Poland: A Ruling That Reached Into Private Life

On 22 October 2020 the Constitutional Tribunal of Poland held that one of the three grounds on which a pregnancy could lawfully be ended in that country was incompatible with the Constitution. The ground was the foetal one: a high risk that the foetus would be severely and irreversibly damaged, or would suffer from an incurable illness threatening its life. The judgment was not published for three months. It took effect on 27 January 2021, the day publication finally occurred.

What happened to the official count afterwards is the part everybody knows. Lawful terminations in Polish hospitals fell from just over a thousand a year to a hundred and seven. That is a direct effect of a rule, it is in the state’s own annual report to parliament, and nobody on either side of the argument disputes it.

The part that earns an essay is what happened next. Between 2022 and 2024 the same official count rose again by roughly a factor of eight, and during those years the statute did not change by a single word. No parliament amended it. No court revisited it. What changed was a set of ministerial guidelines, a set of instructions to prosecutors, and a contractual requirement on publicly funded hospitals — none of which altered what was lawful, all of which altered what doctors believed they were permitted to do.

That is the subject here: not the rule, but the margin. The distance between what a law forbids and what the people who must apply it will risk. When the legal margin narrows, the effect is not confined to the people who wanted the procedure the law names. It changes what a doctor will say, what a hospital will schedule, what a couple will plan around, and what can be raised in a consulting room at all. The chilling effect is the story, and in Poland, unusually, it leaves a countable trace.

This Library takes no position on abortion, and this essay does not offer one. Its position is narrower, and it hopes to be available to readers who disagree with each other completely: laws that govern intimate life have effects beyond their stated scope, and those effects can be documented. A reader who thinks the Tribunal was right and a reader who thinks it was a catastrophe should both be able to finish this piece without finding it has quietly argued the other case.

The governing statute is the Act of 7 January 1993 on family planning, protection of the human foetus and conditions permitting the termination of pregnancy — the text Polish debate calls, with varying degrees of irony, the compromise. As enacted it allowed a lawful termination on three grounds: danger to the pregnant woman’s life or health; a high risk of severe and irreversible foetal damage or incurable life-threatening illness; and strong grounds for believing the pregnancy resulted from a criminal act. A 1997 amendment added difficult material or personal circumstances, and the Constitutional Tribunal struck that addition down within months. The 2020 judgment was therefore the second time the same court had removed a ground from the same statute.

The road to it ran twice. In June 2017 a group of members of the Sejm asked the Tribunal to declare the foetal ground unconstitutional; those proceedings were discontinued in July 2020 because the application had been lodged in an earlier parliamentary term. A fresh application by a further group of deputies, lodged in November 2019, became case K 1/20.

On 22 October 2020 the Tribunal, sitting as a plenary formation of thirteen judges, held by eleven votes to two that the foetal ground, and the sentence of the statute setting its time limit, were incompatible with Article 38 of the Constitution read together with Article 30 and Article 31(3) — the right to legal protection of life, read with the inherent and inalienable dignity of the human person, and with the clause governing when constitutional freedoms may be limited. The core of the reasoning was that a ground defined by the anticipated characteristics of the being to be born could not be reconciled with the protection the Constitution owes to that being’s dignity.

Two things survived. A termination remains lawful where the pregnancy endangers the woman’s life or health, with no gestational limit, performed in a hospital and certified by a physician other than the one performing it except where the threat to life is direct. And it remains lawful, until the end of the twelfth week, where there are strong grounds for believing the pregnancy resulted from a criminal act.

One further point of law is stated here because commentary abroad gets it wrong in both directions. Poland does not punish the pregnant woman. Under the Penal Code the offence is committed by a person who terminates a pregnancy in breach of the statute, or who helps or induces a pregnant woman to do so, and carries up to three years. A woman who ends her own pregnancy commits no offence at all. Readers who believe Polish women risk prison for their own decisions are mistaken; so are readers who conclude from that fact that nothing much followed from the ruling. What followed fell on the people around her.

