Wednesday, June 18, 2014

Reviewing Noah's Review Of Big Ideas In Macroeconomics

Noah Smith has put up in five installments and now as a whole a review of _Big Ideas in Macroeconomics_  by Kartik Athreya.  He had previously blogged on the book without reading it, which led to criticism, arguing that mainstream macro people are attracted to the field because they like the sorts of models and techniques currently used in "modern macro," aka mostly DSGE models (dynamic stochastic general equilibrium).  Now he has read the book and gives his extensive review.

I have not read the book (just as I have not read Piketty yet, so no comments from me on it, yet), so I am not going to comment on it directly.  But, I will comment on his review, which after making a number of critical remarks, most of which look like they might be reasonable, he in the end comes down defending mainstream macro basically on the grounds that Margaret Thatcher used to put out for capitalism: There Is No Alternative ("TINA").  He declares that after spending time around heterodox economists he finds "much less evil" in the mainstream models.  He lists Austrians, Post Keynesians, and MMT people, and charges all with being "too political" and makes somewhat more disparaging specific remarks about each, although he is less harsh on the Post Keynesians (although the MMT approach is generally viewed as a sub-category of Post Keynesianism).  He says nothing in his review about Agent-Based Models (ABM), and I suspect this reflects Athreya saying nothing, or little, about them.  In comments he expresses sympathy for ABM, but says it has not yet delivered.  He recognizes that the mainstream macro has a tendency to be a closed shop ignoring outsiders, but in the end says, "if it's going to be reformed, it's going to be reformed from the inside," noting the rush by the insiders to develop DSGE models with all kinds of financial market formulations, and with Athreya admitting that the mainstream did not do  well  modeling the crisis, with weaknesses in modeling "household finances" and "interfirm financial contracts" as problems.

I will make only three further comments, beyond noting that Noah finds the book not to be very well written and probably hard to understand for anybody who is not already well versed in mainstream macro, even though apparently Athreya has sought a wider audience for his defense of mainstream macro. 

1)  In a discussion of what makes a theory scientific, apparently Athreya argues that it must be mathematical.  Noah does not really disagree with this, although he notes that ideas or theories may be initiated without explicit math.  Where I have a problem is with Noah earlier in his review where he identifies groups that might have trouble with reading Athreya because they are insufficiently mathematical.  He lists these as those who learned their macro from "literary" texts, particularly by Keynes, Hayek, and Minsky.  Well, excuse me.  OK, there were some usual types basically making him look right about this, going on about one or another of these and attacking math modeling in general in econ.  However, I think it should be made clear that for all of these, although perhaps more for some than others, there are people working strongly in one or another of those traditions, let me name Minsky in particular, who are using mathematical models.  Indeed, I would note that some of the tougher criticisms of DSGE come from those who say its math is bad (apparently Athreya worries about chaos theory some, and fights back against the importance of the Sonnenschein-Mantel-Debreu arguments).

2) Following up on that last point, a core part of the review, and I guess the book as well, is how Arrow-Debreu-MacKenzie (ADM, I give Athreya credit for giving MacKenzie credit) fits in to  all this.  According to Noah, Athreya argues that it is the centerpiece of the microfoundations of DSGE macro (the GE part indeed).  The world is assumed to be in general equilibrium, and that is that.  A first defense provided, although I am a bit uncertain if this is Noah himself or Athreya, is to point to experiments in which subjects are able to get to equilibria quickly in certain settings.  The most famous such experiments were done initially by Vernon Smith over a half century ago, and involve double auctions settings.  I note that not all markets have such settings in the real world, and I also note that experiments on multiple market setups have been rare and very limited in their scope.  This is a thin reed to jump to concluding that general equilibrium not only holds, but the stronger point that it is unique and stable, which apparently Athreya spends a lot of time defending, although on this matter Noah is more critical, saying that DSGEers should face up to the possibility of multiple equilibria more readily (and I agree).  Curiously, apparently Athreya refers to GE as an "emergent" idea, which is definitely a complexity argument that some ABMers make.

3)  They use assumptions of agent rationality to rule out speculative bubbles, and bubbles are there in the real world and create serious problems for the existence, or at least uniqueness and stability of general equilibria.  Indeed, starting with work by Vernon Smith but since widely supported, experiments find bubbles to be ubiquitous in many market settings, including ones with fairly short time horizons and where agents are fully informed.  Apparently Athreya does spend some time in one chapter sort of admitting that there might be problems with this, that there might be bubbles and heterogeneous expectations (meaning some not strictly rational).  But in the end, this gets swept aside in the glory of general equilibrium, not to mention the supposed "evil" of the alternatives.  The law of one price also gets dragged into this discussion.

There is much more in the review, which is quite interesting, whether or not the book is as interesting as the review may well be quite another matter, and on that, well Noah says it is not very well written or organized or very clear.  Oh well.

Barkley Rosser

1 comment:

kevin quinn said...

I just read Noah. Here's something that stuck in my craw. He agrees with the author that 'mathematics allows us to be precise about our meaning." This is ABSOLUTE RUBBISH, and needs to be exposed as such. Mathematics is patently neither necessary and is certainly far from sufficient for precision in meaning. Would mathematics have made Veblen's analysis of the leisure class more precise? Among other things, the mathematical statement that Y=C+I+G is might be an identity or an equilibrium condition - nothing about the math makes the distinction. And if I had a nickel for every economist who confuses the two, with absurd results, I'd be as rich as Croesus (Mankiw's intermediate text is a treasure trove in this respect). When Noah goes on to to say "Without math, discussions about theory often degenerate into arguments about 'What Keynes really meant', I can't fail to suspect that Noah hasn't read much of what he and Atheyra are pleased to call 'literary economics.' I don't recall much in the General Theory, which is full of precise thinking, about what so-and-so really meant, do you? Ditto Veblen, Mill, Hirschman, Smith, Schelling, Buchanan, Rawls, Sidgwick, Minsky, Hayek, Kaldor, Robertson, etc. etc.