Phillip W. Magness

U.S. Economic & Political History
  • .: Phil Magness’ Blog :.

    Personal blog of Dr. Phil Magness, historian of the American Civil War and the 19th Century United States. Here I will post my thoughts and commentary on current research topics, upcoming events, and the general state of academia.
  • October 2017
    M T W T F S S
    « Sep    
  • On Intellectual History and MacLean’s Missing Leviathan

    Posted By on July 17, 2017

    Today I’ll offer a quick observation on the ongoing controversy about Nancy MacLean’s Democracy in Chains.

    In several recent interviews MacLean has presented her work as an “intellectual history” of James Buchanan. A few historians have come to her defense as well, taking a similar line and also suggesting that MacLean’s critics either don’t understand or are “misreading” the methods and “best practices” of intellectual history by focusing upon her thin documentation of the figures she presents as Buchanan’s intellectual influences.

    This line of argument reveals a critical oversight in MacLean’s treatment of Buchanan. It also shows that the claim about “intellectual history” methods is largely hollow. That oversight is the 17th century English philosopher Thomas Hobbes.

    If you are even minimally familiar with the work of Buchanan, you should know that one of the most important, recurring, and indeed ubiquitous thinkers that he engages across his vast body of scholarship is Thomas Hobbes. I’d even go so far as to suggest that it’s impossible to accurately write an intellectual history of Buchanan without understanding the deep complexities of his decades-long engagement with Hobbes’ work and his adopted role as Hobbes’ frequent interlocutor (yes, there are other figures like Frank Knight and Knut Wicksell who warrant similar notice for their formative influences on Buchanan’s political economy. Hobbes is a central figure – both utilized and engaged – in Buchanan’s political theory).

    Nor is any of this a big secret. Buchanan probably refers to Hobbes a hundred times or more in his collected academic works, and three of his major books centrally engage what he calls the Leviathan model of government in a direct and obvious reference to Hobbes’ famous work.

    So where is Hobbes in Nancy MacLean’s purported “intellectual history” of James M. Buchanan? Almost completely absent. He appears only once – a passing reference on page 33, where he is quickly cast aside and replaced by a completely imaginary connection to the obscure segregationist Agrarian poet Donald Davidson as the supposed source of Buchanan’s Leviathan concept.

    So not only does MacLean appear to have invented a non-existent connection to Davidson. In doing so she unintentionally jettisoned a central figure – Hobbes – from Buchanan’s corpus of scholarship.

    Nancy MacLean’s segregationist sins of omission…and commission

    Posted By on July 15, 2017

    One of the most inflammatory charges of Nancy MacLean’s new book Democracy in Chains holds that James M. Buchanan, and by extension his department and research center at the University of Virginia, served as something of an intellectual buttress to the segregationist forces of 1950s and 1960s Virginia politics after Brown v. Board. MacLean has very little direct evidence for this charge – in fact she’s even conceded in a couple of interviews that she has no direct documentation of Buchanan ever writing anything in favor of segregation. Her footnotes are similarly flimsy on this point and she resorts to misreading and misrepresenting Buchanan’s work on school choice to make her argument (Steve Horwitz documents the issues here).

    To bolster her non-existent case, MacLean resorts to playing a game of six degrees of separation in which she deploys a heavy stream of innuendo and unfounded supposition to write Buchanan into the pro-segregation political apparatus of Harry Flood Byrd, Sr. and a Richmond newspaper editor. As I’ve documented in my previous posts, she also fabricates claims out of thin air that allege Buchanan’s intellectual debts to the pro-segregation Vanderbilt Agrarians and to the 19th century pro-slavery politician John C. Calhoun. Remarkably, there’s almost no evidence for any of these claims – just a fanciful tale that is increasingly taking on conspiratorial overtones in the way that MacLean has mounted her defense.

    Sadly, a number of historians have displayed a remarkable credulity for MacLean’s claims on this point, even refusing to engage the evidence. One recent exception appears on the blog of John P. Jackson, who at least attempts to mount an actual defense of MacLean’s interpretation of Buchanan’s time at UVA. The whole piece is worth reading and engaging, but the core of Jackson’s argument on Brown and segregation appears in the following excerpt:

    “So, if we take the book as a whole, we find that MacLean shows that Buchanan was embedded in a state power structure where only those folks who were reliably segregationist were allowed to work (the famed “Byrd Machine” of Virginia), that he did nothing to rock that boat, and that his theories and arguments were welcomed by those looking to preserve segregation.”

    In assessing this argument, it’s important to recognize that Jackson accepts MacLean’s own portrayal of her evidence at face value. He does not get into the matter of whether she’s even proven her case, or whether her footnotes support her claims. He treats them as if they have been demonstrated as true.

