Research

Abstract: Science, Technology, Engineering, and Mathematics (STEM) are key drivers of a fast-changing world and provide skills to solve challenges in such an environment. During the high school-to-college transition, significant racial-ethnic disparities emerge in adolescents’ STEM pathways, but the reasons for this divergence remain unclear. This study explores two key concepts: critical experience, defined as in-school educational experiences that shape adolescents’ STEM academic pathways, and critical place, defined as out-of-school neighborhood-based institutional resources that support STEM learning. We investigates how critical experiences and places during high school are associated with adolescents’ postsecondary STEM education pursuits, and whether STEM racial-ethnic minority status (i.e., Black and Latinx) moderates these associations. Using the Educational Longitudinal Study of 2002 data, tracking15,370 sophomores until 2012, multilevel models addressing selection into postsecondary education and omitted variable bias suggest adolescents’ participation in more advanced-level courses, higher standardized math test scores, and GPAs in math and science facilitate their postsecondary STEM pursuits. Furthermore, minority status moderated associations between adolescents’ postsecondary STEM pursuits and critical experiences and places. Advanced math and science coursework may help narrow racial-ethnic gaps in STEM pursuits, though some experiences can widen them for students stronger in math than English. With regard to critical places, we found local minority- and immigrant-serving NPOs, unemployment rates, and the concentration of STEM and high-skilled jobs having pronounced associations with racial-ethnic minority adolescents’ postsecondary STEM education. The findings contribute to a deeper understanding of strategies to promote equity in STEM education and the participation of underrepresented adolescents in STEM fields.

Abstract: What are the social determinants of the socioemotional skills that guide economic decisions? I examine how neighborhood contexts shape adolescents’ intertemporal preferences over schooling decisions. When choosing to finish or leave high school, adolescents weigh immediate earnings against delayed returns to a diploma, discounted at their individual rate. To estimate these discount rates, I infer ex-ante incomes if they were to drop out or to graduate using structural equations of human capital investment. The Restricted-use National Longitudinal Survey of Youth 1997 reveals three main findings: first, high-school dropouts are more patient than graduates (measured by dropouts having lower average discount rates); second, dropouts anticipate higher income in both the dropout and completion states than comparable graduates; third, greater institutional resources and higher shares of high-skill employment promote adolescent patience, whereas high unemployment and concentrations of low-skill work increase discounting.

Abstract: This project examines how policy-induced changes in neighborhood institutional resources affect adolescents’ academic and socioemotional outcomes and racial-ethnic disparities. The Supplemental Educational Services (SES) provision of the No Child Left Behind Act expanded the role of local nonprofit and community organizations in providing out-of-school academic support, generating differential growth in youth-serving institutional capacity across areas. I combine individual-level data from the National Longitudinal Survey of Youth 1997 with newly constructed measures of local institutional resources from administrative sources. To identify causal effects, I exploit cross-area variation in post-NCLB resource expansion using a modern difference-in-differences framework that accounts for staggered treatment timing and heterogeneous treatment effects, avoiding biases in traditional two-way fixed effects estimators. The analysis examines impacts on academic achievement, educational attainment, and socioemotional skills related to self-regulation and educational expectations, providing evidence on how place-based institutional investments shape human capital development.

Oh, S. H. & Mitnik, P. Validating the Bayesian Improved First Name and Surname Geocoding (BIFSG) method Bayesian Improved First Name and Surname Geocoding (BIFSG) method with economic data: Sub-state level analyses

Mitnik, P. * & Oh, S. H.* Validating the Bayesian Improved First Name and Surname Geocoding (BIFSG) method Bayesian Improved First Name and Surname Geocoding (BIFSG) method with economic data: National and state level analyses with different populations

Mitnik, P.* & Oh, S. H.* Extensions of the Bayesian Improved First Name and Surname Geocoding (BIFSG) method: Using income and marital status information.

* These authors contributed equally to this work.

Oh, S. H., Tang, J., & Mitnik, P. reprobs: Bayesian Improved First Name and Surname Geocoding-style probabilistic imputation of racial/ethnic group affiliation.

OECD (2022), Staff teams in early childhood education and care centres, OECD Education Policy Perspectives, No. 53, OECD Publishing, Paris, https://doi.org/10.1787/2b913691-en.