Never Let a Serious Crisis Go to Waste

by Philip Mirowski

  One recent instance of EMH denial that has attracted the endorsement of reformers such as Adair Turner in Britain and Ezra Klein in the United States is the already-mentioned work of the Harvard economist Andrei Shleifer.65 Shleifer is a neoliberal who passes as a left-liberal economist in America, in the Lawrence Summers mold. Shleifer’s model initially “seems” to call into question the validity of certain classes of dubious artificial derivatives by suggesting someone is being misled by the market; but as usual, the devil is in the details. Shleifer combines some “behavioral” themes with a standard neoclassical model to blame the crisis on the investors, who unaccountably ignore the extreme tail risk of newly invented derivatives when they are minted. In other words, the crisis happens because the suitably adjusted agents are blind to its possibility. (Rabbit in hat; rabbit out.) Actually, the EMH is not “refuted” so much as reinforced in the model, since “the market” still emits the correct signals (which really do come out of nowhere in the mathematics); as in most neoliberal scenarios, the crash is the fault of the victims beset with “local thinking” (Shleifer’s terminology) rushing to dump their wonky assets all at the same time. Everything that actually happened, from the waves of neoliberal deregulation to the ratings hand jive to the whitewash of crude evasion of existing rules as financial “innovation” to accounting travesties to outright fraud leave no trace in the model: when in doubt, blame the victims. This is trumpeted to the world as neoclassical economics getting more true to the facts on the ground.

  In the Odyssey, Proteus assumed a plethora of shapes to escape Menelaus; in the EMF, “information” had to be gripped tight by neoclassical theory, because it kept squirming and changing shape whenever anyone tried to confine it within the framework of a standard neoclassical model. Few have been sensitive enough to the struggle to attend to its twists and turns, but for present purposes it will be sufficient that three major categories of cages to tame the beast have been information portrayed as “thing” or object, information reified as inductive index, and information as the input to symbolic computation.66 For numerous considerations here bypassed, they cannot in general be reduced one to another. The reason this matters to journalists’ convictions that the crisis has invalidated the EMH is that the detractors mostly conform to the literature that treats information like a commodity, whereas the defenders repulse them from battlements of legitimation built largely from information as an inductive index. This may seem a distinction that only a pedant could love, but once clarified it goes a long way to demonstrating that the crisis will never induce the majority of neoclassical economists to give up on the EMH.

  The most prominent standard-bearer for the denial of the Kenntnisnahme über alles EMH has been Joseph Stiglitz. Here it is important to acknowledge that Stiglitz is treated by the mainstream media as a stalwart of the American left because he has repeatedly taken political positions that have not ingratiated him with those in power; he often has been steadfast in his pessimistic evaluations of the crisis, when all the journalists wanted to hear was how the crisis was done, dusted, and under control. In stark contrast to most of the figures encountered in this history, he has repeatedly gone on record stating that economists should bear some responsibility for the crisis. He even said complimentary things about French demonstrations in October 2010 against raising the age of retirement. By these lights, Stiglitz has been an exemplary contrarian economist. Nonetheless, Stiglitz has simultaneously been a major defender of neoclassical economics, suggesting that the EMH is not all that central to the core doctrines of orthodoxy:

  Normally, most markets work reasonably well on their own. But this is not true when there are externalities . . . The markets failed, and the presence of large externalities is one of the reasons. But there are others. I have repeatedly noted the misalignment of incentives—bank officers’ incentives were not consistent with the objectives of other stakeholders and society more generally. Buyers of assets also have imperfect information . . . The disaster that grew from these flawed financial incentives can be, to us economists, somewhat comforting: our models predicted that there would be excessive risk-taking and shortsighted behavior . . . In the end, economic theory was vindicated.67