    It’s a problematic position for him to begin with and again I’d encourage any reader to review the aforementioned links in which MacLean’s misuse of evidence is documented. But I want to focus on another aspect of Jackson’s argument, as I believe it does raise an important point: what was Buchanan’s relationship to the segregationist political environment around him in 1950s and 1960s Virginia? Jackson appears to believe that Buchanan lent it support through a combination of indirect policies such as school choice and silent acquiescence to the political machine around him. His source for this claim is, again, a mostly uncritical acceptance of MacLean’s own narrative, which asserts more or less the same thing.

    There’s another problem though: MacLean’s narrative about UVA is badly flawed. In order to portray Buchanan as a collusive and acquiescing partner of Virginia’s segregationist political machine, she omitted a critical piece of evidence that contradicts her narrative.

    In 1965 Buchanan recruited an economist by the name of William H. Hutt to serve as a visiting professor at the Thomas Jefferson Center, his hub of operations at UVA. Hutt was a natural fit for the role. He had recently retired from his position as chair of the economics department at the University of Capetown in South Africa. He was also an early contributor to the public choice school of thought, and his work drew heavily upon Buchanan and Gordon Tullock’s The Calculus of Consent. Hutt’s own academic reputation is noteworthy though because he was one of the leading academic opponents in South Africa of that country’s notorious Apartheid regime.

    Before he came to UVA, Hutt spent almost three decades criticizing the Apartheid government of his own country. His work repeatedly drew the ire of the South African government. In one notable instance from 1955, the Apartheid regime even suspended Hutt’s passport in an attempt to prevent him from presenting on the barbarism of this policy abroad. He regained his travel rights after a public controversy over his academic freedom, and remained undeterred in criticizing the South African government. Hutt’s work on Apartheid eventually culminated in a book length treatment of the subject entitled The Economics of the Colour Bar, which he published in 1964. The work notably employs an early version of public choice theory to explain the origins of Apartheid in South Africa as a form of regulatory capture to the benefit of white labor unions over black workers.

    When Buchanan recruited Hutt the following year, his international reputation as an Apartheid critic was near its peak. Hutt joined the department at UVA over the winter of 1965-66 and remained there for about two years on an extended stay. Drawing upon his recent book, Hutt delivered multiple lectures at Virginia and other universities in the region about the economics of Apartheid. During his stay Hutt also noticed an alarming similarity between the Apartheid regime in South Africa and the segregation in the southern United States. His home country’s brutal laws were more overt and severe, but the two only differed in degree. In fact, Hutt noticed that many segregationist laws had similar origins to South Africa. Both aimed to keep the black workforce out of competition and other forms of economic association with whites, and both used race to achieve this end.

    In short order, Hutt began extending his analysis of Apartheid to what he saw around him in the segregationist United States. While under Buchanan’s sponsorship at Virginia, he gave multiple lectures on this subject and penned a short article for the journal Modern Age describing their similarities. A  choice excerpt follows:

    “My own contribution to the gathering was a comparison of the not altogether dissimilar situation in my own country (the Republic of South Africa). I explained the origins of injustices to which the non-whites of South Africa have for long been subject; and general discussion suggested that, in several respects, the real disabilities of the American Negro population can be traced to the identical ultimate causes.”

    Hutt preceded to explain the differences of the two countries. Importantly, South Africa was majority-black while blacks were a minority in the United States. This, in part, explained the severity and virulent racism of the former regime when compared to the latter’s preference for more indirect forms of discrimination, among them the “separate but equal” doctrine. The similarities were nonetheless pronounced. Hutt continued:

    “These and other differences do not prevent us, however, from recognizing that in both countries, and for not very different reasons, non-whites are condemned to inferiority of productive opportunities, income, status, and respect. The reasons are rooted deeply in history. Yet many of the whites of this generation must share responsibility for perpetuation of the inferiority; for there are deliberately imposed man-made barriers to equality of economic opportunity barriers which are, I suggest, by all odds the most important ultimate cause of inequality of civil rights.”

    MacLean is certainly aware of Hutt’s presence at UVA because she mentions that Buchanan recruited him on p. 59 of her book. But she also conveniently leaves out any references whatsoever to Hutt’s research and activities during his time the Jefferson Center. In fact, she twists and contorts it in an opposite direction that even goes so far as to imply Hutt’s complicity in the same blatantly fabricated segregationist conspiracy she uses to tar Buchanan. Note the underlined segments in the excerpt below:

    MacLean makes no mention of Hutt’s better-known book on Apartheid, or the fact that his academic work and lectures while at UVA explicitly targeted both Apartheid and the parallel segregation regime in the United States. Not content to stop at this sin of omission though, MacLean converts it into a sin of commission. She highlights Hutt’s criticism of labor unions in a decade-old book and contorts it into an implicit nod of support for Virginia’s segregationist political class.

