Andy Blunden August 2004
The problem of inequality, and the exclusion of a large proportion of the world’s population from a share of enjoyment of the world’s products and a share in deciding how the world’s products should be distributed, is as great a problem today as it has ever been.
However, movements for distributive justice have been demobilised and even marginalised; movements for cultural recognition, which have been most prominent in recent decades, seem no longer able to speak to the problems of the most downtrodden sections of the world.
A new paradigm of justice centred around concepts of democracy and freedom seems to be emerging but the relation between a number of very different paradigms remains unclear.
Amartya Sen is a development economist who has been conducting a relentless “internal criticism” of concepts of distributive justice and equality over the past thirty years. Whilst retaining the form of a distributive theory, he has successively interrogated what it is which ought to be distributed more equally, to a point where the content is now closer to that of the politics of recognition and discourse ethics, than to that of a traditional theory of distributive justice.
As a youngster in Bengal in 1943, Amartya Sen witnessed India’s last famine, during which two to three million people died; he also witnessed the sectarian murder of a poor Muslim labourer, who had been forced to risk death by seeking work in a Hindu area.
Although Sen has lived most of his life in Britain and the U.S., he has never been away from India for longer than six months, and there is no doubt that the problems and achievements of independent India have been the central concern of his life.
Sen was a PhD student at Cambridge when Kenneth Arrow published his famous theorem on “social choice” theory in 1951. Sen has remained fascinated with this highly mathematical theory ever since. His academic supervisors at Trinity College, Cambridge however — Joan Robinson and Maurice Dobbs — were not so “excited” by this theory, and Sen wrote his PhD thesis, Choice of Techniques, on the alternative paths of development open to the newly independent former colonies. In this work, completed in only twelve months, Sen advocated using their relatively low labour costs to promote basic health care and literacy, rather than pursuing rapid industrialisation through capital investment. All his subsequent work has continued to focus on these, his “old obsessions,” and his other abiding concerns, inequality, the emancipation of women, and democratic government.
“In line with the importance I attach to the role of public discussion as a vehicle for social change and economic progress ... I have, throughout my life, avoided giving advice to the ‘authorities.’ Indeed, I have never counselled any government, preferring to place my suggestions and critiques — for what they are worth — in the public domain.” [p. xiii-xiv, Sen 1999]
I consider Sen’s express position to be a misrecognition of the actual positioning of his theoretical interventions. In this article, I argue that there is a theoretical tension in Sen’s work, reflected in the ambiguity of delivering advice to government via the public domain, speaking both from the standpoint of social movements and, hypothetically so to speak, from the standpoint of government; a tension between ethical principles which correspond to two different subject positions. This ambiguity is shown, for example, in Sen’s contradictory positions in relation to social choice theory, in his concept of “comprehensive outcome” in which process is included as part of outcome, in his attempt to introduce agency into the utilitarian conception of the person and an element of consequentialism into otherwise deontological libertarian ethics. I think Sen’s ambiguous subject position reflects the contradictory subject positions inherent in a social justice movement which has become the government, and in that sense, is an ambiguity with a thoroughly objective and progressive basis.
I will return to the question of “social choice” theory and Sen’s subject position later. The first issue on which I want to focus is his conception of social justice and human needs.
After a period of work on social choice theory, urged by his wife to involve himself in more “practical issues,” in 1973 Sen wrote On Economic Inequality. This work focused entirely on real income as the measure of advantage and well-being, examining problems like the setting of the “poverty line” and measuring degrees of poverty and inequality. Sen’s idea here was that if a society was to make a decision about the degree of inequality it would tolerate, then it needed a suitable, agreed measure of inequality.
In using real income as the measure of well-being, Sen was only doing what everyone else in the field was doing, and in most cases still is doing. Indeed, there can be no escaping the fact that, in societies where most goods are acquired through the market, real income is indeed a good first approximation to social welfare and the freedom a person has in determining their own life.
Sen had taught with John Rawls at Harvard from 1970, and partly in response to the debate over what constituted the “basic goods” a person needed in order to participate in democratic social life (which Rawls had conceived in terms of basic human needs), Sen set out to more closely investigate the nature of what people really need by way of “basic goods.” (The debate between Sen and Rawls led to both writers developing their conceptions.)
This led to the seminal paper published in 1980 entitled Equality of What? in which Sen put forward the concept of “functioning.” The commodities over which a person had command were, after all, only a means to an end, and that end was a level of functioning in life, being able to live the kind of life that one values. This “functioning” was subject to objective measurement as well: life expectancy, infant mortality, literacy, morbidity, political participation and so on. An example of the indicators to which this approach draws attention is the statistic that the average longevity of a resident of a poor neighbourhood in New York is less than that of a citizen of Bangladesh, despite the fact that not only incomes, but real incomes, are many times higher in Harlem than in Dhaka.
