The Piketty balance sheet

Piketty’s much discussed book, while imperfect, is a timely and valuable contribution to the current debate on inequality says Robert Rowthorn.

Thomas Piketty’s Capital in the Twenty-First Century has been a stunning success. It comes at a time of rising concern about poverty and inequality. This concern has been given added impetus by the 2008 financial crisis and ensuing recession, but it predates the crisis. For example, the IMF has produced a stream of documents over the past decade about globalisation, economic growth and inequality.

Piketty’s contribution is two-fold. Using data compiled by a number of scholars, including himself, he documents with elegance long-run trends in the distribution of income and wealth, going back in some cases to the 18th century. His argument is illustrated by means of numerous well-chosen graphs, the underlying data for which and supplementary graphs are available online. Piketty also provides a simple theoretical framework for analysing the dynamics of wealth and income distribution, and thinking about future trends.

In its 685 pages, Piketty’s book covers a wide range of topics pertaining to income and wealth, but his main thesis is simple. The long march towards equality during the mid-20th century has gone into reverse and, unless something is done about it, things will get worse in the future.

In the 19th and early 20th centuries wealth and income were highly concentrated in what are now the advanced capitalist economies. There was a patrimonial society in which a minority lived off its wealth while the rest of the population worked for a living. Between 1914 and 1945 the wealth and incomes of the rich were squeezed by a combination of war, economic disruption and high taxes. Their shares of national wealth and income then stabilised for a time and since around 1980 have been increasing. Looking forward, Piketty argues that we are witnessing the emergence of a new form of patrimonial society in which as in the past there will be a rich elite that owns a large fraction of national wealth and enjoys the income thereby generated.

Amongst the rich, there will also be super-managers and the like who receive gigantic salaries or bonuses. In addition, there will be a substantial patrimonial middle class that lives a moderately or very comfortable life combining earned income with inherited wealth. A large number of people at the bottom will own very little. Indeed, his data suggest this is already the picture in some countries.

Most reviewers have praised Piketty for the quality of his data even when they do not share his analysis or concern. One exception is Chris Giles of the Financial Times, who has challenged head-on Piketty’s claim that the distribution of wealth is becoming more unequal. Giles accepts Piketty’s estimates for France and Sweden but argues that he is wrong about the rising concentration of wealth ownership in the UK and the USA. Using other data sources, Giles argues that the share of top-wealth owners in national wealth has been roughly stable in the latter two countries.

Piketty has replied to this criticism. He does not address Giles’ claim about the UK, but in the case of the USA he cites recent work by Saez and Zucman. These authors use a novel way of estimating the distribution of wealth. Most estimates of wealth distribution are derived from asset surveys or information on the estates of deceased individuals. These are not a reliable way of estimating the distribution of wealth at the top end of the scale.

Saez and Zucman use instead an alternative, and arguably more reliable, method based on income tax returns. They find that in the USA, there has been a huge rise in the share of the top 0.1 per cent in national wealth from around 8 per cent in the late 1970s to around 22 per cent today. The increase has been mainly driven by the top 0.01 per cent whose share has risen from less than 3 per cent to around 11 per cent. Most other groups have seen their share of national wealth decline. Saez and Zucman have not applied their method to the UK, but if they did so, it might show a similar picture here, though one cannot be sure.

The People at the Bottom

The claim that wealth distribution has become more unequal is significant partly in its own right and partly because past trends may be a guide to the future. However, even if the distribution of wealth remains stable in the future, this is no reason for complacency. In the UK around 70 per cent of adults own their own homes, either outright or with a mortgage, but there are many millions who do not. Some of the latter will buy a home in the future, but for many this will be impossible because house prices are now so high, especially in London and parts of the South East.

The greatest inequality in wealth ownership is observed in the case of financial assets. According to the ONS Wealth and Assets Survey, 25 per cent of households in 2010-2012 had negative net financial assets. Their debts exceeded their assets. At the other end of the scale, 12 per cent had net financial assets of more than £100,000. These figures exclude mortgages and private pension entitlements. The distribution of pension wealth was equally skewed. Around 24 per cent of households had no private pension wealth at all.

Some of these contrasts reflect life-cycle effects. Younger people go into debt when they take out a loan to buy a house or to finance a university degree. Most of them will eventually repay their loan and become net wealth owners. However, there is a large underclass, many of them unemployed, economically inactive and/or lone parents, who will never reach this stage. There are also the estimated 5 million working poor who earn less than the ‘living wage’. Many of them will never buy a house or build up a significant pension pot for their old age. They will leave little or nothing for their children, who will have to fend for themselves, unlike the majority of children who belong to the large ‘patrimonial’ class who can count on some form of help from their parents in the form of gifts or inheritance. The most important inequality both now and in the future may not be between the super-rich and the rest, but between the bottom one third and the fortunate majority who are able to accumulate or inherit at least a modest amount of wealth.

The Wealth-income Ratio

Piketty’s theoretical analysis relies heavily on the ‘wealth-income’ ratio, which is denoted by the symbol ‘β’. Wealth consists of all assets (stocks and shares, housing etc.) net of debt; income refers to total national income. All items are valued at market prices. This ratio was around 600-700 per cent in the 19th century, and then fell sharply in the mid-20th century to around 200-300 per cent in 1980. It has been rising since then and is now around 500-600 per cent.

Movements in this ratio are explained by the relationship between the real rate of return on capital (net of tax and depreciation, adjusted for inflation), which is denoted by ‘r’, and the growth rate of national income, denoted by ‘g’. During the 19th century r was much greater than g so that wealth-owners could have the best of both worlds, saving part of their investment income so as to build up their wealth, whilst at the same time enjoying a high level of consumption. In the mid-20th century, r fell sharply because of taxation and g increased with the result that r < g. Even if wealth owners had saved all of their investment income during this period, their wealth would have increased more slowly than the rapidly growing national income. More recently, taxes on capital have been cut causing r to increase, with the result that once again r > g. Once again, the richer wealth-owners can add to their wealth by saving part of their investment income and at the same time enjoy a high level of consumption.

Looking to the future, Piketty argues that global economic growth will slow down as population growth comes to an end and productivity increases more slowly. His projection for the period 2050-2100 is that r = 4.3% p.a. and g = 1.5% p.a. A similar combination was observed in the late 19th century. Its implications for the wealth-income ratio β depend on what proportion of income is saved. Piketty assumes that this proportion will be sufficiently high to ensure that total wealth grows faster than national income. He expects β to return to its 19th century level or even higher.

This projection is, of course, highly speculative. It depends on assumptions about future economic growth, the return on capital and savings behaviour that may turn out to be false. Even so, Piketty’s use of the r–g relationship as a way of thinking about the future points us in the right direction. If we are now in an era of slow economic growth and higher after-tax returns on investment, as Piketty believes, then it seems likely that the wealth-income ratio will continue rising.

The importance of Inheritance

One implication of a rising wealth-income ratio is the growing importance of inherited wealth, much of it in the form of housing. Piketty calculates that in France the ratio of gifts and bequests to national income has risen from 6 per cent in 1960 to 11 per cent in 2000. He predicts that this ratio will eventually stabilise at between 16 per cent and 23 per cent of national income. The trend is similar in other countries. If inheritances were uniformly spread throughout the population such a trend would not matter. Indeed, a larger role for inheritance might be a good thing because it would strengthen inter-generational ties. However, inheritances are not evenly spread and many people can expect to receive little or nothing from their parents. The greater the role of inheritance as a source of wealth, the more difficult will it be for such people to escape their situation. Piketty’s discussion of inheritance in advanced economies is one of the most interesting and most disturbing parts of his book. NB the US share does not include African Americans

Capital income absorbs between 15% and 25% of national income in rich countries in 1970, and between 25% and 30% in 2000 – 2010. Source: Piketty, Capital in the Twenty-First Century, Figure 6.5

The Capital–Labour Split

Piketty documents how the share of capital in national income has increased in rich countries over the past thirty years or so (see figure 1). The share of labour has decreased by an equivalent amount. Piketty argues that this development is a consequence of the rising wealth–income ratio. His explanation relies on what is known to economists as the ‘neoclassical theory of factor shares’. This theory postulates a link between the share of profits in private industry and the ‘capital–output ratio’, conventionally written K/Y. This ratio is the amount of physical capital (buildings and equipment) used in production divided by annual output. Depending on the prevailing technology, a given change in K/Y will be associated with a determinate increase or decrease in the share of profits in output. When applying this theory to modern experience, Piketty assumes that the capital–output ratio and the wealth–income ratio are equal, i.e. K/Y = β. Having shown that β has increased by a considerable amount, he concludes that K/Y has increased by the same amount. This supposedly explains why the share of capital in national income has risen.

