Cutting medical school places: Poland has done it before.

The dispute over whether Poland trains too many doctors has a precedent. At the turn of the 1980s and 1990s admissions to medicine were cut from over 6,000 places to about 2,000 and kept at that level for more than a decade. The pre-cut level returned after 31–36 years, depending on which places are counted. France and Portugal did the same, and the effects of those decisions are visible in the medical workforce to this day.

We describe how much is known about the Polish cut and where the numbers come from, what trace it left in the register of doctors and what similar decisions looked like in two other countries. At the end we return to the analysis published by the Supreme Medical Chamber (NIL) in 2024, which proposed cutting a second time, and to the question of what basis is used to set this number at all.

How Poland cut admissions

According to Rynek Zdrowia, the number of first-year places in medicine was reduced gradually from over 6,000 in 1987 to about 3,300 in 1990, and as late as 2003 there were 2,090 places in day programmes in medicine.¹ Polityka Zdrowotna writes that in 1994–2000 just over 2,000 people a year were admitted.² The largest part of the cut therefore fell within three years: between 1987 and 1990 admissions shrank by about 45%. Then for more than a decade they stayed near the bottom. From 1987 to 2003 the fall was at least 65% — “at least”, because the starting point is “over 6,000” places.

Why was the cut made? In the same Rynek Zdrowia article a former president of the Supreme Medical Council (in 2010–2018) explained that politicians considered the number of graduates too large. This “overproduction” was linked to Poland being treated as the “medical back-up for the operations of the Warsaw Pact” (our translation), while the country was emerging from poverty and assessed health needs differently.¹ This is an explanation given almost three decades later, not a document from the time of the decision.

The recovery took more than three decades

The Ministry of Health writes that it has been increasing admission limits for medicine since 2005.⁴ In full-time programmes the recovery was slow: 2,360 places in 2005, 3,023 in 2010⁵ and 3,589 in 2015/16.⁶

In full-time programmes in Polish, the 1987 level returned only in 2023/24

1. 1987 · first-year places, according to the pressover 6,000
2. 1990 · first-year places, according to the pressabout 3,300
3. 1994–2000 · according to the press, no annual datajust over 2,000
4. 2003 · day programmes, according to the press2,090
5. 2005 · full-time programmes2,360
6. 2010 · full-time programmes3,023
7. 2015/16 · regulation of the Minister of Health3,589
8. 2019/20 · regulation of the Minister of Health5,042
9. 2022/23 · regulation of the Minister of Health5,836
10. 2023/24 · regulation of the Minister of Health, first year above 6,0006,707
11. 2026/27 · regulation of the Minister of Health7,158

How many years the return took depends on which places are compared with 1987. Full-time programmes in Polish, together with the pool for foreigners, exceeded 6,000 places only in 2023/24 — there were 6,707 of them then, 36 years after 1987.⁶ If fee-paying part-time programmes in Polish are added, the threshold was crossed in 2018/19, after 31 years: the limit after the autumn amendment gave 6,123 places in Polish.⁷ Which measure corresponds better to the 1987 number is not known, because we do not know its definition.

The total limit, including part-time programmes and programmes in a foreign language, is 10,730 places in 2026/27, of which 8,892 are in Polish.⁸ How it grew year by year is shown in a table in the analysis English Division, and why there is still a shortage of doctors — in the analysis Why Poland is short of doctors. Who sets the limit and what role the medical self-government plays in it is described in the text on restricted access to the profession.

