Inequality

What Can Be Done?

Anthony B. Atkinson

• Publisher: Harvard University Press
• Year: 2015
• Print ISBN: 9780674979789, 9780674504769
• eBook ISBN: 9780674287037
• Topic: Economics
• Subtopic: Economic Theory
• Index: Economics

Winner of the Richard A. Lester Award for the Outstanding Book in Industrial Relations and Labor Economics, Princeton University

An Economist Best Economics and Business Book of the Year

A Financial Times Best Economics Book of the Year

Inequality is one of our most urgent social problems. Curbed in the decades after World War II, it has recently returned with a vengeance. We all know the scale of the problem—talk about the 99% and the 1% is entrenched in public debate—but there has been little discussion of what we can do but despair. According to the distinguished economist Anthony Atkinson, however, we can do much more than skeptics imagine.

"[Atkinson] sets forth a list of concrete, innovative, and persuasive proposals meant to show that alternatives still exist, that the battle for social progress and equality must reclaim its legitimacy, here and now… Witty, elegant, profound, this book should be read."
—Thomas Piketty, New York Review of Books

"An uncomfortable affront to our reigning triumphalists. [Atkinson's] premise is straightforward: inequality is not unavoidable, a fact of life like the weather, but the product of conscious human behavior.
—Owen Jones, The Guardian

Part One

DIAGNOSIS

Chapter 1

Setting the Scene

This book is concerned with ways of reducing the extent of inequality, and we need to be clear at the outset exactly what is, and what is not, meant by this goal. Let me begin by removing one possible misconception. I am not seeking to eliminate all differences in economic outcomes. I am not aiming for total equality. Indeed, certain differences in economic rewards may be quite justifiable. Rather, the goal is to reduce inequality below its current level, in the belief that the present level of inequality is excessive. I have stated this proposition deliberately in terms of the direction of movement, not of the ultimate destination. Readers may well disagree as to how much inequality is acceptable while agreeing that the present level is intolerable or unsustainable.

In this chapter, I explore the reasons we should be concerned about inequality and its relation with underlying social values. I then take a first look at the empirical evidence. Just how unequal are our societies? By how much has inequality increased? Once we have seen the broad patterns, however, it is necessary to probe more deeply. Just what is being included in the statistics and what is missing? Who is where in the distribution?

Inequality of Opportunity and Inequality of Outcome

On hearing the term “inequality,” many people think in terms of achieving “equality of opportunity.” This phrase occurs frequently in political speeches, party manifestos, and campaign rhetoric. It is a powerful rallying call with long roots in history. In his classic essay Equality, Richard Tawney argued that all people should be “equally enabled to make the best of such powers as they possess.” In the recent economics literature, following the work of John Roemer, the determinants of economic outcomes are separated into those due to “circumstances” that are beyond personal control, such as family background, and “effort,” for which an individual can be held responsible. Equality of opportunity is achieved when the former variables—circumstances—do not play any role in the resulting outcome. If some people work harder at school, pass their exams, and get into medical school, then at least part (but not necessarily all) of their higher salary as a doctor can be attributed to effort. If, on the other hand, their place at medical school is secured through parental influence (for example, preference being given to the children of alumni), then there is inequality of opportunity.¹

The concept of equality of opportunity is an attractive one. However, does it mean that inequality of outcome is irrelevant? In my view, the answer to this question is “no.” Inequality of outcome is still important, even for those who start from concern for a “level playing field.” To see why, we need to start by noting the difference between the two concepts. Inequality of opportunity is essentially an ex ante concept—everyone should have an equal starting point—whereas much redistributional activity is concerned with the ex post outcomes. Those who think inequality of outcome is irrelevant regard concern for ex post outcomes as illegitimate and believe that, once a level playing field for the race of life has been established, we should not enquire into the outcomes. To me this is wrong for three reasons.

First, most people would find it unacceptable to ignore completely what happens after the starting gun is fired. Individuals may exert effort but have bad luck. Suppose that some people trip and fall into poverty. In any humane society help will be provided to them. Moreover, many believe that this help should be offered without enquiring into the reasons the person fell on hard times. As the economists Ravi Kanbur and Adam Wagstaff note, it would be morally repugnant to “condition the doling out of soup on an assessment of whether it was circumstance or effort which led to the outcome of the individual . . . to be in the soup line.”² The first reason, then, that outcomes matter is that we cannot ignore those for whom the outcome is hardship—even if ex ante equality of opportunity were to exist.

