Comments Guidelines

All comments are pre-moderated. No spam, slurs, personal attacks, or foul language will be allowed.

Thursday, August 21, 2014

Understanding Piketty, part 3

Part 3 of Thomas Piketty's Capital in the Twenty-First Century is the longest section of the book (230 pages out of 577), providing his analysis of inequality at the level of individuals. Notably, Piketty largely avoids the use of the familiar Gini index because, in his view, it obscures the issue by combining the effects of inequality based on income with those of inequality based on wealth. He treats the two sources of inequality separately throughout this analysis.

The first point Piketty emphasizes is one regular readers will be familiar with from my previous discussions of the Crédit Suisse wealth reports: Wealth is always more unequally distributed than income. The disparity is stark (p. 244):
...the upper 10 percent of the labor income distribution generally receives 25-30 percent of total labor income, whereas the top 10 percent of the capital income distribution always owns more than 50 percent of all wealth (and in some societies as much as 90 percent). Even more strikingly, perhaps, the bottom 50 percent of the wage distribution always receives a significant share of total labor income (generally between one-quarter and one-third), or approximately as much as the top 10 percent), whereas the bottom 50 percent of the wealth distribution owns nothing at all, or almost nothing (always less than 10 percent and usually less than 5 percent of total wealth, or one-tenth as much as the wealthiest 10 percent).
 One finding from his data which impressed me is that the inequality of wealth is virtually the same in all age cohorts, refuting the view that advanced industrialized societies are riven by inter-generational warfare (pp. 245-6). In other words, Baby Boomers may be less wealthy than their parents, but their incomes are just as unequally distributed, on average, and will continue to be so after they build up a few more assets and retire.

For my American readers, it is worth noting that Piketty claims that the United States today (meaning 2010) has the highest level of wage inequality ever seen in history, with 35% going to the top 10% and only 25% going to the bottom 50%. If current trends continue, though obviously an iffy proposition, the share of the top 10% would rise to 45% vs. just 20% for the bottom 50% (Table 7.1). In terms of wealth inequality the U.S. has as of 2010 almost reached the astronomical levels only seen by Europe circa 1910, with the top 10% owning 72% of all wealth, versus a 90% share in 1910 Europe (p.257). As this is survey-based data, Piketty says that 75% is a more likely figure, since surveys understate the top income and wealth shares. The bottom 50% of Americans own just 2% of the country's wealth. As I noted before, if r>g, this concentration of wealth is likely to become even more uneven over time.

Piketty then turns to the evolution of inequality over the course of the 20th century. He takes the French case and the U.S. case as representatives of "two worlds," one where inequality is largely caused by differentials in capital income (France) and one where it is caused more by wage inequality.

In France in 1910, the top 10% received over 45% of total income from both labor and capital, whereas its share of labor income was under 30%. Over the course of World War II, the share of the top decile (10%) dropped by about 15 percentage points, which rose and fell between 1945 and 1982, but since then is on the upswing once again (Figure 8.1). Wage inequality has been much more stable for France from 1910 to 2010. The higher you go in the income hierarchy, particularly within the top 1%, income from capital becomes more and more important, fully 60% of the income of the top .01%  of the French population (Figure 8.4). Piketty points out that while the upswing in capital income since 1982 does not yet appear large (only a few percentage points), it is built on an increase of capital as a percentage of national income, of inherited wealth, that will cause capital income to explode in the near future.

In the United States, we see the same decline in the income share of the top decile after 1940 as in France, but unlike France, since 1980 the United States has seen the share of the top 10% fully restored (Figure 8.5). Moreover, the vast majority of that recovery in the wealthy's share of income is due solely to the increasing income share going to the top 1% (Figure 8.6). Behind this is an increase in what he calls "supersalaries," the vast majority (60-70%) of which come from top managers and less than 5% from superstars in sports, acting, and the arts. Of course, though the route is different from that of France, it again causes an increase in capital's share of national income, making it likely that inherited wealth will rise in importance in the United States as well.

What caused this explosion of labor income inequality in the United States? Piketty cites several factors. First, he highlights the findings of Claudia Goldin and Lawrence Katz that wage inequality began to grow in the 1980s, "at precisely the moment when for the first time the number of college graduates stops growing, or at any rate grows much more slowly than before" (p. 306). Second, institutions matter in the labor market. The biggest factor holding down the low end of wages in the United States is the low level of the minimum wage, which peaked in 1969 at $10.10 in 2013 dollars (p. 309).