A gap in the register of doctors

The effects of the cut can be looked for in the age structure of doctors. The Central Register of Doctors gives the number of practising doctors in each five-year age group. Among doctors aged 26–70 the smallest group is those aged 46–50, the cohorts that started their studies roughly in the second half of the 1990s, when admissions were at their lowest.⁹

Among doctors aged 26–70, those aged 46–50 are the fewest — the cohorts that started their studies when admissions were lowest

1. aged 26–30 · studies from about 2015–1920,346
2. aged 31–35 · studies from about 2010–1419,834
3. aged 36–40 · studies from about 2005–0917,008
4. aged 41–45 · studies from about 2000–0412,787
5. aged 46–50 · studies from about 1995–9911,607
6. aged 51–55 · studies from about 1990–9412,563
7. aged 56–60 · studies from about 1985–8915,963
8. aged 61–65 · studies from about 1980–8415,183
9. aged 66–70 · studies from about 1975–7914,830

There are 11,607 doctors aged 46–50. That is 27% fewer than among those aged 56–60 (15,963), who started their studies before the cut, and 43% fewer than among those aged 26–30 (20,346), who entered medical school after the limits were rebuilt.⁹ The gap falls in the middle of the professional career.

The smaller number of doctors is not due to smaller birth cohorts. The 41–45 age group, also one of the least numerous (12,787 doctors), consists of people born roughly in 1981–1985, at the peak of the last baby boom. In 1983 nearly 724,000 children were born in Poland.¹⁰ The constraint was therefore not the number of young people but the number of university places.

The cut in limits is not, however, the only explanation for the gap. The same former president of the Supreme Medical Council pointed out that a clear staffing collapse came after Poland joined the European Union, when doctors began leaving more often.¹ In 2004 today’s 46–50-year-olds were 24–28, finishing their studies or internship, and the 41–45-year-olds were 19–23 and at university. Part of the gap is therefore emigration, not only admissions. Public data do not allow these two causes to be separated, because they do not show how many doctors from these cohorts left for good.

Portugal: admissions down by three-quarters

In Portugal in the early 1970s it was judged that there were too many students in medical schools, which made, among other things, hospital-based teaching difficult. The government introduced a numerus clausus, and according to a study by EDULOG admissions were restricted too much and for too long. In 1979 and 1980 the global limit for medicine was 805 places, and in 1986 it fell to 190 — by 76%. In 1999/00 it stood at 70% of the 1979 level, and it exceeded that level for the first time in 2000/01, after 21 years. Since 2010/11 it has remained at about 1,700 places a year.¹¹

The trace of this decision is visible in the workforce. A 2013 study by the University of Coimbra for the Portuguese medical association called it a “generational gap”: doctors aged 40–49 made up only 15% of all doctors.¹² According to data from the Portuguese patient register of 24 August 2026, about 1.36 million people were waiting to be assigned a family doctor.¹³ The number cannot be directly compared with earlier, higher estimates, because in 2026 the way patients are classified in the register was changed — part of the fall is due to that change.

Portugal also teaches caution with the “doctors per 1,000 population” indicator. In OECD statistics it has 5.8, but this is the number of doctors licensed to practise, including retired people and those who do not practise. The OECD itself notes that this number overstates the system’s capacity.¹⁴ When the Portuguese statistical office counted practising doctors, it found 46,131 in 2024.¹⁵ The same pitfall applies to Poland — see Why Poland is short of doctors.

France: a 61% cut and a correction ten years late

France introduced a numerus clausus in medicine by an act of 1971, with limits from 1972. According to the OECD the official aim was to match the number of students to the training capacity of hospitals, and a less openly stated one — to limit the number of doctors and the pressure on spending in a system that pays for each service.¹⁶ Admissions fell from nearly 9,000 people a year in the mid-1970s to 3,500, or by 61%.¹⁶ According to the French National Academy of Medicine, the bottom, 3,500 places, came in 1992. The limit should have been rising from the late 1980s, but it was raised only ten years later — to 7,500 places in 2007.¹⁷ By 2011 it had returned to 8,000.¹⁶

The effects are visible to this day. According to the French health statistics office (DREES), there were 318 doctors per 100,000 population in 2021, compared with 325 in 2012. Standardised density, which takes population ageing into account, is projected to fall until 2028 and to return to the 2021 level only around 2035. A scenario with admissions 20% higher speeds up the way out of the trough by about three years — it does not remove it.¹⁸

