Archive for the 'microfoundations' Category

A Tale of Two Syntheses

I recently finished reading a slender, but weighty, collection of essays, Microfoundtions Reconsidered: The Relationship of Micro and Macroeconomics in Historical Perspective, edited by Pedro Duarte and Gilberto Lima; it contains in addition to a brief introductory essay by the editors, and contributions by Kevin Hoover, Robert Leonard, Wade Hands, Phil Mirowski, Michel De Vroey, and Pedro Duarte. The volume is both informative and stimulating, helping me to crystalize ideas about which I have been ruminating and writing for a long time, but especially in some of my more recent posts (e.g., here, here, and here) and my recent paper “Hayek, Hicks, Radner and Four Equilibrium Concepts.”

Hoover’s essay provides a historical account of the microfoundations, making clear that the search for microfoundations long preceded the Lucasian microfoundations movement of the 1970s and 1980s that would revolutionize macroeconomics in the late 1980s and early 1990s. I have been writing about the differences between varieties of microfoundations for quite a while (here and here), and Hoover provides valuable detail about early discussions of microfoundations and about their relationship to the now regnant Lucasian microfoundations dogma. But for my purposes here, Hoover’s key contribution is his deconstruction of the concept of microfoundations, showing that the idea of microfoundations depends crucially on the notion that agents in a macroeconomic model be explicit optimizers, meaning that they maximize an explicit function subject to explicit constraints.

What Hoover clarifies is vacuity of the Lucasian optimization dogma. Until Lucas, optimization by agents had been merely a necessary condition for a model to be microfounded. But there was also another condition: that the optimizing choices of agents be mutually consistent. Establishing that the optimizing choices of agents are mutually consistent is not necessarily easy or even possible, so often the consistency of optimizing plans can only be suggested by some sort of heuristic argument. But Lucas and his cohorts, followed by their acolytes, unable to explain, even informally or heuristically, how the optimizing choices of individual agents are rendered mutually consistent, instead resorted to question-begging and question-dodging techniques to avoid addressing the consistency issue, of which one — the most egregious, but not the only — is the representative agent. In so doing, Lucas et al. transformed the optimization problem from the coordination of multiple independent choices into the optimal plan of a single decision maker. Heckuva job!

The second essay by Robert Leonard, though not directly addressing the question of microfoundations, helps clarify and underscore the misrepresentation perpetrated by the Lucasian microfoundational dogma in disregarding and evading the need to describe a mechanism whereby the optimal choices of individual agents are, or could be, reconciled. Leonard focuses on a particular economist, Oskar Morgenstern, who began his career in Vienna as a not untypical adherent of the Austrian school of economics, a member of the Mises seminar and successor of F. A. Hayek as director of the Austrian Institute for Business Cycle Research upon Hayek’s 1931 departure to take a position at the London School of Economics. However, Morgenstern soon began to question the economic orthodoxy of neoclassical economic theory and its emphasis on the tendency of economic forces to reach a state of equilibrium.

In his famous early critique of the foundations of equilibrium theory, Morgenstern tried to show that the concept of perfect foresight, upon which, he alleged, the concept of equilibrium rests, is incoherent. To do so, Morgenstern used the example of the Holmes-Moriarity interaction in which Holmes and Moriarty are caught in a dilemma in which neither can predict whether the other will get off or stay on the train on which they are both passengers, because the optimal choice of each depends on the choice of the other. The unresolvable conflict between Holmes and Moriarty, in Morgenstern’s view, showed that the incoherence of the idea of perfect foresight.

As his disillusionment with orthodox economic theory deepened, Morgenstern became increasingly interested in the potential of mathematics to serve as a tool of economic analysis. Through his acquaintance with the mathematician Karl Menger, the son of Carl Menger, founder of the Austrian School of economics. Morgenstern became close to Menger’s student, Abraham Wald, a pure mathematician of exceptional ability, who, to support himself, was working on statistical and mathematical problems for the Austrian Institute for Business Cycle Resarch, and tutoring Morgenstern in mathematics and its applications to economic theory. Wald, himself, went on to make seminal contributions to mathematical economics and statistical analysis.

Moregenstern also became acquainted with another student of Menger, John von Neumnn, with an interest in applying advanced mathematics to economic theory. Von Neumann and Morgenstern would later collaborate in writing The Theory of Games and Economic Behavior, as a result of which Morgenstern came to reconsider his early view of the Holmes-Moriarty paradox inasmuch as it could be shown that an equilibrium solution of their interaction could be found if payoffs to their joint choices were specified, thereby enabling Holmes and Moriarty to choose optimal probablistic strategies.

I don’t think that the game-theoretic solution to the Holmes Moriarty game is as straightforward as Morgenstern eventually agreed, but the critical point in the microfoundations discussion is that the mathematical solution to the Holmes-Moriarty paradox acknowledges the necessity for the choices made by two or more agents in an economic or game-theoretic equilibrium to be reconciled – i.e., rendered mutually consistent — in equilibrium. Under Lucasian microfoundations dogma, the problem is either annihilated by positing an optimizing representative agent having no need to coordinate his decision with other agents (I leave the question who, in the Holmes-Moriarty interaction, is the representative agent as an exercise for the reader) or it is assumed away by positing the existence of a magical equilibrium with no explanation of how the mutually consistent choices are arrived at.

The third essay (“The Rise and Fall of Walrasian Economics: The Keynes Effect”) by Wade Hands considers the first of the two syntheses – the neoclassical synthesis — that are alluded to in the title of this post. Hands gives a learned account of the mutually reinforcing co-development of Walrasian general equilibrium theory and Keynesian economics in the 25 years or so following World War II. Although Hands agrees that there is no necessary connection between Walrasian GE theory and Keynesian theory, he argues that there was enough common ground between Keynesians and Walrasians, as famously explained by Hicks in summarizing Keynesian theory by way of his IS-LM model, to allow the two disparate research programs to nourish each other in a kind of symbiotic relationship as the two research programs came to dominate postwar economics.

The task for Keynesian macroeconomists following the lead of Samuelson, Solow and Modigliani at MIT, Alvin Hansen at Harvard and James Tobin at Yale was to elaborate the Hicksian IS-LM approach by embedding it in a more general Walrasian framework. In so doing, they helped to shape a research agenda for Walrasian general-equilibrium theorists working out the details of the newly developed Arrow-Debreu model, deriving conditions for the uniqueness and stability of the equilibrium of that model. The neoclassical synthesis followed from those efforts, achieving an uneasy reconciliation between Walrasian general equilibrium theory and Keynesian theory. It received its most complete articulation in the impressive treatise of Don Patinkin which attempted to derive or at least evaluate key Keyensian propositions in the context of a full general equilibrium model. At an even higher level of theoretical sophistication, the 1971 summation of general equilibrium theory by Arrow and Hahn, gave disproportionate attention to Keynesian ideas which were presented and analyzed using the tools of state-of-the art Walrasian analysis.

Hands sums up the coexistence of Walrasian and Keynesian ideas in the Arrow-Hahn volume as follows:

Arrow and Hahn’s General Competitive Analysis – the canonical summary of the literature – dedicated far more pages to stability than to any other topic. The book had fourteen chapters (and a number of mathematical appendices); there was one chapter on consumer choice, one chapter on production theory, and one chapter on existence [of equilibrium], but there were three chapters on stability analysis, (two on the traditional tatonnement and one on alternative ways of modeling general equilibrium dynamics). Add to this the fact that there was an important chapter on “The Keynesian Model’; and it becomes clear how important stability analysis and its connection to Keynesian economics was for Walrasian microeconomics during this period. The purpose of this section has been to show that that would not have been the case if the Walrasian economics of the day had not been a product of co-evolution with Keynesian economic theory. (p. 108)

What seems most unfortunate about the neoclassical synthesis is that it elevated and reinforced the least relevant and least fruitful features of both the Walrasian and the Keynesian research programs. The Hicksian IS-LM setup abstracted from the dynamic and forward-looking aspects of Keynesian theory, modeling a static one-period model, not easily deployed as a tool of dynamic analysis. Walrasian GE analysis, which, following the pathbreaking GE existence proofs of Arrow and Debreu, then proceeded to a disappointing search for the conditions for a unique and stable general equilibrium.

It was Paul Samuelson who, building on Hicks’s pioneering foray into stability analysis, argued that the stability question could be answered by investigating whether a system of Lyapounov differential equations could describe market price adjustments as functions of market excess demands that would converge on an equilibrium price vector. But Samuelson’s approach to establishing stability required the mechanism of a fictional tatonnement process. Even with that unsatisfactory assumption, the stability results were disappointing.

Although for Walrasian theorists the results hardly repaid the effort expended, for those Keynesians who interpreted Keynes as an instability theorist, the weak Walrasian stability results might have been viewed as encouraging. But that was not any easy route to take either, because Keynes had also argued that a persistent unemployment equilibrium might be the norm.

It’s also hard to understand how the stability of equilibrium in an imaginary tatonnement process could ever have been considered relevant to the operation of an actual economy in real time – a leap of faith almost as extraordinary as imagining an economy represented by a single agent. Any conventional comparative-statics exercise – the bread and butter of microeconomic analysis – involves comparing two equilibria, corresponding to a specified parametric change in the conditions of the economy. The comparison presumes that, starting from an equilibrium position, the parametric change leads from an initial to a new equilibrium. If the economy isn’t stable, a disturbance causing an economy to depart from an initial equilibrium need not result in an adjustment to a new equilibrium comparable to the old one.

If conventional comparative statics hinges on an implicit stability assumption, it’s hard to see how a stability analysis of tatonnement has any bearing on the comparative-statics routinely relied upon by economists. No actual economy ever adjusts to a parametric change by way of tatonnement. Whether a parametric change displacing an economy from its equilibrium time path would lead the economy toward another equilibrium time path is another interesting and relevant question, but it’s difficult to see what insight would be gained by proving the stability of equilibrium under a tatonnement process.

Moreover, there is a distinct question about the endogenous stability of an economy: are there endogenous tendencies within an economy that lead it away from its equilibrium time path. But questions of endogenous stability can only be posed in a dynamic, rather than a static, model. While extending the Walrasian model to include an infinity of time periods, Arrow and Debreu telescoped determination of the intertemporal-equilibrium price vector into a preliminary time period before time, production, exchange and consumption begin. So, even in the formally intertemporal Arrow-Debreu model, the equilibrium price vector, once determined, is fixed and not subject to revision. Standard stability analysis was concerned with the response over time to changing circumstances only insofar as changes are foreseen at time zero, before time begins, so that they can be and are taken fully into account when the equilibrium price vector is determined.

Though not entirely uninteresting, the intertemporal analysis had little relevance to the stability of an actual economy operating in real time. Thus, neither the standard Keyensian (IS-LM) model nor the standard Walrasian Arrow-Debreu model provided an intertemporal framework within which to address the dynamic stability that Keynes (and contemporaries like Hayek, Myrdal, Lindahl and Hicks) had developed in the 1930s. In particular, Hicks’s analytical device of temporary equilibrium might have facilitated such an analysis. But, having introduced his IS-LM model two years before publishing his temporary equilibrium analysis in Value and Capital, Hicks concentrated his attention primarily on Keynesian analysis and did not return to the temporary equilibrium model until 1965 in Capital and Growth. And it was IS-LM that became, for a generation or two, the preferred analytical framework for macroeconomic analysis, while temproary equilibrium remained overlooked until the 1970s just as the neoclassical synthesis started coming apart.

The fourth essay by Phil Mirowski investigates the role of the Cowles Commission, based at the University of Chicago from 1939 to 1955, in undermining Keynesian macroeconomics. While Hands argues that Walrasians and Keynesians came together in a non-hostile spirit of tacit cooperation, Mirowski believes that owing to their Walrasian sympathies, the Cowles Committee had an implicit anti-Keynesian orientation and was therefore at best unsympathetic if not overtly hostile to Keynesian theorizing, which was incompatible the Walrasian optimization paradigm endorsed by the Cowles economists. (Another layer of unexplored complexity is the tension between the Walrasianism of the Cowles economists and the Marshallianism of the Chicago School economists, especially Knight and Friedman, which made Chicago an inhospitable home for the Cowles Commission and led to its eventual departure to Yale.)

Whatever differences, both the Mirowski and the Hands essays support the conclusion that the uneasy relationship between Walrasianism and Keynesianism was inherently problematic and unltimately unsustainable. But to me the tragedy is that before the fall, in the 1950s and 1960s, when the neoclassical synthesis bestrode economics like a colossus, the static orientation of both the Walrasian and the Keynesian research programs combined to distract economists from a more promising research program. Such a program, instead of treating expectations either as parametric constants or as merely adaptive, based on an assumed distributed lag function, might have considered whether expectations could perform a potentially equilibrating role in a general equilibrium model.

The equilibrating role of expectations, though implicit in various contributions by Hayek, Myrdal, Lindahl, Irving Fisher, and even Keynes, is contingent so that equilibrium is not inevitable, only a possibility. Instead, the introduction of expectations as an equilibrating variable did not occur until the mid-1970s when Robert Lucas, Tom Sargent and Neil Wallace, borrowing from John Muth’s work in applied microeconomics, introduced the idea of rational expectations into macroeconomics. But in introducing rational expectations, Lucas et al. made rational expectations not the condition of a contingent equilibrium but an indisputable postulate guaranteeing the realization of equilibrium without offering any theoretical account of a mechanism whereby the rationality of expectations is achieved.

The fifth essay by Michel DeVroey (“Microfoundations: a decisive dividing line between Keynesian and new classical macroeconomics?”) is a philosophically sophisticated analysis of Lucasian microfoundations methodological principles. DeVroey begins by crediting Lucas with the revolution in macroeconomics that displaced a Keynesian orthodoxy already discredited in the eyes of many economists after its failure to account for simultaneously rising inflation and unemployment.

The apparent theoretical disorder characterizing the Keynesian orthodoxy and its Monetarist opposition left a void for Lucas to fill by providing a seemingly rigorous microfounded alternative to the confused state of macroeconomics. And microfoundations became the methodological weapon by which Lucas and his associates and followers imposed an iron discipline on the unruly community of macroeconomists. “In Lucas’s eyes,” DeVroey aptly writes,“ the mere intention to produce a theory of involuntary unemployment constitutes an infringement of the equilibrium discipline.” Showing that his description of Lucas is hardly overstated, DeVroey quotes from the famous 1978 joint declaration of war issued by Lucas and Sargent against Keynesian macroeconomics:

After freeing himself of the straightjacket (or discipline) imposed by the classical postulates, Keynes described a model in which rules of thumb, such as the consumption function and liquidity preference schedule, took the place of decision functions that a classical economist would insist be derived from the theory of choice. And rather than require that wages and prices be determined by the postulate that markets clear – which for the labor market seemed patently contradicted by the severity of business depressions – Keynes took as an unexamined postulate that money wages are sticky, meaning that they are set at a level or by a process that could be taken as uninfluenced by the macroeconomic forces he proposed to analyze.

Echoing Keynes’s famous description of the sway of Ricardian doctrines over England in the nineteenth century, DeVroey remarks that the microfoundations requirement “conquered macroeconomics as quickly and thoroughly as the Holy Inquisition conquered Spain,” noting, even more tellingly, that the conquest was achieved without providing any justification. Ricardo had, at least, provided a substantive analysis that could be debated; Lucas offered only an undisputable methodological imperative about the sole acceptable mode of macroeconomic reasoning. Just as optimization is a necessary component of the equilibrium discipline that had to be ruthlessly imposed on pain of excommunication from the macroeconomic community, so, too, did the correlate principle of market-clearing. To deviate from the market-clearing postulate was ipso facto evidence of an impure and heretical state of mind. DeVroey further quotes from the war declaration of Lucas and Sargent.

Cleared markets is simply a principle, not verifiable by direct observation, which may or may not be useful in constructing successful hypotheses about the behavior of these [time] series.

What was only implicit in the war declaration became evident later after right-thinking was enforced, and woe unto him that dared deviate from the right way of thinking.

But, as DeVroey skillfully shows, what is most remarkable is that, having declared market clearing an indisputable methodological principle, Lucas, contrary to his own demand for theoretical discipline, used the market-clearing postulate to free himself from the very equilibrium discipline he claimed to be imposing. How did the market-clearing postulate liberate Lucas from equilibrium discipline? To show how the sleight-of-hand was accomplished, DeVroey, in an argument parallel to that of Hoover in chapter one and that suggested by Leonard in chapter two, contrasts Lucas’s conception of microfoundations with a different microfoundations conception espoused by Hayek and Patinkin. Unlike Lucas, Hayek and Patinkin recognized that the optimization of individual economic agents is conditional on the optimization of other agents. Lucas assumes that if all agents optimize, then their individual optimization ensures that a social optimum is achieved, the whole being the sum of its parts. But that assumption ignores that the choices made interacting agents are themelves interdependent.

To capture the distinction between independent and interdependent optimization, DeVroey distinguishes between optimal plans and optimal behavior. Behavior is optimal only if an optimal plan can be executed. All agents can optimize individually in making their plans, but the optimality of their behavior depends on their capacity to carry those plans out. And the capacity of each to carry out his plan is contingent on the optimal choices of all other agents.

Optimizing plans refers to agents’ intentions before the opening of trading, the solution to the choice-theoretical problem with which they are faced. Optimizing behavior refers to what is observable after trading has started. Thus optimal behavior implies that the optimal plan has been realized. . . . [O]ptmizing plans and optimizing behavior need to be logically separated – there is a difference between finding a solution to a choice problem and implementing the solution. In contrast, whenever optimizing behavior is the sole concept used, the possibility of there being a difference between them is discarded by definition. This is the standpoint takenby Lucas and Sargent. Once it is adopted, it becomes misleading to claim . . .that the microfoundations requirement is based on two criteria, optimizing behavior and market clearing. A single criterion is needed, and it is irrelevant whether this is called generalized optimizing behavior or market clearing. (De Vroey, p. 176)

Each agent is free to optimize his plan, but no agent can execute his optimal plan unless the plan coincides with the complementary plans of other agents. So, the execution of an optimal plan is not within the unilateral control of an agent formulating his own plan. One can readily assume that agents optimize their plans, but one cannot just assume that those plans can be executed as planned. The optimality of interdependent plans is not self-evident; it is a proposition that must be demonstrated. Assuming that agents optimize, Lucas simply asserts that, because agents optimize, markets must clear.

That is a remarkable non-sequitur. And from that non-sequitur, Lucas jumps to a further non-sequitur: that an optimizing representative agent is all that’s required for a macroeconomic model. The logical straightjacket (or discipline) of demonstrating that interdependent optimal plans are consistent is thus discarded (or trampled upon). Lucas’s insistence on a market-clearing principle turns out to be subterfuge by which the pretense of its upholding conceals its violation in practice.

My own view is that the assumption that agents formulate optimizing plans cannot be maintained without further analysis unless the agents are operating in isolation. If the agents interacting with each other, the assumption that they optimize requires a theory of their interaction. If the focus is on equilibrium interactions, then one can have a theory of equilibrium, but then the possibility of non-equilibrium states must also be acknowledged.

That is what John Nash did in developing his equilibrium theory of positive-sum games. He defined conditions for the existence of equilibrium, but he offered no theory of how equilibrium is achieved. Lacking such a theory, he acknowledged that non-equilibrium solutions might occur, e.g., in some variant of the Holmes-Moriarty game. To simply assert that because interdependent agents try to optimize, they must, as a matter of principle, succeed in optimizing is to engage in question-begging on a truly grand scale. To insist, as a matter of methodological principle, that everyone else must also engage in question-begging on equally grand scale is what I have previously called methodological arrogance, though an even harsher description might be appropriate.

In the sixth essay (“Not Going Away: Microfoundations in the making of a new consensus in macroeconomics”), Pedro Duarte considers the current state of apparent macroeconomic consensus in the wake of the sweeping triumph of the Lucasian micorfoundtions methodological imperative. In its current state, mainstream macroeconomists from a variety of backgrounds have reconciled themselves and adjusted to the methodological absolutism Lucas and his associates and followers have imposed on macroeconomic theorizing. Leading proponents of the current consensus are pleased to announce, in unseemly self-satisfaction, that macroeconomics is now – but presumably not previously – “firmly grounded in the principles of economic [presumably neoclassical] theory.” But the underlying conception of neoclassical economic theory motivating such a statement is almost laughably narrow, and, as I have just shown, strictly false even if, for argument’s sake, that narrow conception is accepted.

