‘Shorthand’ for Steve Keen’s contribution to R4 econ program=’made up’?

I had an interesting exchange with Steve Keen last night on Twitter, about things he said on Aditya Chakrabortty’s program on the state of economics and economics teaching.

On the program, Steve Keen said a number of things I contest are false.  He said that mainstream economics ignores banks;  ignores money, or when it doesn’t, simply treats money as a veil.  He said that you can’t publish in top journals without the assumption of rational expectations.  All of these things are false, as a quick Google with some names I suggested in an earlier post will reveal.

The exchange got to the heart of why Steve said these things.

He had two somewhat different answers.

One was that his contribution was ‘edited down’.  Implying that there was a fuller, qualified set of statements not all of which were broadcast, but whose totality could be said to be fair comments.

A second, however, was that his words were ‘shorthand’.  Specifically, when he said ‘you can’t publish papers in top journals without rational expectations’ he didn’t mean that.  He meant ‘you can’t publish papers in top journals without the whole neoclassical edifice.’ This response was to cover my comment that top journals are full of papers that don’t have rational expectations [by Sargent, Marcet, Nicolini, Ellison, Williams, Evans, Honkapohja, McGough, Mitra, Bullard, Brock, Hommes….].

However, these journals are also brimming with empirical finance papers trashing modern finance theory;  of empirical macro papers trashing RBC and New Keynesian theory.  And they are full of behavioural economics and behavioural finance theory papers.  Are those papers that are counted as having ‘the whole neoclassical edifice’?  If we get rid of rational choice, do we still have ‘the whole neoclassical edifice?’  If we dump entirely the project of erecting a theory and have an econometrician demolish one, do we still have ‘the whole neoclassical edifice?’

And when he said ‘mainstream economics ignores money’ [which would be somewhat mysterious for Messrs Kiyotaki, Wallace, Wright, Lagos, Moore, Karecken, Williamson to grasp] he meant ‘mainstream economics has models of money that I think are incorrect’.  [I guess this because he cited my own blog posts questioning whether we yet had a proper model of money].  Presumably the same goes for the claim that ‘mainstream economics ignores banks’ [which Bernanke, Gertler, Gilchrist, Brunnermeir, Carlstrom, Fuerst, Diamond, Dybvig, Keister, Gale, Allen…. would also find peculiar].

Steve calls this ‘shorthand’.  I don’t think this is respectable intellectual discourse on his part.  In my own ‘shorthand’, I’d say that Steve’s characterisation of mainstream economics is ‘made up’ to wage war.  But it’s a desperate tactic.  There are lots of good points to debate about the state of economics and economics teaching, [many made by Karl Whelan, and by Diane Coyle and Andy Haldane, for example, in January’s Prospect] but, by association, Steve weakens the movement he hopes to lead.

 

 

 

Advertisements
This entry was posted in Uncategorized. Bookmark the permalink.

23 Responses to ‘Shorthand’ for Steve Keen’s contribution to R4 econ program=’made up’?

  1. Ian Wright says:

    Steve Keen is trying to define and criticize the hardcore (in Lakatosian sense) research programme of neoclassical economics. You seem to be getting twitchy because he used the narrower phrase “rational expectations” in an edited radio programme. If you look at his published (and very popular) work you get his intent pretty quickly. So it’s somewhat histrionic to then accuse Keen of not engaging in “respectable intellectual discourse”, and “making things up” in order to “wage war”. Also, it helps to take Popper’s advice, and really try to get into the mind of your opponent, rather than assuming they have ill intent. When Keen says neoclassical models “ignore money” you should try to understand what he means by that. Clearly, and prima facie, lots of economics papers deal with money. But Keen is not daft. So what does he mean by that?

    • Tony Yates says:

      Thanks for trying with the advice, but it’s wasted on someone as low-brow as myself. I’m just pointing out he said things that would mean things to your average reader that weren’t actually true, and calling him on it. I don’t care one bit what he means or if he has any subtle intent. What I care about is that Joe Public doesn’t think us crackpots for the wrong reasons. We may, obviously, be crackpots, but not for the made-up reasons in Steve’s remarks [and those of Soros, of course].

    • ivansml says:

      “When Keen says neoclassical models “ignore money” you should try to understand what he means by that.”

      Really? Keen has been mocking “neoclassical” economics for ignoring money, banks and debt for years. And not in some sophisticated sense – rather in the “look how stupid neoclassical economists are” sense. It’s hard to see how else than literally could one interpret his remarks.

