Gilles Dufrénot : gilles.dufrenot[at]sciencespo-aix.fr
Kiyotaka Sato : sato[at]ynu.ac.jp
In this paper, we focus on the French national system for allocation of residency positions (NSAR). At the end of the 6th year of medical studies, a national exam ranks all medical students. Then, depending on the ranking, medical students sequentially choose from the available residency positions a specialty in a training region. In 2022, before the national exam, we conducted a unique survey of participants of the NSAR to analyze preferences regarding specialty and training region (N= 3269, 35% of the total 6th year medical studies population). We aimed to simulate the welfare impacts of letting the students choose their preferred specialty and training region compared to the actual process based on students ranking. Our main variable of interest is the residency position that students declared they would choose if all residency positions were available (“the preferred residency position”). We compare this stated preference with the residency position actually chosen (the “real” residency position). Based on this comparison we answer three research questions: what would the care supply look like if we were giving every 6th year students their first choice of residency position and how strong is the discrepancy compared to the result based on national ranking? Among those whose aspirations do not perfectly overlap with their possibilities: what second best choices they made and how region is traded against specialty? Finally, do medical students finally choose what they stated they would prefer (stated preference) ? Finally, we discuss potential mechanisms explaining the discrepancy between stated and observed choices as well as policy implications.