Duarte provides an informative historical account of the process whereby most mainstream Keynesians and former old-line Monetarists, who had, in fact, adopted much of the underlying Keynesian theoretical framework themselves, became reconciled to the non-negotiable methodological microfoundational demands upon which Lucas and his New Classical followers and Real-Business-Cycle fellow-travelers insisted. While Lucas was willing to tolerate differences of opinion about the importance of monetary factors in accounting for business-cycle fluctuations in real output and employment, and even willing to countenance a role for countercyclical monetary policy, such differences of opinion could be tolerated only if they could be derived from an acceptable microfounded model in which the agent(s) form rational expectations. If New Keynesians were able to produce results rationalizing countercyclical policies in such microfounded models with rational expectations, Lucas was satisfied. Presumably, Lucas felt the price of conceding the theoretical legitimacy of countercyclical policy was worth paying in order to achieve methodological hegemony over macroeconomic theory.

And no doubt, for Lucas, the price was worth paying, because it led to what Marvin Goodfriend and Robert King called the New Neoclassical Synthesis in their 1997 article ushering in the new era of good feelings, a synthesis based on “the systematic application of intertemporal optimization and rational expectations” while embodying “the insights of monetarists . . . regarding the theory and practice of monetary policy.”

While the first synthesis brought about a convergence of sorts between the disparate Walrasian and Keynesian theoretical frameworks, the convergence proved unstable because the inherent theoretical weaknesses of both paradigms were unable to withstand criticisms of the theoretical apparatus and of the policy recommendations emerging from that synthesis, particularly an inability to provide a straightforward analysis of inflation when it became a serious policy problem in the late 1960s and 1970s. But neither the Keynesian nor the Walrasian paradigms were developing in a way that addressed the points of most serious weakness.

On the Keynesian side, the defects included the static nature of the workhorse IS-LM model, the absence of a market for real capital and of a market for endogenous money. On the Walrasian side, the defects were the lack of any theory of actual price determination or of dynamic adjustment. The Hicksian temporary equilibrium paradigm might have provided a viable way forward, and for a very different kind of synthesis, but not even Hicks himself realized the potential of his own creation.

While the first synthesis was a product of convenience and misplaced optimism, the second synthesis is a product of methodological hubris and misplaced complacency derived from an elementary misunderstanding of the distinction between optimization by a single agent and the simultaneous optimization of two or more independent, yet interdependent, agents. The equilibrium of each is the result of the equilibrium of all, and a theory of optimization involving two or more agents requires a theory of how two or more interdependent agents can optimize simultaneously. The New neoclassical synthesis rests on the demand for a macroeconomic theory of individual optimization that refuses even to ask, let along provide an answer to, the question whether the optimization that it demands is actually achieved in practice or what happens if it is not. This is not a synthesis that will last, or that deserves to. And the sooner it collapses, the better off macroeconomics will be.

What the answer is I don’t know, but if I had to offer a suggestion, the one offered by my teacher Axel Leijonhufvud towards the end of his great book, written more than half a century ago, strikes me as not bad at all:

One cannot assume that what went wrong was simply that Keynes slipped up here and there in his adaptation of standard tool, and that consequently, if we go back and tinker a little more with the Marshallian toolbox his purposes will be realized. What is required, I believe, is a systematic investigation, form the standpoint of the information problems stressed in this study, of what elements of the static theory of resource allocation can without further ado be utilized in the analysis of dynamic and historical systems. This, of course, would be merely a first-step: the gap yawns very wide between the systematic and rigorous modern analysis of the stability of “featureless,” pure exchange systems and Keynes’ inspired sketch of the income-constrained process in a monetary-exchange-cum-production system. But even for such a first step, the prescription cannot be to “go back to Keynes.” If one must retrace some steps of past developments in order to get on the right track—and that is probably advisable—my own preference is to go back to Hayek. Hayek’s Gestalt-conception of what happens during business cycles, it has been generally agreed, was much less sound than Keynes’. As an unhappy consequence, his far superior work on the fundamentals of the problem has not received the attention it deserves. (p. 401)

I agree with all that, but would also recommend Roy Radner’s development of an alternative to the Arrow-Debreu version of Walrasian general equilibrium theory that can accommodate Hicksian temporary equilibrium, and Hawtrey’s important contributions to our understanding of monetary theory and the role and potential instability of endogenous bank money. On top of that, Franklin Fisher in his important work, The Disequilibrium Foundations of Equilibrium Economics, has given us further valuable guidance in how to improve the current sorry state of macroeconomics.

 

Filling the Arrow Explanatory Gap

The following (with some minor revisions) is a Twitter thread I posted yesterday. Unfortunately, because it was my first attempt at threading the thread wound up being split into three sub-threads and rather than try to reconnect them all, I will just post the complete thread here as a blogpost.

1. Here’s an outline of an unwritten paper developing some ideas from my paper “Hayek Hicks Radner and Four Equilibrium Concepts” (see here for an earlier ungated version) and some from previous blog posts, in particular Phillips Curve Musings

2. Standard supply-demand analysis is a form of partial-equilibrium (PE) analysis, which means that it is contingent on a ceteris paribus (CP) assumption, an assumption largely incompatible with realistic dynamic macroeconomic analysis.

3. Macroeconomic analysis is necessarily situated a in general-equilibrium (GE) context that precludes any CP assumption, because there are no variables that are held constant in GE analysis.

4. In the General Theory, Keynes criticized the argument based on supply-demand analysis that cutting nominal wages would cure unemployment. Instead, despite his Marshallian training (upbringing) in PE analysis, Keynes argued that PE (AKA supply-demand) analysis is unsuited for understanding the problem of aggregate (involuntary) unemployment.

5. The comparative-statics method described by Samuelson in the Foundations of Econ Analysis formalized PE analysis under the maintained assumption that a unique GE obtains and deriving a “meaningful theorem” from the 1st- and 2nd-order conditions for a local optimum.

6. PE analysis, as formalized by Samuelson, is conditioned on the assumption that GE obtains. It is focused on the effect of changing a single parameter in a single market small enough for the effects on other markets of the parameter change to be made negligible.

7. Thus, PE analysis, the essence of micro-economics is predicated on the macrofoundation that all, but one, markets are in equilibrium.

8. Samuelson’s meaningful theorems were a misnomer reflecting mid-20th-century operationalism. They can now be understood as empirically refutable propositions implied by theorems augmented with a CP assumption that interactions b/w markets are small enough to be neglected.

9. If a PE model is appropriately specified, and if the market under consideration is small or only minimally related to other markets, then differences between predictions and observations will be statistically insignificant.

10. So PE analysis uses comparative-statics to compare two alternative general equilibria that differ only in respect of a small parameter change.

11. The difference allows an inference about the causal effect of a small change in that parameter, but says nothing about how an economy would actually adjust to a parameter change.

12. PE analysis is conditioned on the CP assumption that the analyzed market and the parameter change are small enough to allow any interaction between the parameter change and markets other than the market under consideration to be disregarded.

13. However, the process whereby one equilibrium transitions to another is left undetermined; the difference between the two equilibria with and without the parameter change is computed but no account of an adjustment process leading from one equilibrium to the other is provided.

14. Hence, the term “comparative statics.”

15. The only suggestion of an adjustment process is an assumption that the price-adjustment in any market is an increasing function of excess demand in the market.

16. In his seminal account of GE, Walras posited the device of an auctioneer who announces prices–one for each market–computes desired purchases and sales at those prices, and sets, under an adjustment algorithm, new prices at which desired purchases and sales are recomputed.

17. The process continues until a set of equilibrium prices is found at which excess demands in all markets are zero. In Walras’s heuristic account of what he called the tatonnement process, trading is allowed only after the equilibrium price vector is found by the auctioneer.

18. Walras and his successors assumed, but did not prove, that, if an equilibrium price vector exists, the tatonnement process would eventually, through trial and error, converge on that price vector.

19. However, contributions by Sonnenschein, Mantel and Debreu (hereinafter referred to as the SMD Theorem) show that no price-adjustment rule necessarily converges on a unique equilibrium price vector even if one exists.

20. The possibility that there are multiple equilibria with distinct equilibrium price vectors may or may not be worth explicit attention, but for purposes of this discussion, I confine myself to the case in which a unique equilibrium exists.

21. The SMD Theorem underscores the lack of any explanatory account of a mechanism whereby changes in market prices, responding to excess demands or supplies, guide a decentralized system of competitive markets toward an equilibrium state, even if a unique equilibrium exists.

22. The Walrasian tatonnement process has been replaced by the Arrow-Debreu-McKenzie (ADM) model in an economy of infinite duration consisting of an infinite number of generations of agents with given resources and technology.

23. The equilibrium of the model involves all agents populating the economy over all time periods meeting before trading starts, and, based on initial endowments and common knowledge, making plans given an announced equilibrium price vector for all time in all markets.

24. Uncertainty is accommodated by the mechanism of contingent trading in alternative states of the world. Given assumptions about technology and preferences, the ADM equilibrium determines the set prices for all contingent states of the world in all time periods.

25. Given equilibrium prices, all agents enter into optimal transactions in advance, conditioned on those prices. Time unfolds according to the equilibrium set of plans and associated transactions agreed upon at the outset and executed without fail over the course of time.

26. At the ADM equilibrium price vector all agents can execute their chosen optimal transactions at those prices in all markets (certain or contingent) in all time periods. In other words, at that price vector, excess demands in all markets with positive prices are zero.

27. The ADM model makes no pretense of identifying a process that discovers the equilibrium price vector. All that can be said about that price vector is that if it exists and trading occurs at equilibrium prices, then excess demands will be zero if prices are positive.

28. Arrow himself drew attention to the gap in the ADM model, writing in 1959:

29. In addition to the explanatory gap identified by Arrow, another shortcoming of the ADM model was discussed by Radner: the dependence of the ADM model on a complete set of forward and state-contingent markets at time zero when equilibrium prices are determined.

30. Not only is the complete-market assumption a backdoor reintroduction of perfect foresight, it excludes many features of the greatest interest in modern market economies: the existence of money, stock markets, and money-crating commercial banks.

31. Radner showed that for full equilibrium to obtain, not only must excess demands in current markets be zero, but whenever current markets and current prices for future delivery are missing, agents must correctly expect those future prices.

32. But there is no plausible account of an equilibrating mechanism whereby price expectations become consistent with GE. Although PE analysis suggests that price adjustments do clear markets, no analogous analysis explains how future price expectations are equilibrated.

33. But if both price expectations and actual prices must be equilibrated for GE to obtain, the notion that “market-clearing” price adjustments are sufficient to achieve macroeconomic “equilibrium” is untenable.

34. Nevertheless, the idea that individual price expectations are rational (correct), so that, except for random shocks, continuous equilibrium is maintained, became the bedrock for New Classical macroeconomics and its New Keynesian and real-business cycle offshoots.

35. Macroeconomic theory has become a theory of dynamic intertemporal optimization subject to stochastic disturbances and market frictions that prevent or delay optimal adjustment to the disturbances, potentially allowing scope for countercyclical monetary or fiscal policies.

36. Given incomplete markets, the assumption of nearly continuous intertemporal equilibrium implies that agents correctly foresee future prices except when random shocks occur, whereupon agents revise expectations in line with the new information communicated by the shocks.
37. Modern macroeconomics replaced the Walrasian auctioneer with agents able to forecast the time path of all prices indefinitely into the future, except for intermittent unforeseen shocks that require agents to optimally their revise previous forecasts.
38. When new information or random events, requiring revision of previous expectations, occur, the new information becomes common knowledge and is processed and interpreted in the same way by all agents. Agents with rational expectations always share the same expectations.
39. So in modern macro, Arrow’s explanatory gap is filled by assuming that all agents, given their common knowledge, correctly anticipate current and future equilibrium prices subject to unpredictable forecast errors that change their expectations of future prices to change.
40. Equilibrium prices aren’t determined by an economic process or idealized market interactions of Walrasian tatonnement. Equilibrium prices are anticipated by agents, except after random changes in common knowledge. Semi-omniscient agents replace the Walrasian auctioneer.
41. Modern macro assumes that agents’ common knowledge enables them to form expectations that, until superseded by new knowledge, will be validated. The assumption is wrong, and the mistake is deeper than just the unrealism of perfect competition singled out by Arrow.
42. Assuming perfect competition, like assuming zero friction in physics, may be a reasonable simplification for some problems in economics, because the simplification renders an otherwise intractable problem tractable.
43. But to assume that agents’ common knowledge enables them to forecast future prices correctly transforms a model of decentralized decision-making into a model of central planning with each agent possessing the knowledge only possessed by an omniscient central planner.
44. The rational-expectations assumption fills Arrow’s explanatory gap, but in a deeply unsatisfactory way. A better approach to filling the gap would be to acknowledge that agents have private knowledge (and theories) that they rely on in forming their expectations.
45. Agents’ expectations are – at least potentially, if not inevitably – inconsistent. Because expectations differ, it’s the expectations of market specialists, who are better-informed than non-specialists, that determine the prices at which most transactions occur.
46. Because price expectations differ even among specialists, prices, even in competitive markets, need not be uniform, so that observed price differences reflect expectational differences among specialists.
47. When market specialists have similar expectations about future prices, current prices will converge on the common expectation, with arbitrage tending to force transactions prices to converge toward notwithstanding the existence of expectational differences.
48. However, the knowledge advantage of market specialists over non-specialists is largely limited to their knowledge of the workings of, at most, a small number of related markets.
49. The perspective of specialists whose expectations govern the actual transactions prices in most markets is almost always a PE perspective from which potentially relevant developments in other markets and in macroeconomic conditions are largely excluded.
50. The interrelationships between markets that, according to the SMD theorem, preclude any price-adjustment algorithm, from converging on the equilibrium price vector may also preclude market specialists from converging, even roughly, on the equilibrium price vector.
51. A strict equilibrium approach to business cycles, either real-business cycle or New Keynesian, requires outlandish assumptions about agents’ common knowledge and their capacity to anticipate the future prices upon which optimal production and consumption plans are based.
52. It is hard to imagine how, without those outlandish assumptions, the theoretical superstructure of real-business cycle theory, New Keynesian theory, or any other version of New Classical economics founded on the rational-expectations postulate can be salvaged.
53. The dominance of an untenable macroeconomic paradigm has tragically led modern macroeconomics into a theoretical dead end.

Graeber Against Economics

David Graeber’s vitriolic essay “Against Economics” in the New York Review of Books has generated responses from Noah Smith and Scott Sumner among others. I don’t disagree with much that Noah or Scott have to say, but I want to dig a little deeper than they did into some of Graeber’s arguments, because even though I think he is badly misinformed on many if not most of the subjects he writes about, I actually have some sympathy for his dissatisfaction with the current state of economics. Graeber wastes no time on pleasantries.

There is a growing feeling, among those who have the responsibility of managing large economies, that the discipline of economics is no longer fit for purpose. It is beginning to look like a science designed to solve problems that no longer exist.

A serious polemicist should avoid blatant mischaracterizations, exaggerations and cheap shots, and should be well-grounded in the object of his critique, thereby avoiding criticisms that undermine his own claims to expertise. I grant that  Graeber has some valid criticisms to make, even agreeing with him, at least in part, on some of them. But his indiscriminate attacks on, and caricatures of, all neoclassical economics betrays a superficial understanding of that discipline.

Graeber begins by attacking what he considers the misguided and obsessive focus on inflation by economists.

A good example is the obsession with inflation. Economists still teach their students that the primary economic role of government—many would insist, its only really proper economic role—is to guarantee price stability. We must be constantly vigilant over the dangers of inflation. For governments to simply print money is therefore inherently sinful.

Every currency unit, or banknote issued by a central bank, now in circulation, as Graeber must know, has been “printed.” So to say that economists consider it sinful for governments to print money is either a deliberate falsehood, or an emotional rhetorical outburst, as Graeber immediately, and apparently unwittingly, acknowledges!

If, however, inflation is kept at bay through the coordinated action of government and central bankers, the market should find its “natural rate of unemployment,” and investors, taking advantage of clear price signals, should be able to ensure healthy growth. These assumptions came with the monetarism of the 1980s, the idea that government should restrict itself to managing the money supply, and by the 1990s had come to be accepted as such elementary common sense that pretty much all political debate had to set out from a ritual acknowledgment of the perils of government spending. This continues to be the case, despite the fact that, since the 2008 recession, central banks have been printing money frantically [my emphasis] in an attempt to create inflation and compel the rich to do something useful with their money, and have been largely unsuccessful in both endeavors.

Graeber’s use of the ambiguous pronoun “this” beginning the last sentence of the paragraph betrays his own confusion about what he is saying. Central banks are printing money and attempting to “create” inflation while supposedly still believing that inflation is a menace and printing money is a sin. Go figure.

We now live in a different economic universe than we did before the crash. Falling unemployment no longer drives up wages. Printing money does not cause inflation. Yet the language of public debate, and the wisdom conveyed in economic textbooks, remain almost entirely unchanged.

Again showing an inadequate understanding of basic economic theory, Graeber suggests that, in theory if not practice, falling unemployment should cause wages to rise. The Philips Curve, upon which Graeber’s suggestion relies, represents the empirically observed negative correlation between the rate of average wage increase and the rate of unemployment. But correlation does not imply causation, so there is no basis in economic theory to assert that falling unemployment causes the rate of increase in wages to accelerate. That the empirical correlation between unemployment and wage increases has not recently been in evidence provides no compelling reason for changing textbook theory.

From this largely unfounded and attack on economic theory – a theory which I myself consider, in many respects, inadequate and unreliable – Graeber launches a bitter diatribe against the supposed hegemony of economists over policy-making.

Mainstream economists nowadays might not be particularly good at predicting financial crashes, facilitating general prosperity, or coming up with models for preventing climate change, but when it comes to establishing themselves in positions of intellectual authority, unaffected by such failings, their success is unparalleled. One would have to look at the history of religions to find anything like it.

The ability to predict financial crises would be desirable, but that cannot be the sole criterion for whether economics has advanced our understanding of how economic activity is organized or what effects policy changes have. (I note parenthetically that many economists defensively reject the notion that economic crises are predictable on the grounds that if economists could predict a future economic crisis, those predictions would be immediately self-fulfilling. This response, of course, effectively disproves the idea that economists could predict that an economic crisis would occur in the way that astronomers predict solar eclipses. But this response slays a strawman. The issue is not whether economists can predict future crises, but whether they can identify conditions indicating an increased likelihood of a crisis and suggest precautionary measures to reduce the likelihood that a potential crisis will occur. But Graeber seems uninterested in or incapable of engaging the question at even this moderate level of subtlety.)

In general, I doubt that economists can make more than a modest contribution to improved policy-making, and the best that one can hope for is probably that they steer us away from the worst potential decisions rather than identifying the best ones. But no one, as far as I know, has yet been burned at the stake by a tribunal of economists.

To this day, economics continues to be taught not as a story of arguments—not, like any other social science, as a welter of often warring theoretical perspectives—but rather as something more like physics, the gradual realization of universal, unimpeachable mathematical truths. “Heterodox” theories of economics do, of course, exist (institutionalist, Marxist, feminist, “Austrian,” post-Keynesian…), but their exponents have been almost completely locked out of what are considered “serious” departments, and even outright rebellions by economics students (from the post-autistic economics movement in France to post-crash economics in Britain) have largely failed to force them into the core curriculum.

I am now happy to register agreement with something that Graeber says. Economists in general have become overly attached to axiomatic and formalistic mathematical models that create a false and misleading impression of rigor and mathematical certainty. In saying this, I don’t dispute that mathematical modeling is an important part of much economic theorizing, but it should not exclude other approaches to economic analysis and discourse.

As a result, heterodox economists continue to be treated as just a step or two away from crackpots, despite the fact that they often have a much better record of predicting real-world economic events. What’s more, the basic psychological assumptions on which mainstream (neoclassical) economics is based—though they have long since been disproved by actual psychologists—have colonized the rest of the academy, and have had a profound impact on popular understandings of the world.

That heterodox economists have a better record of predicting economic events than mainstream economists is an assertion for which Graeber offers no evidence or examples. I would not be surprised if he could cite examples, but one would have to weigh the evidence surrounding those examples before concluding that predictions by heterodox economists were more accurate than those of their mainstream counterparts.

