Are Women-Only University Admissions Reparative Justice or Reverse Discrimination? Bentham, Rawls, Sandel, and Nozick

The controversy over women-only places in university admissions is a collision between two ideas of justice.

On one side is the argument that Japanese society excluded women from scientific education and professional careers for generations. On this view, a women-only admissions route is a form of reparation for women, and men should accept some present cost as part of correcting a structure from which men collectively benefited. Giving everyone the same examination today does not erase differences in expectations, guidance, visible role models, university culture, or later working conditions.

On the other side, former House of Representatives member Shiori Yamao argued in a July 2026 post that young men raised to expect sex equality suddenly encounter a barrier defined by sex. She objected that the burden of discrimination created by an older generation is being transferred to people who did not create it, and warned that importing an inherited conflict into a younger generation may be counterproductive.

Both claims contain something morally important. Women's present opportunities are shaped by a history that began before today's applicants were born. Yet the male applicant who may lose a place is not the official, employer, or professor who created that history.

Who should bear the cost of repairing structural injustice?

This article approaches that question through four political philosophers: Jeremy Bentham, who judges institutions by their consequences for aggregate welfare; John Rawls, who criticizes sacrificing individuals to an aggregate and asks what rules people would choose behind a veil of ignorance; Michael Sandel, who criticizes the image of a self detached from community and history; and Robert Nozick, who warns that individuals may not be used as instruments for collective goals.

The four theories do not yield a single automatic answer. They reveal that people who all appeal to “equality” may be trying to protect different things.

Two questions that must be separated: goal A and method B

Debate becomes confused when a social objective and a particular policy are treated as one proposition.

Question A is whether increasing the proportion of women in science and engineering is itself a legitimate goal for universities and society. Does a sharply uneven sex ratio indicate a loss of equal opportunity, educational diversity, or future talent? Or, if discriminatory barriers have been removed and individuals choose freely, is an unequal ratio morally unobjectionable?

Question B is whether, even if goal A is justified, reserving places that only women may apply for is a just means. Could the same goal be pursued through information, scholarships, earlier science education, role models, university reform, or changes in scientific careers? Is excluding male applicants from designated places necessary and proportionate?

Agreement with A does not imply agreement with B. A person may consistently say: “More girls should experience science as a real option, but universities should change earlier education and campus conditions rather than exclude men from an admissions category.” Conversely, if no persuasive reason supports A, even a highly effective quota lacks a justified objective.

Goal A: increase women in STEM	Method B: women-only places	Position
Yes	Yes	A limited quota may be needed to break a self-reinforcing imbalance and improve substantive opportunity.
Yes	No	The objective is legitimate, but less exclusionary measures should be preferred.
No	No	A freely produced sex ratio should not be engineered by universities.
Conditional	Conditional	The answer depends on evidence of barriers, effectiveness, scale, duration, and alternatives.

Bentham asks about the gains and losses produced by both A and B. Rawls asks whether the purpose and the method are fair from every possible social position. Sandel examines A through the purpose and history of the university community. Nozick warns that even a desirable A may not license a B that violates individual rights.

What are women-only admissions intended to change?

A women-only admissions route does not ordinarily grant admission solely because an applicant is a woman. In Japan, many such policies establish a women-only category within comprehensive or school-recommendation admissions and then assess grades, activities, motivation, interviews, or other criteria within that category. Nevertheless, reserving capacity that men cannot contest is a direct distinction by sex.

Kyoto University introduced women-only places in selected science and engineering admissions for the 2026 academic year. It presents the policy as a response to very low female representation and as a way to create an educational environment in which students with different attributes and backgrounds learn together. Its official 2026 results report 96 applicants for 39 designated places and 29 final acceptances. Some units did not fill the announced capacity, showing that the route is not an unconditional guarantee of admission.

National policy also encourages a substantial increase. Japan's Priority Policy for Women's Empowerment and Gender Equality 2026 set a goal of raising the share of women in university engineering from 17.9 percent, based on the 2025 School Basic Survey, to 36 percent by 2040.

