Wednesday, July 16, 2014

The Evolution of Inequality (3): Piketty's Historical Data

The debate over the history of economic inequality has been one of the fundamental disputes in political philosophy and the social sciences.  Until recently, Thomas Piketty complains, this has been a debate without data. 

In his Discourse on the Origin of Inequality among Men, Jean-Jacques Rousseau thought that he had explained the evolutionary origins of war, property, and inequality.  While he relied on the best anthropological evidence available to him, he looked forward to the time when scientists would take long voyages around the world to collect evidence for human evolution that would allow them to write "the natural, moral, and political history" of humanity (1964: 213).  Over the past two centuries, that work has been done by biologists, anthropologists, and archaeologists, so that now we can write the evolutionary history of humanity foreseen by Rousseau, which allows us to judge whether Rousseau's account of the origin of inequality is true or not.  In some previous posts, I have argued that the evidence supports Locke's evolutionary history as more accurate than Rousseau's.

Economists like Piketty are less interested in the ancient evolution of inequality than in the more recent evolutionary history over the past 250 years, because they want to understand whether modern capitalism inevitably concentrates wealth into ever fewer hands.  And they would like to see quantitative data that can be analyzed systematically to produce a statistical history of inequality.  The most impressive feature of Piketty's book is that he does that. 

Piketty's data analysis is available online in his "World Top Incomes Database" and in the website for his book, which includes all of the tables and figures from the book.  Piketty's visual presentation of his argument in these tables and figures is remarkably engaging.

And yet there has been a lot of criticism of his data analysis.  One can see that, for example, in the webcast of the panel on Piketty's book at the Tax Policy Center on April 15, which included Piketty and one of his leading critics--Kevin Hassett of the American Enterprise Institute.

This debate over Piketty's book continues a debate that began with Karl Marx, who predicted that capitalism would inevitably concentrate wealth into ever fewer hands, so that a few capitalists would own almost all the wealth, and their workers would sink into miserable poverty as their wages were reduced.  In doing that, however, capitalism would produce the conditions for its revolutionary overthrow, because the workers would be provoked into revolutionary violence, and the communists could lead them into abolishing private property and then establishing a classless socialist society in which all people could cooperate without a ruling class.

Marx's prediction that workers would become ever more impoverished turned out not to be true.  During the last decades of the nineteenth century, the purchasing power of workers' wages began to increase, and thus many workers were happy enough to be turned away from revolutionary activity.

But Marx's prediction about the growing inequality of wealth seemed to be true.  By the end of the nineteenth century and the beginning of the twentieth century, the top 5-10% of the richest people in the richest countries owned most of the wealth, and the bottom 50% of the people owned almost nothing.  Thus, capitalism seemed to promote an unjust inequality of wealth.

By the 1950s, however, some economists thought they saw evidence that inequality was declining.  In his presidential address to the American Economic Association in 1954, Simon Kuznets surveyed income tax data from the United States, England, and Germany that showed that economic inequality had declined since World War I ("Economic Growth and Income Inequality," The American Economic Review 45 [March, 1955]: 1-28).  Kuznets claimed that while inequality increased in England in the first half of the nineteenth century, Marx mistakenly assumed that this would continue into the future.  Kuznets saw a long historical swing in income inequality: early in the industrial revolution, inequality grew; but later, as more people benefited from economic growth, this inequality would decline.  For many economists, this became part of the intellectual defense of capitalism against Marxist criticisms.  And yet, Kuznets admitted that his empirical evidence was so limited that his general theory was mostly speculation, in which he was generalizing from a limited stretch of historical experience.

Piketty enters this debate over the history of economic inequality with more quantitative evidence over a longer period of history than has been offered by any other economic historian.  Piketty and his colleagues have collected and analyzed relevant data for over 25 countries, mostly from income tax returns and national economic accounts.  Most of this evidence is from the twentieth century, but some of it goes back into the eighteenth century.  This allows him to construct historical time series, such as the following for the United States.






The first figure shows the share of total wealth in the United States held by the wealthiest 10% of the people.  Wealth or capital is defined as the sum total of all nonhuman assets that can be owned and exchanged on some market, which includes real estate and financial and professional capital (such as plants, infrastructure, machinery, and patents).  This stock of capital owned at a particular point in time comes from the wealth appropriated or accumulated in all prior years.  From 1917 to 1932, the share of the top 10% ranged from 75% to 83% of the total wealth.  That share dropped to a low of 64% in 1986.  But then it started rising so that in 2012 it was back up to 75%.  The top 1% of the wealthiest Americans today own about 35% of all the wealth.  By contrast, the bottom 50% of the American people today own only 2% of the wealth.

