The big book about inequality around the world is Thomas Piketty’s classic, Capital in the Twenty-First Century. Piketty worked mainly with tax data from various countries, and his sample was restricted to those developed countries that had good records extending back into history. In terms of figure 7, he focused on the growth of the global elite. And to show that politics matter, he contrasted the experiences of France and the United States. None of the growth in inequality that happened in the United States since 1980 happened in France! This counterexample shows that the effects of technological change and globalization can be altered by political actions.1
This is too short a summary of the book that bought inequality into the center of political debates in the United States. It may help some economists reading this book to explain how Piketty saw the tension between economic and political fortunes and how his analysis relates to the approach used here. He summarized his main conclusion as a race between the rate of growth of the economy and the rate of interest. When the economy grows at a higher rate than the rate of interest, then inequality decreases. When the interest rate exceeds the growth rate, inequality grows. French policies opposed this economic pressure; American policies encouraged it.
The logic behind this conclusion is that the rich save much of their incomes. As Keynes said long ago, consumption rises with income, but not as fast. When the interest rate exceeds the growth rate, these savings increase the capital of rich people faster than income grows. And, of course, a large capital stock protects rich people from the risks of life described in chapter 8 and allows them to live well whether times are good or bad. Piketty went further than other writers on inequality to stress the centrality of the distribution of capital ownership in an economy’s inequality, stating this importance in the title of his book.
Piketty’s insight helps clarify policy choices in the United States today. As noted earlier, the richest people do not want their taxes to rise. If that objection could be overcome, it is necessary to differentiate taxes on income from taxes on capital. Increasing taxes on America’s highest incomes will help fund some of the social programs that a democracy expects. However, it will not affect the capital stock of rich people, enabling them to continue their political programs. Only taxes on capital can result in durable changes in the distribution of income. Piketty supported capital taxation at very low rates as a result of these considerations and to make tax avoidance harder.
Older people today recall the years between the end of the Second World War and the start of income inequality in the 1970s and 1980s. It was a period of rapid economic growth, partly in recovery from the world war, and with a growing middle class. People living then thought of these years as normal, but Piketty asserted that these years were highly unusual in the history of the past two centuries—and even beyond. While there is no inevitable winner in the race between the interest rate and the growth rate, the interest rate normally wins.
I used the Lewis model and Piketty’s data as frameworks for the analysis of income inequality in this book. Piketty appealed to “fundamental laws of economics” to structure the data in his book, and I compare the models of inequality used by Piketty and in this book to clarify my arguments.
Piketty’s fundamental laws come from the model of economic growth created by Robert M. Solow about the same time as W. Arthur Lewis published his model. The Solow growth model is even more well known than the Lewis dual-economy model, and Solow—like Lewis—won a Nobel Prize for his model. It is a curious coincidence that both Piketty and I reached back half a century to get frameworks for our contributions to the study of inequality.2
Piketty’s first law connects wealth and income through the interest rate. His second law shows how the ratio of wealth to national income changes over time—rising with the savings rate and falling with the economy’s growth rate. This is a restatement of the equilibrium in the Solow model, reversed to make the main focus the ratio of capita to output instead of the growth rate of income. A third relation, not quite a law, strengthens the second law by looking at the return on savings. The ratio of wealth to income rises when the rate of interest, defined in the first law, is greater than the economy’s rate of growth.3
I used the Lewis model to frame my account of class and race in America. I emphasized Lewis’s insight that a two-sector model was the key to understanding how far inequality has progressed in the United States. I relied heavily on the part of the Lewis model showing that the upper sector will aim to keep incomes in the lower sector low. This is a darker aspect of the Lewis model than many economists remember, and it is integral to the way the model works.
The two models were formulated more than a half century ago. Lewis and Solow wrote soon after the Second World War ended when the progress of nations seemed promising. Piketty and I wrote as the growth of high incomes is threatening the stability of political structures in advanced economies. These are the countries invested with great hopes in the period when Lewis and Solow were writing.
One way to see how the world has changed is to look carefully at Piketty’s use of Solow’s model. Solow considered growth as his main interest; the primary economic issue in the 1950s was economic growth. He found that the ratio of capital to labor played a crucial role in the determination of the growth rate. Piketty reversed this to make the ratio of capital to income his focus. (Note the different denominators of the two ratios.) The formal treatment of capital by Solow and Piketty was similar, but the implications were far different. While Solow was determining the rate of aggregate economic growth, Piketty was chronicling the progress of the economically entitled.
This book used the Lewis model of developing countries to describe conditions in the most successful industrial country—the United States of America. This seems paradoxical as the Lewis model was designed for developing economies, not for leading ones. But as I described earlier, the United States has undergone deindustrialization that has made the economy more like that of developing countries than industrial ones. American roads and bridges are more like those in developing countries than those in Western Europe. This is not the progress implied by Lewis and Solow.