But a low proportion alone does not prove discrimination or the necessity of a quota. The result may reflect interacting causes: preferences, expectations at home and school, access to prerequisite subjects, occupational images, future working conditions, and the psychological burden of being a small minority. If the main cause arises before university applications or after graduation, changing admissions alone may adjust a visible number without removing the cause. Yet the reverse causal loop is also possible: without a critical mass at university, visible role models and institutional culture may never change.

A responsible policy therefore has to answer four questions:

1. At what stage does what barrier operate?
2. Does a women-only route actually weaken that barrier?
3. Can a similarly effective result be achieved through a less exclusionary measure?
4. Why is the burden placed on the individuals who lose access to designated places?

Without these questions, a quota can become numerical symbolism. Without asking whether competition was shaped before examination day, however, formal neutrality can preserve accumulated inequality.

Question A: should the proportion of women in STEM be increased?

Before judging the instrument, we need to examine why a higher female share might be desirable.

Argument for A: repairing the continuing effects of exclusion

The first justification is reparative. Women in Japan did not historically receive equal encouragement or access to higher education and technical careers, and current representation cannot be assumed to be independent of that history. Identical admissions rules today do not erase accumulated differences in expectations, information, role models, and professional conditions.

This argument may support a public responsibility to widen women's real options. But it does not by itself settle method B. Society's responsibility to repair an inherited structure is distinct from concentrating its cost on a particular group of current male applicants.

Argument for A: diversity has educational and social value

The second justification is diversity. Kyoto University's own explanation emphasizes learning among students with different backgrounds.

Diversity can mean at least three things. Representative diversity concerns who receives access to publicly supported institutions that train future researchers, engineers, and policymakers. Educational diversity concerns whether students encounter experiences and assumptions unlike their own, and whether members of a small minority can participate without isolation. Epistemic diversity concerns which problems a research group notices in the first place. Experience can affect the questions asked in product design, medicine, information technology, and public policy.

None of this means that every woman embodies a single “female viewpoint.” Women differ by class, region, culture, disability, and family background, while men also have diverse experiences. A university that changes only its sex ratio without enabling genuine intellectual disagreement may achieve demographic appearance without educational diversity.

Argument for A: low representation can reproduce itself

A third argument points to a feedback loop. When classrooms, laboratories, faculty, and future occupations contain very few women, a girl may know that she is legally allowed to choose science yet find it harder to imagine herself living that choice. The absence of peers can operate as social information: perhaps the route is unusually risky, lonely, or not meant for someone like her.

Japan's White Paper on Gender Equality cites a survey in which 33.8 percent of girls taught mathematics or science by at least one female teacher in junior high described themselves as science-oriented or somewhat science-oriented, compared with 22.5 percent of girls taught both subjects by men. This association does not prove causation, but it is consistent with the possibility that visible examples influence self-conception.

A quota may therefore be defended not as compensation for a deficiency in individual women, but as a temporary intervention to break an equilibrium in which women stay away because women are absent. The claim remains empirical. If campus culture and later work remain unchanged, or if quota entrants are stigmatized, increasing admissions may fail to break the loop.

Argument against A: the divergence is formed before university admissions

A strong objection says that the disparity largely appears at the stage of high-school subject selection. Material published by Japan's Ministry of Education reports that 66 percent of high schools separate humanities and science tracks, and that by the third year 27 percent of boys but 16 percent of girls take a science track. By the time university applications open, the populations who have taken physics and advanced mathematics and can realistically apply to engineering are already different.

From this perspective, a low university share need not result from women being rejected at a higher rate within the same applicant pool. It may reflect choices made upstream. If so, information in junior high school, subject access, counseling, the difficulty of changing tracks, family expectations, and occupational reform are closer to the cause than university admissions.

This objection does not conclusively defeat quotas. More women at university may change the aspirations of younger students, so a downstream intervention can feed back upstream. That effect, however, should be demonstrated rather than assumed.

Argument against A: average differences in subject achievement or interest may matter

Another objection is that discrimination and role models may not explain the entire ratio. Average differences in achievement profiles, interests, and occupational preferences may influence choices. If people choose the field in which they have the greatest comparative strength, an unequal group outcome is not automatically unjust.

