Challenging the empirical contribution of Thomas Piketty's Capital in the Twenty-First Century.

Author:Magness, Phillip W.
Position:Book review
  1. Introduction

    Thomas Piketty's Capital in the Twenty-First Century (2014b) begins with a bold claim. The ensuing work, he promises, is "based on much more extensive historical and comparative data than were available to previous researchers, data covering three centuries and more than twenty countries, as well as a new theoretical framework that affords a deeper understanding of the underlying mechanisms" of his subject matter (2014b, p. 1). Though he qualifies this assertion with an acknowledgment of an imperfect and incomplete data set, this concession should not be mistaken for modesty. As Piketty repeatedly reminds his readers over the ensuing 700 pages, it is his unprecedented assemblage of data that supposedly sets his work apart from a literature on wealth inequality that--he contends--frequently suffers from "an abundance of prejudice and a paucity of fact" (2014b, p. 2).

    When paired with an unconventional theoretical argument rooted in hypothesized "laws of capitalism," and, perhaps more so, radical policy recommendations in the form of an 80 percent top marginal income tax rate and an annual 5 percent global wealth tax on the biggest fortunes (2014b, pp. 512, 530), Piketty's claim to an empirically robust and data-heavy narrative has always been the strongest ecumenical feature of his work. (1) Empirics are also the root of much of the book's claimed novelty, as well as its self-stated purpose of "patiently searching for facts and patterns and calmly analyzing economic, social, and political mechanisms that might explain them" (2014b, p. 3) in order to better inform the public discourse about the causes and consequences of global wealth inequality.

    Data, and more specifically the story in those data--"as extensive as possible a set of historical data" (2014b, p. 16) as can be gathered--thus become the main evidentiary tool on which Piketty predicates his work. Indeed, he goes on to extol his own "novel historical sources" and lays claim to a patient, empirically driven search for "facts and patterns" within them twice more before the conclusion of the first chapter (2014b, pp. 20, 31-32). While Piketty's product is part theoretical argument, part empirical exercise, and part policy recommendation, its unifying rationalization is an overarching historical narrative about the characteristics of human wealth accumulation, derived from and purportedly sustained in data.

    Given these extensive claims, not to mention the heavy criticism directed toward certain other works in the wealth inequality genre, it might come as some surprise to learn that Piketty's reported "three centuries" of empirics infrequently predate 1900 beyond a stray data point or two connected by a century's worth of linear interpolation. His claimed global analysis only consistently examines three countries--France, Britain, and the United States--with more than passing rigor, with only occasional forays into Sweden and Germany beyond that. Even many of his twentieth-century figures, presumably constructed from better records and more readily available data sources, are often products of further interpolation and decennial averaging around multiyear and decade-long gaps. Taken alone, these circumstances might only attest to the inherent difficulties of amassing a large, continuous economic time series. A more serious problem emerges, though, when an author attempts to interpret highly specific historical events through data points that are substantially less thorough or conclusive than their initial presentation suggests.

    Finally, the investigator may become downright alarmed when discovering the dubious foundation of some of Piketty's "novel" data sets, because Piketty's charts do not convey such weakness to the innocent reader. Furthermore, Piketty's narratives are occasionally peppered with wildly inaccurate historical "facts" that, coincidentally, seem to bolster his desired interpretation of the surrounding data. In this context, the various leaps and judgment calls that Piketty often makes in his historical reconstructions should raise alarm bells.

  2. History: Misconstrued and Missing

    At its most basic descriptive level, Piketty's presentation of major historical events at the center of his argument is laced with factual error. In addition to suggesting an inattentiveness to detail, a recurring problem of factual inaccuracy with historical events indicates that interpretative extrapolations from these errors, as well as more sophisticated data claims that appear throughout the book, may suffer from a basic fault in their underlying historical assumptions. While we will not endeavor to pick apart his most extensive historical recounting--the twentieth-century French economy--it is fair to note that he struggles, and struggles mightily at that, in many instances where he takes up the economic history of the United States. The book's favorable portrayal of FDR's New Deal policy initiatives, which function as a seminal event in Piketty's twentieth-century narrative as well as an important precedent for his prescription of confiscatory tax rates, is illustrative. Consider Piketty's descriptive retelling of Depression-era tax policy:

