Workforce-Focused Financial Aid and Career Outcomes: Evidence from the National Defense Student Loan Program

• In Event: Aligning Dollars with Demand: Effects of Workforce-Aligned Financial Aid on Postsecondary and Labor Market Outcomes

Abstract

Background: The National Defense Student Loan (NDSL) Program, created under Title II of the 1958 National Defense Education Act, was the first federal college loan program. A response to Sputnik, it aimed to expand access to STEM and foreign language (FL) education. NDSL provided low-income students up to $1,000 per year, covering about two-thirds of public college tuition. From 1959 to 1964, 600,000 students borrowed $443 million, with Black students receiving loans at higher rates than White students. By 1963, 90% of full-time students attended participating institutions, with 5% borrowing through NDSL. Though historical data suggests NDSL increased college access and B.A. attainment, its impact has not been studied using modern causal inference. We aim to examine its effects on educational attainment, college majors, and labor market outcomes, exploring differences across student subgroups. 

Data Collection and Analysis: We digitize historical records to compile state-level NDSL funding and high school graduate data from the 1950s-60s, merging it with ACS (2005-2022), Census (1970-2000), and HRS (1992-2010) data. The final dataset includes NDSL funding, birth year, state of birth, gender, race/ethnicity, childhood SES, educational attainment, college majors, and occupations. Our analysis leverages two sources of variation. First, NDSL funding was allocated based on each state’s share of national college enrollment, giving high school graduates in high-enrollment states more access to funds. We use a difference-in-differences (DID) design to estimate:

Yist​=β+δNDSLs​×Exposureist​+γs​+λt​+ϕw​+ϵist​

where NDSLsNDSL_s​ is per-high school graduate state-level federal contributions (in $1,000), and ExposureistExposure_{ist}​ measures the fraction of a student's college years exposed to NDSL funding (ranging from 0 to 1). Fixed effects for state, cohort, and survey wave control for unobserved heterogeneity. Under standard DID assumptions, δ\delta captures the causal effect of NDSL funding.

Second, we explore variation in childhood SES, hypothesizing that students from lower-SES backgrounds benefited more. We re-estimate the model using a continuous SES measure, replacing state fixed effects with SES fixed effects. Analyses are conducted for all students and subgroups by race and gender.

Preliminary Findings: We find suggestive effects on college enrollment and B.A. completion, particularly in FL majors among women, but inconclusive effects on STEM and education majors or careers. Childhood SES variation does not show clear effects on educational attainment. This is the first causal study of NDSL, the first federal college loan program. Our findings suggest NDSL increased college access but did not significantly reduce SES gaps. While FL majors rose, effects on STEM and career outcomes remain mixed.

Authors

• Joshua Goodman, Boston University
  Non-Presenting Co-Author

• Tadeja Gracner, University of Southern California
  Non-Presenting Co-Author

• Mingyan Ma, Boston University
  Presenting Author

• Christine Mulhern, RAND Corporation
  Non-Presenting Co-Author