The second figure shows the share of total yearly pre-tax income in the United States received by the 10% of the people with the highest incomes.  Income includes both payments to workers and others who contribute to the production of goods and services and payments to the owners of capital (such as profits, dividends, interest, rents, and royalties).  From 1917 to 1928, the share of the top 10% with the highest incomes ranged from 38% to 49%.  That share dropped in later years to a low of 33%.  But then it started rising in the late 1970s up to a high of over 50% in 2012.  The top 1% of Americans (with yearly incomes above $350,000) took almost 25% of total income.

Notice that for both wealth and income, economic inequality declined after World War I and then started rising in the 1970s or 1980s.  Piketty shows the same pattern in Europe.  He explains this as a consequence of the economic and political shocks of the two world wars and the Great Depression, which destroyed much of the wealth of the richest people, and of the high progressive tax rates that arose during this period. 

Only such severe shocks could reduce the rate of return on capital below the economic growth rate, which reduces inequality.  The general tendency of capitalism without such shocks, Piketty argues, is to foster a rate of return on capital (r) that is higher than the economic growth rate (g), which creates steadily increasing inequality.  Thus, Marx was right about the tendency of capitalism to increasingly concentrate wealth in the hands of a few capitalists, and thereby create an unjust economic inequality.  This must be so, Piketty argues, as long as > g, which is the central contradiction of capitalism.  Here's how Piketty charts the history of this inequality > g:


This Figure 10-10 of his book shows the rate of return on capital versus growth rate at the global level from antiquity to the present.  For 2,000 years or more, Piketty assumes, the average rate of return on capital has been 4.5-5%, while there has been essentially no annual economic growth until the eighteenth century.  Since then, the economic growth rate has been 1.5-4%.  We can see here that in the first half of the twentieth century, the economic and political shocks of that period created a situation where for the first time in history, the net return on capital was less than the growth rate, which brought a decline in inequality.  But if this is reversed in the rest of the 21st century, and once again r > g then we can expect that capitalism will resume its normal tendency to increase inequality.  To avoid this future turn to capitalist oligarchy, Piketty insists, we will need to adopt his policy proposals: a generous social welfare state, steeply progressive tax rates, and a global tax on wealth.

Piketty's critics have offered many criticisms of his presentation of historical data.  I will mention ten of the criticisms here.

1.  As Kevin Hassett indicated, some economists think that a huge part of the capital growth since 1980 has been from the growth in housing prices, which Piketty ignores.  Until the housing bubble burst, this contributed a lot to the growth in middle class wealth.

2.  Notice that in measuring income as recorded on tax returns, Piketty reports only "pre-tax, pre-transfer" income.  He does not consider the effects of taxation and government transfers in raising the incomes of the poor and the middle class.  When Hassett raised this point, Piketty responded by saying that the biggest transfer programs in the U.S.--Medicaid, Medicare, and Social Security--helped mostly retirees and health providers.  But the point remains that including the value of such transfers would lower inequality.

3.  Hassett argued that if we look at studies of consumption expenditures in the U.S., there hasn't been much increase in inequality since 1980.  Piketty responded by pointing out that the household surveys Hassett was citing relied on self-reporting, which probably underestimates the consumption of upper class people.

4.  Hassett and others have pointed out that Piketty looks mostly at inequality within nations, which ignores the importance of global inequality.  There is a lot of evidence that the inequality between the richest nations and the poorer nations has declined dramatically since 1980.

5.  Writing in National Review, Scott Winship has criticized Piketty for relying on income tax data.  This means that a lot of the income for middle class and working class people is invisible, because it is not reported on tax returns--for example, non-taxable capital gains from home sales, 401 (k) and IRA investments, and employer-provided health insurance.  Another problem with Piketty's income tax data is that it includes tax filings by young dependents with part-time jobs, which makes the bottom level of income look worse than it really is.

