sources: Christie (1975,
pp. 256–57); Falck, von Hofer, and Storgaard (2003), table 13; updated. b, Norway: Number
of convicted offenders, 1835–1972, and number of persons convicted for offenses against the
Penal Code, 1950–2008, per 100,000 population. Petty offenses are excluded. Sources: Christie
(1975, pp. 242–51); Falck, von Hofer, and Storgaard (2003), table 13; updated. c, Sweden: Theft.
Number of convicted persons, 1841–2008, per 100,000 population. Source: von Hofer (2008),
table 3.1; updated.
68 Hanns von Hofer

FIG. 20.—Sweden: Repeated theft. Number of convicted persons, 1832–2005 (5-year average),
per 100,000 population. There is a change of definition from 1916. Data from 1995 are esti-
mated. Source: von Hofer (2008), table 4.1.

trols for changes in the opportunity structure (and that Denmark’s

curve up until 1950 relates only to men, and not to men and women),
it is likely that the level of stability is exaggerated. The more likely
scenario is one of declining trends in both Norway and Denmark until
the late 1950s.

B. Repeated Theft and Robbery

Because of shifting classifications in the official Swedish statistics,
repeated theft has been defined as the number of persons convicted
three or more times for theft (1832–1915) and as the number of per-
sons sentenced for theft who have previously been sentenced three or
more times for so-called crime register offenses (starting in 1916). The
trend in repeated theft offenses and robbery is shown in figure 20.
After 1850, convictions for repeated theft offenses declined until the
outbreak of World War II. Convictions for robbery offenses also follow
a similar pattern as shown in figure 21.
After World War II, the number of convictions for repeated theft
and robbery increased sharply. No leveling off can be observed before
 Punishment and Crime in Scandinavia 69

FIG. 21.—Sweden: Robbery. Number of convicted persons, 1841–2008, per 100,000 popu-
lation. Source: von Hofer (2008) table 4.2; updated.

the 1970s. The increase in robberies at the very end of the period is
mainly explained by an increase in “mugging” offenses among young
males.
Compared with the trend in other theft offenses, the trend in re-
peated theft is distinctive in one essential respect: the sharp increase
occurs about 15 years later than it does for other thefts: the theft curve
starts rising after 1925, while repeated thefts increase after 1940. There
is a good deal of evidence to indicate that the statistics reflect this trend
correctly—since the dark figure of repeat offenders can be assumed to
be small.28 In the nineteenth century, rates of persistent recidivism were
already high among persons with several previous convictions, even if
the number of recidivist offenders was small. This continued to be the
case from the first half of the twentieth century until the present (von
Hofer 2008, p. 120).
The notion that the improved welfare of most segments of the pop-

28
The argument refers to offenders, not to offenses. Even if a lot of repeat offenses
go undetected, the individual risk of being arrested increases with the number of
offenses committed (cf. Killias 2002, p. 65).
70 Hanns von Hofer

FIG. 22.—Sweden: Persons convicted of theft offenses (aged 15–17), 1921–90, lagged 15
years, and persons convicted of serious offenses who have been registered at least six times in
the crime register (or equivalent), 1936–2005. Data are in part based on estimations. Source:
von Hofer (2008), tables 3.3 and 4.1.

ulation would reduce criminality among the poorest seemed to have

been confirmed during the period up until World War II. Extensive
substance use and its political control play an important role in ex-
plaining the overall trend. It is unrealistic to view alcohol, however, as
the sole explanation for the sudden change that occurred following the
outbreak of World War II. Instead, it cannot be ruled out29 that the
development was also determined by the increasing “input” of young
offenders (as measured by the number of youths aged 15–17 convicted
of theft offenses) half a generation earlier (Tham and von Hofer 2009).
Figure 22 shows a substantial increase—in the right causal order—
in both groups following the end of World War II and a leveling off

29
For recent studies on adverse effects of juvenile justice interventions, see McAra
and McVie (2007); Gatti, Tremblay, and Vitaro (2009); and Petrosino, Turpin-
Petrosino, and Guckenburg (2010). But see also Lotta Vikström’s (2008) study on the
limited negative effects of incarceration on young offenders in nineteenth-century
northern Sweden.
 Punishment and Crime in Scandinavia 71

FIG. 23.—Police-reported robberies in Denmark, Finland, Norway, and Sweden, 1950–56

to 2008, per 100,000 population. Source: Falck, von Hofer, and Storgaard (2003), table 4;
updated.

at the end of the twentieth century. This stagnation came after a delay
of about 15 years for the group with large numbers of repeat offenses.30
Denmark, Finland, and Norway. Although historical data on repeat
theft are not easily available in the other Scandinavian countries, trends
in robbery offenses are available and are characterized by a massive
increase after World War II (see fig. 23).31 “At the end of the 1950s
robbery was more or less unheard of in these countries, with a total of
only 1,200 robberies being registered in the four Scandinavian coun-
tries. As in Sweden, in part, the increase is probably linked to the
emergence of a group of socially marginalized older males and in part,

30
The convictions relating to youths (15–17 years) have been shifted 15 years in
fig. 22 to improve the clarity of the comparison. The 1921–47 series relates to court
convictions. Data are missing for the years 1948–54, 1957, and 1972. The 1955–71
series relates to court convictions and certain types of waiver of prosecution issued in
connection with serious offenses. This series was discontinued in 1971. The series for
1973–90 relates to all types of findings of guilt (court convictions, summary fines, and
waivers of prosecution). Missing data have been interpolated. Note that the same
individuals may be counted more than once (particularly in the 6# group). To keep
the comparison simple, no rates per 100,000 of the population have been calculated.
31
Compare Eisner (2008, p. 306) for similar patterns in England and Wales, Italy,
and Switzerland.
72 Hanns von Hofer

FIG. 24.—Sweden: All offenses. Number of convicted persons, 1831–57 to 2007 (5-year
average), per 100,000 population. Source: von Hofer (2008), table 6.3; updated.

more recently, to robberies between young males” (Falck, von Hofer,

and Storgaard 2003, p. 8). However, in Finland reported robberies have
been markedly stable since the early 1970s.

III. Convictions and Sanctions

The most striking features of the Swedish trend relating to all criminal
convictions (see fig. 24) are: the sharp increase in the 1950s, the equally
sudden break in this trend immediately thereafter, and the decrease
since the end of the 1970s. The figure presents a somewhat different
picture of the trend from the preceding discussions relating to specific
offense types. The reason is that all offense types—from murder to
traffic violations—are given equal weight. Most of the increase during
the 1950s results from the rising number of traffic violations—mirror-
ing the expansion in car ownership in Sweden—and a doubling of ar-
rests for public drunkenness following the repeal of alcohol rationing
in 1955. The increases before World War II also reflect control of
public drunkenness and road traffic offenses.
The leveling off during the 1960s and the decrease since the mid-
1970s largely result from the decriminalization of minor traffic viola-
 Punishment and Crime in Scandinavia 73

tions and public drunkenness. Figure 24 is instructive in that it clearly

shows that criminality is not simply “given” but results from political
decisions. A decision made in parliament may lead to a sudden increase
or decline in the level of recorded criminality. Even if traditional in-
terests and values almost universally imply that certain acts should re-
sult in punishment (cf. van Dijk 1999, p. 42), there is a wide margin
for discretion in relation to what else is to be controlled by means of
the penal system (through punishment and the threat of punishment)
or by other means.
Figure 24 also indicates that a significant part of the population re-
ceives a penal sanction each year. Over time the cumulative effect
means that almost 40 percent of men and 10 percent of women in a
given birth cohort can expect to receive at least one penal sanction
between the ages of 15 and 37 years (minor traffic offenses not in-
cluded).32 Lower but still substantial rates are also reported from Den-
mark (Kyvsgaard 2003) and Norway (Skardhamar 2006).

