NB19-01: Disability Benefit Receipt by State and County, 1970-2017: Description and Dataset

Research using geographic variation in the receipt of federal disability benefits has been hampered by the lack of readily available longitudinal data. SSA publishes OASDI Beneficiaries by State and County and SSI Recipients by State and County each year, but these are underutilized because they are in paper form until 1998 and subsequently as online state-specific PDF/Excel files. We have collected and digitized these publications back to 1970. This project will make use of this digitized information, along with more recent data, to create a dataset measuring the percent of DI beneficiaries and SSI recipients in the working-age population for each county and year from 1970 to 2017. DI/SSI payment information will also be included. An article will be written to showcase the dataset and discuss key insights and puzzles that arise from looking at disability incidence across counties over a long period of time. The dataset and a codebook will be made freely available (e.g., via the SSA and/or NBER websites). Project steps and deliverables include:
• Data will be checked for accuracy, and variables and county codes made consistent over time;
• Detailed population estimates, compiled by the Surveillance, Epidemiology, and End Results (SEER) program of the National Cancer Institute, will be merged in and used to as the denominator for calculating disability benefit recipiency rates;
• A codebook will be written describing the structure of the dataset and variables;
• Data will be analyzed to examine characteristics such as: the counties/regions with the highest/lowest DI, SSI, and disability benefit rates in different periods (e.g., 1970s, 1990s, 2010s); the persistence of disability benefit rates within counties; the correlation between DI and SSI incidence rates within counties; and the regions and characteristics of counties with high/low disability benefit rates;
• An article will be written describing the most important and interesting facts that come out of the analysis. It will also describe the dataset. We plan to publish it as a NBER working paper and in a widely accessible policy journal.