But the significance of outcomes goes much deeper than this, leading to the second reason that inequality of outcome matters. We need to distinguish between competitive and noncompetitive equality of opportunity. The latter ensures that all people have an equal chance to fulfill their independent life projects. To pursue the athletic analogy, all can have the opportunity to acquire swimming certificates. In contrast, competitive equality of opportunity means only that we all have an equal chance to take part in a race—a swimming competition—where there are unequal prizes. In this, more typical case, there are ex post unequal rewards, and this is where inequality of outcome enters the picture. It is the existence of a highly unequal distribution of prizes that leads us to attach so much weight to ensuring that the race is a fair one. And the prize structure is largely socially constructed. Our economic and social arrangements determine whether the winner gets a garland or $3 million (the top prize in the U.S. Open Tennis tournament in 2014). The determination of the prize structure is the principal concern of this book.

Finally, the third reason for concern about inequality of outcome is that it directly affects equality of opportunity—for the next generation. Today’s ex-post outcomes shape tomorrow’s ex ante playing field: the beneficiaries of inequality of outcome today can transmit an unfair advantage to their children tomorrow. Concern about unequal opportunity, and about limited social mobility, has intensified as the distributions of income and wealth have become more unequal. This is because the impact of family background on outcome depends both on the strength of the relationship between background and outcome and on the extent of inequality among family backgrounds. Inequality of outcome among today’s generation is the source of the unfair advantage received by the next generation. If we are concerned about equality of opportunity tomorrow, we need to be concerned about inequality of outcome today.

Instrumental and Intrinsic Concerns for Inequality

Reducing inequality of outcome matters, therefore, even to those for whom equality of opportunity is the ultimate objective. It is a means to an end. In the same way, influential books such as The Price of Inequality by Joseph Stiglitz and The Spirit Level by Kate Pickett and Richard Wilkinson have identified other instrumental reasons we should be concerned about inequality of outcome.³ They argue that we should reduce inequality of outcome because it has bad consequences for today’s society; they blame increased inequality for lack of social cohesion, increased crime, ill-health, teenage pregnancy, obesity, and a whole range of social problems. Political scientists have identified a two-way relationship between income inequality and the role of money in determining the outcome of democratic elections, characterised by the “dance of ideology and unequal riches.”⁴ Economists have placed worsening economic performance at the door of increased inequality. In her speech to the 2012 Annual Meetings of the IMF and the World Bank, Christine Lagarde spoke of her “third milestone: inequality and the quality of growth in our future world.” She went on to say that “recent IMF research tells us that less inequality is associated with greater macroeconomic stability and more sustainable growth.” The extent of consequential benefits from reducing inequality can be much debated, and I return to the relation between inequality and economic performance in Chapter 9.

The case for reducing inequality does not, however, depend solely on its having adverse consequences of the kind described above. There are intrinsic reasons for believing that the current degree of inequality is excessive. These reasons may be framed in terms of a broader theory of justice. For economists writing on these issues a hundred years ago, it was natural to think in utilitarian terms. Summarising individual well-being in terms of the utility level attributed to each person, they argued that excessive inequality reduced the sum of total utility, since the value of an additional unit of income (or economic resources more generally) was lower for the well-off. As it was put by Hugh Dalton, British economist and postwar Labour Chancellor of the Exchequer, transferring £1 from a rich person to a less well-off person would, other things the same, reduce inequality and raise the sum of utility for society as a whole.⁵