      Being charitable to ones opponent is a great thing, but it has its limits. With people like Keen, who make a career out of misrepresenting targets of their “debunking”, enough is enough. And of course good faith should be a two-way street. No, I don’t feel the need to be charitable to someone who produces stuff like this, thank you very much. (of course, I’m speaking for myself, not necessarily for our host)

      • Tony Yates says:

        You speak very well for me, thanks.

      • Ian Wright says:

        Your link to the cartoon is very funny. I have no problem with this kind of satire and (entertaining) public engagement. But two things are worth saying.

        First, don’t blame the messenger. Steve Keen is not responsible for the low opinion that Joe Public (and educated scientists outside the field of economics) have of economics. The actual content of the majority of economic research, coupled with recent economic events, is responsible for that. For instance, I entered economics from a background is in computer science and mathematics, and I was appalled by the kinds of assumptions and modelling approaches that (in general) are accepted as OK within economics. E.g., analysis of equilibrium states and trajectories without proper analysis of adjustment mechanisms and their stability properties. Steve Keen is simply saying what a lot of people have been thinking for some time, and therefore there’s a demand for his kind of critique.

        Second, you say “enough is enough” and clearly have lost patience with Keen’s sometimes irreverent tone. But tone is irrelevant to an argument. You need to put the emotions aside and criticise the content of his arguments. Otherwise, you do become uncharitable, and project all kinds of bad intent onto your opponent (e.g., “makes a career out of misrepresenting targets”). Whatever you make of Keen’s tone or conclusions, it must be acknowledged that he has devoted a considerable amount of intellectual energy to seriously engaging with neoclassical economics. Also, he was one of the few economists to correctly predict the great recession, in marked contrast to most neoclassical economists. In other words, he deserves way more respect than you’re paying him here.

      • ivansml says:

        @Ian Wright:

        “First, don’t blame the messenger. Steve Keen is not responsible for the low opinion that Joe Public (and educated scientists outside the field of economics) have of economics. ”

        One reason why Joe Public has low opinion about economics is exactly due to all the silly criticisms by various charlatans. So I will blame the messenger, if that messenger makes stuff up.

        “Whatever you make of Keen’s tone or conclusions, it must be acknowledged that he has devoted a considerable amount of intellectual energy to seriously engaging with neoclassical economics.”

        And this is where you are wrong. Keen’s criticisms of neoclassical economics are superficial, irrelevant, often simply wrong. He doesn’t engage or even cite any of the recent mainstream macroeconomic literature, except for a couple of token examples (Bernanke, Krugman), which he usually misrepresents anyway. Frankly, I don’t think he actually understands neoclassical theory even at an undergraduate level.

        Just one out of many examples, which you’ve touched upon: Keen has repeatedly claimed that neoclassical economists ignore dynamics. Not that their modelling of dynamics is inappropriate or insufficient, but that they literally don’t know how to use differential equations. This is just plainly false, as any graduate student can attest.

        Then in a recent paper, he has finally discovered that we actually do use differential equations, and discusses the Ramsey growth model. He criticizes neoclassicals for not imposing initial condition in the consumption equation and for treating consumption as a “jump” variable, thus departing from “standard mathematical dynamics”. So apparently in Keen’s world, standard mathematical dynamics involves only initial value problems. Boundary value problems? Optimal control theory? Pontryagin’s maximum principle? Forget about those. Yeah, it’s hard to take such people seriously, especially given the air of mathematical superiority they love to project.

      • Ian Wright says:

        Just one point. Keen is absolutely correct in pointing out the (in general) conspicuous absence of differential and difference equations in economic models. Yes there are exceptions. But this is a stand-out oddity to anyone entering from any other scientific field. In addition, your examples of the use of optimal control theory, Hamiltonians etc. make Keen’s point because these techniques are misused by many economists in order to depart from “standard mathematical dynamics”. E.g., only in the economics literature have I seen saddle-point trajectories described as “equilibrium paths” because an optimal control policy jumps around with infinite speed. The assumptions necessary to stay on such a path are often ridiculous. The computational requirements to compute such a policy often require non-polynomial time. In other words, quite mad stuff. I’ll read Keen’s paper with interest.