Sen frequently uses the well-being of someone with a disability to illustrate the point that what someone can achieve with a given amount of wealth, depends on certain conditions, and any measure of inequality has to take these “conversion factors” into account, focusing on outcomes rather than means.
Further, even though we value real income as a means to overcome challenges, a person who spends money in order to fend off the dangers of rampant crime and endemic disease, is obviously less well-off than a person who enjoys good public health and security, and has no need for such expenditure. Thus, the perspective of functionings also brings into account the benefits a person receives from public goods, not just private labour and the market.
However, on closer inspection, the measure of functionings misses an important dimension of well-being, namely freedom. When Gandhi chose to fast, he was clearly not suffering the same level of deprivation as someone starving as a result of poverty, because although Gandhi had the opportunity to eat, he chose to use his freedom to not eat. Sen may have also had in mind comparisons between the India and China: even though, overall, pre-reform China had far surpassed India in its achievements in overcoming illiteracy, ill-health and hunger, this had been achieved at the cost of choice.
Thus, the real measure of well-being had to be not the actual functioning which a person exercised, but capability — the set of functionings from which one can choose. So for example, the university graduate who is serving tables has an unmistakable advantage over their uneducated colleague, for they have a choice, just as the adventurer who suffers exposure while mountain climbing is obviously more advantaged than the slum-dweller who freezes out of necessity.
How to measure capability? Sen’s relentless concern with measurement, even when the problem appears insoluble, is a great strength; measurement is integral to all conceptions of distributive justice and social choice, and Sen never tires of subjecting the most intractable concepts to quantification and ranking.
There are two basic approaches to measurement of capability. On one hand, the functioning which a person chooses from those available to them can be taken, ipso facto, to be the most valued functioning, and therefore a measure of the value of the capability set from which it was eventually chosen. However, a person may choose a functioning for all sorts of reason (for example, a person may choose to give up their freedom in order to care for a sick friend, but this hardly proves that such a life is their favoured choice). Consequently, capability needs to be measured as a set according to the whole range of functionings it contains. This is a challenging technical task of course, but conceptually it is clear enough: well-being is properly measured neither by wealth (which is but a means to an end) nor by functioning (which fails to reflect the valued choices which have been forgone) but by capability.
It is these concepts (functioning and capability) for which Sen is most famous, but in the late 1990s, driven perhaps by his critique of utilitarianism and a growing conviction that women’s emancipation is the central issue in development, Sen took this determination of human needs two steps further.
Sen’s central critique of utilitarianism is that by reducing human motivation to the maximisation of a person’s utility (however defined), utilitarianism effectively eliminates agency. This capacity of a person to choose to do one thing and not another Sen saw as an essential ingredient of well-being. But so long as choice was confined to selection between options determined by others — so long as a person’s capability set was determined by social arrangements in which one had no say — then there is no real freedom.
In Development as Freedom (1999), Sen further determines advantage from wealth to functioning to capability to voice:
“... the general enhancement of political and civil freedoms is central to the process of development itself. The relevant freedoms include the liberty of acting as citizens who matter and whose voices count, rather than living as well-fed, well-clothed and well-entertained vassals. The instrumental role of democracy and human rights, important as it undoubtedly is, has to be distinguished from its constitutive importance.” [Sen 1999, p. 288]
In India: Development as Participation (2002), shows that women’s well-being, fertility and child survival all depend on women’s agency (which I am taking as synonymous with “voice” at this point) including access to employment, women’s literacy and property rights — independently of the overall level of opulence, industrialisation or literacy. This advantage shows up, for example, in gender differences in child survival and longevity. Women’s voice (as for example in the Indian state of Kerala where there is a long tradition of women’s education and property rights) proved more effective in lowering fertility than China’s one-child policy, and more effective in increasing longevity than the greater wealth and industrialisation of northern India.
With Development as Freedom, Sen moved from including freedom as instrumental to well-being, and seeing freedom as an essential ingredient of well-being, to conceiving of well-being as freedom: freedom to lead a life that one has reason to value, including both positive freedom (real opportunities) and negative freedom (freedom from constraints and interference), actualised as achievement.
In the recent book on India, Sen shows how those sections of society which have more than their fair share of voice in the determination of government priorities (men rather than women, city people rather than rural people, the middle classes, the military elite) enjoy capability sets larger than others, because they are able to see to it that social arrangements are geared to meeting their needs and provide them with opportunities. While voice is therefore instrumental in the formation of real freedom, it is also constitutive of freedom, an achievement, and end in itself. We thus have a fourth concept in the series of determination of human needs: wealth (or opulence), functioning (or real living standards), capability (or real opportunity) and voice — the say that someone has in determining the social arrangements to which they are subject.
Sen’s observation that no people which has had the vote has ever suffered from famine, aptly illustrates the point.
