The capabilities approach: Sen, Nussbaum, and justice as the freedom to live a valued life

Intuition Beginner

How should we measure whether a country is doing well? For most of the twentieth century the default answer was gross domestic product — the total value of goods and services a country produces. But GDP says nothing about who gets the money, whether people are healthy, or whether they are free. A country can post rising GDP while average lifespan falls, as the United States did across the years when opioid and suicide epidemics cut life expectancy at birth. The GDP answer measures wealth, not lives.

A second answer says: measure happiness. If people report themselves satisfied, the country is doing well. But people adapt to deprivation. A person who has never had clean water, schooling, or political voice may report satisfaction, because they have lowered their expectations to match what once seemed possible. Philosophers call this the problem of adaptive preferences, or more sharply, the happy-slave problem. Happiness is a poor measure when deprivation reshapes what people allow themselves to want.

Amartya Sen, an Indian economist who won the 1998 Nobel Memorial Prize, and Martha Nussbaum, an American philosopher, proposed a third answer. A country is doing well when its people have real freedoms — capabilities — to live lives they have reason to value. The question shifts from "do people have money?" or "are they happy?" to "what can they actually do and be?": be healthy, be educated, vote, move freely, love without fear. This capability approach is the framework behind the United Nations Human Development Index.

Visual Beginner

The picture reads left to right. On the left sit resources — income, food, schools, clinics. An arrow passes through a box labelled "conversion factors", where personal traits (disability, age, sex), social conditions (norms, laws, discrimination), and environment (climate, roads, water) reshape what those resources yield. On the right are two columns: functionings (what the person actually does and is — being nourished, being literate, voting) and capabilities (the set of functionings they are really free to achieve). A band along the bottom marks Nussbaum's ten central capabilities; a lower strip marks the Human Development Index and the Multidimensional Poverty Index as policy instruments.

The load-bearing image is the middle box: the same resource produces different capabilities in different people, depending on conversion factors. Strip away the box and the approach collapses back into a resource-based view.

Worked example Beginner

The canonical application is India versus China in the Human Development Index.

Step 1. In 1990, when the United Nations Development Programme first published the index, China scored roughly 0.50 and India roughly 0.40. Both fell in the "low human development" band.

Step 2. By 2022, China had climbed to about 0.77, the "high human development" band, while India had reached about 0.63, still "medium". The gap had widened even though both economies had grown.

Step 3. The gap is not accidental. Between 1949 and 1976, before market reforms, China invested heavily in basic literacy, primary healthcare, land reform, and public sanitation. India, after independence in 1947, underinvested in the same capabilities for decades, especially among women and lower castes.

Step 4. Sen's reading is direct. China's later growth rested on a capability base — literate, healthy workers — that India had not yet built. Growth without prior capability investment is fragile and uneven; capability investment both enables growth and widens what people can do and be.

What this tells us: development is the expansion of capability, and GDP growth is one downstream consequence of that expansion, never a substitute for it.

Check your understanding Beginner

Formal definition Intermediate+

The capabilities approach is the thesis that the proper space of interpersonal comparison for justice and development is the set of functionings a person is really free to achieve. Its core notions are functionings, capabilities, conversion factors, and agency.

Definition (functionings). A functioning is an achievement — a state of being or doing that constitutes a person's life. Being well-nourished, being literate, participating in political deliberation, moving about freely, having self-respect, and being free from violence are functionings. Functionings are observable doings and beings, not subjective reports of satisfaction.

Definition (capability). A capability is the real freedom to achieve a functioning — the set of functioning vectors a person can achieve given their personal characteristics and social and environmental setting. If $b$ denotes a vector of functionings, person $i$'s capability set is

$$C_i=\{b\mid i\text{ can achieve }b\}.$$

Capability is functioning plus the option to choose it; the person who fasts and the person who starves may share the same functioning (undernourished) but differ utterly in capability.

