:: Article

Paul Krugman’s Philosophy of Economics, and What It Should Be

By Alex Rosenberg.

Like most successful economists, Paul Krugman doesn’t have much time for methodology. “I’ve never liked the notion of talking about economic “science” — it’s much too raw and imperfect a discipline to be paired casually with things like chemistry or biology.”

But ever since the financial crisis of 2007-2008 he has found it unavoidable. In “How did economics get it so wrong?” Krugman had a great deal to about where economics went astray, substantively and methodologically, but not much about how in general the discipline should proceed, little to say about what epistemic problems and limits it faces, and less about how it might be improved. In academic papers he has written both before and after his 2009 broadside, and even more so in his blog, “The Conscience of a Liberal,” in the New York Times, Krugman has gone some of the way towards filling out the picture of how he thinks economics should proceed. The view he articulates will strike observers, and even most economists, as reasonable, non-controversial, and hard to disagree with.

The trouble is that Krugman’s positive picture of how economics as its best is done undermines his substantive economic arguments. It provides just what is needed by those who dissent from his analysis of the economy and reject the policies he advocates to End This Depression Now (as his latest book is titled). Chicago school “extremists,” freshwater ideologues, and other free-market fundamentalists can help themselves to Krugman’s methodology to defend the substance of their disagreements with him. This raises the question of what should Krugman’s philosophy of economics should really be?

1. Models — simple and complex

Krugman tells us that economic theory is a sub-discipline of models, of varying degrees of mathematical complexity. All of them are incomplete—i.e. leaving out one of more important features of their targets, many of them incompatible—since they illuminate differing features of common targets by making differing simplifications, none of them are accurate, owing to their idealizations. But the good ones among them have some real explanatory power and some predictive content worthy of policy makers paying head to.

There are according, to Krugman, two kinds of models: simple ones proceeding from a variety of more and less ad hoc assumptions, designed to tell a story that combines some predictive success with an appeal to educated intuitions about how reasonable people act. And there are more complex models, whose mathematical character is driven by the need to extract consequences—sometimes surprising–from their assumptions and to check the validity of inferences. Simple models with vast simplifications and huge idealizations make intuitive sense to the economist, and mathematically elegant ones prove theorems that nail down the conditions on which the simple models work well.
Between simple models and complex ones Krugman is clear that the former are of greater importance in the discipline.

‘The point is not that these models are accurate or complete, or that they should be the only models used. Clearly they are incomplete, quite inadequate to examining some questions, and remain as full of ad hoc [sic] as ever. But they are easy to use, particularly on real-world policy questions, and often seem to give more or less the right answer.‘ (emphasis added.)

Krugman argues for his preference with an example. Contrast a simple model of the relationship between exchange rates, money supply and investment developed by Mundell and Fleming with a mathematically sophisticated model of the same relationship:

‘One of the most influential macro models of the 1990s, and deservedly so, is the revisitation of Mundell–Fleming–Dornbusch by Obstfeld and Rogoff (1995). It’s a beautiful piece of work, integrating a new Keynesian approach to price stickiness (albeit with the ad-hoc assumption that prices are set for only one period) with a full intertemporal approach to aggregate demand; it addresses the classic question of the effect of a monetary shock on output, interest rates, and the exchange rate. It’s also pretty hard work: as restated in their 1996 book, the model and its analysis take 76 equations to describe, even though the authors restrict themselves to a special case, and to log-linearized analysis around a steady state.’

Krugman describes the theorem as a “gorgeous welfare result” that is “immediately understandable in terms of the welfare economics of monopoly in general.” By “immediately understandable” Krugman means ‘makes sense in a story that explains by exploiting normal human motivations.’ A worthwhile economic model must tell a story, come with a narrative, harness human expectations and wants. Even when complex models do their work, the results have to make economic sense: the story that the economist can tell about the model’s results arise may have been unnoticed before they extracted from the model by the mathematics. But the model must show how the choices of rational agents could somehow bring them about.

In many blog posts Krugman has insisted that

‘…economics is about how people (the word “agents” is itself a kind of tribal marker) are motivated to take actions, and how those actions interact. Equilibrium is often a very convenient way to think through all of that, and all of us sometimes use wording about what the economy “needs” or “requires” as shorthand. When I talk about the Dornbusch overshooting model of the exchange rate, for example, I might say something like “the currency has to overshoot its long-run value, so that investors expect appreciation that offsets the interest differential.” But behind that verbal shorthand is a story about people doing stuff: investors selling the currency because yields are down, the currency falling until it’s so low that people figure it has nowhere to go but up.’