However, these factors do not explain what is going on at the very top of the wage distribution, according to Piketty. One piece of evidence is that the supermanager phenomenon is a characteristic of the Anglo-Saxon countries after 1980, whereas there are only very small increases for the top 1% in Continental European countries or Japan (Figures 9.2-9.4). But another important point is that, among the English-speaking countries, the United States has the largest increase in the income of the top 1%, and the same is true for the income share of the top 0.1% of incomes (Figures 9.5-9.6). So, something is going on specifically in the United States. It is not, however, that there is some fantastically higher level of productivity growth in the U.S. All of the countries considered are at the world's technological frontier; indeed, the evidence suggests that telecommunications infrastructure and cost are worse in the United States than many other industrialized countries.

Instead, what appears to account for the sharply divergent incomes of supermanagers are the difficulty of determining the marginal productivity of top managers (Piketty cites evidence that firms with the highest-paid executives do not perform better than those with lower-paid executives); the ability of top managers to "set their own salaries," for example by appointing the compensation committees that will judge them; and, as Part 4 will show, the decrease in top marginal tax rates and the change in compensation norms related to that (pp. 334-5).

Piketty then turns to inequality in capital ownership, beginning with the treasure trove of data compiled in France since 1790. One of his most amazing findings is that in France, despite the 1789 Revolution, the inequality of wealth increased throughout the 19th century, to the point where the top 1% owned 60% of all wealth on the eve of World War I (Figure 10.1). As noted before, the World Wars, the Great Depression, and higher taxation brought about a dramatic fall by 1970 in the share of of both the top 10% (to 60%) and the top 1% (20%). Here, too, the data yield a striking finding: essentially none of this went to the bottom 50% of the wealth distribution, but was redistributed downward only to those below the top 10% but in the top 50% (p. 342). This "patrimonial middle class," as Piketty calls it, has largely maintained its wealth share, with the top 10% and top 1% gaining a few percentage points from their 1970 low point.

Contrasting with France and other European countries for which there is reasonable wealth inequality data, the United States has seen a stronger recovery of the top shares of wealth holders, and in fact U.S. wealth inequality passed European wealth inequality in the mid-1950s, and has been higher ever since. In none of these countries, however, has wealth inequality returned to its highest levels. Piketty has three main explanations for why we have not (yet) seen a return to 1910 levels of wealth inequality. First, there simply hasn't been enough time to rebuild from the 1945 low point; indeed, wealth inequality did not even start increasing again until the 1970s in most industrialized countries. Second, the decline in wealth inequality was accompanied and intensified by a sharp increase in taxation on capital. Third, there was a high rate of economic growth for thirty years after 1945. However, all these factors may reverse in this century. 1945, obviously, is steadily receding from view. Tax competition may well spell the end of capital taxation. Finally the slowdown of demographic growth will reduce economic growth as well, exacerbating the key r>g relationship.

Piketty then turns to inheritance, again with data going back to 1790. Here his foil is economist Franco Modigliani, who posited that people save in order to finance their retirement, but bequeath essentially 0 wealth. Piketty finds that "the desire to perpetuate a family fortune has always played a central role" in savings decisions (p. 392). As evidence, he points out that "annuitized wealth" (non-heritable, such as Social Security but not a 401-k account) makes up a tiny fraction of private wealth (under 5% in France, and 15-20% in "English-speaking countries, where pension funds are more developed"). In other words, the desire to leave a bequest to one's heirs is the predominant fact behind large-scale savings behavior.

Piketty contends that a society dominated by inherited wealth becomes less democratic over time. Not only do the wealthy have increased mechanisms to influence political outcomes, but unearned income is an "affront" to the meritocratic story we tell ourselves. If high incomes are not based on merit, an important justification for this inequality disappears. He goes on to note that "rent" is not originally a pejorative term (as in, for example, "monopoly rent"). If capital is used in production, it yields income, and this is not due to monopoly and cannot be "solved" by greater competition. As Piketty says, universal suffrage [which in the 19th century had often been posed as fatal to the economy - KT] "ended the legal domination of politics by the wealthy. But it did not abolish the economic forces capable of producing a society of rentiers" (p. 424).

I have left out some of the technical detail of Piketty's argument, but one more point here needs emphasizing. As he shows (Figure 11.12), inheritance flows are increasing in Europe, and have been since 1980. The situation is not quite as bad in the United States, because the U.S. population is still growing, while Europe's is stagnating. However, long-term forecasts point to an eventual slowdown in population growth in the United States as well, in which case inherited wealth would emerge just as strongly as it already has in Europe.