Graeber returns to the topic of monetary theory, which seems a particular bugaboo of his. Taking the extreme liberty of holding up Mrs. Theresa May as a spokesperson for orthodox economics, he focuses on her definitive 2017 statement that there is no magic money tree.

The truly extraordinary thing about May’s phrase is that it isn’t true. There are plenty of magic money trees in Britain, as there are in any developed economy. They are called “banks.” Since modern money is simply credit, banks can and do create money literally out of nothing, simply by making loans. Almost all of the money circulating in Britain at the moment is bank-created in this way.

What Graeber chooses to ignore is that banks do not operate magically; they make loans and create deposits in seeking to earn profits; their decisions are not magical, but are oriented toward making profits. Whether they make good or bad decisions is debatable, but the debate isn’t about a magical process; it’s a debate about theory and evidence. Graeber describe how he thinks that economists think about how banks create money, correctly observing that there is a debate about how that process works, but without understanding those differences or their significance.

Economists, for obvious reasons, can’t be completely oblivious to the role of banks, but they have spent much of the twentieth century arguing about what actually happens when someone applies for a loan. One school insists that banks transfer existing funds from their reserves, another that they produce new money, but only on the basis of a multiplier effect). . . Only a minority—mostly heterodox economists, post-Keynesians, and modern money theorists—uphold what is called the “credit creation theory of banking”: that bankers simply wave a magic wand and make the money appear, secure in the confidence that even if they hand a client a credit for $1 million, ultimately the recipient will put it back in the bank again, so that, across the system as a whole, credits and debts will cancel out. Rather than loans being based in deposits, in this view, deposits themselves were the result of loans.

The one thing it never seemed to occur to anyone to do was to get a job at a bank, and find out what actually happens when someone asks to borrow money. In 2014 a German economist named Richard Werner did exactly that, and discovered that, in fact, loan officers do not check their existing funds, reserves, or anything else. They simply create money out of thin air, or, as he preferred to put it, “fairy dust.”

Graeber is right that economists differ in how they understand banking. But the simple transfer-of-funds view, a product of the eighteenth century, was gradually rejected over the course of the nineteenth century; the money-multiplier view largely superseded it, enjoying a half-century or more of dominance as a theory of banking, still remains a popular way for introductory textbooks to explain how banking works, though it would be better if it were decently buried and forgotten. But since James Tobin’s classic essay “Commercial banks as creators of money” was published in 1963, most economists who have thought carefully about banking have concluded that the amount of deposits created by banks corresponds to the quantity of deposits that the public, given their expectations about the future course of the economy and the future course of prices, chooses to hold. The important point is that while a bank can create deposits without incurring more than the negligible cost of making a book-keeping, or an electronic, entry in a customer’s account, the creation of a deposit is typically associated with a demand by the bank to hold either reserves in its account with the Fed or to hold some amount of Treasury instruments convertible, on very short notice, into reserves at the Fed.

Graeber seems to think that there is something fundamental at stake for the whole of macroeconomics in the question whether deposits created loans or loans create deposits. I agree that it’s an important question, but not as significant as Graeber believes. But aside from that nuance, what’s remarkable is that Graeber actually acknowledges that the weight of professional opinion is on the side that says that loans create deposits. He thus triumphantly cites a report by Bank of England economists that correctly explained that banks create money and do so in the normal course of business by making loans.

Before long, the Bank of England . . . rolled out an elaborate official report called “Money Creation in the Modern Economy,” replete with videos and animations, making the same point: existing economics textbooks, and particularly the reigning monetarist orthodoxy, are wrong. The heterodox economists are right. Private banks create money. Central banks like the Bank of England create money as well, but monetarists are entirely wrong to insist that their proper function is to control the money supply.

Graeber, I regret to say, is simply exposing the inadequacy of his knowledge of the history of economics. Adam Smith in The Wealth of Nations explained that banks create money who, in doing so, saved the resources that would have been wasted on creating additional gold and silver. Subsequent economists from David Ricardo through Henry Thornton, J. S. Mill and R. G. Hawtrey were perfectly aware that banks can supply money — either banknotes or deposits — at less than the cost of mining and minting new coins, as they extend their credit in making loans to borrowers. So what is at issue, Graeber to the contrary notwithstanding, is not a dispute between orthodoxy and heterodoxy.

In fact, central banks do not in any sense control the money supply; their main function is to set the interest rate—to determine how much private banks can charge for the money they create.

Central banks set a rental price for reserves, thereby controlling the quantity of reserves into which bank deposits are convertible that is available to the economy. One way to think about that quantity is that the quantity of reserves along with the aggregate demand to hold reserves determines the exchange value of reserves and hence the price level; another way to think about it is that the interest rate or the implied policy stance of the central bank helps to determine the expectations of the public about the future course of the price level which is what determines – within some margin of error or range – what the future course of the price level will turn out to be.

Almost all public debate on these subjects is therefore based on false premises. For example, if what the Bank of England was saying were true, government borrowing didn’t divert funds from the private sector; it created entirely new money that had not existed before.

This is just silly. Funds may or may not be diverted from the private sector, but the total available resources to society is finite. If the central bank creates additional money, it creates additional claims to those resources and the creation of additional claims to resources necessarily has an effect on the prices of inputs and of outputs.

One might have imagined that such an admission would create something of a splash, and in certain restricted circles, it did. Central banks in Norway, Switzerland, and Germany quickly put out similar papers. Back in the UK, the immediate media response was simply silence. The Bank of England report has never, to my knowledge, been so much as mentioned on the BBC or any other TV news outlet. Newspaper columnists continued to write as if monetarism was self-evidently correct. Politicians continued to be grilled about where they would find the cash for social programs. It was as if a kind of entente cordiale had been established, in which the technocrats would be allowed to live in one theoretical universe, while politicians and news commentators would continue to exist in an entirely different one.

Even if we stipulate that this characterization of what the BBC and newspaper columnists believe is correct, what we would have — at best — is a commentary on the ability of economists to communicate their understanding of how the economy works to the intelligentsia that communicates to ordinary citizens. It is not in and of itself a commentary on the state of economic knowledge, inasmuch as Graeber himself concedes that most economists don’t accept monetarism. And that has been the case, as Noah Smith pointed out in his Bloomberg column on Graeber, since the early 1980s when the Monetarist experiment in trying to conduct monetary policy by controlling the monetary aggregates proved entirely unworkable and had to be abandoned as it was on the verge of precipitating a financial crisis.

Only after this long warmup decrying the sorry state of contemporary economic theory does Graeber begin discussing the book under review Money and Government by Robert Skidelsky.

What [Skidelsky] reveals is an endless war between two broad theoretical perspectives. . . The crux of the argument always seems to turn on the nature of money. Is money best conceived of as a physical commodity, a precious substance used to facilitate exchange, or is it better to see money primarily as a credit, a bookkeeping method or circulating IOU—in any case, a social arrangement? This is an argument that has been going on in some form for thousands of years. What we call “money” is always a mixture of both, and, as I myself noted in Debt (2011), the center of gravity between the two tends to shift back and forth over time. . . .One important theoretical innovation that these new bullion-based theories of money allowed was, as Skidelsky notes, what has come to be called the quantity theory of money (usually referred to in textbooks—since economists take endless delight in abbreviations—as QTM).

But these two perspectives are not mutually exclusive, and, depending on time, place, circumstances, and the particular problem that is the focus of attention, either of the two may be the appropriate paradigm for analysis.

The QTM argument was first put forward by a French lawyer named Jean Bodin, during a debate over the cause of the sharp, destablizing price inflation that immediately followed the Iberian conquest of the Americas. Bodin argued that the inflation was a simple matter of supply and demand: the enormous influx of gold and silver from the Spanish colonies was cheapening the value of money in Europe. The basic principle would no doubt have seemed a matter of common sense to anyone with experience of commerce at the time, but it turns out to have been based on a series of false assumptions. For one thing, most of the gold and silver extracted from Mexico and Peru did not end up in Europe at all, and certainly wasn’t coined into money. Most of it was transported directly to China and India (to buy spices, silks, calicoes, and other “oriental luxuries”), and insofar as it had inflationary effects back home, it was on the basis of speculative bonds of one sort or another. This almost always turns out to be true when QTM is applied: it seems self-evident, but only if you leave most of the critical factors out.

In the case of the sixteenth-century price inflation, for instance, once one takes account of credit, hoarding, and speculation—not to mention increased rates of economic activity, investment in new technology, and wage levels (which, in turn, have a lot to do with the relative power of workers and employers, creditors and debtors)—it becomes impossible to say for certain which is the deciding factor: whether the money supply drives prices, or prices drive the money supply.

As a matter of logic, if the value of money depends on the precious metals (gold or silver) from which coins were minted, the value of money is necessarily affected by a change in the value of the metals used to coin money. Because a large increase in the stock of gold and silver, as Graeber concedes, must reduce the value of those metals, subsequent inflation then being attributable, at least in part, to the gold and silver discoveries even if the newly mined gold and silver was shipped mainly to privately held Indian and Chinese hoards rather than minted into new coins. An exogenous increase in prices may well have caused the quantity of credit money to increase, but that is analytically distinct from the inflationary effect of a reduced value of gold or silver when, as was the case in the sixteenth century, money is legally defined as a specific weight of gold or silver.

Technically, this comes down to a choice between what are called exogenous and endogenous theories of money. Should money be treated as an outside factor, like all those Spanish dubloons supposedly sweeping into Antwerp, Dublin, and Genoa in the days of Philip II, or should it be imagined primarily as a product of economic activity itself, mined, minted, and put into circulation, or more often, created as credit instruments such as loans, in order to meet a demand—which would, of course, mean that the roots of inflation lie elsewhere?

There is no such choice, because any theory must posit certain initial conditions and definitions, which are given or exogenous to the analysis. How the theory is framed and which variables are treated as exogenous and which are treated as endogenous is a matter of judgment in light of the problem and the circumstances. Graeber is certainly correct that, in any realistic model, the quantity of money is endogenously, not exogenously, determined, but that doesn’t mean that the value of gold and silver may not usefully be treated as exogenous in a system in which money is defined as a weight of gold or silver.

To put it bluntly: QTM is obviously wrong. Doubling the amount of gold in a country will have no effect on the price of cheese if you give all the gold to rich people and they just bury it in their yards, or use it to make gold-plated submarines (this is, incidentally, why quantitative easing, the strategy of buying long-term government bonds to put money into circulation, did not work either). What actually matters is spending.

Graeber is talking in circles, failing to distinguish between the quantity theory of money – a theory about the value of a pure medium of exchange with no use except to be received in exchange — and a theory of the real value of gold and silver when money is defined as a weight of gold or silver. The value of gold (or silver) in monetary uses must be roughly equal to its value in non-monetary uses. which is determined by the total stock of gold and the demand to hold gold or to use it in coinage or for other uses (e.g., jewelry and ornamentation). An increase in the stock of gold relative to demand must reduce its value. That relationship between price and quantity is not the same as QTM. The quantity of a metallic money will increase as its value in non-monetary uses declines. If there is literally an unlimited demand for newly mined gold to be immediately sent unused into hoards, Graeber’s argument would be correct. But the fact that much of the newly mined gold initially went into hoards does not mean that all of the newly mined gold went into hoards.

In sum, Graeber is confused between the quantity theory of money and a theory of a commodity money used both as money and as a real commodity. The quantity theory of money of a pure medium of exchange posits that changes in the quantity of money cause proportionate changes in the price level. Changes in the quantity of a real commodity also used as money have nothing to do with the quantity theory of money.

Relying on a dubious account of the history of monetary theory by Skidelsky, Graeber blames the obsession of economists with the quantity theory for repeated monetary disturbances starting with the late 17th century deflation in Britain when silver appreciated relative to gold causing prices measured in silver to fall. Graeber thus fails to see that under a metallic money, real disturbances do have repercussion on the level of prices, repercussions having nothing to do with an exogenous prior change in the quantity of money.

According to Skidelsky, the pattern was to repeat itself again and again, in 1797, the 1840s, the 1890s, and, ultimately, the late 1970s and early 1980s, with Thatcher and Reagan’s (in each case brief) adoption of monetarism. Always we see the same sequence of events:

(1) The government adopts hard-money policies as a matter of principle.

(2) Disaster ensues.

(3) The government quietly abandons hard-money policies.

(4) The economy recovers.

(5) Hard-money philosophy nonetheless becomes, or is reinforced as, simple universal common sense.

There is so much indiscriminate generalization here that it is hard to know what to make of it. But the conduct of monetary policy has always been fraught, and learning has been slow and painful. We can and must learn to do better, but blanket condemnations of economics are unlikely to lead to better outcomes.

How was it possible to justify such a remarkable string of failures? Here a lot of the blame, according to Skidelsky, can be laid at the feet of the Scottish philosopher David Hume. An early advocate of QTM, Hume was also the first to introduce the notion that short-term shocks—such as Locke produced—would create long-term benefits if they had the effect of unleashing the self-regulating powers of the market:

Actually I agree that Hume, as great and insightful a philosopher as he was and as sophisticated an economic observer as he was, was an unreliable monetary theorist. And one of the reasons he was led astray was his unwarranted attachment to the quantity theory of money, an attachment that was not shared by his close friend Adam Smith.

Ever since Hume, economists have distinguished between the short-run and the long-run effects of economic change, including the effects of policy interventions. The distinction has served to protect the theory of equilibrium, by enabling it to be stated in a form which took some account of reality. In economics, the short-run now typically stands for the period during which a market (or an economy of markets) temporarily deviates from its long-term equilibrium position under the impact of some “shock,” like a pendulum temporarily dislodged from a position of rest. This way of thinking suggests that governments should leave it to markets to discover their natural equilibrium positions. Government interventions to “correct” deviations will only add extra layers of delusion to the original one.

I also agree that focusing on long-run equilibrium without regard to short-run fluctuations can lead to terrible macroeconomic outcomes, but that doesn’t mean that long-run effects are never of concern and may be safely disregarded. But just as current suffering must not be disregarded when pursuing vague and uncertain long-term benefits, ephemeral transitory benefits shouldn’t obscure serious long-term consequences. Weighing such alternatives isn’t easy, but nothing is gained by denying that the alternatives exist. Making those difficult choices is inherent in policy-making, whether macroeconomic or climate policy-making.

Although Graeber takes a valid point – that a supposed tendency toward an optimal long-run equilibrium does not justify disregard of an acute short-term problem – to an extreme, his criticism of the New Classical approach to policy-making that replaced the flawed mainstream Keynesian macroeconomics of the late 1970s is worth listening to. The New Classical approach self-consciously rejected any policy aimed at short-run considerations owing to a time-inconsistency paradox was based almost entirely on the logic of general-equilibrium theory and an illegitimate methodological argument rejecting all macroeconomic theories not rigorously deduced from the unarguable axiom of optimizing behavior by rational agents (and therefore not, in the official jargon, microfounded) as unscientific and unworthy of serious consideration in the brave New Classical world of scientific macroeconomics.

It’s difficult for outsiders to see what was really at stake here, because the argument has come to be recounted as a technical dispute between the roles of micro- and macroeconomics. Keynesians insisted that the former is appropriate to studying the behavior of individual households or firms, trying to optimize their advantage in the marketplace, but that as soon as one begins to look at national economies, one is moving to an entirely different level of complexity, where different sorts of laws apply. Just as it is impossible to understand the mating habits of an aardvark by analyzing all the chemical reactions in their cells, so patterns of trade, investment, or the fluctuations of interest or employment rates were not simply the aggregate of all the microtransactions that seemed to make them up. The patterns had, as philosophers of science would put it, “emergent properties.” Obviously, it was necessary to understand the micro level (just as it was necessary to understand the chemicals that made up the aardvark) to have any chance of understand the macro, but that was not, in itself, enough.

As an aisde, it’s worth noting that the denial or disregard of the possibility of any emergent properties by New Classical economists (of which what came to be known as New Keynesian economics is really a mildly schismatic offshoot) is nicely illustrated by the un-self-conscious alacrity with which the representative-agent approach was adopted as a modeling strategy in the first few generations of New Classical models. That New Classical theorists now insist that representative agency is not an essential to New Classical modeling is true, but the methodologically reductive nature of New Classical macroeconomics, in which all macroeconomic theories must be derived under the axiom of individually maximizing behavior except insofar as specific “frictions” are introduced by explicit assumption, is essential. (See here, here, and here)

The counterrevolutionaries, starting with Keynes’s old rival Friedrich Hayek . . . took aim directly at this notion that national economies are anything more than the sum of their parts. Politically, Skidelsky notes, this was due to a hostility to the very idea of statecraft (and, in a broader sense, of any collective good). National economies could indeed be reduced to the aggregate effect of millions of individual decisions, and, therefore, every element of macroeconomics had to be systematically “micro-founded.”

Hayek’s role in the microfoundations movement is important, but his position was more sophisticated and less methodologically doctrinaire than that of the New Classical macroeconomists, if for no other reason than that Hayek didn’t believe that macroeconomics should, or could, be derived from general-equilibrium theory. His criticism, like that of economists like Clower and Leijonhufvud, of Keynesian macroeconomics for being insufficiently grounded in microeconomic principles, was aimed at finding microeconomic arguments that could explain and embellish and modify the propositions of Keynesian macroeconomic theory. That is the sort of scientific – not methodological — reductivism that Hayek’s friend Karl Popper advocated: a theoretical and empirical challenge of reducing a higher level theory to its more fundamental foundations, e.g., when physicists and chemists search for theoretical breakthroughs that allow the propositions of chemistry to be reduced to more fundamental propositions of physics. The attempt to reduce chemistry to underlying physical principles is very different from a methodological rejection of all chemistry that cannot be derived from underlying deep physical theories.

There is probably more than a grain of truth in Graeber’s belief that there was a political and ideological subtext in the demand for microfoundations by New Classical macroeconomists, but the success of the microfoundations program was also the result of philosophically unsophisticated methodological error. How to apportion the share of blame going to mistaken methodology, professional and academic opportunism, and a hidden political agenda is a question worthy of further investigation. The easy part is to identify the mistaken methodology, which Graeber does. As for the rest, Graeber simply asserts bad faith, but with little evidence.

In Graeber’s comprehensive condemnation of modern economics, the efficient market hypothesis, being closely related to the rational-expectations hypothesis so central to New Classical economics, is not spared either. Here again, though I share and sympathize with his disdain for EMH, Graeber can’t resist exaggeration.

In other words, we were obliged to pretend that markets could not, by definition, be wrong—if in the 1980s the land on which the Imperial compound in Tokyo was built, for example, was valued higher than that of all the land in New York City, then that would have to be because that was what it was actually worth. If there are deviations, they are purely random, “stochastic” and therefore unpredictable, temporary, and, ultimately, insignificant.

Of course, no one is obliged to pretend that markets could not be wrong — and certainly not by a definition. The EMH simply asserts that the price of an asset reflects all the publicly available information. But what EMH asserts is certainly not true in many or even most cases, because people with non-public information (or with superior capacity to process public information) may affect asset prices, and such people may profit at the expense of those less knowledgeable or less competent in anticipating price changes. Moreover, those advantages may result from (largely wasted) resources devoted to acquiring and processing information, and it is those people who make fortunes betting on the future course of asset prices.

Graeber then quotes Skidelsky approvingly:

There is a paradox here. On the one hand, the theory says that there is no point in trying to profit from speculation, because shares are always correctly priced and their movements cannot be predicted. But on the other hand, if investors did not try to profit, the market would not be efficient because there would be no self-correcting mechanism. . .

Secondly, if shares are always correctly priced, bubbles and crises cannot be generated by the market….

This attitude leached into policy: “government officials, starting with [Fed Chairman] Alan Greenspan, were unwilling to burst the bubble precisely because they were unwilling to even judge that it was a bubble.” The EMH made the identification of bubbles impossible because it ruled them out a priori.

So the apparent paradox that concerns Skidelsky and Graeber dissolves upon (only a modest amount of) further reflection. Proper understanding and revision of the EMH makes it clear that bubbles can occur. But that doesn’t mean that bursting bubbles is a job that can be safely delegated to any agency, including the Fed.

