The views expressed here are those of the authors and do not represent the official position of the Social Security Administration or any other federal agency.

Using Social Security Administration administrative records matched to the 1993 and 1996 SIPP panels, we assess discrepancies in reports of benefit receipt and benefit amounts for four sample months. The Supplemental Security Income (SSI) “payment” records are excellent sources of “true” Federally administered SSI payments. The Old-Age, Survivors, and Disability Insurance (OASDI) “payment eligibility” records used in this analysis are imperfect proxies for “true” OASDI payments, but we provide evidence suggesting that over half of the observed absolute error is attributable to the SIPP, and the true SIPP error may even be larger than the discrepancy we observe. We present evidence of confusion between the OASDI and SSI as sources of payments, and we observe a higher incidence of reporting errors for SSI than for OASDI. After wave 1 of the 1993 SIPP, survey responses tend to reflect OASDI benefits net of the Medicare Supplementary Medical Insurance (SMI) premium. While this is consistent with revised questionnaire wording, the use of survey data without adjusting for the SMI premium could substantially bias SIPP-based estimates of total income and poverty status. We find that reporting errors for both SSI and OASDI differ dramatically by imputation status and provide evidence that errors may be systematically related to sample attrition and interview status (self, proxy, and refusal). Finally, we provide a brief assessment of the effect of the lack of social security numbers in a nontrivial fraction of cases and find clear evidence of selectivity.