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.
  • March 2019
    M T W T F S S
    « Dec    
  • The Decline of the Adjunct Workforce

    Posted By on May 28, 2018

    A little under three years ago I posted a brief comment about some stats I was compiling for an article on the higher ed workforce. The post investigated a myth that was popular at the time and remains so today, namely that adjunct employment had grown to encompass an astounding 76% of the higher ed workforce. The claim is empirically false on its face, and rests upon the conflation of adjunct, or part-time, faculty with all non-tenure track appointments. It nonetheless serviced a popular political slogan, suggesting that adjuncts were the “new faculty majority” and predicting the impending “adjunctification” of higher ed.

    My 2015 post cast doubt not only on this overall number, but the actual source of statistical trends. Adjunct employment had peaked in 2011 at near-parity with full time positions, but the most recent available statistics at the time – 2013 – hinted at the beginnings of a retreat in adjunct numbers. After digging a little deeper, I discovered that these trends were primarily being driven by an unnoticed source: for-profit higher ed.

    For-profit institutions use an almost exclusively adjunct faculty pool. It therefore followed that as the for-profit university sector boomed between the mid 1990s and 2011, adjunct hiring would follow suit. And it did. While adjunct use certainly grew elsewhere in this period, the single biggest contributor to the sector’s growth was for-profit higher ed. But in 2015, the for-profit sector was undergoing a contraction that has only intensified since then.

    So I made a prediction.

    Not only were we already past the adjunct workforce peak, but I further predicted that the total number of adjunct professors would continue to shrink in the wake of the for-profit contraction.

    This prediction and my other stats work quickly drew the ire of a cadre of adjunct union activists who were entirely uninterested in statistical analysis that ran counter to their political cause. It also prompted a particularly nasty article by Aaron Barlow of the AAUP, who falsely accused me of statistical misrepresentation. I confronted Barlow at the time about his charge, providing evidence that he was deeply mistaken. It turned out he was no more interested in an accurate or honest assessment of the stats than the adjunct activists of the twittersphere. Needless to say, my prediction was a lone voice on the subject in an activism-obsessed higher ed sector that had already concluded its error before even considering the evidence I brought to the table.

    The reason I bring this up is that the U.S. Department of Education released a report last week containing its most recent set of faculty workforce statistics, covering 2016. The stats fully bear out the continued decline I predicted 3 years ago. The adjunct workforce shed 30,000 jobs between 2011 and 2016, while full time faculty positions grew by 50,000. It also just so happens that the decline in adjunct jobs directly parallels the decline in the for-profit higher ed workforce.

    The AAUP would be wise to temper its activist rhetoric on the adjunct issue in light of this developing trend. For-profit higher ed has continued to contract since amidst ongoing accreditation and fiscal solvency problems, suggesting the decline will play out over several more reporting cycles.

    How communism affects the inequality U-curve

    Posted By on May 7, 2018

    U-curves are all the rage in inequality studies. Increasingly viewed as a global phenomenon, the distribitional U-curve attempts to show changes in top income and wealth shares over the course of the last century. In other words, it tracks the holdings of the top 1%, 5%, or 10% and so forth over time.

    The common story behind the U-curve is one of a great levelling in top income shares up to and including the World War II era, followed by a rebound that’s supposedly taken place since the 1980s and that manifests today in political claims about rising inequality. The posited causes for the movements on the U-curve are complex, but the levelling is usually attributed to some combination of disasters in the first half of the 20th century (two world wars, the Great Depression) and to specific policies (progressive income taxes, redistribution, and the modern welfare state). These causal claims are more often asserted than demonstrated, particularly on the fiscal policy side. As a result, the U-curve upswing since 1980 is almost always attributed to a resurgence of fiscal conservatism since the Reagan-Thatcher era that cut into the mid century’s highly progressive income tax structure. Tax hikes, by implication, then become a favored policy to reduce inequality among those who fret about its alleged rise in recent years.

    I’ve discussed the measurement problems with the U-curve elsewhere on this blog, but let’s assume for the moment that it reflects the past 100 years and there was indeed a “great levelling” at the mid-century mark. While scholars have focused almost entirely upon tax policy as a driver of this pattern in the United States and western Europe, one other factor on the world scene has received comparatively little study.

    Specifically, the rise and fall of communism, and particularly its totalitarian derivatives, appears to have precipitated a dramatic levelling in top income shares at the mid-century. The following chart, generated from the World Inequality Database, shows all currently available estimates for the top 1% income share from communist countries.

    The figure includes century long estimates for the Soviet Union/Russia and two eastern bloc countries, as well as evidence of the rebound since the 1970s in Poland and China. Note that the leveling effect in recorded top income shares is almost instantaneous with the adoption of communism. This much may be seen as well in Russia, which levels some 25 years before the rest of the world. The levelling/rebound pattern seen in the communist world is also atypically severe.

    Of course one purpose in my pointing this out is to show that measured income inequality is not an accurate representation of the distributional conditions of material well-being. During the mid-century period, practically all of the countries reflected on this chart were places of severe economic and political repression. Even as they appeared “equal” on paper, that equality entailed a race to the bottom – unless, of course, you were a member of the political elite and therefore enjoyed the non-income luxuries afforded to the communist party leadership. For the rest, communism entailed an equality of shared misery.

