18 JANUARY 2003, Page 29

Blair's lack of 'process'

Samuel Brittan

RATIONALITY AND FREEDOM by Amartya Sen Harvard, 426.50, pp. 752, ISBN 0674009 What is really wrong with the Blair government? The unease it excites is at least as strong on the articulate political Left as on the Right. Indeed the grounds for anxiety may overlap across the political spectrum. Until now it has been difficult to verbalise this sense of malaise. The citation of particular policies that are disliked, or even of the Downing Street style, is not sufficient.

It is only after looking at Atriartya Sen's new book of essays that the penny suddenly dropped for me. Unlike his earlier hook, Development as Freedom, which I reviewed here on 31 January 2000, this is a technical volume containing some of the papers for which he received his Nobel prize. It is concerned with major issues of freedom, welfare and human achievements, but at the rarefied level where political and economic theory and formal philosophy all meet. Yet unlike so many writers in this field, who are mainly concerned with their reputation among academics, Sen never forgets the more general reader looking over his shoulder. It is not an accident that one can skip most of the equations. In many of the essays, they are deliberately segregated to smoothen the path of the reader more concerned with substance than technique.

The last three chapters of this book are a revised version of the Arrow Lectures he gave in 1995 and which have never been previously published. It is here that he divides the goals of public policy into two aspects. The first is what he calls 'opportunities'. This refers to the range of choices that people have and corresponds roughly to the traditional idea of economic welfare, but goes much beyond what is usually measured in estimates of Gross Domestic Product. The second aspect consists of what he calls 'process' — matters such as the constitutional and legal system, which determine how decisions are taken and the scope of personal freedom.

Much of Sen's argument is a critique of other thinkers who put too much emphasis on process at the expense of opportunities. Such thinkers are what the vulgar have in mind when they talk about 'market fundamentalists', but if they knew what they were talking about they would direct their criticisms, as Sen does, not towards politicians, but towards writers such as Nozick, Buchanan and Hayek. As Sen points out, it is absurd to judge a social system entirely by concepts such as freedom and the rule of law without any attention to welfare or the distribution of opportunities among citizens.

The opposite error is utter indifference to process and concern only with results. There could hardly be a clearer example than Tony Blair. He is not mentioned in Sen's book; but he is surely a demonstration example in his stated preoccupation with 'what works'. This comes out in his indifference to the way in which the House of Lords is reformed, to the fairness or otherwise of the electoral system and in his uncritical support for policies that cut across traditional protections of citizens' rights such as trial by jury, the rule against double jeopardy, the presumption of innocence and even habeas corpus. It comes out too in his insouciant support for imprisoning parents whose children play truant and much else besides. Maybe one or two of these policies may be justified as regrettable expedients: but any critical scrutiny is brushed aside in a concern with hoped-for immediate results. In philosophical terms Blair could be described as an 'extreme act utilitarian', whether he uses such terms or not.

A defender of Blair could point out that at international level the British prime minister is extremely concerned with process. For instance he has used all his influence with George W. Bush to persuade the latter to take the UN route in the Middle East rather than unilaterally attempt to eliminate Saddam Hussein. If this were one of the surfeit of articles probing Blair's psychology, it would be necessary to ask whether there really are two different Blairs in internal and foreign policy or whether the Prime Minister just thought that he could not get away politically with supporting a hawkish stance on Iraq without attempting the UN route first.

But I use the Blair illustration mainly to show that there can be practical applications for even the most abstract of studies in this volume. If the reader feels that some of the discussion is hair-splitting, the quick response would be that Sen in practice agrees. The subject of Welfare Economics was established near the beginning of the 20th century by a Cambridge professor, A. C. Pigou. Although extremely unworldly himself, Pigou used rudimentary economics to analyse some of the defects of a pure market system in down-to-earth fashion. He cited elementary examples of spillover effects, such as smoking chimneys and ash from locomotives, as well as the perennially absorbing subject of income distribution. Today his work has many applications both in commonsense environmental regulation as well as in limiting the excesses of extreme 'greens'.

Later economists observed that many of the environmental defects which preoccupied Pigou were due to lack of markets and property rights rather than excess of them. The obvious contemporary example is the incentive to overfishing given by lack of ownership of sea-beds. A different criticism is that market failures need to be balanced against governmental failure, due to the imperfections of the political market.

But the reason why Pigou's book was out of print for so many decades had nothing to do with these relevant qualifications. It was to do rather with the doctrine that interpersonal comparisons of welfare were impossible — even though all of us make such comparisons every day. This was a piece of amateur philosophy by economists who confused sceptical doubts in the higher range of metaphysics with doubts about the realities of everyday life. It was nevertheless accepted as dogma for many decades; and economists tied themselves in knots trying to work out assumptions on which judgments could be made. The leading example was the so-called 'Pareto improvement', which is supposed to make some people better off and no one worse off and does not exists in any large-scale modern society.

