Divergent Effects of Beliefs in Heaven and Hell on
National Crime Rates
Azim F. Shariff1*, Mijke Rhemtulla2
1Department of Psychology, University of Oregon, Eugene, Oregon, United States of America, 2Center for Research Methods and Data Analysis, University of Kansas,
Lawrence, Kansas, United States of America
Though religion has been shown to have generally positive effects on normative ‘prosocial’ behavior, recent laboratory
research suggests that these effects may be driven primarily by supernatural punishment. Supernatural benevolence, on the
other hand, may actually be associated with less prosocial behavior. Here, we investigate these effects at the societal level,
showing that the proportion of people who believe in hell negatively predicts national crime rates whereas belief in heaven
predicts higher crime rates. These effects remain after accounting for a host of covariates, and ultimately prove stronger
predictors of national crime rates than economic variables such as GDP and income inequality. Expanding on laboratory
research on religious prosociality, this is the first study to tie religious beliefs to large-scale cross-national trends in pro- and
Citation: Shariff AF, Rhemtulla M (2012) Divergent Effects of Beliefs in Heaven and Hell on National Crime Rates. PLoS ONE 7(6): e39048. doi:10.1371/
Editor: Tiziana Zalla, Ecole Normale Supe ´rieure, France
Received February 13, 2012; Accepted May 17, 2012; Published June 18, 2012
Copyright: ? 2012 Shariff, Rhemtulla. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The authors have no funding or support to report.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: firstname.lastname@example.org
A growing program of research from across the social sciences
now supports the long-held claim that religion positively affects
normative behavior (see  for a review). Religiosity shows
consistent positive correlations with charity and volunteerism ,
and negative relations with lax attitudes about the justifiability of
moral transgressions . Moreover, experimental work has shown
that religious priming increases ‘prosocial’ generosity and co-
operation, and decreases cheating [4–6].
However, recent studies suggest that not all religious beliefs are
equal in this respect. Though supernatural punishment is
associated with increases in normative behavior, laboratory
research reveals the concept of supernatural benevolence to be
associated with decreases in normative behavior. For example,
university students with stronger beliefs in in God’s punitive and
angry nature tended to be the least likely to cheat on an academic
task, whereas stronger beliefs in God’s comforting and forgiving
nature significantly predicted higher levels of cheating . These
results remained robust after controlling for plausible third
This pattern of results is consistent with theories highlighting the
effectiveness of supernatural punishment–specifically–at regulating
moral behavior and, as a result, group cooperation [8–9,1]. These
theories argue that human punishment is a highly effective
deterrent to anti-social behavior within groups, but one that faces
inevitable limitations of scale. Human monitors cannot see all
transgressions, human judgers cannot adjudicate with perfect
precision, and human punishers are neither able to apprehend
every transgressor, nor escape the potential dangers of retribution.
Divine punishment, on the other hand, has emerged as a cultural
tool to overcome a number of those limitations. Unlike humans,
divine punishers can be omniscient, omnipotent, infallible, and
untouchable-and therefore able to effectively deter transgressors
who may for whatever reason be undeterred by earthly policing
Supernatural benevolence, however, is not theorized to be
similarly effective at stabilizing cooperation within groups.
Moreover, the evidence thus far suggests that though the more
‘positive’ religious attributes may provide their own benefits, such
as better self-esteem  or health coping , their role in
encouraging moral behavior may be, at best, minimal and, at
Indeed, recent social psychological research has used priming
experiments to establish causality in this negative relationship.
Christian participants spent ten minutes writing about God’s
forgiving nature, God’s punitive nature, a forgiving human,
a punitive human, or a neutral control. In a subsequent and
purportedly unrelated task, participants were given the opportu-
nity to overpay themselves for the study using a common measure
of petty theft where participants ostensibly privately assign
themselves payment for correctly answered anagrams. Though,
on average, participants in the punishing God and both human
conditions overpaid themselves less than 50 cents more than what
they deserved for their anagrams, and did not statistically differ
from the neutral condition, those who wrote about a forgiving God
overpaid themselves significantly more-nearly two dollars (Un-
fortunately, because the level for both the control and the
punishing god prime conditions were at a statistical floor in this
study, the ability of punishing god primes to reduce levels of theft
could not be reliably assessed) .
The important question that arises from all of these religion and
prosocial behavior studies is to what degree do these laboratory-
based effects translate to large-scale societal effects? Ultimately,
PLoS ONE | www.plosone.org1 June 2012 | Volume 7 | Issue 6 | e39048
what is the relation between the punitive and benevolent aspects of
religion and pro/anti-social behavior in the world beyond the
laboratory? If the theories and findings described above reflect
substantial psychological effects, then differences in supernatural
punishment and benevolence beliefs should also differentially
predict transgressive actions such as criminal behavior at the
societal level. Using large international datasets, we investigated
this prediction by examining the role of beliefs in heaven and hell.
Materials and Methods
Data for belief in hell, belief in heaven, belief in God, and
religious attendance were taken from the 1981–1984, 1990–1993,
1994–1999, 1999–2004, and 2005–2007 waves of the World
Values Surveys (WVS) and European Value Surveys . Some
countries included the question in multiple survey years (individual
survey participants only participated once); others included the
question at only one data collection wave. In total, these data were
based on participants from 67 countries (N=143,197; mean N per
country=2137, range=36229016). Weighted means were com-
puted for each country based on a proportional weighting variable
supplied with the WVS.
Belief in heaven, hell and God was assessed with the oral
question, ‘‘Which, if any, of the following do you believe in?’’,
followed by a list of concepts including ‘‘Heaven,’’ ‘‘Hell’’ and
‘‘God’’. Accepted answers were Yes and No. Religious attendance
was assessed with the question, ‘‘Apart from weddings, funerals,
and christenings, about how often do you attend religious services
these days?’’; the response options were 1=More than once
a week, 2=Once a week, 3=Once a month, 4=Only on special
6=Once a year, 7=Less than once a year, 8=Never or
practically never. A weighted average was computed for each
country across all available data, using the supplied individual
Mean standardized crime rates were computed from the 10
crimes for which the United Nations Office on Drugs and Crime
(UNODC)  had reliable statistics: homicide (N=67 countries),
robbery (N=51), rape (N=48), kidnapping (N=46), assault
(N=48), theft (N=47), drug crime (N=47), auto theft (N=28),
burglary (N=43), and human trafficking (N=39). These data were
compiled by the UNODC from national government sources,
including police and court records, national statistics ministries,
and other national government bodies. For each individual crime,
the annual data from 2003–2008 were averaged to form a single 5-
year measure, except for homicide data, which were taken from
the latest available single year (range 2004–2010).
Per capita GDP data were taken from the 2007 CIA World
Factbook . Dominant religion, Gini coefficient data and life
expectancy data were 2011 estimates (or the latest available
estimate where 2011 estimates were not available) drawn from the
CIA World Factbook . Urban density data reflect the
percentage of each country’s population residing in urban areas
in 2005; these data are from the United Nations Department of
Economic and Social Affairs . Population imprisonment data
were taken from the World Prison Population List .
Personality data were taken from Schmitt, Allik, McCrae &
Benet-Martı ´nez , who administered translated versions of the
Big Five Inventory of personality differences to 17,837 individuals
in 56 nations. We elected to add only conscientiousness,
agreeableness and neuroticism to the analysis as they have been
previously linked to pro- and anti-social behavior, however, results
do not dramatically change if all five personality factors are
entered into the regression. All the above data are publicly
All variables were entered into a series of linear regression
equations, with each crime regressed on beliefs in heaven and hell
as well as all covariates. Standardized beta coefficients for each of
these analyses, as well as for the aggregate average of the 10 crimes
are presented in Table 1.
As predicted, rates of belief in heaven and hell had significant,
unique, and opposing effects on crime rates. Belief in hell
predicted lower crime rates, b
hell=21.941, p,.001; whereas
belief in heaven predicted higher crime rates, b
p,.001 (Note that these are standardized regression coefficients, so
they may be interpreted as effect sizes). Controlling for the effect of
belief in heaven, a 1 SD increase in belief in hell resulted in an
almost 2 SD decrease in national crime rate; conversely,
controlling for the effect of hell, a 1 SD increase in belief in
heaven resulted in an almost 2 SD increase in national crime rate.
Analyzing each crime individually revealed the same significant
pattern of effects for 8 of the 10 individual crimes (kidnapping and
human trafficking excepted; see Table 1).