    Income inequality in the United States: it’s flatter than you probably realize

    Posted By on May 1, 2018

    Most economic discussions of inequality in the United States begin with a U-shaped curve. More specifically, they begin with historical estimates of top income shares (e.g. the top 10%, 5%, and 1%) as depicted in Thomas Piketty and Emmanuel Saez’s famous 2003 paper on the subject. When these figures are displayed across the entire 20th century they look like a giant U. They show a period of relatively high inequality prior to World War II, followed by a massive levelling at the mid-century mark, followed by an inequality rebound that begins around 1980 and persists to this day.

    The Piketty-Saez estimates were a groundbreaking example of historical work that forms the basis for many subsequent studies of inequality today. They also suffer from data limitations, owing to problems of both accuracy and consistency in historical IRS records.

    For the past several months I’ve been working on an empirical project (along with Vincent Geloso, John Moore, and Phil Schlosser) that investigates some of these problems. In particular, we look at the portion of the U-curve where these data issues are most severe: 1917 to 1943. Our results produce a series of new estimates for top income shares in this period, accounting for a number of oversights, errors, and problematic adjustments that appeared in the original Piketty-Saez paper.

    We focus on three issues in particular, and offer corrections.

    The first one derives from the way(s) the IRS reported total income earnings prior to 1944, when it switched to a standardized accounting definition of Adjusted Gross Income (AGI). Think of AGI as your total earnings for the year, minus certain personal and work expenses that may be itemized on your tax form. Prior to 1944, the IRS tabulated tax returns a different way using “Net Income,” which is typically a smaller amount than AGI. The main difference between Net Income and AGI is that the latter includes several additional deduction-eligible income categories such as charitable contributions, interest payments on debt, and state and local taxes. What this means is that pre-1944 tax statistics generally undercount income relative to the AGI standards. As a result you have to harmonize the statistics from these years to get an apples-to-apples comparison with 1944 to the present day.

    Piketty-Saez took a shortcut to harmonize the two sets of numbers. They essentially guesstimated how much earners in each tax bracket deducted and adjusted them upward using a single stable multiplier that tiers up its deduction levels to the wealthiest earners. The problem with this adjustment is that taxpayers in the 1910s-1940s actually did not claim deductions at the same rates from year to year. There were even some shifting patterns in deduction rates between high and lower income earners. To correct this, we went back into the data files and extracted totals for the main AGI-eligible deduction categories. Since IRS records were sporadic, we had to estimate missing data in some years by imputation from partial data. Other years in this period have full records of deduction amount by tax bracket. The result is an improved data set that includes reported Net Income plus the main AGI-eligible deductions, and that reflects the year-to-year fluctuations in deduction patterns that are missing from the original Piketty-Saez guesstimation. We then use this improved series to re-estimate top income shares prior to 1944.

    Second, we corrected an error in one of Piketty & Saez’s adjustments to account for another recognized problem in the pre-1940 data. Until 1940, the income tax only applied to upper-middle class incomes and higher, but it did so at a distorted rate. Personal exemption levels varied depending on whether you filed as a married couple or a single filer. To account for the “missing” filers under the much higher married couple threshold, Piketty-Saez uses a ratio of married to single filers from after 1940 when the issue was no longer a problem due to expansions of the income tax base. Unfortunately, Piketty-Saez picked a poor year to calculate this ratio. They used 1942. The problem with 1942 stems from the United States’ entry into World War II, which also precipitated a large influx of young single male filers onto the tax roles as they entered into military service. As a result, any married filer adjustment using 1942 as its base ratio will tend to skew inequality upward in earlier years. To fix this problem, we calculated a new base ratio from 1941 (the U.S. did not enter the war until mid-December). This ratio is then used to generate multipliers to perform the necessary adjustment to our results from the first step.

    Third, we identified a further distortion in the Piketty-Saez calculations that comes from their selection of a suitable income denominator to actually run the calculations for top income shares. The denominator comes from national income accounts, and is not as much of an issue in recent years owing to standardized accounting practices and high quality data. The 1910s and 20s in particular are a different story, as formal attempts to estimate national accounts were just coming into existence back then. Simon Kuznets (who also pioneered national accounting in the 1930s) produced a denominator estimate in his own 1953 study of income inequality. Piketty-Saez reject Kuznets’ numbers though as being too high and offer a substantially lower denominator. This is a discretionary call, but we believe it is in error because they based it on a comparison with post-1944 tax data, after the aforementioned switch to AGI. This justification does not hold up as well before 1944 under the old Net Income standards used in tax reporting. We accordingly return to the Kuznets estimates and offer a range of plausible scenarios for selecting a denominator while estimating the pre-World War II years.

    When all of these adjustments are brought together, we end up with a substantially revised historical series for top income shares. We present a range of denominator options for comparison, but the effect is an across-the-board lowering of estimated inequality prior to 1944. For example, using a 90% Kuznets denominator base, the income share of the top 10% drops by roughly three to seven percentage points in most years. The only exception is an acute inequality spike in 1928-29, consistent with the stock market bubble and crash at the outset of the Great Depression. Our results still show an overall leveling trend between 1929 and 1945, but it is much more gradual than the sharp U-shaped drop depicted in Piketty-Saez.