      • Luis Enrique says:

        I would find it helpful if Keen didn’t start talking about expectations in the middle of discussing the deterministic Ramsey model. Consumption is treated as a jump variable for perfectly sensible reasons.

        afaik, in the Ramsey model, given k, once you have made the initial choice of c, then you just slavishly follow the Euler equation for c and that takes you to balanced growth path. I do no think it is of much interest to ask what would happen if the initial choice of c is ‘wrong’ but you nonetheless slavishly follow the Euler equation and spiral off into the distance. I mean either look at the combination of optimal initial c and following the optimal rule for c, or look at non-optimal initial c and them some rule for subsequent choices of c that makes sense in that context. I wouldn’t be surprised if there are papers out there that have something to say about that.

        no economists think real people can compute the optimal initial c or slavishly follow equations for subsequent c (real people don’t inhabit a universe made of two variables) nor do they think the Ramsey model is a satisfactory description of how countries with capital stocks below their long-run level take consumption / investment decisions. Some economists think the model is a useful starting point for thinking about these questions, and contain some insights that would survive in a more realistic model, so long as people are comparing what they get out of consuming against what they get out of investing and making a decision in the light of that.

        if you think superficially highly unrealistic models cannot say something interesting that helps us think about reality, then you must throw away all of Keen’s own models too.
        .

      • ivansml says:

        @Ian Wright:

        “Keen is absolutely correct in pointing out the (in general) conspicuous absence of differential and difference equations in economic models.”

        Almost all of recent, DSGE-style macroeconomics runs on systems of difference equations, as did the older Keynesian-style large-scale models, so I don’t know what absence you’re talking about.

        “e.g., only in the economics literature have I seen saddle-point trajectories described as “equilibrium paths” because an optimal control policy jumps around with infinite speed.”

        Call them “mutually consistent best response paths” then. Or “marshmallow paths”, if you prefer. Semantic arguments about the meaning of the word “equilibrium” must be one of most wasteful activities in the econoblogosphere.

        “The computational requirements to compute such a policy often require non-polynomial time.”

        It’s a good thing then that we’re studying people, not computers.

      • Ian Wright says:

        I say most, not all, swans are white. You point out a few black ones …

        My point about non-polynomial time complexity has got nothing to do with computers. My point is, rather, that many economic models assume that something computes an optimal control in order to jump with infinite speed to maintain the “marshmallow” path. But the time complexity of that computation implies that no physical process, whether computer or human etc., can terminate with the solution in reasonable time. So it’s not just that the models are unrealistic — they are impossible.

      • Luis Enrique says:

        for God’s sake, nobody thinks people go about solving inter-temporal optimisation problems. You really are taking the model too literally if you worry that choosing the optimal control instantaneously as perturbations occur is impossible.

        Yes it is interesting to think about how people really take decisions. I think one or two economists might already be working on that. In the meantime, the question you need to be asking is whether models of optimal decisions that are not realistic depictions of how people actually take decisions have anything useful to teach us.

      • Luis Enrique says:

        read this

        http://sims.princeton.edu/yftp/Gerzensee/info.pdf

        notice that economists did not start thinking about this yesterday. Compare and contrast with Keen’s depiction of economics.

      • Ian Wright says:

        Two quick responses:

        If you want to flag the class of impossible models as strictly normative, that’s fine by me.

        The paper you link to considers constraints on information flow. So I agree it makes a step towards considering physically possible decision making, albeit from a very unrealistic starting point. However, it does not address the computational complexity of the optimization problem it defines. So I do not understand why you think it’s relevant to my point.

        More broadly, the whole approach of focussing on “rational” individual decision-making and then aggregating to macroeconomic outcomes is, in my view, a non-starter. But that’s another discussion.

    • Luis Enrique says:

      “If you look at his published (and very popular) work you get his intent pretty quickly.”

      yes. although perhaps not in the way you have in mind.

  2. A P Webster says:

    Apologies in advance for the slighly muddled nature of the below.

    Regardless of the niceties around what specifically is the problem with economics and economics teaching, I don’t think anyone would disagree with the claim that Economics (with a capital E) has a serious PR problem post-crash.

    In my (not-particularly well informed) view, part of this problem is the fact that economists themselves are often guilty of saying one thing and meaning another (call this ‘shorthand’ if you like), particularly when it comes to the more mathematical end of Macroeconomics.

    For example, an outsider listening to two Macro PhD students discussing their model (or sitting in on a third-year Macro lecture) would often be left with the impression that they actually believe that model reflects reality all of the time, when the model is itself ‘shorthand’ for a particular aspect of an economy, under a particular set of conditions. This is fine when the conversation involves people who intuitively grasp the limits of the models, but the habit of talking in those terms often extends to conversations that involve people who look to economists as ‘experts’ (or ‘boffins’, or whatever term is currently in vogue with the sub-editors of newspaper websites).

