Italy’s Medical-School Admissions Row Is Arguing About the Wrong Thing
The country is fighting over whether the entrance test is too hard. The real question — unasked — is what we are trying to select for.
Editorial — April 2026
A note for readers outside Italy. For decades, access to Italian medical schools has been rationed by a notoriously contentious system: a single multiple-choice national test taken at 19, immediately after secondary school, deciding in ninety minutes who gets one of roughly 24,000 first-year places. In 2024 the government replaced the single exam with a semestre filtro — a “filter semester” in which tens of thousands of students enrol together, sit a common first term of physics, chemistry and biology, and only those who clear a performance threshold continue in medicine. The 2025–26 cycle is the first full run. The debate has been loud, partisan, and almost entirely focused on whether the tests were too easy or too hard. The editorial below argues that this is the wrong argument.
For weeks the Italian debate over the semestre filtro has sounded like a domestic quarrel overheard through a wall. The minister calls the protesters “junior Landinis and Schleins.” A prominent virologist jokes on social media that he would have passed the tests in his second year of secondary school. Students and families speak of “a year thrown away.” Everyone is angry. Almost everyone is concentrating on the wrong detail.
The question is not whether the exam was easy or difficult. The question is that we are arguing about how much to select — without having clarified what we are trying to select for.
A sense of scale, without rhetoric
In 2025–26, 63,079 students enrolled in the Italian filter semester: 52,847 for medicine and surgery, 4,366 for dentistry, 5,866 for veterinary medicine. Against roughly 24,000 places, the numerical outcome is unavoidable. Two-thirds of the candidates will not continue in medicine, whatever system is used to choose among them.
Meanwhile, here is what medical-school selection looks like in other high-income systems.
United States, AAMC 2025–26. Applicants: 54,699. First-year matriculants: 23,440. Admission rate: roughly 43%. And that 43% arrives after three or four years of pre-med undergraduate study, with a mean entering GPA of 3.81, a mean MCAT of 512, and an average of 717 hours of certified community service per matriculant. Nobody walks into a U.S. medical school at nineteen with a secondary-school diploma.
Canada, University of British Columbia, 2025–26 cycle. Applications reviewed: 3,339. Places: 328. Overall admission rate: 9.8%. Candidates invited even to interview: 882 — which is to say, three out of four applicants never meet a member of the admissions panel.
Canada, University of Toronto, 2025. Applications: 4,722. Places: 303. Admission rate: 6.4%. Interviewed: 700, about 14.8%.
At Toronto, 93 candidates out of every 100 do not get in. Nobody organises protests. No virologist mocks them on social media. No minister calls them ideologues. They try again the next year, or they move on to something else. This is not a scandal; it is the ordinary structure of competitive admission to a scarce, long, expensive professional training.
In Italy, by contrast, a semester of serious study of physics, chemistry and biology has become “a stolen year,” and everyone is looking for someone to blame: the minister, the test, the system, fate. Something has gone wrong with the national conversation, not merely with the policy.
“I’ve lost a year” is very often an immature reading
Let us be direct. A year spent studying physics, chemistry and biology is not, in any meaningful sense, lost. It enlarges what the Italian tradition once called cultura generale — general intellectual formation. It builds study discipline. It tempers fortitudo — the capacity to bear fatigue and frustration, which is not optional equipment for a clinical career. Anyone who has studied seriously for a year, even without being admitted to medicine, emerges better prepared for adult life.
The legitimate problem begins at a different point: when the system renders that effort sterile. When the credits accumulated in the filter semester cannot be transferred to any adjacent programme — biology, biotechnology, nursing, biomedical engineering. When there is no structured redirection toward alternative paths. When the asymmetry between candidates with resources (private tutoring, preparation courses, time) and those without is neither monitored nor corrected. A filter is acceptable if it orients. It is acceptable if it recognises the work done and converts a “no” into a “not here, but perhaps over there.” If the semester is merely a diluted version of the old multiple-choice test, after which everything restarts from zero, then we are not selecting. We are consuming energy, money and trust — and, worse, we are feeding a posture of victimhood instead of a culture of maturity.
The question nobody is asking: what does a good doctor actually do?
Here is where the Italian debate misses its own subject. We are confusing the measurement of factual knowledge with the measurement of fitness for the work.
A good doctor is not merely someone who is “fast at answering questions.” A good doctor combines clinical reasoning and quantitative literacy, effective communication, ethical judgment and a sense of responsibility, the capacity to work in teams, stress management and priority-setting, reliability and professionalism. The trouble with the filter semester — as with the old multiple-choice test before it — is not that it selects too much. It is that it selects almost exclusively along one dimension.
When a virologist reduces the whole matter to “it was easy, stop complaining,” he earns applause on social media — but he does not answer the foundational question. Are we selecting good future clinicians, or good future answerers of multiple-choice questions?
What the better-performing systems do
Some countries have tried to answer the harder question: not “who knows most right now?”, but “who will be a trustworthy doctor?” Two instruments recur in the international literature.
Multiple Mini Interviews (MMI). Structured station-based interviews with trained assessors and explicit rubrics. In a multi-centre study across five California medical schools, MMI scores correlated more strongly with downstream outcomes — clinical performance, licensing examination scores — than traditional unstructured interviews.
Situational Judgement Tests (SJT). Scenario-based assessments designed to measure non-cognitive attributes — teamwork, communication, coping, professionalism. They are used in combination with the academic component in contexts such as entry to clinical practice in the United Kingdom.
There is no such thing as “the perfect test.” But there are selection batteries that are measurably better than a single high-stakes multiple-choice quiz.
The ghost of the 1970s: “natural selection” that was nothing of the sort
Some voices in the Italian debate invoke a return to the past: let everyone in, and let the strong survive. This slogan confuses the absence of selection with equity.
In the 1970s, out of 1,200 first-year students at a typical Italian medical school, only about 150 graduated on time. Another 100 to 150 finished within ten years. The rest dropped out. And those who survived were not necessarily the most capable. They were the ones who could afford to fail exams for years without having to work.
That was not natural selection. It was social selection, slightly disguised. And the children of poorer families — the very ones whom some today would “liberate” from testing — were the first to be pushed out.
The political sleight of hand: calling the relocated gate a “reform”
When a minister claims to have “abolished the numero chiuso” — the cap on first-year places — and in the same breath speaks of “filling the 24,000 places,” she is implicitly conceding that the number remains. What has changed is only when and how the decision about who gets through is made.
And when politicians — who require votes — respond from the gut to a complex problem, chasing consensus rather than pausing to study, the outcome is exactly what Italy is now witnessing: cosmetic reforms, sterile polemics, and thousands of young people treated as test subjects in poorly designed experiments.
A problem of this kind demands reflection, analysis of educational systems, examination of what has worked and what has failed elsewhere. It does not demand electoral slogans.
Five concrete choices, instead of slogans
If we genuinely wish to leave the tribal war behind, five adult decisions present themselves.
• Define the profile of the doctor we want to train — cognitive and non-cognitive competencies alike — with universities, clinicians, educational psychologists, the medical order and student representatives as co-designers rather than as factions to appease.
• Turn the filter semester into a bridge semester: credits fully transferable to adjacent programmes (biology, biotechnology, nursing, biomedical engineering), real tutoring, structured remedial support. “Losing a year” must become rare; more often it must become intelligent redirection.
• Evolve the cognitive component: less rote knowledge, more reasoning, data interpretation, graph comprehension, foundations of scientific method. And, crucially, psychometric transparency about the test items themselves.
• Add a standardised non-cognitive component: well-constructed SJTs, MMIs with trained assessors and shared rubrics, run through regional hubs where necessary.
• Measure and correct for equity: monitor socioeconomic impact, provide free preparation materials, fund tutoring for candidates who start at a disadvantage, enforce serious controls against irregularities.
The real alternative
The choice is not between “selection” and “no selection.” It is between blind selection — one dimension, often socially regressive — and intelligent selection — multiple dimensions, real orientation, genuine possibilities of reallocation, less human waste.
This is why the current row is a distraction. It shifts attention from the end (what kind of doctor) to the means (how hard was the physics paper). As long as we continue this way, we will continue to produce anger, appeals, suspicion, and cosmetic reform.
The grown-up question is a single one. What kind of doctor do we want to form — and with what serious instruments can we recognise that doctor before they set foot on a ward?
Because if we do not define the good doctor ourselves, a sixty-second multiple-choice test will do it for us — or a chaos dressed up as “natural selection” will. And both will be called “merit.”
