Winner of the New Statesman SPERI Prize in Political Economy 2016

Sunday 16 December 2012

Mistaking models for reality

In a recent post, Paul Krugman used a well known Tobin quote: it takes a lot of Harberger triangles to fill an Okun gap. For non-economists, this means that the social welfare costs of resource misallocations because prices are ‘wrong’ (because of monopoly, taxation etc) are small compared to the costs of recessions. Stephen Williamson takes issue with this idea. His argument can be roughly summarised as follows:

1) Keynesian recessions arise because prices are sticky, and therefore 'wrong', so their costs are not fundamentally different from resource misallocation costs.

2) Models of price stickiness exaggerate these costs, because their microfoundations are dubious.

3) If the welfare costs of price stickiness were significant, why are they not arbitraged away?

I’ve heard these arguments, or variations on them, many times before.[1] So lets see why they are mistaken, taking the points in roughly reverse order.

Keynesian recessions arise because of deficient demand. If you want to think of this as being because some price is wrong, in my view that price is the real interest rate. Now flexible wages and prices might get you the right real interest rate, either because they encourage monetary policy to do the right thing (by changing inflation), or because a particular monetary policy combines with inflation expectations to generate the appropriate real interest rate. However when nominal interest rates hit zero and there are inflation targets, flexible prices may not be enough (as argued here), so there may be no flexible price solution that gets rid of the costs of recession. At the very least, that suggests that recessions are a bit different from, say, the costs of monopoly or distortionary taxation. It also tells you why they cannot be arbitraged away by the actions of individuals.(See also Nick Rowe on this.)

What we have in a recession is a coordination problem. If everyone were to spend more, the additional output would generate incomes that matched the spending. If monetary policy cannot induce that coordination, then individuals could try and persuade someone with a great deal of spending power who could borrow freely and very cheaply to embark on additional spending. The obvious someone is the government, and the real puzzle is why governments have been so reluctant to arbitrage away recessions in this way.  

The second point is horribly wrong, and it explains the title of this post. The problem with modelling price rigidity is that there are too many plausible reasons for this rigidity - too many microfoundations. (Alan Blinder’s work is a classic reference here.) Microfounded models typically choose one for tractability. It is generally possible to pick holes in any particular tractable story behind price rigidity (like Calvo contracts). But it does not follow that these models of Keynesian business cycles exaggerate the size of recessions. It seems much more plausible to argue completely the opposite: because microfounded models typically only look at one source of nominal rigidity, they underestimate its extent and costs.

I could make the same point in a slightly different way. Lets suppose that we do not fully understand what causes recessions. What we do understand, in the simple models we use, accounts for small recessions, but not large ones. Therefore, large recessions cannot exist. The logic is obviously faulty, but too many economists argue this way. There appears to be a danger in only ‘modelling what we understand’ that modellers can go on to confuse models with reality.

Lets move from wage and price stickiness to the major cost of recessions: unemployment. The way that this is modelled in most New Keynesian set-ups based on representative agents is that workers cannot supply as many hours as they want. In that case, workers suffer the cost of lower incomes, but at least they get the benefit of more leisure. Here is a triangle maybe (see Nick Rowe again.) Now this grossly underestimates the cost of recessions. One reason is  heterogeneity: many workers carry on working the same number of hours in a recession, but some become unemployed. Standard consumer theory tells us this generates larger aggregate costs, and with more complicated models this can be quantified. However the more important reason, which follows from heterogeneity, is that the long term unemployed typically do not think that at least they have more leisure time, so they are not so badly off. Instead they feel rejected, inadequate, despairing, and it scars them for life. Now that may not be in the microfounded models, but that does not make these feelings disappear, and certainly does not mean they should be ignored.

It is for this reason that I have always had mixed feelings about representative agent models that measure the costs of recessions and inflation in terms of the agent’s utility.[2] In terms of modelling it has allowed business cycle costs to be measured using the same metric as the costs of distortionary taxation and under/over provision of public goods, which has been great for examining issues involving fiscal policy, for example. Much of my own research over the last decade has used this device. But it does ignore the more important reasons why we should care about recessions. Which is perhaps OK, as long as we remember this. The moment we actually think we are capturing the costs of recessions using our models in this way, we once again confuse models with reality.

