A shrug at the dole queue

Paul Ormerod ponders the harsh economic realities of a problem of our times.

This book originated in Nobel prize-winning economist Robert Solow's 1997 Tanner lectures on human values at Princeton University. It consists of the text of the lectures themselves, plus commentaries from four distinguished American scholars whose work spans the social sciences: Gertrude Himmelfarb, Anthony Lewis, Glenn Loury and John Roemer.

Solow addresses a major social and economic problem of our times: how to get people off welfare and into jobs. His examples of policies are drawn from America, but his analysis is just as relevant to Continental Europe, where unemployment is considerably higher, and to the UK, where the government has put a large amount of effort into its New Deal for the unemployed.

Despite its brevity, this book justifies completely the use of the phrase "it cannot be recommended too highly". The clarity of Solow's thought and the elegance of his exposition mean that the work is accessible to anyone with an interest in this important issue. It is an object lesson in how economics should be done.

Solow works firmly in the tradition of Alfred Marshall, in which economics is seen as "an engine for the discovery of concrete truth", and not as a branch of mathematics. He sets out some general principles, supports them with empirical evidence and follows through their implications. In the course of his arguments, Solow summarises neatly a large amount of technical econometric literature while avoiding the use of mathematics himself. Indeed, he is politely sceptical in his concluding remarks about the value of the only piece of maths in the book, the model set up in Roemer's commentary, even though it supports his arguments.

The theoretical underpinnings of policy in Britain and Europe are based on the view that, left to its own devices, the free market will solve the problem. The existence of unemployment is therefore due mainly to obstacles that prevent the market mechanism from working, hence the emphasis on the need for "flexibility" in labour markets.

Solow does not deny that in principle such a policy might work, but he is sceptical about its practical chances: "The standard discussion rests on the tacit belief that all the problems lie on the supply side of the labour market; kennel dogs need merely to act like bird dogs, and the birds will come. But that is a Panglossian error." He argues that the number of jobs will not respond one-for-one to the number of offers to work, and to the extent that it responds at all, it will do so by forcing low wages even lower.

A key empirical question is therefore the responsiveness of the overall demand for labour to the real wage. In other words, if labour markets can be made more "flexible" and the price of labour pushed down a bit, how many extra jobs will be created? The theory behind the influential work of Stephen Nickell and Richard Layard of the London School of Economics suggests that the answer is - a lot. An important forthcoming article by Bob Rowthorn in the Cambridge Journal of Economics shows that this conclusion rests on a specific assumption that lacks empirical support, and that in general the answer is - only a few.

Solow points out that there has been a lot of research on this very question and, in a phrase typical of his measured style, he states: "I think it is fair to say that the measured responsiveness has been disappointingly small." The evidence suggests that a 2 or 3 per cent fall in the overall real wage level is needed to increase the demand for labour by 1 per cent -a figure very close to that obtained separately by Rowthorn. This has important implications for Europe, where unemployment stands at well over 10 per cent of the labour force. To increase the number of jobs by just 5 per cent would, according to Solow's view, require reductions in average wages of between 10 per cent and 15 per cent.

But the problem with this policy approach goes even deeper for, as Solow makes clear, the required change in the average wage is not the figure that matters. The unemployed are considerably less skilled than many of those in work, and they compete potentially not with, say, the partners of Goldman Sachs, but with relatively unskilled members of the labour force. It is the real wages of the unqualified that would have to fall in practice to price people back into work. So to produce a reduction in average wages of only 2 to 3 per cent, a fall in unskilled wages is required of, in Solow's laconic phrase, "considerably more than 5 per cent".

In Europe, the harsh reality of true "flexibility" is usually disguised behind schemes to re-train the unemployed, to increase their overall skill levels. The British government, with its obsession with education, believes that in this way unemployment can be reduced painlessly. Solow's evidence is a salutary and depressing reminder that things are not that simple. He points out that many US states have operated schemes like this for the unemployed. Even more important from a scientific perspective, a number have carried out experiments with controls. Some of the unemployed have been given training and others have not, and the subsequent labour market experiences of the two groups have been compared.