6.  Look again at Figure 10-10 above.  Notice that Piketty assumes that the return on capital was fixed at 4.5-5% for 1,800 years or more, up to 1820.  How does he know that?  Piketty writes: "For early periods, I have used a pure return of 4.5 percent, which should be taken as a minimum value (available historical data suggest average returns on the order of 5-6 percent)" (354).  In the footnote to this passage, he writes: "For land rent, the earliest data available for antiquity and the Middle Ages suggest annual returns of around 5 percent.  For interest on loans, we often find rates above 5 percent in earlier periods, typically on the order of 6-8 percent, even for loans with real estate collateral.  See, for example, the data collected by S. Homer and R. Sylla, A History of Interest Rates (New Brunswick, NJ: Rutgers University Press, 1996)" (613, n. 16).

Notice that Piketty does not cite any specific pages of Homer and Sylla's book.  If you look at their book, you will see that nowhere do they specify returns on capital "around 5 percent" or "typically on the order of 6-8 percent."  Actually, Homer and Sylla report wildly variable annual interest rates from less than 1% to over 100%.

Hunter Lewis at the Mises Institute website has pointed out the absurdity of Piketty's assumption of a steady compounded return on capital of 4.5% a year over 1,800 years.  If we started with $10 in year one, this fixed rate of return for 1,800 years would have generated a trillion times the entire wealth of the world today!

This points to a fundamental problem for Piketty's historical data analysis, a problem indicated by Tyler Cowen in his review of Piketty's book for Foreign Affairs.  He generally assumes that the return on capital is fixed and risk-free, so that the wealth of those with high capital assets will always grow and never decline.  Of course, that is not true.  In various passages of his book, Piketty admits that "the return on capital is in practice extremely volatile" (488).  (See also pages 6, 115-16, 353, 362, 411, 414, 446, 449-52, 488, 527.)

Capital has no value unless it is invested in productive activity; and since no one knows for sure what is going to be productive in the future, the owners of capital are always engaged in risky activity.  So, for example, as Piketty indicates, David Ricardo mistakenly assumed that all wealth would become concentrated in the hands of landowners, because of the rising price of land (6).  This was a mistake, because he did not anticipate how technological progress and industrial growth would reduce the value of farm land relative to other forms of wealth.  Contrary to what Piketty assumes in much of his book, capital is not a homogeneous blob of stuff that uniformly and inevitably earns a high rate of return.  In 2009, Circuit City was the second largest electronics retailer (behind Best Buy), and it went bankrupt.  Apparently, the capital of Circuit City had not been directed to its most productive uses.

Oddly, Picketty concedes this point when he comments on how the largest university endowments (Harvard, Yale, and Princeton) earn greater returns than smaller endowments, because the schools with the largest endowments can afford to spend $100 million a year to employ the best financial advisors and money managers to identify the high-risk but potentially high-return investments.  "To be sure, financial advisors and money managers are not infallible (to say the least), but their ability to identify more profitable investments is the main reason why the largest endowments obtain the highest returns" (450).

7.  Looking again at Figure 10-10, one should notice that Piketty's own data show a narrowing of the gap between the return on capital and economic growth at the end of the 20th century as compared with the 19th century.  If this is so, then the logic of his own argument should dictate that inequality is unlikely to increase in the future.

8.  So how does Piketty arrive at his prediction in Figure 10-10 that r will rise above g over the next few decades?  Well, he has to make lots of assumptions about the future.  He has to assume that the rate of economic growth will slow from 3.5-4% a year in the second half of the 20th century to 1.5% a year between 2050 and 2100.  He also has to assume that by 2050 there will be no taxes on capital (355).  But why should we make such assumptions about the future?

This points to a fundamental problem.  As James Q. Wilson used to say, we can't predict the future in the social sciences, because it's hard enough to predict the past.  As Piketty indicates, social scientists like Malthus, Ricardo, Marx, and Kuznets all failed to predict the future in any precise way.  So why should we expect Piketty to succeed where they failed?  He even admits that the title of his book--Capital in the Twenty-First Century--is false, and that a more accurate title would be Capital at the Dawn of the Twenty-First Century (35). But then after admitting that he cannot predict the future, he predicts the future based on lots of dubious assumptions; and it's that prediction about the future that has made his book a best-seller.

9.  Piketty predicts that by the middle of the 21st century, we will see a "terrifying" class structure of wealth distribution in which a small group of oligarchic capitalists will rule over the rest of society, and he thinks this is a realistic projection of trends that we already see today (571).  This implies that the class structure of income and wealth is becoming fixed, with no movement between classes.  But it's not clear that the evidence supports this.