    Even this charge is absurd on its face. As Hutt repeatedly noted in his work, white labor unions were a major source of political support for Apartheid in South Africa. But MacLean subscribes to a worldview where labor unions are sacrosanct. And that, apparently, includes a license to violate the most basic evidentiary norms of the historian’s trade out of service to her political argument.

    Nancy MacLean’s Calhounite Imagination

    Posted By on June 29, 2017

    In my last post I documented how Nancy MacLean, the author of the new book Democracy in Chains, misused evidence to depict a non-existent intellectual debt between the economist James M. Buchanan and a group of pro-segregation Agrarian poets from Vanderbilt University. MacLean’s primary purpose in doing so was to prop up her own narrative, which portrays Buchanan’s role in the development of Public Choice economics as having been motivated by resentment over the Brown v. Board decision. This claim is not supported by any evidence in Buchanan’s works either.

    I’d like to turn to another similarly unsavory claim in MacLean’s book next, that of John C. Calhoun. As with the Agrarians, MacLean attempts to place Calhoun at the center of Buchanan’s work and the Public Choice school of thought more broadly. She minces no words in advancing this claim. At the outset of the book she declares Calhoun the “intellectual lodestar” of Buchanan and his heirs. She similarly describes Calhoun as a “recurrent theme in the brain trust” of scholars funded by the Koch brothers and devotes the prologue of her book to belaboring the different ways in which Buchanan is allegedly a secret Calhounite.

    MacLean’s argument runs into an immediate problem though. As with her claims about Donald Davidson and the Agrarians, Calhoun’s name does not appear anywhere in the 20 volume Collected Works of James M. Buchanan.

    To get around this complete deficiency of evidence, MacLean again resorts to her own imagination to forge the link nonetheless. She starts by enlisting a highly conspiratorial reading of a scholarly article by Tyler Cowen and Alex Tabarrok on a few similarities between Public Choice voting rule insights and Calhoun’s theoretical concept of the concurrent majority. The article itself is innocuous and steeped in textual analysis of this narrow point. Simply discussing Calhoun in a scholarly argument does not, after all, make one a Calhoun acolyte. Even Richard Hofstadter, the progressive historian who MacLean cites for her chapter title (and then conveniently neglects to read) conceded that Calhoun was “the last American statesman to do any primary political thinking.” Surely that recognition does not make Hofstadter a secret aficionado of plantation slavery.  MacLean is almost wholly uninterested in an intellectual discussion of the complexities of Calhoun though, preferring instead to enlist it as “proof” that Buchanan had somehow instilled Calhoun’s pro-slavery principles into his colleagues and proteges – and all despite never once mentioning his name.

    There is a far simpler explanation, of course. Buchanan’s Calculus of Consent was rooted in an exploration of James Madison’s theory of federalism, and in particular the use of constitutional mechanisms to constrain factional captures of legislative majorities. Calhoun similarly developed his concept of the concurrent majority as an extended commentary on Madison. Quite simply, any similarities between the two arise from the fact that they were both engaging the same well known arguments of the same author.

    Still, a single 1992 article about Calhoun’s concurrent majoritarian theories by two faculty colleagues of Buchanan falls far short of a basis from which to write Calhoun into Buchanan’s own work from the 1960s. This is where MacLean gets creative.

    As a substitute for Buchanan’s own works – which, again, are completely silent on Calhoun – she turns instead to Murray Rothbard, who wrote a few passages in the 1960s in which he spoke favorably of Calhoun’s concurrent majority concept. Since Rothbard was (a) supported by Koch and (b) a libertarian contemporary of Buchanan, this juxtaposition is apparently sufficient in her mind to link Calhoun into Buchanan’s own work. The Rothbard passage appears below:

    From there it is but a small step to point out Calhoun’s radical defenses of slavery, and thus tarnish Buchanan and Public Choice theory with that brush.

    MacLean’s argument neglected to account for a substantial complication. Rothbard, it turns out, was no fan of Buchanan or The Calculus of Consent, his main work (along with Gordon Tullock) on majoritarian constitutional mechanisms. In a 1960 essay for a book reviewer, Rothbard harshly criticized the work and specifically argued that it fell outside of the main body of libertarian political theory.

    “I am so out of sympathy with James M. Buchanan and Gordon Tullock’s The Calculus of Consent that I don’t think a particularly detailed critique to send to them would be worthwhile. I recognize that there are some merits to the piece: a searching for methodological individualism in political science, an emphasis upon unanimity rather than majority rule, and a harking back to the constitutional system of 1900 as better than the situation today. But these merits are, I believe, more ad hoc than integral to the main body of work. In considering the work as a whole, they are far overshadowed by the numerous flaws and fallacies.”

