NOTE-256: Capabilities as Fundamental Entitlements: Sen and Social Justice

Contribution

The paper is Nussbaum’s case, addressed to Sen, for turning the capability approach from a space of comparison into a partial theory of justice. That requires a definite list of central capabilities, each owed to every citizen up to a threshold, as “a set of basic entitlements without which no society can lay claim to justice” (p. 5).

• §II. It argues that capabilities give rights language precision. Securing a right means securing the capability, which takes “affirmative material and institutional support, not simply a failure to impede” (p. 7).
• §III. It presents the ten Central Human Capabilities (pp. 12–13). It then attacks Sen’s treatment of freedom as a general good in Development as Freedom (pp. 16–20) and argues that pluralism itself requires a list (pp. 21–23).
• §IV. It argues that the social-contract tradition cannot handle care for dependants. Recasting Rawls’s primary goods as capabilities does not fix this unless the list has definite content that includes care, and unless there is an Aristotelian rather than Kantian conception of the person (pp. 24–31).

Key insight

The capability space alone decides nothing about justice. A theory of justice needs content: which capabilities are entitlements, and to what level.

Freedom cannot supply that content. Freedoms conflict. Some freedoms are trivial, such as riding without a helmet. Others are unjust by their very definition, such as marital-rape immunity or the right to hang a “No Blacks here” sign. So any just society must evaluate freedoms rather than promote freedom as such (pp. 16–19).

A short list, targeting capability rather than functioning, is also what respects reasonable pluralism. It asks citizens to endorse a few freedoms for political purposes, not autonomy as a way of life (pp. 22–23).

Assumptions

• Starting point. A conception of human dignity and of “a life that is worthy of that dignity”, with “truly human functioning” in Marx’s 1844 sense (p. 11). The list is justified as the “central requirements of a life with dignity”. The justification is not given here; it is referred to Nussbaum 2000a, ch. 1.
• Capability, not functioning, is the political goal (pp. 14, 22–23, 30). This leaves citizens free not to function, as with the Amish and voting (p. 14).
• Political liberalism in Rawls’s sense. The list is a free-standing “partial moral conception” and a “module” that holders of different comprehensive doctrines can attach to their own views (p. 14, fn 19). Justification is separated from implementation. The list is a basis for persuasion, and sanctions are licensed only for grave crimes against humanity (p. 15).
• Each person is an end. The capabilities are held for “each and every person”, none “as a mere adjunct or means to the ends of others” (p. 11).
• Non-tradeability. The ten are “mutually supportive and all of central relevance”. Neglecting one to promote others is “a failure of justice” (p. 11), although temporary practical priorities may be set.
• Pluralism safeguards (pp. 14–15):
  • (1) the list is open-ended and revisable;
  • (2) items are stated abstractly, for national specification;
  • (3) the list is political only, with no metaphysical grounding;
  • (4) capability, not functioning, is the target;
  • (5) the pluralism-protecting liberties are on the list itself;
  • (6) justification is kept separate from implementation.

Key results

What the paper argues, in order.

• What she grants Sen (§I, pp. 1–5):

  • the case against GNP per capita and against utility as measures of development and equality (distribution, heterogeneity, non-commensurability, adaptive preferences, and the neglect of agency and freedom);
  • the case against equality of resources (conversion differences: the pregnant woman, the child’s protein, the wheelchair user);
  • the special force of these arguments for gender justice.

  “Not surprisingly, I endorse these arguments. But I think that they do not take us very far in thinking about social justice” (p. 3).

  Elsewhere she also grants:

  • Sen’s effective argument that rights ideas are not uniquely Western (p. 9);
  • his emphasis on leaving room for choice (p. 9);
  • that he and she both hold that Rawls’s primary goods would do better as capabilities (pp. 24–25, 28);
  • that Sen’s practical writing on women is “extremely definite” (p. 21);
  • that her care argument and her Aristotelian conception of the person are “squarely in line with his ideas” (pp. 30–31).

• First criticism: no substance, no threshold (pp. 3–4). She makes three points.

  • Because of “Sen’s reluctance to make commitments about substance”, development guidance “remains but an outline”. It gives “no sense of what a minimum level of capability for a just society might be”. Its use is “typically comparative merely”, as in the Human Development Reports.
  • On equality, “Sen never says to what extent equality of capability ought to be a social goal, or how it ought to be combined with other political values” (p. 4).
  • Footnote 5: whether equality of capability is the right goal depends on the capability, for example voting versus basketball, or “access to the social bases of health” versus “the ability to be healthy”. That cannot be answered “without specifying a list”.