So, why are simpler models to be preferred? First, the more complex model rarely if ever provides better predictions or even explanations about what is going on in the economy than a simple one. And even as a guide to policy their welfare implications are highly limited: “Do you really believe that welfare result [the Obstfeld and Rogoff (1995) theorem]? Certainly not—indeed, it is driven to an important extent by the details of the model, and can quite easily be undone.” What is the use of the more complex model, then? They provide reassurance about simpler models when they confirm the simpler models conclusions; and where they contradict the simpler models they are warning signs that something is amiss. But they don’t, it seems, enhance explanation or prediction, unlike more complex models in natural science.

Some will ask why more complex models don’t improve on simpler ones, as they often do in natural science. Krugman has little interest in this question. He needs to answer it, as we shall see.

2. That was then, this is now.

Krugman tells us that a model gives “more or less the right answer” when it makes predictions that are borne out by events. When the model does so and is simple, it is particularly useful in guiding policy. This is right as far as it goes. And Krugman thinks it is enough to sustain his models and undermine ones that give different, or even opposite answers to our questions about where the economy is heading.

In the context of the period from 2008 the questions relevant for judging the adequacy of models were ones about the consequences of macroeconomic policy, especially the austerity measures of the European nations and the limited stimulus policies of the US government. The right answers about what these consequences were are given by Krugman’s simple models—Keynesian IS-LM curve based models. In Krugman’s own work they were used to predict that austerity wouldn’t work and that the stimulus would only do so in part. Because they rightly predicted the course of events in Europe and the US over several years, Krugman is prepared to attach explanatory weight to the models. Thus he diagnoses the recession as resulting from decline in demand and the persistence of a liquidity trap.

By contrast with Krugman’s Keynesian models, those of the New Classical macroeconomists faired very badly in their predictions over the same period. These models are so called because they all begin with the same strong assumptions that characterized Keynes’ predecessors, economists he labeled ‘classical’: that economic agents are fully rational, and that prices respond smoothly, quickly and efficiently, to changes in quantities supplied and demanded. Now, there’s nothing wrong with strong assumptions, Krugman’s models make them too. But the New Classical models predicted that European austerity would result in growth, that the US stimulus would be ineffective, and that the Fed’s quantitative easing policy would be inflationary. All of these New Classical predictions failed. What is more, Krugman’s Keynesian models explain fairly simply why the New Classical models don’t work—liquidity traps, price (i.e. wage) rigidities.

So, Krugman’s models win, right? Not so fast, the New Classical Macroeconomists will respond. Wind the clock back about 40 years and the shoe was on the other foot. Back then, the very same IS-LM models that Keynesians had developed and Krugman has recycled, were giving more or less the wrong answers. This was the period of “Stagflation” in the ‘70s when the economy refused to behave the way Keynes’ models told us it would—increasing unemployment together with increasing inflation.

Back then it was the New Classical macrotheory that gave the right answers and explained what the matter with the Keynesian models was. The Keynesian models assumed that economic agents were irrational or ignorant or both. They were built on the idea that when the government incurs a deficit in order to stimulate the economy now, the tax payers, will forget that to make good that deficit they’ll have to pay more taxes later; they’ll be fooled by the stimulus and the inflation it causes into thinking they are richer than they really are. So they will spend more now, refloating the economy. The government can try that trick repeatedly. But, the New Classical economists argued, eventually people will catch on. After a while they will be able to tell the difference between prosperity and inflation. At that point deficit spending (fiscal policy) will become ineffective. The result will be both increased inflation and unemployment, the worst of both worlds. When that began to happen (just after Nixon announced, “we’re all Keynesians now”) the Keynesian model will have been shown to be more or less the wrong answer.

The New Classical macromodels had solid microfoundation: they assumed that taken one at a time people are rational, that they used their rational faculties to frame expectations about the future, and in doing so they were at least as well informed about the economy, on average, as macroeconomists themselves were. So adding them all up together they couldn’t be fooled by government policies. The ineffectiveness of government policy to fine-tune the economy was one of the consequences of the New Classical macrotheory dear to the hearts of laissez-faire economists, politicians and businessmen.

Whose models are right, Keynes’ and Krugman’s? Or is it the “Rational Expectations” models that treated the macroeconomy as simply the sum of all people rolled into one absolutely rational “representative agent” making decisions for the whole economy on the basis of accurate expectations of what the government was going to do?
The answer to this question seems to be it depends were we come into the story: 40 years ago it looked like the New Classicals had more or less the right answers to the predictive questions and so the best explanations of economic processes. Today, it looks like the Keynesian models are in the saddle.

