1 Amartya Sen is often portrayed as a different economist. As a Nobel Prize recipient in 1998, he serves as a reference for both mainstream economists (“neoclassical”) and their adversaries (“heterodox”). He is accredited with reconciling the unemotional approach of economists with the humanity of philosophers, as is reflected in the titles of his works (On Ethics and Economics, Inequality Reexamined, or Development as Freedom, to name just a few). The “capability approach,” which is at the heart of Sen’s analyses of poverty, development, and inequality, is portrayed as uniting these two disciplines in a way that surpasses conventional approaches to economic analysis.
2 Given the consensus this author garners, there is a paucity of critiques of the theories he propounds. Yet such critiques would be merited. This is the objective of the present paper, which summarizes Sen’s main analyses and positions, and seeks to identify the problems they raise on both the theoretical and the practical level. The main difficulty in studying Amartya Sen stems in particular from the terms he uses for the core concepts in his theories, which are often words of his own invention. While these may bestow his writings with an aura of scholarship, they make reading and understanding his work particularly arduous. This paper will first provide a brief overview of Sen’s two core notions, namely “functionings” and “capability,” and then seek to determine what distinguishes his approach from conventional approaches. Next, the focus will turn to Sen’s ethical approach. Lastly, Sen’s positions regarding economic policy will be examined, with consideration given to his underlying ethical and theoretical perspectives.
The “Capability” Approach
3 One of the tenets of Sen’s analysis is his refusal to assimilate well-being and utility (which Sen sometimes calls happiness, sometimes satisfaction, and sometimes describes as a ranking according to a scale of preferences). Sen considers utility-based approaches to be reductionist as they only take into account the psychological or mental consequences of owning goods in terms of the happiness or satisfaction they provide,  and not the actual well-being of individuals (the standard of living attained thanks to these goods).
From Goods to Functionings
4 The concept of capability, which was first proposed by Sen in 1979 (Sen 1982), is intended to account for human characteristics more accurately than the conventional approach in economics, which is based on utility. To define capability, Sen begins by representing individuals through a set of goods they may acquire (an “entitlement set”). However, Sen rejects the standard reference to “commodities” and prefers instead that of the “characteristics” of commodities.  For example, instead of considering goods such as apples, peanuts, rice, or beef, Sen prefers to base his analysis on their nutritional value, their taste, and so on.
5 However, this change in perspective does not make it possible to account for what individuals are, or do, thanks to the characteristics of these goods. Sen therefore also endows each individual with a set of what he calls “utilization functions,” which convert the characteristics of goods into “functionings” (Sen 1985), which are defined as the “doings and beings” of an individual. The effective functioning of a particular individual will therefore depend on that individual’s choice of a set of goods (transformed into characteristics), as well as on that individual’s choice of utilization functions. Thus, Sen intends to incorporate human diversity into the analysis of well-being as these utilization functions translate inter-individual differences into doings and beings that can be attained thanks to the characteristics of goods. 
6 This may sound rather obscure, which is indeed the case. Yet it may also be what explains Sen’s popularity (we all think it is profound, although we do not understand much of it). Nevertheless, those familiar with neoclassical theory may recognize such concepts, as they merely reproduce the conventional neoclassical approach, albeit dressed in a different vocabulary.
Capability and Choice of a Particular Lifestyle
7 Functionings therefore replace the goods (or their characteristics) found in standard approaches in economics, and in particular in microeconomics. As in microeconomics, these functionings can be combined by a person who makes a choice, and it is based on this combination of functionings that Sen defines “capability,” which is conceived as the freedom to choose functionings. As Sen writes in The Quality of Life,
The life that a person leads can be seen as a combination of various doings and beings, which can be generically called functionings. These functionings vary from such elementary matters as being well-nourished and disease-free to more complex doings or beings, such as having self-respect, preserving human dignity, taking part in the life of the community, and so on. The capability of a person refers to the various alternative combinations of functionings, any one of which (or rather any combination of which) the person can choose to have. In this sense, the capability of a person corresponds to the freedom that a person has to lead one kind of life or another.
