Date of Award

Spring 3-1-2023

Degree Type


Degree Name

Doctor of Philosophy (PhD)


Public Management and Policy

First Advisor

Ross Rubenstein

Second Advisor

Tim Sass

Third Advisor

Audrey Leroux

Fourth Advisor

Diana Hicks

Fifth Advisor

Juan Rogers


The dissertation includes three essays contributing to our understanding of human capital development and student talent allocation. The first essay provides insights into the impact of algorithmic student advising programs, while the second essay highlights the role of higher education agencies in promoting international student mobility. The third essay evaluates the cognitive development trade-offs entailed by technical coursework. The first essay reports results of the Graduation and Progression (GPS) program, the largest deployment of an algorithmic student advising platform based on learning analytics to date. Implemented by Georgia State University in A.Y. 2012-2013, the platform aims to improve student outcomes by tightening the fit between students and majors. The study uses matching techniques to mitigate observational bias between advised students and students who did not get advised, and builds an index of Relative Academic Strength (RAS) based on SAT/ACT math and reading test scores. Results show that the benefits from the program are clustered among higher-ability students who terminated their studies in majors that are closer fits to their inclinations, as measured by the academic aptitude index, except for students entering Humanities majors. Furthermore, students who intended to major in STEM Computation fields took over 15 more credit hours in their relevant subject area when advised by GPS. The second essay aims to evaluate the influence of higher education agencies on international student mobility. It examines the relationship between the 1996-2016 expansion of the German agency DAAD from 11 to 64 outbound offices and international student enrollment in Germany. The gravity equation framework is utilized through synthetic difference-in-differences and fixed effects negative binomial regression to account for the overdispersion in student counts. The use of two-way time and unit weights in the synthetic difference-in-differences estimates helps to address concerns about potential endogeneity and staggered office foundations. The findings suggest that an increase in the number of DAAD offices has a positive impact on international student enrollment in Germany (case rate ratio = 1.11, p-value