Do National Homicide Rates Follow Supranational Trends?

• Author: Meghan L. Rogers,William Alex Pridemore
• Date: 01 November 2018
• Published date: 01 November 2018
• DOI: 10.1177/0022427818785210
• Subject Matter: Articles

Article

Do National Homicide

Rates Follow

Supranational Trends?

Meghan L. Rogers

1

and William Alex Pridemore

2

Abstract

Objectives: We explored supranational trends in national homicide rates.

We searched for a global trend, regional trends, and trends specific to other

theoretically relevant groups of nations. We also tested two common

metanarratives—modernization and conflict—as potential explanations for

any global trend present in homicide rates. Method: We obtained annual

homicide victimization rates for 94 nations between 1979 and 2013. We

examined year-to-year differences, squared semipartial correlation coeffi-

cients to search for supranational trends, and pooled cross-sectional mixed

models to test potential explanations of any global trend. Results: There was

a very weak global homicide trend. We found strong regional trends in

Eastern Europe and in Northern Europe, a weak trend for South and

Central America, and n o trend for Asia. Both we althy and nonwealthy

nations exhibited weak trends. Transitional nations shared a strong homi-

cide trend. Modernization and conflict theories fared poorly as explanations

for the weak global trend. Conclusions: Thepresenceorabsenceof

1

Department of Sociology and Criminology, University of North Carolina–Wilmington,

Wilmington, NC, USA

2

School of Criminal Justice, University at Albany–State University of New York, Albany, NY,

USA

Corresponding Author:

Meghan L. Rogers, Department of Sociology and Criminology, University of North Carolina–

Wilmington, 601 S College Rd., Bear Hall 207, Wilmington, NC 28403, USA.

Email: rogersml@uncw.edu

Journal of Research in Crime and

Delinquency

2018, Vol. 55(6) 691-727

ªThe Author(s) 2018

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/0022427818785210

journals.sagepub.com/home/jrc

supranational homicide trends holds significant implications for theory. A

weak global trend is evidence against widely held metanarratives such as the

modernization, civilizing, and conflict perspectives. Strong subregional

homicide trends in Eastern Europe and Northern Europe demand further

exploration and should shift popular attention away from Western Europe.

The lack of a homicide trend in developed or developing nations and the

presence of a strong trend among transitional nations are curious features

requiring further consideration.

Keywords

homicide, supranational trends, aggregate time series, squared semipartial

correlation coefficients

The possibility of a global homicide trend is provocative. Criminologists

tend to believe structural causes of crime act locally due to the character-

istics of neighborhoods (Sampson, Raudenbush, and Earls 1997), cities

(Land, McCall, and Cohen 1990), or nations (Pridemore and Trent 2010).

We may even consider forces such as the impact on U.S. crime rates of

decades-long changes to the structure of routine activities (Cohen and Fel-

son 1979) or the effect on the Russian homicide rate of the social dereg-

ulation accompanying the collapse of the Soviet Union (Pridemore,

Chamlin, and Cochran 2007). It would be remarkable, though, if nations

distributed geographically throughout the world and with ve ry different

histories, cultures, levels of development, state institutions, polities, and

violence rates followed the same global trend. Recent work hypothesizes

global and regional violence trends over both the long (Pinker 2011) and

short terms (Karstedt 2015; Tonry 2014).

Analyses of historical records, mostly of Western nations, provide per-

suasive evidence violence declined substantially over the last millennium

(Eisner 2003; Gurr 1981; Pinker 2011). The existence of a recent global

homicide trend, however, remains unclear. The presence of a global

homicide trend would have important implications for existing social

theory—from modernization theories to conflict theories to the demo-

graphic transition model—and perhaps demand new explanations from

paradigms as wide ranging as evolutionary psychology and global culture.

In spite of strong statements about the presence of a single homicide trend in

recent decades among Western nations (Tonry 2014), which may be

expanding throughout the world early in the twenty-first century (Baumer

692 Journal of Research in Crime and Delinquency 55(6)

and Wolff 2014; Weiss et al. 2016), a handful of careful studies showed no

support for a global homicide trend or provided evidence for a weak trend

only among a subset of nations (Baumer and Wolff 2014; LaFree 2005;

LaFree, Curtis, and McDowall 2015).

Our aim was to determine whether nati onal homicide rates followed

global, regional, or other supranational trends over the last three decades.

We first took the same philosophical and analytical approach McDowall

and Loftin (2009) used to learn whether city crime rates follow a national

trend in the United States. Thus, this portion of our study was descriptive in

nature, not explanatory, and meant to discover whether patterns exist. We

answered this question by computing year-to-year differences and squared

semipartial correlation coefficients for annual homicide victimization rates

from a sample of 94 nations between 1979 and 2013. The number of

observations (i.e., nation-years) in our sample was substantially larger than

all but one of the few prior studies attempting to discern a supranational

homicide trend, and we employed similar analytical methods to McDowall

and Loftin (2009) and to LaFree et al. (2015) in their assessment of country-

level declines in homicide rates since 1950.

After finding evidence for various supranational trends and a weak glo-

bal trend, we tested for potential explanations of the latter. We focused on

modernization and conflict theories, which are popular explanations of

long-term trends in violence rates across nations, in an attempt to determine

whether these metanarratives accounted for differences in homicide victi-

mization across waves and between nations over the several decades.

Literature Review

The Existence of Supranational Homicide Trends

The first objective of our study was to determine whether supranational

homicide trends exist (McDowall and Loftin 2009).

1

The second objective

was to determine whether core concepts from two theories account for the

weak trends in homicide victimization we observed in our exploratory

analyses. We took this approach for two reasons.

The first reason is that compelling theoretical statements with persuasive

narratives about sweeping homicide rate changes over time make us sus-

ceptible to assuming the presence of a trend. The simple presences of a local

trend, as observed by Tonry (2014), may not support a universal trend in

homicide victimization. An aggregate time series is a function of the num-

ber and character of the individual time series that compose it. The

Rogers and Pridemore 693