9 Similarly, in Development as Freedom, Sen explains that, “A person’s ‘capability’ refers to the alternative combinations of functionings that are feasible for her to achieve. Capability is thus a kind of freedom, or the substantive freedom to achieve alternative combinations of functionings (or, to put it less formally, the freedom to achieve various lifestyles)” (Sen 2000, 74–75).
10 In fact, the essence of Sen’s approach by no means constitutes a departure from mainstream economics (whether microeconomics or neoclassical economics) since all of these assume that individuals make free choices (of “functionings” rather than goods) subject to constraints (through “capability” instead of income).
The Capability Approach and Neoclassical Theory
11 Moreover, Sen does not hide his intention to situate his analysis within the neoclassical theoretical framework developed by Kenneth Arrow and Gérard Debreu (1954). For example, in Development as Freedom, Sen explains that, “I have, in fact, demonstrated elsewhere that in terms of some plausible characterizations of substantive individual freedoms, an important part of the Arrow-Debreu efficiency result readily translates from the ‘space’ of utilities to that of individual freedoms, both in terms of freedom to choose commodity baskets and in terms of capabilities to function” (Sen 2000, 117–119). This begs the question of what prices—as well as supply and demand—may be (and mean) in such a market. Nevertheless, the main point here is that Sen fully and unreservedly adopts the Arrow-Debreu model. As Sen stated at a conference held by the French Economic Observatory (Observatoire français de conjoncture économique—OFCE) in Paris, he is a mainstream economist, so we should not expect too much from him in terms of methodology or analytical frameworks, since his approach does not provide an alternative path to mainstream economics.
12 But what about Sen’s approach in terms of philosophy and, more specifically, the solutions he offers from an ethical perspective?
Multiple Ethical Criteria
13 Sen’s widespread popularity stems not from his academic writings, which are not particularly accessible, but from his work intended for the general public, in which he adopts an engaging viewpoint, in particular with respect to the issue of poverty and to ways of resolving it. Ethics boils down to forming criteria that enable a choice to be made between various situations, after having ranked these situations according to these criteria. One of the major ethical doctrines is utilitarianism, which values any action or institution according to its propensity to increase—or decrease—the happiness of the community. Drawing solely on one criterion—general happiness—thus makes utilitarianism a monist ethical doctrine.
14 Sen departs from this doctrine, which, implicitly or explicitly, is adhered to by most economists, because he considers its vision of well-being to be excessively one-dimensional. According to Sen, the happiness criterion neglects individuals’ other values, which he purports to incorporate with the use of functionings and capabilities, thereby providing a better assessment of well-being. This is equivalent to adopting an ethical framework comprising several criteria, for which Sen uses the term “pluralism.”
Critique of Monism
15 In Inequality Reexamined, Sen writes that, “While being happy may count as an important functioning, it cannot really be taken to be all there is to leading a life (i.e., it can hardly be the only valuable functioning)” (Sen 1982, 54). Thus valuable functionings are seen to be the basis for a life worth living. While happiness features among the factors that constitute a “valuable life,” it is not the only factor—hence Sen’s critique of utilitarianism.