  It can sometimes be hard for outsiders to detect just where Stiglitz is coming from, since in the course of the very same lecture he can lead with a rousing call to battle like “There ought to be a crisis in economics” and yet follow with the oxymoronic “fortunately, we don’t need to rewrite the textbooks.”68 Stiglitz has been known to make very insightful comments about how to judge crisis narratives, such as the observation that the crisis was not the consequence of “shocks” impinging on the system, but rather was man-made; the stretch comes in understanding how one was supposed to know this from the standard neoclassical model.69

  This is what Krugman has called “flaws-and-frictions” economics, and it propounds the bromide that “we already had models that told us the crisis was coming.” It follows that outsiders should ask: So why weren’t these models well represented in macro or micro textbooks and graduate pedagogy? Stiglitz is fully aware there exists a tradition of oxymoronic “New Keynesianism” that reprised a boring old pre-Keynesian story of sticky wages and prices in a neoclassical equilibrium, but he suggests there exists something else on offer more compelling. In Stiglitz’s case, there is a special caveat: the models he has in mind are found mostly in his own previous publications. While there could be no academic prohibition against tooting your own horn, there is something less than compelling about claiming a grand generality for some idiosyncratic models where the novelty quotient is distinctly low. While Stiglitz has certainly earned his Nobel, he has not effectively staunched the intellectual trend of treating markets as prodigious information processors; nor has he provided a knock-down refutation of the EMH. This has led to the distressing spectacle of Stiglitz, the great hope of the “legitimate left,” openly defending the neoclassical approach to the crisis, while not really changing it all that much.

  Stiglitz has admitted that his mission all along was to undermine free-market fundamentalism from within:

  [I]t seemed to me the most effective way of attacking the paradigm was to keep within the standard framework as much as possible . . .While there is a single way in which information is perfect, there are an infinite number of ways that information can be imperfect. One of the keys to success was formulating simple models in which the set of relevant information could be fully specified . . . the use of highly simplified models to help clarify thinking about quite complicated matters.70

  The way he sought to do this is to produce little stripped-down models that maximize standard utility or production functions, with a glitch or two inserted up front in the setup. He has been especially partial to portraying “information” as a concrete thing to be purchased, and “risk” as standard density function with known parameters. There is no canonical Stiglitz “general model,” but rather a number of specialized dedicated exercises, one for each flaw and/or friction explored. Macroeconomics then simply becomes microeconomics with the subscripts dropped. This distinguishes Stiglitz from the small cadre of researchers in the third section below, who are convinced that this “representative agent” trick does not constitute serious macroeconomic theory.71

  In Stiglitz’s academic writings, he stakes his claim to have refuted the EMH primarily on two papers, one co-authored with Sanford Grossman in 1980, the other with Bruce Greenwald in 1993.72 The takeaway lesson of the first was summarized in his Nobel lecture:

  When there is no noise, prices convey all information, and there is no incentive to purchase information. But if everybody is uninformed, it clearly pays some individual to become informed. Thus, there does not exist a competitive equilibrium.73

  The second is proffered as the fundamental cause of the crisis:

  It perceives the key market failures to be not just in the labor market, but also in financial markets. Because contracts are not appropriately indexed
, alterations in economic circumstances can cause a rash of bankruptcies, and fear of bankruptcy contributes to the freezing of credit markets. The resulting economic disruption affects both aggregate demand and aggregate supply, and it’s not easy to recover from this—one reason that my prognosis for the economy in the short term is so gloomy.74

  Both of his crucial “findings” are in fact based upon very narrow versions of what is a much more diversified neoclassical orthodoxy. It would indeed have been noteworthy if Stiglitz or his co-workers had provided a general impossibility theorem, say, along the lines of Gödel’s incompleteness theorem or Turing’s computability theorem, but Stiglitz has explicitly rejected working with full Walrasian general equilibrium, or Chicago’s resort to transactions costs, and doesn’t seriously consider the game theorists’ versions of strategic cognition. Indeed, it seems a rather heroic task to derive any blanket general propositions from any one of his individual toy models. Stiglitz himself admits this when he is not engaged in wholesale promotion of his information program.75

