Assessing Prominent Explanations for U.S. Homicide Trends: A Meta-Analysis

• In Event: 4448 - New Directions and Innovation in the Study of Crime and Justice

Abstract

A substantial body of research has accumulated over the past quarter century to explain recent fluctuations in U.S. homicide rates, and several plausible explanations have been offered. This research seeks to assess the relative strength of the most debated and empirically tested of these explanations, with a focus on changes in the economy, age structure, policing, and incarceration. In addition to assessing the relative strength of these explanations, special emphasis is placed on the impact of methodological variation, including variable operationalization and unit of analysis, on the results. Meta-analytic techniques were used to quantitatively synthesize 2,453 effect sizes from 91 different studies examining the relationship between these and other possible explanations and post-WWII U.S. homicide trends published between 1990 and 2016. Contrary to the emphasis in the extant literature, meta-analytic results reveal strong support for the role of changes in structural features, including family dynamics (via both single parent households and divorce), disadvantage, and racial composition. Results also reveal strong support for the role of incarceration on declines in homicide trends. Other common explanations, including indicators of policing and age structure only received moderate support, and changes in the economy received little to no empirical support. These results indicate that, overall, many of the most often-debated explanations are not receiving strong empirical support and suggest other possible explanations for future research to consider. Results also reveal that for some of the explanations examined, the strength of the relationship is conditioned on methodology. The implications of these findings are discussed.

Authors

• Ashley Michelle Mancik, University of South Carolina

• Karen F. Parker, University of Delaware

• Graham C. Ousey, College of William & Mary