Winner of the New Statesman SPERI Prize in Political Economy 2016

Friday 30 August 2013

The New Keynesian model in grad school macro teaching

Despite appearances to the contrary, I do not think Krugman and DeLong actually disagree with me on the dominance or otherwise of the New Keynesian business cycle model in academia. I wrote:

“Yet in many who were part of the New Classical revolution, or who were taught by its leaders, there remains a deep antagonism to Keynesian ideas…. As a result, in certain places NK theory was tolerated rather than embraced, or was quietly marginalised.”

That was a key point of the post, although I admit I did take rather a long time getting there.

So I do not think there is any qualitative disagreement. The New Keynesian model is the dominant model of business cycles in central banks. In academia it is not dominant. Is it still the framework of first choice for the majority, as I suspect, or only about half of academics, as DeLong suggests? Impressions gained from non-random contacts or gatherings are not good evidence, so I was relying more on this survey, but that is just one particular sample. And it would be interesting to know results for other countries besides the US - particularly those that have chosen to embrace austerity or advocate it for others.

I did find some evidence from a survey of grad students undertaken by David Colander in 2005. This only covered the major US schools, but from Table 5 you will find that 22% of students from Chicago believed price rigidities were unimportant, while the equivalent number for Harvard was zero. 13% of Chicago students and 20% from Stanford disagreed that ‘fiscal policy can be an effective stabilizer’ (Table 6). But perhaps more revealing are some of the observations Colander makes in the text. Here is one.

“The students told me that the differences in policy views on macro that showed up in the survey did not reflect what they were taught about policy in macro, since they were taught almost nothing about macro policy, but reflected their undergraduate training. When asked about survey results showing that students at his school had changed their views on policy, one student stated, “I think that in the macro course we never talked about monetary or fiscal policy, although it might have been slipped in as a variable in one particular model, but that wasn’t the focus, so it didn’t come from the courses.” Another stated, “Monetary and fiscal policy are not abstract enough to be a question that would be answered in a macro course.”

I guess it might be possible to teach the NK model as if you were not talking about policy, but it would be an odd thing to do. 

For me this reaffirms Paul Krugman comment that, “It would be interesting to know how many graduate departments were in fact teaching New Keynesian macro in 2008”. Even if records on this are not immediately available, memories should still be reasonably fresh. And knowing what is taught in 2013 should at least give us a lower bound. A good indicator of academics’ views is what they teach or what they were taught. I don’t think this information has been collected by anyone, and I think it would be a valuable thing to do if anyone has the time. [1] What I am pretty sure about is that, in twenty years time, some people will write about this, and it would be nice for them to have some more concrete evidence: the time to collect it is now.

[1] I think the key number here should be the proportion of a core masters macro course that is devoted to New Keynesian analysis (i.e. any DSGE model with sticky prices, and policy analysis using that model). Obviously zero would be a very interesting finding, but one week out of twenty is also somewhat suggestive. Measurement is tricky of course, but my quick estimate of this number for the current Oxford MPhil, which would also apply to 2008, is around 20-30%.
Information on current teaching should either be available from the web, or available on request from the departments. However it would be interesting to try and collect information from recent students as a cross check. As one academic once told me: “I intended to cover NK economics but I just ran out of time”.    


  1. You need to whip up a questionnaire and have the blogosphere economists have a look if they think questions are right, and then someone needs to send it out.

    Is there an socio-economics department anywhere that would revel in doing this?

  2. This would be a very interesting survey indeed. I've almost completed my MSc Economics at QMUL and out of 22 lecutres in macro, only 1 covered the Fischer and Taylor models. Even the Calvo model was ignored. I think that also prespective students would benefit from such a survey in order to have more information when deciding which department of economics join.

  3. At Humboldt University Berlin, we have 2 undergraduate courses, which teach the NK model. Both courses are mainly based on Gali's work. Nominal rigidities were modeled with Calvo and Rotemberg.