In 2019 France abolished the numerus clausus. Five years later the French Court of Audit (Cour des comptes) found that there is still no common method for estimating health needs, so it cannot be guaranteed that the current targets are right.¹⁹

Three countries, the same pattern

Numbers of places from different countries cannot be compared directly, because education systems and population sizes differ. What can be compared is the course of the decisions: how deep the cut was, how it was justified and how long the return took. The comparison is based on the sources from the previous sections.¹¹¹¹⁶

In none of the three cases was the stated reason a calculation of the population’s health needs. What decided was the capacity of universities and hospitals, spending or a political judgement, and the effects appeared after a dozen or more years, because that is how long it takes from admission to medical school to the title of specialist (see Why Poland is short of doctors). In each of these countries the limit was set by the state, not by the medical self-government.

The 2024 proposal for a second cut

In June 2024 the Supreme Medical Chamber published an analysis prepared by the then chair of its team for the transformation of the healthcare system. The chamber stated that it was not the position of the Supreme Medical Council. According to the team’s model, in 2048 Poland will need about 5.14 doctors per 1,000 population (in a range of 4.8–5.4), and with the limits proposed by the ministry at the time it would reach 6.3, about 20% above need. The cost of training the surplus doctors was estimated at up to PLN 15.0 billion (≈ €3.4 billion) over the forecast period.²⁰

The analysis presented three variants and did not say which one it recommended:²⁰

• Variant A: keep the 7,797 places in programmes in Polish from the draft regulation for 2024/25, and from 2027 “cut the number of places abruptly by almost half” (our translation), to about 4,100 a year.
• Variant B: reduce the number of places every year by about 9%, to about 4,200 a year from 2030.
• Variant C: reduce the limit by 1,500 places already in 2024, to about 6,300, and then by 4% a year in 2026–2034, to about 4,200.

Poland passed the level of about 4,100 places during the first cut roughly in 1988–1989, between over 6,000 in 1987 and about 3,300 in 1990.¹ This is an approximate comparison: the analysis counts all places in Polish, while the numbers from the 1980s concern the first year of studies.

The text of the analysis says nothing at all about doctors’ pay — we checked its version on the chamber’s website on 16 October 2026. Its arguments are the cost of training and the balance between the supply of doctors and need. The analysis itself also refers to the experience of other countries: it warns against a “yo-yo effect”, that is fluctuations in limits set without prior analysis, and calls for a plan of the number of places 5–10 years ahead.²⁰

None of the variants came into force. The final version of the regulation for 2024/25 gave 8,226 places in Polish instead of the 7,797 in the draft; in 2025/26 there were 8,616, and in 2026/27 — 8,892.²¹ In June 2026, however, the Ministry of Health announced that the limit no longer needed to be increased and that it might fall in the coming years.²² The dispute over the direction continues, and the medical self-government gives its opinion on successive drafts — we describe this in the text on restricted access to the profession.

The “show the model” test

The chamber and the ministry reach a similar conclusion. In November 2025 the Ministry of Health forecast that by 2034 the supply of doctors would exceed need, assuming that a doctor works on average 161 hours a month. It wrote: “The challenge for healthcare in the future will therefore not be staff shortages among doctors” (our translation).⁴ It added the caveat that shortages may occur in some regions and fields. We write more about this forecast in the analysis Why Poland is short of doctors.

The history of the three countries yields a simple test for any proposal to change the limit, up or down: show the model. What health needs do you assume, on what data, with what working time per doctor, with what breakdown by region and field — and what happens when these assumptions are changed.

• The chamber’s 2024 analysis describes results and variants, but the chamber’s website has neither the input data nor a calculation that would allow it to be reproduced. Its author wrote, at the same time, that the discussion should be about the assumptions adopted and the mechanics of the model.²⁰
• The ministry does not give the forecasting method in its November 2025 announcement.⁴ On the BASiW portal, however, it publishes an interactive report, “Model popytowo-podażowy lekarzy” (“Supply and demand model for doctors”), with definitions, data sources and scenarios.²³ Whether it allows the result to be recalculated on one’s own assumptions, we have not established.