In India: Development as Participation, Sen goes one step further as a result of his study of “son preference.” Son preference is the tendency of people in certain cultures to prefer a son to a daughter, resorting to abortion of female foetuses or simply neglect of the health of young girls. As a result of these practices, India and China are each “missing” about 40 million women in their current populations. Sen observed that this tendency not only increases with industrialisation and rising real incomes, but increased even in those societies where women had a voice. Even educated women and women who have full control over the decision whether or not to abort a female foetus, may be active participants in exercising son-preference because they share their husband’s preference for a son.
“This type of gender inequality [son preference] cannot be removed, at least in the short run, by the enhancement of women’s empowerment and agency, since that agency is itself an integral part of the cause of natality inequality. This recognition demands an important modification — and indeed an extension — of our understanding of the role of women’s agency in eliminating gender inequality in India. The enhancement of women’s agency which does so much to eliminate sex differentials in mortality rates (and also in reducing fertility and mortality rates in general) cannot be expected, on its own, to produce a similar elimination of sex differentials at birth and abortion, and correspondingly in the population of children. What is needed is not merely freedom and power to act, but also freedom and power to question and reassess the prevailing norms and values. The pivotal issue is critical agency. Strengthening women’s agency will not, by itself, solve the problem of ‘son preference’ when that works through the desires of the mothers themselves.” [Sen 2002, p. 258.]
“... the agency of women is effective in promoting those goals which women tend to value. When those values are distorted by centuries of inequality, for example yielding the perception that boys are to be welcomed more than girls, then the empowerment of women can go hand in hand with persistent inequality and discrimination in some fields, in particular ‘boy preference’ in births (with possibly brutal results in the form of sex-specific abortions). Indeed, the agency of women can never be adequately free if traditionally discriminatory values remain unexamined and unscrutinised. While values may be culturally influenced (we have provided some evidence corroborating this presumption), it is possible to overcome the barriers of inequality imposed by tradition through greater freedom to question, doubt, and — if convinced — reject. An adequate realisation of women’s agency relates not only to the freedom to act but also to the freedom to question and reassess. Critical agency is a great ally of development.” [Sen 2002, p. 274.]
To reflect the fact that recognition as an equal participant in the social and political life of a society still leaves the person trapped within dominant customs, beliefs and modes of living, which for example, may include misrecognition of their personality or unjust constraints on their activity, Sen introduced the term “critical voice.”
This concept of critical voice is thus the fifth in a series of determinations of advantage: wealth, functioning, capability, voice and finally, critical voice.
Critical voice is the capacity of a person living “inside” a society to form views available from a position “outside” that society:
“... virtually every society tends to have dissenters, and even the most repressive fundamentalist regimes can — and typically do — have dissenters .... Even if the perspective of the dissenters is influenced by their reading of foreign authors, the viewpoints and critical perspectives of these members are still ‘internal’ to the society.” [Sen 2002a, p. 476-77.]
Critical agency is thus “not only to the freedom to act but also to the freedom to question and reassess.” The answer to the question Sen asked in 1980 – Equality of what? – seems increasingly to be “critical voice.” This does not imply that the demand for equality of critical voice necessarily has traction as a normative demand, any more than does equality of wealth. But “critical voice” does more truly determine the essence of human need and is the true measure of inequality in a society.
Critical voice is both instrumental, in that it is needed in order to sustain the other elements of well-being, and constitutive, in that only the person with critical voice is truly free.
That which is the means to well-being, not just apparently, but essentially, comes to be an end in itself, constituent of well-being. Thus, for example, the while education is valued initially for its contribution to job-seeking, over time it comes to be valued for itself.
Conversely, that which is formally the end, can only be real to the extent that it is supported by appropriate means. Thus, for example, even though everyone in a parliamentary democracy formally has an equal voice, without an adequate capability set, without an adequate functioning and wealth, this right is no more than formal.
So it is not just a question of Sen having “changed his mind,” and abandoned an economic conception of well-being and a paradigm of distributive justice in favour of a political conception of justice and a recognition paradigm. Each step in the further determination of well-being both overcomes and maintains the previous determination, including it within a yet deeper determination.
Thus advantage conferred by command over commodities is by no means done away with in the determination of well-being in terms of critical voice — witness the influence exerted in academia, the media and the legislature by big business! The capability to choose the functioning of one’s own choice retains functioning as the substance of well-being. The voice needed to secure an adequate capability for oneself is actualised only in the enjoyment of that capability. And conversely, voice can only be exercised to the extent that a person enjoys a wide capability set, of which the exercise of voice turns out to be the essential component.
Critical voice is the truth of voice: critical voice can exist only in and through voice, but voice proves to be worthless unless it is critical.
Sen’s critical examination of distributive justice took the form of asking himself the question: “Equality of what?” But “critical voice” is surely the key claim of the “politics of recognition.”
In arriving at a determination of well-being which is appropriate to the subject position of the representative of a social movement who demands to be heard, Sen retains his concern for the consequences of policy for the majority and for the need to uncritically weigh the preferences and well-being of all citizens, irrespective of their social position — an ethical concern appropriate for the representative of a democratic government.