The Polish Constitution requires that judgments of the Constitutional Tribunal be published immediately in the official journal in which the reviewed act appeared, and provides that a judgment takes effect on the day of its publication. The judgment of 22 October 2020 was announced, reported, protested against across the country — and not published. Publication came on 27 January 2021, together with the reasoning and the dissents.

For those three months the legal position was, in the strict sense, knowable: the old provision remained in force until publication. In the practical sense it was not, because nobody could say when publication would come, and a hospital scheduling a procedure four weeks ahead was scheduling into a period whose law had not yet been settled. Procedures were cancelled in that window. That is the chilling effect in its purest and most checkable form: behaviour changing in response to a rule that did not yet have legal force.

It is also, now, a finding of a court. In A.R. v. Poland, decided unanimously by the First Section of the European Court of Human Rights on 13 November 2025, the Court held that the prolonged uncertainty as to the applicable law, caused by the delayed publication and entry into force of the Tribunal’s judgment, lacked the foreseeability that Article 8 of the Convention requires, and that the resulting interference with the applicant’s private life was therefore not in accordance with the law.

Notice precisely what that holding is and is not. It is not a ruling that a right to abortion was denied; the Strasbourg court has consistently declined to say any such thing. It is a ruling that a period in which nobody could state the law was itself an interference with private life. For the argument of this essay that is the more useful finding of the two, because it locates the harm in uncertainty rather than in prohibition — which is exactly where a chilling effect lives.

The composition of the Tribunal that decided K 1/20 is contested, and this essay reports the dispute instead of resolving it, because resolving it is not within the competence of a foreign editorial page.

The contested facts are these. In December 2015 the President of Poland declined to swear in three judges elected to the Constitutional Tribunal by the outgoing parliament, and the incoming parliament then elected three judges to seats that had already been filled. In 2021, in Xero Flor w Polsce sp. z o.o. v. Poland, the European Court of Human Rights held that there had been serious irregularities in that election procedure. In M.L. v. Poland, decided by five votes to two on 14 December 2023, the same court found a violation of Article 8 on the ground, among others, that the bench which issued the 2020 abortion ruling had included a judge elected in that procedure and replacements for two others, so that the interference had not been issued by a body meeting the Convention’s rule-of-law requirements. It is also a matter of record that one of the parliamentarians who signed the 2017 application to the Tribunal was elected a judge of that Tribunal in December 2019 and sat on the bench that decided the 2020 case.

The opposing position is not frivolous and is stated here at its own strength. The Constitution provides that judgments of the Tribunal are universally binding and final. On that view a judgment cannot be unmade by a parliamentary resolution, by a change of government, or by a court in Strasbourg whose jurisdiction runs to the Convention and not to the composition of national constitutional courts; and when the Sejm adopted a resolution in March 2024 declaring the Tribunal incapable of performing its constitutional tasks, the Tribunal held that resolution unconstitutional on precisely that reasoning. Polish lawyers of good faith are on both sides of this, and the disagreement turns on questions of domestic constitutional law that do not translate cleanly.

What can be said without adjudicating is threefold. The dispute exists. It is not a fringe position on either side. And the European Court has now twice treated it as legally consequential, which means that whatever one concludes about its merits, it is not merely rhetorical. The essay goes no further, and a reader who thinks the Tribunal was properly constituted loses nothing of the argument that follows, because that argument does not depend on the point.

It is worth adding that the judgment was not unanimous. Two judges dissented. One argued that the application should have been rejected as formally inadmissible; the other that the Tribunal had considered only one of the two perspectives the case contained, weighing the protection of life before birth without weighing the dignity of the woman. Their opinions, together with the separate opinions on the reasoning, ran to several times the length of the judgment itself.

Poland’s government reports annually to parliament on the operation of the 1993 Act, and the Ministry of Health maintains the underlying count. The series is the most solid thing in this essay and it is worth setting out plainly.

In 2019 there were 1,110 lawful terminations in Polish hospitals, of which 1,074 were performed on the foetal ground. In 2020 there were 1,076, of which 1,053 were on that ground. In 2021, the first year of the new position, there were 107. In 2022 there were 161. In 2023 there were 425. In 2024 there were 885, of which 882 rested on the threat to the woman’s life or health and three on the criminal-act ground.

