Working papers

Academic version - NBER working paper

Abstract - Declining U.S. college enrollments over the past 15 years have triggered questions about the health of the postsecondary sector. Using institution-level data, we make four points. First, such declines are driven not by the four-year sector but by two-year community colleges, which have apparently shrunk by over 30% since the peak of the Great Recession. Second, over one-third of this apparent decline is an artifact of some community colleges being reclassified as offering four-year degrees. Third, pre-Great Recession data shows a 1 percentage point increase in the local unemployment rate increases first-time community college enrollment by 2 percent, suggesting many students are on the margin between community college and job opportunities. For-profit college enrollments are similarly countercyclical, while public and private four-year college enrollments appear acyclical. Our estimates suggest that strengthening labor markets explain about 60% of the post-Great Recession decline in first-time community college enrollment. Fourth, students whose enrollment decisions are most sensitive to labor market conditions appear unlikely to have completed a degree. Though declining community college enrollments are a challenge for postsecondary institutions, it is less clear whether they signal a problem for students on the margin of enrollment.

Media highlights - Chalkbeat, Chronicle of Higher Ed, Inside Higher Ed, Education Next (podcast)

Academic version - Working paper

Abstract - Special education serves more than one in seven U.S. students and constitutes a major public investment in human-capital development, yet its causal impact remains unclear. Using longitudinal administrative data from states X, Y, and Z, we estimate the achievement effect of receiving individualized educational supports with an event-study design that traces within-student changes around the timing of classification. Achievement declines before placement and rises afterward, producing a consistent V-shaped trajectory across states. Three years after classification, test scores are 0.30–0.40σ higher than they would have been absent services. Effects are robust across disability categories, student subgroups, and specifications and are not driven by testing accommodations. Applying the Rambachan–Roth (2023) bounding framework shows these gains persist even under highly conservative assumptions. The results indicate that individualized supports substantially raise learning productivity, yielding large returns to targeted educational investment.

Academic version - NBER Working Paper 20221

Summary version - Education Next

Media highlights - Washington Post, CNN, BBC, Slate, The Atlantic, Vox, Weather Channel, WGBH, Brian Lehrer, WBUR, WNYC