After fifty years of regulating admissions, France still has no common method for estimating needs.¹⁹ Without a public model it is impossible to check whether a proposed number follows from patients’ needs or from something else — and a mistake comes to light after a dozen or more years. How a change in admissions translates into the number of doctors in the following decades can be checked in the doctor supply simulator (in Polish).

The same principle — decisions on public money based on public data — underlies our proposal: open pseudonymised data, that is the total paid from public money to each person, under a permanent identifier, without names.

Notes 23laws, data and statements we relied on

1. Rynek Zdrowia, “Limity przyjęć na medycynę wyraźnie rosną. Czy odbudujemy lekarskie kadry?” (“Admission limits for medicine are clearly rising. Will we rebuild the medical workforce?”, in Polish), 16.09.2019 — first-year places: over 6,000 in 1987 and about 3,300 in 1990; 2,090 places in day programmes in 2003 (without naming a source); statement by the president of the Supreme Medical Council in 2010–2018 on the reasons for the cut and on emigration after 2004. ↩ ↩² ↩³ ↩⁴ ↩⁵

2. Polityka Zdrowotna, article on admissions to medical school (in Polish) — over 6,000 students in 1987 and just over 2,000 in 1994–2000, without naming a source. ↩

3. Regulation of the Minister of Health of 24 April 2003 on admission limits for medical studies, Dz.U. 2003 No. 99 item 917 (in Polish) — the annexes with the limits were published as a separate annex to that issue of the Journal of Laws; they are not in the file of the act. ↩

4. Ministry of Health, “Kształcenie i liczba lekarzy — trendy i wyzwania” (“Medical education and the number of doctors — trends and challenges”, in Polish), 7.11.2025 — increasing the limits since 2005, a forecast of a surplus by 2034 assuming on average 161 working hours a month, the caveat about shortages in some regions and fields. ↩ ↩² ↩³

5. Forum Akademickie, “Lekarskie limity” (“Medical limits”, in Polish), No. 9/2022 — places in full-time programmes in Polish: 2,360 in 2005 and 3,023 in 2010. ↩

6. Ministry of Science and Higher Education, reply to parliamentary question No. 9177 (in Polish), 28.04.2025 — the limit for medicine broken down into four types of programme: in 2015/16, 3,529 full-time places in Polish and 60 for foreigners; in 2023/24 (Dz.U. 2023 items 1322 and 2027), 6,598 and 109. ↩ ↩²

7. Regulation of the Minister of Health of 16 July 2018, Dz.U. 2018 item 1381, amended by the regulation of 26 September 2018, Dz.U. 2018 item 1845 (in Polish), annex 1 — after the amendment, 4,691 full-time places in Polish, 87 for foreigners and 1,345 part-time; totals — our calculations. In the original version of the act there were 5,955 places, and the 6,000 threshold would have been crossed a year later. ↩

8. Regulation of the Minister of Health of 28 August 2026 on the admission limit for medicine and dentistry, Dz.U. 2026 item 1167 (in Polish), annex 1 — totals across 39 universities, our calculations: 7,051 full-time places in Polish, 107 for foreigners, 1,734 part-time and 1,838 in a foreign language. ↩

9. Supreme Medical Chamber, Zestawienie liczbowe lekarzy i lekarzy dentystów według wieku, płci i tytułu zawodowego, stan na 31.12.2025 (“Doctors and dentists by age, sex and professional title, as of 31.12.2025”, in Polish) — practising doctors (excluding dentists) by age; percentages — our calculations. ↩ ↩²