The problem is that “giving more or less the right answer” to questions about prediction is not enough to decide between the usefulness of simple (or complex) models. In economics as in other domains, it’s too easy to be right by accident, to be right for the wrong reasons, to be right for one period and wrong for another. Those of us who want to believe that Krugman is right about the economy need more. We need to know that Krugman’s models have been right about the last half decade for the right reasons. We need some confidence that the success of Krugman’s models is not accidental. We need some reason to suppose that the temporary success of the New Classical models during the “great moderation” of the ‘80s and ‘90s was accidental, or at least transitory. There are only two ways to acquire those reasons: systematic, secular improvements in the predictive power of Krugman’s models over the medium term future, and/or some greater confidence in their explanatory power than in the New Classical models’.

Improvement in predictive power means getting things more and more right over time, to more decimal places, over a wider range of phenomena, not immediately but eventually. Betting on any economic theory, or its set of models to do this, is to say the least a long shot, given the track record since Adam Smith. What about increasing our confidence in the explanatory power of Krugman’s models?

3. Mr. Krugman and the New Classics

Krugman writes:

‘So how do you do useful economics? In general, what we really do is combine maximization-and-equilibrium as a first cut with a variety of ad hoc modifications reflecting what seem to be empirical regularities about how both individual behavior and markets depart from this idealized case.’

But if you ask the New Classical economists, they’ll say, this is exactly what we do—combine maximizing-and-equilibrium with empirical regularities. And they’d go on to say it’s because Krugman’s Keynesian models don’t do this or don’t do enough of it, they are not “useful” for prediction or explanation. This is the whole point of the demand for microfoundations that Krugman and the Keynesians reject: it’s only by taking maximizing and equilibrium seriously combining that you get an explanation of the empirical regularities about stagflation of the ‘70s. Krugman, on the other hand, thinks models should incorporate empirical regularities about how individual behavior departs from maximization and markets move among local equilibria. And he has no time for microfoundations:

‘What we call “microfoundations” are not like physical laws. Heck, they’re not even true. Maximizing consumers are just a metaphor, possibly useful in making sense of behavior, but possibly not. The metaphors we use for microfoundations have no claim to be regarded as representing a higher order of truth than the ad hoc aggregate metaphors we use in IS-LM or whatever; in fact, we have much more supportive evidence for Keynesian macro than we do for standard micro.’

But notice, first, if maximizing consumers are just metaphors, convenient but dispensable, they can hardly be described they way Krugman does, as part of the general method of useful economics, along with equilibrium. Second, having insisted that models and their assumptions be judged for usefulness, Krugman can’t just suddenly switch to judging them by evidential support—disputed support at that, as we have seen.

When he accepts maximizing and equilibrium as the (only?) way useful economics is done Krugman makes a concession so great it threatens to undercut the rest of his arguments against New Classical economics:

‘Specifically: we have a body of economic theory built around the assumptions of perfectly rational behavior and perfectly functioning markets. Any economist with a grain of sense — which is to say, maybe half the profession? — knows that this is very much an abstraction, to be modified whenever the evidence suggests that it’s going wrong. But nobody has come up with general rules for making such modifications.’

The trouble is that the macroeconomic evidence such as it is, can’t tell us when and where maximization-and-equilibrium goes wrong, and there seems no prospect for improving the assumptions of perfect rationality and perfect markets from behavioral economics, neuroeconomics, experimental economics, evolutionary economics, game theory, etc. These research areas provide, as Krugman says, “more a collection of interesting and sometimes useful observations than a general…paradigm that can offer guidance across a wide range of cases.” There are, Krugman notes, too many “ways of being slightly stupid, so it’s hard to come up with a general theory about which of these ways [economic agents] will choose in any given situation.”

All the New Classical economists need to defend the dominant “paradigm” in economics against Krugman and other dissenters are the tools he grants them—maximization and equilibrium.

4. Uncertainty and reflexivity

One thing that’s missing from Krugman’s treatment of useful economics is the explicit recognition of what Keynes and before him Frank Knight, emphasized: the persistent presence of enormous uncertainty in the economy. Most people most of the time don’t just face quantifiable risks, to be tamed by statistics and probabilistic reasoning. We have to take decisions in the prospect of events–big and small–we can’t predict even with probabilities. Keynes famously argued that classical economics had no role for money just because it didn’t allow for uncertainty. Knight similarly noted that it made no room for the entrepreneur owing to the same reason. That to this day standard economic theory continues to rules out money and excludes entrepreneurs may strike the noneconomist as odd to say the least. But there it is. Why is uncertainty so important? Because the more of it there is in the economy the less scope for successful maximizing and the more unstable are the equilibria the economy exhibits, if it exhibits any at all. Uncertainty is just what the New Classical neglected when they endorsed the efficient market hypothesis and the Black-Scholes formulae for pumping returns out of well-behaved risks.