• Capabilities and rights (§II, pp. 5–10).

  • Why capabilities are clearer. Rights language hides disagreement: about the basis of rights, whether they are prepolitical, whether groups hold them, side-constraints versus goals, correlative duties, and what rights are rights to. The capability approach “tak[es] clear positions” on these questions (p. 6).
  • Securing a right. A right is secured only when the capability is. So a nominal right, for example women’s political participation under threat of violence, is not a right “in the sense that matters for judging that the society is a just one” (p. 7).
  • The constitutional comparison. The approach sides with the affirmative phrasing of the Indian Constitution, for example Art. 19, over the negative phrasing of the US Constitution (1st and 14th Amendments) (pp. 7–8). It reaches inequality inside the family (pp. 8–9).
  • What rights language still adds. It keeps a role in two respects. It imports the idea of an “urgent claim based upon justice”, and it emphasises choice and autonomy (p. 9). She lists “four important roles” but names only these two.

• The ten Central Human Capabilities (pp. 12–13, “the current version”), quoted closely:

  1. Life. Living to the end of a normal-length human life, not dying prematurely or before life is “so reduced as to be not worth living”.
  2. Bodily Health. Good health, including reproductive health; adequate nourishment; adequate shelter.
  3. Bodily Integrity. Moving freely; security against violent assault, including sexual assault and domestic violence; opportunities for sexual satisfaction and choice in reproduction.
  4. Senses, Imagination, and Thought. Using the senses, imagining, thinking and reasoning in a “truly human” way, cultivated by adequate education (literacy, basic mathematics and science). Using imagination in producing and experiencing works of one’s choice. Guarantees of freedom of political and artistic expression and of religious exercise. Pleasurable experiences, and avoiding non-beneficial pain.
  5. Emotions. Attachments to things and people; to love, grieve, long, feel gratitude and “justified anger”; not having one’s emotional development “blighted by fear and anxiety”. This means supporting the forms of association crucial to that development.
  6. Practical Reason. Forming a conception of the good and reflecting critically on planning one’s life. This entails liberty of conscience and religious observance.
  7. Affiliation.
     • (A) Living with and toward others, showing concern, imagining another’s situation, which means protecting institutions of affiliation and freedom of assembly and political speech.
     • (B) The social bases of self-respect and non-humiliation, and being treated as of equal worth. This entails non-discrimination on race, sex, sexual orientation, ethnicity, caste, religion and national origin.
  8. Other Species. Living with concern for and in relation to animals, plants and nature.
  9. Play. Laughing, playing, recreation.
  10. Control over one’s Environment.
      • (A) Political: effective participation in political choices, the right of political participation, and free speech and association.
      • (B) Material: holding property on an equal basis; seeking employment on an equal basis; freedom from unwarranted search and seizure; working “as a human being”, with practical reason and relationships of mutual recognition.

• The threshold conception, as far as this paper states it.

  • All ten are part of “a minimum account of social justice”. A society that does not guarantee them “to all its citizens, at some appropriate threshold level, falls short of being a fully just society, whatever its level of opulence” (p. 11).
  • Institution design starts from “what it would take to get citizens up to an acceptable level on all these capabilities” (p. 30).
  • The threshold is a minimum for a partial account of justice (pp. 5, 11). The paper leaves open what justice requires above it.

• Second criticism: Sen’s refusal to list, and the “puzzling tension” (pp. 15–16).

  • The tension. Sen writes as though health, education, political and civil liberties and free choice of occupation are “absolutely central and non-negotiable”. He gives liberty “considerable priority” without enumerating the liberties. His Human Development Report measures endorse health and education capabilities. Yet he “has conspicuously refused to endorse any account of the central capabilities”, so these examples “remain in limbo”.
  • The reason she reports. It is respect for democratic deliberation, per Sen at Bielefeld (fn 15).
  • Her objection. Sen goes further than she does. He suggests that endorsing central entitlements in international debate, as feminists do with CEDAW, inhibits democracy.

• Third criticism: “freedom” as an all-purpose good in Development as Freedom (pp. 16–20).

