Association for Public Policy Analysis & Management

The 2014 Survey of Income and Program Participation (SIPP) asks questions about two programs administered by the Social Security Administration (SSA): Supplemental Security Income (SSI) and Old-Age, Survivors and Disability Insurance (OASDI or Social Security). Matching SIPP data to SSA administrative records revealed that nearly one-half of respondents who reported SSI receipt in the first wave of the 2014 SIPP panel did not have a corresponding administrative record indicating receipt of SSI payments. Moreover, the linked data suggested respondents reported Social Security benefits when answering questions that were actually about SSI. This “program confusion” between SSI and Social Security likely contributed to the high level of SSI misreports. Program confusion was also identified when respondents reported receiving OASDI but administrative records showed only SSI receipt. The OASDI misreports was a much smaller percentage compared to the SSI misreports.

Comparing administrative records to reported data in the 2014 SIPP allows us to correct for much of this program confusion. However, this correction is only applied to the ‘yes/no’ indicator of whether someone received SSI during the reference year. For people who said ‘yes I received SSI’ but administrative records show they received OASDI, we lose any information they provided beyond the ‘yes.’ Similarly, for people who said ‘yes I received OASDI’ but administrative records show receipt of only SSI we lose information beyond the ‘yes.’ This information can include months of receipt, payment amounts, and reason for receipt. This research examines the amount of reported data that could be retained by moving downstream reported responses that were misreported in the SSI question fields to the corresponding Social Security question fields (and vice versa). Moreover, this research assesses the data quality of the reported responses to downstream questions from the initial ‘yes/no’ indicator where respondents reported receiving SSI when, according to administrative records, they were receiving OASDI (and vice versa). This research is especially important for determining best practices for correcting detailed questions about SSI and OASDI in future SIPP panels when program confusion is identified using administrative records.