A. Convictions for Serious Offenses

When one examines “criminality” as a whole, there is a considerable
risk that the picture will primarily reflect trends in minor offenses (be-
tween 85 and 95 percent of all convictions during the period under
study resulted in a fine). When seeking an alternative description of
the “crime” trend in Sweden, one might instead use the number of
convictions for so-called serious offenses as the unit of measurement:
that is, crimes leading to a more severe sanction than a fine.
Figure 25 (which is dominated by the development of theft crimes)
shows once again that industrialization and urbanization did not neg-
atively influence overall conviction trends in Sweden in any appreciable
degree before World War II. The same picture emerges from the
trends in police-registered criminality in Stockholm (Grabosky, Pers-
son, and Sperlings 1977) or Gothenburg. The big drop in the 1850s
occurred long before the start of industrialization in Sweden, which is
usually dated around the 1870s.
Because it is always risky to analyze criminality as a monolith, I

32
This statement refers to a Swedish cohort born in 1960 and followed up between
1975 and 1997 (Brå 2000; Svensson 2002). Four out of five offenses were, in rank
order, theft, major traffic offenses (e.g., drunken driving), drug offenses, criminal as-
sault, disturbing the public peace, fraud, and violence/threat to public servant.
74 Hanns von Hofer

FIG. 25.—Sweden: All serious offenses. Number of convicted persons, 1842–2007, per
100,000 population. Data from 1995 are estimated. Source: von Hofer (2008), table 5.4; updated.

analyzed trends for 25 separate types of crime. The findings can be

summarized as follows.
Even when all the crimes resulting in convictions are divided into
subgroups and examined separately, the evidence still indicates that
urbanization and industrialization did not as a rule result in any sub-
stantial increase in the level of convictions before World War II. Al-
most all crime types were at lower levels during the interwar period
than before the First World War.
Three types of crime have steadily decreased since the 1870s: crimes
against religion, defamation, and crimes against the family. At the same
time, three other groups of offenses (“economic crimes”) have in-
creased steadily: fraud, counterfeiting, and crimes associated with
debts.
The use of custodial sanctions in association with non–Penal Code
offenses was very rare in the 1800s. In 1885, for example, only 13 such
prison sentences were recorded, whereas the corresponding number in
2005 was approximately 7,300 (47 percent of all prison sentences).
These latter sentences were mainly for drug-related crimes (drunken
driving, sale of alcohol and other drugs), tax crimes, and illegal pos-
session of weapons.
 Punishment and Crime in Scandinavia 75

FIG. 26.—Sweden: All serious offenses. Number of convicted persons, 1857–2007, by gender
(5-year averages), per 100,000 population; semilog scale. Data from 1995 are estimated. Source:
von Hofer (2008), table 5.5; updated.

With a few exceptions, levels of convictions for all groups of offenses

increased after World War II, the watershed for conviction trends in
Sweden. In a matter of just two decades, Sweden was transformed from
a country with low conviction rates in Scandinavia to one with high
rates. A stable development evolved from the mid-1970s onward.

B. Specific Groups Convicted of Serious Offenses

The statistics on serious offenses also allow us to summarize some
central tendencies associated with the trends in crime that result in
convictions among specific groups.
Women and Men. During the period under study, convictions of
women and men have largely followed the same pattern, but the total
conviction rate for men is consistently much higher than for women
(see fig. 26). The proportion of women among convicted persons has
varied between 18 percent (1866–70) and 6 percent (1956–60), and
stood at 13 percent in recent years.
A similar development—with declining female conviction rates—has
been described for Norway (see fig. 27; Christie 1975, p. 288) and for
76 Hanns von Hofer

FIG. 27.—Norway: Percentage of female offenders convicted for serious offenses, 1860–2007.
Source: Statistisk sentralbyrå (1995), table 8.7; updated.

Denmark (Hurwitz and Christiansen 1983, pp. 35–36) where the per-
centage of female offenders declined from 25 percent in 1866–70 to 7
percent by 1972.33 This trend was upset for a brief period in World
War II when the number of female convictions rose at a higher rate
than the number of male convictions. The development relating to
female prison inmates in Finland is likewise characterized by a histor-
ical decline. The lowest level was recorded during the 1970s (Aho and
Karsikas 1980, fig. 10).
Young and Old. From a historical perspective, the focal point of the
Swedish conviction rate has shifted downward through the age range.
In the middle of the nineteenth century, 15–17-year-olds had the low-
est conviction rate of all age groups in Sweden. In the 2000s they have
the highest. This top ranking had been achieved by the 1930s—which
may partly indicate a fundamental shift in the way social control is
exercised over juveniles: from (relatively effective?) informal to (rela-
tively ineffective?) formal control.

33
For similar results as regards Toronto, 1859–1955, see Boritch and Hagan (1990).
For a general overview on gender and crime, see Fagan (2002).
 Punishment and Crime in Scandinavia 77

FIG. 28.—Norway: Young offenders (under 21 years) convicted for serious offenses, 1860–
2007, per 1,000 of the age group. The years 1923–29 and 1993–98 are missing. Source: Christie
(1975, pp. 278–81); updated.

The development is similar in Norway (see fig. 28), although the

increase is delayed by about 20 years (also reflected in the general crime
trend in Norway, where the increase begins later than it does in Swe-
den). There also seems to be a similar trend in Denmark—at least in
relation to convictions for theft (Hurwitz and Christiansen 1983, p.
278).
Foreign Citizens. The presence of foreign citizens among persons
convicted of serious offenses is for the most part a postwar phenom-
enon (see fig. 29).34 Sweden and the other Scandinavian countries were
all characterized by net emigration during the latter half of the nine-
teenth and the beginning of the twentieth centuries.35 Before World
War II, the proportion of foreign citizens with convictions stood at
under 2 percent in Sweden. In the mid-1990s it was around 17 percent,

34
See also von Hofer, Sarnecki, and Tham (1997). For international research round-
ups, see Tonry (1997) and Grafl (2009).
35
Finland experienced its most recent wave of emigration as late as the 1960s. This
emigration was to Sweden.
78 Hanns von Hofer

FIG. 29.—Sweden: All serious offenses. Number of convicted persons, 1842–2005, by citi-
zenship (from 1956), per 100,000 population. Data from 1995 are estimated. Sources: von Hofer
(2008), table 5.4; updated; von Hofer (1985), table 5.9; updated.

of whom 40 percent came from neighboring countries: Denmark, Nor-

way, Finland, Russia, Poland, the Baltic states, and Germany.
In the mid-1950s, 5 percent of all persons convicted for so-called
serious offenses were foreign citizens. By 1994, this had increased to
17 percent. Roughly half of this increase is a function of the demo-
graphic growth of the foreign (mostly immigrant) population in Swe-
den. The other half is due to the overrepresentation36 of the foreign
population in the criminal justice system.
The level of overrepresentation has remained more or less stable
throughout the period. This stability has not been affected by the de-
crease in the proportion of citizens from other Scandinavian and Eu-
ropean countries in the total non-Swedish population during the pe-
riod. Stable proportions, albeit high, are also found for the most serious
crimes such as homicide, rape, and robbery since the 1980s.
Denmark and Norway report higher convictions rates among im-
migrants with regard to violence and property offenses (Kardell and

36
There is no consensus among Swedish researchers as to whether the overrepre-
sentation is behavioral or a result of discrimination.
 Punishment and Crime in Scandinavia 79

Carlsson 2009). The proportion of foreign inmates is today substantial

in Scandinavian prisons, with the exception of Finland. According to
data from the Council of Europe (Aebi and Delgrande 2010, table 3)
this proportion amounts to 25 percent in Norway, 23 percent in Den-
mark, 22 percent in Sweden, and 9 percent in Finland.37
Prior Criminal Records. In addition to the statistics on persons con-
victed for repeated theft offenses, somewhat sketchy data on prior
criminal records (referring to the number of prior convictions accu-
mulated by an offender) also exist in Sweden.
Until the outbreak of World War I, the number of first-time of-
fenders increased, while the group with a prior criminal record re-
mained at a constant and low level after an initial decrease. After World
War I, both groups of offenders became larger, but the group of re-
cidivists increased most (particularly after World War II). There has
been a pronounced leveling off in the rate of increase since the 1970s,
first among the debutants and later among the recidivists. This leveling
off also holds true for the group with the lengthiest criminal records
(see fig. 30).
Available recidivism statistics for Sweden (1917–94) and Norway
(1957–89) give no indication that the penal system has produced any
discernible reduction in the recorded risk of recidivism on the aggre-
gate level.38 In Norway, the statistics show clear increases. In Sweden
there appears to have been relative stability, but with a marked increase
during the 1940s and the beginning of the 1950s (see fig. 31).39

C. Sanctions
Imprisonment was formally introduced in Sweden in 1855 as an al-
ternative to corporal punishment, which was abolished that year as a
criminal sanction for adults. The first alternatives to imprisonment

37
Among other things as a result of Finland’s special geopolitical position, lying
close to the former Soviet Union, the number of foreign citizens living in Finland
remained very low until the beginning of the 1990s. At the end of 2008, foreign citizens
accounted for 2.7 percent of the Finnish population. For a general account of foreign
citizens in European prisons, see van Kalmthout, Hofstee-van der Meulen, and Dünkel
(2007).
38
This seems to be in accordance with the body of current (cross-sectional, English-
speaking) research on the effects of sanctions on recidivism (Lipsey and Cullen 2007).
However, the notion of the “interchangeability of sanctions” as regards recidivism is
still contested (cf. Meier [2007, p. 987] for a German-speaking account).
39
For data, covering the period 1991–2003, see Brå (2009), table 6.2. No decreases
are discernible holding gender and the number of previous convictions constant.
 FIG. 30.—Sweden: All serious offenses. Number of convicted persons, 1936–2005, persons
with six or more prior convictions. Data are estimated for 1940–54, 1961–65, and 1995–2005.
Sources: Statistiska centralbyrån (1958), table 5; Statistiska centralbyrån (1967), table 5; Brå
(1997), table 3.15.