Another difference is in the treatment of economic growth. Piketty emphasized the critical role of savings as a contributor to economic growth. The growth of wealth, defined by Piketty as equivalent to capital, only happens if people save from their income. Lewis argued that the capitalist sector grows from retained earnings. He assumed that members of the capitalist sector reinvest their retained earnings. In other words, both models rely on savings, but the determinants of investment are quite different. Lewis and Solow were working within a Keynesian framework in which capital referred to the means of production: factories and machines are the prime examples.
Simon Kuznets, a third Nobel Laureate in economics, also was focused on economic growth in the 1950s. Using the data available to him, he formulated what came to be called the Kuznets Curve that asserted that income inequality would first rise and then fall during economic growth. He was reacting to the declining income inequality he observed around him and a political-economic view that richer countries would choose policies that increased equality.
Piketty argued that Kuznets was living in a very unusual economic period during the years after the Second World War. Only in that period, Piketty observed, was the growth rate of income higher than the interest rate, promoting equality. Since then, the American economy reverted to its more usual pattern where the interest rate is larger than the growth rate. I argued on different grounds that the “golden age of economic growth” was unusual, that it was a protracted recovery from the preceding thirty years of war and depression. Either way, the rapid growth of income inequality in the United States since the 1970s contradicted Kuznets’ optimistic view.4
Piketty and I defined capital in very different ways. Piketty equated capital and wealth, including the value of both public and private financial assets in his definition of capital. He was interested in gathering data on income inequality over several centuries, and he restricted his data on wealth to assets traded on markets. He then could sum varied forms of capital at market prices to get national capital stocks by adding their prices.
Piketty drew on a literature that argued for the role of finance in economic growth. Economists collected data on the growth of financial intermediation and showed that while finance grew as a result of economic growth, there was good evidence that financial development caused economic growth. Piketty did not discuss this research directly, focusing instead on more concrete issues such as the imputed rent from owner-occupied houses.5
This procedure raises an important question of national income analysis. We do not have a good way of measuring the productivity of modern finance, even though all economists agree that financial activity is important for economic growth. Does that mean that the growth of financial assets in the housing boom before the financial crisis of 2008 was all productive? This question is important, but it is too complex to be pursued further here.6
The literature on the role of capital in economic growth and development went in other directions after Lewis and Solow published their models. Solow discovered a startling fact in the empirical work that accompanied the publication of his model. The growth of capital as defined by Lewis, while important for the dynamics of the Solow model, accounted for only a tiny part of the economic growth of the United States in the twentieth century.7
Economists initially reacted to this finding by denying its importance, saying the unaccounted part of growth was simply a residual. But labor economists at the same time introduced a new kind of capital—human capital. This kind of capital was embodied in people; it was increased by formal education and informal training. Jacob Mincer used the term “human capital” in the title of an article in the late 1950s, and Gary Becker published a book by that title in the 1960s.8
Education was hardly new in the 1950s, and many authors had noted over the years how useful it was in economic affairs. The new element lay in the formalization of its effect and insertion into economic models. The contribution of education to growth then could be compared with the contribution of other kinds of capital to provide a more detailed and complete analysis of economic growth and account for much of the economic growth not explained by the accumulation of capital in Solow’s model.9
A third kind of capital came into general use several decades later. It was popularized by Robert Putnam, who wrote first about the differences between Northern and Southern Italy and then about the United States. He emphasized the role of social capital, defined to be the networks of relationships among people who live and work in a particular society that enable their society to function effectively. As with human capital, this was a formalization of a concept that was widely noted in previous work.10
A generation after Solow found that physical capital alone could not explain the progress of the United States in the twentieth century, economists found that they could explain the differences of output per worker in different countries far more completely than Solow if they used all three kinds of capital—physical, human, and social. They altered Solow’s question in two ways: they expanded the kinds of capital they used, and they compared different countries at the same time instead of one country at different times. This concept has penetrated economics to the extent that when Partha Dasgupta wrote a very short introduction to economics, he wrote it entirely about the difference between social capital in developed and in undeveloped countries. Economics in this view was a tool to answer this critical question: “Under what circumstances would the parties who have reached agreement trust one another to keep their word?” Social capital has become as important as physical capital.11
Piketty gathered prodigious amounts of data for capital that could be “evaluated in terms of market value: for example, in the case of a corporation that issues stock, the value depends on the share price.”12 Since human and social capital are not traded on public markets, they could not be measured this way and were relegated to the sidelines.13
I incorporated these two new kinds of capital into my modern adaptation of the Lewis model. I compared two sectors of a single dual economy, just as others compared different countries, with the illumination provided by all three kinds of capital. Human and social capital differ sharply between the two sectors of the American dual economy, with important effects on economics and politics. Despite this different approach, Piketty seems sympathetic to the analysis presented here. Early in his massive compilation of data, he wrote, “The resurgence of inequality after 1980 is due largely to the political shifts of the past several decades, especially in regard to taxation and finance.”14