To discount the role of obvious third variables, we conducted
a second analysis with several covariates. Dominant religion was
included in the form of three dummy coded variables that
indicated a country’s predominant religious group as Roman
Catholic, Other Christian, and Muslim . We included two
standard economic factors relevant to crime rates: income
inequality (measured by the Gini coefficient,  and GDP per
capita ; national imprisonment rates as a measure of
a country’s punitive nature ; two demographic factors that
reflect important differences between nations: life expectancy 
and urban density ; three of the ‘‘Big Five’’ personality
variables that have been previously tied to pro- and anti-social
behavior: conscientiousness, neuroticism and agreeableness ;
and finally two factors specifically focused on the religiousness of
the different nations: belief in God and religious attendance 
(Reported results use maximum likelihood estimation to deal with
missing data. Listwise deletion (N=53) gave the same pattern of
Despite many of these variables–especially poverty and income
inequality–being frequently discussed as determinants of crime
[20,21], only belief in God had a significant effect on average
crime rates over and above the effects of belief in heaven and hell,
which remained highly significant (both ps ,.001). Moreover,
when analyzing the individual crimes with these covariates, beliefs
in heaven and hell still emerged as the strongest predictors for 5 of
the 10 crimes (see Table 1).
The strength of these effects is made clear by examining crime
rates as a function of the degree to which a greater percent of
people in a nation believe in heaven than in hell. As Figure 1
depicts, rates of belief in heaven are virtually always higher than
rates of belief in hell; however, as the degree of that discrepancy
increases (from roughly equal proportions of the population
believing in the two concepts, to up to 40% more believing in
heaven than in hell), so too do crime rates.
The pattern of results is robust to spatial and cultural variability.
The same pattern of results emerges in three out of the four
continental zones for which there were sufficient data-namely,
Africa, South and Central America, and Europe plus Canada,
USA, Australia and New Zealand. The same pattern also emerges
for three out of the four religious groups that form national
majorities–predominantly Roman Catholic, predominantly non-
Heaven and Hell Beliefs Predict Crime
PLoS ONE | www.plosone.org2June 2012 | Volume 7 | Issue 6 | e39048
Catholic Christian, and predominantly ‘Other,’ which comprises
of either unaffiliated majorities, or more localized majority
religions such as Hinduism, Shintoism and syncretic religions that
combine Islam and Christianity with traditional indigenous
religions (see Figure 1). The only exception to this observation is
predominantly Muslim countries in Asia, for which the uniformly
high levels of both belief in heaven and hell (Ms=93% and 91%
respectively), produce insufficient variance for prediction.
The present analysis has uncovered two strong, unique, and
reliable relations between religious belief and national crime rates.
The degree to which a country’s rate of belief in heaven outstrips
its rate of belief in hell significantly predicts higher national crime
rates. Statistically, this finding manifests in two independent
effects: the strong negative effect of rates of belief in hell on crime,
and the strong positive effect of rates of belief in heaven on crime.
Both of these effects follow from predictions based on recent
laboratory findings [7,12] and on theories that ascribe socio-
cultural functions of religions . Indeed, these findings coalesce
with theoretical and empirical work suggesting that beliefs in
punishing and omniscient supernatural agents spread across
historical societies primarily because of their ability to foster
cooperation and suppress anti-social behavior among anonymous
Limitations and Future Directions
First and foremost, these findings are correlational, and thus
reverse-causation and third variable explanations need to be
discounted before causal claims can be firmly endorsed. However,
at least two reasons suggest that a causal effect of these religious
beliefs on crime is a plausible explanation for the pattern of results.
First, obvious third variable candidates such as differences
between countries in national personality, wealth, wealth distri-
bution, and general religiosity show no indication of driving the
effects. Second, numerous lab studies have established direct
causal effects for religious beliefs on both pro- and anti-social
behaviors. The possibility remains that the lab effects and the
international crime rate effects are entirely unrelated, but
parsimony suggests that both are, at least to some degree,
a reflection of the same underlying causal story. Nevertheless,
future research would be beneficially directed towards addressing
possible alternative explanations.
Understanding the mechanisms underlying these effects is
another important area of future inquiry. As discussed in the
introduction, much research now supports the conclusion that the
prosocial effects of religion are due, at least in part, to the fear of
supernatural punishment serving as a deterrent to transgressive
behaviors. The antisocial effects of religious benevolence, which,
counting the current findings, have been now documented in
several studies (e.g. [7,12]), are, however, less well understood.
