You might think that there is nothing left to be said about this year’s liberal darling liber, Capital in the Twenty-First Century, but Piketty’s opus has just been named the Financial Times’ Business Book of the Year, not so long after that same publication also questioned a variety of Piketty’s empirical conclusions.
It turns out the FT hardly exhausted the problems with Piketty’s data. One issue that has not been discussed at length: the French economist’s work on the source of wealth in the United States.
Affluent Americans, Piketty believes, broadly belong to two main categories. “Rentiers” have inherited wealth and live off the capital returns of their dynastic fortunes, while “supermanagers” are salaried corporate executives whose pay has grown spectacularly by negotiating large compensation packages.
There is one problem with this categorization: It’s wrong.
Even accounting for the fact that economic models simplify things, Piketty has misunderstood the American data and cites the relevant research on top earners incorrectly.
Wealthy rentiers and salaried corporate executives may be vaguely unsympathetic groups, but they do not constitute the bulk of rich Americans. In particular, Piketty underestimates the importance of entrepreneurs and business owners.
It’s not a minor oversight: Self-employed business owners who actively manage their firms own around 70 percent of the wealth of the top 0.1 percent. With top earnings, too, business owners are far more important than salaried executives. This doesn’t change the fact that inequality is high and rising, but it undermines Piketty’s explanation for why inequality is increasing in the United States.
While the exact magnitude is debated, Piketty is indisputably right that economic inequality has increased rapidly. According to the Congressional Budget Office, the pre-tax income share of the top 1 percent increased from an already high 9 percent in 1979 to 15 percent in 2010. I sympathize with Piketty’s concerns and entirely agree that the explosion of income disparity that we are witnessing is problematic for society. According to the CBO, real median incomes increased by less than 40 percent between 1980 and 2010, while earnings of the top 1 percent tripled.
Decades of middle-class stagnation was hardly what the adherents of the Chicago school were hoping for. The economy is not a zero-sum game, but nor is it an infinite-sum game. With competition for status, positional goods, mates, and permanently scarce real-estate, both absolute and relative income influence well-being.
So what is Piketty getting so wrong? Economists’ standard theories on rising inequality include skill-biased technological change, globalization, financialization, tax cuts for the rich, low-skilled immigration, and winner-take-all-markets.
With the exception of tax cuts, Piketty rejects these causes. He is particularly hostile toward explanations focusing on higher returns to skill, which he believes “legitimize” inequality.
Instead, Piketty’s controversial “r > g” theory views capital concentration as the fundamental force in capitalism. Piketty believes that most top wealth is inherited and that the rich tend to pass on their growing wealth to the next generation. It goes like this: Assume that the wealthy receive a return of 6 percent of their capital, spend half and reinvest the remaining 3 percent perpetually while the growth rate of the economy is 2 percent; the fortunes of the rich will outgrow the economy by 1 percent per year and eventually take over the economy.
But Piketty is aware that the force driving his r > g theory — rentiers with ever-growing inherited fortunes — is ill-suited for the United States. The book therefore introduces a second force behind inequality to better account for trends here. Piketty believes that most top earning Americans are senior managers of large firms, a group that he labels “supermanagers”:
The increase in very high incomes and very high salaries primarily reflects the advent of “supermanagers,” that is, top executives of large firms who have managed to obtain extremely high, historically unprecedented compensation packages for their labor. One possible explanation of this is that the skills and productivity of these top managers rose suddenly in relation to those of other workers. Another explanation, which to me seems more plausible and turns out to be much more consistent with the evidence, is that these top managers by and large have the power to set their own remuneration, in some cases without limit and in many cases without any clear relation to their individual productivity, which in any case is very difficult to estimate in a large organization.
One of Piketty’s central arguments is that the earnings of corporate executives are not determined by market forces but through bargaining in hierarchical organizations. Social norms, corporate politics, corruption and nepotism by “hierarchical superiors” determine what supermanagers are paid. The transfer of hundreds of billions of dollars in these negotiations, Piketty thinks, has caused stagnating middle class incomes and rising top incomes without contributing to growth. In Piketty’s view, rich Americans have raised their incomes through rent-seeking rather than value creation.