Although it is not spelt out at this stage, it appears that Sen has arrived, by means of a successive critique of egalitarianism, at an ethic which places recognition within a paradigm of distributive justice.
In the light of this reconstruction of Sen’s theoretical agenda, we can understand why, as the dominant current of “welfare economics,” utilitarianism has necessarily been Sen’s nemesis and the subject of a sustained critique over several decades. Sen’s critical relationship to utilitarianism is quite complex however, and can only be understood in relation to how utilitarianism has changed in parallel with economic science, in response to positivist criticism.
Utilitarianism is political economy translated into the language of ethics. Utilitarianism originated with the “first positivism” of Auguste Comte, John Stuart Mill and Herbert Spencer, the positivism of law-governed social progress. Accordingly, Bentham formulated the principle of utilitarianism effectively as an apology for political economy, identifying capital accumulation with social progress.
As Sen points out, the early literature of utilitarianism is replete with careful distinctions between economic and ethical value, but such declarations cannot detract from the factual affinity between utilitarianism and political economy.
For example, the definition John Stuart Mill gives of “utility” in Utilitarianism is entirely consistent with the definition he gives in his Principles of Political Economy, and in the chapter on exchange, he demonstrates that every act of free-market exchange increases the sum of utility, thus establishing the fundamental principle of utilitarianism as a law of political economy.
It might be observed that since a dollar makes more difference to a pauper than to a millionaire, it necessarily follows that utilitarianism demands greater and greater inequality of wealth [See Sen 1973 p. 19.], but this is, after all, nothing but what is required by the laws of political economy.
In the late 1860s, a new wave of positivism, the positivism of anti-metaphysics, began within political economy itself, giving rise to the marginal revolution in economics and ultimately the Mach/Einstein revolution in physics. The “second positivism” banished the concept of “value” from economic science, confining economics to the phenomenon of price. The result was a significant narrowing of the conception of welfare to real income.
This was the state of utilitarianism at the time of the Great Depression when the absurdity of the criterion of the sum-of-utility became starkly obvious. The illusion that the sum-of-wealth could form a valid criterion for welfare economics could not now be harboured even by the most hardened and cynical apologist of capitalism.
John Maynard Keynes proved the fallacy of the theory of marginal utility as a macro-economic theory and the inadequacy of utilitarianism, as it then was, as the basis for welfare economics. Strange as it would seem, the reaction of utilitarianism in abandoning interpersonal sum-of-utility, was to take, not an egalitarian turn, but in effect the completely opposite direction.
Rather than querying the sum-of-utility as a criterion of the good, the “third positivism,” Logical Positivism, denied the very legitimacy of interpersonal comparison of utility — it was declared “unscientific” to suggest that the well-being of one person could be greater or less than that of any other person.
Economic science overcame the challenge of Logical Positivism by drawing on the mathematical apparatus named after Vilfredo Pareto. This new economic science eschewed reference to utility, basing itself exclusively on acts of exchange and choice. In logical positivist terms, “utility” can only be defined in terms of each agent’s preference ranking as objectively manifested in the choices they make.
The crucial concept here is the “Pareto optimum” — that state of the market where there exists no possible exchange, deemed by both parties to be beneficial, which remains to be executed. By this move, a new utilitarianism is established, in continuity with the old, in which the good is a Pareto optimum. The Pareto optimum could be a universal famine, but so long as no mutually beneficial exchange is left unmade, it is “the best of all possible worlds” in the terms of a logical positivist Pangloss, who, incidentally, does not believe in the maxim that things have to get worse before they can get better.
If I could coin a term, I think Kenneth Arrow’s economics, together with his contributions to complexity theory and social-choice theory marks the beginning of a “fourth positivism” based on the concepts of information and communication science. And the contemporary form of utilitarianism is what we call neo-liberalism or “economic rationalism.” It is self-evident that this ethic is as different from the utilitarianism of Jeremy Bentham as the economics of Kenneth Arrow is different from that of David Ricardo.
“Critique of utilitarianism” therefore has to be seen in this sense: criticism which traces the outlines of the development of political economy is what could be called “internal criticism,” whilst only criticism which pits itself as much against political economy as against its ethical expression, could be called critique in the proper sense of the word. In Sen, we find both kinds of criticism of utilitarianism mixed up together.
There are of course more varieties of utilitarianism than there are species of song bird; however, I believe that the above sketch illustrates the essential development of utilitarianism, reflecting the development of bourgeois science.
The enormous value of utilitarianism therefore is that it expresses the real ethic of participants in political economy, with preferences accepted uncritically like the “externals” of political economy. “Real” means: within the bounds of the “assumptions” of political economy.
There are however at least three distinct subject positions from which the ethics of the market can be viewed: (i) that of the agent (obeying the law, keeping promises, telling the truth, etc.), (ii) that of a government (regulating market outcomes and legislating), and (iii) that of a participant in a social movement, who could be critical of the very existence of political economy.
