A Reflection on Capital in the 21st Century

Philip Arestis and Malcolm Sawyer

The recent book Capital in the 21st Century by Thomas Piketty (2014) has attracted an enviable amount of attention with its detailed history of income and wealth inequality. A central idea in this book comes from (r > g); that is, the idea that the rate of return on wealth (r) exceeds by a considerable margin the rate of economic growth (g) so far as Western industrialised countries are concerned. This, Piketty argues, has implications for rising inequality especially of wealth and of rentier income.

The basic mechanism is that when the rate of return is greater than the income growth rate, then savings made out of the return on wealth received in the form of dividends, rents, interest and capital gains adds to wealth, which then rises at a rate determined by these savings (and equal to savings propensity out of wealth (s), multiplied by the rate of return (r), (i.e., s∙r), and then the stock of wealth rises faster than income. The wealth to income ratio then rises, and the gains from wealth rise faster than other incomes. Wealth inequality also rises on the basis that rich wealth owners can achieve higher returns on their wealth than the poorer wealth owners. This is a substantial, though not complete, part of the general rises in inequality over the past three or more decades (and a large part of his book is concerned with documenting those rises in inequality).

It is interesting to compare Piketty’s analysis with others who have considered the relationship between the rate of return on wealth and the rate of growth (which are largely ignored by Piketty). Before undertaking this comparison it is worth noting, as King (2014) suggests, that the relationship in question is an accounting identity and the question is should one treat it as a “fundamental law of capitalism”? A notable example of another view of the relationship between rate of return and rate of growth comes from what was called the Cambridge equation based on the equality between savings and investment (applied to a closed economy without a government). Savings are largely based on profits, and investment is closely linked with the growth of output with investment enabling the capital stock to grow in line with output. Then savings equal to investment yields the equality s∙r = g. In contrast, Piketty’s approach would argue that s∙r > g. In other words, Piketty is indicating that the tendency to save exceeds the tendency to invest (on the basis that investment will be undertaken to keep the capital stock growing in line with output).

There are two problems with Piketty’s approach. The first is that his focus is on savings without paying any regard to the other side of the equation, namely investment. He views wealth to income ratio to be rising; but that also means the capital to income ratio rising. What is the consequence of that? From a neo-classical perspective, the marginal product of capital, and thereby the rate of profit would decline (there are well-known objections to that view from the work of Sraffa and others). From a heterodox perspective, if savings intentions exceed investment intentions, deflation results with high levels of unemployment. If the capital-output ratio rises, the rate of profit will decline unless there is an increase in the share of profits, and as the capital-output ratio continues to rise, increasing share of profits emerges.

The second is that the intention to save out of wealth exceeds investment intentions: savings to take place has to be matched by investment. The excess of savings over investment provides a deflationary situation, which can be met by the government running a budget deficit or which results in high levels of unemployment.

Thus there must be doubts as to whether the scenario portrayed by Piketty could continue with r > g. Something would have to give. It may be “rioting on the streets” as unemployment rockets; or it may be that the rate of profit declines, though budget deficits could play a role of limiting those effects. Consequently, the conclusions of the book need a great deal more work to be substantiated. A further example, as King (2014) comments, is that the current problems most economies are faced with are different. Namely, that the problem is not r > g but that the rate of interest is below the growth rate, and this is particularly so with regard to the rate of interest on government bonds. This does not necessarily lead to the same conclusions in terms of distribution as Piketty suggests. Another issue which has been pointed to is the use by Piketty of a single average rate of return, whereas there are important differences between, for example, the rate of return on government bonds and the rate of return on equity with regard to their level, their implications for distribution along with the risks and uncertainties involved.

There would be many reasons to think that growth rates in industrialised countries will continue to slow as they have tended to do in the past few decades (as Piketty, 2014, p. 97) suggests) since the end of the “Golden Age”. The causes of this slower growth can be variously ascribed to the end of technological catch-up (by industrialised countries with the United States), the impacts of financialisation on investment, the failures of neo-liberalism to maintain growth through increasing inequality; to these Piketty adds slower population growth (and sometimes decline). The ecological limits on growth, though strangely absent from his discussion, reinforce the view that growth will be lower than during the “golden age” of capitalism. But whatever the causal factors behind slowing growth, it would seem likely that future growth rates in industrialised countries may well be of the order suggested by Piketty or slower.

It is then imperative to consider the macro-economic and distributional consequences of that slower growth. Piketty’s preferred solution is that of a global wealth tax, which would need to be substantial—in effect sufficient to reduce savings out of rentier income to around 1 to 1 ½ per cent of wealth. To boost the level of demand such a wealth tax would need to be spent, and public expenditure increased or other taxes decreased. There is much to be said for such a wealth tax, which is in no way to underestimate the practical and political difficulties of securing such a tax. However, we would argue that the implications of the r > g assumption are more widespread and serious than Piketty envisages, and that there can be other ways of addressing the issues. Taxation on corporate profits can be raised; the power of workers enhanced to shift the distribution of income from profits to wages.

Piketty’s book has documented the widening inequality, which scars industrialised economies, and generally draws the implication that rising inequality has not been justified by enhanced economic performance. He has identified the relationship between the rate of return on wealth and the rate of growth as a major issue. In our view, an excess of savings out of return on wealth and the rate of growth is unsustainable. It may lead, following Piketty, to rising wealth inequality. But we have argued that the difference would be deflationary and cause high levels of unemployment. There is a need to reduce the effective rate of profit—and this can be attempted through enhanced worker power or a corporation tax (on a co-ordinated basis to reduce tax competition).


Piketty, T. (2014), Capital in the 21st Century, (translated by Arthur Goldhammer), Harvard University Press.

King, M. (2014), “Capital in the Twenty-First Century by Thomas Piketty, Review”. Available at: http://www.telegraph.co.uk/culture/books/bookreviews/10816161/Capital-in-the-Twenty-First-Century-by-Thomas-Piketty-review.html

Triple Crisis welcomes your comments. Please share your thoughts below.