Linking survey data with administrative records will enable the federal government to better gauge the effectiveness of social welfare programs such as food stamps and federal housing assistance, as well as the overall reach of federal antipoverty initiatives, Bruce D. Meyer today told a session audience at the 2015 Joint Statistical Meetings (JSM 2015) in Seattle.
During a presentation titled "Using Linked Survey and Administrative Data to Better Measure Income: Implications for Poverty, Program Effectiveness, and Holes in the Safety Net," Meyer shared the findings of a study he conducted with Nikolas Mittag.
For the study, they combined data from the Current Population Survey (CPS)--the government's source of official poverty and inequality statistics--with New York administrative records for four public assistance programs to better measure the receipt of government transfers and the effect of the programs on reducing the nation's poverty rate. The studied programs are the Supplemental Nutrition Assistance Program (SNAP), more commonly called food stamps; Temporary Assistance for Needy Families (TANF); General Assistance; and Housing and Urban Development Housing Assistance.
Meyer is professor of economics in The University of Chicago's Harris School of Public Policy Studies; Mittag is assistant professor of economics at Charles University in Prague, Czech Republic.
The response rates of household surveys, which are the source of official rates of unemployment, poverty, health insurance coverage, inflation and other statistics that guide public-policy decisions, are in decline. Individuals have become reticent to answer surveys, and those who respond are less likely to answer certain questions or give accurate answers. To combat this worrying trend, federal statistical agencies are looking to administrative data as a way to maintain the quality of their survey products.
To obtain a better measure of government transfers to families in need, Meyer and Mittag linked household survey data from the New York state sample of the 2008-2013 Current Population Survey Annual Social and Economic Supplement with administrative records from the New York State Office of Temporary and Disability Assistance (OTDA) and administrative data on housing assistance programs under the U.S. Department of Housing and Urban Development's purview.
The former contained records of monthly payments from SNAP, TANF and General Assistance for all individuals in the state from 2007 through 2012. The latter contained data on all recipients of federal housing assistance such as addresses, number and ages of family members and rent paid by the tenant from April 2008 to March 2012. Administrative data were matched to the CPS survey data by the U.S. Census Bureau using social security numbers and provided to Meyer and Mittag without personal identifiers. The data were analyzed at a secure data center and the results were screened to prevent disclosure of personal information.
Their analysis identified widespread underreporting of social welfare program payments in the CPS household survey data and a significant understatement of the impact of these programs on reducing the nation's poverty rate. Meyer cited the following examples in his presentation:
The results show using administrative data combined with survey data sharply alters the picture of well-being at the bottom of the income distribution and the effects of transfer programs. "Inaccurate reporting has badly distorted the understanding of policymakers of income distribution, poverty and the effectiveness of federal government antipoverty programs," said Meyer.
"Correcting for misreporting sharply changes key results derived from only using survey data. By linking the survey and administrative data, poverty and inequality are better than officially reported, program effects are larger and fewer individuals have fallen through the public safety net."
Meyer added that although the study is limited to New York state, it is likely the results carry over to the entire country. "This study's findings on the importance of underreporting of SNAP benefits and the understatement of the value of public assistance almost certainly will apply to the rest of the United States," he said.
In making the case for linking administrative and survey data to improve policymaker opinions of poverty and inequality statistics, Meyer said: "The most productive step toward improving the quality of data available for social science research and policy decisions is to increase the availability of administrative data sets and find ways to link these to household survey data."
Cite This Page: