Winner of the New Statesman SPERI Prize in Political Economy 2016

Sunday 9 February 2014

Speaking as an Old New Keynesian …

Labels are fun, and get attention. They can be a useful shorthand to capture an idea, or related set of ideas. But is there really a New Old Keynesian school of thought? I don’t think so. Here are a couple of bold assertions, which I think I believe, and which I will try to justify. First, in academic research terms there is only one meaningful division, between mainstream and heterodox. (Of course the heterodox divide themselves up into various ‘schools’, but their size is small and their influence is also small.) Second, in macroeconomic policy terms I think there is only one meaningful significant division, between mainstream and anti-Keynesians.

But before trying to justify these statements, I want to defend being a killjoy. As I said, putting people into categories can be fun - why spoil it by taking the exercise seriously? Two reasons. First, I want to make some points which do not get said often enough on economics blogs. Second, labels can lead to confusion or worse. Just think about the label Keynesian. Any sensible definition would involve the words sticky prices and aggregate demand. Yet there are still some economists (generally not academics) who think Keynesian means believing fiscal rather than monetary policy should be used to stabilise demand. Fifty years ago maybe, but no longer. Even worse are non-economists who think being a Keynesian means believing in market imperfections, government intervention in general and a mixed economy. (If you do not believe this happens, look at the definition in Wikipedia.)

So what do I mean by a meaningful division in academic research terms? I mean speaking a different language. Thanks to the microfoundations revolution in macro, mainstream macroeconomists speak the same language. I can go to a seminar that involves an RBC model with flexible prices and no involuntary unemployment and still contribute and possibly learn something. Equally an economist like John Cochrane can and does engage in meaningful discussions of New Keynesian theory (pdf). [1]

Of course all academic macroeconomists have their own idea of how the world actually works and will probably do research using models that roughly conform to that. Yet I think you would be hard put to draw meaningful boundaries here. Take John Quiggin’s Old/New Keynesian post, for example (which followed this from Tyler Cowen). He characterises New New Keynesians as those still working with DSGE models who are now attempting to add financial frictions. He wants to argue (and labels New Old Keynesian) the idea that following this recession, there “is no unique long-run equilibrium growth path, determined by technology and preferences, to which the economy is bound to return. In particular, the loss of productive capacity, skills and so on in the current depression is, for all practical purposes, permanent.” Now listening with my mainstream ears, this sounds like a combination of hysteresis effects and endogenous growth, which sounds interesting. Yet I also think we can learn a lot from adding financial frictions to DSGE models. Does this make me a middle aged Keynesian?

What I suspect Quiggin is getting at here is that New New Keynesians are still following a microfoundations research programme (using DSGE), whereas he would not. Now many mainstream macroeconomists, myself included, can be pretty critical of the limitations that this programme can place on economic thinking, particularly if it is taken too literally by microfoundations purists. But like it or not, that is how most macro research is done nowadays in the mainstream, and I see no sign of this changing anytime soon. (Paul Krugman discusses some reasons why here.) My own view is that I would like to see more tolerance and a greater variety of modelling approaches, but a pragmatic microfoundations macro will and should remain the major academic research paradigm.

When it comes to macroeconomic policy, and keeping to the different language idea, the only significant division I see is between the mainstream macro practiced by most economists, including those in most central banks, and anti-Keynesians. By anti-Keynesian I mean those who deny the potential for aggregate demand to influence output and unemployment in the short term. [2] Why do I use the term anti-Keynesian rather than, say, New Classical? Partly because New Keynesian economics essentially just augments New Classical macroeconomics with sticky prices. But also because as far as I can see what holds anti-Keynesians together isn’t some coherent and realistic view of the world, but instead a dislike of what taking aggregate demand seriously implies.