In Japan's PISA 2022 results, boys scored nine points above girls on average in mathematics, while girls scored seventeen points above boys in reading. A girl who is capable in mathematics but even stronger in language may rationally select a field reflecting that comparative advantage.

The data do not justify the conclusion that men are innately suited to science and women to the humanities. Group distributions overlap extensively, so an average cannot determine an individual's capacity or admissibility. Scores at age fifteen reflect education, confidence, expectations, study time, and testing behavior as well as any biological factor. Sex gaps also vary across countries: in PISA 2022, boys led in mathematics in 40 systems, girls led in 17, and 24 showed no statistically significant difference. Social conditions clearly affect the pattern.

Moreover, the ministry materials describe no comparably large sex gap in high-school mathematical and scientific literacy while showing a much wider difference in track choice. Achievement may explain part of the divergence, but it does not necessarily explain a nearly two-to-one difference in selection.

The proper conclusion is that average differences are a hypothesis to investigate, not a verdict about individuals or a complete explanation of representation.

Argument against A: equal freedom need not produce equal numbers

The most fundamental objection distinguishes the removal of barriers from the engineering of outcomes. If people have adequate information and genuine educational opportunities, a ratio other than fifty-fifty may reflect legitimate differences in preference.

Turning a numerical target into the measure of progress can also imply that a woman who selects engineering is advancing equality while one who chooses the humanities or care work is somehow behind. A policy intended to respect women's freedom may impose an official image of the correct modern woman.

Supporters reply that preferences are formed socially and cannot simply be treated as independent choices. Opponents answer that if every choice is indefinitely reinterpreted as structural conditioning, women's own stated decisions will never be recognized as authentic. The philosophical issue in A is therefore not whether choices are influenced—they always are—but when the conditions are fair enough for a choice to count as genuinely one's own.

Question B: if increasing participation is justified, why use a quota?

Policies commonly grouped under affirmative action vary greatly in their effect on others' rights.

The first level includes outreach, science programs for girls, encounters with women researchers, scholarships, mentoring, and campus support. These measures encourage application and persistence without directly barring men from places.

The second level lets all candidates enter the same process but considers sex as one element in individualized evaluation or as a tiebreaker between similarly qualified candidates. Sex affects selection, but no applicant is categorically excluded at the outset.

The third level reserves a fixed number of places for women, so men cannot apply for them. Many Japanese policies called “women's quotas” resemble this stronger form.

The difference between support, flexible preference, and fixed reservation matters. So does the difference between converting existing places and adding capacity. Converting places can intensify competition in the general pool. Adding capacity may reduce that direct burden but requires teachers, facilities, and public funds. Justice concerns not only an objective but the allocation of its cost.

Formal equality and substantive equality

Formal equality requires that similarly situated people be treated by the same rule. If men and women sit the same examination and face the same score threshold, sex does not directly determine access.

Substantive equality asks whether people arrived at that competition with a genuinely fair range of options. A student repeatedly told that science is “for boys” and a student for whom engineering was always presented as normal may face the same paper at eighteen without having enjoyed the same opportunity.

Affirmative action is an active measure designed to increase participation by a group historically excluded or disadvantaged. Article 2 of Japan's Basic Act for Gender Equal Society expressly defines positive action as providing, within necessary limits, opportunities to one sex in order to improve disparities in opportunity between women and men.

Japanese law therefore contemplates both nondiscrimination and active correction. The real question is what distinctions are justified, necessary, and limited enough to advance equality rather than replace one arbitrary barrier with another.

Opportunity itself has several layers. There is equality of the application rule; equality of preparation, including schooling, advice, resources, and role models; and equality in the practical ability to use an option after admission, including belonging, safety, and viable working conditions. A quota at the final gate affects only one layer.

Substantive equality does not mean that every outcome must be numerically identical. A disparity can be evidence prompting investigation, but it is not proof of a single cause. Equally, calling a result “personal choice” without examining how the options were formed can transfer social constraints back onto individuals.

Bentham's utilitarianism: does the policy increase total welfare?

Jeremy Bentham evaluates an act or institution by the balance of pleasure and pain it produces. In the principle associated with Bentham, the good of each affected person enters the calculation; the happiness of a prestigious person does not count more merely because of status.

