Abstract:

Human capital theory advocates that a student's college choice decision is based on the expected net benefits of attending a specific Higher Education institution (HEI). Apparently, the higher the cost of attending a given higher education institution, the higher the quality of the education provided and the higher the expectations regarding future labor market outcomes. This seems to be the case in the US, but this might not be the case in most European countries.

The purpose of this paper is twofold. First, it represents an attempt to complement the existing analysis of HEI choice by explicitly modelling two choices together: university or polytechnic and public or private institution. Second, it also aims at providing an updated look of the effect of the spatial distribution of HEI on the choice of prospective students.

This study brings in some knowledge that can be used to characterize the student profile in each higher education sector (public or private), as well as provide some insights into the degree to which some types of students select themselves to private higher education institutions. In particular, it allows for testing whether lower quality students self-select to private institutions.

It also aims at providing an updated look of the effect of the spatial distribution of HEI on the choice of prospective students. Such higher education student expansion has resulted in a new group of students: namely, low-income students, with access to higher education, but who might be more geographically constrained in their choices than other students. On the other hand, HEI are bundles of education, leisure, social networks, and (in some cases) religious education. When choosing a given HEI, the student gets a package, including not only the institution itself and its features, such as its quality, but also the attributes of the higher education institution location: namely, the local supply of leisure activities and the local housing market.

Thus, depending on the spatial distribution and accessibility of higher education institutions, students may decide to attend higher education in their home region, or instead move to another region. Price differentiation may also have an influence on student decisions on whether to move or not.

We pursue these objectives using the example of Portuguese higher education students. We explicitly model student choices of HEI type (university versus polytechnic institute and private versus public institution) based on socioeconomic background, spatial distribution of higher education institutions and student geographic mobility. We estimate a bivariate probit model using a comprehensive data set on all first time/first year students entering higher education in a specific academic year.

The empirical strategy involves the model estimation by maximum likelihood with robust standard errors. In order to justify the use of a bivariate probit model (rather than two independent and separately estimated probit models), a test for the absence of correlation is then performed.

Conclusion point out to higher income students prefer university over polytechnic institutes, but they favour private institutions. The value students put on institutional attributes like prestige, location, employability and leisure offer at the institution location play an effective role in the institution choice. Geographic accessibility to each type of institution favors the choice of that type of institution.

Keywords:

Higher education choice, university, polytechnic institute, public, private, accessibility.