    The next interesting implication of our new series is what it tells us about the century-long story. Since the publication of Piketty-Saez in 2003, other scholars have focused upon the accuracy of their income share estimates from 1980 to the present. One major point of contention revolves around the Tax Reform Act of 1986, which eliminated several tax shelter loopholes that were popular among the wealthy in the mid 20th century as ways to relieve their tax burdens. Piketty-Saez do not account for the effects of this tax code change on how wealthy persons reported their income. The result of this oversight may give the illusion of a more rapid rebound in the U-curve than actually happened. A new study by Gerald Auten and David Splinter offers a series of corrections to the Piketty-Saez series from 1960 to the present that account for this change in the tax code as well as other related scenarios that affect reported income shares. Their results generally flatten the depicted post-1980 upswing of the original U-curve, although they still show modestly rising inequality.

    So what happens when you combine the adjustments to the pre-WWII period and more recent times? The chart below depicts the century-long results (in our case using the 90% adjusted Kuznets denominator). The original Piketty-Saez U-curve may be seen in blue. Our pre-1944 revised figures are in red. And the most conservative adjustment by Auten & Splinter is in yellow.

    What they show is a century-long pattern that looks more like a tea saucer shape than a U-curve. In fact, the top 10% U.S. income share has remained in a fairly stable 10 percentage point range for the last 100 years, except for the brief aforementioned inequality spike leading up to 1929. If you take the cumulative effect of the adjustments into account, inequality patterns appear to be much flatter and less prone to fluctuation than what we previously believed.



    On the intellectual uselessness of a certain method of socioeconomic analysis

    Posted By on March 27, 2018

    A few brief observations:

    1. The Labor Theory of Value is incapable of functionally explaining even basic economic relationships. See Menger 1871.

    2. The notion that class identity functionally drives political or any other type of collective action is hopelessly incoherent and undermined by a pervasive free rider problem. See Olson 1965, Section IV.

    3. Even if one were to assume that the initial allocation of all property is by mere theft (and it is not), its effectual consequences are entirely subordinate to the question of whether property rights exist in the first place. See Coase 1960.

    4. The predictive ability of historical materialism in the ~150 years since its formulation is practically zero, although the cost of attempting to force its predictions into being is a hundred million bodies.

    5. In practice, the concept of alienation is indistinguishable from subjective emoting about things that the individual exhibiting “estrangement” envies, and envy is a difficult concept to defend as the basis of a system of social allocation as it reduces to little more than subjective valuation executed by forcible acquisition.

    Now note that one might still dispute if these particular observations are true. I’d entertain an argument as much, although they seem to be uphill battles if indeed the associated references hold. But let’s explore the implications by asking a question. Grant for a moment that the five observations are true. We may then ask what, exactly, remains of Marxist thought that is of any value in explaining anything?


    Abolitionists, Communists, and Jacobin magazine’s fake Jeff Davis quote

    Posted By on March 4, 2018

    One of the stranger recent currents in historical discussions around slavery involves an increasingly common attempt by scholars and activists on the far left to enlist historical abolitionists as ideological ammunition to the causes of anti-capitalism or even socialism. A primary motivating factor appears to be the so-called “New History of Capitalism” (NHC) genre of books by Sven Beckert, Ed Baptist and others who have all advanced arguments that portray capitalism and slavery as deeply intertwined and interdependent economic systems.

    The abolitionist extension of the NHC argument may be seen at several places in Manisha Sinha’s recent book The Slave’s Cause. While not a primary thesis of Sinha’s book, she does indulge the claimed link between abolitionism and anti-capitalism, citing the NHC literature. The argument as she presents it is a strange one, relying largely on a rhetorical sleight of hand that notes “if slavery is capitalism, as the currently fashionable historical interpretation has it, the movement to abolish it is, at the very least, its obverse.”

    Of course for that claim to be true, the currently fashionable NHC interpretation must first be accepted, and that remains a highly dubious proposition (for those interested, I critique this genre at length in What Is Classical Liberal History?, my new book with Michael Douma). But it is also wanting on another front. Establishing abolitionism as the obverse of capitalism requires – at minimum – some evidence of the abolitionists’ alleged hostility to capitalism.

    Abolitionism was a complex and multifaceted movement spanning not only multiple political factions but also several continents, so providing substantiation for the claim is no small task. It must also get around the unambiguous philosophical affinities for laissez-faire political economy that several of the more prominent abolitionist leaders incorporated into their philosophical systems (William Lloyd Garrison in the U.S. and Richard Cobden in Britain, for example, both had strong philosophical proclivities toward free market or laissez-faire economics). And finally it must also address the pronounced anti-capitalist undercurrents that many of the late antebellum proslavery theorists put forth as their own (see here in particular).

    A full discussion of these topics could probably fill an entire book, but one certain obstacle that the NHC reinterpretation of the abolitionists (as well as its more brazen variants among self-described socialist claimants of abolitionism) is a shortage of evidence in favor of the claimed link. One has to reach pretty deeply into the abolitionist barrel to find socialist adherents of the cause, and few if any of them had comparable stature to Garrison, Douglass, Sumner, Higginson or similar figures who operated more comfortably within classical liberal philosophical frameworks.

    To get around this problem, a fairly recent article in the far-left magazine Jacobin attempted to make the case for “ideological connections between abolitionism and socialism” by turning to abolitionism’s slaveholding critics to make their argument for them. The meandering piece offers very little in the way of historical context to substantiate its argument, preferring instead to assert it by declamation and pad it a bit with armchair intersectionality theory that supposedly demonstrates an overlapping of various social causes. But they also bolster this claim with a list of cherrypicked quotations from slaveholders.

