Choosing colleges: Identifying and modeling choice sets

Although largely ignored in the growing literature about college choice that adopts a random utility modeling framework, there is reason to believe that the estimates are sensitive to the criteria used to define the choice sets. Accordingly, this paper examines the methodological and substantive implications of using merit (class rank) and socioeconomic (high school strata) to define college choice sets. Using Texas as a case study, results show that criteria used to constrain choice sets—and type of high school attend in particular—not only produce quite different post-secondary institutional profiles, but also different estimates of institutional attributes on students’ top choice. These findings have methodological implications for future research about college choice and substantive importance for the prospects that the Texas top 10% law, which guarantees automatic admission to students who graduate in the top decile of their high school class, will equalize access to the state’s competitive public institutions.