    Rothbard concluded by stating that the thought “the nub of the entire analysis of the book…is utterly and absolutely wrong.” Buchanan, for his own part, was not particularly fond of Rothbard’s contributions either as Don Boudreaux points out.

    Once again, Nancy MacLean went casting about for a link between James Buchanan and unsavory connections to white supremacist viewpoints – this time by way of Calhoun. Once again, she was unable to find any evidence of that connection in Buchanan’s own works. So what did she do instead? She invented one by stringing together other sources that do not actually show what she claims they show. And in fact, she failed to realize that the the author she uses to build a link from Buchanan to Calhoun – Murray Rothbard – was deeply hostile to Buchanan’s own work, directly contradicting her claimed link.

    (HT Michael Makovi for alerting me to the source of the Rothbard essay on Buchanan)

    How Nancy MacLean went whistlin’ Dixie

    Posted By on June 27, 2017

    If you read Duke University historian Nancy MacLean’s new book Democracy in Chains, you will probably come away thinking that the late economist James M. Buchanan believed himself to be something of an intellectual heir to the Vanderbilt Agrarians of the 1930s. According to MacLean, these now-obscure southern literary figures were a main reason Buchanan wanted to go to Vanderbilt University.

    Even though Buchanan’s family ultimately could not afford to send him to the prestigious university, MacLean claims that Buchanan owed these men a direct intellectual debt. They allegedly “stamped his vision of the good society and the just state.”  One of the Agrarians in particular, she claims, had a “decisive” influence on “Jim Buchanan’s emerging intellectual system” – the poet Donald Davidson.

    MacLean has a very specific reason for making this claim, and she returns to it at multiple points in her book. The Agrarians, in addition to spawning a southern literary revival (the novelist Robert Penn Warren was one of their members), were also segregationists. By connecting them to Buchanan, she bolsters one of the primary charges of her book: an attempt to link Buchanan’s economic theories to a claimed resentment over Brown v. Board and the subsequent defeat of racial segregation in 1960s Virginia.

    MacLean’s argument presents a challenge. Buchanan wrote very little on Brown or the ensuing school desegregation, and the archival evidence she presents from his papers is both thin and far short of the smoking gun she implies it to be. Instead, she sets out to strengthen her portrayal of Buchanan as a segregationist by tying him to other known segregationists. The Agrarians, and specifically Davidson, serve this purpose in her narrative by becoming formative intellectual influences on Buchanan.

    There’s a problem with MacLean’s story though: it appears to be completely made up.

    Her footnotes to the passages on the Agrarians don’t actually check out, and the Davidson link in particular appears to be a figment of her own imagination. I’ll walk through the sources in detail, starting with the passage where Davidson appears:

    MacLean’s purpose here is to identify Davidson as the font for one of Buchanan’s most frequently enlisted concepts from his academic work – the all-powerful Leviathan state. Of course most students of political philosophy will automatically recognize that this metaphor is a famous one. It derives from the 17th century English philosopher Thomas Hobbes, as MacLean begrudgingly concedes. But Buchanan’s version of the Leviathan is different, she contends – a product of Davidson’s “new and distinctive” use to describe a northern-dominated post-Civil War federal government and thus a code-word for racially tinged “states rights” and other nefarious purposes.

    There’s another problem with MacLean’s evidence. Donald Davidson’s name does not appear anywhere in Buchanan’s academic works. The massive 20 volume Collected Works of James Buchanan is searchable online. It contains most of his major books and papers and it does not yield a single hit for the name. Thomas Hobbes, by contrast, is one of the most frequently discussed figures in Buchanan’s works according to the index:

    MacLean nonetheless presses ahead with her invented connection and attempts to tar Buchanan with a litany of vices from the Agrarians: sympathy with the Confederacy, voter suppression, and racial animosity toward African-Americans. These and other charges may be seen in the passage below from MacLean, including a quotation that she claims to show Buchanan’s endorsement of the Agrarians’ vision:

    This passage points us to footnote 12 for the chapter for a list of its sources, which – again – purportedly link Buchanan to this literary group in ways that reflect all the aforementioned claims and charges. Except that’s not what the reader actually finds in footnote 12, or any of its neighboring notes on the Agrarians…

    Along with the citations to a couple Agrarian tracts, what we find instead is a fairly boilerplate list of secondary literature on 20th century racism and its links to the Agrarians. The only reference to Buchanan at all is not an archival source but rather a citation to page 126 of his autobiography, Better than Plowing. Not recalling any passages that would support what MacLean claims here about the Agrarians, I turned to Buchanan’s autobiography to check the reference. The page appears below and consists of a single passing reference to the Southern Agrarians having been influenced by Thomas Jefferson’s famous concept of the yeoman farmer.

