Uniform Crime Reports

Every month, local law enforcement agencies are asked to record for

their jurisdictions the total number of murders, rapes, robberies, aggravated

assaults, burglaries, larcenies, motor vehicle thefts, and arsons on a

form known as UCR Return A.2 For robberies and aggravated assaults,

counts broken down by type of weapon (firearms; knives or cutting instruments;

other weapons; and personal weapons, such as hands, feet, fists,

etc.) are requested. Participation in the UCR program is voluntary.

The UCR Return A data offer a relatively long monthly time series of

robberies and assaults by firearms and other weapons occurring in local

police jurisdictions across the country. However, administrative data such

as UCR have a different set of problems than the NCVS. Foremost among

them is that these data alone cannot be used to draw inferences about

firearms use or victimization in the general population.3 The UCR is a

sample of crimes reported to and recorded by local law enforcement agen-

2The UCR program excludes jurisdictions covered by federal law enforcement agencies.

3In fact, the NCVS was created to address this problem by capturing data on both reported

and unreported crimes, to develop a clearer picture of national crime trends.

cies in the United States. Ideally, they reveal the number of crimes per

month for each of the reporting jurisdictions. Of course, many crimes are

not reported to the police, so increases or decreases in reports for certain

offenses, such as burglary and auto theft, can result in large differences in

outcomes and misleading conclusions about crime trends.

Other reporting problems may further limit the usefulness of these

data. First, the accuracy of UCR data can be compromised by differences in

definitions of crimes and reporting protocols. Local authorities, for example,

might choose criminal charges to achieve certain objectives (e.g.,

increasing plea bargains by downgrading what might otherwise be a charge

of aggravated assault, armed robbery, or rape to a lesser charge that then

gets reported in the UCR).

Second, participation in the UCR program is voluntary, with smaller,

more rural police agencies less likely to submit reports than larger, urban

departments. A review of the preliminary 2000 UCR data posted on the FBI’s

web site indicates that in one large midwestern state, only six cities with

10,000 and over population reported arrest data between January and June.

In all, there were six states that could provide only limited data. For example,

rape data were unavailable for two states because the state reporting agencies

did not follow the national UCR guidelines (http://www.FBI.gov/ucr/

99cius.htm). The committee is not aware of research that details how this

nonresponse problem affects inferences in firearm-related research. Maltz

and Targonski (2002) argue that underreporting in the UCR data may bias

the results of research on right-to-carry laws, but they do not document the

magnitude of these biases (see Chapter 6 for further details).

Finally, because these data are based on monthly counts and not on

individual incidents, only limited detail is available regarding crime circumstances.

There is no information, for example, on the nature or severity of

the injuries inflicted. The Supplemental Homicide Report (which is part of

the UCR program) provides limited information on the relationship between

victim and offender and event circumstances (e.g., whether the homicide

is related to an argument or the commission of another felony).

National Incident-Based Reporting System

The National Incident-Based Reporting System (NIBRS) is designed to

provide detailed incident-level information on crimes, including firearm-related

crimes. It is administered through the FBI’s UCR program and augments

the crime reports of local law enforcement agencies in several key

respects: offense categories are greatly expanded; attributes of individual

crime incidents (offenses, offenders, victims, property, and arrests) can be

collected and analyzed; arrests and clearances can be linked to specific inciMEASURING

dents or offenses; and all offenses in an incident can be recorded and

counted.4 NIBRS is intended to replace the UCR as the nation’s comprehensive,

standardized crime data source based on crimes known to the police.

However, since its blueprint was published in 1985 (Poggio et al.,

1985), only 16 percent of the U.S. population is covered by NIBRS data

(Bureau of Justice Statistics, 2001), with few large cities or urban areas

participating. Thus, at this time, NIBRS is not an effective data set for

studying firearms violence.