For Bentham, good intentions are insufficient. A quota may be motivated by equality but produce stigma, resentment, and institutional complacency. If so, it can fail on utilitarian grounds. Conversely, a limited cost to some applicants may be justified if the policy substantially widens talent, improves education and research, changes younger students' options, and produces large long-term gains.

Possible benefits include reduced isolation, more visible role models, wider perspectives in research and design, use of talent discouraged by social expectations, and perhaps less occupational segregation. Possible costs include lost opportunities for male applicants, suspicion that women were admitted without merit, the reduction of women's achievements to their sex, intensified conflict between young people, and numerical gains without later institutional reform.

The hard problem is that these effects differ in form and timing. A denied place is an immediate and concentrated loss to a known individual. A future innovation or role-model effect is dispersed and uncertain. Aggregating them can hide distribution: a large total gain may coexist with a severe, uncompensated burden on a few.

Utilitarianism therefore demands measurement rather than slogans: application behavior, academic progress, persistence, graduation, subsequent careers, classroom climate, stigma, and effects on general-pool competition. It also requires comparison with alternatives. Even if a quota produces benefits, it is not the best method if earlier education or nonexclusive support produces more benefit with less harm.

Rawls's veil of ignorance: would we choose the rule without knowing our sex?

John Rawls objects to the idea that an individual's basic prospects may simply be traded for a larger social sum. In the original position, people choose principles of justice behind a veil of ignorance: they do not know their sex, class, talents, family wealth, social status, or conception of the good.

Applied to admissions, imagine not knowing whether you will be a girl whose scientific aspiration is weakened by stereotypes, a boy excluded from reserved places despite excellent results, a woman admitted through a quota and then stigmatized, or a student of either sex whose poverty prevents application altogether.

Rawls's first principle protects equal basic liberties. His second regulates social and economic inequalities: offices must be open under fair equality of opportunity, and inequalities must benefit the least advantaged. Fair equality of opportunity is stronger than a formally identical examination. People with similar ability and motivation should have comparable prospects regardless of social starting point.

This gives Rawlsian force to A. If girls' scientific options are narrowed by circumstances unrelated to their capacity, society should not be satisfied with a neutral examination at the end of an unequal developmental process.

But B is more difficult. A women-only category improves opportunity for some women while reducing access for men in the same generation. It may also miss the least advantaged if beneficiaries are predominantly affluent women while poor or rural applicants of both sexes remain excluded. Sex is not the only morally arbitrary circumstance.

Behind the veil, people might accept a temporary, evidence-based corrective if it protects them against being born into a systematically constrained group. They are less likely to choose an indefinite fixed quota with no test of necessity, no attention to class and region, and no protection for those who bear concentrated costs.

A Rawlsian policy should therefore define the disadvantage precisely, consider multiple sources of unequal opportunity, use the least restrictive effective instrument, publish reasons and results, and include review and termination conditions. Rawls does not mechanically endorse or reject every quota. He changes the question from “who deserves the seat in isolation?” to “what system of opportunity could everyone accept without knowing where they will stand?”

Sandel's communitarianism: can we detach ourselves from inherited responsibility?

Michael Sandel and other communitarian critics argue that Rawls's choosing self is too detached from the communities, histories, and commitments that partly constitute identity. We do not begin as unencumbered individuals and only later choose every obligation. We are born into families, languages, political societies, and institutions whose achievements and wrongs we inherit.

This perspective strengthens the argument that a present university cannot say, “The exclusion occurred before us, so it has nothing to do with us.” An institution has a history and a public purpose. A national university is not merely a market that distributes private credentials to the highest score. It may understand its common good as producing knowledge for society, cultivating citizens and professionals, and correcting exclusions that distort its intellectual community.

Sandel's discussion of affirmative action also highlights that “merit” is defined by a social purpose. A university decides which abilities and experiences serve the kind of institution it seeks to be. Admission is not a cosmic verdict on a person's worth. If a university reasonably regards a more diverse class as part of its educational mission, diversity can enter admissions without declaring that a rejected applicant is less worthy as a human being.

