The impact of postsecondary education on state-level economic and social characteristics has been widely considered. Of interest has been its role in attracting non-residents and retaining in-state students. However, the current body of research on student residency, migration patterns, and their impact on states’ “brain drain/gain” has substantive and methodological flaws. First, student residency and migration data is not readily available and is limited to state and institutional level counts. The data’s structures are often misaligned, and misused as a crude proxy of intra- and inter-state migration of students and graduates. For example, student A from state i may elect to attend a postsecondary institution located in state j. If student A remains in state j for employment or graduate school then state j’s postsecondary education system can be viewed a stimulus for state-level brain gain. Conversely, if student A returns to state i or moves to another state, then state j’s postsecondary system should only be viewed of as sustaining temporary immigration. The failure to capture a student’s transition after degree completion hinders conclusions on brain gain/drain policies.

Despite these concerns, scholars have analyzed the impact of state-level educational policies such as merit-based scholarship programs (i.e. Georgia’s HOPE Scholarship) on student retention and migration patterns. Findings from these studies have concluded that these policies encourage residents to remain within the state for postsecondary education. These conclusions attempt to relate these patterns to a larger conversation about brain gain/drain dynamics; however these studies are limited to a student’s time in college only. The lack of consideration of post-graduation outcomes (i.e. graduate school or employment) prevents a more realistic understanding of the issue at hand.

This study is the first to account for migration and residency patterns using individual level data coming from two nationally representative surveys of college students over the past two decades: The National Education Longitudinal Study 1988:00 and the Education Longitudinal Study 2002:12. Geostatistics will be used to account for geographic movement and spatial dependence among a state’s economic and educational conditions. Specifically, we model the mobility patterns of student transitions from high school to college and from college to a variety of post-graduation outcomes (i.e. graduate school and employment). Borrowing from population ecology, this study adapts a matrix population model design to examine a state’s population dynamically by accounting for the role of its postsecondary education system. The goal of this paper is to analyze student decisions and state dynamics that retain residents and attract non-residents during and after degree completion. More specifically, this study uses Bayesian inference to explore a student’s “state” in “space” and incorporates state and individual level variables to draw inferences about postsecondary education’s impact on states being importers (brain gain) or exporters (brain drain). Geographical social network analysis will be used to provide graphical and statistical support for further analysis. The study’s implications can provide stronger evidence for policy makers about their state’s post-graduation outcomes, diffusion of residents, and effectiveness of current educational policies.