SUMMARY

Since 1930, the Federal Bureau of Investigation (FBI) has administered the Uniform Crime Reporting Program (UCR). The UCR was designed to provide nationally representative estimates of the level and change in level of crimes known to and recorded by the police as well as arrested persons. To obtain these data, the FBI surveys state and local police organizations. Because not all types of crime are equally well reported to the police, the UCR focuses on seven types of crime that are (1) prevalent, (2) serious, and (3) well reported to the police. These seven types of crime are known as Part I offenses and constitute the Crime Index; they are homicide, rape, robbery, aggravated assault, burglary, larceny, and motor-vehicle theft. (Arson was added in 1978, but is not reported as part of the Crime Index.) Participants in the UCR report all of these crimes coming to their attention. The reporting of other offenses (Part II offenses) is not required of jurisdictions that participate in the program. Arrests made in both Part I and Part II categories are reported, as well as data on law enforcement personnel.

The amount of data on juveniles in the UCR is quite limited. Coverage problems in the system further limit the usefulness of the available data. Moreover, because the system is jurisdiction-based rather than incident-based, the data cannot be manipulated extensively to provide estimates useful in national estimation or policy research.

SAMPLING

The UCR is intended to be a census of state and local police agencies. The FBI says that agencies participating in the system provide service to approximately 94% of the residential population of the U.S. This coverage varies across time and data items. In recent years coverage on arrests ranged from 80% of the residential population in 1994 to just 68% of the residential population in 2001.

DATA COLLECTION PROCEDURES

The UCR is an administrative series of official police records that are reported at the jurisdiction level.

INSTRUMENTATION

The UCR employs a set of forms on which jurisdictions can report aggregate counts of offenses, arrests, personnel, and other information. These forms and their content are listed below:

• Return A--Index offenses known to the police, unfounded offenses, clearances, and number of clearances involving persons under 18.
• Supplement to Return A--counts of Index Offenses that involve certain attributes such as monetary value of stolen property or type of residence entered forcibly.
• Monthly Return of Arson Offenses Known to the Police.
• Age, Race, Sex, and Ethnic Origin of Persons Over 18 Arrested.
• Age, Race, Sex, and Ethnic Origin of Persons Under 18 Arrested—includes juvenile disposition data.
• Supplementary Homicide Report--includes information on victims, offenders, and context of homicide offenses.
• Law Enforcement Officers Killed or Assaulted.
• Number of Full-Time Law Enforcement Employees.

KEY VARIABLES

The UCR system collects data on all crimes except traffic offenses and federal crimes. The offenses are grouped into two categories: Part I and Part II. Counts of the number of offenses reported to the police, as well as arrest and clearance rates, are collected for Part I offenses, whereas only arrest data are reported for Part II offenses. The Part I offenses include criminal homicide, forcible rape, robbery, aggravated assault, burglary, larceny-theft, motor-vehicle theft, and, since 1978, arson. All other offenses are categorized as Part II offenses. For the purposes of the Uniform Crime Reports, offenses are reported by a uniform definition rather than by a state's own penal code.

For the Part I offenses, UCR collects information on all offenses known or reported to the police, regardless of whether an arrest resulted. In addition to the number of offenses reported or known, data are collected on the number of these reports proved to be unfounded (false or baseless), the number of offenses cleared by arrest or exceptional means, and the number of clearances involving only persons under 18. The monetary value of property stolen and recovered is recorded for Part I offenses. Also, details relevant to specific Part I offenses are collected, such as weapon, offender, and victim characteristics, and circumstances of murders; the type of weapon and type of premise in robberies; and the time and premise type (residential or nonresidential) of burglaries.

Information on arrests is collected for both Part I and Part II offenses, including information on the offender's age, sex, race, and ethnic origin. For persons under 18, arrest information includes data on police disposition, that is, whether the offender was handled within the department and released or referred to a court or other agency.

Finally, the UCR gathers information on law enforcement personnel. It collects data on officers killed and assaulted and conducts a yearly count of law enforcement employees by department. Employees are distinguished according to their sworn or civilian status, full- or part-time status, and sex.