FIG. 31.—Sweden: Calculated risks of recidivism, 1917–94. 0#p zero prior convictions;
1#p one prior conviction. Separate data on two or more prior convictions are not available.
Source: Compiled by the author from yearly publications; cf. Tham (1979), table 27 (1917–
77).
 Punishment and Crime in Scandinavia 81

FIG. 32.—Sweden: Convictions for serious offenses, 1842–2007, by sanction, per 100,000
population. The total from 1995 is estimated. Source: von Hofer (2008), table 6.3; updated.

came in their turn 50 years later, with the introduction of the sus-
pended sentence in 1907 and probationary sentences in 1919. Norway
introduced the suspended sentence in 1894 (following the Belgian ex-
ample) and Denmark in 1905. As can be seen from figure 32, much of
the increase following World War II is related to noncustodial sen-
tences (in the form of suspended sentences, probation, waivers of pros-
ecution, special treatment for young offenders, and, since the late
1980s, civil commitment and community service),40 which in turn is
strongly connected to the deflated penal value of theft offenses.
From a historical perspective, the Swedish and other Scandinavian
prison systems, in line with European counterparts, are young insti-
tutions. The prison’s rise to prominence can be dated to the first half
of the nineteenth century (von Hofer 1975; Snare 1992; R. Nilsson
2003). At that time, imprisonment was introduced as a substitute for
the death penalty and corporal punishment. With their roots in the
early modern workhouses, prisons originally functioned as a congre-
gation point for the poor, the jobless, and the marginalized. This func-

40
Since 1994, short prison sentences can be served at home by means of electronic
monitoring, which is used today in all Scandinavian countries.
82 Hanns von Hofer

tion is still very much alive. According to recent level-of-living surveys

among prison inmates, inmates still compare very unfavorably with the
remainder of the population as regards their living conditions before
incarceration (A. Nilsson [2003] for Sweden; Skardhamar [2003] for
Norway).
From a structural perspective, the development of police-recorded
criminality and the use of imprisonment (measured in terms of the
daily prison population) appear to constitute two independent pro-
cesses in Sweden. Even if the data shown in figure 33a are partially
based on estimates, it is clear that the size of the Swedish prison pop-
ulation has not been determined by the level of recorded crime (see
fig. 33b). Nor does the widely discussed idea of the “stability of pun-
ishment” apply to the Swedish data.41
Whether the fluctuating use of imprisonment has influenced the
course of crime trends is a more difficult question to answer, but figure
33b suggests that there has been no effect. Clearly, before World War
II there is no relationship at all. The trend in the inmate population
was on the decline, while offense rates remained more or less stable.
After World War II, the picture altered quite drastically, with increas-
ing prison trends (until the mid-1960s) and soaring offense rates (until
the late 1980s). In the case of theft, which determines the shape of the
offense curve, a comprehensive analysis of the data over the course of
the twentieth century has shown that there is no support to the deter-
rence/incapacitation hypothesis in a longitudinal perspective (von
Hofer and Tham 1989; see also figs. 17a and 17b, above).
In Norway too the crime trends and the trends describing inmate
numbers diverge markedly at both the beginning and the end of the
observation period (see figs. 34a and 34b). The period 1880–1965 is
characterized by a parallel and substantial level of stability, with the
exception of the end of the World War I. A similar stability also ap-
pears to characterize Denmark during the period from about 1870 to
1965 (see figs. 35a and 35b).
In summary, the marked decline in the prison populations in Den-
mark, Norway, and Sweden during the latter half of the nineteenth

41
For a critical summary of the discussion, see Zimring and Hawkins (1991, pp. 14–
37) and Killias (2002, p. 381). To bolster their case, Blumstein and Cohen (1973, pp.
203–4) and Blumstein (2007, p. 3) also used Norwegian prison data starting in 1880,
although complete and contradictory data from 1814 on were available, as well as data
from the other Scandinavian countries (Christie 1968, fig. 3).
 FIG. 33.—Sweden: a, Estimated crime curve, 1841–2008, and daily prison population, 1861–
2008, per 100,000 population; index 1950 p 1. b, Estimated crime curve, 1841–2008, and daily
prison population, 1861–2008, per 100,000 population; index 1950 p 1; shown as a Phillips-
type curve. Source: von Hofer (2008), tables 3.1, 5.3, and 7.1; updated.
 FIG. 34.—Norway: a, Estimated crime curve, 1835–2008, and daily prison population, 1810–
2008, per 100,000 population; index 1957 p 1. b, Estimated crime curve, 1835–2008, and daily
prison population, 1810–2008, per 100,000 population; index 1957 p 1; shown as a Phillips-
type curve. Source: Christie (1975), tables 3.3.-1A and 6.2-1; updated.

84
 FIG. 35.—a, Denmark: Trend in prison inmate numbers, 1867–1972, per 100,000 population.
Source: Adapted from Greve (1996, p. 35). b, Denmark: Number of convicted male offenders,
1867–1972, per 100,000 of the male population above age of criminal responsibility. Source:
Christie (1975), table 3.3.-2.

85
86 Hanns von Hofer

century was not connected with concurrent increases in conviction

rates. A similar picture emerged in Finland during the period 1927–
66, for which data on police recorded Criminal Code offenses exist
(see figs. 36a and 36b).
The Finnish prison population was, as early as the end of the nine-
teenth century, larger than those of its Scandinavian neighbors (Chris-
tie 1968). World War I, the civil war, and the subsequent period of
repression as well as World War II, all made a major impression on
the Finnish prison statistics (cf. Lenke 1980; Kekkonen 1999). The
remarkable reduction in the Finnish prison population by almost three-
quarters during the postwar period (see fig. 37) has been described by
Patrik Törnudd (1993) and Tapio Lappi-Seppälä (2007).
In sum, all Scandinavian countries reduced their prison populations
in the long-term historical perspective, and no Scandinavian country
matched the increase of reported criminality after World War II with
corresponding increases of their prison populations. The underlying
reasons for this development are widely discussed among scholars (see,
e.g., Christie 1968, 2000; von Hofer 2003b; Sutton 2004; Bondeson
2005; Cavadino and Dignan 2006, p. 149; Lappi-Seppälä 2007; Pratt
2008a, 2008b).

IV. Major Findings

The preceding survey and the data on which it is based support a
number of generalizations. They are drawn, of course, from develop-
ments in the Scandinavian countries but possibly mirror developments
in many Western countries.
1. The stable or declining conviction trends during the second half
of the nineteenth century and the first half of the twentieth century
are striking, particularly since they occurred despite profound changes
in the social and political structure of the Scandinavian countries.
2. The proportion of women punished declined at least until the
1950s. Thereafter it began to increase. While women are less often
convicted than men, the long-term conviction trends are parallel. The
modern conviction rate for women in Sweden equals the male rate of
the early 1920s. Male crime then was clearly defined as a social prob-
lem, whereas female crime today generally is not.
3. Young people (predominantly young men) emerged increasingly
 FIG. 36.—a, Finland: Daily prison population, 1881–2008, per 100,000 population. The years
1947–49 are missing. b, Finland: Police-recorded Penal Code offenses, 1927–2008, per 100,000
population. Sources: Aho and Karsikas (1980), tables 1 and 2; Falck, von Hofer, and Storgaard
(2003), tables 8 and 14; updated.
88 Hanns von Hofer

FIG. 37.—Prison populations in Scandinavia, 1950–2008, per 100,000 population. Source:

Falck, von Hofer, and Storgaard (2003), table 14; updated.

clearly as a central focus of criminal law controls during the twentieth

century.
4. The presence of foreign citizens among convicted persons is, for
the most part, a postwar phenomenon. In all Scandinavian countries,
the share of convicted foreign citizens is higher than their comparable
shares in the general population. The reasons for the overrepresenta-
tion are contested.
5. Long-term data on repeat offenders are scarce in Scandinavia.
Swedish data suggest, however, that the number of repeat offenders
soared during a 30-year period following World War II—possibly trig-
gered by an increased input of young offenders and exacerbated by
accelerated substance use.
6. The significance of alcohol and drugs—both use and control—
for the development of punishment and crime in Scandinavia cannot
be overestimated.
7. In the long run, the penal value of violent offenses has been up-
graded, while theft offenses have been downgraded from serious crimes
to mass offending (“volume crime”). The proportion of prison inmates
in 2008 who were convicted of theft offenses as the principal offense
(excluding robbery) amounted to 7 percent in Sweden, 9 percent in
 Punishment and Crime in Scandinavia 89

FIG. 38.—Persons convicted per 100,000 population in Finland, Denmark, Sweden, and 31
other European countries, average 2000–2003. Norway is missing. Source: Aebi et al. (2006,
p. 99).

