Distributive justice and utilitarianism

Distributive justice and utilitarianism
Utilitarianism’s handling of the reality of heterogeneous populations and consequent inequality is perverse
Utilitarianism is one of several ethical theories addressing the question of how to assess the “goodness” of any state of affairs. In the history of ideas, the most distinguished proponents and defenders of utilitarianism have been the great English thinkers Jeremy Bentham (1748-1832) and John Stuart Mill (1806-73).

Bentham’s is one of the foremost names in British jurisprudence: A man who held the notion of “natural rights” to be “nonsense upon stilts”, he nevertheless also fought for the abolition of slavery. J.S. Mill, a child prodigy who grew into an adult genius, is known (apart from his contributions to the principles of political economy) for his tracts On Liberty and On Utilitarianism. Utilitarianism, as encompassed in what is called the “Benthamite Social Welfare Function”, is a staple feature of contemporary economic theory. Among economist-philosophers in the modern era, a notable proponent of utilitarianism has been the Nobel Prize-winning game theorist John Harsanyi.

In summary form, utilitarianism is well-served by that clichéd description of it as demanding “the greatest good of the greatest number”. Stated more rigorously, the “goodness” of any state of affairs is taken to be reflected by the sum of the utilities derived from that state by all the individuals in a community (this sum is what is called the Utilitarian or Benthamite Social Welfare Function). In comparing the “goodness” of alternative states of affairs, utilitarianism will recommend that state in which the sum of individual utilities is larger. “Utility”, for present purposes, may simply be interpreted as “desire fulfilment”. (It is said that the famously irascible Cambridge economic theorist Frank Hahn, growing weary of philosophical definitions, once told his class that “utility is the damned thing you maximize”!) In order to be able to sum utilities across individuals, some special assumptions about personal utility must be made—in particular, that utility is “cardinal” (which permits statements such as “my utility is twice as much in state x as in state y”) and also “interpersonally comparable”, that is, commensurable across individuals.

What does utilitarianism have to say about distributive justice? To fix ideas, let us ask: What is the optimal distribution of income that utilitarianism would prescribe in a two-person world? Typically, the assumption is that both individuals share the same utility function defined on income—and further, that utility increases as income increases, though at a declining rate, to accommodate the thesis of “diminishing marginal utility”. Notice that social welfare under utilitarianism is a sum of individual utilities. When will this sum be maximized? Suppose we have a fixed amount of income, say Rs100, to distribute between the two individuals, what is the optimal distribution?

Suppose the welfare-maximizing distribution is an unequal one, say, Rs40 to individual 1 and Rs60 to individual 2. Then, because of diminishing marginal utility of income, if we were to transfer some income from person 2 to person 1, the gain in utility to person 1 would be greater than the loss in utility to person 2, that is, on net, social welfare would increase with a transfer of income from the richer to the poorer person. And this must continue to be the case as long as one person has more income than the other. This is just another way of saying that an unequal distribution can never be a welfare-maximizing distribution. The optimal distribution of income, under utilitarianism, must always be an equal one.

The above equality result, some thought will reveal, is driven by the assumption of identical utility functions for the two individuals which increase with income at a diminishing rate. The condition for maximum social welfare is equalization of the two individuals’ marginal utilities—and given the assumption of identical utility functions, the accidental by-product of equal marginal utilities at the optimum is equal incomes. Take away the assumption of identical utility functions, and what do we have?

This is the problem that Amartya Sen addressed in his Radcliffe Lectures at the University of Warwick, subsequently published in book form in 1973 with the title On Economic Inequality. Specifically, he considered a situation in which, given the two-person world invoked earlier, person 1 experiences exactly half as much utility as person 2 at each level of income, because, let us say, person 1, unlike person 2, is physically handicapped. What would the utilitarian welfare-maximizing distribution of income now look like? At the optimum, the two persons’ marginal utilities would have to be equalized. This will happen when person 2’s income is twice the income of person 1.

Elementary considerations of “need” would prompt us to prescribe a greater share of income for the disadvantaged person 1. Utilitarianism does precisely the opposite. As Sen has pointed out, utilitarianism rewards the more efficient “pleasure machine”, instead of compensating the more needy disabled individual. By 1973, Sen was already anticipating his subsequently-developed “capability theory”, wherein well-being is seen to be a matter of human capability rather than desire satisfaction. Utilitarianism’s assumption of identical utility functions depends on the postulation of homogeneous populations. But the world as we know it has populations that are heterogeneous, that is, people are not identical in their non-income characteristics. Heterogeneity is of the essence of inequality, a problem which utilitarianism handles inadequately, indeed perversely.