Why do Britain's have-nots have less?

Tony Atkinson deplores the UK's record on income distribution and the relative neglect of the problem by economists.

Income inequality has increased more sharply in the United Kingdom in the 1980s than in other western countries. The gap between rich and poor has widened much more dramatically than in the United States. France, Germany and Italy have not seen the same transformation of the distribution of income as Britain. Increased inequality is not a worldwide trend.

Increases in inequality on the British scale are a new phenomenon. In the past, changes in income distribution have been dismissed as too insignificant to be worth attention. American economist Henry Aaron remarked that observing income distribution data "was like watching the grass grow". But this is no longer the case. Changes in the distribution are at least as large as changes in aggregate variables like gross national product. Corrected for the distributional changes, Britain's much vaunted economic performance in the 1980s turns out to be poorer than in the 1960s. The income of the bottom half of the population (not of course always the same people) grew slower in the 1980s than in the 1960s, and only a little faster than in the much criticised 1970s.

Why has inequality risen so much in Britain? Why have the benefits of growth been unequally divided? Why do some people today get paid Pounds 1.50 an hour and others Pounds 1.50 a minute? Why is wealth still concentrated? Why has the proportion of the population with incomes below half the national average more than doubled?

These are questions to which the ordinary person quite reasonably wants answers. But they are not questions that are central to present-day economics. In my presidential address to the Royal Economic Society conference in Swansea recently, I took the profession to task for paying only limited attention to the subject of income distribution. The society's journal, the Economic Journal, over the past half century has published on average only 1.4 articles a year on income distribution, compared with about four times as many on international economics. International economics is important for an economy such as that of the United Kingdom, but the distribution of income seems to me to warrant at least equal attention.

Moreover, few of the Economic Journal articles address the fundamental issue of explaining the distribution of income. The same is true of economics textbooks. One finds sections labelled "income distribution", but these are usually about how national income is divided between wages and profits. This is an interesting topic but is only indirectly relevant to understanding how the share of the bottom 50 per cent or the top 10 per cent is likely to have changed. What we want to know is how to use economic theory to explain what is happening to the incomes of individuals and families. What we need is research that ties income distribution centrally to analysis as to how the economy works.

I am not suggesting that the subject has been totally neglected, and there are signs that the situation is improving. Recently, there has been interesting research on one significant component of increased income inequality: widening earnings differentials. Much of this research agrees that there is a straightforward explanation of rising earnings dispersion: a shift in demand away from unskilled labour in favour of skilled workers. In the US and the UK, this has led to a fall in the relative wage of unskilled workers. In other European countries, the wage structure has not changed, in part because of minimum wages, and the result has been, on this explanation, increased unemployment of unskilled workers. The shift in demand is in turn explained by some economists by the liberalisation of international trade and increased competition from the countries where unskilled labour is abundant. Other economists argue that technical change has been biased towards skilled labour with the introduction of information technology.

Increased wage differentials can, apparently, be explained by the simplest of economic tools: supply and demand. But it is not clear that it is all of the story. What is striking about the change in wages since the 1970s is what Paul Krugman has called the "fractal" quality of increased dispersion: however narrowly one defines groups, one still finds increased differentials in almost all cases (the main exceptions are the public sector). Even in the army, the pay of privates has fallen by 15.3 per cent relative to average earnings since 1979, whereas that of brigadiers has risen by 18.1 per cent more than the average.

Economists may argue that such a pervasive increase in dispersion is due to skills being more highly rewarded in all occupations, but it is also possible that there is a more general set of social forces at work. Sociologists may argue that pay differentials are governed as much by social custom and social norms as by supply and demand. What we are observing may have been a shift in social norms regarding acceptable pay differentials. Such shifts in norms may have been encouraged by government policy, particularly with regard to trade unions.

Even if we accept the economists' interpretation in terms of the supply and demand for marketable skills, we have to ask what lies behind the acquisition - or lack - of these skills. Why are people not investing in acquiring "human capital"?

One reason that has received attention from economic theorists is the limited capacity of young people to borrow to finance investment in skills, especially where they come from low income families. In this way, disadvantage may be transmitted from generation to generation. This worries not just economic theorists. The Organisation for Economic Cooperation and Development emphasised how, in the UK, "economic inequality in general hampers education and training reform I high income inequality can act to constrain pupil achievements in the lower tail of the distribution I continued increases in educational participation and attainment may be difficult to achieve so long as inequality is high and/or rising".

The lesson for the Government and the Opposition is that it is not only training programmes that are needed but also redistribution of family incomes.

Moreover, it is not just a matter of pay. Unemployment, sickness, and the withdrawal from the labour force of discouraged workers, all play a part in explaining increased inequality. There has been over the past two decades a substantial fall in the proportion of families with incomes from work. Overall, the proportion living in families supported by a person in work has fallen from four in five in 1975 to two in three in 1993. The impact on family incomes of this decline could have been offset by the welfare state, but here we come to a second important part of the puzzle. Progressive income tax and social security benefits are designed to moderate the impact of rising inequality of market incomes. If some people are earning more, while others are earning less or becoming unemployed or early retired, then the tax and benefit system should cause the differences to be less marked after tax and benefits. But inequality in income after tax and benefits actually increased more than inequality in market income between 1984 and 1989.

Government policy changes have contributed to the rise in income inequality. The results of a study by Gerry Redmond and Holly Sutherland of the Microsimulation Unit at Cambridge show that the difference between the present-day and the 1978-79 tax and benefit system, indexed in line with per capita GDP, could account for about half of the recorded increase in inequality. There have been deliberate policy choices which have weakened the safety net. Unemployment insurance has been decimated since 1979; the new Jobseeker's Allowance provides much less satisfactory coverage. In the case of retirement pensions, the Government has, as a matter of deliberate policy, allowed the value of the basic state pension to fall as a percentage of net average earnings: since 1982, it has fallen from around 32 per cent to around 22 per cent. To quote Michael Portillo, the basic pension "is going to be worth a nugatory amount in the coming century".

To explain the distribution of income, we have therefore to explain the behaviour of the Government. Given the importance of state transfers in the distribution of personal income, we need to look, as some economists have been doing, at the theory of public choice, or "political economy". It is not just market income which should be the focus of attention. Since I began here by criticising the economics profession for what I feel to have been its long-standing neglect of a central subject, I would like to end on an upbeat note.

The first ground for optimism is that there has been - very recently - an upsurge of interest. There are welcome signs that income distribution is beginning to receive again the attention which it merits. The second is that there is evidence that economics is beginning to learn in this area from other disciplines. I have touched on social norms, where we can learn from sociology and social psychology; I have referred to public choice, which abuts on political science. A subject so central to social science as income distribution is unlikely to be one that economists can solve on their own, and I take a receptiveness to outside ideas to be a sign of a discipline in good health.

Tony Atkinson is warden of Nuffield College, Oxford. His latest book, Incomes and the Welfare State, has just been published by Cambridge University Press. He is president of the Royal Economic Society.