The largest and best-documented example is in California, where an experiment known as Gain (Greater Avenues for Independence) has been carried out. The training given to the unemployed was intensive. The results were disappointing. About a third of welfare recipients held a job at one time or another during any year; participation in Gain increased that fraction by four to six percentage points. Solow points out that similar studies have been carried out in a dozen other states, but "none of them offers grounds for optimism about the ability of welfare recipients to find and hold jobs, or to earn a decent living".

An important aspect of the book is the interchange between Solow, Loury and Himmelfarb on the issue of social norms. This question is usually neglected by economists, who assume that the tastes and preferences of individuals are fixed. Solow, however, recognises that this may well not be the case. The larger the fraction of any given population on welfare, the weaker the social norm that favours work becomes. Any social norm is strengthened by frequent observation and weakened by frequent violation. Solow argues that, as a result, "any induced weakening of the norm of self-reliance will tip some marginal cases into the welfare pool". But his emphasis is very much on the word "marginal". He is reluctant to give much weight to arguments that regard poverty and unemployment as being in part due to social norms, with the implication that problems stem from the attitudes of some individuals in these positions.

This aspect of the debate is where Solow's position is potentially at its weakest. Himmelfarb draws on her own extensive work of the history of welfare systems to advocate the Victorian attitude towards the work ethic. She points out that a prime tenet of Victorian liberalism was that the more effective the internal exercise of morality, the less need there would be for the external, coercive instruments of the state. Himmelfarb essentially endorses the long-standing concept of a deserving and undeserving poor, and argues that policy should be based on this. In similar vein, Loury is sceptical about the ability even of a policy of direct publicly-funded job provision to motivate those trapped in the inner-city ghettos. He states that attachment to the labour force is low among residents of such areas, but so too is their attachment to marriage, school and obeying the law. "Which is cause and which is effect?" he asks.

The scientific problem here is a fundamental one of a lack of effective evidence with which to arbitrate between the two positions. Economists have not built the analytical models needed to give a theoretical framework in which to judge the empirical evidence. For they have been very reluctant to acknowledge the fact that individual behaviour can be affected directly by the behaviour of others, rather than merely indirectly via the price mechanism. And while social scientists from other disciplines have provided a huge amount of evidence to show that such effects exist, they are not in general inclined, whether by training or by aptitude, to set such results in a formal, analytical framework. Here is a real multidisciplinary challenge for the social sciences.

Solow's own proposals are based on the recognition that any set of policies "will have to deal with a substantial minority of current welfare recipients capable of only sporadic work, and with a larger group whose earning power is very low by the standards of society". This is a polite way of making points that may seem brutal, but are supported by the evidence. In the current phase of capitalism, the market mechanism is incapable of creating sufficient numbers of jobs for the number of unqualified workers that exists. And, while a small number of individuals may be able to better their positions by training and education, this is not possible for the unskilled as a group. The overall distribution of ability can be altered, but only slowly, and even in the medium term it can be regarded as effectively fixed.

The only realistic solution for Solow is to create low-skill jobs in the public sector. He is sceptical, for good reasons, about the ability of subsidies to private-sector employers to make much difference and argues that "there will have to be a determined and expensive effort to increase the demand for unskilled labour". He is at pains to emphasise the cost of such an approach. Moreover, in his elegant phrase, "the supply of altruism is limited". In other words, taxpayers are reluctant to finance such schemes. He is therefore willing to endorse aspects of the welfare-to-work approach, of an element of harshness in the choice between welfare and work, to try to increase the willingness of the electorate to pay for the jobs.

He is careful not to overstate his case, unlike many scholars who operate in the policy world, and recognises its potential limitations. But the real value of the book is the devastating way in which Solow combines analysis and empirical evidence to undermine many of the ideas fashionable in western policy circles.

Paul Ormerod is chairman, Post-Orthodox Economics.