Definition (conversion factors). A conversion factor is any personal, social, or environmental characteristic that determines how a resource is converted into a functioning for a given person. Personal conversion factors (metabolism, disability, age, sex), social conversion factors (public-health norms, gender roles, legal status, discrimination), and environmental conversion factors (climate, altitude, infrastructure) together mediate the map from resources to functionings. A bicycle, as Sen puts it, is the same commodity for all but yields very different mobility functionings for the able-bodied commuter, the disabled person, and the person in a region without roads.

Definition (agency). Agency is the capacity to pursue and realize goals one has reason to value. Agency freedom is the freedom to bring about outcomes one values; well-being freedom is the freedom to achieve good functionings. The two come apart: a comfortable dependent may have high well-being and low agency; an impoverished activist may have low well-being and high agency. Sen treats both as normatively irreducible.

Counterexamples to common slips Intermediate+

• Capabilities are just happiness in disguise. No. Functionings are observable doings and beings (being literate, being mobile, being free from violence), not subjective reports. The happy-slave case has low capability and high reported welfare precisely because the two spaces are distinct.

• Sen and Nussbaum agree completely. No. Sen is a pluralist who refuses to fix a closed list of central capabilities, arguing that the relevant set is context-dependent and emerges from democratic deliberation. Nussbaum commits to a definitive list of ten central capabilities as a threshold for constitutional guarantees. The disagreement is philosophical, not stylistic.

• The approach is only for poor countries. No. It applies to wealthy societies wherever conversion factors block capability: disability rights, gender equality, elderly autonomy, political participation, the treatment of migrants. The HDI's focus on developing countries is a policy choice, not a theoretical limit.

• The approach is anti-market. No. Sen treats markets as one instrument among many for capability expansion, praising their efficiency while criticizing market-fundamentalism that ignores what markets cannot deliver (basic education, public health, political rights).

• Capabilities are unmeasurable. Some are (practical reason is hard to survey), but many are not. Literacy, life expectancy, nutrition, political participation, and safety are all measurable, and the Alkire-Foster Multidimensional Poverty Index measures ten indicators across health, education, and living standards.

• Rawls and Sen are opponents. Mostly they are allies within political liberalism who disagree on the comparison space. Rawls focuses on the institutional design of the basic structure; Sen focuses on the individual outcomes the structure produces. Their views are complementary more than opposed.

Key argument: capabilities correct utilitarian and resource-based justice Intermediate+

Argument (the "Equality of What?" correction, after Sen 1979).

1. Premise (the launching question). Every substantive egalitarian theory must specify the space in which equality is assessed — Sen's "Equality of What?" question. Different answers yield incompatible policy prescriptions, so the choice of space is not a detail but the load-bearing decision.

2. Premise (the utilitarian answer and its failure). Utilitarianism answers: equality of subjective welfare (happiness, desire-satisfaction). This fails because welfare is adaptive: persons in chronic deprivation adjust their expectations downward, so a society can register high average welfare while its members lack basic functionings. This is the adaptive-preferences failure: the measure is deformed by what it tries to measure.

3. Premise (the resource-based answer and its failure). Rawls answers: equality of primary goods (income, wealth, opportunities, the social bases of self-respect). Dworkin answers: equality of resources. Both fail because resources convert into functionings at different rates for different persons. A disabled person requires more resources than an able-bodied person to achieve the same functioning of mobility, so equality of resources leaves them unequal in what they can do and be. This is the conversion-factor failure: means are mistaken for ends.

4. Intermediate conclusion. Neither welfare nor resources is the right space, because each ignores a structural feature of human life — the first ignores adaptation, the second ignores conversion. Any space that ignores either will misrank two persons who ought to be judged equal (or vice versa).

5. Conclusion (the capability space). The proper space of interpersonal comparison for justice is the capability — the real freedom a person has to achieve the functionings they value. Capability absorbs conversion factors (it is defined relative to the person's characteristics and environment) and is not distorted by adaptive preferences (it is an objective measure of what the person can do and be, not what they report wanting).

Reconstruction. The argument is an eliminative induction over candidate answer-spaces, not a deduction from first principles. Sen does not claim capabilities are the only relevant space; he claims they are the space that avoids the two structural failures — adaptation and conversion — that defeat utilitarian and resource-based accounts. The move is from means (resources) and ends-experienced (welfare) to ends-available (capability), and it preserves the egalitarian commitment while relocating it.