Moreover, the housing bubble peaked in early 2006, two and a half years before the financial crisis in September 2008. The financial crisis was not unrelated to the housing bubble, which undoubtedly added to the fragility of the financial system and its vulnerability to macroeconomic shocks, but the main cause of the crisis was Fed policy that was unnecessarily focused on a temporary blip in commodity prices persuading the Fed not to loosen policy in 2008 during a worsening recession. That was a scenario similar to the one in 1929 when concern about an apparent stock-market bubble caused the Fed to repeatedly tighten money, raising interest rates, thereby causing a downturn and crash of asset prices triggering the Great Depression.

Graeber and Skidelsky correctly identify some of the problems besetting macroeconomics, but their indiscriminate attack on all economic theory is unlikely to improve the situation. A pity, because a focused and sophisticated critique of economics than they have served up has never been more urgently needed than it is now to enable economists to perform the modest service to mankind of which they might be capable.

Dr. Popper: Or How I Learned to Stop Worrying and Love Metaphysics

Introduction to Falsificationism

Although his reputation among philosophers was never quite as exalted as it was among non-philosophers, Karl Popper was a pre-eminent figure in 20th century philosophy. As a non-philosopher, I won’t attempt to adjudicate which take on Popper is the more astute, but I think I can at least sympathize, if not fully agree, with philosophers who believe that Popper is overrated by non-philosophers. In an excellent blog post, Phillipe Lemoine gives a good explanation of why philosophers look askance at falsificationism, Popper’s most important contribution to philosophy.

According to Popper, what distinguishes or demarcates a scientific statement from a non-scientific (metaphysical) statement is whether the statement can, or could be, disproved or refuted – falsified (in the sense of being shown to be false not in the sense of being forged, misrepresented or fraudulently changed) – by an actual or potential observation. Vulnerability to potentially contradictory empirical evidence, according to Popper, is what makes science special, allowing it to progress through a kind of dialectical process of conjecture (hypothesis) and refutation (empirical testing) leading to further conjecture and refutation and so on.

Theories purporting to explain anything and everything are thus non-scientific or metaphysical. Claiming to be able to explain too much is a vice, not a virtue, in science. Science advances by risk-taking, not by playing it safe. Trying to explain too much is actually playing it safe. If you’re not willing to take the chance of putting your theory at risk, by saying that this and not that will happen — rather than saying that this or that will happen — you’re playing it safe. This view of science, portrayed by Popper in modestly heroic terms, was not unappealing to scientists, and in part accounts for the positive reception of Popper’s work among scientists.

But this heroic view of science, as Lemoine nicely explains, was just a bit oversimplified. Theories never exist in a vacuum, there is always implicit or explicit background knowledge that informs and provides context for the application of any theory from which a prediction is deduced. To deduce a prediction from any theory, background knowledge, including complementary theories that are presumed to be valid for purposes of making a prediction, is necessary. Any prediction relies not just on a single theory but on a system of related theories and auxiliary assumptions.

So when a prediction is deduced from a theory, and the predicted event is not observed, it is never unambiguously clear which of the multiple assumptions underlying the prediction is responsible for the failure of the predicted event to be observed. The one-to-one logical dependence between a theory and a prediction upon which Popper’s heroic view of science depends doesn’t exist. Because the heroic view of science is too simplified, Lemoine considers it false, at least in the naïve and heroic form in which it is often portrayed by its proponents.

But, as Lemoine himself acknowledges, Popper was not unaware of these issues and actually dealt with some if not all of them. Popper therefore dismissed those criticisms pointing to his various acknowledgments and even anticipations of and responses to the criticisms. Nevertheless, his rhetorical style was generally not to qualify his position but to present it in stark terms, thereby reinforcing the view of his critics that he actually did espouse the naïve version of falsificationism that, only under duress, would be toned down to meet the objections raised to the usual unqualified version of his argument. Popper after all believed in making bold conjectures and framing a theory in the strongest possible terms and characteristically adopted an argumentative and polemical stance in staking out his positions.

Toned-Down Falsificationism

In his tone-downed version of falsificationism, Popper acknowledged that one can never know if a prediction fails because the underlying theory is false or because one of the auxiliary assumptions required to make the prediction is false, or even because of an error in measurement. But that acknowledgment, Popper insisted, does not refute falsificationism, because falsificationism is not a scientific theory about how scientists do science; it is a normative theory about how scientists ought to do science. The normative implication of falsificationism is that scientists should not try to shield their theories by making just-so adjustments in their theories through ad hoc auxiliary assumptions, e.g., ceteris paribus assumptions, to shield their theories from empirical disproof. Rather they should accept the falsification of their theories when confronted by observations that conflict with the implications of their theories and then formulate new and better theories to replace the old ones.

But a strict methodological rule against adjusting auxiliary assumptions or making further assumptions of an ad hoc nature would have ruled out many fruitful theoretical developments resulting from attempts to account for failed predictions. For example, the planet Neptune was discovered in 1846 by scientists who posited (ad hoc) the existence of another planet to explain why the planet Uranus did not follow its predicted path. Rather than conclude that the Newtonian theory was falsified by the failure of Uranus to follow the orbital path predicted by Newtonian theory, the French astronomer Urbain Le Verrier posited the existence of another planet that would account for the path actually followed by Uranus. Now in this case, it was possible to observe the predicted position of the new planet, and its discovery in the predicted location turned out to be a sensational confirmation of Newtonian theory.

Popper therefore admitted that making an ad hoc assumption in order to save a theory from refutation was permissible under his version of normative faslisificationism, but only if the ad hoc assumption was independently testable. But suppose that, under the circumstances, it would have been impossible to observe the existence of the predicted planet, at least with the observational tools then available, making the ad hoc assumption testable only in principle, but not in practice. Strictly adhering to Popper’s methodological requirement of being able to test independently any ad hoc assumption would have meant accepting the refutation of the Newtonian theory rather than positing the untestable — but true — ad hoc other-planet hypothesis to account for the failed prediction of the orbital path of Uranus.

My point is not that ad hoc assumptions to save a theory from falsification are ok, but to point out that a strict methodological rules requiring rejection of any theory once it appears to be contradicted by empirical evidence and prohibiting the use of any ad hoc assumption to save the theory unless the ad hoc assumption is independently testable might well lead to the wrong conclusion given the nuances and special circumstances associated with every case in which a theory seems to be contradicted by observed evidence. Such contradictions are rarely so blatant that theory cannot be reconciled with the evidence. Indeed, as Popper himself recognized, all observations are themselves understood and interpreted in the light of theoretical presumptions. It is only in extreme cases that evidence cannot be interpreted in a way that more or less conforms to the theory under consideration. At first blush, the Copernican heliocentric view of the world seemed obviously contradicted by direct sensory observation that earth seems flat and the sun rise and sets. Empirical refutation could be avoided only by providing an alternative interpretation of the sensory data that could be reconciled with the apparent — and obvious — flatness and stationarity of the earth and the movement of the sun and moon in the heavens.

So the problem with falsificationism as a normative theory is that it’s not obvious why a moderately good, but less than perfect, theory should be abandoned simply because it’s not perfect and suffers from occasional predictive failures. To be sure, if a better theory than the one under consideration is available, predicting correctly whenever the one under consideration predicts correctly and predicting more accurately than the one under consideration when the latter fails to predict correctly, the alternative theory is surely preferable, but that simply underscores the point that evaluating any theory in isolation is not very important. After all, every theory, being a simplification, is an imperfect representation of reality. It is only when two or more theories are available that scientists must try to determine which of them is preferable.

Oakeshott and the Poverty of Falsificationism

These problems with falsificationism were brought into clearer focus by Michael Oakeshott in his famous essay “Rationalism in Politics,” which though not directed at Popper himself (whose colleague at the London School of Economics he was) can be read as a critique of Popper’s attempt to prescribe methodological rules for scientists to follow in carrying out their research. Methodological rules of the kind propounded by Popper are precisely the sort of supposedly rational rules of practice intended to ensure the successful outcome of an undertaking that Oakeshott believed to be ill-advised and hopelessly naïve. The rationalist conceit in Oakesott’s view is that there are demonstrably correct answers to practical questions and that practical activity is rational only when it is based on demonstrably true moral or causal rules.

The entry on Michael Oakeshott in the Stanford Encyclopedia of Philosophy summarizes Oakeshott’s position as follows:

The error of Rationalism is to think that making decisions simply requires skill in the technique of applying rules or calculating consequences. In an early essay on this theme, Oakeshott distinguishes between “technical” and “traditional” knowledge. Technical knowledge is of facts or rules that can be easily learned and applied, even by those who are without experience or lack the relevant skills. Traditional knowledge, in contrast, means “knowing how” rather than “knowing that” (Ryle 1949). It is acquired by engaging in an activity and involves judgment in handling facts or rules (RP 12–17). The point is not that rules cannot be “applied” but rather that using them skillfully or prudently means going beyond the instructions they provide.

The idea that a scientist’s decision about when to abandon one theory and replace it with another can be reduced to the application of a Popperian falsificationist maxim ignores all the special circumstances and all the accumulated theoretical and practical knowledge that a truly expert scientist will bring to bear in studying and addressing such a problem. Here is how Oakeshott addresses the problem in his famous essay.

These two sorts of knowledge, then, distinguishable but inseparable, are the twin components of the knowledge involved in every human activity. In a practical art such as cookery, nobody supposes that the knowledge that belongs to the good cook is confined to what is or what may be written down in the cookery book: technique and what I have called practical knowledge combine to make skill in cookery wherever it exists. And the same is true of the fine arts, of painting, of music, of poetry: a high degree of technical knowledge, even where it is both subtle and ready, is one thing; the ability to create a work of art, the ability to compose something with real musical qualities, the ability to write a great sonnet, is another, and requires in addition to technique, this other sort of knowledge. Again these two sorts of knowledge are involved in any genuinely scientific activity. The natural scientist will certainly make use of observation and verification that belong to his technique, but these rules remain only one of the components of his knowledge; advances in scientific knowledge were never achieved merely by following the rules. . . .

Technical knowledge . . . is susceptible of formulation in rules, principles, directions, maxims – comprehensively, in propositions. It is possible to write down technical knowledge in a book. Consequently, it does not surprise us that when an artist writes about his art, he writes only about the technique of his art. This is so, not because he is ignorant of what may be called asesthetic element, or thinks it unimportant, but because what he has to say about that he has said already (if he is a painter) in his pictures, and he knows no other way of saying it. . . . And it may be observed that this character of being susceptible of precise formulation gives to technical knowledge at least the appearance of certainty: it appears to be possible to be certain about a technique. On the other hand, it is characteristic of practical knowledge that it is not susceptible of formulation of that kind. Its normal expression is in a customary or traditional way of doing things, or, simply, in practice. And this gives it the appearance of imprecision and consequently of uncertainty, of being a matter of opinion, of probability rather than truth. It is indeed knowledge that is expressed in taste or connoisseurship, lacking rigidity and ready for the impress of the mind of the learner. . . .

Technical knowledge, in short, an be both taught and learned in the simplest meanings of these words. On the other hand, practical knowledge can neither be taught nor learned, but only imparted and acquired. It exists only in practice, and the only way to acquire it is by apprenticeship to a master – not because the master can teach it (he cannot), but because it can be acquired only by continuous contact with one who is perpetually practicing it. In the arts and in natural science what normally happens is that the pupil, in being taught and in learning the technique from his master, discovers himself to have acquired also another sort of knowledge than merely technical knowledge, without it ever having been precisely imparted and often without being able to say precisely what it is. Thus a pianist acquires artistry as well as technique, a chess-player style and insight into the game as well as knowledge of the moves, and a scientist acquires (among other things) the sort of judgement which tells him when his technique is leading him astray and the connoisseurship which enables him to distinguish the profitable from the unprofitable directions to explore.

Now, as I understand it, Rationalism is the assertion that what I have called practical knowledge is not knowledge at all, the assertion that, properly speaking, there is no knowledge which is not technical knowledge. The Rationalist holds that the only element of knowledge involved in any human activity is technical knowledge and that what I have called practical knowledge is really only a sort of nescience which would be negligible if it were not positively mischievous. (Rationalism in Politics and Other Essays, pp. 12-16)

Almost three years ago, I attended the History of Economics Society meeting at Duke University at which Jeff Biddle of Michigan State University delivered his Presidential Address, “Statistical Inference in Economics 1920-1965: Changes in Meaning and Practice, published in the June 2017 issue of the Journal of the History of Economic Thought. The paper is a remarkable survey of the differing attitudes towards using formal probability theory as the basis for making empirical inferences from the data. The underlying assumptions of probability theory about the nature of the data were widely viewed as being too extreme to make probability theory an acceptable basis for empirical inferences from the data. However, the early negative attitudes toward accepting probability theory as the basis for making statistical inferences from data were gradually overcome (or disregarded). But as late as the 1960s, even though econometric techniques were becoming more widely accepted, a great deal of empirical work, including by some of the leading empirical economists of the time, avoided using the techniques of statistical inference to assess empirical data using regression analysis. Only in the 1970s was there a rapid sea-change in professional opinion that made statistical inference based on explicit probabilisitic assumptions about underlying data distributions the requisite technique for drawing empirical inferences from the analysis of economic data. In the final section of his paper, Biddle offers an explanation for this rapid change in professional attitude toward the use of probabilistic assumptions about data distributions as the required method of the empirical assessment of economic data.

By the 1970s, there was a broad consensus in the profession that inferential methods justified by probability theory—methods of producing estimates, of assessing the reliability of those estimates, and of testing hypotheses—were not only applicable to economic data, but were a necessary part of almost any attempt to generalize on the basis of economic data. . . .

This paper has been concerned with beliefs and practices of economists who wanted to use samples of statistical data as a basis for drawing conclusions about what was true, or probably true, in the world beyond the sample. In this setting, “mechanical objectivity” means employing a set of explicit and detailed rules and procedures to produce conclusions that are objective in the sense that if many different people took the same statistical information, and followed the same rules, they would come to exactly the same conclusions. The trustworthiness of the conclusion depends on the quality of the method. The classical theory of inference is a prime example of this sort of mechanical objectivity.

Porter [Trust in Numbers: The Pursuit of Objectivity in Science and Public Life] contrasts mechanical objectivity with an objectivity based on the “expert judgment” of those who analyze data. Expertise is acquired through a sanctioned training process, enhanced by experience, and displayed through a record of work meeting the approval of other experts. One’s faith in the analyst’s conclusions depends on one’s assessment of the quality of his disciplinary expertise and his commitment to the ideal of scientific objectivity. Elmer Working’s method of determining whether measured correlations represented true cause-and-effect relationships involved a good amount of expert judgment. So, too, did Gregg Lewis’s adjustments of the various estimates of the union/non-union wage gap, in light of problems with the data and peculiarities of the times and markets from which they came. Keynes and Persons pushed for a definition of statistical inference that incorporated space for the exercise of expert judgment; what Arthur Goldberger and Lawrence Klein referred to as ‘statistical inference’ had no explicit place for expert judgment.

Speaking in these terms, I would say that in the 1920s and 1930s, empirical economists explicitly acknowledged the need for expert judgment in making statistical inferences. At the same time, mechanical objectivity was valued—there are many examples of economists of that period employing rule-oriented, replicable procedures for drawing conclusions from economic data. The rejection of the classical theory of inference during this period was simply a rejection of one particular means for achieving mechanical objectivity. By the 1970s, however, this one type of mechanical objectivity had become an almost required part of the process of drawing conclusions from economic data, and was taught to every economics graduate student.

Porter emphasizes the tension between the desire for mechanically objective methods and the belief in the importance of expert judgment in interpreting statistical evidence. This tension can certainly be seen in economists’ writings on statistical inference throughout the twentieth century. However, it would be wrong to characterize what happened to statistical inference between the 1940s and the 1970s as a displace-ment of procedures requiring expert judgment by mechanically objective procedures. In the econometric textbooks published after 1960, explicit instruction on statistical inference was largely limited to instruction in the mechanically objective procedures of the classical theory of inference. It was understood, however, that expert judgment was still an important part of empirical economic analysis, particularly in the specification of the models to be estimated. But the disciplinary knowledge needed for this task was to be taught in other classes, using other textbooks.

And in practice, even after the statistical model had been chosen, the estimates and standard errors calculated, and the hypothesis tests conducted, there was still room to exercise a fair amount of judgment before drawing conclusions from the statistical results. Indeed, as Marcel Boumans (2015, pp. 84–85) emphasizes, no procedure for drawing conclusions from data, no matter how algorithmic or rule bound, can dispense entirely with the need for expert judgment. This fact, though largely unacknowledged in the post-1960s econometrics textbooks, would not be denied or decried by empirical economists of the 1970s or today.

This does not mean, however, that the widespread embrace of the classical theory of inference was simply a change in rhetoric. When application of classical inferential procedures became a necessary part of economists’ analyses of statistical data, the results of applying those procedures came to act as constraints on the set of claims that a researcher could credibly make to his peers on the basis of that data. For example, if a regression analysis of sample data yielded a large and positive partial correlation, but the correlation was not “statistically significant,” it would simply not be accepted as evidence that the “population” correlation was positive. If estimation of a statistical model produced a significant estimate of a relationship between two variables, but a statistical test led to rejection of an assumption required for the model to produce unbiased estimates, the evidence of a relationship would be heavily discounted.

So, as we consider the emergence of the post-1970s consensus on how to draw conclusions from samples of statistical data, there are arguably two things to be explained. First, how did it come about that using a mechanically objective procedure to generalize on the basis of statistical measures went from being a choice determined by the preferences of the analyst to a professional requirement, one that had real con-sequences for what economists would and would not assert on the basis of a body of statistical evidence? Second, why was it the classical theory of inference that became the required form of mechanical objectivity? . . .

Perhaps searching for an explanation that focuses on the classical theory of inference as a means of achieving mechanical objectivity emphasizes the wrong characteristic of that theory. In contrast to earlier forms of mechanical objectivity used by economists, such as standardized methods of time series decomposition employed since the 1920s, the classical theory of inference is derived from, and justified by, a body of formal mathematics with impeccable credentials: modern probability theory. During a period when the value placed on mathematical expression in economics was increasing, it may have been this feature of the classical theory of inference that increased its perceived value enough to overwhelm long-standing concerns that it was not applicable to economic data. In other words, maybe the chief causes of the profession’s embrace of the classical theory of inference are those that drove the broader mathematization of economics, and one should simply look to the literature that explores possible explanations for that phenomenon rather than seeking a special explanation of the embrace of the classical theory of inference.

I would suggest one more factor that might have made the classical theory of inference more attractive to economists in the 1950s and 1960s: the changing needs of pedagogy in graduate economics programs. As I have just argued, since the 1920s, economists have employed both judgment based on expertise and mechanically objective data-processing procedures when generalizing from economic data. One important difference between these two modes of analysis is how they are taught and learned. The classical theory of inference as used by economists can be taught to many students simultaneously as a set of rules and procedures, recorded in a textbook and applicable to “data” in general. This is in contrast to the judgment-based reasoning that combines knowledge of statistical methods with knowledge of the circumstances under which the particular data being analyzed were generated. This form of reasoning is harder to teach in a classroom or codify in a textbook, and is probably best taught using an apprenticeship model, such as that which ideally exists when an aspiring economist writes a thesis under the supervision of an experienced empirical researcher.

During the 1950s and 1960s, the ratio of PhD candidates to senior faculty in PhD-granting programs was increasing rapidly. One consequence of this, I suspect, was that experienced empirical economists had less time to devote to providing each interested student with individualized feedback on his attempts to analyze data, so that relatively more of a student’s training in empirical economics came in an econometrics classroom, using a book that taught statistical inference as the application of classical inference procedures. As training in empirical economics came more and more to be classroom training, competence in empirical economics came more and more to mean mastery of the mechanically objective techniques taught in the econometrics classroom, a competence displayed to others by application of those techniques. Less time in the training process being spent on judgment-based procedures for interpreting statistical results meant fewer researchers using such procedures, or looking for them when evaluating the work of others.