16 In Development as Freedom, Sen states that, “To insist that there should be only one homogeneous magnitude that we value is to reduce drastically the range of our evaluative reasoning. It is not, for example, to the credit of classical utilitarianism that it values only pleasure without taking any direct interest in freedom, rights, creativity, or actual living conditions. To insist on the mechanical comfort of having just one homogeneous ‘good thing’ would be to deny our humanity as reasoning creatures. It is like seeking to make the life of the chef easier by finding something which—and which alone—we all like (such as smoked salmon, or perhaps even french fries), or some one quality which we must all try to maximize (such as the saltiness of the food)” (Sen 2000, 77)
17 Instead, Sen proposes a system that would make room for a set of values to reflect the diverse aspirations of individuals. In The Quality of Life, Sen writes that, “Because of the nature of the evaluative space, the capability approach differs from utilitarian evaluation . . . in making room for a variety of human acts and states as important in themselves (not just because they may produce utility, nor just to the extent that they yield utility)” (Nussbaum and Sen 1993, 33). Moreover, as Sen states in On Ethics and Economics, “These ‘functionings’ may cover a diverse range of achievements, varying from being free from under-nourishment and avoidable morbidity to achieving self-respect and creative fulfillment. In fact, it is in this list that the functioning of ‘being happy,’ which some utilitarians see as the basis of all valuation, can—not unreasonably—figure (inter alia)” (Sen 1987, 63-64).
18 At first sight, all this seems obvious. No one, not even utilitarians, can deny the importance of “self-respect,” “creative fulfillment,” or “avoiding morbidity,” unless we believe that we are incapable of the slightest introspection or unable to observe the behavior of those around us. So why would we persist in upholding a single criterion when it seems easier—and more relevant—to consider the diverse facets (or “functionings”) of the human mind and therefore to adopt an approach comprising several criteria? This question is as old as ethics itself, yet Sen seems to ignore it. He merely criticizes the “arbitrary” and “defective” nature (to use his own terms) of monist approaches, despite the fact that they belong in the domain of great philosophers such as Hume, Mill, and Kant.
The Problem of Multiple Ethical Criteria
19 If the use of multiple ethical criteria has been rejected by all great philosophers, whether utilitarian or not, it is for the very simple reason that it does not allow us to settle all situations with which a philosopher, or even a man of action, may be confronted.
20 In A System of Logic, John Stuart Mill puts the problem as follows: “There must be some standard by which to determine the goodness or badness, absolute and comparative, of ends, or objects of desire. And whatever that standard is, there can be but one; for if there were several ultimate principles of conduct, the same conduct might be approved by one of those principles and condemned by another; and would be needed some more general principle, as umpire between them” (Mill 1843/1979, 951). In his essay on utilitarianism, Mill further states that, “If utility is the ultimate source of moral obligations, utility may be invoked to decide between them when their demands are incompatible. Though the application of the standard may be difficult, it is better than none at all: while in other systems, the moral laws claiming independent authority, there is no common umpire entitled to interfere between them; their claims to precedence one over the other rest on little better than sophistry, and unless determined (as they generally are) by the acknowledged influence of considerations of utility, afford a free scope for the actions of personal desires and partialities” (Mill 1861/1991, 157–158).
21 Similarly, in The Theory of Moral Sentiments, Adam Smith posits that “All constitutions of government . . . are valued only in proportion as they tend to promote the happiness of those who live under them” (Smith 1790/1976, 185). For his part, Immanuel Kant, a non-utilitarian philosopher with whom Sen claims affinity, is also very clear on this subject, when he writes in The Metaphysics of Morals that, “Considered objectively, there can be only one human reason. . . . So the moralist rightly says that there is only one virtue and one doctrine of virtue, that is, a single system that connects all duties of virtue by one principle” (Kant 1796/1996, 81).
22 The use of a single criterion makes it possible to choose, in all cases, between two actions, rules, or institutions that are in conflict with one another.  However, this is not the case when several criteria are present. To cite an example frequently used by Sen, if “living healthily” and “being able to read and write” are taken as essential criteria for a “valuable” life, it is impossible to choose between recruiting a doctor, a nurse, or a teacher without resorting to a higher criterion (that of collective happiness, for example), which would enable such a trade-off to be made.
23 As Séverine Deneulin explains, “Sen gives a reason for not specifying what is to be counted as relevant capabilities, which is his concern for pluralism . . . . [However,] if one refuses to take any position regarding the ends that are to be promoted, how then can we know which opportunities have to be given to people in order to improve their quality of life? How can we give people conditions for a better human life without knowing what a better life consists of?” (Deneulin 2002, 500–501).
