The Causal Effect of Family Income on College Entry and College Completion: A Feasible Semiparametric Approach for Isolating the Margin

Stephen L. Morgan, Cornell University

Through an analysis of NLSY data, this paper examines whether or not family income shortfalls have long-run deprivation effects and/or short-run credit constraint effects on college entry and college completion patterns. Based on the logic of propensity score analysis, the favorite status attainment variable of educational expectations is used to develop a latent variable model of the propensity to enter and complete college. Using this estimated single dimension as a ranking of individuals, the conditional association between family income and educational transition rates is then analyzed for different strata of that population, some of which can be narrowly represented as students on the margin of college entry. This approach enables an inside-out approach to modeling the causal effect for those students most likely to be subject to the influence of short-run fluctuations in available credit.