[1] A classic example comes from Robert Lucas. This includes the rather unfortunate statement that the “central problem of depression prevention has been solved”, but I don’t think that should be used as evidence against the more substantive claim of the paper, which is that the gains from stabilising the business cycle are relatively small. This assertion has been criticised even if we stick with New Keynesian representative agent models (see this paper by Canzoneri, Cumby and Diba), but the problems I outline below are more fundamental.

[2] For non-economists: twenty years ago most Keynesian analysis measured the success of policy (social welfare) by how well it stabilised inflation and the output gap, and the relative importance of inflation compared to output was a ‘choice for policy makers’. Since work by Michael Woodford, a similar measure of social welfare can be derived from the utility of individual agents, often using pages of maths, but the importance of output compared to inflation is then a function of this utility and the model’s structure and parameters.    


  1. "the long term unemployed typically do not think that at least they have more leisure time, so they are not so badly off. Instead they feel rejected, inadequate, despairing, and it scars them for life."

    Unemployment also kills, not just by suicide, but by physical illness.

    See, for instance, Unemployment as a disease and diseases of the unemployed. Scandinavian Journal of Work, Environment & Health, 23(suppl 3),79-83. ( ).

    Unemployment is a public health problem that affects not only the unemployed but people close to them.

  2. "In terms of modelling it has allowed business cycle costs to be measured using the same metric as the costs of distortionary taxation and under/over provision of public goods, which has been great for examining issues involving fiscal policy, for example. "

    I think something can be measured if the resulting measurements approximately reflect reality -- in most contexts a measurement is something one does with a real thing when trying to describe it. A yardstick* is not put up against a theory to measure its length.

    In the rest of the post, you argue that the costs of business cycles are grossly miss-measured if one uses representative agent models. [rudeness deleted -- I am honestly trying].

    I think that the representative agent approach to assessing (not measuring) welfare is much less appropriate in micro even than in macro. There is simply no way to understand why we have progressive taxation using a representative agent. How do you model social insurance assuming complete perfect insurance markets to begin with ? Yes the analysis is used a lot and is influential. In that sense it is great. But I see nothing good in it.

    * or "meter stick" "metre stick" "metre" ??? I'm drowning in the Atlantic even though I'm sitting inItaly)

  3. "Keynesian recessions arise because of deficient demand."

    Wrong. That language is not descriptive of what is going on in those models. It's either sticky wages and prices or its a coordination failure (multiple equilibrium) phenomenon. People use the "deficient demand" language for marketing purposes. To your average lay person, it sounds good.

    "...that price is the real interest rate."

    Sure, the relative prices are screwed up on that margin too, for example in a New Keynesian model.

    " However when nominal interest rates hit zero and there are inflation targets, flexible prices may not be enough..."

    That's not correct in the standard NK model.

    "What we have in a recession is a coordination problem."

    Different model now. Measuring the welfare loss is now a different problem of course. Note that, in coordination failure models, the policy solutions tend to look very "non-Keynesian." The ones that work well are those that kill off the bad equilibria - doesn't look anything like "demand management" for example.

    "The problem with modelling price rigidity is that there are too many plausible reasons for this rigidity - too many microfoundations."

    No, the problem is that there are no microfoundations. No one has written down a serious model of nominal contracting with stickiness. I know some models, however, that exhibit stickiness, and where money is neutral - e.g. Berentsen/Menzio/Wright.

    1. Steve: if barter is costly, so people use monetary exchange, and if prices are sticky in terms of money, and the money supply falls, so there's an excess demand for the medium of exchange, why can't we call that "deficient demand" (for non-money goods)?

    2. "Men cannot be employed when the object of desire (i.e., money) is something which cannot be produced and the demand for which cannot be readily choked off” (GT,1936, 235).