Utilitarianism is an ethic of type (i). Historically, utilitarians have been very prominent in advising governments, of course, but this is simply evidence of the capacity of political economy to manage government rather than the reverse.
Egalitarianism, as an ethical determination of the activity of the market regulator, is an ethic of type (ii). A democratic government, which recognises the equal moral worth of all its citizens, must at least to some extent, be egalitarian.
Egalitarianism has a history as well, perhaps most succinctly summed up in Marx’s 1844 three-stage schema: primitive or crude communism, political communism — “democratic or despotic,” and humanism as the transcendence of private property. The egalitarianism of our age is democratic-political egalitarianism, and accordingly, Sen formulates the problem for egalitarianism in terms of social-choice theory.
At whatever historical juncture, the ethical dispute between egalitarianism and utilitarianism is a struggle between two different subject positions. Utilitarianism is at best indifferent to equality, but its significance comes from the fact that utilitarianism informs us about the “real” ethic of the market. The interest that egalitarianism has in utilitarianism is that utilitarianism describes the conditions which egalitarianism aims to manage or overcome. It is from this point of view that egalitarianism is interested in the “internal” criticism of utilitarianism. Exactly the same issue confronts those who address themselves to the ecological or social fall-out from the market.
With these clarifications, let us look at Sen’s critique of utilitarianism.
Sen’s most crucial criticism of utilitarianism is its failure to recognise human agency. There are two aspects of a person, he says: their agency (the person as “actor”) and their well-being (the person as “patient”); utilitarianism “simplifies” this conception by assuming that a person’s agency always functions simply to maximise the person’s well-being (or “utility”). Consequently, agency and utility are reduced to the same thing, since the person is “obliged” to do whatever is objectively necessary to maximise their utility. By the converse argument of the third positivism, utility is by definition whatever it is that a person is maximising by their choices. Utilitarianism thus fails to distinguish between the ‘well-being aspect’ and the ‘agency aspect’ of a person, employing an impoverished conception of the person, lacking in subjectivity and any capacity for moral conflict — I would say, an unethical person. If for example, a person chooses to act in accordance with a moral imperative, they may do so at cost to their own comfort and happiness, but for utilitarianism, this is a contradiction in terms: doing the right thing is just something that this agent enjoys.
Part of Sen’s program therefore is to introduce agency, or “freedom,” into utilitarianism. This can be done for example, by taking freedom as a distinct component of well-being over and above utility. Thus, the expansion of functioning to capability allows the agency of a person to be taken into account. Whether this “works” is another question however: could economic science rely on the principle of agents acting to maximise their own capability? As an improvement of egalitarianism it makes sense, and it certainly makes sense as a critique from the standpoint of a social movement, but it is doubtful that it can be realised as an improvement of political economy. To suppose otherwise is like the “triple bottom line” movement, which believes that maximisation of the sum-of-utility is the result of economic agents being utilitarians.
Drawing on Sen’s introduction to the 1999 Utilitarianism and beyond, we can summarise his criticism of utilitarianism as follows.
“Utilitarianism as a moral principle can be seen to be a combination of three more elementary requirements: (1) Welfarism, requiring that the goodness of a state of affairs be a function only of the utility information regarding that state; (2) Sum-ranking, requiring that utility information regarding any state be assessed by looking only at the sum-total of all the utilities in that state; (3) Consequentialism, requiring that every choice, whether of actions, institutions, motivations, rules, etc., be ultimately determined by the goodness of the consequent state of affairs.” [Sen 1998 p 39]
As explained above, I do not accept (2) as essential to utilitarianism, since sum-raking applies exclusively to the utilitarianism of the first and second positivisms. The Pareto theoretic, in my view, retained the essential features of utilitarianism whilst disposing of the sum-total, and Sen at times supports this view as well:
“If interpersonal comparisons of utility are dropped, but nevertheless utility is regarded as the only thing of intrinsic value, then Pareto optimality would be the natural surviving criterion, since it carries the utilitarian logic as far forward as possible without actually making any interpersonal comparisons of utility.” [Sen 1998 p. 38]
Though Sen also refers to this move as ‘post-utilitarian’:
“This particular tradition has been carried into the post-utilitarian phase of welfare economics, concentrating on Pareto optimality and efficiency.” [Sen 1998 p. 49]
Sen actually proposes a “return to a more full-blooded utilitarian conception” of “interpersonal comparisons of utility” because, rightly in my view, he does not accept that interpersonal comparisons of well-being or advantage are in-principle ruled out. I think Sen’s work on functioning provides an objective measure of well-being and his argument that partial ordering presents no barrier to objectivity are entirely convincing. Taken together, I think these considerations dispense with the logical positivist ban on interpersonal comparison of well-being.
