Friday, 27 January 2012

The Return of Schools of Thought Macro

When I first studied macro, it was all about ‘schools of thought’. Keynesians, Monetarists, New Classicals, and probably many more I cannot remember. Macroeconomists tended to take sides. Antagonists often talked across each other, and anyone not already on one side just got totally confused. I recall reading one textbook on international macro where each chapter represented an alternative ‘view’, with no clear idea of how each view or school was related to another. One thing that was pretty clear, however, was that most schools of thought could be identified with a particular ideological position.
                But then things began to change. The discipline appeared to become much more unified. It would be going much too far to suggest that there was a general consensus, but to use a tired cliché, most macroeconomists started talking the same language, even if they were not saying the same thing. I think there were two main reasons for this. The first was microfoundations: deriving the components of macro models from standard optimisation applied to representative agents. This gave macroeconomics the potential to achieve the same degree of unity as microeconomics. The second was the development of New Keynesian theory, which allowed an analysis of aggregate demand within a microfounded framework, and which integrated ideas like rational expectations and consumption smoothing into Keynesian analysis. To use the jargon, all models were now DSGE models.
                Goodfriend and King coined the term ‘New Neoclassical Synthesis’ (call it ‘synthesis’ for short), and other authors wrote along similar lines. So, even as recently as five years ago, I told masters students starting a macro course to forget anything they might have been told about alternative schools of thought: they were going to learn a unified framework that most macroeconomists – the mainstream – would sign up to. It was like the first half of David Romer’s popular textbook: start with Solow, but quickly replace a fixed savings propensity by an optimising intertemporal consumer to get the basic Ramsey model. Add endogenous labour supply to get RBC. Probably talk a bit about overlapping generations. Hopefully add to what was in Romer by doing some open economy stuff. Then add New Keynesian theory built around sticky prices. If the student went on to work in a central bank, they would probably encounter this framework as a central part of that institution’s forecasting and policy analysis.
I think this synthesis and the reasons behind it may have had one or two unintended and unfortunate consequences. Sometimes the emphasis on microfoundations became a bit of a fetish. (I have written about the shaky methodological grounds on which ‘microfoundations purists’ sometimes stand here.) Some have suggested that the emphasis on microfoundations meant too much time was devoted to elements that were easy to model within that framework rather than the things that really mattered. But I personally thought this synthesis had many more positive than negative consequences. I would not wish for every single macroeconomist to sign up to the synthesis: there is an important role for heterodox economists. However I liked the fact that the majority of macroeconomists used the same analytical framework.
                Just five years later and it seems rather different. I’ve encountered a much larger range of economics blogs in the last week or two (you can guess why), and it does feel like going back in time. Schools of thought in macro are definitely back. Since the recession it has become clear that the synthesis had not been adopted everywhere. In particular, in sections of the profession there remained a suspicion (to put it mildly) of New Keynesian theory, and partly as a consequence of this the amount of this theory that was taught to graduates differed widely.
                It is true that for some, schools of thought can be quite fun. Some students find the idea that academic macroeconomics consists of opposing forces locked in combat adds a degree of interest and motivation that might otherwise be lacking. However I am not persuaded that this spice was sufficient to offset misunderstanding. Personally when I was a student I found all the motivation I needed from socially destructive inflation, and widespread unemployment should do the same today. I do think that the schools of thought approach leads to an inexactness which can be misleading and annoying.
                Take the label Keynesian. Look up Wikipedia, and in the third paragraph you will find ‘Keynesian economics advocates a mixed economy — predominantly private sector, but with a significant role of government and public sector...’. Now I have a much more limited idea of what Keynesian economics is. For me, Keynesian macro is business cycle analysis based on aggregate demand and sticky prices. By this definition, the only ‘significant role for the public sector’ required is a central bank. Even in the rather unusual (I hope) times of a zero lower bound, Keynesian advocacy of fiscal stimulus implies is that the government brings forward its bridge building, and not that it permanently build more bridges. Does my preferred definition make me narrow-minded?
 Keynesian analysis as I define it implies that you need monetary policy, and occasionally countercyclical fiscal policy, to stabilise the economy, but that is not what is generally meant by a mixed economy! The extent of the public sector’s involvement in the economy will depend on microeconomics, not macro. Now it is true that those who tend to be antagonistic to state intervention may be uncomfortable with monetary and fiscal stabilisation policy, as I have suggested, but I am against ideology clouding economics, and I certainly do not want this connection hard wired as a school of thought.
                School of thought thinking also tends to bracket ways of looking at the economy with policy prescriptions, even when they are not inextricably linked. Take countercyclical fiscal policy, for example. Is that an intrinsic part of Keynesian thinking? For some time there has been general agreement among most macroeconomists that monetary policy was the stabilisation tool of choice, because of issues like implementation lags. This view has been strengthened by analysis over the last ten years that explicitly looks at welfare derived from a representative agent’s utility: the analysis of a simple basic case is contained in the Woodford paper I referenced here, and some of my own collaborative work has shown this result is surprisingly robust. So linking the routine use of countercyclical fiscal policy to what I think of as Keynesian theory is just misleading.
                  This sort of bundling together of ideas under one label at the very least causes confusion. (Here Jonathan Portes gives one recent example.) Worse still, it leads people to take sides on issues not because of the merits of the case, but because that case is associated with a school of thought whose other elements they do or do not like. I also miss the synthesis. I very much liked the idea that disagreements could be clearly located within a common framework. With the synthesis, I felt macroeconomics began to look more like a unified discipline - more like micro, and dare I say it, more like a science than a belief system. 


