With funding from the Rockefeller Foundation, the Urban Institute and GeoLytics, Inc., a private firm specializing in the development of demographic and geographic data products, have produced a new data resource that will enable policy makers, community organizations, and researchers to examine and analyze changes that have occurred in U.S. neighborhoods over the past three decades. The Neighborhood Change Database (NCDB) combines tract-level data from the 1970 - 2000 decennial censuses into one easy-to-use product. It is literally the only source of census data with variables and tract boundaries that are consistently defined across census years.
History
The NCDB builds upon a previous Urban Institute effort called the Under Class Database (UDB), which was created in 1989 by Isabel Sawhill and Erol Ricketts, also with the support of the Rockefeller Foundation. The UDB contained social, demographic, economic, and housing data for census tracts in the United States from the 1980 census. It was later expanded under the supervision of Ronald Mincy and Susan Wiener to include 1970 and 1990 data.
The NCDB combines data from the UDB with new information from Census 2000. A standard set of indicators is provided for each of the 65,232 census tracts in the U.S. Census tracts are locally-determined geographic units, ranging in size from 2,500 to 8,000 persons. Tracts are meant to approximate "neighborhoods" by capturing a group of residents with similar population characteristics, economic status, and living conditions. Tracts can be used by themselves as units of analysis, or as the building blocks to create larger neighborhood areas.
Ease of Use
Although it is possible to download much of the Census 2000 data for free from the Census Bureau's web site, the NCDB provides a valuable service by lowering the barriers to accessing and using these data for many users. For instance, to obtain the complete summary tabulations from the census "short form," it would be necessary to download several hundred megabytes of data in up to 40 separate files for each state. These individual files would then have to be read using database or data analysis software and combined with each other to create a complete set of indicators for a particular area. Even if one has the technical capabilities and skills to accomplish all of this, it still would be a very time-consuming task.
In addition, there are a number of challenges to using historical census data that are addressed by the NCDB. First, census tract-level data from 1970 and 1980 are no longer obtainable directly from the Census Bureau. The NCDB is one of the few sources for this information available anywhere. Second, because of changes made to questionnaires and summary tabulations from census to census, it can be very challenging to attempt to define the same indicators consistently across census years. The NCDB is comprised of a set of standard, uniformly named data elements that will make it easy for users to track the same indicators over time.
Geographic Changes
Along with changes in data definitions there are also changes in census tract and other geographical boundaries that must be taken into account. As was noted earlier, census tracts are meant to be fairly homogenous areas. As population changes occur over time, tracts get redefined so that a reasonably uniform distribution of people in tracts can be maintained. If, for instance, a tract grows in population between censuses, it may be split into two or more separate tracts in later years. Or, several existing tracts may be reconfigured into new tracts for subsequent censuses. This problem is not insignificant. An analysis of 1990 and 2000 census tract boundaries showed that about 46 percent of all tracts were redefined between these two census years.
To make accurate comparisons of neighborhoods over time, one must adjust for the way tract boundaries get redefined. Using GIS technology and taking advantage of GeoLytics' access to geographic boundary files for previous censuses, a methodology has been developed to "remap" earlier data to a standardized set of 2000 census tract boundaries. As a result, the NCDB provides data which makes it possible to compare neighborhood conditions consistently across time. Furthermore, each tract is identified according to the city, county, and metropolitan area in which it is located. Like tracts, the definitions of these areas also change from time to time, making comparisons across years difficult. The NCDB therefore allows consistent analysis of data for these larger geographies, too.
Potential Uses
The NCDB has been designed to be easy to use by a variety of audiences. NCDB data products are being released on CD-ROM using GeoLytics' proprietary data compression and mapping technology. The data can be accessed using the menu-driven, user-friendly mapping and analysis software included on the same CD-ROM, or the data can be extracted for use in external database, mapping, and analysis packages.
It is envisioned that the NCDB will serve any number of useful purposes for those interested in the well-being of neighborhoods and communities. For instance, it can be used to measure changes in the concentration of poverty in neighborhoods over time, along with other measures of family risk, such as single-parent families and unemployment. The NCDB can also be used to study patterns of racial and ethnic segregation in cities and metropolitan areas. Or, it can be a source of data for neighborhood "report cards" or to help identify areas for service delivery or targeted interventions.
Release Schedule
The NCDB was released in two stages, corresponding to the release of Census 2000 short and long form data. The Short Form NCDB contained only the basic population and housing characteristics from the short form questions and was replaced by the Long Form NCDB in October 2003. The Long Form NCDB contains data from the 1970, 1980, 1990 and the 2000 censuses and is based on the "long form" questions that were asked of about one out of every six households in the census. These data include additional detailed population, household, and housing characteristics, including income, poverty status, education level, employment, housing costs, immigration, and other variables of interest for research, social policy, and community building. Along with other normalized census data sets, the NCDB can be purchased via the GeoLytics web site, http://www.geolytics.com.
For More Information
The complete NCDB Data Users Guide is available here in PDF.
Appendices:
Appendix A: State Codes
Appendix B: County Codes
Appendix C: Metropolitan Area Codes
Appendix D: Aggregation Error for
New England Metro Areas and for Places
Appendix E: Data Dictionary
Appendix F: Census Source
Tabulation Matrices - 1970
Appendix G: Census Source
Tabulation Matrices - 1980
Appendix H: Census Source
Tabulation Matrices - 1990
Appendix I: Census Source
Tabulation Matrices - 2000
Appendix J: Description of
Tract Remapping Methodology
The NCDB project is co-directed by Peter Tatian and Tom Kingsley at the Urban Institute. For more information, please contact them at ncdb@ui.urban.org.
To purchase the NCDB, go to the Normalized Data section of the Geolytics web site at http://www.geolytics.com.
For additional information, e-mail NNIP at nnip@ui.urban.org