Norway, 13 percent in Denmark, and 16 percent in Finland (Aebi and

Delgrande 2010, table 7).
8. The declines in the prison population that occurred in Denmark,
Norway, and Sweden took place for the main part during the second
half of the nineteenth century and thus long before the welfare state
became established in the Nordic countries—a fact that is not properly
recognized in research to date. Finland deviates from the other coun-
tries until the end of the 1980s. The reduction in the number of prison
inmates in Finland subsequent to World War II is likewise remarkable.
9. The periods surrounding wars have affected trends in both the
short- and long-term perspective.
10. With some degree of time lag between the countries, the period
after World War II brought with it a steep increase in the level of
criminal justice interventions across Scandinavia, in part a result of
increases in offending, in part a result of the expansion of penal law
control (regarding, e.g., motoring, violence, sexuality, drugs).
11. No Scandinavian country appears to have been able to control
rising crime trends after World War II, though the Scandinavian crim-
inal justice systems have been “good” at punishing people (see fig. 38).
90 Hanns von Hofer

The number of individuals punished is high by European standards

(Aebi et al. 2006, p. 99). Both Denmark and Norway make large use
of very short prison sentences.
12. Notwithstanding the introduction of a wide range of alternative
sanctions to imprisonment and fines during the twentieth century,42
recidivism statistics (in Norway and Sweden) do not indicate that the
criminal justice interventions have produced any major decline in the
risk of recidivism. In Norway, clear increases are reported, in Sweden
relative stability, although with a marked increase during the 1940s and
the beginning of the 1950s.
In sum, there is ample evidence that traditional crime policy mea-
sures—the reliance on police and punishment—have not produced the
desired effects of protecting ordinary citizens and reforming criminals.
Punishing individuals who commit crimes seems to be of little marginal
value as a means of preventing crime.

V. Crime Policy Implications

Advocating the use of punishment and the threat of punishment as
means of crime control is commonsensical and easily understood. It is
a “good story,” as Cullen, Wright, and Chamlin (1999, p. 196) have
pointed out, since crime and punishment are highly emotive issues.
But referring to punishment as a response to crime too often prevents
any search for really effective alternative measures. Such effective al-
ternative measures do exist, however. This can be illustrated with six
examples concerning Sweden, all involving serious matters of life and
death. The examples chosen relate to long-term trends in infant mor-
tality, suicide, fatal accidents at work and on the roads, homicide, and
mortality among drug addicts. In the first four cases, penal law either
no longer has any role to play at all (infant mortality and suicide) or
it plays only a secondary role (fatal accidents at work or on the roads),
whereas penal law is still ascribed major importance in relation to the
control of homicide and in the area of drugs.
If one considers trends across all six areas together (see figs. 39a–f),
it becomes very clear how little effect is produced by general measures
based primarily on punishment and the threat of punishment (i.e., in
the areas of homicide and drugs). The effects of specific nonpenal law

42
Swedish criminal justice statistics, e.g., count at least 14 different sanction cate-
gories at present (Brå 2009, table 4.8).
 FIG. 39.—a, Sweden: Rate of infant deaths within the first year of life, 1950–2008, per 1,000
live births. Source: Statistiska centralbyrån (2011), table 4.30, and previous editions. b, Sweden:
Rate of suicides, 1950–2008, per 100,000 population. Source: Statistiska centralbyrån (2011),
table 20.15, and previous editions. c, Sweden: Rate of persons killed in road traffic accidents,
1950–2008, per 100,000 population. Source: Statistiska centralbyrån (2011), table 10.20, and
previous editions.
 FIG. 39 (continued).—d, Sweden: Rate of persons killed in workplace accidents, 1955–2008,
per 100,000 population. Source: Statistiska centralbyrån (2011), table 12.17, and previous edi-
tions. e, Sweden: Rate of homicides, 1950–2008, per 100,000 population. Source: Statistiska
centralbyrån (2011), tables 20.16–17, and previous editions. f, Sweden: Rate of drug-related
deaths, 1969–2008, per 100,000 population. Drug use is the primary or underlying cause of
death. Source: CAN (2010), table 89.
 Punishment and Crime in Scandinavia 93

measures within the childcare sector, at work places, on the roads, and
in the prevention of suicide, are by contrast impressive.
One can also observe that the use of punishment is being withdrawn
from many other areas of modern (Swedish) society.43 For example, the
labor market has long been completely free of the use of punishment.
Military penal justice, which in earlier periods of (Swedish) history
constituted the motor that drove developments in the field of penal
law more generally, was abolished in 1986 in Sweden. The field of
education has also long been free of the use of punishment and parental
punishment of children has been banned. Thus the incongruous ad-
herence to the use of punishment in the field of criminal policy is
striking.
It is therefore reasonable to conclude that the use of penal law in
practice has little to do with the conduct of an efficient crime policy
whose principal objective is to prevent crime. The political attraction
of penal law appears rather to lie in its symbolic significance and the
visible exercise of power which reinforces widely shared attachments
to punishment as a proper response to wrongdoing (cf. Roberts and
Hough 2002, pp. 5–10).
Power is usually exercised by someone, against someone, and for
someone or something. The actor’s perspective in itself is relatively
unproblematic in the Scandinavian countries. They are constitutional
states in which the government and parliament determine the general
focus of the police, the prosecution service, the courts, and the prison
bureaucracies by means of legislation and financial resources. The ap-
plication of this legislation is conducted for the most part (although
not always) within the prescribed limits. A system of judicial and po-
litical controls exists. However, the questions for and against whom
penal justice power is exercised are more difficult to answer.
In theory, criminal law applies to all. In practice, criminal law is
directed primarily against men (as unorganized individuals).44 To an
increasing extent, it has also become young men—particularly after

43
Compare Haferkamp (1984) and also Braithwaite’s (2003) discussion of alternative
strategies of regulation.
44
Compare Martin J. Wiener (1998, p. 209, with further references): “Th[e] ‘mas-
culinization’ of crime and punishment was one of the most notable, but least noticed,
facts of nineteenth-century (and, indeed, western) criminal justice history.” To belong
to an organization—legal or illegal—appears by contrast to produce good opportunities
for protection against penal law control (cf. Wheeler and Rothman 1982, p. 1422).
94 Hanns von Hofer

World War II.45 If one looks at prison populations, the focus is still on
socially disadvantaged men, both those with domestic and overseas or-
igins. It was previously a question of poor, unemployed, alcohol de-
pendent, vagrant property offenders (cf. Emsley 2007, p. 13). Today it
is poor, unemployed, alcohol and drug dependent, homeless, violent,
sex, and drug offenders (cf. Wacquant 2009).The penal law is thus still
focused on the control of segments of the marginal population. The
criteria for intervention have changed, however from stealing (and va-
grancy) to violence, sex, and substance use. At the same time, penal
justice control, which is exercised first and foremost through the use
of fines, has undergone a very substantial increase after World War II.
This development is of course due, to some extent, to the development
in motoring, but conviction rates remain high within the Scandinavian
countries even when motoring offenses are excluded (Aebi et al. 2006,
pp. 99–100).
The penal justice system is thus being allowed to play an expanding
role in relation to the control of the (male) population at large, while
the police apparatus itself is small in Scandinavia by comparison with
the rest of Europe (Aebi et al. 2006, p. 74).
The groups intended to benefit from the penal law as it is applied
are to a growing extent women and children (as victims of violence
and sex offenses) at the same time as the penal law is used to satisfy
demands from inter- and transnational organizations (IGOs, such as
various bodies of the United Nations and the European Union), na-
tional pressure groups (NGOs), and parts of the media.46 The latter
groups’ significance is particularly prominent in the area of drug policy.
Thirty percent of prison inmates in Norway and Sweden today have
been convicted of drug offenses, and 24 percent in Denmark (Aebi and
Delgrande 2010, table 7). Only Finland stands out, with 16 percent,
which is mainly47 explained by historical geopolitical factors.
Drugs probably constitute the largest single crime policy problem
in the Scandinavian countries today (Träskman 2004; Kinnunen 2008;
Stene 2008). That the drugs phenomenon has been defined primarily

45
But see also John Muncie’s illuminating “Histories of Youth and Crime” (2009,
chap. 2).
46
Unfortunately, the decision-making process in criminal policy is an underre-
searched area in Scandinavia.
47
See also Törrönen (2004) as to the rather pragmatic position of Finnish newspapers
on drug policy questions. On Finnish crime media research in general, see Smolej and
Kivivuori (2008).
 Punishment and Crime in Scandinavia 95

in moral and penal law terms (and not in terms of realpolitik as harm
reduction) constitutes the crux of the matter. The Scandinavian coun-
tries (with the exception of Denmark) have been countries with pow-
erful moralistic temperance movements, whose ideology lay very close
to the values of the (U.S.) American drug policy that came to dominate
the international scene (Bruun, Pan, and Rexed 1975, p. 132; Kinnunen
2008, p. 136).48 A drug policy that instead defines drug dependency as
a public health problem would naturally reduce the level of penal law
repression substantially. But the “War on Drugs” or, rather, the “War
on Drug Users” ( Jepsen 1996) is likely to continue for some time,
because it is difficult49 for politicians and pressure groups to concede
that what they have established as “right” in fact can be “wrong.”
In 1956, the influential Swedish Strafflagberedningen (Criminal Law
Committee) observed that “it must be emphasized that the possibilities
of penal policy to prevent crime in many ways are limited and that the
presence of crimes does not depend only on, or even substantially on,
shortcomings in the penal legislation. It is a misconception, when many
. . . seem to think that by reforming the sanction system one might
stop an increase in the number of offenses. This goes against all ex-
perience” (Strafflagberedningen 1956, p. 28; English translation by
Janson 2004, p. 436). Thirty years later, the Finnish Criminal Law
Committee in 1986 wrote in the same spirit that the criminal justice
system is not society’s only “or even principal” instrument for con-
trolling behavior (Törnudd 1996, p. 176). Nevertheless, the penal
codes have prevailed as a pivotal means of Scandinavian criminal pol-
icy.50
The main lesson that can be drawn from the preceding historical
analyses of criminal justice statistics is that the criminal justice systems
in Scandinavia generally have not offered efficient ways of controlling
the development of crime in the twentieth century: many types of tra-