    That’s it. There are no references to Donald Davidson. No segregationist visions, or pining over the Confederacy. No claims about wanting to study with the Agrarians at Vanderbilt. No intellectual nods to them at all, aside from a brief factual statement that they espoused a well known Jeffersonian argument about the agricultural lifestyle.

    MacLean’s book has already caught some flak for factual misrepresentations of her sources. In this case she appears to have simply made up an inflammatory association and tacked it onto Buchanan in an effort to paint him as a racist. When scrutinized though in her own sources, it becomes quickly apparent that she has no actual evidence to sustain her many detailed and specific claims. When one actually searches for the link and checks her sources, it quickly becomes apparent that there is none. In fact, one could legitimately note that there are more references to the pro-segregation Vanderbilt Agrarians on Nancy MacLean’s own CV than in the entire Collected Works of James M. Buchanan.

    On Tariffs and the American Civil War

    Posted By on May 26, 2017

    A new piece that I wrote on the role of tariffs in the American Civil War era is now available at the Essential Civil War Curriculum, hosted by Virginia Tech. This article is an encyclopedia-style overview of my research on the subject as well as what other scholars have written, but it provides a short primer on a subject that is often muddled in confusion and erroneous claims.

    If you ever encounter the argument that “tariffs caused the Civil War,” I’d simply urge to you read this piece first to see why the line of reasoning behind that claim is both in error and representative of a false “Lost Cause” historiography that came out of the Reconstruction era. At the same time though, I detail the history of the tariff issue’s role in antebellum economic debates and show how this culminated in an ancillary controversy, after slavery, on the eve of the Civil War.

    Philanthropy and the Great Depression: what historical tax records tell us about charity

    Posted By on May 19, 2017

    As part of my ongoing investigation into early 20th century tax policy, I recently compiled a data series to track patterns in charitable giving during the 1920s and 1930s. As a result of tax code changes in 1917, the IRS began allowing federal income tax payers to deduct up to 15% of their taxable income for donations to recognized philanthropic causes. Eligible donations included charities for the poor, as well as certain contributions to the arts, scientific study, and education. The policy was intended to incentivize private giving and its successor program persists to the present day in the form of tax deductible donations to eligible non-profit organizations.

    The IRS required tax filers to report their charitable deductions on their tax forms and tabulated the total amounts in their annual report on the income tax system. Surprisingly, little work has been done with the resulting data series on annual charitable contributions in this period.

    The chart above illustrates the total amount (inflation-adjusted) of charitable contributions in the period between the implementation of the deductions policy and the eve of World War II. Several patterns are noticeable. First, the Great Depression caused a precipitous decline in charitable giving that persisted into the late 1930s. While this effect is partially attributable to the economic decline caused by the Depression, its persistence also attests to the documented ‘crowding out’ effect that New Deal era spending had upon private charities. Earlier work by Jonathan Gruber and Daniel Hungerman noticed a similar drop-off in church-based charitable giving during the New Deal era, as the federal government picked up the tab for Depression-relief programs.

    Further evidence of this phenomenon may be seen in the raw IRS figures showing where charitable deductions came from. The chart below depicts earners in the $1 million+ income tax bracket, extended through the middle of World War II. Notice several patterns. A sharp spike in charitable giving followed the introduction of charitable giving exemptions, indicating the incentive structure worked. Giving among the wealthiest Americans also spiked dramatically after the 1924 tax cut, reducing high World War I-era rates to a top marginal rate of 25%. Charitable giving among the wealthy dropped off though during the New Deal, and especially following another income tax hike enacted by Herbert Hoover in 1932. It never really recovered until some point after World War II.

    Now compare this second chart to the first, depicting overall charitable giving. While donations by the wealthy are evident in the mid 1920s on this chart as well, something else is missing. The total amount of deductions actually began to accelerate around 1939-1940 in the first chart. This acceleration occurred even though deduction patterns for the wealthiest earners remained essentially flat during the same years. 

    What explains this somewhat counter-intuitive pattern? The answer may be seen in the next chart, showing raw charitable deduction amounts claimed by lower level income earners – specifically tax brackets for incomes between $5,000 and $10,000 per year (note: an almost identical pattern appears for earners in tax brackets below the $5,000 mark, although IRS records did not report these brackets individually in most years – only in cumulative).

    As we can see in this chart, donations from the bottom of the income ladder actually drove the spike in charitable giving in the early 1940s. The reason has to do with yet another set of changes to the federal tax code. Beginning on the eve of the war and continuing until 1945, Congress rapidly expanded the federal tax base onto lower income earners and simultaneously increased income tax enforcement to control for tax evaders. This was done in part to finance the war, though it also involved major administrative reforms such as the addition of automatic payroll withholding in 1943 to increase tax compliance. Faced with a newfound tax burden, lower income individuals began taking advantage of the same charitable deduction allowance that the wealthy utilized to alleviate their own tax burdens in the 1920s.