Bridge. The "Equality of What?" correction builds toward 20.02.01 Rawls and Nozick, where the primary-goods and entitlement theories Sen criticizes are set out in their strongest forms, and appears again in 30.04.01 social stratification as the measurement framework that distinguishes inequality of resources from inequality of capability. The foundational reason the correction works is that capability is defined at the intersection of person and environment — the same resource yields different capabilities depending on conversion factors — and this is exactly the structure that identifies justice with the expansion of real freedom rather than the redistribution of objects. The central insight is that what a person can do and be is constitutively sensitive to their individual characteristics and social setting; the bridge is from the question "how much do they have?" to "what are they actually free to do?", and the pattern generalises from individual policy assessment (disability accommodation, gender equity) to international development (the Human Development Index) and back to first-order theories of justice.

Exercises Intermediate+

Interpretive debates Master

Debate 1 (the completeness question: Sen's pluralism versus Nussbaum's list). The deepest internal disagreement in the literature concerns whether the capability approach should be a complete theory of justice with a fixed list of central capabilities, or an open framework whose content is fixed contextually by public reasoning. Nussbaum, from Nature, Functioning, and Capability (1988) onward and definitively in Women and Human Development (2000) and Creating Capabilities (2011), argues that the approach can ground constitutional guarantees only if it commits to a threshold list of ten central capabilities: life; bodily health; bodily integrity; senses, imagination, and thought; emotions; practical reason; affiliation; other species; play; and control over one's environment (political and material). Sen, in Inequality Reexamined (1992) and throughout Development as Freedom (1999), refuses to close the list, arguing that the relevant capabilities depend on context and value formation, that pre-empting public reasoning would itself violate the agency the approach protects, and that partial agreement on capability as the space suffices for many policy purposes.

Debate 2 (adaptive preferences and the measurement of the good). Sen's happy-slave argument holds that subjective welfare is deformed by deprivation and is therefore unsuitable as the space of comparison. Jon Elster and Martha Nussbaum extend the diagnosis; Sugden and others press the reply that all preference rankings are endogenous to context, so adaptation cannot by itself defeat the utilitarian without also defeating the capability theorist's appeal to "what the person has reason to value". The reply's force turns on whether there is an objective, non-adaptive standard of functioning to which the capability theorist can appeal — the position Nussbaum takes with her Aristotelian account of human functioning, and that Sen declines to take in fully general form, preferring context-specific lists.

Debate 3 (the Human Development Index and its operational limits). The HDI, introduced in the UNDP Human Development Report 1990 under Mahbub ul Haq with Sen's framework, aggregates life expectancy, education, and gross national income per capita into a single index by geometric mean. The index is a policy triumph but a philosophical compromise: it measures only three dimensions, uses imperfect proxies (life expectancy for the capability to live long and healthy, mean and expected years of schooling for the capability of knowledge, log GNI for the capability of a decent standard of living), and collapses national averages that can conceal deep internal inequality. The Alkire-Foster Multidimensional Poverty Index (2010 onward) extends the framework to ten indicators across health, education, and living standards, capturing overlapping deprivations the income-poverty line misses. The interpretive debate is how much philosophical weight such indices can bear: defenders treat them as honest proxies for a richer underlying capability space; critics (Pogge, Reddy) argue that the choice and weighting of indicators smuggles in contestable value judgments that Sen's pluralism should expose rather than hide.

Debate 4 (capability versus Rawlsian primary goods). Rawls himself, in Political Liberalism (1993) and Justice as Fairness: A Restatement (2001), conceded the force of Sen's conversion-factor objection and proposed that primary goods be reconceived as the goods needed by normal, fully cooperating members of society over a complete life — explicitly setting aside the cases of disability, illness, and dependency that Sen's framework highlights. The exchange (Rawls 2001 §7; Sen 1992 ch. 5; Nussbaum 2006 Frontiers of Justice) is the cleanest case in the literature of the resource-based and capability-based positions meeting head-on. Anderson's What Is the Point of Equality? (1999) and Wolff's work on disability argue that the capability theorist's critique overreaches where citizens' own choices, rather than conversion factors, produce disadvantage; the reply (Sen, Nussbaum) is that choices, too, are made within a capability set and cannot be assessed independently of it.

