Association for Public Policy Analysis & Management

Panel: Using Linked Data to Advance Evidence-Based Policymaking for Public Programs Serving U.S. Families

(Tools of Analysis: Methods, Data, Informatics and Research Design)

This cross-cutting panel presents four papers that utilize linkages to federal data to advance evidence-based policy making for programs serving families and children.  Findings are drawn from demonstration projects linking local data to federal data sources as a strategy for providing new evidence on public programs addressing education, health, poverty, and criminal justice.

 

Government policymakers and program managers at the Federal, state, and local levels are under increasing pressure to ensure that programs achieve better results at lower cost.  In many public programs, decision makers lack access to information about program outcomes and to evaluations that can inform policy decisions and strategies for improvement.  Currently, there is not sufficient capacity for researchers in government or academia to retrieve the federal data that is needed to rigorously evaluate government programs. One significant barrier is a lack of access to key datasets that can reliably measure important outcomes for program participants.  To address this issue, the U.S. Census Bureau and Chapin Hall at the University of Chicago formed a partnership to demonstrate innovative strategies for linking data across programs and levels of government to advance evidence-based policymaking. The goals of the project were to gauge demand from researchers and state and local governments for linking their data to the Census Bureau infrastructure; demonstrate efficient ways to link local data to Census-held data; create compelling use cases for strengthening the Census Bureau linkage infrastructure to serve multiple levels of government; and inform strategies for facilitating data linkage across programs.  

 

To achieve these goals, pilot projects were selected through a competitive proposal process and began implementation in 2016. This panel will feature four of the selected pilots which cover a diverse range of policy topics, research methodologies, and local data sources. Paper one uses a regression discontinuity design to assess outcomes of state public aid for post-secondary education focusing on success in college and returns to college.  Paper two uses state birth records and exploits time and geographic variation to examine returns of policy-driven early health investments in infant and perinatal health on adult health and achievement. Paper three uses a quasi-experimental design to identify impacts of eviction for individuals and families, including their future earnings, use of public assistance, neighborhood choice, housing stability, and school outcomes for children. Paper four uses linked state agency data on juvenile justice, criminal justice and recidivism to identify patterns of service utilization among formerly criminalized youth and correlations to recidivism, employment, income, health, and housing stability.

 

Taken together, the papers demonstrate the important gains from linking existing state and local data at the individual level to restricted federal data sources held by the Census Bureau presents as well highlighting challenges with existing mechanisms for executing this type of research, and will facilitate a robust discussion on key strategies for advancing data linkage infrastructures for evidence-based policy making in the current era.