This process, if indeed it happened, would not explain why the classical theory of inference was the particular mechanically objective method that came to dominate classroom training in econometrics; for that, I would again point to the classical theory’s link to a general and mathematically formalistic theory. But it does help to explain why the application of mechanically objective procedures came to be regarded as a necessary means of determining the reliability of a set of statistical measures and the extent to which they provided evidence for assertions about reality. This conjecture fits in with a larger possibility that I believe is worth further exploration: that is, that the changing nature of graduate education in economics might sometimes be a cause as well as a consequence of changing research practices in economics. (pp. 167-70)

The correspondence between Biddle’s discussion of the change in the attitude of the economics profession about how inferences should be drawn from data about empirical relationships is strikingly similar to Oakeshott’s discussion and depressing in its implications for the decline of expert judgment by economics, expert judgment having been replaced by mechanical and technical knowledge that can be objectively summarized in the form of rules or tests for statistical significance, itself an entirely arbitrary convention lacking any logical, or self-evident, justification.

But my point is not to condemn using rules derived from classical probability theory to assess the significance of relationships statistically estimated from historical data, but to challenge the methodological prohibition against the kinds of expert judgments that many statistically knowledgeable economists like Nobel Prize winners such as Simon Kuznets, Milton Friedman, Theodore Schultz and Gary Becker routinely used to make in their empirical studies. As Biddle notes:

In 1957, Milton Friedman published his theory of the consumption function. Friedman certainly understood statistical theory and probability theory as well as anyone in the profession in the 1950s, and he used statistical theory to derive testable hypotheses from his economic model: hypotheses about the relationships between estimates of the marginal propensity to consume for different groups and from different types of data. But one will search his book almost in vain for applications of the classical methods of inference. Six years later, Friedman and Anna Schwartz published their Monetary History of the United States, a work packed with graphs and tables of statistical data, as well as numerous generalizations based on that data. But the book contains no classical hypothesis tests, no confidence intervals, no reports of statistical significance or insignificance, and only a handful of regressions. (p. 164)

Friedman’s work on the Monetary History is still regarded as authoritative. My own view is that much of the Monetary History was either wrong or misleading. But my quarrel with the Monetary History mainly pertains to the era in which the US was on the gold standard, inasmuch as Friedman simply did not understand how the gold standard worked, either in theory or in practice, as McCloskey and Zecher showed in two important papers (here and here). Also see my posts about the empirical mistakes in the Monetary History (here and here). But Friedman’s problem was bad monetary theory, not bad empirical technique.

Friedman’s theoretical misunderstandings have no relationship to the misguided prohibition against doing quantitative empirical research without obeying the arbitrary methodological requirement that statistical be derived in a way that measures the statistical significance of the estimated relationships. These methodological requirements have been adopted to support a self-defeating pretense to scientific rigor, necessitating the use of relatively advanced mathematical techniques to perform quantitative empirical research. The methodological requirements for measuring statistical relationships were never actually shown to be generate more accurate or reliable statistical results than those derived from the less technically advanced, but in some respects more economically sophisticated, techniques that have almost totally been displaced. One more example of the fallacy that there is but one technique of research that ensures the discovery of truth, a mistake even Popper was never guilty of.

Methodological Prescriptions Go from Bad to Worse

The methodological requirement for the use of formal tests of statistical significance before any quantitative statistical estimate could be credited was a prelude, though it would be a stretch to link them causally, to another and more insidious form of methodological tyrannizing: the insistence that any macroeconomic model be derived from explicit micro-foundations based on the solution of an intertemporal-optimization exercise. Of course, the idea that such a model was in any way micro-founded was a pretense, the solution being derived only through the fiction of a single representative agent, rendering the entire optimization exercise fundamentally illegitimate and the exact opposite of micro-founded model. Having already explained in previous posts why transforming microfoundations from a legitimate theoretical goal into methodological necessity has taken a generation of macroeconomists down a blind alley (here, here, here, and here) I will only make the further comment that this is yet another example of the danger of elevating technique over practice and substance.

Popper’s More Important Contribution

This post has largely concurred with the negative assessment of Popper’s work registered by Lemoine. But I wish to end on a positive note, because I have learned a great deal from Popper, and even if he is overrated as a philosopher of science, he undoubtedly deserves great credit for suggesting falsifiability as the criterion by which to distinguish between science and metaphysics. Even if that criterion does not hold up, or holds up only when qualified to a greater extent than Popper admitted, Popper made a hugely important contribution by demolishing the startling claim of the Logical Positivists who in the 1920s and 1930s argued that only statements that can be empirically verified through direct or indirect observation have meaning, all other statements being meaningless or nonsensical. That position itself now seems to verge on the nonsensical. But at the time many of the world’s leading philosophers, including Ludwig Wittgenstein, no less, seemed to accept that remarkable view.

Thus, Popper’s demarcation between science and metaphysics had a two-fold significance. First, that it is not verifiability, but falsifiability, that distinguishes science from metaphysics. That’s the contribution for which Popper is usually remembered now. But it was really the other aspect of his contribution that was more significant: that even metaphysical, non-scientific, statements can be meaningful. According to the Logical Positivists, unless you are talking about something that can be empirically verified, you are talking nonsense. In other words they were deliberately hoisting themselves on their petard, because their discussions about what is and what is not meaningful, being discussions about concepts, not empirically verifiable objects, were themselves – on the Positivists’ own criterion of meaning — meaningless and nonsensical.

Popper made the world safe for metaphysics, and the world is a better place as a result. Science is a wonderful enterprise, rewarding for its own sake and because it contributes to the well-being of many millions of human beings, though like many other human endeavors, it can also have unintended and unfortunate consequences. But metaphysics, because it was used as a term of abuse by the Positivists, is still, too often, used as an epithet. It shouldn’t be.

Certainly economists should aspire to tease out whatever empirical implications they can from their theories. But that doesn’t mean that an economic theory with no falsifiable implications is useless, a judgment whereby Mark Blaug declared general equilibrium theory to be unscientific and useless, a judgment that I don’t think has stood the test of time. And even if general equilibrium theory is simply metaphysical, my response would be: so what? It could still serve as a source of inspiration and insight to us in framing other theories that may have falsifiable implications. And even if, in its current form, a theory has no empirical content, there is always the possibility that, through further discussion, critical analysis and creative thought, empirically falsifiable implications may yet become apparent.

Falsifiability is certainly a good quality for a theory to have, but even an unfalsifiable theory may be worth paying attention to and worth thinking about.

More on Sticky Wages

It’s been over four and a half years since I wrote my second most popular post on this blog (“Why are Wages Sticky?”). Although the post was linked to and discussed by Paul Krugman (which is almost always a guarantee of getting a lot of traffic) and by other econoblogosphere standbys like Mark Thoma and Barry Ritholz, unlike most of my other popular posts, it has continued ever since to attract a steady stream of readers. It’s the posts that keep attracting readers long after their original expiration date that I am generally most proud of.

I made a few preliminary points about wage stickiness before getting to my point. First, although Keynes is often supposed to have used sticky wages as the basis for his claim that market forces, unaided by stimulus to aggregate demand, cannot automatically eliminate cyclical unemployment within the short or even medium term, he actually devoted a lot of effort and space in the General Theory to arguing that nominal wage reductions would not increase employment, and to criticizing economists who blamed unemployment on nominal wages fixed by collective bargaining at levels too high to allow all workers to be employed. So, the idea that wage stickiness is a Keynesian explanation for unemployment doesn’t seem to me to be historically accurate.

I also discussed the search theories of unemployment that in some ways have improved our understanding of why some level of unemployment is a normal phenomenon even when people are able to find jobs fairly easily and why search and unemployment can actually be productive, enabling workers and employers to improve the matches between the skills and aptitudes that workers have and the skills and aptitudes that employers are looking for. But search theories also have trouble accounting for some basic facts about unemployment.

First, a lot of job search takes place when workers have jobs while search theories assume that workers can’t or don’t search while they are employed. Second, when unemployment rises in recessions, it’s not because workers mistakenly expect more favorable wage offers than employers are offering and mistakenly turn down job offers that they later regret not having accepted, which is a very skewed way of interpreting what happens in recessions; it’s because workers are laid off by employers who are cutting back output and idling production lines.

I then suggested the following alternative explanation for wage stickiness:

Consider the incentive to cut price of a firm that can’t sell as much as it wants [to sell] at the current price. The firm is off its supply curve. The firm is a price taker in the sense that, if it charges a higher price than its competitors, it won’t sell anything, losing all its sales to competitors. Would the firm have any incentive to cut its price? Presumably, yes. But let’s think about that incentive. Suppose the firm has a maximum output capacity of one unit, and can produce either zero or one units in any time period. Suppose that demand has gone down, so that the firm is not sure if it will be able to sell the unit of output that it produces (assume also that the firm only produces if it has an order in hand). Would such a firm have an incentive to cut price? Only if it felt that, by doing so, it would increase the probability of getting an order sufficiently to compensate for the reduced profit margin at the lower price. Of course, the firm does not want to set a price higher than its competitors, so it will set a price no higher than the price that it expects its competitors to set.

Now consider a different sort of firm, a firm that can easily expand its output. Faced with the prospect of losing its current sales, this type of firm, unlike the first type, could offer to sell an increased amount at a reduced price. How could it sell an increased amount when demand is falling? By undercutting its competitors. A firm willing to cut its price could, by taking share away from its competitors, actually expand its output despite overall falling demand. That is the essence of competitive rivalry. Obviously, not every firm could succeed in such a strategy, but some firms, presumably those with a cost advantage, or a willingness to accept a reduced profit margin, could expand, thereby forcing marginal firms out of the market.

Workers seem to me to have the characteristics of type-one firms, while most actual businesses seem to resemble type-two firms. So what I am suggesting is that the inability of workers to take over the jobs of co-workers (the analog of output expansion by a firm) when faced with the prospect of a layoff means that a powerful incentive operating in non-labor markets for price cutting in response to reduced demand is not present in labor markets. A firm faced with the prospect of being terminated by a customer whose demand for the firm’s product has fallen may offer significant concessions to retain the customer’s business, especially if it can, in the process, gain an increased share of the customer’s business. A worker facing the prospect of a layoff cannot offer his employer a similar deal. And requiring a workforce of many workers, the employer cannot generally avoid the morale-damaging effects of a wage cut on his workforce by replacing current workers with another set of workers at a lower wage than the old workers were getting.

I think that what I wrote four years ago is clearly right, identifying an important reason for wage stickiness. But there’s also another reason that I didn’t mention then, but whose importance has since come to appear increasingly significant to me, especially as a result of writing and rewriting my paper “Hayek, Hicks, Radner and three concepts of intertemporal equilibrium.”

If you are unemployed because the demand for your employer’s product has gone down, and your employer, planning to reduce output, is laying off workers no longer needed, how could you, as an individual worker, unconstrained by a union collective-bargaining agreement or by a minimum-wage law, persuade your employer not to lay you off? Could you really keep your job by offering to accept a wage cut — no matter how big? If you are being laid off because your employer is reducing output, would your offer to work at a lower wage cause your employer to keep output unchanged, despite a reduction in demand? If not, how would your offer to take a pay cut help you keep your job? Unless enough workers are willing to accept a big enough wage cut for your employer to find it profitable to maintain current output instead of cutting output, how would your own willingness to accept a wage cut enable you to keep your job?

Now, if all workers were to accept a sufficiently large wage cut, it might make sense for an employer not to carry out a planned reduction in output, but the offer by any single worker to accept a wage cut certainly would not cause the employer to change its output plans. So, if you are making an independent decision whether to offer to accept a wage cut, and other workers are making their own independent decisions about whether to accept a wage cut, would it be rational for you or any of them to accept a wage cut? Whether it would or wouldn’t might depend on what each worker was expecting other workers to do. But certainly given the expectation that other workers are not offering to accept a wage cut, why would it make any sense for any worker to be the one to offer to accept a wage cut? Would offering to accept a wage cut, increase the likelihood that a worker would be one of the lucky ones chosen not to be laid off? Why would offering to accept a wage cut that no one else was offering to accept, make the worker willing to work for less appear more desirable to the employer than the others that wouldn’t accept a wage cut? One reaction by the employer might be: what’s this guy’s problem?

Combining this way of looking at the incentives workers have to offer to accept wage reductions to keep their jobs with my argument in my post of four years ago, I now am inclined to suggest that unemployment as such provides very little incentive for workers and employers to cut wages. Price cutting in periods of excess supply is often driven by aggressive price cutting by suppliers with large unsold inventories. There may be lots of unemployment, but no one is holding a large stock of unemployed workers, and no is in a position to offer low wages to undercut the position of those currently employed at  nominal wages that, arguably, are too high.

That’s not how labor markets operate. Labor markets involve matching individual workers and individual employers more or less one at a time. If nominal wages fall, it’s not because of an overhang of unsold labor flooding the market; it’s because something is changing the expectations of workers and employers about what wage will be offered by employers, and accepted by workers, for a particular kind of work. If the expected wage is too high, not all workers willing to work at that wage will find employment; if it’s too low, employers will not be able to find as many workers as they would like to hire, but the situation will not change until wage expectations change. And the reason that wage expectations change is not because the excess demand for workers causes any immediate pressure for nominal wages to rise.

The further point I would make is that the optimal responses of workers and the optimal responses of their employers to a recessionary reduction in demand, in which the employers, given current input and output prices, are planning to cut output and lay off workers, are mutually interdependent. While it is, I suppose, theoretically possible that if enough workers decided to immediately offer to accept sufficiently large wage cuts, some employers might forego plans to lay off their workers, there are no obvious market signals that would lead to such a response, because such a response would be contingent on a level of coordination between workers and employers and a convergence of expectations about future outcomes that is almost unimaginable.

One can’t simply assume that it is in the independent self-interest of every worker to accept a wage cut as soon as an employer perceives a reduced demand for its product, making the current level of output unprofitable. But unless all, or enough, workers decide to accept a wage cut, the optimal response of the employer is still likely to be to cut output and lay off workers. There is no automatic mechanism by which the market adjusts to demand shocks to achieve the set of mutually consistent optimal decisions that characterizes a full-employment market-clearing equilibrium. Market-clearing equilibrium requires not merely isolated price and wage cuts by individual suppliers of inputs and final outputs, but a convergence of expectations about the prices of inputs and outputs that will be consistent with market clearing. And there is no market mechanism that achieves that convergence of expectations.

So, this brings me back to Keynes and the idea of sticky wages as the key to explaining cyclical fluctuations in output and employment. Keynes writes at the beginning of chapter 19 of the General Theory.

For the classical theory has been accustomed to rest the supposedly self-adjusting character of the economic system on an assumed fluidity of money-wages; and, when there is rigidity, to lay on this rigidity the blame of maladjustment.

A reduction in money-wages is quite capable in certain circumstances of affording a stimulus to output, as the classical theory supposes. My difference from this theory is primarily a difference of analysis. . . .

The generally accept explanation is . . . quite a simple one. It does not depend on roundabout repercussions, such as we shall discuss below. The argument simply is that a reduction in money wages will, cet. par. Stimulate demand by diminishing the price of the finished product, and will therefore increase output, and will therefore increase output and employment up to the point where  the reduction which labour has agreed to accept in its money wages is just offset by the diminishing marginal efficiency of labour as output . . . is increased. . . .

It is from this type of analysis that I fundamentally differ.

[T]his way of thinking is probably reached as follows. In any given industry we have a demand schedule for the product relating the quantities which can be sold to the prices asked; we have a series of supply schedules relating the prices which will be asked for the sale of different quantities. .  . and these schedules between them lead up to a further schedule which, on the assumption that other costs are unchanged . . . gives us the demand schedule for labour in the industry relating the quantity of employment to different levels of wages . . . This conception is then transferred . . . to industry as a whole; and it is supposed, by a parity of reasoning, that we have a demand schedule for labour in industry as a whole relating the quantity of employment to different levels of wages. It is held that it makes no material difference to this argument whether it is in terms of money-wages or of real wages. If we are thinking of real wages, we must, of course, correct for changes in the value of money; but this leaves the general tendency of the argument unchanged, since prices certainly do not change in exact proportion to changes in money wages.

If this is the groundwork of the argument . . ., surely it is fallacious. For the demand schedules for particular industries can only be constructed on some fixed assumption as to the nature of the demand and supply schedules of other industries and as to the amount of aggregate effective demand. It is invalid, therefore, to transfer the argument to industry as a whole unless we also transfer our assumption that the aggregate effective demand is fixed. Yet this assumption amount to an ignoratio elenchi. For whilst no one would wish to deny the proposition that a reduction in money-wages accompanied by the same aggregate demand as before will be associated with an increase in employment, the precise question at issue is whether the reduction in money wages will or will not be accompanied by the same aggregate effective demand as before measured in money, or, at any rate, measured by an aggregate effective demand which is not reduced in full proportion to the reduction in money-wages. . . But if the classical theory is not allowed to extend by analogy its conclusions in respect of a particular industry to industry as a whole, it is wholly unable to answer the question what effect on employment a reduction in money-wages will have. For it has no method of analysis wherewith to tackle the problem. (General Theory, pp. 257-60)

Keynes’s criticism here is entirely correct. But I would restate slightly differently. Standard microeconomic reasoning about preferences, demand, cost and supply is partial-equilbriium analysis. The focus is on how equilibrium in a single market is achieved by the adjustment of the price in a single market to equate the amount demanded in that market with amount supplied in that market.

Supply and demand is a wonderful analytical tool that can illuminate and clarify many economic problems, providing the key to important empirical insights and knowledge. But supply-demand analysis explicitly – but too often without realizing its limiting implications – assumes that other prices and incomes in other markets are held constant. That assumption essentially means that the market – i.e., the demand, cost and supply curves used to represent the behavioral characteristics of the market being analyzed – is small relative to the rest of the economy, so that changes in that single market can be assumed to have a de minimus effect on the equilibrium of all other markets. (The conditions under which such an assumption could be justified are themselves not unproblematic, but I am now assuming that those problems can in fact be assumed away at least in many applications. And a good empirical economist will have a good instinctual sense for when it’s OK to make the assumption and when it’s not OK to make the assumption.)

So, the underlying assumption of microeconomics is that the individual markets under analysis are very small relative to the whole economy. Why? Because if those markets are not small, we can’t assume that the demand curves, cost curves, and supply curves end up where they started. Because a high price in one market may have effects on other markets and those effects will have further repercussions that move the very demand, cost and supply curves that were drawn to represent the market of interest. If the curves themselves are unstable, the ability to predict the final outcome is greatly impaired if not completely compromised.

The working assumption of the bread and butter partial-equilibrium analysis that constitutes econ 101 is that markets have closed borders. And that assumption is not always valid. If markets have open borders so that there is a lot of spillover between and across markets, the markets can only be analyzed in terms of broader systems of simultaneous equations, not the simplified solutions that we like to draw in two-dimensional space corresponding to intersections of stable supply curves with stable supply curves.

What Keynes was saying is that it makes no sense to draw a curve representing the demand of an entire economy for labor or a curve representing the supply of labor of an entire economy, because the underlying assumption of such curves that all other prices are constant cannot possibly be satisfied when you are drawing a demand curve and a supply curve for an input that generates more than half the income earned in an economy.

But the problem is even deeper than just the inability to draw a curve that meaningfully represents the demand of an entire economy for labor. The assumption that you can model a transition from one point on the curve to another point on the curve is simply untenable, because not only is the assumption that other variables are being held constant untenable and self-contradictory, the underlying assumption that you are starting from an equilibrium state is never satisfied when you are trying to analyze a situation of unemployment – at least if you have enough sense not to assume that economy is starting from, and is not always in, a state of general equilibrium.

So, Keynes was certainly correct to reject the naïve transfer of partial equilibrium theorizing from its legitimate field of applicability in analyzing the effects of small parameter changes on outcomes in individual markets – what later came to be known as comparative statics – to macroeconomic theorizing about economy-wide disturbances in which the assumptions underlying the comparative-statics analysis used in microeconomics are clearly not satisfied. That illegitimate transfer of one kind of theorizing to another has come to be known as the demand for microfoundations in macroeconomic models that is the foundational methodological principle of modern macroeconomics.

The principle, as I have been arguing for some time, is illegitimate for a variety of reasons. And one of those reasons is that microeconomics itself is based on the macroeconomic foundational assumption of a pre-existing general equilibrium, in which all plans in the entire economy are, and will remain, perfectly coordinated throughout the analysis of a particular parameter change in a single market. Once you relax the assumption that all, but one, markets are in equilibrium, the discipline imposed by the assumption of the rationality of general equilibrium and comparative statics is shattered, and a different kind of theorizing must be adopted to replace it.