      “If we sought to condense Keynes’s whole thesis concerning employment into a single sentence, “we might say that he ascribes the possibility of involuntary general unemployment to the existence of a liquid asset in a world of uncertainty” (G.L.S. Shackle,1974, 28).

    3. Nick,

      How are you imagining we can have an "excess demand for the medium of exchange?" You're thinking asset prices are sticky too?

    4. I'd really like a reference (or explanation) to how greater price flexibility reduces the ZLB problem under inflation targeting in a NK model. But I also want to ask this. What do you conclude if you think there is no serious microfoundation for price stickiness?
      1) that we have to fall back on less serious models (like Calvo)
      2) that price stickiness does not exist

    5. Neither. More work needs to be done. It seems surprising, but Keynesian economics is very much unfinished business.

    6. That is a dodge. The debate is about the advice we give now, with the knowledge we have.

    7. The problem that New Keynesian models simply refuse to acknowledge is that once you recognize prices are sticky there's no reason to assume quantities are more flexible. If we have all sorts of non walrasian contracts, a lot of the observed price or wage stickiness is likely to be accompanied by sticky quantities. The quantities that adjust need not be affected by stickiness. But what matters for new keyensian economics is not individual level stickiness of some prices, but aggregate price level stickiness. Once you abandon walrasian contracts and supply/demand reasoning the link between the 2 is no longer obvious (see models of long term contracts and the models by Menzio et al mentionned by Williamson). Perhaps another way to think of it. We know at some basic micro level the housing decision (or the large business capital investment)is quite sticky due to fixe adjustment costs. You usually don't move that often. But aggregate investment is among the most volatile componenents of GDP. Oh, and the ZLB is not a problem with sticky prices, because Phillips curve logic does not work with flexible prices (except as an occasional correlation that may appear for example if the central bank follows a Taylor rule and the Fisher equation holds, or in some models with monetary frictions). In a flex price equilibrium supply and demand for financial assets determine real interest rates. Expected inflation adjusts to make this compatible with the central bank policy rate, or there is some decoupling of the relevant risk free rate for the economy and the central bank rate (due to some market segmentation). Also Keynesian economists would not be so dogmatic in their insistence that the output gap is the name of the game if they paid more attention to all the research about how things like financing constraints, market incompleteness and uncertainty or sentiment shocks can cause important, potentially highly inefficient fluctuations in the flexible price output level. But my sense is that Krugman still thinks he's fighting Kydland and Prescott back in 1982. He may be right in the sense that the pure RBC model without any inefficiencies is more friendly to the political positions of the Krugmans's republican/conservative devils, but in doing that he ends up dismissing a lot of relevant and serious macroeconomics.

    8. Steve: "How are you imagining we can have an "excess demand for the medium of exchange?" You're thinking asset prices are sticky too?"

      I have this argument with keynesians all the time!

      Start with a model in which *all* prices are sticky. Cut the money supply (or increase money demand) and we get an excess demand for money, an excess supply of everything else, and a recession.

      Now suppose the price of one little good, say peanuts, is flexible. The peanut market now clears, at a lower price for peanuts. But we still have a recession. Do we say that there is no excess demand for money, because you can always sell peanuts to get money? Do we blame the recession on the price of peanuts being out of line with other prices?

    9. Nick,

      OK. In a Keynesian world, some goods prices get fixed, and then the quantities and asset prices move to give you an equilibrium, with some assumption about how the quantities achieve that. Usually, it's a demand-side rule, which is arbitrary, but important. All the markets are clearing, in the sense nothing is left on the table. But given market prices, some firms are not happy with what they are producing. In the models I know about, the assets - including money - are all willingly held, given the price level and asset prices. I don't know what you could mean.


