Poverty and Violence
The scholarly literature provides extensive and consistent evidence that poverty is associated with violent crime. Most conspicuously, poverty has been associated with homicide rates using both standard measures of the poverty rate (e.g., Bailey, 1984; Kposowa, Breault, & Harrison, 1995; Kovandzic, Vieraitis, & Yeisley, 1998; Parker & Pruitt, 2000; Stretesky, Schuck, & Hogan, 2004; Williams, 1984) and index measures of socioeconomic deprivation (Baller, Anselin, Messner, Deane, & Hawkins, 2001; Hannon, 2005; Land, McCall, & Cohen, 1990; Lee, Maume, & Ousey, 2003; Matthews, Maume, & Miller, 2001; Morenoff et al., 2001; Sampson et al., 1997). Pratt and Cullen (2005) conducted a meta-analysis of all macro-level studies of crime from 1960-1999. They classify poverty as one of only five predictors that are high strength and high stability across types of studies. Although findings of a negative relationship between poverty and homicide do exist (e.g., Sullivan, 1985), they are uncommon. Even authors of studies that challenge a definitive relationship between poverty and homicide acknowledge that the effect may be contingent upon context (Crutchfield, Glusker, & Bridges, 1999; Lee et al., 2003) or type of homicide (Kovandzic et al., 1998) rather than concluding that the association is nonexistent. The often-cited studies of Blau and Blau (1982) and Messner (1982) did find no significant relationship between poverty and homicide. However, when Williams (1984) reestimated their models, he concluded that their null findings were due to model misspecification. Williams found that poverty has a nonlinear effect on homicide. Indeed, summarizing the literature as a whole, Pridemore (2002) concludes that measures of poverty are the most consistent structural predictors of homicide.
Poverty has been consistently associated with other indicators of violent crime as well (e.g., Dunaway, Cullen, Burton, & Evans, 2000; Hsieh & Pugh, 1993; Peterson, Krivo, & Harris, 2000; Sampson et al., 1997). In the Cambridge Study in Delinquent Development, Farrington (1978) reported that low family income was a precursor of violent offending. Forty-four percent of the families of violent offenders had low family income when those offenders were first encountered at age eight (Farrington, 1991). Elliott et al. (1989) show that low- SES subjects in the NYS had a much higher prevalence of assault and robbery than middle class subjects. By contrast, the association between SES and theft was confusing. Felony theft rates were higher among lower class Ss for one of three years, minor theft rates are higher for middle class youths than lower class youths, and there is evidence that the remaining minor crimes were more prevalent among middle class than lower class youths (e.g., public disorder, alcohol use).