The first thing the series shows is why the ruling was decisive rather than marginal. The foetal ground had been carrying almost the entire lawful practice of the country; the other two grounds together accounted for a few dozen procedures a year. Removing it did not narrow lawful abortion in Poland. It very nearly ended it, and it did so without the legislature voting on anything.

The second thing worth saying is what the series cannot do. It counts procedures performed in hospitals under a statute. It is not a count of pregnancies, of decisions, of outcomes, or of what anyone did instead. It says nothing whatever about how many abortions occurred in Poland in any year, and a reader who treats it as a national total has misunderstood what is being counted. It is the numerator of a fraction whose denominator nobody has.

Now hold the two ends of the series together. In 2022, 161. In 2024, 885. Between those two years the Act of 1993 was not amended, the Tribunal’s judgment was not revisited, and the two remaining grounds read exactly as they had read before.

What did change is administrative. In August 2024 the Prosecutor General issued guidelines to prosecutors handling cases connected to termination of pregnancy. At the end of the same month the Ministry of Health published guidelines to hospitals stating that the statutory reference to the woman’s health is general and therefore includes mental health; that one physician’s certificate is sufficient and a hospital may not require further opinions; and that the conscience clause belongs to an individual doctor and cannot be exercised by an institution. A separate requirement obliged publicly contracted obstetric and gynaecological providers to offer lawful terminations, with financial consequences for those that did not, running to a share of the value of the provider’s contract with the national health fund and to a penalty available to the patients’ rights commissioner.

The prime minister said of the guidelines that they could not change the law, but that they could change the attitude of prosecutors and doctors. It is worth taking that sentence at face value rather than as politics. It is a statement, by a head of government, that the behaviour of the medical profession is set by something other than the statute — and it turned out to be correct, by a factor of five in two years.

This is the mechanism of the essay, running in the other direction, and the symmetry is not optional. If an administrative document is credited with an eightfold rise from a floor, then the ruling of 2020 must be credited with the fall — and not only through what it forbade, but through what it left people unsure of. A reader cannot hold the first and deny the second. They are the same claim about the same variable.

And the symmetry cuts against this essay as hard as it cuts for it. If the margin can be widened by a circular, then the chilling effect was never a wall; it was a practice, and practices can be changed without changing law. That is the good news in the Polish record, and it is also why the change is fragile. A guideline can be withdrawn as easily as it was issued, by a ministry, without a vote, and the women affected would have no judgment to point at.

From October 2022 Polish clinicians have been obliged to record pregnancy among the data entered into the national medical information system, under an ordinance signed by the health minister earlier that year. The government’s stated purpose was continuity of care: a doctor seeing a patient should know she is pregnant. The Ombudsman’s objection was not that the purpose was sinister but that the scope of personal data processing ought to be set by statute rather than by ministerial ordinance. Opposition politicians called it a register and read it as an instrument of detection.

This essay cannot settle what it is, and does not need to. What matters for the argument is that the existence of the requirement changed what some people believed they could say in a consulting room, and that no clause of the 1993 Act mentions consulting rooms at all. That belief is the thing the law never legislated and cannot measure. The essay states the requirement, states the objection, and declines to assert an effect it cannot evidence.

Two further effects are asserted constantly in commentary and are not asserted here, because the verification was attempted and failed. The first is that uptake of prenatal diagnostic testing fell after the ruling, on the reasoning that a test whose result cannot change anything is a test not worth having. It is plausible, it is said often, and I could not confirm it against an official statistical series or peer-reviewed work, so it is not claimed. The second is that couples deferred or abandoned planned pregnancies. Registered births in Poland did fall sharply across this period and the fertility indicator reached a modern low, but births were already falling before 2020 and across the region, and attributing a demographic trend to a single ruling is exactly the move this Library asks other people not to make.

What can be supported rests on peer-reviewed work about doctors rather than about patients, and it is more interesting than the missing numbers would have been. Research on abortion provision in Poland published in Social and Legal Studies argues that what governed provision was institutional and intra-professional as much as criminal: hospital exposure, professional sanction, the visibility of the individual clinician within a small specialty. If that is right, then the criminal law was never the binding constraint, and a shock delivered to the statute is transmitted through institutions in ways the statute cannot anticipate — which is another way of saying that the margin was always where the action was.