  • (a) Incoherence. “Some freedoms limit others”: campaign donations against the equal worth of the vote, pollution against a clean environment, landowners against land reform. And every freedom implies constraint on others (fn 16, citing Rawls 1971, p. 202, on the three-place structure of liberty).
  • (b) Not all freedoms are good. Even if the project were coherent, a just politics must say “some freedoms are central … and some are distinctly not”. Some are “positively bad”: campaign contributions, the freedom to pollute, gender-discriminatory landholding. “Gender justice cannot be successfully pursued without limiting male freedom” (p. 18): marital rape, workplace harassment. Some are trivial, like the helmet case.
  • (c) Abridging a core freedom is a distinctive cost. It is “a violation of basic justice”, independent of willingness-to-pay. Many would not pay much to vote; motorcyclists would pay a lot to ride without a helmet (pp. 18–19).
  • (d) Her reply to Sen’s reported answer. Sen answered that freedom is good per se, though it can be badly used, like male strength. She replies that “some freedoms include injustice in their very definition” (p. 19).
  • The conclusion. “Sen cannot avoid committing himself to a core list”. “All citizens are entitled to freedom understood as capability” would be “hopelessly vague. It would be impossible to say whether the society in question was just or unjust” (pp. 19–20).

• Why not leave list-making to each nation? (pp. 20–21).

  • Her answer, drawn from Sen. “some human matters are too important to be left to whim and caprice, or even to the dictates of a cultural tradition”. If capabilities have intrinsic importance, it is right to press nations on girls’ education and health care. Radical moral statement is justified before acceptance (Gandhi, King).
  • She attributes the same stance to Sen. “One cannot say, ‘I’m for justice, but any conception of justice anyone comes up with is all right with me’ … Sen, of course, does not say that” (p. 21).

• Fourth criticism: pluralism requires a list, and political liberalism prefers a short one (pp. 21–24).

  • (1) Real respect for pluralism requires non-negotiable protection of religious freedom, association and speech. It also requires health and education, which a person needs in order to choose a way of life. “we are both, in effect, making a list”, and the question is only what goes on it (p. 22).
  • (2) The second argument is Rawlsian, and “I am not certain that Sen would endorse it”. Endorsing freedom or autonomy as a general good, as Mill and Raz do, “shows deficient respect” for citizens of authoritarian religions. A short capability list, endorsed for political purposes, lets a Roman Catholic, an Amish citizen, or a woman who believes in seclusion endorse it without being told that only autonomous lives are worthwhile (pp. 22–23).
  • “I am not certain whether Sen is in this sense a comprehensive liberal like Raz, or a political liberal like Rawls and me” (p. 23).

• The social contract and care (§IV, pp. 24–31).

  • The diagnosis. Contract theories assume parties who are “free, equal, and independent” (Locke). Gauthier excludes people of unusual need. Rawls’s citizens are “fully cooperating members of society over a complete life” (Rawls 1980, p. 546, quoted at length on p. 27). Rawls puts animals outside justice (TJ 512).
  • Why patches fail. Postponing dependency to the legislative stage is “inadequate”. Adding care to the primary goods (Kittay), recasting them as capabilities, and adding health, work and emotional items still values care only as it serves Kantian moral personality (pp. 28–29).
  • Her proposal. A political conception of the person that is “more Aristotelian than Kantian”, both “capable and needy”, with capabilities as the analogue of primary goods. “this may well mean that the theory cannot be a contractarian theory at all” (p. 29).

• Conclusion (pp. 31–32). The capability approach does this work “only in virtue of having a definite content”. “the bare idea of capabilities as a space within which comparisons are made and inequalities assessed is insufficient.”

Claims

id	claim	strength	support
C1	Sen’s capability arguments (against GNP, utility and resources) are sound and especially apt for gender justice	moderate (agreement plus a summary of Sen’s arguments, which are not re-argued)	§I, pp. 1–4
C2	Without a list, the capability approach gives no minimum level of entitlement and no account of how far equality of capability matters	moderate (informal argument; fn 5’s examples show the answer varies by capability)	pp. 3–4, fn 5
C3	Securing a right means securing the capability, which requires affirmative support, not just non-interference	moderate (conceptual argument plus constitutional comparison)	§II, pp. 6–8
C4	The ten central capabilities are central requirements of a life with dignity	weak here (asserted; justification referred to Nussbaum 2000a, ch. 1)	pp. 11–13
C5	A just society guarantees all ten to all citizens at an appropriate threshold; neglecting one for another is an injustice	weak (stated as the view’s commitment; the threshold is not specified or derived)	p. 11
C6	Treating freedom as a general good is incoherent, because freedoms conflict and every freedom implies constraint	moderate (examples; fn 16 on the three-place structure of liberty)	pp. 16–17
C7	Some freedoms are positively bad or trivial, so any just society must evaluate freedoms by content	moderate (a series of examples: campaign money, pollution, landholding, marital rape, harassment, helmets)	pp. 17–19
C8	”freedom understood as capability” as a blanket goal is “hopelessly vague”: it cannot tell whether a society is just	moderate (follows from C6–C7)	pp. 19–20
C9	Sen’s own commitments (intrinsic importance of capabilities, definite stands on misogyny) already require a list	moderate (argument from Sen’s texts and practice; ad hominem in structure)	pp. 20–21
C10	Respect for pluralism requires a list of non-negotiable liberties plus health and education	moderate (informal argument)	pp. 21–22
C11	Political liberalism favours a short capability list over endorsing freedom/autonomy as a general good	moderate (Rawlsian argument; Sen’s endorsement explicitly uncertain)	pp. 22–23
C12	Contract theories cannot treat care for dependants as a basic issue of justice, even with capability-shaped primary goods	moderate (argument from Rawls’s texts; the conclusion “may well” be non-contractarian is tentative)	§IV, pp. 24–29
C13	Sen said, in discussion, that freedom per se is always good though it can be badly used; that his list-reluctance comes from respect for democratic deliberation; that pluralism is not his reason	weak as evidence of Sen’s view (Nussbaum’s report of oral remarks at Bielefeld, July 2001)	pp. 16, 19, fns 15, 17