Debate 5 (agency freedom and well-being freedom). Sen's distinction between agency freedom (the freedom to bring about outcomes one has reason to value) and well-being freedom (the freedom to achieve one's own good functioning) is load-bearing for the approach's account of the person. A purely well-being-oriented theory collapses into paternalism — the state provides what is good for you; a purely agency-oriented theory loses the normative grip on deprivation (the impoverished activist's agency does not excuse her poverty). Crocker, in Ethics of Global Development (2008), develops Sen's agency concept into a theory of deliberative participation; Alkire, in Valuing Freedoms (2002), operationalizes agency in project evaluation. The debate concerns whether agency is a constituent of every capability (so that functioning achieved without choice is deficient) or an independent value that can trade off against well-being in policy design.

Debate 6 (extensions and boundaries). The approach has been extended in four contested directions. On gender, Nussbaum's Women and Human Development (2000) applies the central-capabilities list to the measurement of women's disadvantage in South Asia, arguing that adaptive preferences are most acute where deprivation is gendered. On disability, Nussbaum's Frontiers of Justice (2006) makes the capability threshold the basis for disability justice, where Rawls's primary-goods framework concedes its own inadequacy. On future generations and the environment, the approach is stretched by the problem of identifying the capability sets of persons who do not yet exist; Holland and drydyk debate whether the central-capabilities list can absorb ecological limits. On non-human animals, Nussbaum extends the framework across the species boundary (treating animal flourishing as a capability matter), while Sen declines to take the extension as anything more than a structural possibility.

Synthesis. The capabilities approach builds toward 20.02.01 Rawls and Nozick as a corrective within political liberalism, refining the resource-based and entitlement views without abandoning their egalitarian motivation, and appears again in 30.04.01 social stratification as the measurement framework that separates inequality of resources from inequality of what people can do and be. The foundational reason the approach is structurally robust is that capability is defined at the intersection of person and environment — the same resource produces different capabilities depending on conversion factors — and this is exactly the structure that identifies advantage with real freedom rather than with means or with felt satisfaction. Putting these together with the adaptive-preferences and conversion-factor arguments, the approach generalises from individual policy (disability accommodation, gender equity) to the international measurement of development (HDI, MPI), and the central insight is that justice compares persons in the space of what they are free to do and be, never in the space of what they own or report wanting; the bridge is between the philosophical critique of utilitarian and resource-based egalitarianism and the operational machinery of the human-development reports, and the pattern recurs wherever advantage must be measured across persons whose circumstances deform both their means and their wants.

Full argument set Master

Argument I (the adaptive-preferences argument against utilitarian egalitarianism). No theory that identifies advantage with subjective welfare (happiness, desire-satisfaction) can be an adequate egalitarian theory, because subjective welfare is endogenous to the very circumstances the egalitarian theory is meant to assess.

Reconstruction. (i) Chronic deprivation tends to lower a person's expectations and wants to the level of what they take to be feasible; this is the adaptive-preferences phenomenon. (ii) Therefore a person in severe deprivation can report satisfaction not because their life is good but because their preferences have been deformed by the deprivation. (iii) A measure of advantage that identifies the good with the satisfaction of such preferences will read a deprived population as satisfied, and so will fail to register the deprivation the egalitarian theory was constructed to detect. (iv) This is a structural feature of the welfare space, not an empirical accident; it holds whenever preferences are endogenous to the circumstances being assessed. (v) Therefore the welfare space is inadequate as a space of interpersonal comparison for an egalitarian theory, and the egalitarian must locate advantage in a space not deformed by adaptation — the space of capability. The argument is an eliminative one: it does not show that welfare is irrelevant, only that it cannot bear the egalitarian's weight. $\square$

Argument II (the conversion-factor argument against resource-based egalitarianism). No theory that identifies advantage with resources (primary goods, income, opportunities) can be an adequate egalitarian theory, because resources convert into functionings at different rates for different persons, so equality in the resource space is compatible with deep inequality in the space that matters.