  4. As an undergraduate, I had 3 courses on macro, which I where taught pretty much the whole Mankiw's Macroeconomics and Bernarnke at al. Macroeconomics textbooks, plus some parts of Romer's advanced macroeconomics textbook. (thus, some old schools IS-LM, basic growth models, plus New Keynesian Theory, mostly)

    I feel kinda lucky since we also had the chance to undertake 2 courses on political economy (which covered smith, ricardo, marx, mill, etc) and some modern aspects of marxian political economy.
    The most interesting course was one which called "history of economic theories". It was all there, from Quesnay to Friedman and Modigliani to Srafa to Leijonhuvfud. It was based on previous courses and even though it was kinda brief and simplistic in terms of maths (in order to cover all competing schools of thought in 1 semester) we had the chance to look the period of each theory, main assumptions and models, proposed policies, etc.
    Since there are departments which teach only NK or only New Keynesian, I consider my self a lucky one.

  5. Funny how when I was an undergrad at QMUL I felt that there was more of a focus on NK-type models in the final years.

    On the M.Sc. at Warwick, for the half of the core macroeconomics module that focused on the short-run, we only looked at the New Keynesian model. We just ploughed straight into Calvo pricing, NKPC and then the monetary transmission mechanism after an introduction to the NK framework. In the optional Monetary Economics module we followed Gali's monograph very closely. That's pretty much all the macro on the Warwick M.Sc. and I barely heard a mention of the words 'New Classical' or 'Real Business Cycle.'

    1. Thanks - the info on Warwick is interesting. I'm mainly interested in graduate rather than undergraduate, as I think the textbooks ensure good coverage of Keynesian ideas in undergrad courses. And its quite possible that undergrad and grad courses in the same department are very different - its often just who happens to teach the grad course and their interests/views/background.

  6. small real story .(.true)
    i know many of some people in my age up-down (east countries)
    they have pay (money ) not exact how AND take one degree (diplome ) by some university (professional ) and fun is they just have finish 12 k .unbelieveble but true .my Q is do real play important role one decoment for work ?or the people who we have to get in our work's (specialy in finnace economic) needed to pass by chack up .?
    and yes needed course specyfice for work NK wanted to make difference btw to day and tomorrow ..this is one step (theory is just to show the road
    thank you

  7. Imho the 'field' is missing the point.
    For teaching things you basically need one model at least to start with.
    For use in a practical situation simply take what gives most likely the best result in THAT particular situation. Is it model A take model A, if it is model B take model B, is it dropping your pants jump on the table and throw darts while holding a beer in the other hand, do that.
    At the end of the day it is about human behaviour. With models daeling with areas in which nearly each time other players determine the outcome. Human behaviour in a particular situation determines the outcome not a model. And human behaviour can change over time and depend on not modelled circumstances.

    Hard to see eg that anybody who has been around in real life can deny pricestickiness. Just look at the real estate market(and a few others) in several countries at the moment.

  8. A last point.
    More interesting than which theory seems to me where actions stop to work or better stop to work effectively.
    A lot of the actions are based on certain things being (stable) are moving in a straight line. While according to other theories or other fields they donot.
    Like interest at present a lot of maco folks seem to assume that CBs can keep these more or less stable. However Finance says otherwise.
    Imho both are right only on different parts of the curve. And of course in modern societies regulation (including taxation) and things like CB policies can have an influence. But in the part we are now at or close to at the end of the day one of the theories will have to give (or sometimes both).

    The most interesting discussion is at the moment imho when do these models start to work less efficient and when are the costs of measures higher than the revenue?

  9. To ask what may seem a silly question - what else would they teach at the grad level, if they are going to teach macro? I used to work in finance as a fixed income analyst, and I only really had contact with central bank research, and NK-style models were the only thing I came across. (Not counting econometric work, which I lump more in with statistics than macro.)

    It was my understanding that the DSGE-style modelling had largely replaced other approaches.


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.