![Sample our Social Sciences journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/110822/TOC-Subject-Sample-2021-Social-Sciences.jpg"})
ABSTRACT
The focus on the institutional control of violent crime has increased over the past few decades, stimulated largely by Messner and Rosenfeld’s “institutional anomie theory.” A related theory, referred to as “institutional legitimacy theory” in this study, has received considerably less attention. This theory, originating in the social control theoretical tradition, is tested in an analysis of homicide rates (circa 2012) across 108 nations. Overall, institutional legitimacy theory receives support as economic, political, familial, and religious institutional legitimacy assist in reducing homicide rates across societies. Most notably, homicide is positively associated with the size of the shadow economy (the measure of waning economic institutional legitimacy), consistent with the hypothesis that parties lacking institutional redress are more likely to use unilateral violence to resolve grievances.
Acknowledgement
The author would like to thank Patricia L. McCall for her helpful comments on previous versions of this manuscript.
Notes
1. While analysing Cook’s distance statistics, it became apparent that Haiti is an influential outlier. Haiti is excluded from the final analysis (originally, there were 109 nations in the sample).
2. To maximise sample size, observations missing for 2012 are imputed using estimates of 2011 and 2010.
3. Traditionally, scholars have drawn upon the United Nations (Citation2014) Demographic Yearbook for divorce and marriage statistics (Greenstein & Davis, Citation2006). However, these data have very limited coverage in developing nations, especially in Africa. For this reason, the survey data outlined above are utilised. When different estimates for the same year are available, surveys conducted directly by the United Nations are utilised. For Albania, Malaysia, Myanmar, and Vietnam, estimates of a total of “divorced” and “separated” (combined) are the only divorce data available but are included nevertheless to avoid losing cases. Estimates for 2013 are used for Sierra Leone and Turkey as older measures were either incomplete or exceeded a 5-year window preceding 2012. While this may appear to violate time ordering of the independent and dependent variables, the imputation of 2013 data is conducted under the assumption that divorce and marriage rates are relatively stable annually and the goal of this study is to capture stable differences between nations in institutional legitimacy and homicide. Substantive results are largely the same without the inclusion of these two cases.
4. Where available, estimates of income shares based upon disposable income were used to maintain consistency between nations. In 53 cases, disposable income statistics are utilised, while 49 cases represent consumption income results. The remaining six cases are of miscellaneous definitions (gross, primary income, and “other”).
5. An additional model (not shown) that does not include the percentage of males aged 15–29 was estimated, finding that HDI was a significant (negative) predictor of the natural logarithm of homicide rates. Another model (not shown) omits the corruption measure and reveals a statistically significant coefficient for HDI, another indication of the partialling fallacy. All other results remained substantively the same (except for Eastern religion failing to reach statistical significance in the first additional model), confirming that institutional control of homicide is a robust predictor of homicide and HDI has an influence on homicide rates that may be obscured in this analysis.
6. The United Nations Office of Drugs and Crime (UNODC) (Citation2013) typically draws upon a mixture of direct public health estimates of mortality, often drawn from the World Health Organization (WHO) (Citation2017), and estimates provided by national criminal justice systems. However, in a few cases, especially in sub-Saharan Africa, homicide estimates were dependent upon indirect regional data or modelled based upon known correlates to homicide. The 18 cases produced by model-based approaches eliminated for the supplemental analysis are Burkina Faso, Burundi, Cameroon, Ethiopia, Ghana, Guinea, Iran, Laos, Malaysia, Mali, Myanmar, Niger, Nigeria, Rwanda, Togo, Tanzania, Vietnam, and Zimbabwe.
Additional information
Notes on contributors
James Tuttle
James Tuttle is a PhD candidate at North Carolina State University. He studies violent crime and social control in comparative and cross-national contexts. A previous version of this manuscript won the Student Paper Award for the Division of International Criminology of the American Society of Criminology.