48
Compare Tonry’s discussion of a conceivable link between “protestant funda-
mentalism and intolerance” and temperance movements and moralistic crusades against
drugs and crime in the United States (Tonry 2009, pp. 382–84). In the current internal
Swedish debate, the fight against drugs is used as an argument to justify the presence
of Swedish troops in Afghanistan (Tolgfors 2009; Bildt 2010; Tolgfors is the Swedish
Secretary of Defense and Bildt is the Swedish Secretary of State).
49
But not impossible; see, e.g., the statement by the Latin American Commission
on Drugs and Democracy (2009) and now also the Norwegian commission report on
drugs (Stoltenbergutvalget 2010).
50
Compare Bohm (1986, p. 199) for possible reasons from a leftist perspective.
96 Hanns von Hofer

ditional crime soared and modern crimes sprawled despite the criminal
justice systems’ significant growth in size and scope.51 The salient point
for this failure seems to be that “doing justice” and “being efficient”
became muddled. The labeling of unwanted behavior as “crime” in-
evitably triggers a series of processes that aim to “do justice,” irre-
spective of whether they are efficient or not. This happens because
crime is defined through punishment,52 but punishment does not by
default prevent crime—neither in its deterrent nor in its norm-
strengthening guise. Thus a politician who wants to reduce unwanted
behavior or events should look in the first place for means of preven-
tion other than those offered by the criminal justice system. Punish-
ment is the deliberate infliction of harm. Harming the offender is in-
tended, but punishment also harms citizens and society at large—given
all crime that the reliance on the ritual use of punishment did not
prevent.

APPENDIX
The Phillips Curve
The Phillips-type curve is a graphical technique to add a third, temporal, di-
mension in a two-dimensional space, that is, the dependent variable (Y ), the
independent variable (X) and time (T), by connecting the observations (x, y) in
consecutive order (t). This is a handy way to screen for contingencies over
time, when theory predicts the direction of the relationship (positive or neg-
ative). The technique was used by William Phillips (1958) in “The Relationship
between Unemployment and the Rate of Change of Money Wages in the
United Kingdom 1861–1957.” Phillips described an inverse relationship be-
tween money wage changes and unemployment in the British economy over
the period examined.
Needless to say, this graphical technique is a screening tool only. It visualizes
the temporal covariance between X and Y, but it cannot handle the importance
of other variables which may also independently account for the covariance of
X and Y.
In the present essay the Phillips-type curve is used to study the relationship
between indicators of punishment and crime. According to deterrence or in-
capacitation theories, there should be a negative relationship between the use
of punishment and crime: more punishment yields less crime and less punish-
ment more crime.
Looking at figure A1, one may perhaps discern an overall negative relation-

51
Unfortunately, systematic scholarly analyses of the growing powers of criminal
justice institutions are lacking in Scandinavia.
52
See, e.g., the Swedish Penal Code (chap. 1, sec. 1): “Crime is an act for which
punishment as stated below is provided by this Code or by other codes or laws.”
 Punishment and Crime in Scandinavia 97

FIG. A1.—Sweden: Estimated crime curve, 1841–2008, and daily prison population, 1861–
2008, per 100,000 population; index 1950 p 1. Source: von Hofer (2008), tables 3.1, 5.3, and
7.1; updated.

ship between punishment (number of prisoners) and crime (number of con-

victions): when the number of prisoners is high, crime convictions are low and
vice versa. However, looking at figure A2, the Phillips-type curve, and follow-
ing the observations in a temporal order from 1861 to 2008, one clearly sees
that there is no negative relationship whatsoever. When nonzero relationships
are discernible, they are positive (see, e.g., 1861–69, World War I, 1935–63,
end of period). In conclusion: in a bivariate frame of reference, the number of
prisoners has not determined the number of crime convictions. Instead, the
number of convictions has determined the number of prisoners. This result is
expected in a criminal justice system—such as the Swedish system—that is
based on the rule of law.
98 Hanns von Hofer

FIG. A2.—Sweden: Estimated crime curve, 1841–2008, and daily prison population, 1861–
2008, per 100,000 population; index 1950 p 1; shown as Phillips-type curve. Source: von Hofer
(2008), tables 3.1, 5.3, and 7.1; updated.

REFERENCES

Aebi, M. F., K. Aromaa, B. Aubusson de Cavarlay, G. Barclay, B. Gruszczyñska,

H. von Hofer, V. Hysi, J.-M. Jehle, M. Killias, P. Smit, and C. Tavares.
2006. European Sourcebook of Crime and Criminal Justice Statistics—2006. 3rd
ed. Onderzoek en beleid 241. The Hague: Boom Juridische uitgevers/
WODC.
Aebi, M. F., and N. Delgrande. 2010. Council of Europe Annual Penal Statistics:
SPACE I. Survey 2008. PC-CP (2010) 07. Strasbourg: Council of Europe.
Aho, T., and V. Karsikas. 1980. Vankien taustaan ja vankilukuun liittyviä tilastoja
1881–1978. Summary: Statistical Data Relating to the Prisoners’ Background and
to the Prison Population from 1881 to 1978. Vankeinhoidon hisatoriaprojektin
julkaisu 3/1980. Helsinki: Oikeusministeriön vankeinhoito-osasto.
Andersson, H. 2008. “Shaming and Economic Crime in Nineteenth Century
Sweden.” In Den ljusskygga ekonomin: Organiserad och ekonomisk brottslighet,
edited by T. Alalehto and D. Larsson. Umeå, Sweden: Umeå Universitet.
Andersson, J., and M. Hilson. 2009. “Images of Sweden and the Nordic Coun-
tries.” Scandinavian Journal of History 34:219–28.
Anttila, I., and O. Heinonen. 1975. Brott och påföljd: Straffrättens grunder. Hel-
sinki: Helsingin yliopiston monistuspalvelu.
Ball, M., and A. Wood. 1996. “Trend Growth in Post-1850 British Economic
History: The Kalman Filter and Historical Judgment.” Statistician 45(2):143–
52.
 Punishment and Crime in Scandinavia 99

Balvig, F. 1987. Den tyvagtige dansker: Brudstykker af kriminalitetens udvikling

gennem tre århundreder, fra 1725 til 2025. Valby, Denmark: Borgens Forlag.
———. 1988. Studies on Theft: English Summary of a Dissertation about the Act
of Theft in Regard to Population, Time and Geographical Space. Kriminalistisk
Instituts Stencilserie no. 44. Copenhagen: Institute of Criminal Science,
University of Copenhagen.
Barak, G. 2001. “Crime and Crime Control in an Age of Globalization: A
Theoretical Dissection.” Critical Criminology 10:57–72.
Bergenlöv, E., and M. Lindstedt Cronberg. 2002. “Sexualmoral och brotts-
statistik.” In Offer för brott: Våldtäkt, incest och barnamord i Sveriges historia
från reformationen till nutid, edited by E. Bergenlöv, M. Lindstedt Cronberg,
and E. Österberg. Lund, Sweden: Nordic Academic Press.
Bildt, C. 2010. “Kamp mot knark och terrorism.” Svenska Dagbladet, July 20.
Blumstein, A. 2007. “The Roots of Punitiveness in a Democracy.” Journal of
Scandinavian Studies in Criminology and Crime Prevention 8(suppl. 1):2–16.
Blumstein, A., and J. Cohen. 1973. “A Theory of the Stability of Punishment.”
Journal of Criminal Law and Criminology 64:198–207.
Bohm, R. M. 1986. “Crime, Criminal, and Crime Control Policy Myths.”
Justice Quarterly 3(2):193–214.
Bondeson, U. V. 2005. “Global Trends in Corrections.” In Crime and Justice
in Scandinavia, edited by U. V. Bondeson. Copenhagen: Thomson.
Boritch, H., and J. Hagan. 1990. “A Century of Crime in Toronto: Gender,
Class, and Patterns of Social Control, 1859 to 1955.” Criminology 28:567–
99.
Brå. 1997. Kriminalstatistik 1995. Sveriges officiella statistik. Brå-rapport 1997:
5. Stockholm: Brottsförebyggande rådet.
———. 2000. Strategic Offences: English Summary and Conclusion. Brå-rapport
2000:3. Stockholm: Brottsförebyggande rådet.
———. 2009. Kriminalstatistik 2008. Sveriges officiella statistik. Brå-rapport
2009:17. Stockholm: Brottsförebyggande rådet.
Braithwaite, J. 2003. “What’s Wrong with the Sociology of Punishment?” The-
oretical Criminology 7:5–28.
Braudel, F. 1980. On History. Chicago: University of Chicago Press.
Bruun, K., L. Pan, and I. Rexed. 1975. The Gentlemen’s Club: International
Control of Drugs and Alcohol. Chicago: University of Chicago Press.
CAN. 2010. Drogutvecklingen i Sverige 2010. CAN Rapport 125. Stockholm:
Centralförbundet för alkohl- och narkotikaupplysning.
Cavadino, M., and J. Dignan. 2006. Penal Systems: A Comparative Approach.
Thousand Oaks, CA: Sage.
Christie, N. 1968. “Changes in Penal Values.” Scandinavian Studies in Crimi-
nology 2:161–72.
———. 1975. Hvor tett et samfunn? Tabeller m m. Copenhagen: Ejlers
———. 1976. “Criminological Data as Indicators on Contemporary Society.”
In Crime and Industrialization. First Seminar for Criminologists from So-
cialist and Scandinavian Countries in Helsinki, Finland, August 26–29, 1974.
Stockholm: Scandinavian Research Council for Criminology.
100 Hanns von Hofer

———. 1982. Hvor tett et samfunn? 2nd ed. Oslo: Universitetsförlaget.