But communitarianism is not a blank cheque for administrators. Which community's good is being defined, by whom, and after what public debate? “Diversity” can become a managerial slogan that substitutes a visible ratio for difficult reform. A community can also oppress in the name of tradition or unity.

Historical responsibility must be shared through political reasoning. If society owes repair, costs should be distributed through education budgets, school reform, scholarships, campus change, and working conditions—not placed almost entirely on a narrow group of eighteen-year-old men. We inherit responsibility to improve institutions, not personal guilt for acts we did not commit.

Nozick's libertarianism: may individuals be used for a social pattern?

Robert Nozick begins from strong individual rights. His political philosophy rejects using persons as resources for an aggregate social goal. A distribution is not just merely because it matches a preferred pattern such as equality, need, or a demographic ratio. Justice depends on the history of acquisition and transfer and on the rectification of identifiable injustice.

From this perspective, goal A is suspect if it is defined simply as producing a particular sex ratio. Individuals are not counters to be arranged into a socially attractive pattern. If a male applicant satisfies the relevant academic criteria, barring him from designated publicly funded places because society wants a different aggregate can treat him as a means rather than an end.

Nozick's famous challenge to patterned distribution is that free individual choices continually disturb the pattern. Preserving a predetermined ratio therefore requires ongoing intervention. Applied to education, if women and men make different voluntary choices, the state would have to keep correcting their outcomes to preserve the target.

Nozick does leave room for rectification where holdings or opportunities result from past injustice. But rectification requires a connection between the injustice and the remedy. A vague appeal to historical discrimination does not automatically show why one present applicant should pay another present applicant. He would likely favor removing current legal barriers, remedying identifiable discrimination, expanding capacity, or assisting disadvantaged individuals through nonexclusive means over a permanent group quota.

Nozick's approach has its own weakness. Present “rights” and entitlements arise within an institutional history. If earlier injustice shaped schools, wealth, and professional networks, simply protecting today's distribution can preserve the effects of the past. His demand for historical analysis can therefore support corrective action as well as constrain it; the dispute is how specific the causal connection must be and what remedy respects current individuals.

Is a women-only quota unconstitutional in Japan?

Claims that the policy is obviously constitutional and claims that it is obviously unconstitutional both move too quickly.

Article 14 of the Constitution of Japan guarantees equality under the law and rejects discrimination because of sex. Article 4 of the Basic Act on Education provides equal educational opportunity and prohibits discrimination in education on grounds including sex.

At the same time, equality does not necessarily forbid every different treatment. Constitutional analysis generally asks whether a classification has a legitimate purpose and whether the means bear a reasonable, proportionate relation to that purpose. Japan's gender-equality statute recognizes positive action, and Article 4 of the Convention on the Elimination of All Forms of Discrimination against Women states that temporary special measures aimed at accelerating de facto equality are not discrimination, while requiring that they be discontinued when equality objectives are achieved.

Legal analysis is clearer in four stages:

1. Classification: the route directly distinguishes by sex.
2. Legitimate aim: the university must identify more than a preferred number—for example, a demonstrated barrier to fair opportunity or a concrete educational purpose.
3. Suitability and necessity: it must show that the policy can address the identified problem and compare less restrictive measures.
4. Proportionality and time limit: the scale, effect on general admissions, individualized assessment, review, and termination conditions matter.

Japanese courts have not supplied a single definitive judgment resolving every current university women-only scheme. The constitutional issue therefore turns on institutional facts and policy design. A carefully limited, evidence-based temporary measure is not legally equivalent to an indefinite automatic reservation justified only by a numerical target.

What European decisions actually rejected

“Europe found quotas unconstitutional” compresses distinct legal systems and cases into a misleading slogan.

Finland's National Non-Discrimination and Equality Tribunal found that sex quotas used in a University of Jyväskylä student-selection process violated the Finnish Equality Act. The tribunal's account is directly relevant because it concerns university admissions.

The Court of Justice of the European Union has also examined positive action in employment. Kalanke rejected automatic and unconditional priority for women in promotion where candidates were equally qualified. Marschall permitted a preference that preserved an individual saving clause allowing objective circumstances to be considered. Abrahamsson rejected a scheme capable of preferring a sufficiently qualified woman over a more qualified man.