What is incoherent about believing in pretty flexible prices, you might ask? Two things. First, as I have argued before, with the demise of the Pigou effect flexible prices do not get you out of a deficient demand problem at the zero lower bound when there are inflation targets. Second, the evidence that prices are not flexible is so overwhelming that you need something else to drive you to ignore this evidence. Or to put it another way, you need something pretty strong for politicians or economists to make the ‘schoolboy error’ that is Says Law, which is why I think the basis of the anti-Keynesian view is essentially ideological.

Of course there are a huge number of policy debates in macroeconomics, and you can attach labels to those if you like. Should we use fiscal stimulus at the zero lower bound, for example. Was austerity a good idea? However, anti-Keynesians aside, I don’t think these debates reveal large fault lines in economic thinking. Economists do not rigidly line up on one side or another, and some even change their mind over time as the facts change. It is possible to have serious discussions about the effectiveness of monetary policy, the dangers of high debt etc. The only group where a discussion can fail to get off the ground is with those who contend that aggregate demand is always irrelevant.

[1] Heterodox economists might argue that they have to be bilingual - they are able to speak mainstream, even if they prefer not to among friends. Those more critical might detect a reluctance to get past certain words.

[2] An alternative, and more positive, way to define the anti-Keynesian group is that they believe macroeconomic outcomes are essentially efficient, and so intervention by a government (or central bank, beyond providing a nominal anchor) is not required. This difference might be important in placing someone like Roger Farmer (who I’m glad to see now has a blog), who is not an anti-Keynesian under this positive definition, but might be using my more negative criteria.


  1. I could put this in a more historically vivid way and ask the question:

    Do you think Chancellor Bruening and the SDP, in retrospect, went in the right direction in economic policy after 1931?

  2. I agree with this a lot more than I disagree.

    One minor problem: By your definition of "Keynesian", most (all?) economists before Keynes were Keynesian, though they might not have used the words "aggregate demand", they would, like Keynes, have invoked money and sticky prices to explain the trade cycle, all they way back to Hume. "Monetarist" vs "anti-Monetarist" would work as well.

    "Or to put it another way, you need something pretty strong for politicians or economists to make the ‘schoolboy error’ that is Says Law, which is why I think the basis of the anti-Keynesian view is essentially ideological."

    Dunno. Say's Law may work empirically, if prices are perfectly flexible. It may be more a refusal to believe anything they cannot model, like sticky prices.

    "Now listening with my mainstream ears, this sounds like a combination of hysteresis effects and endogenous growth, which sounds interesting."

    Yep. Take a standard AK growth model, a standard NK business cycle model, and put them together, and bad monetary policy will have permanent real effects, because it affects investment. (The investment-Euler equation would look a bit weird, because it would keep on hitting a corner solution if the central bank set r either above or below A-d.)

    1. So how would you define Keynesian? If all economists before Keynes believed that output was demand rather than supply determined, what exactly did Keynes bring to the table? And while we are at it, how do you define a monetarist nowadays?

  3. Part of what defines a Keynesian (new or old), is that a Keynesian thinks that his or her views are "mainstream," and that the rest of macroeconomic thought is defined relative to what Keynesians think - Keynesians reside at the center of the universe, and everything else revolves around them. The truth is that there are few macroeconomists who would take the position that it is useful to think about macroeconomics in terms of pure Arrow-Debreu economies. Most of us are concerned with frictions of various kinds - search frictions, private information frictions, all manner of credit frictions, etc. There are also frictions associated with sticky wages and prices, and multiple equilibrium phenomena, which arise because of coordination frictions. So, what frictions should we focus on? My interest is in the frictions that gum up financial markets, the frictions that might give rise to financial crises, that imply a role for collateral and asset trade in the economy. My prior is that those things are very important, especially given our recent experience, and that sticky price and sticky wage frictions probably don't matter much. I'm not "against" Keynesians. They are free to work on those problems if they want, and I think it's important that some people do that. But that's not the only game in town.

    1. How is that different from the financial frictions that Simon mentions? See:

      "Yet I also think we can learn a lot from adding financial frictions to DSGE models. Does this make me a middle aged Keynesian?"