Economists typically cannot measure objective value creation, which makes it hard to theoretically debate whether someone’s income reflects their marginal productivity or not — i.e., whether they’re engaged in rent-seeking.
But we can say this: “Hierarchical relationships” and “relative bargaining power” do not explain rising top incomes in America. (Jim Manzi has also discussed this issue from a similar angle earlier on NRO.) This is because the largest segment of the group Piketty calls “supermanagers” do not work in hierarchical organizations to begin with — they are business owners running their own firms. The earnings of business owners are determined by market forces and equity prices, not bargaining with their “hierarchical superiors.” What then is the source for the claim that the vast majority of the rich are big business executives? It appears that Piketty has misunderstood a 2010 study by economists Bakija, Cole, and Heim. Citing this study, he writes:
Recent research, based on matching declared income on tax returns with corporate compensation records, allows me to state that the vast majority (60 to 70 percent, depending on what definitions one chooses) of the top 0.1 percent of the income hierarchy in 2000–2010 consists of top managers.
The 2010 study by Bakija, Cole, and Heim is the only empirical source in Piketty’s book for his claim that most top U.S. earners are corporate executives. But Bakjia, Cole, and Heim do not match tax returns with corporate compensation and do not look at the 2000–2010 period.
Piketty might be thinking about a 2013 study by Giertz and Mortenson, which matched tax returns with corporate compensation for the years 2000–2010. That study matches Piketty’s methodological description, but is never cited in his book.
It appears that Piketty has confused the two studies and, more important, misunderstood both. As we will see, neither study suggests that two-thirds of the 0.1 percent of top earning Americans are salaried corporate executives.
Bajika, Cole, and Heim (2010) is an important study in this field, relying on IRS income-tax data for the period 1979–2005 to try to identify top earners. This isn’t easy, since American tax returns are anonymous and contain little background data. The authors combine the occupation reported in most tax returns with data on sources of income to guess at the occupation of top earners. “The data demonstrate that executives, managers, supervisors, and financial professionals account for about 60 percent of the top 0.1 percent of income earners in recent years,” the abstract reads.
A superficial reading of this abstract has given rise the myth that 60 percent of top earners in the U.S. are CEOs. This misses the fact that Bajika, Cole, and Heim 2010 includes self-employed business owners among executives and managers. The study cannot determine with certainty whether top earners are employees working for someone else or entrepreneurs and co-owners of their own firms. The tax data does, however, allow the authors to make a rough division between “salaried” and a “closely held business,” in which the owners receive pass-through income.
The more detailed breakdown of the top 0.1 percent:
‐ 21 percent belonged to various groups of professionals (lawyers, doctors, management consultants, computer scientists, engineers, academics and top earners in art, media and sports)
The claim that 60–70 percent of top earners are salaried corporate executives whose compensation is determined by their hierarchical superiors is thus shown to be incorrect. Neither entrepreneurs nor hedge-fund managers fit Piketty’s definition of the “supermanager.” The type of big-business supermanager Piketty describes would almost entirely belong to the category of salaried executives and managers. But this group only includes 20 percent of top earners, and even this category contains many wealthy entrepreneurs. (The occupation description in the tax data used in this study classifies the chairman of Microsoft and the CEO of Amazon as salaried executives.)
Other empirical strategies using other data sources confirm the importance of entrepreneurs among the rich. An influential study by Cagetti and De Nardi used Federal Reserve data in 1989 to estimate the wealth share of self-employed business owners. Self-employed business owners constituted 8 percent of the population, but owned one-third of national wealth and more than half the wealth of the top 1 percent. (Some business owners are passive investors rather than proprietors, and this study does not define them as entrepreneurs.)
I have updated the numbers for 2010 using the same method. In 2010, self-employed business owners account for an astonishing 70 percent of the wealth of the top 0.1 percent. If we look at top earnings rather than top wealth, self-employed business owners accounted for around 50 percent of the total earnings of the top 0.1 percent. These facts are never mentioned in Piketty’s book.
This all seems to be of a piece with the mistake Piketty makes in assuming that entrepreneurs are an economically unimportant group whose role has only been exaggerated by the Right in order to justify inequality.
Don’t believe he got this so wrong? In the discussion of top 0.1 percent earners, Piketty explicitly says the following: “‘Superentrepreneurs’ of the Bill Gates type are so few in number that they are not relevant for the analysis of income inequality.”
