Individual Submission Summary
Share...

Direct link:

Testing What Works in TANF: Experimental Evidence from Seven Tennessee Opportunity Pilots

Friday, November 6, 8:30 to 10:00am, Property: Boston Marriott Copley Place, Floor: 3rd Floor, Room: Boston University

Abstract

Introduction and Background
As the federal government's role in social policy shrinks, states bear increasing responsibility for how they invest Temporary Assistance for Needy Families (TANF) block grant funds to reduce poverty and promote upward mobility. Tennessee invested $175 million in the Tennessee Opportunity Pilot Initiative (TOPI), funding seven community-based pilots designed to support economic mobility and wellbeing among families with low incomes. Tennessee paired this large-scale investment with embedded randomized evaluations across all seven pilots, creating a rare opportunity to test what works within a single state policy and funding framework.

Purpose and Research Questions
A central question in TANF policy is whether flexible block grant funds can do more than maintain a basic safety net. We synthesize findings across the seven pilots to address three cross-cutting questions: (1) To what extent do comprehensive, coordinated service models increase employment and earnings? (2) To what extent do enhanced program components, such as financial incentives, structured parenting curricula, or shared data systems, produce measurable gains beyond core services? (3) To what extent do these models influence public benefit participation?

Data
We draw on baseline and follow-up survey data, program management information systems, and linked administrative records from Tennessee’s longitudinal P20 Connect data system. Administrative data include quarterly employment and earnings from unemployment insurance records, SNAP and TANF participation and benefit amounts, and education enrollment and credential attainment. Follow-up periods range from four to 18 months post-random assignment. Study samples range from 600 to 1,600 participants per pilot.  

Research Design and Methods
We implemented randomized controlled trials in all seven pilots: differential RCTs in four sites to examine the added value of specific program components and delayed-intervention RCTs in three sites to examine the impact of full-service models. We pre-specified confirmatory, secondary, and exploratory outcomes to limit multiple testing concerns. We estimated impacts using regression-adjusted models that controlled for baseline characteristics.

Results
We will present regression-adjusted impact estimates for each pilot’s confirmatory and secondary outcomes. Where feasible, we will present cross-site comparisons examining how effects vary by service model type and follow-up period. We will integrate implementation and cost evidence to contextualize variation across pilots and assess scalability. Analyses are underway and will be complete before the conference. By examining multiple randomized trials within a single statewide initiative, this study offers rare insight into how different approaches perform under a common policy and funding framework.

Conclusions and Policy Implications
Few studies have tested multiple service models head-to-head within a single state TANF framework. This study addresses that gap, providing evidence on whether enhanced components produce gains beyond core services and whether coordinated models generate short-term returns that justify their cost. As federal and state leaders revisit how TANF funds should support workforce participation, family stability, and poverty reduction, this study provides timely evidence on what coordinated service models can and cannot achieve in the short term. Policymakers should consider both the promise and limits of comprehensive service models, particularly when designing scalable strategies within constrained public budgets and shifting federal priorities.

Author