Race and Violent Offender “Propensity”: Does the Intraracial Nature of Violent Crime Persist on the Local Level?

• Author: Sarah Becker
• Published date: 01 December 2007
• Date: 01 December 2007
• DOI: http://doi.org/10.3818/JRP.9.2.2007.53
• Subject Matter: Article

Race and Violent Offender “Propensity” • 53

Sarah Becker

University of Massachusetts-Amherst

* Abstract

In the 1980s, some scholars suggested black violent offenders had a “propensity”

to select white victims. Subsequent research demonstrated, however, that violent

offenders tend to victimize intraracially. One weakness of this research was its

reliance on national-level data. Using national-level estimates to calculate expected

values of interracial offending implicitly assumes people have access to each other

across the nation. Segregation patterns suggest otherwise. This study examines

whether the propensity of violent offenders to select victims intraracially holds up

if expected values are calculated locally. Findings indicate that violent offenders do

tend to select victims intraracially at the local level, but that the intraracial character

of violent offending varies by crime, offender race, and locale. For most cities in the

analysis, assault is predominantly intraracial across offense/offender categories. For

a few cities, however, criminal assault is less intraracial than expected (with white

offenders victimizing interracially more than random selection would predict).

I would like to thank Roland Chilton for teaching me how to work with NIBRS data,

Douglas Anderton for assisting with my analysis, and patiently and intelligently answer-

ing questions too numerous to count, and Jill McCorkel and Robert Faulkner for reading

drafts of this paper and consistently supporting my work.

JUSTICE RESEARCH AND POLICY, Vol. 9, No. 2, 2007

© 2007 Justice Research and Statistics Association

*

Race and Violent Offender “Propensity”:

 Does the Intraracial Nature of Violent

 Crime Persist on the Local Level?

54 • JUSTICE RESEARCH AND POLICY

Race and Violent Offender “Propensity” • 55

In the late seventies and early eighties, some criminological research focused at-

tention on black violent offenders choosing white victims more often than white

violent offenders selected black victims (Chilton & Gavin, 1985; Katz & Ma-

zur, 1979; LaFree, 1982; LeBeau, 1984; Pallone & Hennessy, 1999; Wilbanks,

1985). Some of these studies even claimed that black offenders choosing white

victims in a majority (or near majority) of violent offenses demonstrated a “pro-

pensity” of black offenders to seek out white victims (LaFree, 1982; LeBeau,

1984; Wilbanks, 1985).

Subsequent research demonstrated that though black offenders might

choose white victims more often than white offenders choose black victims, it

is primarily because more white people are in the potential victim pool than

black people (O’Brien, 1987). In fact, studies showed that black offenders

choose white victims less often than we would expect if they randomly se-

lected victims from the potential victim pool (and vice versa) (O’Brien, 1987;

Koch, 1995). This work led to the conclusion that if any “propensity” on the

part of black (and white) offenders exists, it is the propensity to select victims

of one’s own race.

One admitted weakness of these studies, however, was reliance on national-

level data. Models for calculating expected amounts of inter/intraracial crime

were constructed using national-level percentages of offenders reported in the

National Crime Victimization Survey (NCVS) and national-level racial popula-

tion proportions as potential victim pools. This assumes that black and white

offenders and victims are spread homogenously throughout society when they

are not (O’Brien, 1987, p. 825).

Given segregation patterns in the United States, ndings indicating an

overwhelmingly intraracial character of violence at the national level are hardly

surprising. A more powerful nding would require demonstrating that violent

crime remains predominantly intraracial within a bounded locality (like a city

or neighborhood) in which it is reasonable to expect that crime would occur. It

is possible, then, that if residential segregation patterns are taken into account

when calculating expected values of inter/intraracial offending, the observed

propensity of offenders to select victims within race will be diminished. In other

words, if we calculate the salience of race in victim selection at the city level or

(more importantly) the neighborhood level (where, hypothetically, 50% of po-

tential victims are black) instead of at the national level (where, hypothetically,

13% of potential victims are black), it is possible that race will be less salient

in victim selection. It is possible that offenders’ victim selection patterns on the

local level more closely parallel available victim pools or that black offenders

exhibit the greater “propensity” to select white victims, as Wilbanks (1985) and

others have suggested (LaFree, 1982; LeBeau, 1984).

Because it provides city-level data on both victim and offender/arrestee race

and ethnicity, the National Incident-Based Reporting System (NIBRS) allows for

this sort of investigation of violent offending. This study examines the salience

54 • JUSTICE RESEARCH AND POLICY

Race and Violent Offender “Propensity” • 55

of race in non-bias-motivated assault offending for 27 mid-sized cities in the

2002–2004 NIBRS reports in “full color”—including white, black, Asian, Na-

tive American, and Latino/a offenders/arrestees and victims.1

* Interracial Crime Research and the Inter/Intraracial Nature of

Criminal Violence

Macrostructural opportunity theories—rooted in Blau’s (1977) work and

drawing on criminal opportunity theory and routine activities theory—argue

that interracial violence is a function of opportunity and access. Specically,

they suggest that the more heterogeneous, dense, and integrated a population

is (or the more opportunities people of different races have to interact with one

another), the more interracial violence will occur. Several studies have found

relationships between a host of macrostructural opportunity factors (hetero-

geneity of population, population density, degree of urbanization, residential

segregation/integration, social contact/routine activities, unemployment levels,

and income inequality) and levels of interracial rape, robbery, assault, theft,

and homicide (Block, 1985; Jacobs & Wood, 1999; Jang, Messner, & South,

1991; Koch, 1995; Lee, Martinez, & Rosenfeld, 2001; Martinez, 2000; Mar-

tinez & Lee, 2002; Martinez, Nielsen, & Lee, 2003; Messner & South, 1986;

Messner & South, 1992; O’Brien, 1987; Parker & McCall, 1999; Sampson,

1984; Sampson, 1986; South & Felson, 1990; Tremblay & Tremblay, 1998;

Wadsworth & Kubrin, 2004).

Racial threat theory offers an alternate explanation for levels of interracial

violence. Most racial threat research has focused on formal or informal social

control directed at minority populations (Blalock, 1967; Jacobs, 1979; Jacobs

& Britt, 1979; Greenberg, Kessler, & Loftin 1985; Anderson, 1990; Jacobs &

Helms, 1999; Crawford, Chiricos, & Kleck,1998; Jacobs & O’Brien, 1998;

Venkatesh, 2000; Eitle, D’Alessio, & Stolzenberg, 2002; Behrens, Uggen, &

Manza, 2003; Stolzenberg, D’Alessio, & Eitle, 2004) and rare events like riots

1 “Asian” category includes Pacic Islanders and “Native American” includes Alas-

kan Natives. Latino/a category includes those classied as “Hispanic” in NIBRS arrestee

or victim records. Data on Hispanic origin are not available for offenders. They are only

available for a subset of arrestees and victims for whom “Hispanic origin” data were

reported. Submission of Hispanic origin data by agencies is voluntary. Therefore I pres-

ent two analyses: victim-offender dyads (which do not include Latino/a as a separate cat-

egory) and victim-arrestee dyads (which include non-Latino/a white, non-Latino/a black,

non-Latino/a Native American, non Latino/a-Asian, and Latino/a victims and arrestees).