Linking Data to Study the Effects of Social Security and Means-Tested Transfers

November 7, 2018

In 2017, RSF grantees Bruce Meyer (University of Chicago) and James Sullivan (University of Notre Dame) constructed the Comprehensive Income Dataset, linking survey, tax and administrative program data in order to measure more accurately income for households in the U.S. Income data used by social scientists is drawn from many sources—including census surveys, tax records, and other administrative data—which each have their strengths and weaknesses. For instance, while the Current Population Survey (CPS) and American Community Survey (ACS) contain important demographic information, certain income components (such as government benefits, self-employment income, interest, dividends, rents, and royalties) are underreported, especially for individuals in the lower and upper tails of the distribution. On the other hand, tax data are more accurate than self-reported surveys, but lack demographic detail and information on non-taxable income such as food stamps and the earned income tax credit. 

In a new article for the anniversary issue of the ILR Review, Meyer and coauthor Derek Wu (University of Chicago) discuss how they used the Comprehensive Income Dataset to study the effects of key social safety net programs on poverty. They linked data from the Survey of Income and Program Participation (SIPP)—a longitudinal household survey on income and participation in benefits programs—with administrative data on Social Security and five means-tested programs: SNAP (food stamps), Temporary Assistance for Needy Families (TANF), Supplemental Security Income (SSI), housing benefits, and the Earned Income Tax Credit (EITC). By linking this data, they were able to show that in the period between 2008 and 2013, Social Security cut the poverty rate by a third—more than twice the combined effect of the five means-tested transfers. And among the means-tested transfers, the EITC and SNAP had the largest anti-poverty effects.

As the authors note, combining different sources of data produces a more accurate measure of income and program participation for researchers and policymakers. Their ILR article is the one of the first steps of the larger Comprehensive Income Dataset project that seeks to improve income measurement through the linkage of administrative data to household surveys.

Read the article in full.



RSF: The Russell Sage Foundation Journal of the Social Sciences is a peer-reviewed, open-access journal of original empirical research articles by both established and emerging scholars.


The Russell Sage Foundation offers grants and positions in our Visiting Scholars program for research.


Join our mailing list for email updates.