Thus far, researchers have focused their speculation on the idea
that divine forgiveness offers individuals a way to cleanse their
Table 1. Regression of individual crime rates on beliefs in heaven, hell, and 13 covariates.
Traffick-ing Kidnapping RapeRobbery Theft
Analysis with no
Belief in heaven 1.728*** 1.905*** 1.536*** 1.25***1.244***–0.1960.325 1.733***1.29***1.325*** 1.958***
Belief in hell–1.788*** –2.186*** –1.876*** –1.594***–0.947***0.337 –0.174 –1.79*** –1.288***–1.779***–1.941***
Belief in heaven 2.079*** 1.094*0.931 1.314**0.935*–0.370–0.843 2.031***0.2801.020*** 1.820***
Belief in hell –2.075*** –0.979–0.853 –1.413** –0.850*0.7251.175*–2.075*** –0.376–1.048**–1.698***
Roman Catholic–0.2580.085 0.3070.1570.140 0.090 0.242–0.228 0.2830.0850.112
Other Christian0.069 –0.005 0.3510.415* –0.044–0.0490.0910.146 –0.0990.297**0.170
Muslim0.106 0.0210.1270.455* –0.134–0.477* –0.114 0.137 –0.1360.0770.051
–0.444 0.361 0.310 0.055–0.247 0.972*** 0.864*–0.5090.478 0.293*–0.124
Pop. imprisoned–0.031 0.369*** 20.071
20.451* 20.080 0.0290.299 0.0180.077
GDP per capita
20.079 0.022 0.3940.785***
Belief in God
20.069 0.546*** 0.012
Note. As recommended by , regression results are presented with and without covariates. Asterisks indicate significance levels according to Wald tests (*=
significant at a=.05,**= significant a=.005,***= significant at a=.001). Significance tests should be interpreted cautiously, as no correction has been made for
inflated error rates due to performing a large set of analyses. As 12 analyses were performed, only those effects that are significant at a=.005 or below may be
confidently interpreted as significant.
Heaven and Hell Beliefs Predict Crime
PLoS ONE | www.plosone.org3June 2012 | Volume 7 | Issue 6 | e39048
moral palate, and thereby feel more licensed to transgress again.
That is, divine forgiveness, like its earthly variant, may act as
a counter-deterrent. This hypothesis garners support from several
related studies (e.g. [23,24]), but remains open to further testing.
Finally, in the efforts to uncover the mechanisms at play, it will
be important to examine these real-world effects at the level of the
individual. The present findings tie rates of belief at the societal
level to national crime rates; the direct causal explanation for this
effect is that individuals who believe in heaven and not hell take
punishment less seriously and are thus more likely to commit
crimes. It is also possible, however, that an intervening variable or
variables are at work at the societal level. It may be that
widespread belief in a forgiving god may lead a society to value
forgiveness over punishment, and that this secular value in turn
affects crime rates. In this scenario, an individual’s belief in heaven
or hell may not directly affect her proclivity to engage in criminal
behavior. The direct causal explanation is most closely in line with
the experimental findings, but it could well be that both the direct
and indirect mechanisms are at work. To assess individual-level
effects simultaneously with societal-level effects, it will be necessary
to collect data with both national crime rates and individual
tendencies toward immoral behavior.
These findings not only help to explain the differential relations
that supernatural punishment and benevolence have to moral
behavior-a topic of considerable recent interest in the social
sciences-but also raise important questions about the potential
impact of religious beliefs on global crime. Though little research
in economic and social policy concentrates on religion, economists
have observed that hell beliefs may positively impact the economic
growth of developing nations . It is quite possible that the
present findings, which tie belief in hell to lower levels of anti-social
behavior, may serve as one of the key explanatory mechanisms
underlying this economic trend. Indeed, simply given the strength
of the results compared to standard economic indicators such as
GDP and the Gini coefficient, social scientists and policy makers
might more deeply consider the cultural impact of religious beliefs
in future work.
Conceived and designed the experiments: AS. Performed the experiments:
AS MR. Analyzed the data: AS MR. Wrote the paper: AS MR.
1. Norenzayan A, Shariff AF (2008) The Origin and Evolution of Religious
Prosociality. Science 322: 58–62.
2. Monsma SV (2007) Religion and Philanthropic Giving and Volunteering:
Building Blocks for Civic Responsibility. Int J Res Religion 3:1–27.
