Is Rising Income Inequality Inevitable?
by Anthony Atkinson
Income inequality is rising in a large number of industrialised countries. The phenomenon first attracted attention in the United States, where inequality began to rise in the 1970s, but the increase in the United Kingdom is even greater. The figure shows data on the Gini coefficient - the most widely used measure of income inequality. This varies between 0 when everyone has the same income to 100 per cent when one person has all the income. In the UK case, the Gini coefficient rose by some 10 percentage points from around 23 per cent in 1977 to 33 per cent in 1990 (concepts and definitions are discussed in more detail in the full text of the lecture available at www.wider.unu.edu).
However, the experience is not uniform across the OECD - the extent and timing of the increase has varied across countries. Income inequality has increased sharply in New Zealand, and has risen in Germany and the Netherlands, but France shows little upward trend in the inequality of disposable income. This variety of country experiences suggests that policy matters.
It is best to describe changes in inequality in terms of ‘episodes’ when inequality rose or fell, rather to talk of ‘trends’. Thus in the UK case, there has not been a continuous upward trend. Rather, the UK experienced an episode of rising inequality from 1977 to 1989, but from 1990 onwards inequality showed little trend (see figure). The 1990s do not look like the 1980s in the US and the UK. Therefore, to describe recent experience as an inexorable trend is potentially misleading - in particular it leads too readily to explanations which focus on technology and trade, and too little on policy.
The Transatlantic Consensus
There is a widely held belief that rising income inequality is inevitable; technological change and the globalization of world trade are the most frequently cited culprits. This view is common on both sides of the Atlantic, particularly in key policy-making institutions. The ‘Transatlantic Consensus’ holds that increased inequality in the United States and high unemployment in Continental Europe are due to a shift in demand from unskilled workers to skilled workers, whether due to technology, trade or both. For some economists rising inequality is driven by the revolution in information technology. For others, trade - especially imports from low-wage developing countries - weaken the demand for unskilled labour in the developed world.
However, the Transatlantic Consensus is open to question. The rise in inequality is neither universal nor of the same extent across countries. Where there has been a rise, it has not happened at the same time. Moreover, in some countries the rise in inequality takes the form of steps - increases followed by plateaux - rather than a continuing trend. This indicates that the world is working in more complex ways than those described in simple technological and trade explanations.
Indeed the models themselves are too simple. For instance, separating workers into just two categories - skilled and unskilled - reflects the industrial economy of the nineteenth century, rather than that of the century now ending. Rather, there is a continuum of earnings reflecting a continuum of skills. Moreover, the models of the Transatlantic Consensus see wage-differentials as nothing more than the outcome of supply and demand - thereby ignoring the role of conventions and social norms.
More insightful economic theories draw upon the work of sociologists to argue that supply and demand only place limits on possible wage differentials, while social forces determine where wages actually lie between those limits. These models also indicate that support within society can at times shift away from pay norms - for example the adoption of performance-related pay in the public sector can affect private-sector wage setting - leading to widening wage differentials.
Redistribution Can Offset Market Inequality
What people are paid in the market place is not the only factor determining the distribution of income. Market outcomes are significantly modified by income taxation and by social transfers financed out of the government budget.
To see this, let us suppose for the moment that the Transatlantic Consensus is right about the cause of rising inequality. If demand has shifted away from unskilled workers in economies with very flexible labour, then progressive income taxation should moderate the impact on income inequality - the Gini coefficient of disposable income would rise less than that of market incomes. Outcomes would differ across countries depending on the degree of progression of national income-tax systems. Outcomes will also vary depending on national variations in the importance of non-wage income sources in total household income, and the generosity (or not) of unemployment benefits.
How far has fiscal redistribution offset any rise in inequality in market incomes? Canada provides an illuminating contrast with the UK. In Canada, the Gini coefficient for market income rose by some 5 percentage points over 1980 to 1994, whereas that for disposable income did not show a significant increase. In the UK, however, the income taxation system became less progressive after 1984 and benefit levels and coverage were reduced. Thus inequality in disposable income rose much more strongly than that in market income.
Such differences in outcomes suggest that national governments do have room to manoeuvre. While the pressure of international fiscal competition is often cited as limiting the scope for income redistribution by government, this is far from selfevident, as I argue in the 1999 UNU/ WIDER Annual Lecture. Workers who perceive that taxes are lower in other states may react, not by migrating, but by demanding (through the political system) lower taxes at home. This, rather than trade or technology, may be the most important restriction on the freedom of national governments to carry out social protection. Changing social norms regarding pay - a potential cause of rising inequality in market incomes - may also lead societies to shift away from fiscal redistribution. Tax and benefit policy thus reinforces, rather than moderates, rising market inequality.
Rising Inequality is Not Inevitable
The Transatlantic Consensus sees rising inequality as the product of exogenous, inevitable events with an emphasis on technology and trade - including trade with the developing world. There is nothing to be done about rising market inequality. It is pure supply and demand.
But in this Lecture I have described an alternative story which sees the rise in inequality in market incomes as the product of changing social norms resulting in a shift away from a redistributive pay norm to one where market forces dominate. Social conventions in the labour market have changed within countries, and these have spilled out into other spheres, most notably in national tax and benefit systems. But the fact that the driving force is social in origin - rather than trade or technology - means that there is more scope for political leadership. Social norms can be influenced by policy decisions, and thus rising inequality is not inevitable.
Editors Note: This article is a summary of the main arguments of the 1999 WIDER Annual Lecture delivered on 1 November 1999 in Oslo by Professor Atkinson. The full text of the lecture is available on UNU/WIDER’s web site (www.wider.unu.edu). The lecture is also available as a UNU/ WIDER publication (see page 15 for details of how to order our publications).