AEASTAT: Administrative Data

Access to Federal Government Administrative Data

The policy relevance of social science research could be increased substantially with better access to administrative records on federal government program participants or those who report to government entities. Administrative data are critical to the evaluation of public programs, and provide the means by which to better understand socio-economic behavior. Advantages of using administrative data over survey data for research include the larger number of observations, inherently longitudinal structures, the higher quality data contained in administrative files, deep coverage of local areas, and opportunity for novel research design. Linkages of administrative data sets with one another and/or with survey data promise a new frontier in public policy evaluation and development. Yet, there is a relative paucity of economic studies utilizing administrative data for American public policy illumination. Most federal micro-level administrative data is very difficult for researchers to obtain. Privacy and disclosure fears are major constraints.

The Sloan Foundation with help from the American Economic Association’s Committee on Government Relations is supporting a project by the Council of Professional Associations on Federal Statistics (COPAFS) to “expand social scientists’ access to administrative data for research in order to improve the bases for sound and informed public policy design and implementation. The focus is on economic research toward the improvement of evidence based public decision making.”

This project has so far produced inventories of websites for Federal Administrative Data and procedures for accessing restricted-use Federal Administrative Data, conducted a survey of administrative data access priorities of a random sample of AEA members, and organized a panel session at the 2015 ASSA annual meetings to present the results of the survey and discuss the use of administrative data. Presenters in this panel review the utility of administrative data and provide examples of analyses enriched by access to those data, but also provide an overview of the constraints faced by federal agencies in providing access. The panel culminates with a presentation of the results and analysis of a survey of AEA and other economic professional associations' members' demand and preferences for various types of federal administrative data

Session: The Use of Administrative Data in Economic Research: Rewards, Risk, and Demand
Session Organizer: Katherine R. Smith - Council of Professional Associations on Federal Statistics
Presiding: John Haltiwanger - University of Maryland
John Friedman - Harvard University "The Value of Public Administrative Data in Policy-Relevant Economic Research"
Amy Finkelstein – Massachusetts Institute of Technology "The Value of Administrative Data for Randomized Evaluations"
Ron Jarmin - U.S. Bureau of the Census "Confidentiality, Privacy Protection and Other Constraints on Stewards of Public Administrative Records"
Katherine R. Smith - Council of Professional Associations on Federal Statistics "Demand and Preferences for Access to Federal Administrative Data: Results of a Survey