3. Atkinson QD, Bourrat P (2010) Beliefs about God, the afterlife and morality
support the role of supernatural policing in human cooperation. Evol Hum Behav
Figure 1. Crime rate z-scores as a function of how much higher the proportion of a nation that believes in heaven is compared to
the proportion that believes in hell. R2=.54.
Heaven and Hell Beliefs Predict Crime
PLoS ONE | www.plosone.org4 June 2012 | Volume 7 | Issue 6 | e39048
4. Shariff AF, Norenzayan A (2007) God is watching you: Supernatural agent Download full-text
concepts increase prosocial behavior in an anonymous economic game. Psychol
Sci 18: 803-809.
5. Ahmed AM, Salas O (2011) Implicit influences of Christian religious
representations on dictator and prisoner’s dilemma game decisions. J Socio-Econ
6. Randolph-Seng B, Nielsen ME (2007) Honesty: One effect of primed religious
representations. Int J Psy Rel 17: 303–315.
7. Shariff AF, Norenzayan A (2011) Mean Gods Make Good People. Int J Psy Rel
8. Johnson DDP (2011) Why God is the best punisher. Religion Brain Behav 1: 77–84.
9. Johnson DDP, Kru ¨ger O (2004) The good of wrath: Supernatural punishment
and the evolution of cooperation. Pol Theol 5: 159–176.
10. Francis LJ, Gibson HM, Robbins M (2001) God images and self-worth among
adolescents in Scotland. Ment Health Relig Cult 4: 103–108.
11. Ironson G (2011) View of God as benevolent and forgiving or punishing and
judgmental predicts HIV disease progression. J Behav Med 34: 414–425.
12. DeBono A, Shariff AF, Muraven M (2012) Forgive Us Our Trespasses: Priming
a Forgiving (but not a punishing) God Increases Theft. Manuscript under review.
13. World Values Survey, 1981–2005 (2009) Official Aggregate v.20090901, World
Value Survey website. Available: http://www.wvsevsdb.com/wvs/WVSData.
jsp. Accessed 2011 November 28.
14. United Nations Office on Drugs and Crime (2011) Crime and Crim Just Stat.
United Nations Office on Drugs and Crime website. Available: http://www.
unodc.org/unodc/en/data-and-analysis/statistics/crime.html. Accessed: 2011
15. Central Intelligence Agency (2007) The World Factbook. CIA website. Available:
html. Accessed: 2011 November 28.
16. Central Intelligence Agency (2011) The World Factbook. CIA website. Available:
2102rank.html. Accessed 2011 November 28.
17. United Nations Department of Economic and Social Affairs (2010) Population
density and urbanization. United Nations Statistics Division website. Available:
Accessed 2011 November 28.
18. Walmsley R (2011) World Prison Population List, Ninth Edition. King’s College
London, International Centre for Prison Studies. The Institute for Democracy
and Conflict Resolution website. Available: http://www.idcr.org.uk/wp-
content/uploads/2010/09/WPPL-9-22.pdf. Accessed 2012 April 26.
19. Schmitt DP, Allik J, McCrae RR, Benet-Martı ´nez V (2007) The Geographic
Distribution of Big Five Personality Traits: Patterns and Profiles of Human Self-
Description Across 56 Nations. J Cross-Cult Psy 38: 173–213.
20. Patterson EB (1991) Poverty, Income Inequality, and Community Crime Rates.
Crim 29: 755–76.
21. Wilkinson R, Pickett K (2009) The Spirit Level (Allen Lane, London).
22. Simmons JP, Nelson LD, Simonsohn U (2011). False-Positive Psychology :
Undisclosed Flexibility in Data Collection and Analysis Allows Presenting
Anything as Significant. Psychol Sci 22: 1359–1366.
23. Lamborn SD, Mounts NS, Steinberg L, Dornbusch SM (1991) Patterns of
competence and adjustment among adolescents from authoritative, authoritar-
ian, indulgent, and neglectful families. Child Dev 62(5): 1049–1065.
24. Zhong C, Liljenquist K (2006) Washing away your sins: Threatened morality
and physical cleansing. Science 313: 1451–1452.
25. Barro R, McCleary R (2003) Religion and economic growth across countries.
Am Soc Rev 68: 760–781.
Heaven and Hell Beliefs Predict Crime
PLoS ONE | www.plosone.org5 June 2012 | Volume 7 | Issue 6 | e39048