Sunday, September 30, 2012

Having fun with economics and Lord Skidelsky

I participated in the 11th International Post Keynesian Conference in Kansas City this weekend and presented a paper on the effect of government intervention in a macro model based on a system of ODE first developed by Steve Keen, who presented just before me.

The conference itself was a bit of a sleepy affair, with most of the other talks I attended being more literary criticism (e.g several quotes from Keynes, Minsky, and others, strung together and compared with recent events) than actual modelling.

My overall impression is that if this is all that heterodoxy has to offer as an alternative to mainstream economics, then the profession is in deeper trouble than I thought.

But the highlight was the keynote after dinner speech by none other than Lord Robert Skidelsky (his business card actually says "Lord Skidelsky", no joking!)

He started his speech with a joke saying that when John Kenneth Galbraith wrote an economic report to Lyndon Johnson, the president allegedly said "You know Ken, talking about economics is like peeing in you pants, it feels hot to you, but leaves everyone else in the cold", then proceeded to say that he was reminded of this story by listening to mathematicians earlier in the afternoon, who appeared to be having to much fun with his models (Steve and me!). Then at the Q&A session, his advice to an economics grad student included "don't use math". Finally when I asked him directly what his advice to  mathematician trying to contribute to economics would be, he offered this gem: "well sometimes you just need to tell what hullaballoo is all about".

So apart from all the fun to be had with economics (which I'll continue to do), I think that between the hyper mathematical (but incorrect!) DSGE guys on the mainstream and the math phobics in the fringes, this is going to be an uphill battle.


  1. Though your criticism is a little fair, it misses the point. First of all, there were other models were only there for one day. But as to why it misses the point, you really have to understand the belief system of heterodox economics (ie it's ontological and epistemological statements) to understand why in many cases modeling in a mathematical fashion is not appropriate. Again, that isn't to say that we shouldn't be formally modeling a lot of what we say (though, as always this area still needs a little more focus), but simply that a lot of things cannot be modeled. And you cannot make the statement then, as many would then say, that that is then not economics. The math only serves to illustrate the ideas for the sake of pedagogy/better understanding/putting complex ideas and beliefs into a model, as you well know. For example, you may think that Steve Keen is the best economist because he tries to formally model things, but his math obscured what were some logical errors. We focus on logical consistency first, and maths second. Neoclassical economics, as you know, ignores the logic and just does math.

  2. Hey thanks for commenting on my blog! It's definitely a rare event. Though it would be much nicer of course if I also knew who you are Anonymous, since you seem to know a lot about me and my schedule during the conference ;)

    And fair enough, I did only attend one day and based my comment on what I saw. But I take your point. As for your other observations, I'll respond in a fresh blog entry, just for the fun again :)

  3. Hi Matheus and Anonymous. What I find interesting about this discussion (and Matheus's subsequent post) is that it is a discussion (at least in part) about reaching the masses (or about effecting a much needed paradigm shift in the face of the virulent opposition of the neo-classical mafia). But it really is a case, through no fault of your own, of one academic speaking to another (or to some other technically advanced thinker such as Neil Wilson). Presumably, I (!) am the typical target of these discussions, though: putatively an economist, but crappy at math (it really is amazing I made it through my Masters). I spent ages trying to figure out where I was going wrong: certain post-Keynesian or circuitist arguments (especially Parguez's work) reinforced the constant unease I always had, say, with IS/LM models or with rational expectations theory. But I was unsettled about this because colleagues and friends could just dismiss my mostly intuitive concerns by just saying: "you really have to understand the math". But even when I succeeded in following their math, I didn't have the intellectual capacity to point out the flaws in it. So in following neo-classical math, I merely further "internalized" their basic assumptions. Reading Steve Keen's book revealed that there were actual mathematical problems with, say, DSGE methodology (I have tried to work my way through some of his models, but I'm not good at understanding ordinary differential equations). Admittedly, Keen greatly simplifies his arguments for people like me, but the point is: simple non-technical expositions of the absurd assumptions of neo-classical economics got me interested, but even at my relatively low level, it took formal models to VALIDATE the heterodox claims.

  4. Oddly, Lord S. is an advocate of Keynes' work, but overlooks his mathematical Treatise on Probability. This underpins his economics and is full of examples of why the classical assumptions are limited. Mathematics should be used to expose and challenge assumptions.

    Unfortunately, as Tom says, it has often been used to obsfucate them.