    A pair of quotes come from the notorious pro-slavery theorist George Fitzhugh in which he ascribes socialist or communist monikers to his adversaries, albeit with very little context. This is done for a reason, as Fitzhugh himself is a poor witness to bring to Jacobin’s cause. They do so through selective editing, but only to omit the fact that Fitzhugh professed a radical anti-capitalist vision that trended strongly toward the philosophical tendencies of the reactionary socialist Thomas Carlyle. It was Fitzhugh who urged slave-owning southerners to “throw Adam Smith, Say, Ricardo & Co., in the fire” and who declared the theory of laissez-faire to be “at war with all kinds of slavery, for they in fact assert that individuals and peoples prosper most when governed least.”

    Although now deprived of Fitzhugh, Jacobin has a few more testimonials for its case. The most prominent remaining example by far is future Confederate president Jefferson Davis himself, or so it would seem. Jacobin’s author accordingly quotes Davis asserting that “the European Socialists, who, in wild radicalism, . . . are the correspondents of the American abolitionists” and proceeding to place them in league with the French socialist-anarchist theorist Pierre-Joseph Proudhon. The crusade against slavery, he continues, would “superinduce attacks on all property, North and South” and, presumably, lead to an ascendance of socialism in America.

    If no less than the president of the Confederacy declared that the abolitionists were in league with the socialists, then surely it’s a powerful case for not only the NHC genre but Jacobin’s own even more extreme spin in which the forces who stood against slavery also stood for socialist upheaval.

    But there’s a problem. The unsourced Jefferson Davis quote is not a Jefferson Davis quote at all. It’s actually a misattributed passage by Caleb Cushing, a Massachusetts politician and diplomat who sided with the Union during the Civil War. Cushing’s address was delivered at a meeting in Bangor, Maine on January 6, 1860 and reprinted shortly thereafter in the newspaper excerpt below:

    Where and how this passage received its misattribution is uncertain, as the Jacobin article gave neither a source nor even a date. The episode is nonetheless highly representative of the quality of the argument that article advances. Along with easily identified misrepresentations of Fitzhugh that obscure his avowed anti-capitalism, Jacobin‘s main evidence for a socialist-aligned “red abolitionism” comes down to a fake Jeff Davis quote.

    The Epistemic Toxins of False Historical Claims

    Posted By on February 27, 2018

    Allegations of racism carry a substantial social stigma in today’s intellectual climate. Provided that the allegation is valid, this may be a desirable effect. Racism is insidiously unethical as it fundamentally devalues the targeted person. This may make it worthy of not only condemnation, but the ostracizing that often follows a racist action. Knowledge that racism will be forthrightly condemned and socially penalized is a highly effective means of making racist beliefs and behaviors costly to maintain – and thereby discouraging their propagation.

    For the same reason though, racism is not a charge to throw around lightly. Accusing an innocent person of racist beliefs or actions can destroy a reputation and unfairly subject its target to the derision and scorn we might legitimately direct at a genuine act of racism. To falsely charge someone with racism, or to even mislead others into making and believing that charge, is therefore an unethical act of another type.

    This brings us back to the subject of Democracy in Chains, a 2017 book by Nancy MacLean about the academic career of the late economist James M. Buchanan. Although Democracy in Chains was published on a trade press that does not utilize peer review, its author has traded in on the currency of her own academic prestige to present it as a scholarly work. MacLean is the holder of a prestigious chair at Duke University, and has used the occasion of this book to lecture widely at universities across the United States. Her paying hosts include several Ivy League institutions and other top ranked history departments.

    Democracy in Chains is a deeply problematic book though, and not simply because it never saw peer review. As I’ve discussed and documented extensively on this blog, its author also misuses and misrepresents historical evidence to portray Buchanan as a complicit figure in the segregationist “massive resistance” to Brown v. Board in the 1950s and 60s. The problem with MacLean’s analysis is that this charge – a central thesis of her book and by far its most bombastic claim – is a historical falsehood. MacLean’s cited documentation simply does not support her contention of complicity, and in fact counter-evidence that she overlooked, ignored, or omitted shows the exact opposite to be true. Buchanan had no interest in segregation, and in fact condemned it at multiple points across his long academic career, starting as early as his doctoral dissertation in 1948 (for a detailed analysis of these findings see the paper-length discussion here).

    Unfortunately, MacLean’s false charges of segregationist complicity and other insinuations of racism have had a poisonous effect upon the academic discourse around Buchanan’s scholarly contributions. Relying upon MacLean’s own academic prestige, numerous other scholars have accepted her depictions of Buchanan at face value, believing them to be the product of meticulous archival research. In turn, they attach the social stigma of racism to Buchanan’s name. His scholarly contributions, it then follows, may be dismissed, discarded, and maligned.

    In her more careful moments, MacLean stops just short of making the charge of racism herself (although numerous other less guarded moments also exist). A few of MacLean’s more tendentious defenders have claimed exoneration of her work on account of hedging its claims with allegedly nuanced and non-committal phrasing by way of qualifiers in her text. Call it the innuendo defense: “she didn’t actually call Buchanan a racist, she simply tricked her readers into believing he was a racist through word games,” as if that somehow makes it better.

    What’s absolutely certain though is that the overwhelming majority of MacLean’s sympathetic readers reached the intended conclusion, and now casually toss around false and reckless claims of racism when discussing Buchanan’s work. The examples are numerous. Citing MacLean, political scientist Richard Saull recently deemed Buchanan “a supporter of Jim Crow.” Legal scholar Ariela Gross places him in league with overt segregationists in a mission “to preserve Jim Crow.” Historian Bethany Moreton, who also invokes the aforementioned innuendo defense of MacLean, nonetheless attributes Buchanan’s work on public choice theory to “racist roots.” Education scholars Eleni Schirmer and Michael W. Apple nonchalantly accuse Buchanan of advancing a “white supremacist” ideology cloaked in the “race-neutral language” of economics.