10. Statistics Poland (GUS), Demographic Surveys Department, “Dzieci w Polsce w 2008 roku. Charakterystyka demograficzna” (“Children in Poland in 2008. Demographic characteristics”, in Polish), November 2009 — 1983 as the peak year of the last baby boom, nearly 724,000 births. ↩

11. EDULOG, “Há falta de médicos em Portugal?” (“Is there a shortage of doctors in Portugal?”, in Portuguese), 2018 — reasons for introducing the numerus clausus; global limit of 805 places in 1979–1980, 190 in 1986, 70% of the 1979 level in 1999/00, first exceeded in 2000/01, about 1,700 since 2010/11. ↩ ↩²

12. University of Coimbra for the Ordem dos Médicos, study on the projected number of doctors (in Portuguese), 2013 — “hiato geracional”, doctors aged 40–49 as 15% of all doctors. ↩

13. Diário de Notícias, data from the Portuguese health service patient register (ACSS) (in Portuguese), August 2026 — as of 24.08.2026: 1,364,561 people waiting to be assigned a family doctor, after the change in the classification of patients in the register. ↩

14. OECD and European Observatory on Health Systems and Policies, “Portugal: Country Health Profile 2025” — 5.8 doctors licensed to practise per 1,000 population, including retired people and those not practising; Portugal as one of three EU countries without data on practising doctors. ↩

15. ECO, INE data on practising doctors (in Portuguese), 24.07.2026 — 46,131 practising doctors in 2024. ↩

16. OECD, “Health Workforce Policies in OECD Countries: Right Jobs, Right Skills, Right Places”, 2016, chapter 3 and box 3.1 — numerus clausus in France since 1971, limits since 1972, official and less openly stated aims; a fall from nearly 9,000 in the mid-1970s to 3,500 in the 1990s; 8,000 by 2011. ↩ ↩² ↩³ ↩⁴

17. Académie nationale de médecine, report on medical demography (in French), 2014 — 3,500 places in 1992; the increase in the limit, needed from the late 1980s, came ten years later; 7,500 in 2007. ↩

18. DREES, press release on projections of the number of doctors, midwives, dentists and pharmacists (in French) — 318 doctors per 100,000 population in 2021 compared with 325 in 2012; projection of standardised density with a trough in 2028 and a return to the 2021 level around 2035; a scenario with admissions 20% higher. ↩

19. Cour des comptes, “L’accès aux études de santé” (“Access to health studies”, in French), December 2024 — abolition of the numerus clausus by the act of 24.07.2019; no common method for estimating health needs. ↩ ↩²

20. Supreme Medical Chamber, “Polska na drodze do marnowania miliarda złotych rocznie na kształcenie nadmiarowych lekarzy” (“Poland on course to waste a billion zloty a year on training surplus doctors”, in Polish), 11.06.2024 — with the caveat that the analysis is not the position of the Supreme Medical Council; model to 2048, variants A–C, cost of up to PLN 15 billion, “yo-yo effect”. Text checked on 16.10.2026. ↩ ↩² ↩³ ↩⁴

21. Regulations of the Minister of Health on the admission limit for medicine, annex 1, in the final version: for 2024/25 Dz.U. 2024 item 1085 amended by item 1419, for 2025/26 Dz.U. 2025 item 1086, for 2026/27 Dz.U. 2026 item 1167 (in Polish). Places in Polish are the total limit excluding programmes in a foreign language; totals — our calculations, as in the table in the analysis English Division. ↩

22. Rynek Zdrowia, “Koniec rekordowych limitów przyjęć na medycynę. To może być ostatni tak liczny nabór” (“The end of record admission limits for medicine. This may be the last such large intake”, in Polish), June 2026 — statement by the deputy minister of health on the admission limit. ↩

23. Ministry of Health, BASiW portal, “Model popytowo-podażowy lekarzy” (“Supply and demand model for doctors”, in Polish) — interactive report (Power BI): definitions, data sources, supply, demand and scenarios; as of 1.09.2026. ↩