Sunday, November 8, 2015

Free Market Dogmatism Still Going Strong at the University of Chicago

On Friday I attended Thomas Piketty’s two appearances at the University of Chicago, my old stomping grounds, along with my good friend (and now, my coauthor!) Bernard Weisberger. I was eager to see the commotion in person because I was curious whether there would be any discernible acknowledgement on the part of his interlocutors of the huge challenges that Capital in the 21st Century poses to their worldview and their research agendas. And there was, if only in the vehemence with which they denounced Piketty’s arguments and insisted on shifting the terms of the conversation. In my experience, you can always tell when an academic feels his expertise has been exposed to public questioning and found wanting by the frequency of absurd pronouncements stated with ever-increasing confidence as they veer further and further from reality.

What follows is my account of the first event, a discussion between Piketty and two notable economists, moderated by a third. The second event, where Piketty more or less had the floor to himself to present Capital in the 21st Century, was by far the more substantive. Interestingly, the crowd at the first was tilted toward graduate students and faculty, in which context all parties articulated long-held positions and the audience was there to express solidarity with one or the other. The crowd was more undergraduate-heavy at the second event, where Piketty’s presentation presumed a deeper engagement with the issues at hand, and where the atmosphere was much closer to the ideal of free, high-level academic inquiry on an important public matter.

Piketty’s interlocutors at the first event were Professors Kevin Murphy and Steven Durlauf, with Jim Heckman acting as moderator. (He opened proceedings by announcing, in classic Heckman self-contradictory style: “It’s a debate! It’s not a debate—it’s a discussion. It might turn into a debate.”) I admire Professor Durlauf’s original research as well as most of his informal commentary, though not his review of Capital in the 21st Century in the Journal of Political Economy. I do not admire Professor Murphy. As was evident from his comments in Friday’s session, he worships free market economics as a religion, without any sense that its truth is contingent on reality. He further thinks that his 1992 paper with Lawrence Katz, which tried to explain the dynamics of the college wage premium in the 1970s and 1980s with reference to the supply and demand for skilled labor in the form of workers with a college degree, constitutes the final word on inequality in its methodological approach, its empirical conclusions, and its policy implications. Even if you consider that paper to have been correct in all of its particulars regarding the identification and interpretation of the fact pattern it aimed at (college wage premiums declined in the 1970s and increased in the 1980s), it simply doesn’t rise to the inequality-conversation-ending status Murphy believes it to have. Its model fails at explaining the labor market outcomes that have transpired since, especially since 2000—a period in which inequality increased vastly more than it did during the time Katz and Murphy covered, while the college wage premium only changed slightly—and that only because life got worse for non-college-educated workers, hardly evidence that the dynamic new economy provides ample benefits to those with the educational credentials to take advantage of them. The paper also begs the question at a fundamental level, a point Murphy reiterated at the UChicago session, when he declared that “technology causes growth, and in response to technology, we invested in physical and human capital.” Skills-biased technical change: the Ghost in the Free Market Economics Machine.

Beyond the empirical and methodological vacuity of that paper, Murphy’s way of presenting it is offensive in the extreme—and he tacitly aimed his arrows at Piketty. Murphy began with “let’s just do a little economics,” which he distinguished from accounting by saying that economists distinguish prices from quantities in calculating expenditure shares, a subtle dig at Piketty’s World Top Incomes Database (a far more important contribution to economics than anything Murphy’s done). To Murphy, “doing economics” means moving supply and demand curves around to tell a nice little story about how The Market solves every problem as long as it’s left alone.