Piketty's final points refer to global dimensions of inequality of wealth. His argument here is that while most people's instinct is to object less to entrepreneurial fortunes than to inherited ones, in fact there is less difference between the two that meets the eye. This is because, he argues, the largest fortunes are able to command the highest rates of return. He gives the example of the fortunes of Bill Gates, who obviously worked to build Microsoft, and that of Lillian Bettencourt, the heir of the L'Oréal fortune, "who never worked a day in her life" (p. 440). As it turns out, from 1990 to 2010, both saw their wealth increase by a factor of 12.5 times in that period. Large fortunes command the highest rate of return. However, the data Piketty presents do not fully support this. Gates' fortune was twice that of Bettencourt in 1990 and 2010, yet his rate of return was no more than hers. Furthermore, when Piketty reports on the returns made by sovereign wealth funds, we can see that Abu Dhabi's, the world's largest, worth more than all U.S. university endowments combined, nonetheless had lower earnings than did Harvard, Princeton, and Yale on their endowments. I think there is much merit to his overall claim, but it would appear to be a little more complicated than he lets on.

Last but not least, a couple of international wealth quick hits. First, will sovereign wealth funds own the world? No, but they could wind up amassing 10-20% of global capital by 2030 or 2040, which would be a much greater percentage of liquid global assets. Second, will China own the world? The short answer is no. Third, why do rich countries so frequently have negative asset positions? Are these counterbalanced by positive asset positions in poorer countries overall? The answer to this last question is no, so the answer to the previous question is that so much wealth has been diverted to tax havens, it makes the rich countries look poor, even though they aren't. Indeed, even the relatively low estimate of tax haven assets by Piketty's colleague Gabriel Zucman comes to more than twice the negative asset position of the rich countries.

This long section of the book is the necessary set-up for Part Four, where Piketty takes on still more received theories, and proposes his own recommendations for what can be done about inequality. I will turn to those questions in my next post.

UPDATE: Thanks to jack for catching a couple of typos.

Cross-posted at Angry Bear.

Sunday, August 17, 2014

Understanding Piketty, part 2

In my last post, I gave an introduction to the massive data work underlying Capital in the Twenty-First Century, as well as two clear results from Piketty's work. First, he showed that the optimistic Kuznets view after World War II that inequality was well on its way to being conquered was wrong, based on putting too much stock into a short-term trend. Inequality in fact has been increasing in the industrialized world since about 1980. Second, Piketty showed that a slow-growth economy is ripe for an increasing concentration of wealth in society, absent government action to counter that trend. He introduced the relationship r>g, which says the private return on capital is greater than the economic growth rate. When this holds true, as it has for most of history, inequality is likely to increase.

In this post, I address Part Two, "The Dynamics of the Capital/Income Ratio. The next two posts will address parts Three and Four, followed by a summation and critique of certain aspects of the book.

One important observation that Piketty makes is that in Europe, capital in the form of agricultural land accounted for 300-400% of gross national income (GNI), and total capital reached about 700% of GNI in the early 1700s. In Britain (Figure 3.1), France (Figure 3.2), and Germany (Figure 4.1), total capital fell below 300% of GNI in the period encompassing World War I and World War II. By 2010, total capital was back up to about 600% of GNI, but its composition had changed, with agricultural land falling to vanishingly low levels, replaced by housing and other domestic capital. In the United States (Figure 4.2), by contrast, farmland in 1770 was plentiful and cheap, making up about just 150% of GNI. Total capital also was much lower than in Europe, only about 300% of GNI. In the twentieth century, however, the U.S. did not suffer the devastation of the World Wars, so it had fewer and smaller dips in the value of total capital, which had risen to about 450% of national income in 2010; like in Europe, however, the value of U.S. agricultural land had also fallen to a tiny fraction of national income. Piketty points out if one includes the value of slaves, total U.S. capital in 1770-1810 rises by another 150% of national income (Figure 4.10), meaning that the capital/income ratio in the United States has been even more stable than it appears at first blush.

For Piketty, the resurgence of the capital/income ratio in the late 20th century is a consequence of slow growth. One important result of this is that capital's share of national income has increased since 1975, and labor's share has consequently fallen. According to his data for eight rich countries (the United States, the United Kingdom, Germany, Japan, France, Canada, Italy [the G-7, as they are usually called] and Australia.), "Capital income absorbs between 15 percent and 25 percent of national income in rich countries in 1970, and between 25 percent and 30 percent in 2000-2010" (Figure 6.5, p. 222). Notably, he raises the important point, central to my own academic work, that the increasing mobility of capital increases capital's bargaining power vis-a-vis both labor and governments (p. 221). He considers it likely that this factor has been mutually reinforcing with  the ability to substitute capital for labor. I would consider it not merely likely, but close to self-evident. Moreover, he omits (though I am sure he is aware) that one use capital mobility has been put to is to substitute less expensive for more expensive labor.

This increase in capital's share of national income shatters another comforting standard economic view, that the relative share of capital and labor is fixed. This assumption is built into a workhorse of neoclassical macroeconomic analysis, the Cobb-Davis production function. Piketty shows that, as with Kuznets work, the results of Cobb and Douglas generalize from a data sample that is far too short in term (p. 219).

My next post will analyze Part Three, "The Structure of Inequality." See you soon.