Part of Sen's achievement has been to clear away some of this logic-chopping and introduce some common sense. He long ago showed, with a homely example involving Lady Chatterley's Lover, that a 'prude' and a 'lewd' could both be made better off in Pareto terms by prohibitions which curtailed basic human freedoms.

The Nobel laureate does not merely engage in these intellectual battles to win his spurs, and therefore his right to be respected on actual policy matters such as development economics and the treatment

of famines on which he has written extensively elsewhere. He takes seriously, for instance, the paradoxes of voting theory. As long ago as the 18th century the French philosophe, the Marquis de Condorcet, showed what a deceptive slogan majority voting could be. It is easy to cite examples in which policy or person A is preferred to B, B is preferred to C, but C is preferred to A. These paradoxes have come to haunt those who devised rules for the election of party leaders, paradoxes which superficial Conservative reformers have tried to sidestep by throwing open the election to the elderly activists of the constituency parties. Two centuries later another Nobel prize-winner, Kenneth Arrow, has shown that if a few simple conditions are required such as non-dictatorship and absence of irrelevant alternatives, no voting rules or any other decision method can guarantee a result which is clearly preferable to any other. Arrow has regarded his theorem not as an excuse for obstructionism, but as a challenge to find ways round it and to which Sen has eagerly responded.

Although Rationality and Freedom is not bedside reading he would indeed be dull of soul who could not find something interesting and challenging in its pages. Anyone who claimed to agree with all of it or to take it as a social democratic bible would show that he had gained little from it. My own long-standing dissent is over his dis missal of Darwinian psychology. What seems to put him off is Darwin's assessment of progress in terms of reproduction and population growth. In that case the most successful species would be ants or possibly some minute creatures of which non-biologists have not even heard. But surely this confusion of progress with mere multiplication was a foible on the part of Darwin and some of his followers. The serious part of Darwinism is reproduction with variation. 'Reproductive fitness' is a necessary but not sufficient condition for moral progress and can in any case be seen in local terms relative to particular environments and situations.

The most intellectually radical aspect of Sen's multifaceted book is his discussion of rationality, which I hope he will take further in the forthcoming companion volume. Logicians have concentrated mainly on deductive logic where statements are true or false by definition and have left the sphere of rationality in choice and action to economists and so-called decision theorists. These have too often slipped into identifying rationality with the maximisation of self-interest. When it is pointed out that many human actions, ranging from the activities of Mother Theresa to the individual act of voting which has a negligible effect on an election outcome, do not make sense on this basis, economists have tended to make their assertions true by definition: in which case it is difficult ever to say that a person has acted irrationally.

The textbook definition of rationality is the ability to compile a complete, consistent and transitive preference order, and to

choose actions accordingly. Sen dissents from such definitions on two grounds. First he insists that rationality requires critical reflection about one's preferences which can go quite far without breaching Hume's law that an 'ought' proposition cannot be derived from an 'is'. Secondly, it is possible, he believes, for some choices to be more rational than others even if preferences are not complete and consistent. The price of such modification is that there is no 'sure-fire' test of personal or political rationality.

A broader concept of rationality suggests that a voter is not being stupid if he goes to the polls to fulfil his duty as a citizen or as a gesture of solidarity with a political party. But the narrower economic analysis comes into its own when he sees how quite modest obstacles such as bad weather or a popular football match on television can reduce turnout and the lengths to which politicians go to prevent at least the latter type of disturbance.

Sen's reflections on rationality had the effect of making me reconsider not my objections to a legal minimum wage, but my grounds for them. It is tempting to regard such legislative interference as an irrationally costly way of achieving a reduction in poverty. But reflection soon leads one on to endless complications. For instance, how do you evaluate the poverty of those forced out of jobs by a sizeable minimum wage against the higher wages of those who retain jobs? If you want to help the latter by ensuring that unemployment pay is at least as high as the minimum wage, what are the incentive effects on economic performance of paying and financing such a dole? On the other hand, if one prefers the apparently more rational course of in-work benefits to top up the pay of those incapable of earning a conventional minimum, one has to ask what effect this will have on pre-benefit market pay and also on the work incentives of those who have to finance the top-ups? In the end one is led to look for an impossible-toachieve model of the whole economic and social system.

It is surely more straightforward to take the 'process' route and to say that a legally enforced minimum wage is a breach of my freedom to make a contract with another person. This is not acceptable when adverse third-party effects can be offset through social-security top-ups which are a reasonable procedure whether or not their pure economic 'optimality' can be proven. It is the great value of Sen's book that he tempts one to investigate fresh examples and to realise that the world is a more complex place than it might appear to the supporter of an ideology, even one's own.