Up to the point of functioning as a second approximation to utility, I think this critique stands up as a valid internal criticism of utilitarianism. Since the conception of functioning facilitates the ethical weighing of non-market choices alongside goods acquired through the market, there is a further aspect in which Sen rehabilitates the original broader horizons of both political economy and utilitarian ethics.
Sen points to three key problems in transcending utilitarianism:
(1) “First, we have to distinguish between the ‘well-being aspect’ and the ‘agency aspect’ of a person;” (2) “Second, the utilitarian conception provides a defective (and systematically biased) view of well-being ... a very inadequate reflection of value — indeed even of what the person himself or herself actually values, not to mention what he or she would value on serious and courageous reflection, freed from the limitations imposed by unfavourable circumstances.” (3) “Third, a person’s freedom can be seen as being valuable in addition to his or her achievements. A person’s options and opportunities can be seen as counting in a normative evaluation, in addition to what the person ends up achieving or securing.” Sen 1998 p. 58-60.
(2) above refers in part to the fact that people may ‘accustom’ themselves to poverty, ill-health, discrimination and unfreedom, and indeed may know of nothing else, but a person’s well-being can and must be judged by what the person would value if they had the capability to enjoy a given freedom.
In relation to the other two points, insofar as the ‘agency aspect’ is dealt with in terms of the value a person attaches to freedom as in (3), for example, acquiring an education for the purpose of enhancing life choices, then such a criticism can function as an “internal” criticism of utilitarianism. However, when Sen proposes in (1) a break with the unitary, uncontradictory, impoverished conception of the person, he goes beyond utilitarianism. It is the objectification of human beings which is essential to utilitarianism
One of the great strengths of Sen’s subject position, is that he takes the consequences of an ethical principle seriously. Utilitarianism is characterised by the fact that it is almost exclusively consequentialist and Sen essentially embraces this consequentialism. But he does not accept that “the ends justify the means,” and a certain compromise with deontological ethics is necessary.
Sen deals with this in characteristic fashion by combining “process” with “culmination outcome” as “comprehensive outcome.” So for example, a politician may wish to be elected, but she wants to be elected fairly, not by fraudulent means; the outcome of “fairly elected” is a “comprehensive outcome” incorporating a deontological element within the consequentialist framework.
The need for some kind of “third point” between utilitarian consequentialism and libertarian deontology seems inescapable, but the manner in which process is to be valued and rolled into the value of the outcome remains somewhat unclear. Can the value of the fairness of an election be added in with the value of success in an election to give a total value for a fair election, compared with the value of a fraudulent election or a fair but unsuccessful election?
What Sen seems to be striving for here is a way to incorporate into a paradigm of distributive justice, concerns which are not normally compatible with conceptions of distributive justice. Isn’t there a danger that the move of “comprehensive outcome” achieves the same result as reducing agency to “utility-maximising"?
Sen is not a utilitarian. He does not believe in the thesis that the market delivers the best of all possible worlds in any of the four historical versions of this thesis. Sen is an egalitarian who is wrestling with the complexities of understanding what it is that we measure when we construct distribution, how do we measure distribution when we demand equality, what agency decides and implements appropriate measures, and how can inequality be ameliorated without making inroads into freedom, when well-being has been determined as freedom?
Sen is more than aware that an egalitarian principle exists only in connection with the agency that ‘enforces’ it. But having determined advantage as voice and ultimately critical voice, does it any longer make sense to talk of “distribution"? It would seem that any principle of distributive justice necessarily invokes an agent, a state in fact, whose responsibility is the management of distribution.
But surely, when it is critical voice, or even voice, which is being “distributed,” we have gone beyond not only utilitarianism, but egalitarianism, for precisely to the extent that the distributive agency (the state) is successful in distributing critical voice, it makes itself historically redundant; the state would be one subject among others.
For more than 50 years, long before the notion of critical voice found its way to the centre of his work, Sen has been wrestling with this paradox in terms of Arrow’s social choice theory.
Sen tells the story of the origin of Kenneth Arrow’s famous Impossibility Theorem as follows:
“In 1948, Olaf Helmer, a logician at the RAND Corporation, wondered about the legitimacy of applying game theory to international relations (’the “players” were countries, not individuals’), and asked young Arrow, a PhD student, ‘In what sense could collectivities be said to have utility functions?’ Arrow replied that ‘economists had thought about that question and that it had been answered by Abram Bergson’s notion of the social welfare function’. As Arrow settled down to writing an exposition for Helmer, he was soon convinced that no satisfactory method for aggregating a set of orderings into one ordering existed. The impossibility theorem and related results and their proofs came within ‘about three weeks.’ Arrow changed his dissertation topic to reflect the new finding, and sent off a brief exposition of the result to the Journal of Political Economy at the request of the editor.” [Sen 2002 p. 330.]
A diplomat (for example) needs to have some way of knowing the preferences of the group she represents, but it is a big step from this legitimate, if specialised, need to the idea of establishing the principles of justice for a whole society.