The search for that different kind of theorizing is the challenge that has always faced macroeconomics. Despite heroic attempts to avoid facing that challenge and pretend that macroeconomics can be built as if it were microeconomics, the search for a different kind of theorizing will continue; it must continue. But it would certainly help if more smart and creative people would join in that search.

The Standard Narrative on the History of Macroeconomics: An Exercise in Self-Serving Apologetics

During my recent hiatus from blogging, I have been pondering an important paper presented in June at the History of Economics Society meeting in Toronto, “The Standard Narrative on History of Macroeconomics: Central Banks and DSGE Models” by Francesco Sergi of the University of Bristol, which was selected by the History of Economics Society as the best conference paper by a young scholar in 2017.

Here is the abstract of Sergi’s paper:

How do macroeconomists write the history of their own discipline? This article provides a careful reconstruction of the history of macroeconomics told by the practitioners working today in the dynamic stochastic general equilibrium (DSGE) approach.

Such a tale is a “standard narrative”: a widespread and “standardizing” view of macroeconomics as a field evolving toward “scientific progress”. The standard narrative explains scientific progress as resulting from two factors: “consensus” about theory and “technical change” in econometric tools and computational power. This interpretation is a distinctive feature of central banks’ technical reports about their DSGE models.

Furthermore, such a view on “consensus” and “technical change” is a significantly different view with respect to similar tales told by macroeconomists in the past — which rather emphasized the role of “scientific revolutions” and struggles among competing “schools of thought”. Thus, this difference raises some new questions for historians of macroeconomics.

Sergi’s paper is too long and too rich in content to easily summarize in this post, so what I will do is reproduce and comment on some of the many quotations provided by Sergi, taken mostly from central-bank reports, but also from some leading macroeconomic textbooks and historical survey papers, about the “progress” of modern macroeconomics, and especially about the critical role played by “microfoundations” in achieving that progress. The general tenor of the standard narrative is captured well by the following quotations from V. V. Chari

[A]ny interesting model must be a dynamic stochastic general equilibrium model. From this perspective, there is no other game in town. […] A useful aphorism in macroeconomics is: “If you have an interesting and coherent story to tell, you can tell it in a DSGE model.  (Chari 2010, 2)

I could elaborate on this quotation at length, but I will just leave it out there for readers to ponder with a link to an earlier post of mine about methodological arrogance. Instead I will focus on two other sections of Sergi’s paper “the five steps of theoretical progress” and “microfoundations as theoretical progress.” Here is how Sergi explains the role of the five steps:

The standard narrative provides a detailed account of the progressive evolution toward the synthesis. Following a teleological perspective, each step of this evolution is an incremental, linear improvement of the theoretical tool box for model building. The standard narrative identifies five steps . . . .  Each step corresponds to the emergence of a school of thought. Therefore, in the standard narrative, there are not such things as competing schools of thought and revolutions. Firstly, because schools of thought are represented as a sequence; one school (one step) is always leading to another school (the following step), hence different schools are not coexisting for a long period of time. Secondly, there are no revolutions because, while emerging, new schools of thought [do] not overthrow the previous ones; instead, they suggest improvements and amendments, that are accepted as an improvement by pre-existing schools therefore, accumulation of knowledge takes place thanks to consensus. (pp. 17-18)

The first step in the standard narrative is the family of Keynesian macroeconometric models of the 1950s and 1960s, the primitive ancestors of the modern DSGE models. The second step was the emergence of New Classical macroeconomics which introduced the ideas of rational expectations and dynamic optimization into theoretical macroeconomic discourse in the 1970s. The third step was the development, inspired by New Classical ideas, of Real-Business-Cycle models of the 1980s, and the fourth step was introduction of New Keynesian models in the late 1980s and 1990s that tweaked the Real-Business-Cycle models in ways that rationalized the use of counter-cyclical macroeconomic policy within the theoretical framework of the Real-Business-Cycle approach. The final step, the DSGE model, emerged more or less naturally as a synthesis of the converging Real-Business-Cycle and New Keynesian approaches.

After detailing the five steps of theoretical progress, Sergi focuses attention on “the crucial improvement” that allowed the tool box of macroeconomic modelling to be extended in such a theoretically fruitful way: the insistence on providing explicit microfoundations for macroeconomic models. He writes:

Abiding [by] the Lucasian microfoundational program is put forward by DSGE modellers as the very fundamental essence of theoretical progress allowed by [the] consensus. As Sanajay K. Chugh (University of Pennsylvania) explains in the historical chapter of his textbook, microfoundations is all what modern macroeconomics is about: (p. 20)

Modern macroeconomics begin by explicitly studying the microeconomic principles of utility maximization, profit maximization and market-clearing. [. . . ] This modern macroeconomics quickly captured the attention of the profession through the 1980s [because] it actually begins with microeconomic principles, which was a rather attractive idea. Rather than building a framework of economy-wide events from the top down [. . .] one could build this framework using microeconomic discipline from the bottom up. (Chugh 2015, 170)

Chugh’s rationale for microfoundations is a naïve expression of reductionist bias dressed up as simple homespun common-sense. Everyone knows that you should build from the bottom up, not from the top down, right? But things are not always quite as simple as they seem. Here is an attempt to present microfoundations as being cutting-edge and sophisticated offered in a 2009 technical report written by Cuche-Curti et al. for the Swiss National Bank.

The key property of DSGE models is that they rely on explicit micro-foundations and a rational treatment of expectations in a general equilibrium context. They thus provide a coherent and compelling theoretical framework for macroeconomic analysis. (Cuche-Curti et al. 2009, 6)

A similar statement is made by Gomes et al in a 2010 technical report for the European Central Bank:

The microfoundations of the model together with its rich structure allow [us] to conduct a quantitative analysis in a theoretically coherent and fully consistent model setup, clearly spelling out all the policy implications. (Gomes et al. 2010, 5)

These laudatory descriptions of the DSGE model stress its “coherence” as a primary virtue. What is meant by “coherence” is spelled out more explicitly in a 2006 technical report describing NEMO, a macromodel of the Norwegian economy, by Brubakk et al. for the Norges Bank.

Various agents’ behavior is modelled explicitly in NEMO, based on microeconomic theory. A consistent theoretical framework makes it easier to interpret relationships and mechanisms in the model in the light of economic theory. One advantage is that we can analyse the economic effects of changes of a more structural nature […] [making it] possible to provide a consistent and detailed economic rationale for Norges Bank’s projections for the Norwegian economy. This distinguishes NEMO from purely statistical models, which to a limited extent provide scope for economic interpretations. (Brubakk and Sveen 2009, 39)

By creating microfounded models, in which all agents are optimizers making choices consistent with the postulates of microeconomic theory, DSGE model-builders, in effect, create “laboratories” from which to predict the consequences of alternative monetary policies, enabling policy makers to make informed policy choices. I pause merely to note and draw attention to the tendentious and misleading misappropriation of the language of empirical science by these characteristically self-aggrandizing references to DSGE models as “laboratories” as if what was going on in such models was determined by an actual physical process, as is routinely the case in the laboratories of physical and natural scientists, rather than speculative exercises in high-level calculations derived from the manipulation of DSGE models.

As a result of recent advances in macroeconomic theory and computational techniques, it has become feasible to construct richly structured dynamic stochastic general equilibrium models and use them as laboratories for the study of business cycles and for the formulation and analysis of monetary policy. (Cuche-Curri et al. 2009, 39)

Policy makers can be confident that the conditional predictions corresponding to the policy alternative under consideration, which are derived from their “laboratory” DSGE models, because those models, having been constructed on the basis of the postulates of economic theory, are therefore microfounded, embodying deep structural parameters that are invariant to policy changes. Microfounded models are thus immune to the Lucas Critique of macroeconomic policy evaluation, under which the empirically estimated coefficients of traditional Keynesian macroeconometric models cannot be assumed to remain constant under policy changes, because those coefficient estimates are themselves conditional to policy choices.

Here is how the point is made in three different central bank technical reports: by Argov et al. in a 2012 technical report about MOISE, a DSGE model for the Israeli economy, by Cuche-Curti et al. and by Medina and Soto in a 2006 technical report about a new DSGE model for the Chilean economy for the Central Bank of Chile.

Being micro-founded, the model enables the central bank to assess the effect of its alternative policy choices on the future paths of the economy’s endogenous variables, in a way that is immune to the Lucas critique. (Argov et al. 2012, 5)

[The DSGE] approach has three distinct advantages in comparison to other modelling strategies. First and foremost, its microfoundations should allow it to escape the Lucas critique. (Cuche-Curti et al. 2009, 6)

The main advantage of this type of model, over more traditional reduce-form macro models, is that the structural interpretation of their parameters allows [it] to overcome the Lucas Critique. This is clearly an advantage for policy analysis. (Medina and Soto, 2006, 2)

These quotations show clearly that escaping, immunizing, or overcoming the Lucas Critique is viewed by DSGE modelers as the holy grail of macroeconomic model building and macroeconomic policy analysis. If the Lucas Critique cannot be neutralized, the coefficient estimates derived from reduced-form macroeconometric models cannot be treated as invariant to policy and therefore cannot provide a secure basis for predicting the effects of alternative policies. But DSGE models allow deep structural relationships, reflecting the axioms underlying microeconomic theory, to be estimated. Because they reflect the deep, and presumably stable, microeconomic structure of the economy, estimates of deep parameters derived from DSGE models, DSGE modelers claim that these estimates provide policy makers with a reliable basis for conditional forecasting of the effects of macroeconomic policy.

Because of the consistently poor track record of DSGE models in actual forecasting (for evidence of that poor track record see the paper by Carlaw and Lipsey and my post about their paper) comparing the predictive performance of DSGE models with more traditional macroeconometric models), the emphasis placed on the Lucas Critique by DSGE modelers has an apologetic character, DSGE modelers having to account for the relatively poor comparative predictive power of DSGE models by relentlessly invoking the Lucas Critique in trying to account for, and explain away, the poor predictive performance of the DSGE models. But if DSGE models really are better than traditional macro models why are their unconditional predictions not at least as good as those of traditional macroeconometric models? Obviously estimates of the deep structural relationships provided by microfounded models are not as reliable as DSGE apologetics tries to suggest.

And the reason that the estimates of deep structural relationships derived from DSGE models are not reliable is that those models, no less than traditional macroeconometric models, are subject to the Lucas Critique, the deep microeconomic structural relationships embodied in DSGE models being conditional on the existence of a unique equilibrium solution that persists long enough for the structural relationships characterizing that equilibrium to be inferred from the data-generating mechanism whereby those models are estimated. (I have made this point previously here.) But if the data-generating mechanism does not conform to the unique general equilibrium upon whose existence the presumed deep structural relationships of microeconomic theory embodied in DSGE models are conditioned, the econometric estimates derived from DSGE models cannot capture the desired deep structural relationships, and the resulting structural estimates are therefore incapable of providing a reliable basis for macroeconomic-policy analysis or for conditional forecasts of the effects of alternative policies, much less unconditional forecasts of endogenous macroeconomic variables.

Of course, the problem is even more intractable than the discussion above implies, because there is no reason why the deep structural relationships corresponding to a particular equilibrium should be invariant to changes in the equilibrium. So any change in economic policy that displaces a pre-existing equilibrium, let alone any other unforeseen technological change or change in tastes or resource endowments that displaces a pre-existing equilibrium will necessarily cause all the deep structural relationships to change correspondingly. So the deep structural parameters upon whose invariance the supposedly unique capacity of DSGE models to provide policy analysis upon which policy makers can rely simply don’t exist. Policy making based on DSGE models is as much an uncertain art requiring the exercise of finely developed judgment and intuition as policy making based on any other kind of economic modeling. DSGE models provide no uniquely reliable basis for making macroeconomic policy.

References

Argov, E., Barnea, E., Binyamini, A., Borenstein, E., Elkayam, D., and Rozenshtrom, I. (2012). MOISE: A DSGE Model for the Israeli Economy. Technical Report 2012.06, Bank of Israel.
Brubakk, L.,Husebø, T. A., Maih, J., Olsen, K., and Østnor, M. (2006). Finding NEMO: Documentation of the Norwegian economy model. Technical Report 2006/6, Norges Bank, Staff Memo.
Carlaw, K. I., and Lipsey, R. G. (2012). “Does History Matter?: Empirical Analysis of Evolutionary versus Stationary Equilibrium Views of the Economy.” Journal of Evolutionary Economics. 22(4):735-66.
Chari, V. V. (2010). Testimony before the committee on Science and Technology, Subcommittee on Investigations and Oversight, US House of Representatives. In Building a Science of Economics for the Real World.
Chugh, S. K. (2015). Modern Macroeconomics. MIT Press, Cambridge (MA).
Cuche-Curti, N. A., Dellas, H., and Natal, J.-M. (2009). DSGE-CH. A Dynamic Stochastic General Equilibrium Model for Switzerland. Technical Report 5, Swiss National Bank.
Gomes, S., Jacquinot, P., and Pisani, M. (2010). The EAGLE. A Model for Policy Analysis of Macroeconomic Interdependence in the Euro Area. Technical Report 1195, European Central Bank.
Medina, J. P. and Soto, C. (2006). Model for Analysis and Simulations (MAS): A New DSGE Model for the Chilean Economy. Technical report, Central Bank of Chile.

Paul Romer on Modern Macroeconomics, Or, the “All Models Are False” Dodge

Paul Romer has been engaged for some time in a worthy campaign against the travesty of modern macroeconomics. A little over a year ago I commented favorably about Romer’s takedown of Robert Lucas, but I also defended George Stigler against what I thought was an unfair attempt by Romer to identify George Stigler as an inspiration and role model for Lucas’s transgressions. Now just a week ago, a paper based on Romer’s Commons Memorial Lecture to the Omicron Delta Epsilon Society, has become just about the hottest item in the econ-blogosophere, even drawing the attention of Daniel Drezner in the Washington Post.

I have already written critically about modern macroeconomics in my five years of blogging, and here are some links to previous posts (link, link, link, link). It’s good to see that Romer is continuing to voice his criticisms, and that they are gaining a lot of attention. But the macroeconomic hierarchy is used to criticism, and has its standard responses to criticism, which are being dutifully deployed by defenders of the powers that be.

Romer’s most effective rhetorical strategy is to point out that the RBC core of modern DSGE models posit unobservable taste and technology shocks to account for fluctuations in the economic time series, but that these taste and technology shocks are themselves simply inferred from the fluctuations in the times-series data, so that the entire structure of modern macroeconometrics is little more than an elaborate and sophisticated exercise in question-begging.

In this post, I just want to highlight one of the favorite catch-phrases of modern macroeconomics which serves as a kind of default excuse and self-justification for the rampant empirical failures of modern macroeconomics (documented by Lipsey and Carlaw as I showed in this post). When confronted by evidence that the predictions of their models are wrong, the standard and almost comically self-confident response of the modern macroeconomists is: All models are false. By which the modern macroeconomists apparently mean something like: “And if they are all false anyway, you can’t hold us accountable, because any model can be proven wrong. What really matters is that our models, being microfounded, are not subject to the Lucas Critique, and since all other models than ours are not micro-founded, and, therefore, being subject to the Lucas Critique, they are simply unworthy of consideration. This is what I have called methodological arrogance. That response is simply not true, because the Lucas Critique applies even to micro-founded models, those models being strictly valid only in equilibrium settings and being unable to predict the adjustment of economies in the transition between equilibrium states. All models are subject to the Lucas Critique.

Here is Romer’s take:

In response to the observation that the shocks are imaginary, a standard defense invokes Milton Friedman’s (1953) methodological assertion from unnamed authority that “the more significant the theory, the more unrealistic the assumptions (p.14).” More recently, “all models are false” seems to have become the universal hand-wave for dismissing any fact that does not conform to the model that is the current favorite.

Friedman’s methodological assertion would have been correct had Friedman substituted “simple” for “unrealistic.” Sometimes simplifications are unrealistic, but they don’t have to be. A simplification is a generalization of something complicated. By simplifying, we can transform a problem that had been too complex to handle into a problem more easily analyzed. But such simplifications aren’t necessarily unrealistic. To say that all models are false is simply a dodge to avoid having to account for failure. The excuse of course is that all those other models are subject to the Lucas Critique, so my model wins. But your model is subject to the Lucas Critique even though you claim it’s not, so even according to the rules you have arbitrarily laid down, you don’t win.

So I was just curious about where the little phrase “all models are false” came from. I was expecting that Karl Popper might have said it, in which case to use the phrase as a defense mechanism against empirical refutation would have been a particularly fraudulent tactic, because it would have been a perversion of Popper’s methodological stance, which was to force our theoretical constructs to face up to, not to insulate it from, empirical testing. But when I googled “all theories are false” what I found was not Popper, but the British statistician, G. E. P. Box who wrote in his paper “Science and Statistics” based on his R. A. Fisher Memorial Lecture to the American Statistical Association: “All models are wrong.” Here’s the exact quote:

Since all models are wrong the scientist cannot obtain a “correct” one by excessive elaboration. On the contrary following William of Occam he should seek an economical description of natural phenomena. Just as the ability to devise simple but evocative models is the signature of the great scientist so overelaboration and overparameterization is often the mark of mediocrity.

Since all models are wrong the scientist must be alert to what is importantly wrong. It is inappropriate to be concerned about mice when there are tigers abroad. Pure mathematics is concerned with propositions like “given that A is true, does B necessarily follow?” Since the statement is a conditional one, it has nothing whatsoever to do with the truth of A nor of the consequences B in relation to real life. The pure mathematician, acting in that capacity, need not, and perhaps should not, have any contact with practical matters at all.

In applying mathematics to subjects such as physics or statistics we make tentative assumptions about the real world which we know are false but which we believe may be useful nonetheless. The physicist knows that particles have mass and yet certain results, approximating what really happens, may be derived from the assumption that they do not. Equally, the statistician knows, for example, that in nature there never was a normal distribution, there never was a straight line, yet with normal and linear assumptions, known to be false, he can often derive results which match, to a useful approximation, those found in the real world. It follows that, although rigorous derivation of logical consequences is of great importance to statistics, such derivations are necessarily encapsulated in the knowledge that premise, and hence consequence, do not describe natural truth.

It follows that we cannot know that any statistical technique we develop is useful unless we use it. Major advances in science and in the science of statistics in particular, usually occur, therefore, as the result of the theory-practice iteration.

One of the most annoying conceits of modern macroeconomists is the constant self-congratulatory references to themselves as scientists because of their ostentatious use of axiomatic reasoning, formal proofs, and higher mathematical techniques. The tiresome self-congratulation might get toned down ever so slightly if they bothered to read and take to heart Box’s lecture.

There Is No Intertemporal Budget Constraint

Last week Nick Rowe posted a link to a just published article in a special issue of the Review of Keynesian Economics commemorating the 80th anniversary of the General Theory. Nick’s article discusses the confusion in the General Theory between saving and hoarding, and Nick invited readers to weigh in with comments about his article. The ROKE issue also features an article by Simon Wren-Lewis explaining the eclipse of Keynesian theory as a result of the New Classical Counter-Revolution, correctly identified by Wren-Lewis as a revolution inspired not by empirical success but by a methodological obsession with reductive micro-foundationalism. While deploring the New Classical methodological authoritarianism, Wren-Lewis takes solace from the ability of New Keynesians to survive under the New Classical methodological regime, salvaging a role for activist counter-cyclical policy by, in effect, negotiating a safe haven for the sticky-price assumption despite its shaky methodological credentials. The methodological fiction that sticky prices qualify as micro-founded allowed New Keynesianism to survive despite the ascendancy of micro-foundationalist methodology, thereby enabling the core Keynesian policy message to survive.

I mention the Wren-Lewis article in this context because of an exchange between two of the commenters on Nick’s article: the presumably pseudonymous Avon Barksdale and blogger Jason Smith about microfoundations and Keynesian economics. Avon began by chastising Nick for wasting time discussing Keynes’s 80-year old ideas, something Avon thinks would never happen in a discussion about a true science like physics, the 100-year-old ideas of Einstein being of no interest except insofar as they have been incorporated into the theoretical corpus of modern physics. Of course, this is simply vulgar scientism, as if the only legitimate way to do economics is to mimic how physicists do physics. This methodological scolding is typically charming New Classical arrogance. Sort of reminds one of how Friedrich Engels described Marxian theory as scientific socialism. I mean who, other than a religious fanatic, would be stupid enough to argue with the assertions of science?