Getting back to the “discussion,” there was no one theme. Piketty started things off by claiming that the received wisdom (at least among economists) for why inequality has increased, globalization and skill-biased technical change, simply don’t explain the phenomena very well. Neither can explain the rise of the top 1%, nor can they explain the international variation in the extent of tail inequality. Piketty did credit the role of educational exclusion in closing off access to the most elite precincts of the economy, as shown by the new Chetty, Saez, et al findings on the extent to which top universities draw their undergraduate students from rich households. But he continued on to a discussion of how the Piketty, Saez, and Stantcheva (2014) findings on wage bargaining and top income shares can’t be squared with a marginal-productivity story of wage-setting. He mentioned norms of corporate governance shifting in favor of managers and owners (the subject of the new excellent report from J.W. Mason and the worthy fellows and fellowesses of the Roosevelt Institute) by way of explaining tail inequality, as well as erosion of unions and the minimum wage as explanations for stagnant or falling wages at the bottom and middle of the distribution. He closed with what I consider a profound restatement of why Capital in the 21st Century is such an important book:
The gap between [the] official discourse and what’s actually going on is enormous. The tendency is for the winner to justify inequality with meritocracy. It’s important to put these claims up for public discussion.

Durlauf spoke next, and at first mostly inoffensively. He shifted the terms of the debate to focus on disadvantage within the US, something he knows a great deal about but which simultaneously takes the focus off Piketty-style tail inequality or the possibility that the two might be linked. He said, quite reasonably, that the key mechanism of inequality is segregation, because it translates individual inequality into entrenched deprivation, and that its policy implications are therefore to foster integration in a variety of contexts. That is all well and good (great, even), but in a Pikettian nightmare of tail inequality and patrimonial capitalism, all that matters is whether you know billionaires and are good at kissing their asses, and one thing we know about billionaires is that the people they like best are those most like themselves. No amount of policy-fostered integration matters in comparison to the social reality of that world. We’ll all be living together—in a slum.

Murphy’s presentation was where the wheels came off, intellectually speaking. He declared “we can do a lot with a little bit of economics,” then proceeded to do nothing at all of substance by regurgitating his 1992 paper. But according to Murphy, “that theory has done an amazing job,” including a cryptic statement about how it explains the rise of tail inequality “if you extrapolate,” whatever that means. Murphy closed by declaring that providing for adequate supply of skilled workers would benefit not only those workers who gain skills, but also those who don’t, since it would scarcify their labor supply—a direct recapitulation of the Hershbein-Kearney-Summers nonsense I addressed with John Schmitt here.

In the discussion that followed, Murphy stepped forward once again to declare that the economy’s “natural supply response of supplying capital” will help workers by reducing the capital share and increasing their productivity. This is a direct restatement of John Bates Clark’s old fairy tale of the rate of return on capital equilibrating in the long run to the benefit of all, so therefore we should never be so injudicious as to get in the way of that mechanism by taxing capital or those whose savings feed into it—i.e. the rich. Murphy’s use of the words “natural” and “technology,” which passed by the panel (and, seemingly, the audience) without comment, is notable here. Also notable is the fact that Durlauf asserted in his JPE review of C21 that no one thinks like Clark anymore, with his quasi-moralistic view of the efficient functioning of capital formation and the adjustment of its rate of return. Unfortunately, Durlauf’s empirical prediction was falsified by Murphy right there on that stage.

Later, Murphy added that in the absence of better education, “The march of technology over time means there’s little for someone with no human capital to do.” Astoundingly, that flagrantly question-begging just-so story garnered a round of applause from the audience, reminding me of the debate I attended in my first year of graduate school about the founding of what was then the Milton Friedman Institute, now the Becker Friedman Institute that hosted Friday’s event. On that occasion, various campus lefties from around the university protested inconsequentially against the institute’s namesake, and the presentation by economics heavyweights Heckman and Lars Hansen garnered competitive applause from the economics partisans present—not exactly a sign of reasoned academic discourse.

Meanwhile, Durlauf raised the oft-heard point that America just has a different ideology when it comes to progressive taxation, which is why expanding education is the route forward. (Durlauf even made the bizarre and obviously factually false assertion that “America never had a Socialist Party.” Eugene V. Debs got one in twelve votes in the 1912 election, as my friend Bernie, noted historian of Progressivism, pointed out afterward.) Fact-based and historically-minded as I am, I considered Piketty’s rebuttal devastating: that progressive taxation was invented in America and that it flourished here as a complement to free and equal quality public education, not a substitute. Together, the two did not destroy capitalism. Quite the contrary: the period of their efflorescence, complete with confiscatory estate taxation, saw the highest aggregate and per-capita growth across the income distribution of any time in American history.