“It is already to make some substantive political assumptions to suppose that there is or should be one sovereign decision centre to determine what is right, even within a limited time span, for society as a whole.” [Sen 1999, p. 2.]
Social choice theory concerns a group representative whose task is to determine a group preference from the ordered lists of choices submitted by each group member. These “ordered lists” are supposedly lists held by each member of the population in which they have placed every possible state of the world in order of preference. These preference lists are then handed to the “returning officer” to compute the group’s preference list.
Unsurprisingly, only the most abstract imaginable problems find any solution in this theory. The observation, which goes back to the Marquis de Condorcet, that there is no consistent way of deciding between any more than two options by majority voting, already contained the essential truth which is extended and generalised in Arrow’s theorem. The “group preference” is in reality nothing more nor less than the preference arrived at by whatever process the participants submitted themselves to.
Once we have disposed of metaphysical illusions about the sovereignty of majorities, the validity of answers to multiple-choice questions as expressions of individual opinion and the mystifications like “public opinion,” there still remains the problem of how, and on what grounds, a group representative can determine their own obligations. Sen suffers from none of the above illusions, but he comes back again and again to Arrow’s social-choice theory.
In a strong sense, social choice theory is to parliamentary politics what utilitarian ethics is to political economy, and all four are closely tied together.
Like utilitarianism, social-choice theory takes the preferences of the agents as given, effectively as “externals.” Social choice theory concerns itself only with the procedures by which any combination of preferences, however absurd or evil, could be processed to produce a group preference. All ideas about agents discussing with one another, changing each others’ mind, doing deals, making promises, sympathising with one another, problem-solving, or whatever, lie outside the bounds of and are irrelevant to social choice theory.
Social choice theory, which is concerned only with the problem of the group representative or “returning officer,” is therefore closely tied to a very specific subject position, the same subject position as that of egalitarian ethics. A critique of social choice theory is therefore a critique of parliamentary politics, but with the same qualifications as we made in respect to the critique of utilitarianism. “Internal criticism” aimed at improving the processing of agent preferences into a group preference is wedded to that type of government which alienates people from government and utilises illusions about parliament and “public opinion” to maintain the fraud. A real critique of social choice theory can only be directed at exposing the fraud of parliamentarism, and re-orienting attention to the formation of genuine social subjectivity.
Sen holds to the point that however people arrive at their preferences, there still remains the problem of processing them into a group preference:
“Analyses of dialogues and exchanges, and of their impact on individual preferences can indeed be important for social choice theory. While there has not been any denial of the importance of such communication in contributions to social choice theory, this has not been a particularly active area of investigation within the discipline. ... As and when the set of individual preferences alters, there would be related alterations in the corresponding social choices, and in understanding this relationship, social choice arguments of the standard kind would continue to be relevant. It should also be noted that the extensions which are called for in investigating preference formation would often require substantive empirical presumptions, regarding what can and cannot be plausibly achieved through dialogues or swaps , taking us beyond the thoroughly analytical format of traditional social choice theory.” [Sen 2002a]
And against James Buchanan’s argument for consensus decision-making, he argues:
“Difficulties in social choice arise precisely because unanimity does not exist on many questions. What do we do then? One answer is to insist on unanimity for a change, and if there is no such unanimity for any proposed change, then to stick to the status quo. ... Marie Antoinette’s opposition to the First Republic would have saved the monarchy in France.” [Sen 2002a]
Sen continues to look for a process by means of which a “returning officer” could determine a group preference without troubling the voters to sort things out amongst themselves, but he argues that more has to be taken into account than simply the list of preferences. In arguing against the use of real income as the measure of well-being and for the use of functioning instead, Sen also argued that the former was lacking in information content. In the 1999 Development as Freedom, Sen sums up his criticism of social choice theory along similar lines:
“The informational base for this class of rules, of which the majority decision procedure is a prominent example, is thus extremely limited, and it is clearly quite inadequate for making informed judgments about welfare economic problems. This is not primarily because it leads to inconsistency (as generalised in the Arrow theorem), but because we cannot really make social judgments with so little information. Acceptable social rules would tend to take notice of a variety of other relevant facts in judging the division of the cake: who is poorer than whom, who gains how much in terms of welfare or of the basic ingredients of living, how is the cake being ‘earned’ or ‘looted’ and so on. The insistence that no other information is needed (and that other information, if available, could not influence the decisions to be taken) makes these rules not very interesting for economic decision making. Given this recognition, the fact that there is also a problem of inconsistency — in dividing a cake through votes — may well be seen not so much as a problem, but as a welcome relief from the unswerving consistency of brutal and informationally obtuse procedures.” [Sen 2002, p 252]
and in perhaps the most damning criticism of all, going to the issue of agency:
“... in arriving at social choice solutions of diverse views on systemic process concern, preferences cannot do all the work. In particular, rules of aggregation are processes too, and they are needed to do the social choice exercise of combining diverse views (even about systemic processes). Rules that fix the constituent features of the overall arrangement for aggregation are sometimes called “the constitution” — in terms of which individual preferences are put together to arrive at a social choice. For example, in the Arrovian system, rules such as the independence of irrelevant alternatives and the Pareto principle are not themselves put to a vote. In fact, if these rules themselves were to be determined by a ‘prior’ voting mechanism or some other social choice process, there would, then, be a need to have other rules governing the choice of these ‘prior’ social choice mechanisms. At some stage or other, some rules would have to come from outside the immediate domain of individual preferences.” [Sen 2002a, p. 626]
With these two paragraphs, Sen effectively damns Arrow’s social choice theory. This belies the fact the Sen has spent 50 years with this theory, including setting up whole academic departments to study and research it, describes Arrow’s Impossibility Theorem as “momentous,” seizing every opportunity to virtually canonise Arrow himself, and even the book from which the last quote is taken (written three years after Development as Freedom) spends about 500 of its 700 pages eulogising the “Arrovian” social choice theory.