Concepts

• Capability — “what people are actually able to do and to be” (p. 1). It is contrasted with functioning, the actual doing and being. The political goal is capability (p. 14).
• Central Human Capabilities — the ten items listed, as “central requirements of a life with dignity” (p. 11).
• Threshold — the “appropriate threshold level” of each central capability that a just society must secure to every citizen (p. 11). It is left for national specification (p. 14).
• Partial account of social justice — the list defines “a set of basic entitlements without which no society can lay claim to justice” (p. 5). It is not a complete theory.
• Fundamental entitlement — a claim based on justice whose abridgment is “a cost of a distinctive kind” (p. 18).
• Partial moral conception / module — Rawls’s terms, applied to the list: political only, attachable to different comprehensive doctrines (p. 14).
• Adaptive preferences — preferences “adjusted to their second-class status” (p. 1, citing Sen 1990, 1995).
• Comprehensive vs political liberal — Mill and Raz endorse autonomy as a general good. Rawls and Nussbaum endorse only a political list (pp. 22–23).

Connections

• Sen, “Equality of What?” (1979). This batch’s primary source. Nussbaum takes its argument against primary goods and resources, and the wheelchair example, as her starting point (pp. 3–4, 24). Footnote 4 adds a correction: the wheelchair case is social (missing ramps), not merely physical.
• SEP “The Capability Approach” (Robeyns & Byskov); IEP “Sen’s Capability Approach” (Wells); Robeyns, Wellbeing, Freedom and Social Justice (2017). These are the batch’s secondary accounts of the Sen–Nussbaum list dispute. This paper is the primary text on Nussbaum’s side of it. Sen’s reply, “Capabilities, Lists, and Public Reason” (Feminist Economics 10(3), 2004), is not reachable here (see the scoping report).
• Valentini, “A Paradigm Shift in Theorizing about Justice?” (2011); Zwolinski’s review of The Idea of Justice. Nussbaum’s complaint that Sen’s use of capabilities is “typically comparative merely” (p. 3) anticipates the comparative-versus-transcendental framing that The Idea of Justice (2009) later made explicit. Nussbaum wants a threshold, a partial transcendental standard. This is my connection, not hers: the paper predates the book.
• LIT-202 (Pettigrew, “Consent and the Formation of Preferences”). Both take adaptive preferences formed under unjust background conditions to disqualify preference-based measures: here, “entitlements that are to some extent independent of the preferences that people happen to have” (p. 1). Pettigrew refines when such preferences still count. Nussbaum sidesteps them by fixing entitlements.
• LIT-146 (SEP “The Moral Status of Animals”). Per NOTE-181, that entry lists Nussbaum’s capabilities as an objective-list theory applied to animals (Nussbaum 2006). This paper contains the seeds. Capability 8, “Other Species”, is on the list, and §IV objects that Rawls puts animals “outside the scope of the theory of justice” (p. 27).
• LIT-098 (Veit, “Health and Disease Concepts Cannot Be Grounded in Social Justice Alone”). A parallel in structure only. Veit argues that leaving “social justice” undefined empties a definition. Nussbaum argues that leaving the capability list undefined empties a theory of justice. There is no citation link.

Bearing on the record

• For the capability-approach cluster being assembled. This is the primary text for:

  • (a) Nussbaum’s ten capabilities, in the form current in 2001–2003;
  • (b) the charge that Sen lacks a list and a threshold;
  • (c) the argument that freedom is not a general good;
  • (d) the political-liberal case for a short list.

