Taylor, Ralph, Elizabeth Groff, and David Elesh. Forecasting Municipality Crime Counts in the Philadelphia [Pennsylvania] Metropolitan Area, 2000-2008. ICPSR35319-v1. [distributor], 2017-06-26. https://doi.org/10.3886/ICPSR35319.v1
Persistent URL: https://doi.org/10.3886/ICPSR35319.v1
- RIS (generic format for RefWorks, EndNote, etc.)
- EndNote XML (EndNote X4.0.1 or higher)
crime control policies,
crime control programs,
ecology of crime,
metropolitan statistical areas
Smallest Geographic Unit:
Date of Collection:
Unit of Observation:
All jurisdictions in Philadelphia [Pennsylvania] metro area between 2000 and 2008
geographic information system (GIS) data
Data Collection Notes:
These data are part of NACJD's Fast Track Release and are distributed as they were received from the data depositor. The files have been zipped by NACJD for release, but not checked or processed except for the removal of direct identifiers. Users should refer to the accompanying readme file for a brief description of the files available with this collection and consult the investigator(s) if further information is needed.
The purpose of this study was to examine four types of questions: the size and direction of the impacts of community fabric and law enforcement coverage on crime; the spatial patterning of crime and crime links, the spatiotemporal patterning of crime and crime links; and the predictability of one or three-year-look-ahead crime rates using the available variables.
This study aggregates jurisdiction level crime data in order to compare it to numerous findings about socioeconomic status, residential stability, and race composition factors.
The data compiled for this study began with the most basic form of Unified Crime Reporting (UCR) data, obtained directly from the FBI: Return A. Pennsylvania State Police data were derived from the Pennsylvania UCR webpage. New Jersey data were taken from the Crime Reports and Statistics webpage of the New Jersey State Police. In addition, not all police agencies provide 12 months of crime data to the FBI. Counts were not adjusted for missing months. GIS shape files have been provided to show jurisdiction areas.
For the permeability data, a variety of publicly available indicators were collected which the researchers agreed represented the four permeability domains. Internally consistent indices were created tapping three of the domains of interest. In constructing each index, individual variables were z scored and then averaged. Scores on indices were mapped to verify that they provided geographically differentiated scores.
Demographic indices were contructed form individual variables. A population weighted percentile (PWP) form of an indicator or index captures the position of the municipal or civil division (MCD), relative to the entire population in the rest of the Metropolitan Statistical Area (MSA), in that year. Each PWP equals the percent of the population, at the MCD level, in the entire MSA, with scores equal to or less than the PWP of the target MCD.
The UCR data used are fixed length records with monthly counts, by crime category, of unfounded offenses, actual offenses, total offenses cleared by arrest, and juvenile arrests. The most common coverage style was that an MCD maintained its own police department, responsible for the population of the MCD. During the period 276 out of 355 MCDs had this arrangement (76.6 percent). Both Pennsylvania and New Jersey have state police agencies which provide policing coverage and prepare state-level reports. Because each is a distinct governmental entity, each has the ability to organize its data as it sees fit. This has significant implications. Pennsylvania State Police data were derived from the Pennsylvania Uniform Crime Reporting webpage. New Jersey data are available from the Crime Reports and Statistics webpage of the New Jersey State Police.
Data issues arose from the varied nature of policing arrangements at the MCD level in the Philadelphia MSA.
Except for weighting applied to tables describing demographic factors, all MCDs were treated equally, regardless of their size. The focus is on jurisdiction dynamics within the metro area, not description of one relationship for the entire metro area.
Mode of Data Collection:
Description of Variables:
The study includes four data sets and two geographic base shape files.
- The agency_crime_structure_final dataset (38 variables and 3,195 cases), consists of geographic variables such as county name, jurisdiction name, area key. Additionally, it contains numerous crime variables, for instance, spatially lagged property crime counts, reported property crimes, and reported violent crimes. The file also includes variables about police, such as the ratio of police employees per 1000 people, and natural log 1 plus the number of sworn in police officers.
- The demographics_final data (37 variables and 3,195 cases) consists of variables pertaining to characteristics of homes, such as total housing units, median household income, occupied housing units total, median gross rent, and percent multi-person households. It also contains demographics variables, for example, percent Asian population, percent black population, and percent population aged 50-54.
- The ID_all_final dataset (10 variables and 355 cases) contains location ID variables such as PA or NJ, MCD polygon number, and others.
- The permeability_final dataset (106 variables and 355 cases) contains transportation variables such as dead ends, bus stops, external barriers, highway miles, rail and subway stops, and collector road miles. The file also includes yearly crime variables, for example, spatially lagged violent crimes from 2000 to 2008, and spatially lagged property crime from 2000 to 2008.
- In addition, GIS shape files have been provided to show jurisdiction areas in the Philadelphia metro area.
The MCD shape file contains the boundaries of all of the municipal civil division units (n = 355) within the Philadelphia region. The Phila_metro_counties shape file contains the boundaries of the nine counties in the Philadelphia region.