Harnessing the Power of Administrative Data for Research and Policy Insights
Academic and government collaborations can develop sustainable protocols for long-term partnerships.
by Day Manoli | June 5, 2018
Government agencies collect administrative data for a variety of programs. Partnering with academic researchers to analyze these data can lead to valuable human capital development and policy insights. As an academic economist, I have collaborated with coauthors at the United States Treasury and Internal Revenue Service to study federal income tax administration and tax policy. Following are four examples in which I learned valuable institutional background, provided data analysis skills, and developed academic research and policy-relevant conclusions.
Example 1: Guyton et al (2017) use administrative tax data to identify potentially EITC-eligible individuals who did not file federal income tax returns and then implement a randomized controlled trial to test whether postcard reminders could increase tax filing and EITC participation. The results demonstrated that reminders successfully increased EITC participation and filing of current and past tax returns, and some individuals paid taxes owed to the IRS.
Example 2: Manoli and Turner (2018) use administrative tax data to examine the impacts of tax refunds received in the spring of the high school senior year on college enrollment. The results suggest tax refunds received in the spring of the high school senior year have meaningful effects on college enrollment for students from low-income households.
Example 3: Guyton et al (2018) estimate the impacts of EITC correspondence audits on taxpayer behavior by developing a research design based on random variation within part of the IRS’ audit selection procedures. The results indicate that, in the years after the audits, there are significant decreases in EITC participation, and this is primarily driven by decreases in the likelihood of filing tax returns. Furthermore, self-employed taxpayers increase their likelihood of having wage employment, and wage earner taxpayers decrease their likelihood of having wage employment. Further research may examine strategies to increase response rates to audits and strategies for post-audit follow-up communications.
Example 4: Manoli, Michaelides and Patel (2018) use administrative tax data to examine the long-term effects of an experimental job-search assistance program. The program required randomly-selected unemployed workers who had just started collecting unemployment insurance (UI) benefits to undergo an eligibility review and receive personalized job-counseling services. The analysis demonstrates that the program led to increases in employment, earnings, tax filing, and home ownership.
In conclusion, rather than focusing on producing a single paper or study, successful partnerships between academic researchers and government agencies may focus instead on respecting each party's interests and developing sustainable protocols for longer-term partnerships.
This policy brief was prepared for the UC Center, Sacramento Speaker Series. Correspondence related to it should be addressed to Day Manoli at dsmanoli@austin.utexas.edu.
To learn more about ESSPRI, visit esspri.uci.edu.