  Any record statement that “Sen refuses a list because of respect for democratic deliberation” should cite this paper as reporting Sen’s oral reply (fn 15), or cite Sen 2004. It should not present that as read in Sen.

• Page references. They are to the preprint. A THEORY citing specific pages should say so, or re-locate them in the version of record.

• No THEORY document on capabilities exists in the record yet (per the scoping report). None is contradicted.

• ML practice. It carries no instruction for ML practice. The list-versus-open-space debate is sometimes invoked in AI-alignment writing on “whose values”, but nothing in this paper addresses that, and I note it only as a possible later link.

Note of 2026-10-16: THEORY-025 now cites LIT-288 as a source.

Limitations

• The list is presented, not justified here. Justification is referred to Women and Human Development (2000a).
• The threshold is named but not specified or derived. How thresholds are set, whether they are sufficientarian, and what justice requires above them are all left open in this paper.
• The argument is by example. The freedom critique rests on examples (campaign money, pollution, marital rape, helmets). It shows that unrestricted freedom-promotion is untenable. It does not show that Sen’s actual position is unrestricted: Nussbaum herself grants that Sen ranks some freedoms (p. 16). Some of the force depends on reading Development as Freedom’s rhetoric literally.
• Sen’s side appears only as Nussbaum reports his conference remarks.
• The social-contract section argues against Rawls, not Sen. Its tie to Sen is the shared view that primary goods should be capabilities.
• The paper uses “four important roles” but names two (p. 9). It also has the section-numbering slips noted under corrections.

Open questions

• Do the preprint and the version of record differ? In wording, in the list, or in the section numbering (fn 5’s “Section V”)? Settled by reading the published article.
• Sen’s reply. How does Sen’s published reply (2004, “Capabilities, Lists, and Public Reason”) answer the four criticisms, especially the freedom-is-not-a-general-good argument? Unreachable here. Check the SEP and Robeyns 2017 accounts.
• Is Sen a comprehensive or a political liberal? Nussbaum leaves this open (p. 23). The secondary sources in this batch may take a view.

Corrections to the seeded skim

• none (there was no seed or dossier)

• The threshold conception is thinner here than the task description and the scoping report imply. The paper states it three times:

  • that a just society guarantees all ten “at some appropriate threshold level” (p. 11);
  • that institutions should be designed around “what it would take to get citizens up to an acceptable level on all these capabilities” (p. 30);
  • the complaint that Sen “says little about how a threshold level of each capability would be constructed” (p. 15) and gives “no sense of what a minimum level of capability for a just society might be” (p. 3).

  The paper never says how a threshold is to be set, who sets it, or what happens above it. Specification is left to each nation’s “citizens and their legislatures and courts” (p. 14). It does not use the word “sufficientarian”. The fuller account is in Women and Human Development (2000a, ch. 1), which this paper cites but does not reproduce.

• “Political-liberal overlapping consensus” is attributed accurately, but it is a secondary argument here. The list is called a free-standing “partial moral conception”, a Rawlsian “module” (p. 14). The Rawlsian argument against endorsing freedom as a general good is offered as the second of two pluralism arguments (pp. 22–23), and Nussbaum says “I am not certain that Sen would endorse it” (p. 22). The phrase “overlapping consensus” occurs only on p. 24 and in footnote 19, about Women and Human Development.

• The scoping report’s “Sen is too vague to yield a theory of justice” is right, with two qualifications. The “hopelessly vague” charge is aimed at one specific formulation, “All citizens are entitled to freedom understood as capability” (pp. 19–20). It is not aimed at Sen’s work as a whole. Nussbaum repeatedly grants that Sen’s own practice is “extremely definite” (p. 21).

• Internal slips in the preprint.

  • Footnote 5 promises an argument “in Section V”, but the paper has only §§I–IV. The argument meant is in §III.
  • “Nussbaum (forthcoming d)” is cited for the sanctions view (p. 15), but the reference list’s “forthcoming d” is “The Complexity of Groups”. “Forthcoming e”, cited in footnote 14, has no entry.
  • Whether these were fixed in the journal version is unverified.

• Sen’s replies are reported only through Nussbaum. The paper reports three things Sen said in the Bielefeld discussion (July 2001):

  • that his reluctance to list comes from respect for democratic deliberation (p. 16, fn 15);
  • that “freedom per se is always good, although it can be badly used”, illustrated by an analogy with male strength (p. 19);
  • that pluralism is “not his concern” as a reason against a list (fn 17).

  These are her unrecorded reports of oral remarks, not Sen’s text.