This is not the case. It would be more accurate to write that corporate executives are so few in number that they are not relevant for the analysis of income inequality. As a share of the rich, salaried corporate CEOs are trivial compared with business owners and the financial sector. Last year, the 500 CEOs of the Fortune 500 collectively earned $5 billion. By comparison, the best-performing hedge-fund manager alone earned $4 billion. The total earnings of the 0.1 percent are somewhere around $800 billion per year.
The insignificance of corporate CEOs for income concentration is confirmed by the 2013 study by Giertz and Mortenson (which Piketty seemed to confuse with Bajika, Cole, and Heim). Giertz and Mortenson matched income on tax returns with corporate compensation records in 2000–2010, looking at CEOs and other top corporate executives among the S&P 1500. The collective earnings of these approximately 7,000 American top executives, including equity and stock options, was $18 billion in 2010. $18 billion would equal the annual income of around 400.000 median-income American workers.
But as provocative as the disparity between management and labor is for struggling blue-collar workers, this figure actually shows that executive compensation cannot account for rising income inequality. In the context of extreme top earnings, this is not an impressive number: The $18 billion earned by the 7,000 salaried executives working for the largest American corporations equals the collective earnings of merely the ten best-paid hedge-fund managers last year!
The study by Giertz and Mortenson, as well as a 2013 study by Kaplan and Rauh, confirm that salaried corporate executives are too few in number to account for more than 2 or 3 percent of the earnings of the top 0.1 percent. CEOs with excessive compensation packages negotiated in smoky rooms are an easy target, but they do not constitute an economically significant group.
Inherited or created wealth?
We now know that business owners are important, but do we know if they created their firm or merely inherited it? Piketty argues that the large majority of Americans have inherited their wealth, but, once again, he takes a position that’s not empirically well-supported.
One of the main sources of data on top wealth is the Forbes list of billionaires. According to Forbes’s classification, around 70 percent of American billionaires are self-made in the sense that they did not inherit large fortunes. The Forbes estimates are quite problematic for Piketty’s theory, which predicts that most top wealth should be inherited.
In the book, Piketty vehemently disputes the idea that most billionaires are self-made entrepreneurs. He admits that he has no evidence for this, but suggests that it is the other way around and that inherited wealth constitutes 60 to 70 percent of billionaires. Forbes estimates, he says, are methodologically biased because billionaire entrepreneurs are easier to find.
The Forbes list is obviously not a precision estimate. Forbes itself refers to its estimates as “highly educated guesses.”
But external evaluation using tax data have nevertheless shown that Forbes’s list is surprisingly accurate. Piketty seems to be unaware that researchers working for the IRS have matched the Forbes billionaire list with estate-tax data available to the Internal Revenue Service.
For example, one study by IRS researchers Raub, Johnson, and Newcomb compared the tax returns of the wealthy and their families with the Forbes list. For Piketty’s view to be correct, the Forbes list must suffer from a massive methodological bias that makes it miss the majority of billionaire heirs. But Raub, Johnson, and Newcomb only found a small number of billionaires who should have been on Forbes list but weren’t — just 26 individuals in the tax data who should have been on the list but weren’t, compared with 376 who were in the tax data and the list.
Piketty himself suggest another way to indirectly test whether Forbes systematically misses inherited wealth. Fortunes above $10 billion, he says, are easy to find and therefore fairly accurate, whereas there is a large bias in smaller fortunes. If Piketty is right that most billionaires are heirs and Forbes is just missing it, we should expect the share of inherited wealth to be, ceteris paribus, higher above $10 billion range, where Forbes is supposedly more accurate. This is not the case. In last year’s Forbes 400 list, 32 percent of billionaires with fortunes at or above $10 billion had inherited their wealth. Among those with fortunes in the $1–10 billion range, the share of inherited wealth was also 32 percent.
It is worth noting that Forbes has no problem finding inherited wealth in societies where inheritance is more important: In France, Forbes estimates that 70 percent of the wealth of billionaires is owned by heirs. This is not far from Piketty’s own estimates of the aggregate inherited wealth share in France. Perhaps the problem isn’t Forbes, but Piketty’s theory.