24 Sen generates sympathy in particular for the attention he gives to the diverse facets of the human personality. No one would oppose more health care, more education, or more freedom. The problem is that we cannot have everything at the same time or without limit. A choice must be made, and Sen provides no method for making that choice. Nor does he provide a method by which to achieve the objectives he lays out, as becomes clear upon analysis of what Sen says (or does not say) about the policies that should be implemented to such ends.
Sen and Concrete Proposals
25 What then do Sen’s analyses offer the economist in the field, who studies what happens in the world as it is or who seeks to provide solutions to concrete problems? The answer is nothing, or—at best—very little. 
Sen and Studies of Economic Problems
26 Sen does not hide the fact that his system makes it impossible in practice to take into account all the differences that exist between individuals (i.e., human diversity). For example, in Inequality Reexamined, he notes that, “There are diversities of many different kinds. It is not unreasonable to think that if we try to take note of all the diversities, we might end up in a total mess of empirical confusion. The demands of practice indicate discretion and suggest that we disregard some diversities while concentrating on the more important ones. . . . In fact, general analyses of inequality must, in many cases, proceed in terms of groups—rather than specific individuals—and would tend to confine attention to intergroup variations” (Sen 1992, 115). Thus, as in standard microeconomics, when put into practice, the subtleties of individual behavior are abandoned in favor of group analyses, albeit with an unclear theoretical status.
27 Sen’s studies of concrete economic issues bear no direct relationship to his theoretical system. In fact, there is no need to mention them at all since he invariably proceeds in the same fashion. Take the example of his book Commodities and Capabilities, one of the rare cases where Sen provides (in an annex) details of what he proposes. In fact, the passage merely consists of statistical studies drawing on common indicators such as life expectancy, the infant mortality rate, the literacy rate, and so forth. A comparison of these indicators leads Sen to postulate that, “The capabilities of the Indian masses are enormously inferior to those of the masses in China and Sri Lanka in terms of the ability to live long, the ability to avoid mortality during infancy and childhood, the ability to read and write, and the ability to benefit from sustained schooling. In terms of basic capabilities—of survival and education—China and Sri Lanka are of a different nature” (Sen 1985, 76). In other words, in China and Sri Lanka, life expectancy is higher, infant mortality is lower, and the education system is more developed and yields better results than in India. In fact, this observation can be made without resorting to the jargon of “capabilities,” “utilization functions,” or other “functionings.” 
28 The same type of indicators is used in a book Sen co-authored with Jean Drèze and entitled India: Economic Development and Social Opportunity, in which the authors explain that,
Poverty is, thus, ultimately a matter of ‘capability deprivation,’ and note must be taken of that basic connection not just at the conceptual level but also in economic investigations and in social or political analysis. This broader and more foundational view of poverty has to be kept in mind while concentrating, as we do in this paper, on the deprivation of such basic capabilities as freedom to lead normal spans of life (undiminished by premature mortality) or the freedom to read or write (without being constrained by illiteracy).
30 Sen’s concrete analyses therefore provide practically no link to his theoretical, “microeconomic” analysis. Again, there is no mention of “baskets of goods,” “utilization functions,” “vectors of functionings,” or other “substantive freedoms.” Instead, we return to classical indicators such as education and health, albeit saddled with the term “capabilities,” which adds absolutely nothing to the analysis.
Sen and Economic Policy
31 It is not enough to simply call for higher levels of healthcare and education and for all the grand things on which Sen writes at length. However, how to attain these goals also needs to be specified since any expenditure requires revenue. Yet curiously, Sen hardly ever addresses the issue of government economic policy or of the instruments required to effectively implement these, including taxes, subsidies, income redistribution, public spending, or regulation that might empower each individual with the “freedom to choose” or simply “capability.” Rather, he generally resorts to the standard discourse of mainstream economists, namely “market failures” and the “corrections” needed to “restore efficiency.”
