The Role of the State on Cross-National Homicide Rates

• Author: Rollin F. Tusalem,Doris C. Chu
• Date: 01 September 2013
• Published date: 01 September 2013
• DOI: 10.1177/1057567713500790
• Subject Matter: Articles

Article

The Role of the State

on Cross-National

Homicide Rates

Doris C. Chu

1

and Rollin F. Tusalem

2

Abstract

Although previous research has established important links between economic factors (e.g.,

economic inequality), social support variables, ethnic fractionalization, and cross-national homicides,

one important area remains unexplored. Relatively little is known about how the role of the state

affects cross-national homicide rates. Using data from 183 nations, this study examined the effect of

the role of the state (political instability, regime type, and use of the death penalty) on homicide

rates. The findings indicate that homicide rates are higher in countries that are politically unstable.

In addition, anocracies (transitional states that have a mixture of democratic and autocratic features)

were found to be correlated with higher homicide rates. Suggestions for future research are

discussed.

Keywords

cross-national homicide rates, homicide rates, political stability, inequality, ethnic fractionalization,

homicide, political turmoil, regime type, anocracy

Introduction

In the last three decades, studies examining cross-national homicides have proliferated. Never-

theless, with the exception of a few studies (e.g., LaFree & Tseloni, 2006; Neumayer, 2003; Stama-

tel, 2009), most research on the cross-national determinants of homicides have focused on macro-

social foundations (e.g., Altheimer, 2007; Chon, 2012; Pridemore, 2011), social support theories

(e.g., Chamlin &Cochran, 1997; Pratt & Godsey, 2002, 2003; Worrall, 2005), cultural aspects

(Altheimer, 2012; Stamatel, 2009), and economic indicators (Jacobs & Richardson, 2008; Messner,

1982, 1989; Sun, Chu, & Sung, 2011). Increasingly, there has been an emphasis on how the attri-

butes and policy decisions made by the state can influence and impact social violence (Altheimer,

1

Department of Criminology, Sociology, and Geography, Arkansas State University, Jonesboro, AR, USA

2

Department of Political Science, Arkansas State University, Jonesboro, AR, USA

Corresponding Author:

Doris C. Chu, Department of Criminology, Sociology, and Geography, Arkansas State University, P.O. Box 2003, Jonesboro,

AR 72467, USA.

Email: dchu@astate.edu

International CriminalJustice Review

23(3) 252-279

ª2013 Georgia State University

Reprints and permission:

sagepub.com/journalsPermissions.nav

DOI: 10.1177/1057567713500790

icj.sagepub.com

2012; Stamatel, 2009). A major objective of this article is to assess what specific attributes within

a state may affect variation in homicides using a larger sample of countries.

Most of the extant empirical work on cross-national homicides includes sample sizes of less than

100 (e.g., Jacobs & Richardson, 2008; LaFree & Tseloni, 2006; Messner, 1989; Messner, Raffalo-

vich, & Shrock, 2002; Pratt & Godsey, 2003; Stamatel, 2009). Our article departs from this conven-

tion by incorporating a sample of more than 100 nations/states, which increases the generalizability

of the results. Thus, the major contribution of this article is to avoid the pitfalls of case-selection bias.

More specifically, previous work on cross-national variation in homicides often excludes cases from

the underdeveloped world. Therefore, this study has more theoretical and analytical leverage from

previous work.

In the past three decades, a number of studies have found various predictors and correlates of

cross-national homicide rates. Indicators of economic inequality have been consistently found to

be predictive of cross-national variations in homicide rates (e.g., Avison & Loring, 1986; Lee,

2001; Messner, 1982, 1989; Messner et al., 2002; Neapolitan, 1998; Sun et al., 2011). Neumayer

(2003) extended previous studies by including variables of political governance and the death

penalty to predict cross-national homicide rates. He found that economic growth, income levels,

protection for human rights, and the abolition of the death penalty are all negatively correlated with

cross-national homicide rates. Nevertheless, inconsistent with the findings of most studies on cross-

national homicide rates, Neumayer did not find income inequality as a significant predictor of cross-

national homicides.

Using data derived from 44 countries from 1959 to 2000, LaFree and Tseloni (2006) found that

a shift from autocracies to democracies was associated with increased homicide rates.

1

Although

LaFree and Tseloni find that transitional states are positively associated with high homicide rates,

their sample of 44 countries are mostly composed of industrialized democracies, which largely

excludes the developing world. Analyzing data from nine countries in East-Central Europe from

1990 through 2003, Stamatel (2009) found more wide-ranging political reforms toward democracy

were correlated with lower homicide rates. It is our contention that our study has more empirical

leverage in assessing whether state-level variables have any effect on homicide rates, considering

that we employ more than 120 nation-states with available data providing a larger sample of devel-

oping states from Africa, Asia, and Latin America.

In a recent study, Chamlin and Cochran (2006) advanced previous studies by examining whether

perceived legitimacy of the structure of stratification mitigates the effect of income inequality on

cross-national homicide rates. Although they did not find the perceived legitimacy of stratification

mediated the effect of economic inequality on homicide rates, the findings of their study revealed

that perceived illegitimacy of economic and political order was correlated with higher homicide

rates in more developed countries than in less developed countries (p. 248).

In a study that examined cross-national variation in homicides in the context of Latin America,

Neapolitan (1994) alluded to the potential role of political turmoil in possibly contributing to

extremely high homicide rates in Colombia and El Salvador. Although Neapolitan took the caveat

that this may not apply to the whole region or even globally, he suggested that a culture of violence

may be attributed to some unexamined or inappropriately operationalized economic, political,

social, or demographic variable and made a call for further examination of these aspects.

This study aims to fill the gap in literature in two ways. First, it examines the effect of the role of

the state (political instability, regime type, and use of the death penalty) on homicide rates in a global

context. Second, this study employs a substantial sample size (183 nations) to increase the explana-

tory power of our findings. This article begins with a review of the literature concerning cross-

national variation in homicide rates. This is followed by the methodology section, which explains

how we operationalized our main independent variables (political instability, regime type, and the

use of the death penalty) and control variables. We then present our results showing how certain

Chu and Tusalem 253