    Different Measurements of Income Inequality – the interwar Wisconsin Example

    Posted By on May 16, 2017

    I have a new paper, co-written with Vincent Geloso, on the measurement of inequality in Wisconsin between 1919 and 1941. The discussion’s geography may initially seem obscure, but there’s a method to this investigation. In the early part of the 20th century Wisconsin had a stable state income tax system and, more importantly, generated high quality data about its taxpayers on a semi-regular basis throughout the period. The Wisconsin tax featured low but moderately progressive rates on a scale of 1 to 7%, was applied broadly across the state’s population, and underwent relatively few statutory changes to its tax rates over the period between its inception in 1911 and the end of World War II.

    Wisconsin’s state income tax contrasts greatly with the federal income tax in the same time period. Federal tax rates fluctuated wildly until the end of World War II. Congress frequently tinkered with top marginal rates. They adopted a low of 25% in the 1920s, but raised them to over 60% in both World War I and the Great Depression era. By the end of World War II, it exceeded 90% on the top income earners. The federal income tax was also notoriously inconsistent in this period. Tax avoidance and evasion were both recognized problems of the system, and the tax base itself was constrained to a narrow set of the population – sometimes even less than 10% of all U.S. households were eligible to pay. This latter decision was in keeping with the tax’s intended progressivity, although it ended with a rapid tax base expansion during World War II that gave us our modern income tax system.

    The point in contrasting these two systems – Wisconsin and federal – is to show how the structure of the tax code affected the data each generated. For example, Wisconsin’s low rates and broad tax base ensured that it collected tax returns from the majority of Wisconsin households in most years even as the IRS collected returns from a much smaller tax base (these collections predated automatic payroll withholding and were therefore self-reported). During the late 1920s and 1930s, Wisconsin consistently reported a higher overall net income for the state than did IRS figures.

    Wisconsin also had more tax filers on the local level in every year between 1915 and 1941, when the IRS surpassed them due to the wartime tax expansions. During the 1930s it was not uncommon for Wisconsin to collect 3 to 4 times as many state-level tax returns as the IRS did in the state at the federal level.

    To make a long story short, all of this matters for purposes of data analysis, including the calculation of income shares and inequality. Thomas Piketty and Emmanual Saez’s article on the historical distribution of incomes in the United States uses federal income tax data from the IRS to compile its estimates. Though their method for doing so is an innovative improvement upon numerous prior attempts to calculate income shares, it still remains highly sensitive to the underlying quality of its tax data source.

    The Wisconsin tax system therefore has implications for measuring inequality by serving as a point of comparison against the IRS. Since Wisconsin had a more stable tax regime and collected its returns from a broader portion of the population, it is likely a superior data set to the federal IRS records for the reasons mentioned.

    So what happens when we calculate the income distribution for Wisconsin in the 1920s and 30s? If we use the state income tax records as our data source, we end up with very different results than the federal IRS records. In particular, the IRS records tend to show a higher level of inequality than the state records – which is to be expected considering that the IRS primarily collected returns from the wealthiest income brackets whereas Wisconsin taxed the entire state more broadly. A depiction of this effect in the 1920s appears in the graph above, and our full series of generated income shares for the top 10% of earners may be found in the appendix to the paper.

    On Keynes and Eugenics

    Posted By on April 25, 2017

    My article with Sean J. Hernandez on the “Economic Eugenicism of John Maynard Keynes” is now available at SSRN. This article should be approached as a synthesis of the role that eugenics played across Keynes’ career and in the formation of his economic theories. It is also the proverbial tip of the iceberg as far as new and under-explored evidence goes.

    I intend to write more on this topic in the coming months as part of an ongoing project to provide contextual detail and background to the emergence of Keynesian thought in the 1920s and 30s. I’ll offer a short preview in the form of a quotation of Keynes, recorded by Margaret Sanger at a 1925 conference on population in Geneva:

    “I am discouraged because they are not striking at fundamentals. They do not want to think of one fundamental question, and that is the population question. There is not a city, not a country, in the League of Nations today that will accept it, or discuss it, and until the nations of the world are willing to sit down and talk about their problems from the population point of view, its rate of growth, its distribution, and its quality, they might just as well throw their peace proposals into the waste basket, because they will never have international peace until they do consider that problem.”

    To my knowledge, none of Keynes’ biographers have engaged with his role at this event or the implications of his statement for his views on unemployment, conflict, war, and resource allocation. Stay tuned!