32 As mentioned above, Sen adopts the Arrow-Debreu general equilibrium model as a reference, a model he proposes to adjust in order to incorporate “capabilities” and “functionings.” In a recent book, Development as Freedom, he appears eager (in a chapter dedicated to the State and the market) to remind us of its pertinence, in particular with regard to the results the model obtains. As Sen writes, “This efficiency achievement—the so-called Arrow-Debreu theorem (after the original authors of the results, Kenneth Arrow and Gérard Debreu)—is of real importance despite the simplifying assumptions” (Sen 2000, 117). Yet it is well known that the Arrow-Debreu theorem relates to an economy with perfect competition and organized in a very particular way. It is a system in which agents make choices based on known prices (thus becoming so-called “price takers”), and prices are announced by what is commonly referred to as an auctioneer.  However, the role of the auctioneer does not end here since he groups together the supply and demand of individuals (with bilateral trade between agents being ruled out), compares them, and adjusts the prices in accordance with their differences.
33 Yet Sen makes no mention of this, which is surprising given that the matter in question concerns both markets and government, with the latter being represented in this model by the auctioneer, while it is not too clear what the market consists of). The “efficiency achievement” may be of “real importance,” but not with regard to the real world since it concerns a world with perfect competition that bears no resemblance to the societies we actually live in.
34 Similarly, with regard to underdevelopment and growth, there is nothing original about Sen’s position since the role of the State is reduced to initiating—or “nurturing”—the “market mechanism” until it becomes “self-sufficient.” In India: Economic Development and Social Opportunity, Drèze and Sen state that, “The government may have a major role in initiating and facilitating market-reliant economic growth . . . This role is easy to understand in the light of economic theory—particularly as related to difficulties of initiation, connected with such difficulties of tâtonnement (pre-exchange negotiations about market prices, leading to simultaneous production decisions), economies of scale, the importance of technological externalities, and the integral nature of skills formation. The nurturing of an early market mechanism by an active state does not, of course, preclude a more self-sufficient role of the market later on” (Drèze and Sen 1995, 19).
35 According to Sen, the process merely needs to be “nurtured,” and all will be well in a brave new world. This is the usual (and contradictory) discourse that forms the basis of the structural adjustment plans devised for developing countries, in which the State puts in place “healthy and sound” structures on the basis of which the market flourishes. While Sen obviously does not express things this way, the idea remains the same—at least, he has never publicly and unequivocally criticized this notion (unlike Joseph Stiglitz, another celebrated economist of our times).
Market Failure and Government Failure
36 In the few cases where Sen addresses matters of economic policy, his analysis is confined to the subject of “market failure.” For example, he justifies government intervention in the areas of education and health on the basis of the presence of externalities or public goods. This is the case in Development as Freedom, where Sen writes that,
Some of the most important contributors to human capability may be hard to sell exclusively to one person at a time. This is especially so when we consider the so-called public goods, which people consume together rather than separately. . . . The rationale of the market mechanism is geared to private goods (like apples and shirts), rather than to public goods (like a malaria-free environment), and it can be shown that there may be a good case for the provisioning of public goods, going beyond what the private markets would foster. Exactly identical arguments regarding the limited reach of the market mechanism apply to several other important fields as well, where the provision involved may also take the form of a public good. Defense, policing, and environmental protection are some of the fields in which this kind of reasoning applies. . . . The “public goods” argument for going beyond the market mechanism supplements the case for social provisioning that arises from the need for basic capabilities, such as elementary healthcare and basic educational opportunities.
38 Such arguments justifying minimal state intervention in the name of efficiency, with no distinction made between rich and poor countries (since they are valid at all times and in all places) can be found in any microeconomics textbook. Once again, there is absolutely nothing original here.