    The most bombastic example to date comes from James Brewer Stewart, himself a distinguished historian of abolitionism. Citing MacLean’s discovery of what he alleges to be Buchanan’s “historical DNA,” Stewart repeats the charge of racist motivation and extends it in even more reckless directions than MacLean’s original work. The product not only tarnishes Buchanan with racism, but stops just short of placing him in the vicinity of Ku Klux Klan rallies and accusing him of wanting to revive a modern version of a pro-slavery political ideology:

    “With sit-ins and Klan rallies erupting around him Buchanan, like the nullifiers, fastened his attention on the overreaching, grasping, oppressing government power over highly privileged whites, saying not a word about the menace of militant blacks. The DNA shared by alt-righters and fire eating slaveholders now begins to come into focus and what takes shape in our vision is two ideologies, born in the midst of racial crisis, both advancing the interests of ultra rich white people while never trafficking openly in white supremacy.”

    While Stewart briefly invokes the innuendo defense himself by hedging on whether the evidence could establish the bombastic claims he just made – now extended to both Buchanan and the Koch Brothers. He then removes any doubt of where he personally stands though, asserting a “family likeness” between them and the racist alt-right of today. In a later passage he follows an extended digression into the radical pro-slavery “fire-eater” ideology of antebellum South Carolina by asserting “James M. Buchanan, one strongly suspects, likely would have approved.”

    Stewart since has claimed some defense for his angry rhetoric by pointing to the op-ed format of his piece, yet ethical considerations still remain. Opinion writing certainly encompasses the art of provocation, but it is not a license to make reckless allegations, be they outright or through innuendo (and Stewart, like MacLean, does both).

    Regardless of the medium, historians are still ethically obliged to faithfully represent the past even if they differ as to particulars of interpretation. History is an evidence-based craft. In the cases of Stewart and MacLean, we see instances of two highly regarded historians making claims that disregard their evidentiary obligations to the past and, instead, advance a politically motivated line of attack that aims to paint another recently-deceased scholar with unsubstantiated charges of racism. Note that the effect is the same, whether they make this charge directly or by heavy innuendo (and in MacLean’s case we’ve seen instances of both). The result is that Buchanan, a widely cited and influential contributor to the field of economics, is now casually associated with racist actions that he played no part in and that his accusers have utterly failed to substantiate. The result is epistemically destructive to the scholarly discourse and, sadly, much of that destruction appears to be intentional.

    Eugenics and the Origins of the AAUP

    Posted By on January 13, 2018

    The history behind the founding of the American Association of University Professors (AAUP) almost always begins with the case of Edward A. Ross. A prominent economist and sociologist in his day, Ross was forced to resign from his faculty position at Stanford University in November 1900 after running afoul of the political views of the university’s surviving namesake, Jane Stanford. Ross’s dismissal set into motion the events that culminated in the founding of the AAUP, some fifteen years later.

    The incident prompted John Dewey to compose a formal defense of academic freedom in the January 1902 issue of Educational Review, which later served as the basis for the AAUP’s own founding principles. The American Economic Association also conducted an investigation into the episode under the charge of Columbia University’s Edwin R.A. Seligman, severely chastising Stanford’s actions. Ross’s dismissal even precipitated a wave of faculty resignations from Stanford in protest, starting with his philosophy department colleague Arthur Lovejoy. Dewey, Seligman, and Lovejoy would all go on to become the principal charter members of the AAUP, as well as early presidents of the organization. Lovejoy would later identify the Ross case as a crucial precipitating event in the formation of the organization.  Seligman took charge of organizing what became the organization’s main investigative body, the Committee A on Academic Freedom. And Ross himself would later go on to hold several positions within the leadership ranks of the AAUP.

    While political events certainly precipitated Ross’s forced resignation, a fair amount of confusion and mythology have come to characterize the exact nature of his disagreement with Mrs. Stanford. Ross and Stanford had previously clashed over the former’s support for William Jennings Bryan and the free silver movement during the 1896 election. He also allegedly criticized Jane Stanford’s late husband, California railroad baron and politician Leland Stanford, and further irritated her by advocating for municipal ownership of street railroads and utilities. The most immediate source of their tension though was a speech that Ross gave a few months prior to his dismissal in which the professor argued for the strengthening of immigration restrictions on Chinese and other Asian laborers.

    While acknowledging that Ross’s position on Chinese labor displayed a number of ugly racial undertones, the cause of the dispute itself is normally presented as a political struggle between Ross’s own strident progressivism and the politically conservative, moneyed interests of Mrs. Stanford and the railroad industry. The Stanford fortune, after all, came from a company that widely utilized Chinese laborers to construct its railroad lines. One version of the episode on the AAUP website even portrays Ross’s comments as “statements about railroad monopolies” – a substantially more innocuous cause than barring immigrants from entering the country. Another AAUP account states that Ross was dismissed for speaking out “against privately owned railroads and Chinese immigration” and heavily implies that the latter issue was only a derivative of the former, on account of the railroad tycoons habits of using Asian labor.