Those employment cases do not directly decide Japanese university admissions. They nevertheless show why automaticity, differences in qualification, individual review, and proportionality matter.

European law does not reject positive measures as such. Article 23 of the EU Charter framework recognizes that equality need not prevent advantages for an underrepresented sex. The general lesson is not “never consider sex,” but “design corrective action precisely enough that individuals are not mechanically displaced.”

The strength and weakness of “society owes women reparation”

The moral strength of reparative language is that today's competition did not begin in a vacuum. Treating representation as nothing but accumulated private preference erases the history that helped form those preferences and opportunities.

But “reparation” can refer to three different projects. The first is personal compensation by a wrongdoer to a victim. Current male and female applicants cannot be assigned those identities as groups. The second is institutional repair of a disadvantage that past rules have carried into the present. This is the strongest reparative basis for policy even when individual wrongdoers cannot be identified. The third is investment in a preferred future, such as broadening scientific perspectives. That may be worthwhile, but it is not identical to compensation for a past violation.

Once someone says that male applicants must simply accept the policy, three difficulties arise.

First, the people who caused the wrong and the people made to bear the cost do not coincide. If society shares responsibility, public funding, school reform, scholarships, universities, and employers should also share the burden.

Second, past victims and present beneficiaries do not coincide. A well-resourced woman may benefit while a poor male or female applicant remains excluded. Group averages cannot fully represent individual disadvantage.

Third, reparation needs an endpoint. Is it complete at thirty, forty, or fifty percent? When application rates converge? When later employment changes? A measure without a termination condition can turn sex into a permanent creditor-debtor relation.

The proper inheritance is not personal guilt. It is responsibility for building fairer institutions.

How the four theories evaluate the two questions

Theory	Standard	Goal A: increasing participation	Method B: women-only places
Bentham's utilitarianism	Net welfare of everyone affected	Supports A if diversity, talent use, and role-model effects generate real long-term gains.	Supports B only if its net benefit exceeds alternatives after counting lost opportunities, stigma, and conflict.
Rawls's justice as fairness	Rules acceptable behind the veil of ignorance	Can support removing structural barriers and achieving fair equality of opportunity.	Requires attention to sex, class, and region; evidence of necessity; minimum scope; review; and an endpoint.
Sandel's communitarianism	Shared purposes, history, and obligations of a community	Can treat diversity and institutional repair as part of a university's common good.	Requires public justification and a distribution of reparative costs broader than young male applicants.
Nozick's libertarianism	Individual rights and the history of entitlements	Is skeptical of an official demographic pattern, while allowing rectification of specific injustice.	Strongly resists fixed exclusions that use individuals as means to a collective ratio.

Bentham asks whether the consequences are good. Rawls asks whether people could accept the system without knowing where they will stand. Sandel asks what history and purpose the university community shares. Nozick asks what may not be done to an individual even for a desirable result.

Consider a hypothetical public department with one hundred places, twenty reserved for women for five years, and eighty open to everyone. The university cites low female application and minority isolation but has not yet presented outcome data or comparison with alternatives.

Bentham would demand evidence of gains and harms and would end the trial if benefits fail to materialize. Rawls would ask why twenty, whether poverty and region are ignored, and why the burden is concentrated in the general pool. Sandel would ask whether a diverse learning community is genuinely part of the institution's mission and whether citizens participated in defining that mission. Nozick would object most strongly to forbidding an individual from competing for publicly funded places to create a group pattern.

The disagreement is therefore not only over whether women should increase. It concerns evidence, acceptability from every position, the purpose of an institution, and the limits of collective power.

Seven conditions a university should meet

1. Identify causes, not only a ratio

Show whether the barrier appears before application, within selection, after admission, or in later careers. Each location calls for a different remedy.

2. Measure the policy's effects

Track not only entrants but applications, general-pool effects, academic progress, persistence, graduation, further study, employment, belonging, and stigma. Change or end a policy that does not work.

3. Compare less restrictive alternatives

Examine earlier guidance, parental and teacher outreach, scholarships sensitive to income and region, hiring and mentoring, anti-harassment measures, laboratory culture, and flexible individualized assessment.