Utilitarianism has been much criticised, not least for being concerned solely with the sum of individual utilities, and being, in the words of Amartya Sen, “supremely unconcerned with the inter-personal distribution of that sum. This should make it a particularly unsuitable approach to use for measuring or judging inequality.”⁶ It is for this reason that distributional weights are applied when measuring inequality, with more weight attached to those who are less well-placed. These distributional weights incorporate our social values regarding redistribution and provide an intrinsic basis for concern about inequality. Just what these weights should be is a matter over which people differ, as may be seen from the “leaky bucket experiment” described by the economist Arthur Okun. He asked what would happen if some of Dalton’s £1 transfer were to be lost on the way. From the answer given, Okun deduced how much more weight would have to be attached to the income of the recipient, compared with that of the donor, in order to justify the transfer. If half of the transfer leaked out of the bucket, then we would need to give twice the weight to the income of the recipient compared with that of the donor. People giving greater weight to poorer recipients would favour more redistribution; they would go further towards reducing inequality. In the limit, all the weight would be given to the least well-off, a position often associated with A Theory of Justice by John Rawls, although there is much more to his theory than is captured by this limiting case.⁷

The “Rawlsian” position of favouring the least advantaged may sound quite radical. However, it is not far removed from the statements of politicians who argue for income tax cuts on the basis that these would stimulate economic activity and hence increase revenue that could be used to raise the incomes of the poorest among us. As this argument illustrates, there is nothing intrinsically egalitarian about the Rawlsian objective. Maximising the well-being of the least advantaged may lead to a quite unequal distribution. More radical in this sense than Rawls was Plato, who expressed the view that no one should be more than four times richer than the poorest member of the society.⁸ On this egalitarian view, inequality matters on account of the distance between rich and poor, and there may be a case for action even where there is no gain to the poorest.

A Theory of Justice by Rawls initiated a wide debate among moral philosophers about the nature of social justice. Of particular relevance here is Rawls’s framing of the principles of justice in terms of access to “primary goods”: “things which it is supposed a rational man wants whatever else he wants,” listed in broad categories as “rights and opportunities and powers, income and wealth.”⁹ As Sen has argued, this takes us well beyond utilitarianism but stops short of considering the “wide variations [people] have in being able to convert primary goods into good living.”¹⁰ Sen has proposed that we should move on from primary goods to “capabilities,” defining social justice in terms of the opportunities open to people according to their functioning. The capability approach differs from Rawls’s approach in two respects. It focuses on what goods can do for people in their particular circumstances, taking into consideration, for example, that people with disabilities may have higher travel-to-work costs than able-bodied people. It is concerned not just with the achieved outcomes, but also with the range of opportunities, which Sen regards as an essential element of personal freedom (hence the title of Sen’s book, Development as Freedom).¹¹ In practical terms, the capability approach has broadened the dimensions of social and economic performance under examination, notably influencing the Human Development Index launched twenty-five years ago by Mahbub ul Haq (the index ranks countries according to their level of development, looking at education and life expectancy, as well as income).¹² In the present context, the capability approach brings us back to instrumental reasons for concern about the inequality of economic resources, but now within a coherent set of principles of justice.¹³ Within such a framework, income is only one dimension, and differences in income should be interpreted in the light of differing circumstances and of the underlying opportunities. But it remains the case that achieved economic resources are a major source of injustice. That is my reason for concentrating here on the economic dimension of inequality.

But what do economists have to say about inequality?

Economists and Income Inequality

Some two decades ago, I gave my presidential address to the Royal Economic Society titled “Bringing Income Distribution in from the Cold.”¹⁴ The title was chosen to underscore the way the subject of income inequality had become marginalised in economics. For much of the twentieth century the topic had been ignored, whereas I believed that it should be central to the study of economics. I started that address by quoting the same concern expressed earlier in the century by Dalton, who said that as a student he had been especially interested in the distribution of income: “I gradually noticed, however, that most ‘theories of distribution’ were almost wholly concerned with distribution as between ‘factors of production.’” He went on to say that “distribution as between persons, a problem of more direct and obvious interest, was either left out of textbooks altogether, or treated so briefly, as to suggest that it raised no question, which could not be answered either by generalizations about the factors of production, or by plodding statistical investigations, which professors of economic theory were content to leave to lesser men.”¹⁵ The same remained true when I reviewed the economics literature in the 1990s. In his account of the history of economic thought on income distribution, Agnar Sandmo observes that “the connection between resource allocation and...

Table of contents

• Cover
• Title Page
• Copyright
• Dedication
• Contents
• Acknowledgements
• Introduction
• Part One: Diagnosis
• Part Two: Proposals for Action
• Part Three: Can It Be Done?
• Glossary
• Notes
• Contents in Detail
• List of Tables and Figures
• Figure Sources
• Index