  1. I have never understood the microfoundations attraction. The Kirman critique - that there is no reason why aggregate consumer behaviour should look like the decisions of a single optimising agent -seems to me quite devastating.

    1. Bingo. It's the determined IGNORING of this critique (a fairly obvious one, in truth) is one reason that I think the claim simon makes - that he is "against ideology clouding economics" - is self-deluding.

      If representative agent methods didn't tend to one broad general tendency - that, just as in lots of micro, individual markets (including markets for intertemporal capital allocation) are welfare-maximising and so governments should butt out - this critique would not have been ignored, any more than Lucas' (very effective) critique of Old Keynesianism.

  2. I don't think the schools of thought ever really disappeared. There have always been economists working in the "freshwater" tradition of market-clearing models. Monetarists begat the New Classicals, who gave us RBC theory etc. Keynesians have obviously been influenced by these schools of thought, and today Friedman would have more in common with New Keynesians than with New Classicals.

    As far as I can see, the New Neoclassical Synthesis never really caught on; it is really synonymous with New Keynesian theory. In fact, I see it as attempt to promote New Keynesian ideas as standard economics, whereas freshwater types do the same thing by describing their models as "neoclassical".

    I think having different schools of thought is probably healthy. The problem as I see it is that both schools want to present their views as science. By any reasonable standard economics is not science, so attempts to appear more scientific result in a descent into pseudoscience. Moreover, as macroeconomic research becomes increasingly arcane and technical (in the sense that it uses ever more advanced mathematics and statistics) it becomes harder to scrutinize. The layman is forced to either accept economic research at face value or dismiss it out of hand.

    I think it would be far better if economics debates recognized that there are differences of opinion regarding how the economy operates, just as with politics people have different views about how the world works.

  3. This is a terrific post and I'm going to assign it to my macro students. However, despite the reasonable tone, this viewpoint remains too restrictive. Economics is not just neoclassical economics, and to advocate a consensus that privileges neoclassical theory will exclude other valid points of view, from traditional schools like neo-Marxists or neo-Ricardians, post Keynesians, Austrians, or complexity theorists. We don't need to recreate the intellectual monoculture that failed to explain the crisis (even ex post!). Students should see alternatives, and be challenged to think for themselves.