QUALITY CONTROL

The extent of quality control procedures used at the state and local levels varies considerably. Checks by local contributors range from no review in a summary-based department where one police officer has been designated the UCR expert, to a sophisticated dual-reporting system in which calls for service are entered into both an automated records system and a manual log. The distribution of extensive and minimal quality control efforts among participating agencies is unknown.

Local contributors whose UCR operations are automated rely on programmed edit checks to identify errors. Most automated departments verify keypunched data, that is, information is keypunched twice and the two are compared for transcription errors made during data entry. All state programs perform edit checks of the data submitted to them by local reporting entities. Manual edits include statistical verification when crime trends show large fluctuations. Automated edit checks search for logical inconsistencies in the data, omission of information, duplicate reporting, arithmetic errors, and discrepancies in data across reports.

The Crime Statistics Processing Unit of the FBI-UCR section is responsible for data entry and quality control of the statistics contributed both by state programs and by agencies that report directly to the FBI. When the data are received, they are checked to ensure that the agency identifiers and month codes are correct and to check the reasonableness of the offense counts and monetary values. Consistency checks are performed across forms, e.g., comparing the number of arrests on the Age, Sex, Race, and Ethnic Origin (ASREO) report with the number of clearances on the Return A. Further edit checks include addition, Return A offense counts versus Supplement offense counts, excessive amounts, and others. After the checks are complete, the corresponding UCR master files are updated.

Implemented in 1997 as a pilot program, the Quality Assurance Review (QAR) conducted by the Criminal Justice Information Services (CJIS) Audit Unit became a permanent program in January 2000, intended to ensure that each state UCR program adheres to summary and incident-based reporting methods that are consistent with UCR standards. The QAR assesses the accuracy and consistency of crime reporting with onsite reviews of local case reports and an examination of collection and compilation of local agency statistics by the state repositories. By the year 2002, 315 agencies had been reviewed in 44 states, Guam, and the District of Columbia.

PERIODICITY

The UCR collects data on a monthly basis and issues major reports annually and preliminary reports semiannually.

REPRESENTATIVENESS

Although coverage is extensive, the UCR is not a census of police organizations. The UCR can be used, with considerable caution, in both cross-sectional and longitudinal designs. There is some under-coverage, but more importantly, there seems to be substantial nonresponse across jurisdictions and data items. For example, Florida, Indiana, Kansas, Kentucky, Mississippi, Montana, New Mexico, Ohio, Oklahoma, South Carolina, Wisconsin, and the District of Columbia did not submit all or part of their reported crime and arrest year 2000 data to the FBI in time for inclusion in the year 2000 Crime in the United States. However, many jurisdictions are late reporters rather than nonreporters; therefore, data from them may be included in publicly available data files (see section on Data Access, below).

Missing data also occurs at the agency level within reporting states: individual agencies may fail to report for an entire year or part of a year. Imputation is used to adjust state and national crime and arrest rates for nonreporting and nonresponse, but the imputation techniques are not very sophisticated; how well the imputation reflects the crime and arrest rates of nonreporting/underreporting agencies is unknown. Care must be taken in generalizing UCR results for a specific item to the nation in any given year and across time.

There is evidence of further coverage problems in the arrest and disposition data. The number of agencies reporting arrest data is roughly two-thirds that reporting offenses known to the police. The level of participation of agencies also varies over time. Among the agencies reporting arrest data for some months, it is unclear how many report for all months of the year. From FBI reports, the extent of missing data in submissions by participating agencies is unclear. If there is a substantial amount of missing data, then the undercoverage problem is more severe than indicated above.

DATA ACCESS

The FBI distributes UCR data in a number of reports and in machine-readable form through the National Archive of Criminal Justice Data at the Inter-University Consortium for Political and Social Research (ICPSR) and can be found at http://www.icpsr.umich.edu/NACJD/ucr.html. The UCR age-specific data on arrested persons has been available in a county-level file from public archives in machine-readable form since 1985. These files are routinely available 24 months after the end of the reference year.

PUBLICATIONS

In addition, both the FBI and the state programs prepare special reports and data tapes upon request.

CONTACTS