    Economists need to be more open about the limits of models and assumptions when engaging in discussion with ‘civillians’ (by which I don’t mean politicians or civil servants, but ‘the public’ in general). Aspiring to some level of ‘scientific’ authority is counter-productive.

  3. GrkStav says:

    “Frankly, I don’t think he actually understands neoclassical theory even at an undergraduate level.”

    At least you were frank. If Steve Keen does not actually understand neoclassical theory even at an undergraduate level, I am the Pope.

    I’ve read a large number of preposterously hyperbolic comments in my many years on the internet. The quote one above is definitively in my top 5%. Bravo!

    • Tony Yates says:

      I bow down before you, Pope GrkStav, magnificent in your anonymity. Very brave of you, I must say.
      I make my comment advisedly, since I do understand it, and think about little else. And my undergraduates would know that the statements Steve made on his radio 4 program, and the many laughable remarks he’s made in his public lectures and papers (my particular favourite is the one where he accuses the entire establishment of misunderstanding profit maximisation), were very wide of the mark indeed.
      Remember, that’s all at issue. Not what’s good and bad economics. But whether Steve Keen said false things or not. That’s really easy to answer, and it says more about you than anything that you feel inclined to defend him. Give up a day to do some reading, and you’ll realise that SK was wrong (about what is and is not ignored in ‘neoclassical’ econ [whatever that is]).

  4. root_e says:

    Yates: “4. In the same breath, Keen states baldly that you can’t publish in top journals if you don’t use the assumption of rational expectations. False. I know many dozens of examples. But here’s Nobel laureate Tom Sargent this time using models with learning, a framework he pioneered. [That’s funny, two Nobel laureates flouting the AC stereotype of economics.]”

    Tom Sargent in the article cited by Yates: ” We have examined what our estimates imply for the self-confirming and rational expectations equilibria discussed above and defined formally in appendices C and D. We have emphasized the parts of our story for the dynamics of hyperinflation that require retreating from rational expectations. However, in many cases the retreat is relatively minor. In particular, as shown in Sargent, Williams, and Zha (2006a), the fixed-mSCEs discussed in the previous sections are close to the REE beliefs. In
    most cases, the differences are well within half of a percentage point, with the largest
    differences being slightly more than one percentage point. Similarly, unconditional SCEs and REEs are very close as well”

    Stunning! A full range of models from A to A+delta

    • Tony Yates says:

      Good try, but this paper really proves my point. You get AMAZING differences by making this small departure. The difference between a society that’s immiserated by hyperinflation, and one that isn’t. There could not be a more radical consequence. And, there are dozens of other papers where the departures from RE are even more stark. If you are that interested, have a slow look through the work of myself, Ellison, Brock, Hommes, McGough, Evans, Honkapohja and many others that you will find in their bibliographies. This is why it’s so malevolent that Steve Keen chose to say ‘you can’t publish a paper in top journals without rational expectations’. It’s not true, on any level, literal or otherwise.

      • root_e says:

        Keene’s assertion was that RE is a required article of belief. Sargent offers a minor departure from RE and shows that minor departures from RE have big implications. That’s not a counterexample to Keene’s claim.

        I say that everyone in your department is a white protestant, and you send me some explanation of why Methodists from Angleford are really different from Saxonwold. Baptists. Well, ok.

      • Tony Yates says:

        Yes it is. They aren’t saying ‘we only believe in minor departures from RE’. They offer the tantilsing prospect of all the chaos that economies can experience given that disaster can be wrought with even this “small” departure. But who cares. There are dozens. I could fill days posting links to other papers. I could fill a whole day posting empirical papers looking into whether expecations are rational or not, and, if not, what best explains measures we have of them. [I wrote one of those]. I’m spending two hours lecturing my macro students about these papers. Check out my lecture notes.

  5. BJ says:

    If we are talking about undergraduate economics, then what is your view on a key claim of Keen, that Banks create debt from nothing? Krugman clearly thinks that this is not possible:

    http://krugman.blogs.nytimes.com/2012/03/27/minksy-and-methodology-wonkish/

    Specifically, would you agree that this is a good description from Krugman:

    “If I decide to cut back on my spending and stash the funds in a bank, which lends them out to someone else, this doesn’t have to represent a net increase in demand. Yes, in some (many) cases lending is associated with higher demand, because resources are being transferred to people with a higher propensity to spend”

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s