39 In Radical Needs and Moderate Reforms, one of the few texts in which Sen addresses specific economic policy issues, he does not seem particularly favorable toward extensive state intervention. In reference to the economic reforms aimed at liberalization and deregulation in India, Sen states that, “The departures are too moderate—and too tolerant of the established tradition of economic planning in India” (Sen 1997, 4). Furthermore, “The counter-productive nature of some of the governmental restrictions, controls, and regulations has been clear for a long time. They have not only interfered with the efficiency of economic operations (especially for modern industries), but also have often failed lamentably to promote any kind of real equity in distributional matters” (Sen 1997, 9). Once again, there is nothing original in Sen’s views. Although the notion of “government failure” (government intervention that hampers “economic efficiency”) resurfaces, this concept has been highly fashionable in recent years, in particular among international organizations (its corollary being “good governance,” that is, government needs to be small but efficient).
40 Contrary to a widely-held view, the issue of poverty has always been at the center of economists’ concerns, as reflected in the work of the best-known among them, including Smith, Mill, Walras, Marshall, Pigou, and Keynes. The major question concerns how to resolve this problem and therefore the respective roles to be assigned to the market and the State. This complex question gives rise to complex answers that are never certain. Along comes an author such as Amartya Sen—crowned with a Nobel prize for work on a separate topic (namely, social choice)—who writes attractively-titled books with impressive bibliographies for the general public on subjects such as poverty, inequality, and freedom. We are now far removed from the dreariness of economists’ usual publications, with their figures and their jargon (not to mention their equations). Because Sen addresses the poor (himself being from a developing country), the consensus is quick to extol his work and single him out as ideologically different from other economists.
41 In fact, few ever reach the end of Sen’s books, for the very simple reason that they are incomprehensible for the uninitiated. More than those of any other economist, Sen’s writings, whether on economics or philosophy, are systemically built upon references (for example, the body text of On Ethics and Economics contains 89 pages, followed by 33 pages of bibliography). Moreover, as he uses a gamut of obscure notions (including “capability,” “functioning,” and “utilization function”), considerable effort is required to decipher his thinking. In fact, this is precisely what this paper set out to achieve, and the conclusion it reaches is that Sen is an entirely orthodox economist with respect to his vision of the economy and the role of markets, that he offers nothing specific or original with respect to how to resolve problems such as poverty, and that his multi-criteria ethics does not bear scrutiny for reasons that are as old as philosophy itself. If he garners consensus, it is perhaps because everyone is able to read what they wish in his confused discourse.
In particular, see Sen 1983, 153–169.
Here, Sen draws on a theory developed by Kelvin Lancaster (1966), in which this neoclassical economist proposed to base individuals’ utility not on baskets of goods but on baskets of characteristics of goods.
To continue with the example of food, an individual who is allergic to peanuts will not be represented by the same set of utilization functions as a non-allergic individual.
Although some authors use several ethical criteria, they rank these. That is, if two situations are similar according to one criterion, they refer to a higher-ranked criterion in order to make a choice. Thus the ability of deciding between two situations depends on the existence (or otherwise) of a so-called ultimate ethical criterion.
Sen formulated an index with the intention of more adequately measuring economic inequalities (in particular by taking into account income gaps among the poorest populations). Needless to say, no knowledge of his theoretical framework of “capabilities” or “functionings” is required in order to make use of this quantitative technique.
Note that such cross-country comparison is possible only if the level of each indicator of a country (expressed here by a number) exceeds the level of the corresponding indicator of another country. This remains true whichever ethical criterion is applied. However Sen’ s approach raises the issue of country ranking when the level of illiteracy of one country is lower than that of another country but its level of child mortality is higher.
For example, in the words of orthodox economist Jean-Marc Tallon, in this economy, “no economic agent sets prices.” This is why “an individual, the auctioneer, is called upon” to “announce prices” (Tallon 1999, 72).