Who Advises Whom? Country-Level Patterns in Advisor–Student Recruitment in U.S. Transportation PhD Training

Abstract:

Doctoral training pipelines play a central role in shaping the future transportation research workforce, yet empirical evidence on how international dependence manifests within specific research communities remains limited. This paper examines country-level structure in PhD recruitment within transportation-focused civil engineering groups at highly ranked U.S. institutions. Using publicly available records, we construct a cross-sectional dataset linking faculty advisors to their PhD students and identify undergraduate countries for both advisors and students when available. We characterize advisor-specific recruiting profiles based on the dominant undergraduate country of their students and quantify alignment as the share of advisor–student pairs with matching undergraduate countries.

To assess whether observed alignment exceeds chance expectations, we implement a permutation-based null model that preserves destination-program composition while randomizing student assignment across advisors within the same program. In the stable advisor sample, the observed advisor–student match rate is substantially higher than the null expectation, indicating systematic country-level homophily in PhD recruitment. Advisor-level match rates exhibit wide dispersion, with a subset of advisors showing near-complete alignment and others recruiting more heterogeneously. We further document sharp differences in student composition by advisor undergraduate origin, with particularly strong alignment for advisors trained in China. Sensitivity analyses demonstrate robustness to alternative stability thresholds and exclusion of large advisor rosters, while showing attenuation when Chinese undergraduate backgrounds are excluded.

These findings are descriptive and do not identify causal mechanisms, but they provide a reproducible framework for measuring international dependence in doctoral training at the advisor level. The results highlight concentrated recruitment pipelines within transportation-focused PhD programs and motivate future work incorporating applicant pool data and longitudinal designs to better understand the origins and implications of these patterns.