    The framing of the case as a battle between labor and the capitalist classes owes much to Ross’s own publicity campaign after his resignation, including a vocal self-portrayal as a martyr to free speech and to the values of the progressive political left. Ross organized an effective PR campaign by taking his case to the newspapers and enlisting other prominent academics from around the country to speak out on his behalf. In doing so he painted himself as a defender of the white working class against railroad industrialists and robber barons.

    Somewhere along the way, the retelling of this storied episode became separated from the immediate events that precipitated it. The earlier dispute over Bryan, the fight against railroad monopolies, and Ross’s progressive criticisms of the Stanford family fortune rose in prominence as secondary or even co-equal causes for his forced resignation. Ross’s remarks on Chinese immigration, while still acknowledged as a cause and condemned for taking a position that we now recognize today as racist, have lost their original specificity. His original remarks do not appear to have been reprinted in full since the San Francisco newspapers covered the story shortly after Ross’s resignation.

    This missing context is crucial, as they show that Ross went well beyond simply advancing a general economic argument against Chinese immigration, or expressing concern for worker conditions against the anti-labor antics of the railroad industry. Rather, his comments included an elaborate endorsement of racial eugenics and concluded on a bombastic appeal to Aryan nationalism.

    While we should recognize that Ross’s comments reflected a racial sentiment that was more common in his time than our own, Mrs. Stanford’s objections to them were also directly rooted in an objection to their prejudices and a belief that Ross intended them to inflame racial animosity against persons of Asian descent. They were also the unambiguous and immediate instigator of Ross’s dismissal. Other disputes such as the Bryan campaign and general labor agitation only reflect an ongoing history of tensions between Ross and Mrs. Stanford.

    The precipitating event happened on May 7, 1900 when Ross addressed a large public assembly on the “problem” of Chinese immigration. The meeting was overtly racist in character, featuring a succession of speakers who railed against the “Asiatic hordes” of immigrants. They adopted a resolution calling on Congress to expand the Chinese Exclusion Act, and extend its provisions to the Japanese, who allegedly mimicked American customs which “makes them more dangerous as competitors.” The San Francisco Call published a summary of Ross’s speech on May 8, 1900, as follows:

    “[Ross] declared that primarily the Chinese and Japanese are impossible among us because they cannot assimilate with us; they represent a different and an inferior civilization to our own and mean by their presence the degradation of American labor and American life. We demand a protection for the American workmen as well as for American products, the speaker insisted. And should the worst come to the worst it would be better for us if we were to turn our guns upon every vessel bringing Japanese to our shores rather than to permit them to land.”

    Jane Stanford read this summary in the newspaper and became incensed. As documented by historian James C. Mohr, she penned a letter to Stanford University president David Starr Jordan the next day stating that Professor Ross “should now be dismissed.” Mrs. Stanford made it absolutely clear that her grievance with Ross, though preceded by other disputes, derived in this case from his race prejudice. Using his university title, Ross had ventured “out of his sphere, to associate himself with political demagogues of this city, exciting their evil passions, drawing distinctions between man and man, all laborers and equal in the sight God, and literally plays into the hands of the lowest and vilest elements of socialism.” She also expressed concern that Ross’s inflammatory rhetoric could lead to violence, and compared them to an anti-Chinese racial riot in San Francisco some two decades prior that resulted in the racial murders of four Chinese immigrants and left much of the city’s Chinatown section in flames. It too had started as a “labor rally” to call for greater immigration restrictions. As sole trustee of the university, she instructed Jordan to provide Ross with “notice that he will not be re-engaged for the new year.” (Stanford to Jordan, May 9, 1900, Stanford University archives)

    Ross, Jordan, and Stanford engaged in a protracted negotiation over the course of the summer, with Jordan attempting in various degrees to prod Mrs. Stanford to back down from her demand. By early autumn however they had reached an impasse, and Ross prepared for his resignation – as well as an accompanying publicity campaign to draw attention to his cause. He wrote his mother on September 9 about both the state of affairs surrounding his job and the cause of the dispute, saying nothing of Bryan, of free silver, or of his calls for publicly owned transit and utilities. He stated one cause alone, and that was the Asian immigration speech:

    “Last May at Dr. Jordan’s suggestion I spoke at a mass-meeting to protest against Japanese and Chinese coolie immigration. I argued against letting in the coolies and the big Railroad people who want cheap yellow labor induced Mrs. Stanford, who is now very old and crochety, to dismiss me.”

    Ross noted that he was timing the announcement so as to “make a sensation on the Pacific slope” with the press. Other correspondence uncovered by Mohr suggests that Ross intentionally coordinated his public departure for November so as to generate publicity for the release of his forthcoming book, Social Control.

    It is the elusive text of Ross’s speech at the May 7, 1900 rally though that demonstrates the full severity of his racist and eugenic objectives. Several San Francisco newspapers as well as William Jennings Bryan’s newsletter obtained copies of it from Ross in November around the time of his public resignation. A key excerpt of the text appears below, as reported in the San Francisco Chronicle on November 15:

    Ross’s eugenic and racial objectives are a pervasive theme of his remarks, including a dire warning of a coming white “race suicide” in the face of declining birth rates compared to Asian populations. Notably, Ross is the academic who first popularized the overtly white supremacist theory of “race suicide” in a 1901 address to the American Academy of Political Science. That address was one of the many fruits of his highly effective publicity campaign to extol his own martyrdom the previous year. Therefore one of Ross’s most notorious scholarly contributions actually takes its root in the political speech that led to his dismissal from Stanford. Ross concluded his remarks with an equally inflammatory pledge to preserve California for the white race: “California, this latest and loveliest seat of the Aryan race, shall not become, if we can help it, the theater of such a stern wolfish struggle for existence as prevails throughout the Orient.”