      No, I'm not dodging anything. I give advice all the time, given the knowledge I (we) have. Suppose I'm talking to people in central banks, which I do sometimes. They want to know what they should be doing in the short run. They think, and I think, that money is not neutral in the short run. I know some mechanisms which can give me that, one of which is sticky prices and wages. But I'm not entirely sure that the nonneutralities I am seeing are due to sticky prices and wages. What do I tell the central banker? Some of this depends on how he/she thinks about the problem. Maybe he/she is a hardcore Keynesian, maybe he/she is more open-minded. I just try to be helpful.

    10. Steve:

      1. Small point. The demand side rule *looks* arbitrary, agreed. (Why isn't it the short side rule instead??). To my mind, one of the big advances of New Keynesian macro is that it fixed this problem. If you replace perfectly competitive firms with monopolistically competitive firms, start in equilibrium then hold prices fixed, then suppose demand increases, a monopolistically competitive firm will happily increase sales, as long as P remains above MC. Firms won't hit their MC curves (and ration buyers) unless there is a very big increase in demand. (Perfectly competitive firms would hit their MC curves, and ration buyers, almost immediately).

      2. In my little peanut parable, all stocks of peanuts are willingly held too. And nobody, in equilibrium, will be unable to trade money for peanuts or vice versa. But that does not mean that the stock of money is equal to the desired stock of money. In all the other markets in which money is traded, there will be an excess demand for money/excess supply of the other good.

      My "peanuts" are a metaphor for "those assets and goods which have flexible prices".

  4. Couldnot you solve most of it by making the economic formula/model PART of the decisionmaking and not effectively being the total decisionmaking.
    Nice list with pro and cons, economic model outcome being one (probably important one).
    Pro/cons including how strong is your model (in general and under the current circumstances). Items put at zero (nil) that might be of more importance than that than in the research on which the model was based, etc..

    Subsequently you could make a new model for the decisionmaking, but probably donot have time for that in most situations.

  5. "The moment we actually think we are capturing the costs of recessions using our models in this way, we once again confuse models with reality."

    The idea seems to be that you know what the costs of recessions are, and if the models don't give you the answer you want, they must be wrong.

    1. I just look at the evidence, like the study I linked to in my post.

    2. Click feelings at the end of the penultimate paragraph, or look at any of the happiness literature.

    3. It's not at all obvious where we go with this as economists. We have been successful as a species in part because we have evolved to "feel good" about work, not only because this allows us to eat, but because this is met with approval from other human beings. Of course, society also puts a value on work that is not rewarded in the market. But ideas about that change over time. For example, women seem to feel much differently about market work vs. homework than they did 50 years ago. Some economists think that those ideas are driven by technology, i.e. the social attitudes are not causal. But those are long-run issues, and you're thinking about the short run. The key question is, what is the key inefficiency that exists in a recession, and is there any kind of government intervention that can correct it, or at least mitigate it? I'm not sure the happiness studies are going to lead you where you want to go. This might tell you the government should just be handing out anti-depressants.

    4. Another thought. Suppose we grant that there is something we are not taking into account in standard approaches, related to the misery of unemployment. Suppose also that the inefficiency comes from sticky wages and prices. Then the fact that unemployment is miserable in ways we have not been modeling does not change how you think about the inefficiencies, relative to the distortions from taxation, for example. A tax will reduce the surplus from exchange (between workers and firms), and increase the unemployment rate. It's the same inefficiency.

  6. The key feature of unemployment in recession is not that it is long-term but that it is short-term. Large numbers of people are unemployed, or under-employed, who had previously been fully employed and reasonably expect to be so again. Simon is correct that the long-term unemployed don't regard their unemployment as "leisure", and when short-term unemployment becomes long-term there is a personal cost in physical and mental health problems, relationship breakup and even death. However, those who regard their unemployment as short-term don't view it as leisure either. They are working full-time (unpaid) looking for work. NK models that regard leisure as the positive side of unemployment (the negative side being loss of income) are behaviourally very, very far wrong.


Unfortunately because of spam with embedded links (which then flag up warnings about the whole site on some browsers), I have to personally moderate all comments. As a result, your comment may not appear for some time. In addition, I cannot publish comments with links to websites because it takes too much time to check whether these sites are legitimate.