Reconstruction. (i) Resources are means to functionings, not functionings themselves; a resource $r$ produces a functioning $b$ for person $i$ only via a conversion that depends on personal, social, and environmental factors. (ii) The conversion map $r\mapsto b_i$ varies systematically across persons: disability, metabolism, age, sex, local infrastructure, and social norms all change the functioning yielded by a given resource. (iii) Therefore two persons with identical resources can differ in the functionings they achieve — and so in the lives open to them. (iv) If advantage is what egalitarian justice is meant to equalize, advantage is constituted by the functionings a person is free to achieve, not by the resources they hold, since the same resources leave different persons unequal in the relevant respect. (v) Therefore the resource space is inadequate, and the egalitarian must compare persons in the capability space. As with Argument I, the claim is not that resources are irrelevant but that they cannot bear the egalitarian's weight. $\square$

Connections Master

• Political philosophy — justice, authority, and the state 20.14.01. The capabilities approach is the principal contemporary alternative to the social-contract and rights-based theories surveyed in the chapter anchor, and the two units are designed to be read together: the survey supplies the contractarian and entitlement background against which Sen and Nussbaum define the capability space, and this unit supplies the depth the survey can only gesture at. The chapter's load-bearing questions — what is a just society, what gives the state the right to rule — receive here a distinctively outcome-oriented answer: justice is the expansion of real freedom, and the state's authority is vindicated when its institutions expand the capability sets of those subject to them.

• Theories of justice: Rawls, Nozick, fairness 20.02.01. The Key argument of this unit is a direct critical engagement with the Rawlsian primary-goods framework and the Nozickian entitlement theory set out in 20.02.01. The conversion-factor argument is most forceful precisely against the difference principle applied to primary goods, and the disability and gender cases that drive the capability critique are the ones Rawls's restatement (2001) explicitly sets aside as deviations from the normal, fully cooperating citizen. Reading the two units together reveals where Rawls concedes the critique and where he resists it.

• Kuhn, Lakatos, and Laudan — paradigms and scientific change 20.08.04. The capabilities approach is itself a case study in the structure of philosophical research programmes: the "Equality of What?" question functions as the hard core of a Lakatosian programme, the conversion-factor and adaptive-preferences arguments as its most progressive protective belt, and the operational moves (HDI, MPI) as novel predictions that distinguish it from degenerating competitors. The two units share a meta-level concern with how normative and descriptive frameworks compete, succeed, and accumulate.

• Social stratification — class, race, gender 30.04.01. The capability framework supplies the measurement layer that distinguishes inequality of resources (the traditional concern of stratification research) from inequality of what people can do and be. The Multidimensional Poverty Index and the gender-focused capability measures of Nussbaum 2000 are direct applications of the approach to stratification analysis; the sociology unit supplies the empirical phenomena the capability theorist must explain, and this unit supplies the normative framework by which the phenomena are judged to be unjust rather than merely unequal.

Historical & philosophical context Master

Amartya Sen launched the capabilities approach in his 1979 Tanner Lecture on Human Values at Stanford, "Equality of What?" ^[Sen1979], in which he posed the launching question and ran the eliminative argument over welfare, Rawlsian primary goods, and resources, arriving at capability as the space that avoids both the adaptive-preferences and the conversion-factor failures. The foundational monograph, Commodities and Capabilities ^[Sen1985], set out the formal apparatus of functionings, capability sets, and conversion factors. Sen developed and defended the framework across Inequality Reexamined ^[Sen1992] and the popular synthesis Development as Freedom ^[Sen1999], the last written after the 1998 Nobel Memorial Prize in Economic Sciences, which Sen received in part for this body of work.

Martha Nussbaum arrived at a closely related framework from a different direction. In "Nature, Functioning, and Capability" ^{[Nussbaum1988]} (1988), she developed an Aristotelian account of human functioning as a basis for political philosophy, and in Women and Human Development: The Capabilities Approach ^{[Nussbaum2000]} she elaborated the definitive list of ten central capabilities and applied them to the situation of women in developing countries. Creating Capabilities: The Human Development Approach ^{[Nussbaum2011]} is the canonical survey of the approach as a whole.

