Japan residency quota simulation finds regional physician disparities largely unchanged

Adjusting residency positions in response to regional physician maldistribution had only a limited effect in a simulation of Japan’s medical residency matching system. Across scenarios that reduced positions according to a physician maldistribution index, fill rates declined slightly, rankings of prefectures and medical schools remained broadly similar to the actual 2022 match, and the Gini coefficient for matched applicants was unchanged. Regional disparities therefore persisted under the tested allocation policies, suggesting that quota adjustment alone was insufficient in the model.

The cross-sectional study used anonymized data from the Japan Residency Matching Program and simulated placements with the Gale–Shapley stable matching algorithm. Researchers examined baseline scenarios in which the number of residency positions in each prefecture was adjusted at reduction rates of 5%, 10%, 15% and 20%. The 20% reduction scenario corresponded to a net reduction of approximately 1.5% in the total number of residency positions nationwide. The simulation included 10,207 applicants, allowing researchers to compare estimated placement outcomes under different policy assumptions with the actual 2022 matching results.

As the adjustment rate increased, the number of applicants assigned a residency position fell from 8,909, a fill rate of 82.6%, in the 5% scenario to 8,648, a fill rate of 81.0%, in the 20% scenario. For comparison, the actual 2022 match placed 8,995 applicants, producing a fill rate of 82.9%. Thus, the simulated policies were associated with a modest reduction in the share of applicants matched as the reduction rate rose. The findings describe modeled outcomes rather than changes observed after a real intervention, so they cannot establish how placements would have responded had the policies actually been implemented.

Most of the comparative rankings also remained stable. When simulated scenarios were compared with the 2022 results, Kendall’s tau coefficients for prefectures ranked by fill rate and medical schools ranked by applicant match rate ranged from 0.687 to 0.919. Match rate was defined as the number of matched applicants divided by the number of applicants from each medical school. These results indicated minor changes in the relative ordering of prefectures and schools even as the overall fill rate changed. The persistence of the Gini coefficient for matched applicants likewise showed that the tested scenarios produced little variation in the distribution of successful placements.

Using the 20% reduction scenario, the researchers also tested three alternative approaches to reallocating positions: uniform reduction, reduction based on prior-year matching results, and reduction based on regional physician density. None of these three reallocation scenarios produced substantial changes from the 2022 results, and the Gini coefficient remained unchanged. The analysis therefore did not identify a tested quota method that markedly reduced regional disparities. The study’s scope was limited to simulated matching outcomes based on program data and specified adjustment rules; it did not measure the effects of recruitment incentives, targeted admissions or community engagement. The authors proposed that policies beyond quota adjustment, including targeted admissions, nonmonetary incentives and community engagement, may be more effective, but that possibility was not tested in the study.

Subject of Research: Simulation of residency matching policies and regional physician distribution in Japan

Article Title: Effects of residency matching policies on regional distribution of residents: a simulation study using a stable matching algorithm

Article References: Ide, H., Hiramatsu, Y., Furui, Y., & Koike, S. (2026). Effects of residency matching policies on regional distribution of residents: a simulation study using a stable matching algorithm. BMC Medical Education. https://doi.org/10.1186/s12909-026-10245-3

Image Credits: AI Generated

DOI: 10.1186/s12909-026-10245-3

Keywords: medical residency, physician maldistribution, Japan, stable matching, Gale–Shapley algorithm, quota adjustment, regional disparities, residency matching, medical education, Effects, residency, matching