Peer-Reviewed Publications

Ma, Mingyan, Joshua Goodman, Christine Mulhern, Tiffany Berglund, and Tadeja Gračner. (2026). Data on the Implementation of the 1958 National Defense Student Loan Program in the United States. Scientific Data. https://doi.org/10.1038/s41597-026-07822-y.

• Abstract: We construct a novel dataset documenting the implementation of the National Defense Student Loan (NDSL) Program in the United States. This program began in 1958 and represented the beginning of the U.S. government’s federal student loan program. We digitize measures from historical PDFs of reports produced by the U.S. Department of Health, Education, and Welfare and calculate how exposure to the program varied over time and across states from 1959 to 1963. These data can be used by researchers to understand the short- and long-term impacts of this large and novel program on a variety of individual and state outcomes.

hh

Under Review

^{* Authors listed in alphabetical order.}

Ma, Mingyan, Marcus A. Winters, and Christopher H. Cleveland. (2026). Beginning Teachers’ Self-Reported Challenges: Evidence from Connecticut Induction Narratives. Under Review at Journal of Policy Analysis and Management.

• Abstract: We analyze 66,721 narratives written by beginning teachers in Connecticut’s statewide induction program between 2010–11 and 2023–24, using topic modeling and large language model classification to construct 17 binary challenge indicators spanning classroom environment, planning, and instruction. Linking these measures to administrative records for 14,301 general and special education teachers, we document three findings. First, self-reported challenges concentrate in the classroom environment domain, led by student engagement (63.9%) and behavioral standards (62.5%); planning and instructional challenges are more evenly spread and individually less common. Second, little of this variation is predictable from the information school systems typically collect. Assignment role is the strongest observable predictor; demographics and school context carry little association; and the signals systems use to screen and pay teachers—licensure scores, a master’s degree, and educator preparation program—are essentially uninformative. Third, this unexplained variation carries some information about mobility: beginning teachers who report challenges promoting student engagement or fostering appropriate standards of behavior are less likely to remain in their initial district or school. These results extend text-as-data methods to teacher personnel records, show that conventional screening signals miss what beginning teachers struggle with, and link residual variation to early-career mobility.

Ma, Mingyan, and Abigail Francis. (2026). Full-Day Kindergarten and Enrollment Patterns: Evidence from Massachusetts. Revise and Resubmit at Education Finance and Policy.

• Abstract: Public school districts across the United States face persistent enrollment pressures. Birth rates have declined and school choice has expanded for more than a decade, and enrollment losses deepened sharply during the COVID-19 pandemic. These trends carry direct consequences for funding, staffing, and school viability. This paper examines whether full-day kindergarten (FDK) can serve as a policy lever to attract and retain public school students, using a staggered difference-in-differences design applied to Massachusetts’ FDK expansion. We find FDK implementation increases public kindergarten enrollment by approximately 9%, about 20 additional students per district, a net gain that conversion from half-day programs alone cannot explain, though where these students come from remains an open question our data cannot resolve. Positive enrollment effects persist through grade 3, consistent with FDK building some lasting attachment to public schools rather than a temporary enrollment boost. Beyond its well-documented benefits for children and families, FDK may offer districts one additional tool for stabilizing enrollment, though its value as such depends on weighing these gains against implementation costs.

Chi, Olivia L., and Mingyan Ma. (2026). Paraeducators as Prospective Teachers: Interest, Support, and Challenges. Under Review at Educational Researcher.

• Abstract: Many districts have launched Grow Your Own programs to recruit paraeducators into the teacher workforce. Using survey data from 533 paraeducators in two districts, we examine predictors of teaching interest and barriers to credential completion. We find that district-level professional support significantly predicts both interest in teaching and self-perceived capability of making the transition. Interested paraeducators anticipate benefits related to economic security and professional identity but cite credentialing costs and time constraints as challenges. Findings suggest that effective paraeducator-to-teacher pathways require financial assistance and flexible programming—alongside strong district-level support— to cultivate and convert interest into credentials.

Ma, Mingyan*, and Marcus A. Winters. (2025). Reclassifying English Learners. (EdWorkingPaper: 25-1234). Retrieved from Annenberg Institute at Brown University: https://doi.org/10.26300/ph0k-1t67. Revise and Resubmit at Education Finance and Policy.

• Abstract: Most English learners (ELs) eventually gain enough proficiency to be reclassified and no longer receive language supports. Though well-identified, prior regression discontinuity estimates of reclassification effects are too imprecise to guide policy. Using student fixed effects and Indiana data, we find ELs experience a significant, lasting negative shock after reclassification. We apply insights from a stylized model for changes in EL outcomes as they gain facility with English to bound the causal effect. A difference-in-differences approach making across-student comparisons between treated students and not-yet reclassified ELs yields similar results. Our analysis illustrates the potential for panel data strategies with transparent assumptions to inform the effects of important but understudied educational services when experimental and quasi-experimental strategies are unavailable and/or uninformative.

hh

Work in Progress

Certification at the Margin: Regression Discontinuity Evidence from Connecticut’s Praxis Bridge

Earn-and-Learn into the Classroom: The Labor Market Effects of K-12 Teacher Apprenticeship Programs

Workforce-Focused Financial Aid and Career Outcomes: Evidence from the National Defense Student Loan Program (with Joshua Goodman, Tadeja Gračner, and Christine Mulhern)

Policy Briefs

Ma, Mingyan, and Abigail Francis. (2026). Full-Day Kindergarten Improves Public School Enrollment: Evidence from Massachusetts. Wheelock Educational Policy Center, Policy Brief.

Ma, Mingyan, and Marcus A. Winters. (2024). Reclassifying English Learners. Wheelock Educational Policy Center, Policy Brief.

Huang, Mingxing, Huidi Liu, Mingyan Ma, and Hongmei Yi. (2024). Urban-Rural Disparities in School and Home Reading Resources and Their Impact on Children’s Reading Behaviors and Achievement. The Land of Mobility.

hh