———. 2000. Crime Control as Industry: Towards GULAGS, Western Style. 3rd
ed. New York: Routledge.
Cohen, L. E., and M. Felson. 1979. “Social Change and Crime Rate Trends:
A Routine Activity Approach.” American Sociological Review 44:588–608.
Cullen, F., J. P. Wright, and M. B. Chamlin. 1999. “Social Support and Social
Reform: A Progressive Crime Control Agenda.” Crime and Delinquency 45:
188–207.
Dahl, H. F. 1987. “Dødsstraffen i Norge.” In Krig og moral: Kriminalitet og
kontroll under andre verdenskrig, edited by H. Takala and H. Tham. Oslo:
Universitetsforlaget.
———. 1989. “Scandinavia in the Second World War.” Scandinavian Studies in
Criminology 10:17–26.
Edman, J., and K. Stenius. 2007. On the Margins: Nordic Alcohol and Drug
Treatment, 1885–2007, edited by J. Edman and K. Stenius. Helsinki: Nordic
Centre for Alcohol and Drug Research. http://www.nordicwelfare.org/?id
p119303.
Eisner, M. 2003. “Long-Term Historical Trends in Violent Crime.” In Crime
and Justice: A Review of Research, vol. 30, edited by M. Tonry. Chicago: Uni-
versity of Chicago Press.
———. 2008. “Modernity Strikes Back? A Historical Perspective on the Latest
Increase in Interpersonal Violence (1960–1990).” International Journal of
Conflict and Violence 2:288–316.
Ekström, S. 2003. “Qualification and Disqualification in Rape Cases.” Journal
of Scandinavian Studies in Criminology and Crime Prevention 4(2):204–22.
Emsley, C. 2002. “The History of Crime and Crime Control Institutions.” In
The Oxford Handbook of Criminology, edited by M. Maguire, R. Morgan, and
R. Reiner. 3rd ed. Oxford: Oxford University Press.
———. 2005. “Crime and Punishment: 10 Years of Research (1). Filling in,
Adding up, Moving on: Criminal Justice History in Contemporary Britain.”
Crime, Histoire et Sociétés/Crime, History and Societies 9(1):117–38.
———. 2007. Crime, Police and Penal Policy: European Experiences, 1750–1940.
Oxford: Oxford University Press.
Estrada, F. 2006. “Trends in Violence in Scandinavia according to Different
Indicators: An Exemplification of the Value of Swedish Hospital Data.” Brit-
ish Journal of Criminology 46(3):486–504.
Fagan, A. A. 2002. “Gender.” In Encyclopedia of Crime and Punishment, vol. 2,
edited by D. Levinson. Thousand Oaks, CA: Sage.
Falck, S., H. von Hofer, and A. Storgaard. 2003. Nordic Criminal Statistics,
1950–2000. Report 2003:3. Stockholm: Department of Criminology, Stock-
holm University. http://www.nsfk.org/upload/files/seminarreports/nordic
stats_1950-2000.pdf.
Felson, M. 1983. “Ecology of Crime.” In Encyclopedia of Crime and Punishment,
vol. 2, edited by S. H. Kadish. New York: Free Press.
Gatti, U., R. E. Tremblay, and F. Vitaro. 2009. “Iatrogenic Effect of Juvenile
Justice.” Journal of Child Psychology and Psychiatry 50:991–98.
 Punishment and Crime in Scandinavia 101

Grabosky, P. N., L. Persson, and S. Sperlings. 1977. “Stockholm: The Politics

of Crime and Conflict, 1750 to 1970s.” In The Politics of Crime and Conflict:
A Comparative History of Four Cities, edited by T. R. Gurr, P. N. Grabosky,
and R. C. Hula. Beverly Hills, CA: Sage.
Grafl, C. 2009. “Migration und Kriminalität.” In Internationales Handbuch der
Kriminologie, Band 2, Besondere Probleme der Kriminologie, herausgegeben von
H. J. Schneider. Berlin: De Gruyter Recht.
Granath, S. 2008. “Bättre akutsjukvård—Färre dödsoffer?” Apropå (1):12–13.
Greve, V. 1996. Straffene. Copenhagen: Jurist- og Økonomforbundets Forlag.
Gurr, T. 1981. “Historical Trends in Violent Crime: A Critical Review of the
Evidence.” In Crime and Justice: A Review of Research, vol. 3, edited by M.
Tonry and N. Morris. Chicago: University of Chicago Press.
Hacking, I. 1990. The Taming of Chance. Cambridge: Cambridge University
Press.
Haferkamp, H. 1984. “Herrschaftsverlust und Sanktionsverzicht: Kritische Be-
merkungen zur Theorie des starken Staates, der neuen sozialen Kontrolle
und des ideellen Abolitionismus.” Kriminologisches Journal 16:112–31.
Howard, G. J., G. Newman, and W. A. Pridemore. 2000. “Theory, Method,
and Data in Comparative Criminology.” In Measurement and Analysis of
Crime and Justice, Criminal Justice 2000, vol. 4, edited by D. Duffee. NCJ
182411. http://www.crim.umontreal.ca/cours/cri3013/comparative.pdf.
Hurwitz, S., and K. O. Christiansen. 1983. Criminology. 2nd rev. ed. London:
Allen & Unwin.
Janson, C.-G. 2004. “Youth Justice in Sweden.” In Youth Crime and Youth Jus-
tice: Comparative and Cross-National Perspectives, edited by M. Tonry and A.
N. Doob. Vol. 31 of Crime and Justice: A Review of Research, edited by M.
Tonry. Chicago: University of Chicago Press.
Jepsen, J. 1996. “Copenhagen: A War on Socially Marginal People.” In Eu-
ropean Drug Policies and Enforcement, edited by N. Dorn, J. Jepsen, and E.
Savona. Basingstoke, UK: Palgrave Macmillan.
Kardell, J., and K.-M. Carlsson. 2009. “Lagföringar av invandrare och invan-
drares barn i de nordiska länderna.” Nordisk Tidsskrift for Kriminalvidenskab
96:237–61.
Kekkonen, J. 1999. “Judicial Repression after the Civil Wars in Finland (1918)
and in Spain (1936–39).” In Crime and Control in Europe from the Past to the
Present, edited by M. Lappalainen and P. Hirvonen. Publications of the His-
tory of Criminality Research Project. Helsinki: Hakapaino.
Kick, E. L., and G. D. Lafree. 1985. “Development and the Social Context of
Murder and Theft.” Comparative Social Research 8:37–58.
Killias, M. 2002. Grundriss der Kriminologie: Eine europäische Perspektive. Bern:
Stämpfli Verlag.
———. 2006. “The Opening and Closing of Breaches: A Theory on Crime
Waves, Law Creation and Crime Prevention.” European Journal of Crimi-
nology 3:11–32.
Kinnunen, A. 2008. The Paradox of Criminal Policy: Studies of Drug Crime and
Its Control in Finland. Publication no. 233. Helsinki: National Research In-
102 Hanns von Hofer

