Does ethnic diversity increase violent crime? A global analysis of homicide rates, 1995–2013
| Published date | 01 March 2020 |
| Date | 01 March 2020 |
| Author | Carmen Noel,Indra de Soysa |
| DOI | 10.1177/1477370818775294 |
| Subject Matter | Articles |
https://doi.org/10.1177/1477370818775294
European Journal of Criminology
2020, Vol. 17(2) 175 –198
© The Author(s) 2018
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/1477370818775294
journals.sagepub.com/home/euc
Does ethnic diversity increase
violent crime? A global analysis
of homicide rates, 1995–2013
Indra de Soysa
Norwegian University of Science and Technology, Norway
Carmen Noel
University of Auckland, New Zealand
Abstract
Many scholars argue that diverse preferences and coordination failure stemming from high ethnic
diversity results in high social frictions, leading to socio-political failure. Criminological theories
suggest that crime is driven by very similar processes. The specialized literature on civil war,
however, reports a diversity dividend, arguing that when two large groups (polarization) make
up a society, the risk of armed violence is increased. Using data on global homicide rates from
the period 1995–2013 for over 140 countries, we find that ethnic heterogeneity is associated
with homicide rates in an inverted U-shape relationship. Measures of ethnic polarization confirm
these results directly. The results suggests that ethnic polarization and ethnic dominance rather
than diversity are what matter for personal security measured as homicide rates. The conditional
effect of high diversity and income inequality is associated with lower homicide rates, results that
reject the view that societal heterogeneity and income inequality drive social dislocation. Several
possible intervening variables, such as unemployment among males and youth, ethnic exclusion
and discrimination, good governance and institutional quality, as well as several demographic and
political variables, do not affect the basic results. It seems that the heavy emphasis placed on
ethnic diversity for explaining social dislocation and violence, in so far as it relates to a country’s
homicide rate, seems to be misplaced.
Keywords
Ethnic diversity, ethnic polarization, homicide rates, social inequality
Corresponding author:
Indra de Soysa, Department of Sociology and Political Science, Norwegian University of Science and
Technology, Trondheim 7491, Norway.
Email: indra.de.soysa@svt.ntnu.no
775294EUC0010.1177/1477370818775294European Journal of Criminologyde Soysa and Noel
research-article2018
Article
176 European Journal of Criminology 17(2)
Introduction
Conflict is thought to be pervasive in ethnically divided societies (Horowitz, 2000). Several
scholars of development and governance blame high social diversity, particularly ethnic
diversity, for the failure of development in many poor countries (Alesina et al., 2011;
Alesina and La Ferrara, 2005).1 The issue has even gained salience in Western societies
where immigration has spurred discussions on social stability (Putnam, 2007). In much of
the specialized literature, however, the question of ethnic divisions and societal disarray is
highly contested (Cederman and Girardin, 2007; Fearon and Laitin, 2003), although a few
recent studies report that higher ethnic diversity increases violent crime (Altheimer, 2007;
Cole and Gramajo, 2009). We revisit this issue with the most updated data and more com-
prehensive geographical and temporal coverage than all previous studies, and we employ
several appropriate statistical techniques for panel data. Like others, we believe that the
homicide rate of a country is a very valid measure of the ‘everyday nature’ of physical
security and socio-political stability, rather than large-scale armed violence (Rivera, 2016;
Stretesky et al., 2016). The issue is not just academic but is highly relevant for current
debates about building decent multicultural societies and policy aimed at enhancing the
prospects of socioeconomic development around the world. Is high diversity owing to so-
called ‘artificial borders’ a problem for social stability and everyday physical security?
Using newly available data on homicide rates covering a large set of countries (142)
over the 1995–2013 period and several measures of ethnic diversity and polarization, we
find that ethnic fractionalization relates to the homicide rate in an inverted U-shape,
results similar to studies on civil war that argue that polarization matters more than frac-
tionalization (Esteban and Ray, 2008). A direct measure of polarization and a measure of
the size of the second-largest ethnic group confirm the curvilinear effect of fractionaliza-
tion, which supports the view that relative homogeneity is more problematic than diver-
sity per se. Interestingly, the effect of ethnic diversity is conditioned negatively on the
homicide rate, the higher the income inequality. We also tested several possible interven-
ing variables that are theoretically linked to homicide rates that could have explained why
ethnic polarization matters. There is little support for grievance-based explanations of
homicide as theorized by structural theories, and, quite surprisingly, ethnic discrimination
lowers the homicide rate. There is strong evidence to suggest that good governance and
strong institutions, rather than political freedoms alone, lower the homicide rate. Our
results taken together do not warrant the recent pessimism linking ethnic heterogeneity to
high social frictions, insofar as homicide rates capture aspects of socio-political failure.
Theory
Sociological and criminological theories of crime stress the role of deviance, social
stress, relative deprivation and group discrimination in explaining the prevalence of seri-
ous crime (Durkheim, 1933; Merton, 1968; Palma, 1995). Most studies in criminology,
perhaps correctly, directly focus on the individual level, looking at personality traits,
neighbourhood factors and socioeconomic situations to explain why some people engage
in crime whereas others do not. A few studies in criminology have addressed the question
of how ethnic configuration within a country might explain criminality more broadly, but
To continue reading
Request your trial