    How the AAUP bends statistics to create an adjunct crisis

    Posted By on April 13, 2017

    Earlier this week the American Association of University Professors released its annual report on the economic status of academia. Repeating a theme from prior years, this report heavily emphasizes the position of adjunct faculty and makes a number of bold empirical claims about the alleged growth of the part time academic workforce. For example, the statement released with the report asserts:

    “Faculty on part-time appointments continue to make up the largest share of the academic labor force, and the percentage of faculty jobs that are part time continued to trend higher.”

    Similar claims appear throughout the full report. The AAUP uses the following figure to “prove” assertions about the current size and alleged trajectory of adjunct growth.

    Taken at its face, the figure appears to support their contention of adjunct growth using figures from 1975, 1995, and 2015. This figure, however, is an act of empirical deception. In reality, the adjunct workforce peaked in 2011. It has been on a continuous decline ever since.

    By selectively presenting only 3 actual years of data with 20-year gaps in between, the AAUP’s figure creates an illusion of dramatic and continuing growth in the higher ed adjunct workforce. They make no real effort to understand the complex causes behind these data points, and the trend they do purport to show appears to be almost willfully designed to obscure the actual drop in higher ed’s use of adjunct labor since 2011.

    The actual pattern is revealed in this chart, which I compiled from the AAUP’s own previously published figures. Curiously, they abandoned the practice from previous years of including these data in their most recent report’s figures.

    As this trend line shows, the adjunct workforce percentage peaked in 2011 and was followed by a continued decline in 2013, 2014, and 2015 (the AAUP did not release a figure for 2012). The 2015 decline was even steeper than the previous two years. As it stands right now, the total percentage of adjuncts has dropped below its published figures for 2007. One would have to go back a decade to 2005 to find a lower percentage of adjuncts.

    The reasons for this decline are simple: the adjunct-heavy for-profit higher ed industry bubble collapsed around 2011, initiating a sharp overall drop in the number of adjunct faculty. As the latest stats show, this decline has continued for the past 5 years and shows no signs at the moment of letting up.

    The real question though is this: why is the AAUP, in spite of evidence showing a clear and still-ongoing contraction in the adjunct workforce, asserting the opposite to be true? And why did they utilize an intentionally partial and biased statistical portrayal to obscure the fact that adjunct numbers have been shrinking for the past several years? As is often the case with the AAUP of late, political ideology appears to trump both scientific credibility and intellectual consistency.


    The same AAUP report also obscures another trend about faculty employment. The total percentage of tenured and tenure-track faculty is actually up over the past decade, even though it is below its 1970s level. In addition to the roughly 30% of faculty who are tenured according to the AAUP figures, another 17% are employed in full time non-tenure track positions.

    Most of this spike has also taken place after the 2007-2008 financial crisis, defying another popular but empirically unattested claim about the supposed decline of tenure.

    Note that the AAUP figures also include graduate students as a “faculty” category, even though these positions are actually more akin to apprenticeships and usually come with sizable tuition credits in addition to payments. The removal of the grad student figures from their totals would have the effect of increasing the percentages of tenured and tenure-track faculty.

    Why Piketty-Saez yields an unreliable inequality estimate before World War II

    Posted By on April 8, 2017

    Next week I will be co-presenting a paper at the APEE conference on the reliability of historical estimates of income inequality in the United States. Our paper examines and offers a number of corrections to the widely cited income inequality time series by Thomas Piketty and Emmanuel Saez (2003). This series provides the baseline for multiple subsequent studies of inequality, and is the primary U.S. inequality series in the World Wealth & Income Database.

    The Piketty-Saez series is the primary example of the famous U-shaped inequality trend line for the United States in the 20th century that was prominently featured in Piketty’s 2014 book Capital in the 21st Century. It is calculated using income tax records from the IRS and a variety of complex statistical techniques to extract a distributional measure of income inequality for the top 1% through top 10% of income earners.

    In this post I want to focus specifically on how Piketty & Saez arrive at their estimates for the pre-World War II period, or basically the first half of their U-shape. This period is both interesting and statistically problematic because the IRS data they use as their source has several under-recognized drawbacks. Most American households were not eligible to pay income taxes prior to a rapid expansion of the tax base through new wartime income tax laws in 1941-1945. Before 1941, only about 10% of U.S. households – or even fewer in some years – were required to file their income taxes. In addition, tax enforcement was often deeply inconsistent in those early years, resulting in ample opportunities for both illegal tax evasion and legal tax avoidance. There were even year-to-year inconsistencies in the accounting measurements that the IRS employed to tabulate reported income. Piketty & Saez are aware of some of these issues and attempt to adjust for them (e.g. IRS accounting issues), but also largely inattentive to others (e.g. evasion and avoidance problems). Our paper argues that the cumulative effect of these issues renders their pre-World War II data, or basically the first half of the U-shape, unusable.