  Take, for instance, the famous Grossman-Stiglitz model. The text starts out by positing information as a commodity that needs to be arbitraged, but claims in a footnote that the model of knowledge therein is tantamount to the portrayal of information as inductive index, which is not strictly true, and then defines its idiosyncratic notion of “equilibrium” as equivalence of plain vanilla rational expected utilities of informed and uninformed agents. Of course, “for simplicity” all the agents are posited identical; how this is supposed to relate to any vernacular notions of divergences in knowledge is something most economists have never been poised to address. Many economists of a different political persuasion simply ignored the model, because they deemed that Stiglitz was not taking into account their (inductive, computational) version of “information.” When Grossman offered his own interpretation of their joint effort, he took the position that the rational expectations model was identical to the approach in Hayek, that “when the efficient markets hypothesis is true and information is costly, competitive markets break down,” and that “We are attempting to redefine the Efficient Markets notion, not destroy it.”76 That seems closer to the median interpretation of Stiglitz’s work in the profession as a whole.

  Perhaps the most distressing aspect of Stiglitz’s designated models that he believes starkly refuted neoliberalism has been that, when you really take the trouble to understand them, they end up having nothing cogent to say about the current crisis whatsoever. Start with Grossman and Stiglitz.77 The problems with the financial system in 2007 had nothing to do with participants lacking correct incentives to purchase enough “information” that would have revealed the dodgy nature of the CDOs and other baroque assets which clogged the balance sheets of the financial sector. Rather, the reams of information they did purchase, from ratings agency evaluations to accounting audits to investment advice, was all deeply corrupted by being consciously skewed to mislead hapless clients and evade the letter of the law. Perhaps the “information” was corrupted by the mere fact of being bought and sold. Since Stiglitz never comes within hailing distance of confronting epistemology in any of his models—he disdains philosophy as much as the next neoclassical economist—he never really deals with matters of truth and falsehood. Agents are just machines buying unproblematic lumps of information (or not).

  Stiglitz often mentions the complaints of heterodox economists (mostly without citing them), but curiously, he seems to have no appreciation for the extent to which no amount of mathematical tweaks could adequately reconcile them with any version of neoclassical theory. He concedes that non-ergodicity pervades real-world random variables (ignoring Paul Davidson), but offers no insights as to how they might be shoehorned into a neoclassical model. He allows that the Sonnenschein-Mantel-Debreu results banish the representative agent, but neglects to point out that they also render all aggregation (and hence all macroeconomics) under neoclassical general equilibrium groundless. More tellingly, he admits that the mere fact that prices for the “same” goods might be set in structurally different market formats would by itself account for destabilizing price dynamics; but passes over in silence the fact that this would delegitimize the entire neoclassical approach to pricing and risk, including his own work.78

  And worse, the “market failure” that he repeatedly diagnoses has nothing to do with what people mean by “failure” in the vernacular. Stiglitz (and Krugman and Solow and other guest stars in the New York Review of Books) identify “market failure” with not realizing the full measure of utility that might have occurred in the standard neoclassical model—this is called Pareto optimality in the trade—and exists in an imaginary universe utterly devoid of markets freezing up and the implosion of the assignment of credible prices across the board. Likewise, the Stiglitz-Greenwald paper has nothing whatsoever to do with the collapse of the financial sector in 2008. Using their own words, “we showed that there were essentially always simple government interventions that could make some individuals better off without making anyone worse off. The intuition behind our result was that whenever information was imperfect, actions generated externality-like effects.”79 Stiglitz persistently conflates “welfare loss” with system-wide economic failure and market breakdown: this travesty stands in stark contrast to the model-free occasions wherein Stiglitz perceptively analyzes the inconsistencies of concrete practices in real world institutions, linking them to palpable dire outcomes. Pareto optimality was the last thing one needed to consult in trying to understand the utter confusion and disarray accompanying the mad improvisations at the Fed and the congressional TARP appropriation in the depths of the crisis; it certainly would be impotent to clarify the types of “government intervention” required to stem the collapse. Incredibly, the Greenwald-Stiglitz model doesn’t even explicitly have any money in it, even though one core phenomenon of the 2008 meltdown was a financial credit crisis. Instead, their model identifies the central weakness of the capitalist system as a rational contraction of investment on the part of firms, not financial system collapse.80

