High-dosage tutoring (HDT) is one of the most effective interventions to boost student learning, but cost and staffing remain barriers to scaling. We ran an RCT with ~4,000 students in two large districts (2018–19, 2019–20), testing a hybrid tutoring model where half of student time was spent with a tutor and half on computer-assisted learning (CAL). This reduced costs by ~30% relative to 2-to-1 tutoring, while producing math test score gains nearly as large as those in prior HDT models. The findings suggest technology can help expand HDT while preserving effectiveness.

Learn more: NBER working paper · Saga Tech project page

Policy influence & recognition: 2025 Economic Report of the President; BLS Monthly Labor Review; Evidence Based Policy; CommonWealth Beacon; K-12 Dive

We evaluate Saga Education's early-grade literacy tutoring program in CPS and a charter network. Approximately 300 first-grade students were randomized in 2016–17 and 2017–18. Mid-year gains persisted for two years but faded by year 3, with impacts comparable to Project STAR and math-focused HDT studies. Classroom observation data show increases in time-on-task and instructional quality, consistent with scaffolding as a mechanism. Estimates may be conservative given that control group sizes declined as treatment students were pulled from classrooms.

Learn more: Saga Lit project page

This paper examines the long-term and intergenerational effects of Pakistan's 1960s school construction program in rural Punjab. Building an additional girls' school per 1,000 children substantially raised girls' primary completion and increased schooling by ~0.5 years, with no comparable impacts for boys. Maternal exposure improved children's education but did not translate into better labor market or marriage outcomes, underscoring the limits of schooling gains in settings where social norms constrain women's opportunities.

Learn more: Read the paper · 2-minute video · Blog summary at CDPR

This paper examines how labor demand shocks from the Great Recession affected schooling decisions of young cohorts. Using group-specific labor demand shocks at the metropolitan level, I show that cohorts more exposed to the recession increased their schooling investments more than others. The results underscore how changes in opportunity costs can shape human capital investment.

We evaluate Genesys Works, a national workforce development program that offers high school students an 8-week professional skills bootcamp followed by a paid, year-long corporate internship during senior year. Roughly 460 Chicago Public Schools students were randomized via admissions lottery in 2017, and we track them for up to six years after expected graduation using CPS records, National Student Clearinghouse data, and state unemployment insurance wage records. Participation raised senior-year GPA, driven entirely by elective coursework, and substantially reduced course failures, but had little effect on high school graduation or college enrollment and completion. Analyses of longer-term employment and earnings outcomes are ongoing.

Following ~5,300 Chicago students over multiple years to measure the persistence of early high-dosage tutoring gains on later academic and postsecondary outcomes.

RCT of a tech-enabled personalized learning model implemented in high schools, examining effects on math achievement and student engagement.

Using administrative and survey data to study how Pakistan's community health worker program shaped fertility and reproductive health outcomes.