The policy translation is due primarily to the Pakistani economist Mahbub ul Haq, who, working with Sen, designed the Human Development Index for the inaugural United Nations Development Programme Human Development Report ^[UNDP1990] of 1990. The HDI made the capability approach the measurement framework of the international development establishment. Sabina Alkire and James Foster later extended the framework to multidimensional poverty with the Alkire-Foster dual-cutoff method, operationalized in the Multidimensional Poverty Index from 2010. Sen and Nussbaum remain in philosophical disagreement over the completeness of the approach — Sen refusing to close the list, Nussbaum committing to ten central capabilities — a disagreement that has structured the literature for three decades.

Bibliography Master

@incollection{Sen1979,
  author = {Sen, Amartya},
  title = {Equality of What?},
  booktitle = {The Tanner Lectures on Human Values, vol. 1},
  editor = {McMurrin, Sterling M.},
  publisher = {University of Utah Press},
  address = {Salt Lake City},
  year = {1980},
  note = {Delivered as the Tanner Lecture at Stanford University, 22 May 1979; the launching-question text},
}

@book{Sen1985,
  author = {Sen, Amartya},
  title = {Commodities and Capabilities},
  publisher = {North-Holland},
  address = {Amsterdam},
  year = {1985},
  note = {The foundational monograph: functionings, capability sets, conversion factors},
}

@book{Sen1992,
  author = {Sen, Amartya},
  title = {Inequality Reexamined},
  publisher = {Harvard University Press},
  address = {Cambridge, MA},
  year = {1992},
}

@book{Sen1999,
  author = {Sen, Amartya},
  title = {Development as Freedom},
  publisher = {Alfred A. Knopf},
  address = {New York},
  year = {1999},
  note = {The popular synthesis, written after the 1998 Nobel Memorial Prize in Economics},
}

@article{Nussbaum1988,
  author = {Nussbaum, Martha C.},
  title = {Nature, Functioning, and Capability: {A}ristotle on Political Distribution},
  journal = {Oxford Studies in Ancient Philosophy},
  year = {1988},
  note = {Supplementary Volume; the early Aristotelian-capability statement, later reprinted in {\it Essays on Aristotle's De Anima} (Oxford, 1992)},
}

@book{Nussbaum2000,
  author = {Nussbaum, Martha C.},
  title = {Women and Human Development: The Capabilities Approach},
  publisher = {Cambridge University Press},
  address = {Cambridge},
  year = {2000},
}

@book{Nussbaum2011,
  author = {Nussbaum, Martha C.},
  title = {Creating Capabilities: The Human Development Approach},
  publisher = {Harvard University Press},
  address = {Cambridge, MA},
  year = {2011},
}

@techreport{UNDP1990,
  author = {{United Nations Development Programme}},
  title = {Human Development Report 1990},
  institution = {UNDP},
  address = {New York},
  year = {1990},
  note = {Inaugural HDR under Mahbub ul Haq, introducing the Human Development Index built on Sen's framework},
}

@book{Alkire2002,
  author = {Alkire, Sabina},
  title = {Valuing Freedoms: {S}en's Capability Approach and Poverty Reduction},
  publisher = {Oxford University Press},
  address = {Oxford},
  year = {2002},
}

@book{Crocker2008,
  author = {Crocker, David A.},
  title = {Ethics of Global Development: Agency, Capability, and Deliberative Democracy},
  publisher = {Cambridge University Press},
  address = {Cambridge},
  year = {2008},
}

@article{Anderson1999,
  author = {Anderson, Elizabeth},
  title = {What Is the Point of Equality?},
  journal = {Ethics},
  volume = {109},
  number = {2},
  year = {1999},
  pages = {287--337},
}

@book{Nussbaum2006,
  author = {Nussbaum, Martha C.},
  title = {Frontiers of Justice: Disability, Nationality, Species Membership},
  publisher = {Harvard University Press},
  address = {Cambridge, MA},
  year = {2006},
  note = {Extends the capability approach to disability, cross-border justice, and non-human animals; direct engagement with Rawls},
}