This is strange indeed, but the contradiction is inherent in Sen’s own subject position, a contradiction inherent in a society in which the major decisions affecting everyone’s lives are taken not by the person themself, but by governments, governments posing as “group representatives.”
Beginning from an immersion in economics, through a painstaking critique of existing notions of utilitarianism and egalitarianism, Sen has arrived at the conclusion that the essence of well-being and advantage is not wealth, or functioning, or capability or even voice, but critical voice.
Egalitarianism must therefore address itself to the distribution of voice and critical voice in particular, rather than just real incomes, functioning or even capability. But this does not and cannot produce any kind of magic formula for a just distribution.
“Foundational ideas of justice can separate out some basic issues as being inescapably relevant, but they cannot plausibly end up, I have argued, with an exclusive choice of some highly delineated formula of relative weights as being the unique blueprint for ‘the just society’.” [Sen 1999, p 286]
The whole point is that to the extent that people have a critical voice in the social arrangements determining their own life, then they can determine those arrangements in discussion with others affected by those same social arrangements.
Through his critique of distributive justice, by introducing into the heart of a fundamentally distributive conception, a concept of recognition (critical voice), Sen seems to have transcended the opposition between distributive and recognition paradigms of justice.
The conception of critical voice at which Sen has arrived has only a sketchy, sociological elaboration in Sen’s own work, but there does exist a vast literature of social theory and moral philosophy on critical voice.
The result is by no means unproblematic. Equitable distribution of critical voice is a kind of utopian conception akin to the “withering away of the state.” Nevertheless, the concept can still function as a regulative ideal. The social measures which tend to give people critical voice are well-known — education and literacy, a free press and communication media, property rights, access to the labour market, freedom of belief and association, freedom to travel, as well as security, public health, food, land and shelter. These factors can be assessed from the point of view: how does a measure contribute to the distribution of critical voice?
Perhaps Sen’s work points a new conception of the relation between the politics of distributive justice, the politics of recognition and the new anti-corporate movements? Sen’s conception is not just of paradigms or social movements which are antagonistic to one another, but rather that the “new” social movements arose by way of critique of redistribution as a paradigm of justice, and that the notion of “critical voice” is emerging, again, by way of critique of recognition notions of justice.
The relation between these competing paradigms of justice is therefore one of sublation or transcendance, i.e., negation in the Hegelian sense, rather than negation in the simple sense of being competing social forces.
Amartya Sen, Choice of Techniques, Basil Blackwell 1960;
Amartya Sen, Collective Choice and Social Welfare, Holden-Day Inc. 1970;
Amartya Sen, On Economic Inequality, Clarendon Press 1973;
Amartya Sen, Poverty and Famines. Essay on Entitlement and Deprivation, Clarendon Press, 1981;
Amartya Sen, Equality of What?, first published in S. McMurrin (ed.) Tanner Lectures on Human Values, Cambridge University Press 1980, also in Amartya Sen, Choice, Welfare and Measurement, Blackwell 1982;
Amartya Sen, Commodities and Capabilities, Oxford University Press 1987;
Amartya Sen, Inequality Re-examined, Oxford University Press 1992;
Amartya Sen, On Economic Inequality, Enlarged Edition, Oxford University Press 1997;
Amartya Sen, On Ethics and Economics, Oxford University Press 1998;
Amartya Sen Autobiography, Les Prix Nobel 1998;
Amartya Sen, Introduction in Utilitarianism and beyond, Cambridge University Press 1999a;
Amartya Sen, Development as Freedom, Oxford University Press 1999;
Amartya Sen and Jean Drèze, India: Development and Participation. Oxford Uni. Press, 2002;
Amartya Sen, Rationality and Freedom, Harvard University Press, 2002a;
The Capability Approach, at the Asia Centre, Harvard University.