stitute of Legal Policy. http://www.optula.om.fi/uploads/nnak8h7ky8ywr64

.pdf.
Kivivuori, J., and M. Lehti. In this volume. “Homicide in Finland and Swe-
den.”
Knafla, L. A. 1996. “Structure, Conjuncture, and Event in Historiography of
Modern Criminal Justice History.” In Crime History and Histories of Crime:
Studies in the Historiography of Crime and Criminal Justice in Modern History,
edited by C. Emsley and L. A. Knafla. Westport, CT: Greenwood.
Kujala, A. 2009. “Illegal Killing of Soviet Prisoners of War by Finns during
the Finno-Soviet Continuation War of 1941–44.” Slavonic and East European
Review 87:429–51.
Kyvsgaard, B. 2003. The Criminal Career: The Danish Longitudinal Study. Cam-
bridge: Cambridge University Press.
Landtman, G. 1932. “Dödsstraffet.” Finsk Tidskrift (1):363–72.
Lappi-Seppälä, T. 2007. “Penal Policy in Scandinavia.” In Crime, Punishment,
and Politics in Comparative Perspective, edited by M. Tonry. Vol. 36 of Crime
and Justice: A Review of Research, edited by M. Tonry. Chicago: University of
Chicago Press.
Latin American Commission. 2009. Drugs and Democracy: Toward a Paradigm
Shift. Statement by the Latin American Commission on Drugs and Democ-
racy. http://www.drogasedemocracia.org/Arquivos/declaracao_ingles_site.pdf.
Leigh, B. C., and R. Room. 2002. “Alcohol and Crime: Behavioral Aspects.”
In Encyclopedia of Crime and Justice, vol. 1, edited by J. Dressler. 2nd ed. New
York: Macmillan Reference.
Lenke, L. 1980. “Criminal Policy and Repression in Capitalist Societies: The
Scandinavian Case.” Scandinavian Studies in Criminology 7:5–30.
———. 1990. Alcohol and Criminal Violence: Time Series Analyses in a Compar-
ative Perspective. Stockholm: Almqvist & Wiksell.
———. 2009. Leif Lenke in memoriam: Valda skrifter, edited by H. von Hofer.
Rapport 2009:1. Stockholm: Kriminologiska institutionen, Stockholms Uni-
versitet.
Lindstedt, J. 2000. “Dödsstraffen i Finland under andra världskriget.” Nordisk
Tidsskrift for Kriminalvidenskab 87:302–9.
Lipsey, M. W., and F. T. Cullen. 2007. “The Effectiveness of Correctional
Rehabilitation: A Review of Systematic Reviews.” Annual Review of Law and
Social Science 3:297–320.
Liu, J. 2005. “Crime Patterns during the Market Transition in China.” British
Journal of Criminology 45(5):613–33.
Lovett, J., and L. Kelly. 2009. Different Systems, Similar Outcomes? Tracking
Attrition in Reported Rape Cases across Europe. London: Child and Women
Abuse Studies Unit, London Metropolitan University. http://www.cwasu
.org/filedown.asp?filepdifferent_systems_03_web(2).pdf.
McAra, L., and S. McVie. 2007. “Youth Justice? The Impact of System Contact
on Patterns of Desistance from Offending.” European Journal of Criminology
4:315–45.
 Punishment and Crime in Scandinavia 103

McDowall, D. 2002. “Test of Non-linear Dynamics in U.S. Homicide Time

Series, and Their Implications.” Criminology 40(3):711–36.
McDowall, D., and C. Loftin. 2005. “Are U.S. Crime Rate Trends Historically
Contingent?” Journal of Research in Crime and Delinquency 42:359–83.
Meier, B.-D. 2007. “Sanktionsforschung.” In Internationales Handbuch der Kri-
minologie, Band 1, Grundlagen der Kriminologie, edited by H. J. Schneider.
Berlin: De Gruyter Recht.
Morris, R. M. 2001. “‘Lies, Damned Lies and Criminal Statistics’: Reinter-
preting the Criminal Statistics in England and Wales.” Crime, Histoire et
Sociétés/Crime History and Societies 5(1):111–27.
Muncie, J. 2009. Youth and Crime. 3rd ed. Thousand Oaks, CA: Sage.
Næshagen, F. L. 2005. “Den kriminelle voldens U-kurve fra 1500-tall til nå-
tid.” Historisk Tidsskrift [Norway] 84:411–27.
Neuman, W. L., and R. J. Berger. 1988. “Competing Perspectives on Cross-
National Crime: An Evaluation of Theory and Evidence.” Sociological Quar-
terly 29:281–313.
Nilsson, A. 2003. “Living Conditions, Social Exclusion and Recidivism among
Prison Inmates.” Journal of Scandinavian Studies in Criminology and Crime
Prevention 4:57–83.
Nilsson, R. 2003. “The Swedish Prison System in Historical Perspective: A
Story of Successful Failure?” Journal of Scandinavian Studies in Criminology
and Crime Prevention 4:1–20.
Norström, T. 1988. “Theft Criminality and Economic Growth.” Social Science
Research 17:48–65.
———. 1998. “Effects on Criminal Violence of Different Types and Private
and Public Drinking.” Addiction 93:689–99.
Petrosino, A., C. Turpin-Petrosino, and S. Guckenburg. 2010. “Formal System
Processing of Juveniles: Effects on Delinquency.” Campbell Systematic Re-
views 2010:1. DOI: 10.4073/csr.2010.1. http://campbellcollaboration.org/lib/
download/761/.
Phillips, W. 1958. “The Relationship between Unemployment and the Rate
of Change of Money Wages in the United Kingdom 1861–1957.” Economica
25:283–99.
Pratt, J. 2008a. “Scandinavian Exceptionalism in an Era of Penal Excess. Part
I: The Nature and Roots of Scandinavian Exceptionalism.” British Journal
of Criminology 48:119–37.
———. 2008b. “Scandinavian Exceptionalism in an Era of Penal Excess. Part
II: Does Scandinavian Exceptionalism Have a Future?” British Journal of
Criminology 48:275–92.
Quetelet, A. 1842. A Treatise on Man and the Development of His Faculties. Ed-
inburgh: William and Robert Chambers.
Reichel, P. L. 2005. Comparative Criminal Justice Systems: A Topical Approach.
4th ed. Upper Saddle River, NJ: Pearson Education.
Reinke, H. 2009. “Crime and Criminal Justice History in Germany: A Report
on Recent Trends.” Crime, Histoire et Sociétés/Crime, History and Societies 13
(1):117–37.
104 Hanns von Hofer

Roberts, J. V., and M. Hough. 2002. “Public Attitudes to Punishment: The

Context.” In Public Opinion, Crime and Justice, edited by J. V. Roberts and
M. Hough. Cullompton, UK: Willan.
Rossow, I. 2001. “Alcohol and Homicide: A Cross-Cultural Comparison of the
Relationship in 14 European Countries.” Addiction 96(suppl. 1):77–92.
Rousseaux, X. 2006. “Historiographie du crime et de la justice criminelle dans
l’espace français (1990–2005). Partie II: De la Révolution au XXIe siècle.”
Crime, Histoire et Sociétés/Crime, History and Societies 10(2):123–61.
Shelley, L. I. 1981. Crime and Modernization: The Impact of Industrialization and
Urbanization on Crime. Carbondale: Southern Illinois University Press.
———. 1986. “Crime and Modernization Re-examined.” International Annals
of Criminology 24(1–2):7–21.
Sirén, R. 2002. “Trends in Assault: On the Relationship between the Assault
Rate and Selected Social Indicators in Post-war Finland.” Journal of Scan-
dinavian Studies in Criminology and Crime Prevention 3:22–49.
Skardhamar, T. 2003. “Inmates’ Social Background and Living Conditions.”
Journal of Scandinavian Studies in Criminology and Crime Prevention 4:39–56.
———. 2006. Kriminalitet gjennom ungdomstiden blant nordmenn og ikke-vestlige
innvandrere: En analyse av fødselskullet 1977. SSB-Notat 2006/33. Oslo:
Statistisk sentralbyrå. http://www.ssb.no/emner/03/05/notat_200633/notat
_200633.pdf.
Smolej, M., and J. Kivivuori. 2008. “Crime News Trends in Finland: A Review
of Recent Research.” Journal of Scandinavian Studies in Criminology and Crime
Prevention 9:202–19.
Snare, A. 1992. Work, War, and Welfare: The Control of the Laboring Poor in
Sweden. 2nd ed. Copenhagen: Institute of Criminology and Criminal Law.
Soman, A. 1980. “Deviance and Criminal Justice in Western Europe 1300–
1800: An Essay in Structure.” Criminal Justice History 1:1–18.
Stack, S. 1982. “Social Structure and Swedish Crime Rate: A Time-Series
Analysis, 1950–1979.” Criminology 20:499–513.
Statistisk sentralbyrå. 1995. Historisk statistikk 1994. Norges officielle statistikk.
Oslo-Kongsvinger: Statistisk sentralbyrå.
Statistiska centralbyrån. 1958. Brottsligheten 1955. Sveriges officiella statistik.
Rättsväsen. Stockholm: Statistiska centralbyrån.
———. 1967. Den grövre brottsligheten 1965. Statistiska meddelanden R 1967:
29. Stockholm: Statistiska centralbyrån.
———. 2011. Statistisk årsbok för Sverige 2011. Sveriges officiella statistik.
Stockholm: Statistiska centralbyrån.
Statistiska Kommitténs Betänkande. 1910. Sveriges officiella statistik och dess all-
männa organisation. Stockholm: Norstedt.
Stene, R. J. 2008. Narkotikasiktedes lovbrudd preger rettssystemet. Oslo: Statistisk
sentralbyrå. http://www.ssb.no/ssp/utg/200801/03/.
Stoltenbergutvalget. 2010. Rapport om narkotika. Oslo: Helse- og omsorgsde-
partementet. http://www.regjeringen.no/upload/HOD/RappOmNarkotika
_nettversjon.pdf.
 Punishment and Crime in Scandinavia 105

Strafflagberedningen. 1956. Skyddslag. Strafflagberedningens slutbetänkande.