One small study is worth naming for a single point. A cross-sectional survey of forty Polish obstetrician-gynaecologists performing prenatal ultrasonography, published in 2022, found the physicians surveyed considerably more permissive than the law they were working under, with around nine in ten treating severe or lethal foetal defects as an indication for termination. Forty respondents in one recruitment channel is not a representative picture of Polish medicine and the authors say so. It is cited here for the gap it illustrates rather than for its magnitude: the people who must apply a rule are not the same population as the people who wrote it, and the distance between them is where a chilling effect either forms or fails to.

Four objections, each of which this essay considers serious, and the first of which it considers unanswered.

The first is that a state may legitimately legislate here, and that a chamber which treats the narrowing of a legal margin as self-evidently a harm has smuggled in a conclusion it claimed not to hold. The Tribunal’s reasoning is a constitutional argument, not a pretext: that a ground for termination defined by the anticipated characteristics of the being to be born stands in tension with the equal dignity the constitution extends to that being. Serious jurists in many countries have made versions of it, and nothing in the record about how the bench was elected answers it. Disliking who ruled is not the same as refuting what was ruled, and a reader who holds that the foetal ground was indefensible can accept every factual claim in this essay and keep that position intact.

The second is that chilling effect is easy to assert and hard to measure, and that the strongest evidence against the population-level version comes from the very court that has twice found individual violations. In 2023 the European Court declared inadmissible applications brought by 927 women of childbearing age who argued they were potential victims of the new position, following the same reasoning it had applied in an earlier group of cases: they had not shown they were at real risk of being directly affected. Strasbourg has been willing to find a violation where a particular woman’s scheduled procedure was cancelled. It has consistently refused to accept the general claim that a population was harmed by a climate. That refusal is a serious evidential objection and not merely a procedural one.

The third concerns the figures this essay has deliberately not used. Counts of people assisted to obtain terminations abroad or by pills sent through the post are published annually by organisations that campaign on the issue. They are self-reported, unaudited, not comparable with a hospital register, and produced by bodies whose funding and standing are connected to their size. This does not make them false. It makes them unusable for the only purpose that would matter, which is magnitude. They establish that a route exists and is used. They do not establish how many use it, and this essay makes no claim that depends on them.

The fourth is about this publication rather than about Poland. A Japanese-language editorial page attached to a private service in Tokyo has no standing to referee a dispute among Polish constitutional lawyers, and should be suspicious of the ease with which a foreign story arranges itself into a moral. It is also worth noticing that Polish opinion does not match what either camp abroad assumes. The country’s public opinion research centre found in a survey conducted in February 2024 that a clear plurality of respondents held that a termination should be lawful where a child is known in advance to be seriously impaired, with a smaller group opposed, while support for termination simply because a woman does not want a child stood far lower — although it had risen by eleven points in a single year. A country can hold both of those at once. Most do, including this one.

The reason Poland can be written about in this way is that something happened on a day. There is a before and an after, a series that falls and a series that rises, a ministerial document with a publication date sitting next to a count that moved afterwards. Discontinuity is what makes an effect measurable, and measurability is what makes an argument possible.

Japan’s constraints on the same territory have no date. This chamber has already set out what the Maternal Health Act permits, on whose consent, and what an international committee asked Japan to report back on, in its essay on a right won in court and a danger that did not move; and it has set out at length, in the essay on the vacuum where a law should be, what happens to a field of intimate life governed not by statute but by professional guidelines, a single judgment and a long legislative silence. The point here is not that Japan is more restricted or less. It is that a continuous constraint produces no discontinuity, therefore no measurable effect, therefore no series anyone can put in front of a reader — and therefore, in the narrow and specific sense that matters to a publication like this one, no argument. The constraint that has always been there is the one nobody can demonstrate.

