Rural and Urban Trends in Family and Intimate Partner Homicide in the United States, 1980-1999 (ICPSR 4115)

Time Period(s) View help for Time Period(s)

Date of Collection View help for Date of Collection

Data Collection Notes View help for Data Collection Notes

Study Purpose View help for Study Purpose

Research has shown that family and intimate partner murders differ from other types of murders in several ways. However, the difference between family and intimate partner and stranger/acquaintance murders as a function of place, specifically the degree of urbanicity or rurality, remained to be explored. This research project examined rural and urban trends in family and intimate partner homicide for the 20-year period from 1980 through 1999. The construct of place served as a backdrop against which changes in trends in family/partner homicide were tracked, and against which various independent measures that purportedly explain variation in the rates were tested. "Place" was operationalized by population and proximity to a metropolitan area. The overall research questions were: (1) How do the rates of family and intimate partner murder differ by place -- specifically, by a place's population size and proximity to a metropolitan area? and (2) What are the independent measures that explain the differences in rates by place, and how do those measures affect changes over time? Finding some support for the hypothesis that family and intimate partner murders are more prevalent in rural than urban communities, an additional question emerged: (3) Did any of these factors worsen over time in rural communities within the time period of interest?

Study Design View help for Study Design

For this study, intimate partners were defined as current and former spouses (including common-law) and current and former boyfriends and girlfriends (including same-sex relationships). Family members were defined as parents, siblings, aunts/uncles, stepparents and stepchildren, in-laws, and "other" family. The project merged data from several sources. The offender data file from the Federal Bureau of Investigation's (FBI) Supplementary Homicide Report (SHR) series for 1980 through 1999 was the primary data source. These agency-level data contain details about murders and non-negligent homicide in the United States, including information about the geographic location and, when known, the relationship between the victim and the perpetrator. Data were weighted to adjust for missing homicide records in the SHR data (as compared to data from the FBI's Offenses Known and Clearances by Arrest series) and for missing victim-offender relationship information. Data for arrests for violent crime (rape, robbery, and assault) and drug- and alcohol-related offenses were obtained from the agency-level FBI Report A Arrest File. Population, population density, and race (and racial segregation) data from the decennial United States Census for 1980, 1990, and 2000 were interpolated for intercensal years. United States Census population data were used to calculate all population-based rates: per 100,000 age 15 and over for intimate partner homicide, and per 100,000 for all ages for all other homicide types. Data on hospitals, educational attainment (persons 25 or older with 12 or more years of education), unemployment, and per capita income were obtained from the 2002 Area Resource File (ARF). The total number of proprietors (farm and non-farm) in the United States by state and county for each year were provided by the Regional Economic Profiles data. The project's population and proximity indicator collapsed the 10-point rural-urban continuum codes (Beale codes) into four categories: metropolitan, nonmetropolitan populations adjacent to a metropolitan area, nonmetropolitan populations not adjacent to a metropolitan area, and rural. The Law Enforcement Agency Identifiers Crosswalk file was used to match the FBI data with the other public use data by matching FBI agency codes with Federal Information Processing Standard (FIPS) codes. Data were aggregated in five-year averages to account for potential instabilities in the annual rates, particularly for the low-population counties.

Sample View help for Sample

Not applicable.

Universe View help for Universe

All counties in the United States from 1980 to 1999

Unit(s) of Observation View help for Unit(s) of Observation

Data Source View help for Data Source

Data Type(s) View help for Data Type(s)

Description of Variables View help for Description of Variables

Data include homicide rates for 1980 through 1999 for intimate partner homicide, family homicide, all other homicide, and all homicide. Variables measuring community socioeconomic distress include annual percent total population change, annual percent population change of individuals ages 18-34, percent poverty, percent unemployment, per capita income (in 1982-1984 dollars), annual percent change in number of establishments, arrests per 100,000 for violent crime other than murder and non-negligent homicide (forcible rape, robbery, or assault), racial segregation (as a dissimilarity index), and percent nonwhite. Additionally, residential overcrowding was measured by persons per housing unit, isolation by persons per square mile, traditionalist views of women and family by the percentage of adults with 12 or more years of education, lack of access to health care by the number of hospitals per 100,000 population, and substance abuse by arrests per 100,000 population for alcohol-related offenses (drunk driving, liquor law offenses, and public drunkenness), and arrests per 100,000 population for drug-related offenses. Five-year averages are included for each of the rates and measures listed above.

Response Rates View help for Response Rates

Not applicable.

Presence of Common Scales View help for Presence of Common Scales

None.

Original Release Date View help for Original Release Date

2005-04-07

Version History View help for Version History