Then things got weird. Durlauf replied by asserting that he didn’t mean to say that Americans had never been moved to oppose inequality, but that what mattered was their perception of its source: whether justified by merit, as in the case of Bill Gates, or extracted through monopolization, as with John D. Rockefeller. At that, Piketty quipped that Bill Gates certainly agrees. But Gates is rich because he has a government-mandated monopoly to sell operating systems and business software used almost universally, while Rockefeller was rich because of his interlocking monopolies on oil, railroads, and banking, and the fortunes amassed by both men endure because of subsequent capital accumulation. One of the best parts of Capital in the 21st Century is where he asserts (and, to my mind, carries) the argument that there’s no moral hierarchy of wealth.

Following a discussion of social mobility as the empirical analog of meritocracy, in which Murphy had his one good argument of the session when he pointed out, contra Chetty and friends, that there’s no reason to believe that a meritocratic society would generate perfect intergenerational rank-rank mobility, discussion shifted to policy. Piketty took the opportunity to return to the ghost of John Bates Clark by saying that we need a Plan B in case the “natural” capital supply response isn’t forthcoming or that it works against rather than in favor of workers—namely progressive taxation of wealth. “Otherwise, it’s a bit magical.” That set Murphy off. He interjected that he never meant to say problems of inequality would solve themselves. But, on the other hand, progressive taxation is the worst thing you could do—shut down that supply response by destroying the incentive to save to form physical capital or to get educated to form human capital. Referring to some version of Mitt Romney’s old friends, the Makers and the Takers, Murphy quipped “we have to keep people engaged”—presumably meaning that without the goad of vast inequality to spur them to action, the left-behind might just drift out of the economy altogether. That roused his followers in the audience once again. Continuing on his lifelong quest to preach the gospel of the First Welfare Theorem to the expectant congregation gathered at the camp revival, Murphy continued [referring to the poor and marginalized] “People aren’t perfect. They don’t all look out for their own interest but most do.”

The intellectual and doctrinal muddle is almost too tangled to tease out here, but the idea seems to be that the poor, benighted though they are, will adopt the morally correct position of looking out for their own interest by acquiring an education, so long as the incentive to do so is preserved by avoiding progressive taxation. Usually the fallacy in the moral philosophy of economics, for those who partake in such things, is to argue that whatever reality exists is for the best, a classic Panglossian is/ought conflation. In this case, though, the “ought” is a priori: people should be selfish. For that reason, they probably will be, so long as the status quo is maintained as an instructive lesson in the disaster befalling anyone not born rich. At least Murphy disagrees with the late Franklin Giddings, the Columbia professor who said in 1893 that following 20 years of populist agitation, American farmers hadn’t managed to improve their economic position and therefore must have only themselves to blame. Murphy, on the other hand, holds out hope of future bootstrap-pulling-up behavior so long as we avoid the pitfall of directly addressing inequality as we find it. Piketty himself said it best: “The idea that we need to keep inequality to preserve incentives is just not consistent with reality.”

Durlauf made a final, inscrutable point in this debate by saying that we should directly address the harms caused by inequality, by which he was referring to capture of the political system by the wealthy—as opposed to dealing with inequality itself, through progressive taxation, presumably. He asserted that the former approach is what motivated Progressive-era reformers, a remark that elicited an audible scoff from Bernie.

That ends my account of the first session of the day—disappointing in the extreme, an opportunity to relive the worst aspects of my experience in graduate school. [Thankfully, a tiny minority of my time was spent listening to Kevin Murphy lecture, but I honestly was scarred by and still resent the nights I spent working very late to gin up the answers he wanted for his flagrantly ideological problem sets.] The saving grace of Friday’s event was of course Bernie’s company—to which I fled more than once in intellectual desperation during my PhD.