As Mark Rank has indicated in the New York Times, the popular image of a static class division between the top and the bottom, or between the top 1 percent and the 99 percent, is false.  From studying 44 years of longitudinal data in the United States, Rank has shown that 12 percent of the population will be in the top 1 percent of the income distribution for at least one year.  39 percent will reach the top 5 percent, 56 percent will reach the top 10 percent, and 73 percent will reach the top 20 percent.

On the one hand, this shows the economic vulnerability of Americans in that even when they have reached affluence in one year, they might fall down in other years.  The title of Rank's article is "From Rags to Riches to Rags."  On the other hand, this does show the remarkable mobility and flux in the American economic class system, with lots of movement into and out of the top 1 percent or 5 percent or 10 percent.  This doesn't look like the sort of rigidly fixed class divisions that are implied by Piketty.

Moreover, Piketty concedes that one of the major innovations of capitalism has been the creation of a "patrimonial middle class" (260-62, 346-47).  In the United States, according to Piketty's numbers, the top 10% (the "upper class") receive 35% of the total labor income, and they own 70% of the capital; but the middle 40% (the "middle class") receive 40% of the labor income, and they own 25% of the capital (246-50).  That the middle class has such a large portion of the total income and wealth is probably unique in human history.

But still we can see a very high level of inequality.  Is this unjust and socially disruptive, as Piketty claims?  Much of this inequality can be explained as a consequence of the emergence of highly technological and cognitively challenging societies in which economic advancement depends largely on education, talent, and cognitive ability.  As the demand for highly skilled people rises, their incomes rise relative to the incomes of unskilled people.  Piketty concedes that this is largely true, but he also disparages the idea that economic success can always be justified as a reward for merit (21, 71, 85, 234, 305-308, 419-429, 443-47, 514).  Of course, he is right that economic success always depends to some large degree on lucky endowments or opportunities that are unearned.  But still, the importance of special cognitive skills and talents in a modern capitalist economy shows that economic success can be meritocratic, at least to some degree. 

As Charles Murray has argued, "the increasing market value of brains" is creating a "cognitive elite"--a new upper class of people who are economically, politically, and culturally dominant because they have the high cognitive ability (the high IQ) required to be successful in highly technological and mentally challenging economies.  That's why the people in the upper class tend to be people with advanced educational training at elite universities, and those in lower classes have less educational achievement.

10.  Finally, many critics have objected that Piketty's obsession with measuring inequality ignores the importance of measuring well-being.  Why should we worry about capitalist inequality if capitalism generally makes all of us better off in the long run?  Piketty admits that capitalism has promoted economic growth that has made most human beings better off on average than ever before in the history of the world.  He observes: "the material conditions of life have clearly improved dramatically since the Industrial Revolution, allowing people around the world to eat better, dress better, travel, learn, obtain medical care, and so on" (89).

But Piketty suggests that no matter how well off people are at the bottom of the social scale, the mere fact that they are at the bottom makes them miserable and angry, because of their envy for those at the top, and at some point this will lead to violent conflict.  Piketty writes: "At Haymarket Square in Chicago on May 1, 1888, and then at Fourmies, in northern France, on May 1, 1891, police fired on workers striking for higher wages.  Does this kind of violent clash between labor and capital belong to the past, or will it be an integral part of twenty-first-century history?" (39)  In asking this ominous question, Piketty clearly implies that we should expect to see such violence if we do not adopt his proposals for redistributing wealth. 

Marx admitted that during periods of economic expansion, capitalists would be able to pay higher wages and increase their profits at the same time. He argued, though, that this would not eliminate the exploitative features of capitalism:  "If capital is growing rapidly, wages may rise; the profit of capital rises incomparably more rapidly. The material position of the worker has improved, but at the cost of his social position. The social gulf that divides him from the capitalist has widened" (Marx-Engels Reader, 211)  Higher wages “would therefore be nothing but better payment for the slave, and would not conquer either for the worker or for labor their human status and dignity” (80).  Even if many modern workers do not suffer from the poverty of physical deprivation, they do suffer from the poverty of social deprivation. Although the workers may benefit from high incomes, they still occupy an inferior position in society as long as most of the wealth is controlled by a small economic elite, and thus inequality is necessarily degrading to those at the bottom.  Piketty seems to agree with Marx about this.

No comments: