Political economy is much, perhaps most, of the time on the side of the angels. It teaches people that resources are scarce, realities cannot be wished away, the consequences of well-meaning government action are often painfully surprising, and that you cannot have it both ways. At times, though, political economy will implant beliefs in the public consciousness that lend legitimacy and moral support to policies that democratic governments are only too inclined to adopt. The “infant industry” argument against free trade is one example, and there are many others. The most important is probably utilitarianism that used to hold the intellectual high ground from the time of Bentham to that of Pigou. It accustomed the public to the strange idea that happiness, satisfaction, or “utility” was a homogenous entity that pervaded society and increased or decreased as desirable goods and services enjoyed by individuals increased or decreased, though not necessarily in the same proportion. As a rule, as an individual’s budget of goods, i.e. his income increased, the “marginal utility” of the income decreased. The marginal dollar given to a poor man generated more “utility” than the same dollar given to the rich man. Simple arithmetic told you that total utility would increase if you gave to the poor and took from the rich, and that total utility reached its maximum when all incomes were equalised. Redistribution enhanced the common good.

The “new welfare economics” of the 1930s and beyond has recognised that adding different individuals’ “utilities” was arrant nonsense and subtracting the losses of some from the gains of others was doubly so. As the English philosopher Philippa Foot tellingly put it, “there was only a black hole where the common good used to be”.

Following Vilfredo Pareto, a change in the social state of affairs was accepted as unequivocally better only if it was worse for none and better for at least one person, (whether “better” and “worse” meant ex ante preference or ex post satisfaction.) Changes that produced some gainers but also some losers were objectively non-comparable. Like any interpersonal comparison, they could be assessed by any observer as matters of his personal value judgment, but could always be challenged by another observer with a different value judgment. Within the wide bounds of grim misery and smiling abundance, which all observers would rank the same way, there could always be legitimate disagreement about the goodness or badness of a change that seemed good for some and bad for others. The common good or the public interest were up for debate and not fit subjects for scientific inquiry and agreement on facts.

It so happened, however, that many or most people who played the role of observer and habitually proffered judgments on matters of welfare, carried and still carry in their subconscious a large residue of the old utilitarian tradition and instinctively reason in the marginal dollar yielding a “greater utility” in the hands of the poor than in those of the rich. They thus draw a false conclusion about how to increase “aggregate” utility (instead of saying, quite correctly, that a dollar makes a greater difference to the poor than to the rich—a finding that is then confusedly transformed into a finding about aggregate utility and how to increase it). The upshot is that despite the Paretian logic of the new welfare economics, there is a climate of opinion that not only favours rich-to-poor redistribution on both altruistic and self-interested grounds, or out of envy and spite, but also imagines that its stand is supported by an obvious rational argument.

While a change involving straightforward redistribution in a definite direction can only be taken as good or bad by choosing between rival value judgments (which in turn involves further value judgments), certain other changes lend themselves to fact-finding as well as value judgment. The standard case is reorganisation of production—an investment project, a technological innovation, a shift in the terms of trade, excise taxes or other causes impacting relative prices—from which some people or groups gain but others lose. Attempts have been made to determine the net effect of such Pareto-non-comparable changes. All are question-begging, though some more so than others.

Friedrich von Hayek, along with his immense learning and magisterial intelligence, is occasionally of an angelic innocence that can be downright startling. In discussing whether the state should produce generally useful goods and services the market could or would not produce, he blandly offers this conclusion:”… the only question which arises is whether the benefits are worth the costs”.1

Discussing the compulsory acquisition of land by the planning authority, his conclusion is no less question-begging: “If they are to be beneficial, the sum of the gains must exceed the sum of the losses”.2

Of course, if the benefits are worth the cost, or the gains exceed the losses, all that needs to be said has been said; it is a tautology that the change is a good change. But the problem is precisely that the answer to the question that is being begged is so controversial and wickedly complicated. It has tormented some of the sharpest minds in economics. It would be reduced to relatively straightforward cost-benefit analysis if the preferences or satisfactions of taxpayers and beneficiaries, gainers and losers could be taken as adequately reflected by the sums of money they paid or received and where a cost of $100 paid by one person were exactly offset by a benefit of $100 received by another. However, nothing permits us to suppose this.

For more on Hicks’s compensation test, see the biography of John R. Hicks in the Concise Encyclopedia of Economics.

The famous Kaldor-Hicks3 theorem proposed a “test” of whether a change induced by a reorganisation of production was a good one. Little, in his equally famous Critique of Welfare Economics, rightly pointed out that the theorem formulated, is not a “test”, but a “definition” of a change being good; but this does not really affect the present argument.

Suppose that in a river valley there are many farms, each tilling some of the valley bottom and grazing sheep on the hillsides. A dam is then built at the valley mouth to produce hydro power for the region’s towns and cities. The arable land of the farmers is flooded and they are reduced to the hillside sheep runs. They lose. The townspeople of the region gain. By Kaldor-Hicks, the dam is a good thing if the gainers could over-compensate the losers, pay for the dam and have something left over.

One objection made to this definition of goodness was that if the change entails a “bad” distribution of income, it would not be good. It may make rich gainers super-rich and leave the poor losers as poor or nearly as poor as before, even if compensation for their loss was actual and not just hypothetical. This objection, of course, relies on a value judgment about income distribution and represents a subjective element in the assessment of the change.

A perhaps more devastating objection, made by Tibor Scitovsky4 was that under certain conditions the gainers could bribe the losers to accept the change and then the losers could bribe the gainers to undo it, thus reducing the “test” to absurdity.

The more arcane and fragile the theorem turns out to be, the more one feels the absence of an obvious question that it does not ask: why does a Pareto-non-comparable change, with its gainers and losers, exist at all? Without abnormally high transactions costs and unspecified obstacles, including incomplete information, the prospective profitability of a hydroelectric dam across the valley would attract entrepreneurs, one of whom would contract with the valley farmers for their bottom lands, with electric utilities for the future power supply, and with builders and engineers for the dam. There would be no losers to begin with; all parties would be made as well off as before and most probably better off. Instead of the Pareto-non-comparable situation that must leave us agnostic about its goodness, we would have a clear and simple Pareto-improvement “tested” by the fact that all parties entered their contracts voluntarily.

Believers in the prevalence of “market failure” could invoke transactions costs and the other “usual suspects” to explain why the state should be called upon to reallocate productive resources when gainers could compensate losers as a result. The burden of proof would lie with them to show why transactions costs and the usual suspects obstruct solutions by voluntary exchange but do not obstruct the state, thanks presumably to its proverbial superior competence.


Footnotes

Hayek F.A., The Constitution of Liberty, (Chicago, Chicago University Press, 1960) p. 222.

Op. cit., p. 351.

Kaldor N. “Welfare Propositions of Economics and Interpersonal Comparisons of Utility,” Economic Journal, September 1939, p. 549, and Hicks J. R. “The Rehabilitation of Consumer Surplus”, Review of Economic Studies, 1940-41, p. 108.

Scitovsky, T., “A Reconsideration of the Theory of Tariffs,” Review of Economic Studies, vol. ix, no. 2.


 

*Anthony de Jasay is an Anglo-Hungarian economist living in France. He is the author, a.o., of The State (Oxford, 1985), Social Contract, Free Ride (Oxford 1989) and Against Politics (London,1997). His latest book, Justice and Its Surroundings, was published by Liberty Fund in the summer of 2002.

The State is also available online on this website.

For more articles by Anthony de Jasay, see the Archive.