Regional, structural, and demographic predictors of violent victimization: A cross-national, multilevel analysis of 112 countries
| Published date | 01 May 2020 |
| Author | Rodney Stark,Katie E Corcoran |
| DOI | 10.1177/0269758019869108 |
| Date | 01 May 2020 |
| Subject Matter | Articles |
Article
Regional, structural,
and demographic predictors
of violent victimization:
A cross-national, multilevel
analysis of 112 countries
Katie E Corcoran
West Virginia University, United States of America
Rodney Stark
Baylor University, United States of America
Abstract
Routine activities and lifestyles theories emphasize structural and demographic predictors that
affect the opportunity to commit crime and the likelihood of being victimized. Past research tends
to focus on either the individual- or country-level with few studies incorporating both. Addi-
tionally, past research primarily draws on the International Crime Victimization Survey, which
results in small country samples that are biased toward developed nations. The current study uses
data from a larger, more diverse sample that allow us to test whether findings from prior studies
are generalizable. We are also able to theorize and test the effect of region on victimization. We
find some differences between our findings and past research; we also identify that region is one of
the strongest predictors of violent victimization.
Keywords
Routine activities theory, victimization, violence, cross-national
There is extensive research in criminology on crime victimization and its correlates. Much of this
research draws on routine activities theory (Cohen and Felson, 1979) and lifestyles theory
(Hindlelang et al., 1978). This work stresses the importance of demographic and structural factors
Corresponding author:
Katie E Corcoran, Department of Sociology and Anthropology, West Virginia University, PO box 6326, Morgantown,
WV 26506-6326, United States of America.
Email: kecorcoran@mail.wvu.edu
International Review of Victimology
2020, Vol. 26(2) 234–252
ªThe Author(s) 2019
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/0269758019869108
journals.sagepub.com/home/irv
that facilitate or impede the opportunity to commit crime (Wilcox et al., 2003). There is a large
literature investigating these factors in the context of the United States or with one or two other
country samples (e.g. Arnold et al., 2005; Cass, 2007; Fisher and Wilkes, 2003; Fisher et al., 1998;
Keane and Arnold, 1996; Kennedy and Forde, 1990; Lynch, 1987; Miethe and McDowall, 1993;
Miethe et al., 1987; Spano and Freilich, 2009). There is considerably less research exploring these
theories of victimization cross-nationally. What research does exist tends to emphasize either the
individual level, typically using fixed-effects models to control for country, or the country level,
often aggregating individual-level data or using country rates (Stein, 2010). In this way, it fails to
consider how individual- and country-level factors may both affect the opportunity for crime (see
Stein, 2010, 2013; Tseloni and Farrell, 2002; van Wilsem et al., 2003 for notable exceptions). The
current study investigates routine activities theory and lifestyles theory on both the individual and
country levels.
Additionally, past research on cross-national victimization almost exclusively draws on the
International Crime Victimization Survey (ICVS), which generally results in small country sam-
ples (fewer than 50 countries) that are heavily biased toward Western developed countries. Thus,
we do not know whether findings from past studies using primarily Western nations apply more
broadly. It is also unclear whether regional indicators may contribute to explaining variations in
victimization cross-nationally. A meta-analysis of cross-national homicide rates identified that a
Latin American dummy variable has the strongest effect of any variable (Nivette, 2011). Yet we
are unaware of any victimization study that models its effect or that of other regions, such as Sub-
Saharan Africa, which has an even higher crime rate than Latin America, the highest of any region
(Morrison et al., 2003). Both regions are likely to provide an increased opportunity for crime
victimization. The failure to investigate regional effects is likely due to the sample size constraints
of the ICVS.
The present study uses victimization data from surveys including large numbers of respon-
dents and nations. Since 2005, the Gallup World Poll (GWP) has conducted annual (or biennial)
nationally representative surveys of adults aged 15 and older in over 160 nations, which repre-
sent more than 99%of the world’s population. A victimization question is a core question on the
survey. Combining country-level variables from prior research with the individual-level GWP
data allows us to test routine activities and lifestyles theories of cross-national victimization with
a diverse sample of 112 countries and over 200,000 individuals. It also allows us to investigate
regional predictors of victimization. Although many of our results are consistent with past ICVS
studies, some findings differ. This suggests the need for more studies using large and diverse
country samples.
Cross-national victimization and routine activities and lifestyles
theories
Central to the study of victimization is identifying what individual and structural characteristics
affect an individual’s likelihood of being victimized. Routine activities and lifestyles are the main
theories used to predict victimization. The main argument of routine activities theory is that the
activities individuals engage in affect their likelihood of being victimized. Routine activities are
‘any recurrent or prevalent activities which provide for basic population and individual needs’,
such as going to work (Cohen and Felson, 1979: 593). These activities contribute to a greater
opportunity for victimization by being in a particular space in which ‘motivated offenders’,
‘suitable targets’, and ‘the absence of capable guardians’ converge (Cohen and Felson, 1979:
Corcoran and Stark 235
To continue reading
Request your trial