    The Great Depression of the 1930s struck the United States with extreme force, and many people blamed the economic and financial elites for having enriched themselves while leading the country to ruin ... Roosevelt came to power in 1933, when the crisis was already three years old and one-quarter of the country was unemployed. He immediately decided on a sharp increase in the top income tax rate, which had been decreased to 25 percent in the late 1920s and again under Hoover's disastrous presidency. The top rate rose to 63 percent in 1933 and then to 79 percent in 1937. (2014b, pp. 506-07) The problem with Piketty's historical narrative in this instance is one of basic fact. Simply put, his dates are all wrong. As readily accessible tax records illustrate, the top marginal income tax rate was actually brought down to 25 percent by the year 1925, which is not "the late 1920s" and which was well within the presidency of Calvin Coolidge (with Hoover taking office on March 4, 1929). (2) More troubling still for Piketty's narrative, it was under Hoover that the rate was raised to a decidedly punitive 63 percent under the Revenue Act of 1932. And just to round out Piketty's tax-error trifecta, the top rate increased under FDR to 79 percent in 1936, not 1937 as Piketty claims. (3)

    We see another example from this playbook--namely, inventing historical "facts" in order to support his narrative--a bit earlier in the book when Piketty informs his readers in a parenthetical remark, "Herbert Hoover, the US president in 1929, thought that limping businesses had to be 'liquidated,' and until Franklin Roosevelt replaced Hoover in 1933, they were" (2014b, p. 472). This claim is simply not true. Herbert Hoover (1952, pp. 30-31) in his memoirs quotes the (in)famous advice given to him by Treasury Secretary Andrew Mellon to "liquidate labor, liquidate stocks, liquidate the farmers, liquidate real estate." But the rhetorical point of Hoover bringing up this advice was to assure his reader that he had rejected such tough love. Hoover was compassionate with his misguided subordinate, though, writing, "Secretary Mellon was not hard-hearted ... He felt there would be less suffering if his course were pursued. The real trouble with him was that he insisted that this was just an ordinary boom-slump." Piketty is not alone in attributing to Hoover the very view that Hoover explicitly renounced, but it is nonetheless one of many examples of demonstrably false statements in the book that conveniently align with Piketty's historical worldview.

    The common theme of these factual errors is that Piketty uses them to augment certain historical political events and figures that align with his own modern prescriptions. In this sense, a specific narrative construction of the past--even though factually erroneous and misconstrued--may be seen to lend favor to desired policies in the present day. (4) We see a comparable episode when Piketty turns to more modern times and the U.S. federal minimum wage, when he writes, "From 1980 to 1990, under the presidents Ronald Reagan and George H. W. Bush, the federal minimum wage remained stuck at $3.35, which led to a significant decrease in purchasing power when inflation is factored in. It then rose to $5.25 under Bill Clinton in the 1990s and was frozen at that level under George W. Bush before being increased several times by Barack Obama after 2008" (2014b, p. 309).

    Here again, this "history" is utterly wrong, as readily available federal sources reveal. (5) Piketty's description is so at odds with actual history that it is easiest if we present the correct information in a table.

    Piketty's breezy discussion of the minimum wage is almost correct--though unbelievably misleading--if one were to look at his treatment up through Clinton. (Even here, he is wrong about 1980 versus 1981, and the minimum wage under Clinton was $5.15, not $5.25.) But for him to claim that the minimum wage was frozen under George W. Bush until being raised under Obama is utter nonsense. If we wanted to be pedantic, we could bring up the fact that the July 24, 2009 increase that occurred under Obama was due to legislation signed by George W. Bush, but that would detract from the more basic point that Piketty cannot even get his years, dollar amounts, and presidential administrations right. There are many problems with Piketty's portrayal, given the ease with which a more conscientious researcher could have verified such basic information from U.S. Department of Labor tables. Indeed, Piketty's own data files indicate his awareness of this source. (6) Yet, Piketty's bizarre errors aren't completely without a pattern: they serve to paint ostensibly market-friendly Republican presidents as ogres, while liberal...

To continue reading