Teasing out the effects of macro-conditions on race-specific male homicide rates: Do distinct predictors vary by racial group and over time?

Richard Stansfield, Karen F. Parker

Research output: Contribution to journalArticle

10 Scopus citations

Abstract

Researchers tend to capture the multiple disadvantages facing urban areas by using an all-encompassing disadvantage index, which combines poverty, joblessness and other economic predictors into a single index. While the use of this index is important for conceptual and methodological reasons, questions remain about whether these city characteristics differ in magnitude and significance when influencing race-specific homicide rates and whether or not there effects exhibit stability or vary over time? This article examines how discrete measures of disadvantage differ in their importance for race-specific groups over three critical time points: 1980, 1990, and 2000. After accounting for problems associated with statistical inferences, cross sectional, Seemingly Unrelated Regression (SUR) analyses reveal that family disruption and poverty status were among the strongest predictors of race-specific homicide rates. Wald tests for the equality of coefficients confirmed significant differences in the influence of many discrete measures of disadvantage for white and black males, but the number of differences declined from 1980 and into the 2000s. That is, along with the crime drop, our research reveals increasing racial parity in structural predictors over time.

Original languageEnglish (US)
Pages (from-to)633-649
Number of pages17
JournalSocial Science Research
Volume42
Issue number3
DOIs
StatePublished - May 1 2013
Externally publishedYes

All Science Journal Classification (ASJC) codes

  • Education
  • Sociology and Political Science

Fingerprint Dive into the research topics of 'Teasing out the effects of macro-conditions on race-specific male homicide rates: Do distinct predictors vary by racial group and over time?'. Together they form a unique fingerprint.

  • Cite this