SOU 1956:55. Stockholm: Justitiedepartementet.
Sulkunen, P., C. Sutton, C. Tigerstedt, and K. Warpenius. 2000. Broken Spirits:
Power and Ideas in Nordic Alcohol Control, edited by P. Sulkunen, C. Sutton,
C. Tigerstedt, and K. Warpenius. Helsinki: Nordic Centre for Alcohol and
Drug Research. http://www.nordicwelfare.org/?idp119316.
Sundin, J. 1976. “Theft and Penury in Sweden 1830–1920.” Scandinavian Jour-
nal of History 1:265–92.
———. 1990. “Current Trends in History of Crime and Criminal Justice:
Some Conclusions, with Special Reference to the Swedish Experience.”
Historical Social Research 15(4):184–96.
Sutton, J. R. 2004. “The Political Economy of Imprisonment in Affluent West-
ern Democracies, 1960–1990.” American Sociological Review 69:170–89.
Svensson, R. 2002. “Strategic Offences in the Criminal Career Context.” Brit-
ish Journal of Criminology 42:395–411.
Takala, J.-P., and Y. Obstbaum. 2009. Looking at Violence in the Nordic Countries:
Statistical Sources, Variations, Improving Measurement. TemaNord 2009:45.
Copenhagen: Nordic Council of Ministers.
Taylor, H. 1998. “The Politics of Rising Crime Statistics in England and
Wales, 1914–1960.” Crime, Histoire et Sociétés/Crime, History and Societies 2(1):
5–28.
Tham, H. 1979. Brottslighet och levnadsnivå: En undersökning av män i krimi-
nalregistret födda 1892–1953. Stockholm: LiberFörlag.
———. 2001. “Law and Order as a Leftist Project? The Case of Sweden.”
Punishment and Society 3:409–26.
Tham, H., A. Rönneling, and L.-L. Rytterbro. In this volume. “The Emer-
gence of the Crime Victim: Sweden in a Scandinavian Context.”
Tham, H., and H. von Hofer. 2009. “Individual Prediction and Crime Trends.”
European Journal of Criminology 6:313–35.
Tolgfors, S. 2009. “Därför strider svenska soldater i Afghanistan.” Dagens Ny-
heter, July 29.
Tonry, M. 1997. “Ethnicity, Crime, and Immigration.” In Comparative and
Cross-National Perspectives on Ethnicity, Crime, and Immigration, edited by M.
Tonry. Vol. 21 of Crime and Justice: A Review of Research, edited by M. Tonry.
Chicago: University of Chicago Press.
———. 2009. “Explanations of American Punishment Policies.” Punishment
and Society 11:377–94.
Törnudd, P. 1993. Fifteen Years of Decreasing Prisoner Rates in Finland. Research
Communication 8. Helsinki: National Research Institute of Legal Policy.
———. 1996. Facts, Values and Visions: Essays in Criminology and Crime Policy.
Helsinki: National Research Institute of Legal Policy.
Törrönen, J. 2004. “Zero Tolerance, the Media and a Local Community.”
Journal of Scandinavian Studies in Criminology and Crime Prevention 5:27–47.
Träskman, P.-O. 2004. “Drug Control and Drug Offences in the Nordic Coun-
tries: A Criminal Political Failure Too Often Interpreted as a Success.” Jour-
nal of Scandinavian Studies in Criminology and Crime Prevention 5:236–56.
106 Hanns von Hofer

United Nations. 1993. Declaration on the Elimination of Violence against Women.

UN General Assembly Resolution 48/104 of 20 December 1993. http://
www.un.org/documents/ga/res/48/a48r104.htm.
van Dijk, J. J. M. 1999. “The Experience of Crime and Justice.” In Global
Report on Crime and Justice, edited by G. Newman. New York: Oxford Uni-
versity Press (for the UN Office for Drug Control and Crime Prevention).
———. 2008. The World of Crime: Breaking the Silence on Problems of Security,
Justice, and Development across the World. Los Angeles: Sage.
van Kalmthout, A. M., F. Hofstee-van der Meulen, and F. Dünkel, eds. 2007.
Foreigners in European Prisons. Vols. 1 and 2. Nijmegen, Netherlands: Wolf
Legal.
Verkko, V. 1948. “Världens äldsta våldbrottsstatistik.” Särtryck ur Statistiska
översikter 1948, nos. 11–12.
———. 1951. Homicides and Suicides in Finland and Their Dependence on National
Character, with an Additional Chapter: The Theories of Morselli and Ferri on
Homicides and Suicides and the Attitude to Them of Tarde and Durkheim. Co-
penhagen: Gad.
Vigarello, G. 2001. A History of Rape: Sexual Violence in France from the 16th to
the 20th Century. Cambridge: Polity.
Vikström, L. 2008. “Illuminating the Labeling Impact of Incarceration: Life-
Course Perspectives of Young Offenders’ Pathways in Comparison to Non-
offenders in Nineteenth-Century Northern Sweden.” Crime, Histoire et So-
ciétés/Crime, History and Societies 12(2):81–117.
von Hofer, H. 1975. “Die Entwicklung der Freiheitsstrafe in Schweden.” Kri-
minologisches Journal 7:278–90.
———. 1985. Brott och straff i Sverige: Historisk kriminalstatistik 1750–1984.
Diagram, tabeller och kommentarer. Urval Nummer 18. Stockholm: Statistiska
centralbyrån.
———. 1990. “Homicide in Swedish Statistics, 1750–1988.” Scandinavian
Studies in Criminology 11:29–45.
———. 2000. “Crime Statistics as Constructs: The Case of Swedish Rape Sta-
tistics.” European Journal on Criminal Policy and Research 8:77–89.
———. 2003a. “Crime and Punishment in Sweden: Historical Criminal Justice
Statistics 1750–2000.” Journal of Scandinavian Studies in Criminology and
Crime Prevention 4:162–79.
———. 2003b. “Prison Populations as Political Constructs: The Case of Fin-
land, Holland and Sweden.” Journal of Scandinavian Studies in Criminology
and Crime Prevention 4:21–38.
———. 2004. “Gewaltsensibilität, Gewaltschutzbewegung und Kriminalstatis-
tik.” In Phänomen Strafanzeige, Jahrbuch für Rechts- und Kriminalsoziologie
03, edited by G. Hanak and A. Pilgram. Baden-Baden, Germany: Nomos
Verlagsgesellschaft.
———. 2008. Brott och straff i Sverige: Historisk kriminalstatistik 1750–2008.
Diagram, tabeller och kommentarer. Rapport 2008:1. Stockholm: Kriminolo-
giska institutionen, Stockholms Universitet. http://www.criminology.su.se/
content/1/c6/04/56/86/Brott_och_straff_i_Sverige_2008.pdf.
 Punishment and Crime in Scandinavia 107

von Hofer, H., J. Sarnecki, and H. Tham. 1997. “Minorities, Crime, and Crim-
inal Justice in Sweden.” In Minorities, Migrants, and Crime: Diversity and
Similarity across Europe and the United States, edited by I. Haen Marshal.
Thousand Oaks, CA: Sage.
von Hofer, H., and H. Tham. 1989. “General Deterrence in a Longitudinal
Perspective: A Swedish Case. Theft, 1841–1985.” European Sociological Review
5:25–45.
———. 2000. “Theft in Sweden 1831–1998.” Journal of Scandinavian Studies
in Criminology and Crime Prevention 1:195–210.
Wacquant, L. 2009. Punishing the Poor: The Neoliberal Government of Social In-
security. Durham, NC: Duke University Press.
Westfelt, L. 2001. Brott och straff i Sverige och Europa: En studie i komparativ
kriminologi. Avhandlingsserie no. 5. Stockholm: Kriminologiska institutio-
nen, Stockholms Universitet.
———. 2008. “Svensk brottsutveckling i internationell belysning.” In Brotts-
utvecklingen i Sverige fram till år 2007. Rapport 2008:23. Stockholm: Brotts-
förebyggande rådet.
Wheeler, S., and M. L. Rothman. 1982. “The Organization as Weapon in
White-Collar Crime.” Michigan Law Review 80:1403–26.
WHO (World Health Organization). 2002. World Report on Violence and Health,
edited by E. G. Krug, L. L. Dahlberg, J. A. Mercy, A. B. Zwi, and R. Lozano.
Geneva: World Health Organization.
Wiener, M. J. 1998. “The Victorian Criminalization of Men.” In Men and
Violence: Gender, Honor, and Rituals in Modern Europe and America, edited by
P. Spierenburg. Columbus: Ohio State University Press.
Wikström, P.-O. 1992. “Context-Specific Trends in Homicide in Stockholm,
1951–1987.” Studies on Crime and Crime Prevention 1:88–105.
Wilkins, L. 1964. Social Deviance. London: Tavistock.
Willner, S. 2001. “The Impact of Alcohol Consumption on Excess Male Mor-
tality in Nineteenth- and Early Twentieth-Century Sweden.” Hygiea Inter-
nationalis 2:45–70.
Zimring, F. E., and G. Hawkins. 1991. The Scale of Imprisonment. Chicago:
University of Chicago Press.