    The severity of these remarks alone is cause to revisit Jane Stanford’s stated concerns with Professor Ross’s activities. While her objection was political in nature, the character of Ross’s statements involved a grossly inflammatory racial appeal – even by the standards of the day. Her concern that this appeal could instigate racial violence was not unfounded.

    That much acknowledged, cases such as the Ross episode are noteworthy precisely because they test the limits of academic freedom. While irresponsible, demagogic, and undeniably made in appeal to the basest racial prejudices of his day, the question of whether Ross’s comments warranted the protections of free speech also directly led to the formalization of academic freedom by Dewey, Seligman, Lovejoy and others under the auspices of a new organization, the AAUP.

    It is important to remember this early precedent, as the cause célèbre that birthed the AAUP involved an incident of extremely offensive speech in support of a cause that we may rightly condemn as not only racist but maliciously so. To put it bluntly, Ross advanced an argument for racial eugenics amidst a call to arms on behalf of Aryan nationalism at a populist labor rally in a city with a very recent history of racial violence. The founding members of the AAUP believed that argument was nonetheless a protected exercise of academic freedom, and subsequently built their organization upon its direct fallout. What that entails for the modern AAUP is necessarily a matter of retrospection, albeit one that should be engaged on account of the organization’s own modern relationship to the ongoing academic freedom debate.

    How John Rawls tried to put Democracy in Chains

    Posted By on December 19, 2017

    I’ve spent the past week at the Hoover Institution, researching the early origins of James M. Buchanan and Gordon Tullock’s 1962 book The Calculus of Consent. While working in the papers of W.H. Hutt, South African economist who spent 1966-67 as a visiting professor at Buchanan and Tullock’s research center at the University of Virginia, I came across a fascinating set of documents. They revealed an early 1960s exchange of correspondence and papers between Buchanan/Tullock and the political philosopher John Rawls.

    The complex but overlapping relationship between Buchanan and Rawls is well known. Buchanan wrote of it at length in his later works. Rawls similarly referenced the Calculus in his own modern classic,  A Theory of Justice (1971). Rawls was, of course, a famously left-liberal thinker while Buchanan and Tullock are typically associated with the libertarian right.

    The exchange in the Hutt papers is distinctive though for its early date. Rawls sent Buchanan and Tullock an early draft of his paper “Constitutional Liberty and the Concept of Justice,” along with comments acknowledging the overlap between the economists’ arguments and his own. The corresponding letter is undated, but textual clues indicate it was written between the date on the draft of Rawls’ paper (December 28, 1961) and its appearance in print as a chapter in Nomos VI, which came out in early 1963. This puts the letter in the immediate vicinity of the Calculus being published in 1962.

    Rawls’ accompanying note contains a fascinating passage where he further develops the relationship between majority rule and justice, essentially concluding that there is none. Rather, majority rule is simply a mechanism to be evaluated by external criteria, and comparatively to other rules:

    “In the second part of the Chapter on Liberty I discuss majority rule and adopt a view in many ways similar to yours, but not I think in all respects. In any case, I agree that majority-rule is just a rule to be adopted on rational grounds like any other, given experience with it etc. Majority rule as a principle of justice I agree is absurd. On my view the principles of justice put constraints on the constitution, & on all political majorities; and majority rule is rational only where it can be supposed that majorities will limit themselves by the principles of justice, and where reasonable differences of opinion may exist and where the disadvantages & injustices of other rules might be worse.”

    How Hutt ended up with this letter is itself a mystery, but most likely relates to his own work on the Calculus by applying its principles to the problem of racial discrimination.

    In any case, Rawls’ repudiation of a claimed inherent link between majoritarianism and justice makes for an interesting juxtaposition to other recent characterizations of Buchanan and Tullock’s work. Based on this evidence and those same characterizations, it would appear that John Rawls approved of the notion of putting “democracy in chains.”


    About MacLean and the matter of John C. Calhoun

    Posted By on December 10, 2017

    Several months ago when I first entered the discussion about Nancy MacLean’s book Democracy in Chains, I called attention to her misuse of historical evidence to write pro-slavery theorist John C. Calhoun into the intellectual lineage of economist James M. Buchanan. MacLean’s claim struck me as odd at the time, because Calhoun’s name does not appear anywhere in Buchanan’s published works, which extend across more than six decades.

    A number of MacLean’s defenders at the time excused her on this point, suggesting that she had somehow discerned and extracted Calhoun’s supposed influence on Buchanan by “reading” it through the unwritten subtext of his works and the surrounding context of the 1950s desegregation debate. This explanation veered suspiciously close to simply making things up, as is MacLean’s apparent penchant. But nonetheless, she has persisted in aggressively pushing the Calhoun claim and a number of her supportive readers have credulously repeated it as if it were now a matter of established fact.

    We can now affirmatively put that claim to rest, showing that it is nothing more than a figment of MacLean’s own imagination. We now know this because Buchanan’s co-author Gordon Tullock specifically addressed the relationship between Calhoun and the Calculus of Consent in a 1975 commentary essay in the American Political Science Review. As Tullock bluntly put it, that relationship was non-existent. As he stated, he had not even read so much as a single page of Calhoun. He further stated that the same was the case of James M. Buchanan, even as Buchanan had a greater interest than himself in other forms of political philosophy.

