Ethics of choice, freedom and utility

Amartya Sen is as at home in philosophy as he is in economics. Peter Singer reviews Sen's latest book, whioch gives an overview of 30 years' thought

Amartya Sen, master of Trinity College, Cambridge, and 1998 Nobel laureate in economics, has been called "the conscience of economics". Reading Development as Freedom , it is easy to see why. Sen is unusual for an economist, not only in his choice of topics to write about - poverty, famine, inequality, women - but also in the extent to which he directly addresses the ethical issues raised by these topics. But then0 Sen was born in Bengal, where as a boy of nine he witnessed the 1943 Bengal famine,in which millions starved. And he is not only an economist. At Harvard, where he taught before he returned to England, he was a professor of philosophy as well as of economics, and his works in philosophy journals are highly regarded in departments of philosophy.

Development as Freedom began life as a series of lectures to the World Bank, a body that, the author tells us in his preface, "has not invariably been my favourite organisation". The volume that has emerged from the rewriting of the lectures focuses on development and is aimed at a broad audience. While those without any background in economics or philosophy may find some of the more abstract sections difficult, the book presupposes no technical skills or prior acquaintance with economics.

The most straightforward chapters are those that present factual material on such issues as inequality, famine, population and food. One of Sen's best-known claims, defended in his 1981 book Poverty and Famines , is that no famine has ever occurred in a democracy, and that this is not mere accident. It reflects, in his view, the fact that famine is not simply, or sometimes not at all, a result of a decline in food production. Whereas a fall of 12 per cent in food production caused a major famine in Ethiopia in 1983-84, a fall of 38 per cent in Zimbabwe in the same period did not, because the Zimbabwe government acted in time to avert a famine. Democratic governments, Sen claims, have incentives to prevent famines that dictatorships lack, and if they have a free press, they will not be able to deny the existence of the famine.

These claims are plausible enough, but few of the world's poorest nations are democratic, and it may still be only a matter of time before a tragic famine in one of the few that are proves the rule false. It could even be argued that at this very moment, the famine developing in Ethiopia shows the link between democracy and the prevention of famine to be illusory. But Sen phrases his claim carefully: "It is certainly true that there has never been a famine in a functioning multiparty democracy." Despite the progress that Ethiopia has made towards democracy since the overthrow of the Mengistu regime, it remains some distance from being "a functioning multiparty democracy", because in the elections held in 1995, one party won more than 80 per cent of the seats. It remains to be seen whether the result will be different in the elections due shortly.

But could a more democratic government have done things to prevent the famine that the present Ethiopian government has not done? Sen writes that "famines are very easy to prevent through regenerating the lost purchasing power of hard-hit groups", and he suggests the creation of emergency employment in short-term public projects. Perhaps, if it were not distracted by the calamitous war with Eritrea, a more democratic Ethiopian government would have done this; but will a poor nation always have the resources for such projects? It is hard to share Sen's confidence that famines are "very easy to prevent". On the other hand, it is certainly true that poor but democratic countries such as India have never seen famines like those that occurred in China after the failure of the "Great Leap Forward".

Sen's argument that democracy and a free press are the best safeguards against famine is part of the broader theme that runs through Development as Freedom : that freedom, far from being at odds with development, is a necessary part of it. As part of this theme, Sen denies the thesis, associated with Singapore's former prime minister Lee Kuan Yew, that civil and political freedoms hamper economic growth. The evidence, he claims, is not sufficient to establish a link between the authoritarian nature of the regimes of China, Singapore and South Korea and their economic success. Moreover, when poor people have been given a choice - as Indians were when they voted on Indira Gandhi's declaration of a state of emergency - they have voted in favour of democratic freedoms, rather than for an allegedly more prosperous authoritarianism. Sen denies, too, that there are "Asian values" that do not include the distinctively liberal values of the West. In a necessarily brief sweep through Confucian, Buddhist, Hindu and Islamic thought, he argues that there is diversity in all cultures, and advocates of liberty in China or Burma are speaking just as authentically for their national traditions as those who defend authoritarianism.

What, though, of population growth, which surely must be limited if economic development is to be sustainable in the long term? Is it not true that authoritarian China, with its coercive policies of refusing housing and other benefits to families with more than the accepted numbers of children, has been more successful than India in reducing the fertility rate? Again, Sen argues that it is not authoritarianism but literacy that is the key factor. Although India's fertility rate is 3.1, well above China's 1.9, the fertility rate of the Indian state of Kerala is only 1.7, which is similar to that of Britain and France. Moreover, Kerala's drop in fertility, no less rapid than that of China, was achieved without coercion,in a climate of free and open public discussion. Why is Kerala different from most other Indian states? Largely because of its high level of literacy, especially among women, where the literacy rate is higher than in China. Female literacy and female labour-force participation seem to be the key factors in reducing the fertility rate.