In Durlauf’s JPE review, he mentions that Charles Murray (and specifically his book The Bell Curve) has no influence in academia, even though it retains a totally undeserved hold on public debate. Durlauf is implying that Capital in the 21st Century merits the same fate, which is a totally ridiculous and offensive comparison. In any case, Murray came to UChicago on his book tour for Coming Apart, and that event garnered no attention whatsoever from the economics establishment (except for me—I wrote it up at the time on my Facebook wall, an account which can be read here.) Indeed, one of the aspects of Washington economics discourse that has surprised me the most is the relatively high stature accorded to Murray, Brad Wilcox, and their “pro-family” nonsense by elites here (notwithstanding the yeoman’s work of the great, and tireless, Phil Cohen, showing what charlatans they are.) The right wing of academic economics doesn’t give Murray and Wilcox the time of day, but apparently it worships the ground Kevin Murphy walks on, despite the fact that Murphy has no greater claim on our attention than Murray. They are both fact-free ideological axe-grinders whispering comforting, apparently-superficially-plausible nonsense to the powerful.


  1. great post

    calling Katz & Murphy '92 empirically and methodologically vacuous is blasphemy though. it may not be the final word on inequality, but just because one of its authors apparently believes it to be doesn't undo the fact that it's an exemplary piece of empirical social science and a crucial paper in the history of the study wage inequality. (although i would love to read a future post on why it begs the question!) also, just because a simple two-skill SBTC model failed to explain wage inequality doesn't mean SBTC is a ghost--the subtler routine/non-routine cognitive/manual model of Autor Levy & Murname 2003 is incredibly empirically successful, yet only a slight tweak on the original SBTC hypothesis.

  2. Thanks. As for a more complete discussion of what's wrong with Katz and Murphy--would be happy to do a blog post on that at some point. I don't agree Autor Levy & Murnane is "incredibly empirical successful"--see the linked Beaudry, Green and Sand paper, and this: http://www.epi.org/publication/technology-inequality-dont-blame-the-robots/

  3. for empirical success, see figure 3 of autor's JEP (http://economics.mit.edu/files/10865.) it could always be something else, but when the same pattern appears in many countries with different institution, we should update our prior in favour of a technological explanation (doesn't mean we can't have an institutional response--just that we have to think about it carefully). that, to me, is empirical economic success. that said, i don't think we have to "choose" between autor, murphy & co and piketty. the former focus on within bottom 99% inequality, the latter has brought inequality between bottom 99% and 1% back into the discussion. both are welcome.

    i'm very familiar with those Beaudry Green & Sand papers. when i was a grad student i recreated the wage and employment profiles from their aer p&p using canadian data and found a similar trend. evidence fits their hypothesis of a bust in demand for cognitive tasks, but as they admit in the earlier & longer nber paper other potential explanations exist and theory does not tell us how to distinguish between them. autor JEP suggests it was just a correction after crash tech bubble, which is consistent with employment polarization being "back" post-2007 (see figure 5 http://economics.mit.edu/files/10865). it's important to be extra skeptical of technology-based explanations for inequality, and the economics establishment is often too smug with them. it's wise of you to kick back against that. i just don't want a piece top shelf empirical work dragged through the mud because one of its authors is a bumbling ideologue!

    1. That last point is fair enough. If K&M 92 just existed out there as an explanation for the dynamics of the college wage premium in the 70s and 80s I would be less exercised about it--though I still think the source of "skills-biased technical change" is a bit mysterious.

  4. The real begging the question issue here, is why meritocracy is be all and end all. It isn't the right goal. See Chris Dillow (Stumbling and Mumbling).

    I do agree that there are potential issues regarding the incentive to invest, but surely Murphy has noticed student debt. The importance of uncertainty and risk (you only get one chance - it is not like you can make a portfolio investment in human capital) are crucial here (making the argument more complicated of course, which Chicago professors avoid like the plague).

  5. This comment has been removed by the author.

  6. This comment has been removed by the author.

  7. This comment has been removed by a blog administrator.