  4. "I’ve encountered a much larger range of economics blogs in the last week or two (you can guess why), and it does feel like going back in time."

    I think the important thing to recognize is that the blogs (not mine, of course) do not reflect what macroeconomists actually do. In modern macro - published research, work presented at conferences and seminars - there are no schools of thought, in the methodological sense. There are however, many competing ideas floating around about how the world works, but it is hard to compartmentalize those ideas into distinct schools of thought. The financial crisis has brought back some rabble-rousers who exist on the fringe of the profession, but that is not an academic movement. Far from it.

    Steve Williamson

    1. The biggest 'rabble rouser' of all beingthe crisis that your models are unable to explain.

  5. Jamie Galbraith wrote a paper some time ago in response to Krugman's admission that some heterodox economists got the developing crisis right (prediction and explanation) while the mainstream economists missed it, cannot explain it, and have not produced policy that will fix the damage. A number or heterodox economists predicted it, explained it, and offered policy prescriptions based on the macro theory.

    The title of Galbraith paper was "Who Are These Economists, Anyway?" He mentions five: 1. Marxians as habitual Cassandras, 2. Dean Baker (Post Keynesian), 3. Wynne Godley (Post Keynesian), 4. Minskyians (who are also Post Keynesians), and 5. Galbraithians-Instititutionalists (JKG cites William K. Black of UMKC, Post Keynesian-MMT). He could also have mentioned Steve Keen as a Post Keynesian that emphases Minsky, as well as L. Randall Wray and Warren Mosler of Modern Monetary Theory (MMT), which is an outgrowth of Post Keynesianism, especially Lerner, Minksy, and Godley.

    Other than the Marxists, these economists are quite close to each other on most matters and significantly in agree over their areas of disagreement with New Classicalism and New Keynesianism. There is quite a large body of literature on this that goes back decades and has exploded recently.

    1. I just read this paper; thanks for the pointer! It appears to point to a basic missing ingredient in DSGE -- the existence of a financial sector capable of generating massive fraud via regulatory arbitrage.

    2. You might also enjoy reading Wyyne Godley, "Why Gordon's Golden Rule is now history" August 2005, where he predicts the impending crisis.

  6. Funny how Steve Williamson dismisses as "rabble rousers" people who did much better than his establishment gang and forecasting and analyzing what has happened in the last few years. Not surprising at all how this crowd grips ever more tightly to control of Official Macro, including denying the existence of schools of thought as something that concerns only fringey "rabble rousers." What a pathetic joke.

  7. I really hate it when commenters mention some sort of 'official macro' as if economic theory were handed down from god on stone tablets. I hate it even more when people say they didn't predict the crisis.

    I mean for pete's sake, it's quite hard to predit a credit crisis if your model of the economy doesn't have a credit market. Stop criticising models that weren't designed for situations like this. It's like criticising newton's laws of motion by saying that they didn't predict that ships would capsize in rough seas.

    And also, this 'crisis'. It wasn't an event. It was lots of events, inter alia: US housing decline june 2007, money market freeze august 2007, northern rock bankruptcy september 2007, bear stearns march 2008, lehman brothers sept 2008, collapse in international trade jan 2009, eurocrisis may 2010 and so on and so forth.

    1. Your views seem to reflect two standard biases that mainstream economists have when presented with the crisis as evidence against their theory:

      (a) The crisis was some sort of unpredictable exogenous shock, unrelated to the previous boom.

      (b)It's impossible to predict these things.

      Of course, Steve Keen's models predicted the crisis pretty well - we're not talking dates, we're talking about the unsustainable build up of private debt that your models ignore. As you say:

      'I mean for pete's sake, it's quite hard to predit a credit crisis if your model of the economy doesn't have a credit market.'

      Except you seem to be using this as a defence?

  8. Newton's laws of motion do predict that ships would capsize in rough seas.