Impact of Employment, Family Structure, and Income on NIBRS Offense, Victim, Offender, and Arrest Rates

• Author: Roland Chilton,Wendy C. Regoeczi
• DOI: 10.3818/JRP.9.2.2007.9
• Date: 01 December 2007
• Published date: 01 December 2007
• Subject Matter: Article

NIBRS Offense, Victim, Offender, and Arrest Rates • 9

* Impact of Employment, Family Structure, and

 Income on NIBRS Offense, Victim, Offender,   

 and Arrest Rates

Roland Chilton

University of Massachusetts-Amherst

Wendy C. Regoeczi

Cleveland State University

* Abstract

In some studies of urban crime, offense, arrest, and victimization counts have been

used as if they were interchangeable. The National Incident-Based Reporting System

(NIBRS), with its provision of offense, victim, offender, and arrest counts lets us avoid

this ambiguity and examine the relationship of social and economic factors to specic

crime measures. Our basic question in this study was, “Does it matter whether we

focus on victims or offenders when we examine and interpret the correlates of violent

crime?” Using NIBRS 2002–2004 robbery and murder counts and census 2000 data

for 166 cities, we examined the importance of male employment, income, racial com-

position, and family structure for an “offense known” measure, a victim measure, an

offender measure, and two arrest measures of crime. Our results indicate that for race-

specic measures even the best murder or robbery victim rate is a poor substitute for

an offender rate, and offense-known measures are even poorer measures of offending.

The NIBRS offender rate is the most logical and useful offender measure because it

provides a direct indication of the age, race, and sex of most offenders whether or not

an arrest occurs. However, arrest rates, and modied arrest rates in particular, provide

reasonably close approximations of offender rates. In addition, and unexpectedly, we

found that race-specic victim and offender counts and rates take much of the mystery

out of the frequently used and variously interpreted variable “percent black.”

Support for this project was provided by the American Statistical Association’s Committee

on Crime, Law, and Justice Statistics.

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

© 2007 Justice Research and Statistics Association

10 • JUSTICE RESEARCH AND POLICY

NIBRS Offense, Victim, Offender, and Arrest Rates • 11

* Appropriate Crime Measures

Does it matter whether we focus on victims or offenders when we examine

and interpret the correlates of violent crime? This unstated question is raised

by the data used in a number of studies of urban violence. It is sometimes

raised openly, as when Krivo and Peterson (1996, p. 624) write, “[O]ur data

are for reported victimizations only. While it is important to study offending,

the links hypothesized should be evident in analyses of these rates.” Krivo and

Peterson used “offenses known to the police” as their victimization measure

and clearly indicate they think the results would not be different if they used

offender rates. Other researchers use offense-known data with other crime

measures, but most writers avoid explicit discussion of the issue (Kposowa,

Breault, & Harrison, 1995; Land, McCall, & Cohen, 1990). Some use victim-

ization rates from the Federal Bureau of Investigation’s (FBI’s) Supplementary

Homicide Reports les (SHR) or from the National Center for Health Statis-

tics cause of death reports as measures of homicide, and some argue that one

would expect to see similar ndings “when using either homicide victimiza-

tion or offending data” (Strom & MacDonald, 2007, p. 56; see also Weaver,

Martin, & Petee, 2004; Krivo & Peterson, 2000; Parker, 2004; Phillips,

1997). Still others use arrest rates or modied arrest rates as offender rates

(LaFree, Drass, & O’Day, 1992; Sampson, 1987; Shihadeh, & Ousey, 1996;

Shihadeh & Maume, 1997; Harer & Steffensmeier, 1992).

Determining the correlates of victimization or offending is not a problem

for those using sample survey procedures to study criminal conduct. If you

ask people to describe themselves and to tell you the type and frequency of

crimes they have committed, you have offenders and offender characteristics.

If you ask people to describe themselves and to tell you about the type and

frequency of incidents in which they were victims of crime, you have victims

and victim characteristics. Generally, self-reported crime surveys tell us little

about victims and victimization surveys tell us something but not much about

offenders. In both cases you have to assume the reports are accurate and that

the sample you use represents a useful segment of the general population. But

linking the offenses reported to other information the respondents provide is

not a problem.

In contrast, identifying the characteristics of those involved in violent

crimes reported to urban police departments is not as straightforward because

traditional reports of offenses known to the police tell us nothing about the

offenders or the victims involved. Therefore, additional information must be

collected to provide offender or victim information. As noted above, one ap-

proach has been to collect and use the age, race, and sex of those arrested as

an indication of the characteristics of offenders. However, only some of those

committing violent offenses are arrested, a limitation that led to at least one

adjustment procedure. Sampson (1987) and others used the offense-known