Piketty’s argument can be criticised on a number of grounds. The neoclassical theory of distribution, on which it is based, is open to question. Moreover, even if we accept this theory, Piketty’s wealth–income ratio is not a good proxy for K/Y. The evidence indicates that the wealth–income ratio β has risen a great deal whereas, properly measured, the capital-output ratio K/Y has been either constant or fallen in most countries. The divergent behaviour of β and K/Y is explained by the behaviour of house prices and share prices.

The term ‘capital’ in the neoclassical theory refers to the physical capital (buildings and equipment) that is used in combination with labour in the production process. It therefore excludes housing, whereas Piketty includes housing in his measure of wealth. Rising house prices explain much of the observed increase in Piketty’s wealth–income ratio. Moreover, ‘capital’ in the neoclassical theory is measured in ‘real’ terms, i.e. corrected for inflation. Piketty’s measure of industrial wealth is based on share prices. This means that a general boom in share prices (capital gains) may lead to an increase in the wealth–income ratio, even though nothing may have altered in real terms. Firms may be using the same amount of equipment and buildings, using the same number of workers and producing the same output as before. When the distortions resulting from housing and capital gains are eliminated, Piketty’s wealth–income ratio shows no increase at all. Thus, his explanation for the rising share of capital in national income falls to the ground.

The debate about what has been driving the shift in income distribution away from labour towards capital is not merely of historical interest. It may provide some clue to the future. Piketty’s argument implies that this shift has been caused by the over-accumulation of capital as indicated by an increasing wealth–income ratio. There has been too much investment. By implication if the wealth-income ratio continues rising, the distribution of income will shift further away from labour to capital. An alternative explanation, which I explore in a forthcoming issue of the Cambridge Journal of Economics, reaches just the opposite conclusion. There has been too little investment and it is this which explains the shift in income distribution away from labour. What we require in the future is more investment not less.

Inequalities of Labour Income

A striking phenomenon in recent decades, especially in Anglophone countries, has been the huge increase in pay for people at the top of the salary scale: top managers, many investment bankers, top football players and so on. The increase is even larger if the effect of lower tax rates for high earners is taken into account. At the other end of the scale, an underclass of ‘working poor’ has emerged who have seen little or no increase in their real incomes over the past forty years.

Developments at the lower end of the income scale are often explained by economists as the outcome of unequal educational performance. Some people are much better qualified than others to compete in our modern technological world. The unqualified get paid very little because they produce so little. The obvious long-term remedy is to improve the quality of education for disadvantaged children. Piketty does not deny that education and skills play a role in wage determination, but he argues that labour market institutions also matter.

One such institution that is the legal minimum wage, which has fallen by 22 per cent in the USA since 1968 but has risen by 260 per cent in France over the same period. This contrast helps to explain why there is such a large pool of working poor in the USA but not in France. Piketty concedes that beyond a certain point, an increase in the minimum wage will reduce the demand for labour and thereby harm the people it is supposed to help. France, he suggests, may have reached this point. However, citing the economists David Card and Alan Krueger, he also suggests that there is plenty of room for the minimum wage in the USA to increase without reducing the demand for labour. On the contrary, an increase in the minimum wage might actually lead to more employment.

Piketty’s discussion of top salaries is focused mainly on the emergence of what he calls ‘super-managers’, whose pay may be several hundred times that of the ordinary employee. This is often justified as a reward for ability and effort. Super-managers are supposedly paid so much because their personal contribution to the firm, or ‘marginal product’, is so large. Piketty challenges this explanation on several grounds. The marginal product of a top manager is virtually impossible to measure, so there is enormous room for subjective judgement.

