"Long-term Impacts of Youth Mentoring and Therapy: Experimental Evidence from Chicago " with Brandon Domash, Kelly Hallberg, and Cristobal Pinto Poehls (Accepted at Journal of Public Economics)
Abstract: We conduct a large-scale, randomized controlled trial of a six-month intervention combining intensive mentoring and group cognitive behavioral therapy (CBT) for youth in Chicago, following study participants for up to five years. The program was designed to engage young people at higher risk of involvement with the criminal justice system, and successfully did so with a take-up rate of 62%. Participants averaged 177 hours with mentors and nine hours of group CBT. After 24 months, youth offered the program experienced a 19% reduction in the probability of being arrested. Notably, this impact persists into adulthood, as effects remain stable through five years post-randomization. A reduction in arrests for violent offenses is driving this overall arrest reduction, with a 23% reduction in the probability of a violent-crime arrest after 24 months. The program does not impact school enrollment persistence (graduation or being enrolled), but shows suggestive improvements in school behavior and school mobility. We conclude that programs that combine mentoring and CBT can serve as a model for engaging a harder-to-reach population of youth, and can be effective in reducing criminal justice contact in the longer run. (older version of paper can be found here)
Press: Probable Causation Podcast
“Overcoming Distrust: Organizational Brokerage, Social Tie Formation, and System Avoidance in a Community Violence Intervention” with Max Lubell and Kelly Hallberg (Accepted at Social Problems).
Labor Market Attachment and Perceived Barriers to Work Among Homeless Families; with John Eric Humphries, Stephen Stapleton, Winnie van Dijk
AEA Papers and Proceedings, vol. 115, May 2025, (pp. 103–07)
We study labor market attachment among homeless families using baseline survey data from a large-scale study in Illinois. Of homeless parents, 40 percent worked in the past month. Among those not working, most report wanting and actively searching for a job. When asked about barriers to employment, parents overwhelmingly cite childcare constraints, lack of housing, lack of transportation, and family responsibilities. Health problems or skills deficits are mentioned far less frequently. The nature of these barriers—each associated with upfront investments—suggests that homeless families lack the initial financial resources needed to obtain stable employment.
After Citizens United: How outside spending shapes American democracy; with Carlo Prato and Stephane Wolton
Electoral Studies, Volume 67, October 2020, 102190
We study the political consequences of lifting restrictions on the funding of groups engaging in outside spending (e.g., independent political advertising) in elections. Theoretically, we assume that outside spending changes the salience of candidate-specific attributes relative to their party labels. Empirically, we employ a difference-in-differences design that exploits the removal of state-level restrictions on the funding of outside spending mandated by the federal-level rulings in both Citizens United and SpeechNow.org v. FEC. We find strong evidence that these regulatory changes increase the electoral success of Republican candidates, thereby leading to more ideologically conservative legislatures. We find no effect on polarization. Consistent with our theory, the size of our estimated effects depends on the power of labor unions and the alignment of business interests with the Republican party.
Press: Monkey Cage post, Vox.com, The Ezra Klein Show