The issue of whether we may use coercive measures to reduce population growth brings out Sen's underlying ethical position. He is at his most persuasive when he argues that freedom, including the freedom to reproduce,is not a "luxury" for the poor, but something that is just as important to them as it is to the more affluent. They are, in fact, more liable to be treated brutally by coercive officials than are better-off people. If, as often seems to be the case, less coercive means could have achieved the same or better outcomes, then this coercion is a great wrong. But Sen does not take the view that coercion is always absolutely wrong. He seeks a middle position, between the utilitarian claim that the question should be decided by asking whether the use of coercion will maximise utility, and the libertarian view that we must not violate individual rights, no matter what the consequences might be.

All this is clear and sound. What is not so clear is whether Sen can succeed in distinguishing his position from utilitarianism, at least if that term is understood broadly. Sen describes his own position as "consequentialist", meaning that, in contrast to the libertarian, he judges actions and policies only by their consequences, not by whether they violate rights or other rules. But the relevant consequences by which the actions and policies are to be judged, he wants to insist, are not merely consequences in terms of utilities. Sen says that he attaches intrinsic importance to rights, and includes them among the goals to be sought. In support of this "goal-rights system" he points out that there is "no parity" between the importance that parents attach to a decision on how many children to have, and the importance that the decision has to others, including "the potentates running the government". This is true, of course, but it does not distinguish Sen's position from that of the utilitarian, who also must take account of the special importance of this decision for the parents. If the utilitarian were a hedonistic utilitarian,she would take into account the effects that being free to make one's own decision would have on the happiness of the parents; if on the other hand the calculation was being done by a preference utilitarian, he would consider the greater strength of the preference the parents have to decide the size of their own family. It is not clear what Sen's view does that the utilitarian cannot do.

The same point can also be made more broadly against other parts of Development and Freedom in which Sen attempts to put forward a new approach to ethics. He argues that "for many evaluative purposes" in assessing whether one state of affairs is better than another, we should not consider utilities but "capabilities", that is, "the substantive freedoms to choose a life one has reason to value". In other words, in deciding which social policy we should choose, instead of asking which policy will lead to greater welfare, Sen suggests we ask which policy will better maximise the capabilities of members of the society. But it is not clear why Sen considers this a better measure than welfare. Is expanding capabilities good in itself, even if it does not lead to greater welfare? None of Sen's examples suggests this; like the case of reproductive freedom mentioned above, they all suggest that greater substantive freedoms lead to greater welfare, at least if we understand welfare in the well-known terms of the preference utilitarian.

What other reason might we have for choosing the capabilities approach? Sen claims that it overcomes one well-known difficulty that utilitarians cannot solve: the problem of interpersonal comparisons of utility. Whether we consider utility to consist of pleasure and the absence of pain, or of the satisfaction of preferences, it seems impossible to judge whether one person's pain is as painful as another's, or whether one person's preference is as strong as that of another person. "Individual functionings," Sen tells us, on the other hand, "can lend themselves to easier interpersonal comparisons than comparisons of utilities (or happiness, pleasures, or desires)." But can they? Sen himself admits that "there are different functionings, some more important than others." How then are we to decide which "substantive freedoms" are the important ones? Sen suggests that we need to arrive at some kind of reasoned consensus on the weights to be given to different functionings. Like a good democrat, he wants this consensus to be achieved through "public discussion and a democratic understanding and acceptance". Why, though, should we need consensus? Why can we not just agree to differ, and say that the value of any given freedoms to any given person depends on how much that person wants to exercise them, or would want to exercise them under certain conditions of adequate information, reflection and the absence of other distorting factors? This seems more reasonable than trying to achieve a consensus by which we can all agree that, no matter what any given individual may want, it is more important for her to be, say, free to have access to uncensored media than to decide how many children she will have - or the other way around. But of course, once we accept that the value of any given freedom to any given person is, ultimately, a matter of that person's values, we are again back to preference utilitarianism.

It is ironic that on the very same page on which Sen argues for the need for a democratic consensus on what substantive freedoms are important, he criticises utilitarianism for "having just one homogenous 'good thing'". This, of course, is a criticism that bites only against hedonistic utilitarianism, and gets no hold against a view that says that what is good is as diverse and individualistic as what human beings are capable of preferring. What Sen might say is that the diverse values different individuals may place on freedom defy measurement, and thus are not useful for judging one state of affairs to be better than another. But if this is the problem with utilitarianism, we should acknowledge that it is not a problem with the ethic itself, but only with attempts to use it, directly, for the purposes of social choice. We might then accept the use of another tool, perhaps even some social consensus on the values of particular capabilities, as a measuring device that roughly gets at the underlying value we are attempting to maximise - namely, utility.

Development and Freedom breaks no major new ground, but its great merit is that it allows us to gain an overview of 30 years of the thought of a distinguished economist and philosopher who has ranged widely in his work. It therefore sometimes moves too swiftly over major issues to be satisfying. It will work best if it leads readers to pick up the much fuller treatments of each topic that Sen has published previously.

Peter Singer is professor of bioethics, Princeton University, United States.