Avon continues with a quotation from David Levine, a fine economist who has done a lot of good work, but who is also enthralled by the New Classical methodology. Avon’s scientism provoked the following comment from Jason Smith, a Ph. D. in physics with a deep interest in and understanding of economics.

You quote from Levine: “Keynesianism as argued by people such as Paul Krugman and Brad DeLong is a theory without people either rational or irrational”

This is false. The L in ISLM means liquidity preference and e.g. here …

http://krugman.blogs.nytimes.com/2013/11/18/the-new-keynesian-case-for-fiscal-policy-wonkish/

… Krugman mentions an Euler equation. The Euler equation essentially says that an agent must be indifferent between consuming one more unit today on the one hand and saving that unit and consuming in the future on the other if utility is maximized.

So there are agents in both formulations preferring one state of the world relative to others.

Avon replied:

Jason,

“This is false. The L in ISLM means liquidity preference and e.g. here”

I know what ISLM is. It’s not recursive so it really doesn’t have people in it. The dynamics are not set by any micro-foundation. If you’d like to see models with people in them, try Ljungqvist and Sargent, Recursive Macroeconomic Theory.

To which Jason retorted:

Avon,

So the definition of “people” is restricted to agents making multi-period optimizations over time, solving a dynamic programming problem?

Well then any such theory is obviously wrong because people don’t behave that way. For example, humans don’t optimize the dictator game. How can you add up optimizing agents and get a result that is true for non-optimizing agents … coincident with the details of the optimizing agents mattering.

Your microfoundation requirement is like saying the ideal gas law doesn’t have any atoms in it. And it doesn’t! It is an aggregate property of individual “agents” that don’t have properties like temperature or pressure (or even volume in a meaningful sense). Atoms optimize entropy, but not out of any preferences.

So how do you know for a fact that macro properties like inflation or interest rates are directly related to agent optimizations? Maybe inflation is like temperature — it doesn’t exist for individuals and is only a property of economics in aggregate.

These questions are not answered definitively, and they’d have to be to enforce a requirement for microfoundations … or a particular way of solving the problem.

Are quarks important to nuclear physics? Not really — it’s all pions and nucleons. Emergent degrees of freedom. Sure, you can calculate pion scattering from QCD lattice calculations (quark and gluon DoF), but it doesn’t give an empirically better result than chiral perturbation theory (pion DoF) that ignores the microfoundations (QCD).

Assuming quarks are required to solve nuclear physics problems would have been a giant step backwards.

To which Avon rejoined:

Jason

The microfoundation of nuclear physics and quarks is quantum mechanics and quantum field theory. How the degrees of freedom reorganize under the renormalization group flow, what effective field theory results is an empirical question. Keynesian economics is worse tha[n] useless. It’s wrong empirically, it has no theoretical foundation, it has no laws. It has no microfoundation. No serious grad school has taught Keynesian economics in nearly 40 years.

To which Jason answered:

Avon,

RG flow is irrelevant to chiral perturbation theory which is based on the approximate chiral symmetry of QCD. And chiral perturbation theory could exist without QCD as the “microfoundation”.

Quantum field theory is not a ‘microfoundation’, but rather a framework for building theories that may or may not have microfoundations. As Weinberg (1979) said:

” … quantum field theory itself has no content beyond analyticity, unitarity,
cluster decomposition, and symmetry.”

If I put together an NJL model, there is no requirement that the scalar field condensate be composed of quark-antiquark pairs. In fact, the basic idea was used for Cooper pairs as a model of superconductivity. Same macro theory; different microfoundations. And that is a general problem with microfoundations — different microfoundations can lead to the same macro theory, so which one is right?

And the IS-LM model is actually pretty empirically accurate (for economics):

http://informationtransfereconomics.blogspot.com/2014/03/the-islm-model-again.html

To which Avon responded:

First, ISLM analysis does not hold empirically. It just doesn’t work. That’s why we ended up with the macro revolution of the 70s and 80s. Keynesian economics ignores intertemporal budget constraints, it violates Ricardian equivalence. It’s just not the way the world works. People might not solve dynamic programs to set their consumption path, but at least these models include a future which people plan over. These models work far better than Keynesian ISLM reasoning.

As for chiral perturbation theory and the approximate chiral symmetries of QCD, I am not making the case that NJL models requires QCD. NJL is an effective field theory so it comes from something else. That something else happens to be QCD. It could have been something else, that’s an empirical question. The microfoundation I’m talking about with theories like NJL is QFT and the symmetries of the vacuum, not the short distance physics that might be responsible for it. The microfoundation here is about the basic laws, the principles.

ISLM and Keynesian economics has none of this. There is no principle. The microfoundation of modern macro is not about increasing the degrees of freedom to model every person in the economy on some short distance scale, it is about building the basic principles from consistent economic laws that we find in microeconomics.

Well, I totally agree that IS-LM is a flawed macroeconomic model, and, in its original form, it was borderline-incoherent, being a single-period model with an interest rate, a concept without meaning except as an intertemporal price relationship. These deficiencies of IS-LM became obvious in the 1970s, so the model was extended to include a future period, with an expected future price level, making it possible to speak meaningfully about real and nominal interest rates, inflation and an equilibrium rate of spending. So the failure of IS-LM to explain stagflation, cited by Avon as the justification for rejecting IS-LM in favor of New Classical macro, was not that hard to fix, at least enough to make it serviceable. And comparisons of the empirical success of augmented IS-LM and the New Classical models have shown that IS-LM models consistently outperform New Classical models.

What Avon fails to see is that the microfoundations that he considers essential for macroeconomics are themselves derived from the assumption that the economy is operating in macroeconomic equilibrium. Thus, insisting on microfoundations – at least in the formalist sense that Avon and New Classical macroeconomists understand the term – does not provide a foundation for macroeconomics; it is just question begging aka circular reasoning or petitio principia.

The circularity is obvious from even a cursory reading of Samuelson’s Foundations of Economic Analysis, Robert Lucas’s model for doing economics. What Samuelson called meaningful theorems – thereby betraying his misguided acceptance of the now discredited logical positivist dogma that only potentially empirically verifiable statements have meaning – are derived using the comparative-statics method, which involves finding the sign of the derivative of an endogenous economic variable with respect to a change in some parameter. But the comparative-statics method is premised on the assumption that before and after the parameter change the system is in full equilibrium or at an optimum, and that the equilibrium, if not unique, is at least locally stable and the parameter change is sufficiently small not to displace the system so far that it does not revert back to a new equilibrium close to the original one. So the microeconomic laws invoked by Avon are valid only in the neighborhood of a stable equilibrium, and the macroeconomics that Avon’s New Classical mentors have imposed on the economics profession is a macroeconomics that, by methodological fiat, is operative only in the neighborhood of a locally stable equilibrium.

Avon dismisses Keynesian economics because it ignores intertemporal budget constraints. But the intertemporal budget constraint doesn’t exist in any objective sense. Certainly macroeconomics has to take into account intertemporal choice, but the idea of an intertemporal budget constraint analogous to the microeconomic budget constraint underlying the basic theory of consumer choice is totally misguided. In the static theory of consumer choice, the consumer has a given resource endowment and known prices at which consumers can transact at will, so the utility-maximizing vector of purchases and sales can be determined as the solution of a constrained-maximization problem.

In the intertemporal context, consumers have a given resource endowment, but prices are not known. So consumers have to make current transactions based on their expectations about future prices and a variety of other circumstances about which consumers can only guess. Their budget constraints are thus not real but totally conjectural based on their expectations of future prices. The optimizing Euler equations are therefore entirely conjectural as well, and subject to continual revision in response to changing expectations. The idea that the microeconomic theory of consumer choice is straightforwardly applicable to the intertemporal choice problem in a setting in which consumers don’t know what future prices will be and agents’ expectations of future prices are a) likely to be very different from each other and thus b) likely to be different from their ultimate realizations is a huge stretch. The intertemporal budget constraint has a completely different role in macroeconomics from the role it has in microeconomics.

If I expect that the demand for my services will be such that my disposable income next year would be $500k, my consumption choices would be very different from what they would have been if I were expecting a disposable income of $100k next year. If I expect a disposable income of $500k next year, and it turns out that next year’s income is only $100k, I may find myself in considerable difficulty, because my planned expenditure and the future payments I have obligated myself to make may exceed my disposable income or my capacity to borrow. So if there are a lot of people who overestimate their future incomes, the repercussions of their over-optimism may reverberate throughout the economy, leading to bankruptcies and unemployment and other bad stuff.

A large enough initial shock of mistaken expectations can become self-amplifying, at least for a time, possibly resembling the way a large initial displacement of water can generate a tsunami. A financial crisis, which is hard to model as an equilibrium phenomenon, may rather be an emergent phenomenon with microeconomic sources, but whose propagation can’t be described in microeconomic terms. New Classical macroeconomics simply excludes such possibilities on methodological grounds by imposing a rational-expectations general-equilibrium structure on all macroeconomic models.

This is not to say that the rational expectations assumption does not have a useful analytical role in macroeconomics. But the most interesting and most important problems in macroeconomics arise when the rational expectations assumption does not hold, because it is when individual expectations are very different and very unstable – say, like now, for instance — that macroeconomies become vulnerable to really scary instability.

Simon Wren-Lewis makes a similar point in his paper in the Review of Keynesian Economics.

Much discussion of current divisions within macroeconomics focuses on the ‘saltwater/freshwater’ divide. This understates the importance of the New Classical Counter Revolution (hereafter NCCR). It may be more helpful to think about the NCCR as involving two strands. The one most commonly talked about involves Keynesian monetary and fiscal policy. That is of course very important, and plays a role in the policy reaction to the recent Great Recession. However I want to suggest that in some ways the second strand, which was methodological, is more important. The NCCR helped completely change the way academic macroeconomics is done.

Before the NCCR, macroeconomics was an intensely empirical discipline: something made possible by the developments in statistics and econometrics inspired by The General Theory. After the NCCR and its emphasis on microfoundations, it became much more deductive. As Hoover (2001, p. 72) writes, ‘[t]he conviction that macroeconomics must possess microfoundations has changed the face of the discipline in the last quarter century’. In terms of this second strand, the NCCR was triumphant and remains largely unchallenged within mainstream academic macroeconomics.

Perhaps I will have some more to say about Wren-Lewis’s article in a future post. And perhaps also about Nick Rowe’s article.

HT: Tom Brown

Update (02/11/16):

On his blog Jason Smith provides some further commentary on his exchange with Avon on Nick Rowe’s blog, explaining at greater length how irrelevant microfoundations are to doing real empirically relevant physics. He also expands on and puts into a broader meta-theoretical context my point about the extremely narrow range of applicability of the rational-expectations equilibrium assumptions of New Classical macroeconomics.

David Glasner found a back-and-forth between me and a commenter (with the pseudonym “Avon Barksdale” after [a] character on The Wire who [didn’t end] up taking an economics class [per Tom below]) on Nick Rowe’s blog who expressed the (widely held) view that the only scientific way to proceed in economics is with rigorous microfoundations. “Avon” held physics up as a purported shining example of this approach.
I couldn’t let it go: even physics isn’t that reductionist. I gave several examples of cases where the microfoundations were actually known, but not used to figure things out: thermodynamics, nuclear physics. Even modern physics is supposedly built on string theory. However physicists do not require every pion scattering amplitude be calculated from QCD. Some people do do so-called lattice calculations. But many resort to the “effective” chiral perturbation theory. In a sense, that was what my thesis was about — an effective theory that bridges the gap between lattice QCD and chiral perturbation theory. That effective theory even gave up on one of the basic principles of QCD — confinement. It would be like an economist giving up opportunity cost (a basic principle of the micro theory). But no physicist ever said to me “your model is flawed because it doesn’t have true microfoundations”. That’s because the kind of hard core reductionism that surrounds the microfoundations paradigm doesn’t exist in physics — the most hard core reductionist natural science!
In his post, Glasner repeated something that he had before and — probably because it was in the context of a bunch of quotes about physics — I thought of another analogy.

Glasner says:

But the comparative-statics method is premised on the assumption that before and after the parameter change the system is in full equilibrium or at an optimum, and that the equilibrium, if not unique, is at least locally stable and the parameter change is sufficiently small not to displace the system so far that it does not revert back to a new equilibrium close to the original one. So the microeconomic laws invoked by Avon are valid only in the neighborhood of a stable equilibrium, and the macroeconomics that Avon’s New Classical mentors have imposed on the economics profession is a macroeconomics that, by methodological fiat, is operative only in the neighborhood of a locally stable equilibrium.

 

This hits on a basic principle of physics: any theory radically simplifies near an equilibrium.

Go to Jason’s blog to read the rest of his important and insightful post.

Representative Agents, Homunculi and Faith-Based Macroeconomics

After my previous post comparing the neoclassical synthesis in its various versions to the mind-body problem, there was an interesting Twitter exchange between Steve Randy Waldman and David Andolfatto in which Andolfatto queried whether Waldman and I are aware that there are representative-agent models in which the equilibrium is not Pareto-optimal. Andalfatto raised an interesting point, but what I found interesting about it might be different from what Andalfatto was trying to show, which, I am guessing, was that a representative-agent modeling strategy doesn’t necessarily commit the theorist to the conclusion that the world is optimal and that the solutions of the model can never be improved upon by a monetary/fiscal-policy intervention. I concede the point. It is well-known I think that, given the appropriate assumptions, a general-equilibrium model can have a sub-optimal solution. Given those assumptions, the corresponding representative-agent will also choose a sub-optimal solution. So I think I get that, but perhaps there’s a more subtle point  that I’m missing. If so, please set me straight.

But what I was trying to argue was not that representative-agent models are necessarily optimal, but that representative-agent models suffer from an inherent, and, in my view, fatal, flaw: they can’t explain any real macroeconomic phenomenon, because a macroeconomic phenomenon has to encompass something more than the decision of a single agent, even an omniscient central planner. At best, the representative agent is just a device for solving an otherwise intractable general-equilibrium model, which is how I think Lucas originally justified the assumption.

Yet just because a general-equilibrium model can be formulated so that it can be solved as the solution of an optimizing agent does not explain the economic mechanism or process that generates the solution. The mathematical solution of a model does not necessarily provide any insight into the adjustment process or mechanism by which the solution actually is, or could be, achieved in the real world. Your ability to find a solution for a mathematical problem does not mean that you understand the real-world mechanism to which the solution of your model corresponds. The correspondence between your model may be a strictly mathematical correspondence which may not really be in any way descriptive of how any real-world mechanism or process actually operates.

Here’s an example of what I am talking about. Consider a traffic-flow model explaining how congestion affects vehicle speed and the flow of traffic. It seems obvious that traffic congestion is caused by interactions between the different vehicles traversing a thoroughfare, just as it seems obvious that market exchange arises as the result of interactions between the different agents seeking to advance their own interests. OK, can you imagine building a useful traffic-flow model based on solving for the optimal plan of a representative vehicle?

I don’t think so. Once you frame the model in terms of a representative vehicle, you have abstracted from the phenomenon to be explained. The entire exercise would be pointless – unless, that is, you assumed that interactions between vehicles are so minimal that they can be ignored. But then why would you be interested in congestion effects? If you want to claim that your model has any relevance to the effect of congestion on traffic flow, you can’t base the claim on an assumption that there is no congestion.

Or to take another example, suppose you want to explain the phenomenon that, at sporting events, all, or almost all, the spectators sit in their seats but occasionally get up simultaneously from their seats to watch the play on the field or court. Would anyone ever think that an explanation in terms of a representative spectator could explain that phenomenon?

In just the same way, a representative-agent macroeconomic model necessarily abstracts from the interactions between actual agents. Obviously, by abstracting from the interactions, the model can’t demonstrate that there are no interactions between agents in the real world or that their interactions are too insignificant to matter. I would be shocked if anyone really believed that the interactions between agents are unimportant, much less, negligible; nor have I seen an argument that interactions between agents are unimportant, the concept of network effects, to give just one example, being an important topic in microeconomics.

It’s no answer to say that all the interactions are accounted for within the general-equilibrium model. That is just a form of question-begging. The representative agent is being assumed because without him the problem of finding a general-equilibrium solution of the model is very difficult or intractable. Taking into account interactions makes the model too complicated to work with analytically, so it is much easier — but still hard enough to allow the theorist to perform some fancy mathematical techniques — to ignore those pesky interactions. On top of that, the process by which the real world arrives at outcomes to which a general-equilibrium model supposedly bears at least some vague resemblance can’t even be described by conventional modeling techniques.

The modeling approach seems like that of a neuroscientist saying that, because he could simulate the functions, electrical impulses, chemical reactions, and neural connections in the brain – which he can’t do and isn’t even close to doing, even though a neuroscientist’s understanding of the brain far surpasses any economist’s understanding of the economy – he can explain consciousness. Simulating the operation of a brain would not explain consciousness, because the computer on which the neuroscientist performed the simulation would not become conscious in the course of the simulation.

Many neuroscientists and other materialists like to claim that consciousness is not real, that it’s just an epiphenomenon. But we all have the subjective experience of consciousness, so whatever it is that someone wants to call it, consciousness — indeed the entire world of mental phenomena denoted by that term — remains an unexplained phenomenon, a phenomenon that can only be dismissed as unreal on the basis of a metaphysical dogma that denies the existence of anything that can’t be explained as the result of material and physical causes.

I call that metaphysical belief a dogma not because it’s false — I have no way of proving that it’s false — but because materialism is just as much a metaphysical belief as deism or monotheism. It graduates from belief to dogma when people assert not only that the belief is true but that there’s something wrong with you if you are unwilling to believe it as well. The most that I would say against the belief in materialism is that I can’t understand how it could possibly be true. But I admit that there are a lot of things that I just don’t understand, and I will even admit to believing in some of those things.

New Classical macroeconomists, like, say, Robert Lucas and, perhaps, Thomas Sargent, like to claim that unless a macroeconomic model is microfounded — by which they mean derived from an explicit intertemporal optimization exercise typically involving a representative agent or possibly a small number of different representative agents — it’s not an economic model, because the model, being vulnerable to the Lucas critique, is theoretically superficial and vacuous. But only models of intertemporal equilibrium — a set of one or more mutually consistent optimal plans — are immune to the Lucas critique, so insisting on immunity to the Lucas critique as a prerequisite for a macroeconomic model is a guarantee of failure if your aim to explain anything other than an intertemporal equilibrium.

Unless, that is, you believe that real world is in fact the realization of a general equilibrium model, which is what real-business-cycle theorists, like Edward Prescott, at least claim to believe. Like materialist believers that all mental states are epiphenomenous, and that consciousness is an (unexplained) illusion, real-business-cycle theorists purport to deny that there is such a thing as a disequilibrium phenomenon, the so-called business cycle, in their view, being nothing but a manifestation of the intertemporal-equilibrium adjustment of an economy to random (unexplained) productivity shocks. According to real-business-cycle theorists, such characteristic phenomena of business cycles as surprise, regret, disappointed expectations, abandoned and failed plans, the inability to find work at wages comparable to wages that other similar workers are being paid are not real phenomena; they are (unexplained) illusions and misnomers. The real-business-cycle theorists don’t just fail to construct macroeconomic models; they deny the very existence of macroeconomics, just as strict materialists deny the existence of consciousness.

What is so preposterous about the New-Classical/real-business-cycle methodological position is not the belief that the business cycle can somehow be modeled as a purely equilibrium phenomenon, implausible as that idea seems, but the insistence that only micro-founded business-cycle models are methodologically acceptable. It is one thing to believe that ultimately macroeconomics and business-cycle theory will be reduced to the analysis of individual agents and their interactions. But current micro-founded models can’t provide explanations for what many of us think are basic features of macroeconomic and business-cycle phenomena. If non-micro-founded models can provide explanations for those phenomena, even if those explanations are not fully satisfactory, what basis is there for rejecting them just because of a methodological precept that disqualifies all non-micro-founded models?