4. Make standards and aggregate results transparent

Publish criteria, application ratios, and anonymized outcomes. Leaving the suspicion that “women are admitted easily” unanswered harms the very students the measure is supposed to support.

5. Use the minimum necessary scale

Prefer a small, testable intervention to a large unexplained conversion of general places. Where feasible, expand educational capacity so the cost is not concentrated on particular applicants.

6. Set review dates and an exit condition

Define when the policy will be reconsidered and what improvement would make it unnecessary. Corrective action should not harden into a permanent entitlement.

7. Do not treat dissent as prejudice

A man who objects to losing eligibility need not oppose women's participation. A supporter need not be indifferent to men. Turning institutional criticism into a judgment of character prevents evaluation.

NOW IN QUESTION: must equality sometimes notice sex?

Women-only admissions notice sex in order to pursue a society in which sex should matter less. That is the policy's unavoidable paradox: present classification in pursuit of future neutrality.

The paradox cannot be resolved by telling men to accept a moral debt. It requires evidence of a barrier, proof that the measure can address it, comparison with gentler alternatives, proportional allocation of costs, and a defined period of review.

Nor is it enough to say that everyone now takes the same test. Identical rules on one day may ignore eighteen years during which options were shaped differently.

On question A, there is a legitimate public interest in ensuring that women with the ability and interest to pursue science do not abandon it because of stereotypes, missing information, or avoidable institutional conditions. Educational diversity and role models can have value. But a fifty-fifty ratio is not an absolute measure of justice, and policy should not override free choices merely to perfect a statistic. The aim should be to prevent sex from unfairly narrowing real options.

On question B, support for A does not automatically establish a quota. Earlier educational intervention, scholarships, information, campus reform, and changes in scientific work often lie closer to the causes and impose fewer direct restrictions. If a quota is used, the university bears the burden of explaining why those measures are insufficient, why this scale is necessary, what the policy improves, and when it will end.

My conclusion is therefore conditional. Increasing women's real capacity to choose STEM has substantial justification. A women-only admissions category is neither always just because its aim is equality nor always impermissible because it uses sex. Only a limited, temporary, evidence-based measure that survives comparison with less restrictive alternatives deserves consideration as a tool of substantive equality.

If society truly owes repair, it must not push the entire invoice across an admissions desk to the next young man in line. Earlier education, families, universities, research institutions, employers, and public budgets should share the work.

Finally, lower Rawls's veil again. You may be the woman whose path into science becomes imaginable. You may be the man who loses by one point after the general pool contracts. You may be a poor student of either sex who never reaches this competition. You may enter through a special route and have your competence questioned for years.

Would we choose the same institution without knowing which person we will be?

The question is not simply whether to favor women or protect men. It is whether we are serious enough about justice to repair one inequality without carelessly placing someone else inside a new one.

References

• Kyoto University: Special Admissions for Women
• Kyoto University: Applications and Selection Results
• Japan Cabinet Office: Priority Policy for Women's Empowerment and Gender Equality 2026
• Japan Cabinet Office: Girls' Educational Choices and Measures to Widen Them
• Japan Ministry of Education: High School Education Working Group Materials
• OECD: PISA 2022 Results for Japan
• OECD: Equity in Education in PISA 2022
• The Constitution of Japan, Article 14
• Basic Act on Education
• Basic Act for Gender Equal Society
• Convention on the Elimination of All Forms of Discrimination against Women
• Kyoto University Press: Considering Women-Only Admissions
• National Center for University Entrance Examinations: Are Women-Only Places Discriminatory?
• Finland Equality Tribunal decision on sex quotas in university selection
• Court of Justice of the European Union: Kalanke
• Court of Justice of the European Union: Marschall
• Court of Justice of the European Union: Abrahamsson
• EU Charter Article 23 and positive action
• Shiori Yamao's post on women-only university places
• Jeremy Bentham, Stanford Encyclopedia of Philosophy
• Original Position, Stanford Encyclopedia of Philosophy
• Communitarianism, Stanford Encyclopedia of Philosophy
• Robert Nozick's Political Philosophy, Stanford Encyclopedia of Philosophy
• Michael Sandel, Arguing Affirmative Action