Top returns from capital or human capital?
Piketty doesn’t deny the importance of entrepreneurs for innovation, but mistakenly believes that entrepreneurs are few in number. Not all wealthy business owners are entrepreneurs by the Schumpeterian definition, which requires innovation. It would similarly be naïve to accept that all business owners are socially productive. Some have grown rich through crony capitalism, rent-seeking, or insider trading. William Baumol has also made the subtle point that we should not assume that talent is synonymous with creating value for the rest of society. The richest Wall Street crooks and Russian oligarchs also tend to be the most talented, and precisely therefore the most socially destructive.
But leave aside the question of whether talent is socially productive or destructive: Piketty’s problem is that he downplays its importance entirely in favor of the role of “capital.”
Among the 0.1 percent top earners, a high and increasing share income is declared on tax returns as capital income, most importantly capital gains and dividends. Piketty relies on this as evidence that the rich are mostly rentiers. More important, he uses the high share of wealthy people’s income derived from capital to combat theories about inequality that emphasize skill-biased technological change or higher returns to talent.
This argument confuses the economic definition of capital with the accounting definition used by the tax system. Among top earners, the line between capital and labor becomes blurry. This group often supplies effort, talent, and capital as an inseparable bundle: Much of earnings is initially reinvested in the firm and used as a complement to talent and effort, and the return to talent can come out in capital gains and dividends.
It is quite reasonable to say that 100 percent of the fortunes created by Gates and Jobs business ventures reflects return to their talent, even if it’s declared on a tax form as a return to “capital.” This distinction between capital and talent also makes more sense when discussing the causes for rising inequality: Piketty’s r > g theory suggests that top incomes to a large extent increase because of passively invested fortunes’ generating ever larger returns, whereas other economists emphasize the return to talent. But as we’ve just seen, some capital-based fortunes are in fact returns to talent.
Piketty himself uses a reasonable and strict definition of capital income:
Today, the rents produced by an asset are nothing other than the income on capital, whether in the form of rent, interest, dividends, profits, royalties, or any other legal category of revenue, provided that such income is simply remuneration for ownership of the asset, independent of any labor.
For self-employed business owners, this definition obviously doesn’t hold. The earnings of this group reflect labor as well as capital. In the entrepreneurial sector, the talent and effort of the founders appear to be far more important than passive investments. Passively invested capital is not a particularly scarce resource, whereas entrepreneurial talent is. This is not a detail when discussing inequality and wealth, given that entrepreneurs constitute the majority of the rich.
The line between capital and labor is also blurry for the financial industry and partners in law firms, consulting agencies and similar sectors. The fact that effective capital gains taxes are significantly lower than income taxes has created an additional incentive to take out the return to talent in the form of capital rather than as labor earnings, artificially making it look like more of the wealthy are rentiers.
Piketty certainly has a point in that the Right tends to use “entrepreneurs” as poster boys to justify inequality. Many of the rich are not, in fact, entrepreneurs. It’s not clear to me that recent the explosion of speculative finance and real-estate constitutes socially productive innovation. Even when it comes to productive entrepreneurship, it is by no means a given that any level of inequality can be justified by entrepreneurial innovation.
Ayn Rand had her protagonists invent amazing sci-fi technology that no one else can replace in order to justify no moral boundaries to the wealth share “creators” claim. That’s all well and good in her stories, but how should we think of distributional issues if the rich don’t invent super-steel and perpetual-motion machines?
But there are also obvious problems associated with the other extreme — totally ignoring entrepreneurs. Piketty’s Capital creates an alternative universe where the rich consist of rentiers born with a silver spoon in their mouths and big-business CEOs who serve themselves raises with the same. This image of the rich is convenient for the Left, but inaccurate. Leaving out entrepreneurs leads Piketty astray in several places.
Besides, incorporating entrepreneurs in the analysis is not merely a right-versus-left question. It also resolves empirical puzzles, such as why the capital share of top incomes can increase at the same time as the return to talent and human capital is rising.
Piketty is a brilliant economist and still a young man. Hopefully, in future work, he will leave more place for entrepreneurs and business owners — what an innovation that would be.
— Tino Sanandaji is a National Review contributor and a fellow at the Research Institute of Industrial Economics in Stockholm.