NB20-04: Comparing Trends in Late-Life Income Across Data Sets

Our team is currently using Internal Revenue Service (IRS) tax data to track Americans’ income trajectories from older age through death, identify the sources of late-life income, and assess the adequacy of that income. We are then measuring how these income trajectories are changing across birth cohorts. Is retirement income adequacy becoming better or worse across time?
We propose to extend this project to compare results from IRS data to those from widely used surveys— the Health and Retirement Study (HRS), the Current Population Survey (CPS), and the Survey of Income and Program Participation (SIPP). Existing research has demonstrated significant underreporting of retirement income in surveys. Our data grants unique visibility into income trends across cohorts, enabling proper analysis of changing Social Security reliance and detailed evaluation of survey data accuracy. This investigation will add to our understanding of how previous research findings on retirement income adequacy and the role of Social Security in ensuring a basic standard of living, which have largely relied upon survey data, should be interpreted. The results of our analysis will be documented and interpreted in a research paper. Completion of this deliverable will involve three key steps: 1) dataset construction – we will build standardized versions of each dataset, with comparable income measures, units of analysis, and selection criteria, a crucial step for ensuring appropriate comparison across the various datasets we plan to analyze; 2) analysis; 3) manuscript preparation.