    The pertinent excerpt is below:

    Note that Tullock also went into the intellectual history of the economic thinkers who most directly influenced the unanimity rule concept in the Calculus of Consent: Vilfredo Pareto and Knut Wicksell. As he explained, there was no reason to believe that either of them were familiar with Calhoun. Therefore Calhoun could not have even been an indirect influence upon Buchanan and Tullock’s own work.

    As a brief aside on this subject, let me also suggest a further explanation for how MacLean came to her unsupported thesis linking Calhoun to Buchanan. Quite simply, she already had her narrative mapped out before she even knew Buchanan’s name. She then discovered Buchanan and attempted to shoehorn him into that narrative, despite having no evidence to support it. The following excerpt derives from a 2007 working paper that MacLean presented when she taught at Northwestern University. Note that Calhoun is identified as the intellectual ancestor of a nefarious “movement” with the very same harmful motives she alleges in her book. Except instead of Buchanan, the guilty party consists simply of nameless “neo-liberals.”

    Quite simply, MacLean retrofitted Buchanan’s name into an unfounded presupposition that she already held. And now as we’ve seen, not only did she lack evidence linking Buchanan to Calhoun – that link is also directly refuted by Buchanan’s co-author Gordon Tullock.

    In short, MacLean’s entire thesis regarding Calhoun is specious and fabricated nonsense.

    It is time for Viking Books to correct or retract Democracy in Chains.

    Why Universities have shifted to the political left in the past 20 years

    Posted By on December 3, 2017

    University faculty ranks have long been thought of as bastions of the political left, and with good reason: sizable majorities of professors self-identify as modern liberal or left-leaning in their politics. This trend dates back to the first comprehensive faculty survey on political leanings in 1969.

    The leftward tilt of college professors used to be more static than it is today. Between 1969 and 1998, self-identified liberals consistently made up between 40 and 45% of faculty in American colleges and universities, but they did not waver much from this range. While these totals outnumbered conservatives and moderates, all three groups were relatively stable in size. In fact, several academic studies published between the 1970s and 1990s pointed to this stable pattern as evidence that long-running conservative grievances about “faculty bias” were overblown.

    Then something changed.

    Starting in the mid-1990s, the number of college professors who self-identified as conservatives or even moderates began to rapidly decline. In 1998, moderates made up 37% of the academy and conservatives made up 18%. In the most recent survey from 2013, moderates and conservatives dropped to 27% and 12% respectively. Meanwhile, self-identified liberals exploded in number. They sat at 45% in 1998, and have grown to 60% in the most recent survey. That’s a shift of over 15 percentage points away from conservatives and moderates and toward self-identified liberals.

    The chart below depicts this trend using the Higher Education Research Institute (HERI) faculty survey (1989-present) combined with its predecessor, the Carnegie Commission Higher Education Survey (1969, 1975, 1984). The once-disputed leftward shift in academia is now unambiguously attested in recent data.

    But why has this shift occurred after several decades of stability? By turning to other measures we may catch a glimpse of the answer. It has to do with faculty composition, as well acute leftward biases in specific academic disciplines.

    As a recent article by political scientist Sam Abrams documented, some disciplines skew substantially further to the left than academia as a whole. While roughly 60% of all professors self-identify as liberals, that number tops 80% among English professors. History, political science, fine arts, and the other humanities and social sciences are all substantially more liberal than the academy as a whole. They have also shifted further left with the overall trends seen in the chart above.

    The leftward skew of the humanities, fine arts, and social sciences matters because these disciplines have also exhibited faster growth in faculty employment than the rest of academia, despite having notoriously saturated job markets in many cases (note that job market saturation does not mean the total number of humanities jobs are declining – it means that the number humanities job seekers continues to outpace the growth in humanities jobs).

    We can see this pattern in a second chart from the American Academy of Arts and Sciences, depicting the percentage of university faculty who work in each academic field:

    By taking a closer look at the trend lines, we may see how left-leaning areas of the academy have grown faster than the rest in the past decade. For our purposes, set aside the dashed line representing “Health Sciences,” which is actually a result of the boom in nursing and other healthcare professional degrees in recent years.

    Among the remaining disciplines, the biggest gains in faculty shares are fine arts (1.82 percentage points), humanities (1.01), and social sciences (0.83). By comparison, the two biggest declines in percentage points were the natural sciences (-1.79) and engineering (-0.87).

    The pattern becomes even more visible when we group together all of the disciplines that skew further to the left of the academy. Humanities, fine arts, and social science faculty all lean further left than academia as a whole. Jointly, they posted a 3.66 percentage point gain in their overall share of faculty. When combined together, the much-discussed STEM disciplines lost 2.65 percentage points from their early 2000s faculty share over the last decade.  Note that while self-identified liberals are still the largest group within the STEM discipline faculty as well, they tend to be less left-leaning than faculty as a whole.

    What this basically means is that the most politically skewed disciplines – the humanities, fine arts, and social sciences – have grown their faculty at a faster rate than other less politicized disciplines. This skewed growth pattern has increased their footprint on campus relative to other faculty, and it has done so at a time of an additional leftward shift in their own ranks. The cumulative effect is to pull the university system as a whole even further to the left, yielding the overall pattern we have seen in the faculty survey at a time when the American public at large has maintained a fairly stable liberal/conservative divide.