  Stiglitz repeatedly pronounces last rites over the EMH, but has little sway on the profession because he cannot see what is sauce for the goose is sauce for the gander. “The Chicago School and its disciples wanted to believe that the market for information was like any other market.” Yet, that is the fundamental initial premise of his own models. “The widespread belief in the EMH played a role in the Federal Reserve’s failure. If that hypothesis were true, then there were no such thing as bubbles.” But this just displays a deficiency of hermeneutic attention. Both the Fed and the profession can accept that the EMH, properly understood, and bubbles are entirely compatible—you just won’t know you are in one till it bursts. And paraphrasing Bill Clinton, it all depends what you mean by “bubble.” Yet Stiglitz never really repudiates the neoliberal doctrine of the Marketplace of Ideas. “The price mechanism is at the core of the market process of gathering, processing and transmitting information.” There is nothing autodestructive about the Marketplace of Ideas. But then, who in the elite of the orthodox economics profession ever thought otherwise?81

  The endless quest to delete the EMH from the Ten Commandments of Neoclassicism almost constitutes the definition of “empty gesture” within orthodox economics.

  3) Abandon the DSGE Model

  A third reaction to the crisis is to refrain from indictment of the global orthodoxy, and instead suggest that since the crisis was eminently a “macroeconomic” event, the onus for failure must be narrowly restricted to that subset of the profession tasked with study of the macroeconomy; and furthermore, the correct response is to simply jettison the paradigmatic model found in contemporary macroeconomic textbooks, the so-called Dynamic Stochastic General Equilibrium (DSGE) model. The crisis, for this cadre, does not portend anything remotely like the “death of economics,” but just a garden-variety “model failure”: so replace the model. Now, I can imagine my audience rolling their eyes—even
those willing to put up with the modicum of technical issues raised so far are not going to countenance a tedious discussion of a specific mathematical model, no matter how crucial to the self-image of the economics profession. And it is true that there is almost no commentary in the general press on the DSGE model, compared with breathless denunciations of “rational economic man” and the EMH. But this option does even more directly call into question the commonplace notion that economists can learn from their mistakes.

  This is exemplified by an event in 2010 that was literally unprecedented in the history of economic thought in America. Congressional testimony is regularly convened on all manner of issues of applied economics, and economists are regularly enjoined to testify. But never before, to my knowledge, has an entire session been convened to hold public hearings on criticism of a mathematical model produced by economic theory, not on its purported applications. Yet, on July 20, 2010, a kind of Star Chamber was convened to pillory the DSGE model.82

  This is how the committee staff described the DSGE model for a lay audience:

  The dominant macro model has for some time been the Dynamic Stochastic General Equilibrium model, or DSGE, whose name points to some of its outstanding characteristics. “General” indicates that the model includes all markets in the economy. “Equilibrium” points to the assumptions that supply and demand balance out rapidly and unfailingly, and that competition reigns in markets that are undisturbed by shortages, surpluses, or involuntary unemployment. “Dynamic” means that the model looks at the economy over time rather than at an isolated moment. “Stochastic” corresponds to a specific type of manageable randomness built into the model that allows for unexpected events, such as oil shocks or technological changes, but assumes that the model’s agents can assign a correct mathematical probability to such events, thereby making them insurable. Events to which one cannot assign a probability, and that are thus truly uncertain, are ruled out.