According to Kevin Hoover, the basis for insisting that only micro-founded macroeconomic models are acceptable, even if the microfoundation consists in a single representative agent optimizing for an entire economy, is eschatological. In other words, because of a belief that economics will eventually develop analytical or computational techniques sufficiently advanced to model an entire economy in terms of individual interacting agents, an analysis based on a single representative agent, as the first step on this theoretical odyssey, is somehow methodologically privileged over alternative models that do not share that destiny. Hoover properly rejects the presumptuous notion that an avowed, but unrealized, theoretical destiny, can provide a privileged methodological status to an explanatory strategy. The reductionist microfoundationalism of New-Classical macroeconomics and real-business-cycle theory, with which New Keynesian economists have formed an alliance of convenience, is truly a faith-based macroeconomics.

The remarkable similarity between the reductionist microfoundational methodology of New-Classical macroeconomics and the reductionist materialist approach to the concept of mind suggests to me that there is also a close analogy between the representative agent and what philosophers of mind call a homunculus. The Cartesian materialist theory of mind maintains that, at some place or places inside the brain, there resides information corresponding to our conscious experience. The question then arises: how does our conscious experience access the latent information inside the brain? And the answer is that there is a homunculus (or little man) that processes the information for us so that we can perceive it through him. For example, the homunculus (see the attached picture of the little guy) views the image cast by light on the retina as if he were watching a movie projected onto a screen.

homunculus

But there is an obvious fallacy, because the follow-up question is: how does our little friend see anything? Well, the answer must be that there’s another, smaller, homunculus inside his brain. You can probably already tell that this argument is going to take us on an infinite regress. So what purports to be an explanation turns out to be just a form of question-begging. Sound familiar? The only difference between the representative agent and the homunculus is that the representative agent begs the question immediately without having to go on an infinite regress.

PS I have been sidetracked by other responsibilities, so I have not been blogging much, if at all, for the last few weeks. I hope to post more frequently, but I am afraid that my posting and replies to comments are likely to remain infrequent for the next couple of months.

Romer v. Lucas

A couple of months ago, Paul Romer created a stir by publishing a paper in the American Economic Review “Mathiness in the Theory of Economic Growth,” an attack on two papers, one by McGrattan and Prescott and the other by Lucas and Moll on aspects of growth theory. He accused the authors of those papers of using mathematical modeling as a cover behind which to hide assumptions guaranteeing results by which the authors could promote their research agendas. In subsequent blog posts, Romer has sharpened his attack, focusing it more directly on Lucas, whom he accuses of a non-scientific attachment to ideological predispositions that have led him to violate what he calls Feynman integrity, a concept eloquently described by Feynman himself in a 1974 commencement address at Caltech.

It’s a kind of scientific integrity, a principle of scientific thought that corresponds to a kind of utter honesty–a kind of leaning over backwards. For example, if you’re doing an experiment, you should report everything that you think might make it invalid–not only what you think is right about it: other causes that could possibly explain your results; and things you thought of that you’ve eliminated by some other experiment, and how they worked–to make sure the other fellow can tell they have been eliminated.

Details that could throw doubt on your interpretation must be given, if you know them. You must do the best you can–if you know anything at all wrong, or possibly wrong–to explain it. If you make a theory, for example, and advertise it, or put it out, then you must also put down all the facts that disagree with it, as well as those that agree with it. There is also a more subtle problem. When you have put a lot of ideas together to make an elaborate theory, you want to make sure, when explaining what it fits, that those things it fits are not just the things that gave you the idea for the theory; but that the finished theory makes something else come out right, in addition.

Romer contrasts this admirable statement of what scientific integrity means with another by George Stigler, seemingly justifying, or at least excusing, a kind of special pleading on behalf of one’s own theory. And the institutional and perhaps ideological association between Stigler and Lucas seems to suggest that Lucas is inclined to follow the permissive and flexible Stiglerian ethic rather than rigorous Feynman standard of scientific integrity. Romer regards this as a breach of the scientific method and a step backward for economics as a science.

I am not going to comment on the specific infraction that Romer accuses Lucas of having committed; I am not familiar with the mathematical question in dispute. Certainly if Lucas was aware that his argument in the paper Romer criticizes depended on the particular mathematical assumption in question, Lucas should have acknowledged that to be the case. And even if, as Lucas asserted in responding to a direct question by Romer, he could have derived the result in a more roundabout way, then he should have pointed that out, too. However, I don’t regard the infraction alleged by Romer to be more than a misdemeanor, hardly a scandalous breach of the scientific method.

Why did Lucas, who as far as I can tell was originally guided by Feynman integrity, switch to the mode of Stigler conviction? Market clearing did not have to evolve from auxiliary hypothesis to dogma that could not be questioned.

My conjecture is economists let small accidents of intellectual history matter too much. If we had behaved like scientists, things could have turned out very differently. It is worth paying attention to these accidents because doing so might let us take more control over the process of scientific inquiry that we are engaged in. At the very least, we should try to reduce the odds that that personal frictions and simple misunderstandings could once again cause us to veer off on some damaging trajectory.

I suspect that it was personal friction and a misunderstanding that encouraged a turn toward isolation (or if you prefer, epistemic closure) by Lucas and colleagues. They circled the wagons because they thought that this was the only way to keep the rational expectations revolution alive. The misunderstanding is that Lucas and his colleagues interpreted the hostile reaction they received from such economists as Robert Solow to mean that they were facing implacable, unreasoning resistance from such departments as MIT. In fact, in a remarkably short period of time, rational expectations completely conquered the PhD program at MIT.

More recently Romer, having done graduate work both at MIT and Chicago in the late 1970s, has elaborated on the personal friction between Solow and Lucas and how that friction may have affected Lucas, causing him to disengage from the professional mainstream. Paul Krugman, who was at MIT when this nastiness was happening, is skeptical of Romer’s interpretation.

My own view is that being personally and emotionally attached to one’s own theories, whether for religious or ideological or other non-scientific reasons, is not necessarily a bad thing as long as there are social mechanisms allowing scientists with different scientific viewpoints an opportunity to make themselves heard. If there are such mechanisms, the need for Feynman integrity is minimized, because individual lapses of integrity will be exposed and remedied by criticism from other scientists; scientific progress is possible even if scientists don’t live up to the Feynman standards, and maintain their faith in their theories despite contradictory evidence. But, as I am going to suggest below, there are reasons to doubt that social mechanisms have been operating to discipline – not suppress, just discipline – dubious economic theorizing.

My favorite example of the importance of personal belief in, and commitment to the truth of, one’s own theories is Galileo. As discussed by T. S. Kuhn in The Structure of Scientific Revolutions. Galileo was arguing for a paradigm change in how to think about the universe, despite being confronted by empirical evidence that appeared to refute the Copernican worldview he believed in: the observations that the sun revolves around the earth, and that the earth, as we directly perceive it, is, apart from the occasional earthquake, totally stationary — good old terra firma. Despite that apparently contradictory evidence, Galileo had an alternative vision of the universe in which the obvious movement of the sun in the heavens was explained by the spinning of the earth on its axis, and the stationarity of the earth by the assumption that all our surroundings move along with the earth, rendering its motion imperceptible, our perception of motion being relative to a specific frame of reference.

At bottom, this was an almost metaphysical world view not directly refutable by any simple empirical test. But Galileo adopted this worldview or paradigm, because he deeply believed it to be true, and was therefore willing to defend it at great personal cost, refusing to recant his Copernican view when he could have easily appeased the Church by describing the Copernican theory as just a tool for predicting planetary motion rather than an actual representation of reality. Early empirical tests did not support heliocentrism over geocentrism, but Galileo had faith that theoretical advancements and improved measurements would eventually vindicate the Copernican theory. He was right of course, but strict empiricism would have led to a premature rejection of heliocentrism. Without a deep personal commitment to the Copernican worldview, Galileo might not have articulated the case for heliocentrism as persuasively as he did, and acceptance of heliocentrism might have been delayed for a long time.

Imre Lakatos called such deeply-held views underlying a scientific theory the hard core of the theory (aka scientific research program), a set of beliefs that are maintained despite apparent empirical refutation. The response to any empirical refutation is not to abandon or change the hard core but to adjust what Lakatos called the protective belt of the theory. Eventually, as refutations or empirical anomalies accumulate, the research program may undergo a crisis, leading to its abandonment, or it may simply degenerate if it fails to solve new problems or discover any new empirical facts or regularities. So Romer’s criticism of Lucas’s dogmatic attachment to market clearing – Lucas frequently makes use of ad hoc price stickiness assumptions; I don’t know why Romer identifies market-clearing as a Lucasian dogma — may be no more justified from a history of science perspective than would criticism of Galileo’s dogmatic attachment to heliocentrism.

So while I have many problems with Lucas, lack of Feynman integrity is not really one of them, certainly not in the top ten. What I find more disturbing is his narrow conception of what economics is. As he himself wrote in an autobiographical sketch for Lives of the Laureates, he was bewitched by the beauty and power of Samuelson’s Foundations of Economic Analysis when he read it the summer before starting his training as a graduate student at Chicago in 1960. Although it did not have the transformative effect on me that it had on Lucas, I greatly admire the Foundations, but regardless of whether Samuelson himself meant to suggest such an idea (which I doubt), it is absurd to draw this conclusion from it:

I loved the Foundations. Like so many others in my cohort, I internalized its view that if I couldn’t formulate a problem in economic theory mathematically, I didn’t know what I was doing. I came to the position that mathematical analysis is not one of many ways of doing economic theory: It is the only way. Economic theory is mathematical analysis. Everything else is just pictures and talk.

Oh, come on. Would anyone ever think that unless you can formulate the problem of whether the earth revolves around the sun or the sun around the earth mathematically, you don’t know what you are doing? And, yet, remarkably, on the page following that silly assertion, one finds a totally brilliant description of what it was like to take graduate price theory from Milton Friedman.

Friedman rarely lectured. His class discussions were often structured as debates, with student opinions or newspaper quotes serving to introduce a problem and some loosely stated opinions about it. Then Friedman would lead us into a clear statement of the problem, considering alternative formulations as thoroughly as anyone in the class wanted to. Once formulated, the problem was quickly analyzed—usually diagrammatically—on the board. So we learned how to formulate a model, to think about and decide which features of a problem we could safely abstract from and which he needed to put at the center of the analysis. Here “model” is my term: It was not a term that Friedman liked or used. I think that for him talking about modeling would have detracted from the substantive seriousness of the inquiry we were engaged in, would divert us away from the attempt to discover “what can be done” into a merely mathematical exercise. [my emphasis].

Despite his respect for Friedman, it’s clear that Lucas did not adopt and internalize Friedman’s approach to economic problem solving, but instead internalized the caricature he extracted from Samuelson’s Foundations: that mathematical analysis is the only legitimate way of doing economic theory, and that, in particular, the essence of macroeconomics consists in a combination of axiomatic formalism and philosophical reductionism (microfoundationalism). For Lucas, the only scientifically legitimate macroeconomic models are those that can be deduced from the axiomatized Arrow-Debreu-McKenzie general equilibrium model, with solutions that can be computed and simulated in such a way that the simulations can be matched up against the available macroeconomics time series on output, investment and consumption.

This was both bad methodology and bad science, restricting the formulation of economic problems to those for which mathematical techniques are available to be deployed in finding solutions. On the one hand, the rational-expectations assumption made finding solutions to certain intertemporal models tractable; on the other, the assumption was justified as being required by the rationality assumptions of neoclassical price theory.

In a recent review of Lucas’s Collected Papers on Monetary Theory, Thomas Sargent makes a fascinating reference to Kenneth Arrow’s 1967 review of the first two volumes of Paul Samuelson’s Collected Works in which Arrow referred to the problematic nature of the neoclassical synthesis of which Samuelson was a chief exponent.

Samuelson has not addressed himself to one of the major scandals of current price theory, the relation between microeconomics and macroeconomics. Neoclassical microeconomic equilibrium with fully flexible prices presents a beautiful picture of the mutual articulations of a complex structure, full employment being one of its major elements. What is the relation between this world and either the real world with its recurrent tendencies to unemployment of labor, and indeed of capital goods, or the Keynesian world of underemployment equilibrium? The most explicit statement of Samuelson’s position that I can find is the following: “Neoclassical analysis permits of fully stable underemployment equilibrium only on the assumption of either friction or a peculiar concatenation of wealth-liquidity-interest elasticities. . . . [The neoclassical analysis] goes far beyond the primitive notion that, by definition of a Walrasian system, equilibrium must be at full employment.” . . .

In view of the Phillips curve concept in which Samuelson has elsewhere shown such interest, I take the second sentence in the above quotation to mean that wages are stationary whenever unemployment is X percent, with X positive; thus stationary unemployment is possible. In general, one can have a neoclassical model modified by some elements of price rigidity which will yield Keynesian-type implications. But such a model has yet to be constructed in full detail, and the question of why certain prices remain rigid becomes of first importance. . . . Certainly, as Keynes emphasized the rigidity of prices has something to do with the properties of money; and the integration of the demand and supply of money with general competitive equilibrium theory remains incomplete despite attempts beginning with Walras himself.

If the neoclassical model with full price flexibility were sufficiently unrealistic that stable unemployment equilibrium be possible, then in all likelihood the bulk of the theorems derived by Samuelson, myself, and everyone else from the neoclassical assumptions are also contrafactual. The problem is not resolved by what Samuelson has called “the neoclassical synthesis,” in which it is held that the achievement of full employment requires Keynesian intervention but that neoclassical theory is valid when full employment is reached. . . .

Obviously, I believe firmly that the mutual adjustment of prices and quantities represented by the neoclassical model is an important aspect of economic reality worthy of the serious analysis that has been bestowed on it; and certain dramatic historical episodes – most recently the reconversion of the United States from World War II and the postwar European recovery – suggest that an economic mechanism exists which is capable of adaptation to radical shifts in demand and supply conditions. On the other hand, the Great Depression and the problems of developing countries remind us dramatically that something beyond, but including, neoclassical theory is needed.

Perhaps in a future post, I may discuss this passage, including a few sentences that I have omitted here, in greater detail. For now I will just say that Arrow’s reference to a “neoclassical microeconomic equilibrium with fully flexible prices” seems very strange inasmuch as price flexibility has absolutely no role in the proofs of the existence of a competitive general equilibrium for which Arrow and Debreu and McKenzie are justly famous. All the theorems Arrow et al. proved about the neoclassical equilibrium were related to existence, uniqueness and optimaiity of an equilibrium supported by an equilibrium set of prices. Price flexibility was not involved in those theorems, because the theorems had nothing to do with how prices adjust in response to a disequilibrium situation. What makes this juxtaposition of neoclassical microeconomic equilibrium with fully flexible prices even more remarkable is that about eight years earlier Arrow wrote a paper (“Toward a Theory of Price Adjustment”) whose main concern was the lack of any theory of price adjustment in competitive equilibrium, about which I will have more to say below.

Sargent also quotes from two lectures in which Lucas referred to Don Patinkin’s treatise Money, Interest and Prices which provided perhaps the definitive statement of the neoclassical synthesis Samuelson espoused. In one lecture (“My Keynesian Education” presented to the History of Economics Society in 2003) Lucas explains why he thinks Patinkin’s book did not succeed in its goal of integrating value theory and monetary theory:

I think Patinkin was absolutely right to try and use general equilibrium theory to think about macroeconomic problems. Patinkin and I are both Walrasians, whatever that means. I don’t see how anybody can not be. It’s pure hindsight, but now I think that Patinkin’s problem was that he was a student of Lange’s, and Lange’s version of the Walrasian model was already archaic by the end of the 1950s. Arrow and Debreu and McKenzie had redone the whole theory in a clearer, more rigorous, and more flexible way. Patinkin’s book was a reworking of his Chicago thesis from the middle 1940s and had not benefited from this more recent work.

In the other lecture, his 2003 Presidential address to the American Economic Association, Lucas commented further on why Patinkin fell short in his quest to unify monetary and value theory:

When Don Patinkin gave his Money, Interest, and Prices the subtitle “An Integration of Monetary and Value Theory,” value theory meant, to him, a purely static theory of general equilibrium. Fluctuations in production and employment, due to monetary disturbances or to shocks of any other kind, were viewed as inducing disequilibrium adjustments, unrelated to anyone’s purposeful behavior, modeled with vast numbers of free parameters. For us, today, value theory refers to models of dynamic economies subject to unpredictable shocks, populated by agents who are good at processing information and making choices over time. The macroeconomic research I have discussed today makes essential use of value theory in this modern sense: formulating explicit models, computing solutions, comparing their behavior quantitatively to observed time series and other data sets. As a result, we are able to form a much sharper quantitative view of the potential of changes in policy to improve peoples’ lives than was possible a generation ago.

So, as Sargent observes, Lucas recreated an updated neoclassical synthesis of his own based on the intertemporal Arrow-Debreu-McKenzie version of the Walrasian model, augmented by a rationale for the holding of money and perhaps some form of monetary policy, via the assumption of credit-market frictions and sticky prices. Despite the repudiation of the updated neoclassical synthesis by his friend Edward Prescott, for whom monetary policy is irrelevant, Lucas clings to neoclassical synthesis 2.0. Sargent quotes this passage from Lucas’s 1994 retrospective review of A Monetary History of the US by Friedman and Schwartz to show how tightly Lucas clings to neoclassical synthesis 2.0 :

In Kydland and Prescott’s original model, and in many (though not all) of its descendants, the equilibrium allocation coincides with the optimal allocation: Fluctuations generated by the model represent an efficient response to unavoidable shocks to productivity. One may thus think of the model not as a positive theory suited to all historical time periods but as a normative benchmark providing a good approximation to events when monetary policy is conducted well and a bad approximation when it is not. Viewed in this way, the theory’s relative success in accounting for postwar experience can be interpreted as evidence that postwar monetary policy has resulted in near-efficient behavior, not as evidence that money doesn’t matter.

Indeed, the discipline of real business cycle theory has made it more difficult to defend real alternaltives to a monetary account of the 1930s than it was 30 years ago. It would be a term-paper-size exercise, for example, to work out the possible effects of the 1930 Smoot-Hawley Tariff in a suitably adapted real business cycle model. By now, we have accumulated enough quantitative experience with such models to be sure that the aggregate effects of such a policy (in an economy with a 5% foreign trade sector before the Act and perhaps a percentage point less after) would be trivial.

Nevertheless, in the absence of some catastrophic error in monetary policy, Lucas evidently believes that the key features of the Arrow-Debreu-McKenzie model are closely approximated in the real world. That may well be true. But if it is, Lucas has no real theory to explain why.

In his 1959 paper (“Towards a Theory of Price Adjustment”) I just mentioned, Arrow noted that the theory of competitive equilibrium has no explanation of how equilibrium prices are actually set. Indeed, the idea of competitive price adjustment is beset by a paradox: all agents in a general equilibrium being assumed to be price takers, how is it that a new equilibrium price is ever arrived at following any disturbance to an initial equilibrium? Arrow had no answer to the question, but offered the suggestion that, out of equilibrium, agents are not price takers, but price searchers, possessing some measure of market power to set price in the transition between the old and new equilibrium. But the upshot of Arrow’s discussion was that the problem and the paradox awaited solution. Almost sixty years on, some of us are still waiting, but for Lucas and the Lucasians, there is neither problem nor paradox, because the actual price is the equilibrium price, and the equilibrium price is always the (rationally) expected price.

If the social functions of science were being efficiently discharged, this rather obvious replacement of problem solving by question begging would not have escaped effective challenge and opposition. But Lucas was able to provide cover for this substitution by persuading the profession to embrace his microfoundational methodology, while offering irresistible opportunities for professional advancement to younger economists who could master the new analytical techniques that Lucas and others were rapidly introducing, thereby neutralizing or coopting many of the natural opponents to what became modern macroeconomics. So while Romer considers the conquest of MIT by the rational-expectations revolution, despite the opposition of Robert Solow, to be evidence for the advance of economic science, I regard it as a sign of the social failure of science to discipline a regressive development driven by the elevation of technique over substance.


About Me

David Glasner
Washington, DC

I am an economist in the Washington DC area. My research and writing has been mostly on monetary economics and policy and the history of economics. In my book Free Banking and Monetary Reform, I argued for a non-Monetarist non-Keynesian approach to monetary policy, based on a theory of a competitive supply of money. Over the years, I have become increasingly impressed by the similarities between my approach and that of R. G. Hawtrey and hope to bring Hawtrey's unduly neglected contributions to the attention of a wider audience.

Archives

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 2,566 other followers

Follow Uneasy Money on WordPress.com