That is the mirror, and it is uncomfortable in a way worth sitting with. A reader here may finish an essay about Poland thinking that this is what it looks like when a state reaches into private life. The honest reading is the reverse: Poland is what it looks like when a state reaches into private life visibly. The invisible version leaves no trace of the kind this essay has been quoting, which does not mean it leaves no trace.

This chamber’s synthesis piece on Japan holds that permission attaches to the act and prohibition to the disclosure — that one may do a great deal and say very little. Poland arrives at the same distinction from the opposite direction. The practical difficulty after January 2021 was not principally that a woman could not act; it was that she could not ask. She could not put a question to a physician and be confident of how the question itself would be received. That is the same structure the Japanese essay describes, produced by a constitutional court rather than by a social convention, and it lands on the same person in the same room.

The Turkish essay in this chamber found that a right which arrives by one instrument can leave by the same one, and that what makes a right durable is not the instrument but whether enough people would notice its absence. Poland supplies the sequel that Turkey could not. In Poland, people did notice, at a scale the country had not seen in a generation; the government changed; and the ruling remained in force, because a judgment of a constitutional court is not undone by an election. What filled the space instead was an administrative document, which holds only as long as the administration does. Noticing is necessary. Poland shows what it is not sufficient for.

One reading should be refused before it forms. It would be easy, and lazy, to make this an essay about the Church. This chamber has already argued, in its essay on what religion was actually doing, that the rules were the least of what religion supplied and that the secular account of it is too narrow. Poland’s abortion politics are conducted in constitutional language, by lawyers, before a court, over the text of a statute that was itself a negotiated settlement between institutions rather than a sermon. Treating it as law and politics is not a concession to one side. It is a description of the file.

This house sells private evenings in Japan, and the thing it actually sells is a room in which a woman can say what she wants and be answered without consequence. An essay arguing that legal and institutional margins determine what can be said out loud is, transparently, an essay that makes this house’s product look necessary. The house benefits from the anxiety the essay describes, and it should say so before it says anything else.

Everything else follows from that admission. Nothing sold here is a substitute for a physician, and the questions at the centre of this essay are medical questions on which the house has no standing whatever. A room in which a question is safe is not a clinic, and a house that blurred the two would be selling something it cannot deliver. Nothing here is guidance to any woman, in Poland or Japan, about her own body, her own pregnancy or her own law. And the house has no view on Polish constitutional law, no business acquiring one, and no way of pretending that its commercial interest is not served by the finding it has just published.

It does not claim a position on abortion, for or against, and it would be a worse essay if it did. The reader who came looking for one has been given, instead, an argument about margins that holds whichever way the underlying question is answered.

It narrates no individual case. There are deaths in Poland after 2021 that were widely reported, that were the subject of national grief and of proceedings, and that this Library has chosen not to recount. The omission is deliberate and it is stated here so that it cannot be mistaken for ignorance. Those were real people with families who are still living, and this Library does not use them as illustration. An argument that requires a death to carry it is an argument that has not been made.

It does not adjudicate the composition of the Constitutional Tribunal. It reports that the question is disputed, that the European Court has twice treated the dispute as legally consequential, and that the Tribunal and its defenders reject that treatment, and it leaves the matter there.

It does not claim that prenatal diagnostic testing fell, that pregnancies were deferred or abandoned, or that the decline in births in Poland was caused by the ruling. Each of those is asserted widely. None of them could be verified here against official statistics or peer-reviewed work, and an essay about the discipline of evidence does not get to relax it at the point where it would be convenient.

It does not use the counts published by organisations that assist women to obtain terminations outside the hospital system, beyond noting that such a route exists. It makes no estimate of how many people in Poland end pregnancies in any year, because no such figure is available from a source this Library would accept.

It does not claim to know what Polish women think, what Polish doctors feel, or what it is like to be either. It has read a court record, a statistical series, a set of ministerial documents and some peer-reviewed research, and those are the only things it is entitled to speak from.

And it does not claim that Japan should copy anything, or that Poland is a warning aimed at anyone. What it claims is one narrow thing, offered to be argued with: that the reach of a law governing intimate life is not the reach the law states, that the difference shows up in the behaviour of the people who must apply it rather than in the text, and that a right which can be widened by a circular can be narrowed by one.