    I will be detailing several of these issues in the coming months, but today I’ll be walking you through some of the issues with one of the most dramatic adjustments that Piketty and Saez make. To reach their initial income distributions for the pre-war period, they begin by taking raw filing data from the IRS’ annual Statistics of Income (SOI) report (most of their post-war data comes from more comprehensive IRS microfile sources that are only available from the 1960s to the present). They use the SOI to calculate distributional estimates using a Pareto interpolation technique that is discussed at length in their paper’s data appendix. The technique itself is fairly standard fare, assuming the source data are accurate. Due to some of the aforementioned problems of IRS accounting inconsistencies and the low number of eligible tax filers before the war, they have to make a few adjustments to its results.

    It’s easiest to see the effects of the adjustments they make through 3 steps in the chart below (showing the calculations for the top 10% income share):

    Model 1, in blue, shows the raw Pareto interpolation from the unadjusted IRS SOI reports. Model 2, in red, attempts to address the problem of insufficient returns due to the low number of eligible tax filers before World War II. To do so it estimates and integrates a modest number of “missing returns” by taking the ratio of married vs. single tax filers in the pre-war years. As you can see, it increases the distributional share of the top 10% slightly before 1940. It does not alter any of the post-1940 results.

    A much larger adjustment comes from what we describe as Model 3, shown here in dark green. This is an accounting adjustment that purports to address the IRS’ switch from Net Income (NI) to Adjusted Gross Income (AGI) in 1943-1944. The difference between the two involves how each handle deductions for charitable giving, local and state tax payments, and some categories of interest payment. It is strictly a feature of the way the tax code handled each. AGI encompasses a share of untaxed but realized income that isn’t present in NI, hence the justification for making an adjustment.

    A problem emerges though with how Piketty and Saez calculate this NI-to-AGI adjustment for Model 3. As you can see in the chart above, the Model 3 adjustments are the most substantial change that Piketty and Saez make to the pre-World War II Pareto calculations. They consistently add about 5 percentage points to the distributional share before 1941, but relatively little thereafter.

    The way that Piketty and Saez go about calculating this adjustment is, unfortunately, opaque. Using their calculation files to replicate the adjustment, it appears that they simply inserted an even, constant, and nicely rounded multiplier across the pre-war income share. The multiplier in turn “bumps” the entire trendline upward until World War II, when Piketty and Saez begin relaxing their multipliers and then bottoming them out to zero with the 1943-1944 NI to AGI switch at the IRS. The weights that Piketty and Saez use for their multipliers are highlighted in blue on their spreadsheet below. The yellow highlighted cells reflect the post-AGI switch at the IRS.

    Notice that all of the adjustments are even, rounded numbers. Also notice two important features: (1) the weights they apply scale upward toward the highest income earning percentiles and (2) the weights are held perfectly constant across the board from 1918 to 1941, and then rapidly reduced from 1941 to 1943. This presumably reflects a number of assumptions that Piketty & Saez make, including the effects of the wartime expansion of the tax base that occurred with a succession of tax hikes after 1940. They also very conveniently create a shape in the resulting time series that looks like the first half of the famous U-shaped pattern.

    Piketty and Saez provide very little indication of where any of these weights even come from, let alone if they accurately reflect the size and distribution of untaxed deductions from the pre-AGI period of IRS accounting. The evenly rounded and constant numbers also strongly suggest that simple “guesstimation” is at play (my readers will remember that Piketty has a bad habit of guesstimating numbers and weights along these lines in historical periods of sparse or insufficient data, usually to create the trend line shape that he wishes to depict).

    Part of the problem comes from the unavoidable issue of insufficient historical data. The IRS records are not sufficiently complete to perform a direct NI-to-AGI adjustment in most pre-war years. The question then becomes one of whether the Piketty-Saez weights, apparently guesstimated, are justifiable. Let me offer one piece of evidence that strongly suggests they are not. While the IRS did not report or differentiate all types of deductions in the pre-war period, they did track tax-exempt donations to charities from the mid 1920s onward.

    Annual charitable deduction totals fluctuated wildly throughout this period, showing deep responsiveness to changes in the tax code and to the Great Depression. The inflation-adjusted totals are depicted below:

    Charitable deductions represent only a part of the NI-to-AGI adjustment, so we cannot make a direct 1-to-1 claim about their effects. Still, the severity of the fluctuations itself suggests at least one strong reason why the stable, constant multiplier that Piketty and Saez employ could be highly problematic.

    This is but one of many similar issues I will be highlighting with their pre-World War II adjustments in the forthcoming paper and future posts. It is a substantial one though, the removal of which completely alters and substantially diminishes the first half of their famous U-shaped distribution.