Moreover, empirical evidence on executive pay indicates that the salaries of top managers are only loosely related to the relative performance of their own firm. It is when sales and profits increase for external reasons that executive pay rises most rapidly. External reasons include exchange rate variations, raw material price shocks and growth in the overall demand for the type of goods or services that the firm supplies. This has been called ‘pay for luck’. Finally, the marginal productivity theory cannot explain why top salaries have increased by much more in the Anglophone countries than in continental Europe and Japan. It is hard to believe that the productivity of top managers in the Anglophone countries has increased so much faster than the productivity of their counterparts elsewhere.

If the pay of top managers is not determined by their marginal productivity, how then is it determined? Piketty argues that it is determined by the social norms which guide remuneration committees in large companies. These norms are, in turn, influenced by the prevailing tax rates on top incomes. When top tax rates were extremely high, there was no point in awarding executives large pay increases because these would by mostly taxed away. The company would spend a lot of money on executive pay with little benefit to its top managers. In such an environment, executive pay was relatively low. However, when top tax rates were drastically cut, as they have been, executives were able to keep a much larger proportion of their salary than before. Firms were therefore able to reward their top managers generously in the knowledge that most of the money would go to the manager and not to the taxman. This new reality led top managers to bargain more vigorously for higher pay. It also led to a shift in social norms so that it became more acceptable to pay very high executive salaries.

This explanation may seem a bit far-fetched, but it has some empirical support. Using a new database on executive pay, Piketty and his colleagues Emmanual Saez and Stefanie Stantcheva investigated the determinants of executive pay in developed economies. They found that variations in the marginal tax rate can explain much better than other factors why executive pay rose so sharply in some countries than in others.

What does this mean for policy?

The closing section of the book is concerned with public policy. There are two main redistributive proposals: a big increase in the marginal rate of tax on top incomes and a global wealth tax. Elsewhere in the book Piketty also supports an increase in the minimum wage in certain countries, in particular the USA.

Piketty’s tax proposals are radical. For the USA he suggests a marginal tax rate of 80 per cent on incomes above $500,000 or $1 million per annum. This is less than the top rate in the 1950s but is twice the present rate. Piketty claims that this would lead to a reduction in pre-tax remuneration thereby releasing funds within the firm for raising pay at the lower end. To see how radical this proposal is, consider the effect of raising the marginal tax rate on salaries above $1 million dollars from the present 40 per cent to 80 per cent. Suppose also that the pre-tax salaries of top managers fall by 50 per cent in response to this tax increase, in line with Piketty’s theory. Following these changes, an executive who currently earns $10 million a year before tax would see his post-tax earnings fall from over $6 million at present to around $1.5 million.

Piketty implies that such reductions would not affect productivity or overall economic performance. Is this realistic? The answer, I suspect, depends on the cultural ambience at the time. If there were a widespread belief, shared by top executives, that huge salaries are unjustified, then a return to the 1950s might be possible. Top salaries could be reduced and top tax rates could be increased, as Piketty advocates, without too much disruption. However, as he concedes, this is unlikely to be anytime soon. Still, it may happen at some time in the future.

The policy reform that has gained the most publicity is Piketty’s proposal for a global wealth tax on large fortunes. As an example, he suggests 5 per cent or 10 per cent per annum on fortunes of over 1 billion Euros, with lower tax rates on smaller wealth holdings and zero tax for net assets below 1 million Euros. Piketty recognises that a global wealth tax is unrealistic at present, but he argues that a start could be made in larger economic blocs, such as the EU, and the tax could afterwards be extended to the rest of the world. To deter tax evasion he advocates automatic transmission of bank information so that foreign holdings or capital flight can be easily identified by the tax authorities.

These are interesting and technically feasible proposals. The confiscatory levy which Piketty proposes for large fortunes seems wildly unrealistic at present, but it is conceivable that a future, more integrated, EU may introduce a less extreme wealth tax. Piketty suggests that the proceeds of such a tax could be used to pay off some of the public debt in Europe. One can think of other uses for the money.

Piketty’s book is a timely and valuable contribution to the current debate on inequality. It provides a clear survey of long-run trends in the distribution of income and wealth. It also provides an insightful analysis of the forces responsible these trends. Some of his data are open to question and his theoretical underpinning is not always convincing. Nevertheless, Piketty’s Capital is a fine piece of work. It will have a lasting impact on how we think about inequality in modern capitalist economies.