If uncertainty is an ever present, pervasive feature of the economy, then we can be confident, along with Krugman, that New Classical models wont be useful over the long haul. Even if people are perfectly rational too many uncertain, “exogenous” events will divert each new equilibrium path before it can even get started.

There is a second feature of the economy that Krugman’s useful economics needs to reckon with, one that Keynes and after him George Soros, emphasized. Along with uncertainty, the economy exhibits pervasive reflexivity: expectations about the economic future tend to actually shift that future. This will be true whether those expectations are those of speculators, regulators, even garden-variety consumers and producers. Reflexiveness is everywhere in the economy, though it is only easily detectable when it goes to extremes, as in bubbles and busts, or regulatory capture. In the run up to the financial crisis expectations that real estate values would continue to increase caused real estate values to continue to increase (e.g., by encouraging banks to lend on future values instead of current collateral).

Bubbles are only the most obvious of these reflexive processes in the economy and the culture. Fashion, whether in clothes or in Iphones, is equally reflexive. Enough people thinking that the 5S (and not the 5c) will be a fashion make it one. Reflexivity is everywhere: the entire monetary economy is based on reflexive expectations: think about everyone’s willingness to accept pieces of paper in exchange for goods and services. I accept your paper money because I expect everyone else will accept it, and every one else will because they believe everyone else will. The reality of money is the product of reflexive expectations. Fortunately the paper money bubble bursts only in times of hyperinflation. How much inflation is needed for the breakdown to occur? That is a matter of uncertainty. It depends on, among other things history, and people’s knowledge of that history. But it can also break down for reasons no one ever expected (consider the scenario of The Walking Dead).

Of course the New Classical economists recognized the importance of reflexivity in the economy. Their whole argument for the ineffectiveness of Keynesian policy turns on it. On their model, inflation and deficit financing might have worked to fool people into spending more for a millenium before the 1970s. But as soon as one set of economic agents—the government—started purposely using it to fool another set of agents—businessmen–into thinking that the economy was growing, the smarter ones among them caught on, and recovery from recessions could no longer be produced by Keynesian fine-tuning. The New Classical economists have the models—simple and complex—to prove it.

Reflexiveness of course played a huge role in bringing on the financial crisis and the Great Recession. It was the New Classical vision of the market as efficiently and instantaneously digesting all information that led to the abrogation of the Glass-Steagall Act, and then made it possible for the banks to employ the risk-allocation formulae of the hedge-fund managers. These equations had no variables for uncertainty, only for risk. Neglecting uncertainty brought on the collapse of the subprime mortgage market.

When combined uncertainty and reflexivity greatly limit the power of maximizing and equilibrium to do useful economics. Reflexive relations between future expectations and outcomes are constantly breaking down at times and in ways that no one can predict—about which there is complete uncertainty. Between them, they make the economy a moving target for the economist. Models get into people’s heads and change their behavior, usually in ways that undermine the model’s usefulness to predict.

Which models do this and how they work is not a matter of quantifiable risk, but radical uncertainty. Of course it’s not just economist’s models that get into people’s heads, lots of other models—“metaphors” Krugman calls them—do so and at rates and with reflexive effects that are radically uncertain. And that’s what makes the economy, and human affairs generally, a target moving too fast for anything more than simple models, whether its in economics, political science, sociology, etc.

Between them reflexivity and uncertainty make economics into a retrospective, historical science, one whose models—simple or complex—are continually made obsolete by events, and so cannot be improved in the direction of greater predictive power, even by more complication. The way expectations reflexively drive future economic events, and are driven by past ones, is constantly being changed by the intervention of unexpected, uncertain, exogenous ones. No bubble ever repeats, because people remember them. Regulators engaged in preventing their recurrence are just setting the stage for new and different ones. Dodd-Frank is just an incentive to figure out new ways to exploit regulations.

There is room for both uncertainty and reflexiveness in Krugman’s philosophy of economics. For he recognizes that economics has to be a historical science.

‘History provides “the real issues and the real experiences that need explaining.” Indeed, his best examples of why history matters incorporate both uncertainty and reflexiveness: “In macro, in particular, you need to know about drastic events”–the German hyperinflation for example. See here. Drastic events, ones no one could see coming—uncertain ones–punctuate economic life. All too often they break up a comfortable predictable reflexive relationship between expectations and reality. They make economics the discipline Krugman thinks it is, like history, one we may look to for lessons but not predictions.’

ABOUT THE AUTHOR

Alex Rosenberg is the R. Taylor Cole Professor of Philosophy and chair of the philosophy department at Duke University. He is the author of Economics — Mathematical Politics or Science of Diminishing Returns, most recently, The Atheist’s Guide to Reality.

First published in 3:AM Magazine: Sunday, September 14th, 2014.