    2. Actually, we could argue that this is now the "mainstream." Every central bank now has a group of people working on quantitative models with financial frictions. In the latest crop of new PhDs doing macro at top schools, there are few sticky price papers, and a lot of work on financial economics. I have seen some financial frictions papers recently that include standard sticky prices, but that's not the focus of the paper - it's about modeling the credit frictions or labor market frictions that will replicate the key features of the financial crisis.

    3. I was careful to distinguish between academic research and policy. What you say about academic research is right - there are lots of frictions to work on, and it is often useful to abstract from some frictions to focus on others. At the moment, financial frictions are the game in town. I did NOT say the mainstream of academic research was Keynesian.

      But the game in town from a policy point of view is the recession, and whether anything that raises aggregate demand will increase output (because it raises demand). I would call those who would agree Keynesian and those who do not anti-Keynesian. From the way that central bankers talk, most appear Keynesian from this point of view.

  4. I'm not an economist, but I read a lot of these blogs (Thoma, Krugman, Delong, etc.). I think a key point you could make in describe in "Keynesians" is the zero lower bound. And default currency. And sticky wages. Of course Say's law is idiotic. Listen I invested a lot in Apple 3 years ago (a lot for me heh). They're sitting on a 150 billion dollars and aren't spending it. For 3 years now.

    But the ZLB: I don't see how this is complicated and not empirically proven. If the fed can lower interest rates below nominal zero, private companies are hording cash, consumers have a lot of debt due to a housing bubble, aggregate demand goes into the toilet. On only fiscal policy can flush that toilet.

  5. «those who contend that aggregate demand is always irrelevant. »

    That is surely a bit exxagerated, because "those" argue that aggregate demand is always *sufficient*, to ensure full employment of all factors, except for the effects of the preference for leisure of lazy workers.

    Equivalently in their view that involuntary unemployment cannot exist, because there is always a wage low enough to ensure full employment of labourers to whom the value of leisure is smaller than that wage.

    If one admits that such a wage may be zero or negative, they are surely right: workers who are not lazy, that is they don't have a preference for leisure, will find employment by working for free or paying employers to let them work.

  6. «multiple equilibrium phenomena, which arise because of coordination frictions.»

    As Sraffa conclusively proved multiple equilibriums and thus multiple distributions of income arise entirely naturally even in ridiculously simple economies. And as Steve Keen has summarized from that and others, it is also because aggregate demand schedules are not monotonic unless all buyers have the same demand schedule.

  7. «as far as I can see what holds anti-Keynesians together isn’t some coherent and realistic view of the world, but instead a dislike of what taking aggregate demand seriously implies.»

    I think that diving ecoomists in two camps based on non-sticky prices or semi equivalently demand always being *sufficient* to ensure full *voluntary* factor employment is a very shallow and un-enlightening position.

    Because that prices are flexible and/or demand is sufficient is an epiphenomenon, a consequence of a more fundamental verity, which is the central verity of aligned, sponsored Economics:

    *the distribution of income is always justified by factor productivity except for distortive government intervention*

    That is the true dividing line, that is what distinguishes true Economists from mere political economists. The belief that theory must prove above everything else the verity that the distribution of income is one, true and holy, unless corrupted by the sins of the government.

    Things like the full flexibility of prices, the veil of money, Say law, the merry lack of a theory of capital, rational expectations, the identical shape of demand schedules, the infinite number of markets with infinite participants, and many other comical shysterisms are all justified by the necessity to have a theory in which it is possible to "prove" the central verity of Economics.

    That is the what makes all anti-keynesian views coherent and realistic of the world of careers in Economics.

    That's why Keynesian teaching was blacklisted by business and property interests, instinctively, in USA academia.

  8. Sounds like a request for two armies to continue fighting over barren land.

  9. "We're All [mostly] Monetarists Now"

    I agree with Nick, only with more emphasis that the debate should be framed in terms of monetarism.


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.