Dumb it up, Tyler!

Tyler Cowen’s got more of his Macroeconomics series up. It’s nothing like as bad as the monetary economics post that I objected to yesterday. Part Three on fiscal policy is OK ..ish. I don’t agree with him on Keynes, and think his comments on deficits and interest rates are naïve (I include by citation Brad Delong posts on this subject passim ad nauseam), but I can see how others would class my disagreements with it as probably political rather than technical. And Four on open economy macroeconomics is actually quite good, although the omission of any discussion of optimal currency areas is a bit of a lacuna. Part 2 has one very serious error, but in being bad, it is actually good, because it’s clued me into what went wrong in the train wreck which was Part One.

First, the really, really serious mistake (I think that he’s misunderstood Paul Krugman’s views on Japan in this addendum, but it’s probably arguable either way). In part 2, however, he makes the following argument in relation to the predictability of business cycles:

“[…] Most modern business cycles are simply bad luck. You can spend your whole life trying to divine the relevant patterns, but you are very likely to fail, no matter how smart you are.

Many macroeconomists argue that the “time series” of most variables is statistically indistinguishable from a random walk […]”

He then links to this working paper, which to me shows that he has massively misunderstood the literature on this subject.

The point is that when econometricians talk about a “random walk” in GDP, they do not mean anything like the same thing as financial economists do when they talk about stock prices being a “random walk”. The random walk debate in econometrics is the debate over whether GDP and similar series have a particular statistical property that shocks have a permanent effect (rather than dissipating over time as the series goes back toward a long term trend). It’s also known as the “unit root” debate, because it can be framed as a question about whether the equations which make up a model of GDP have roots which lie outside the unit circle (or something; it’s ‘king ages since I did this and memory is hazy). In any case, GDP being a “random walk” in this sense is not at all the same thing as being unpredictable or the same thing as saying that the business cycle is impossible to model; it’s just a matter of whether time series of interest have the Markov property or not. It’s a loose usage of “random walk” to refer to something which isn’t random, which is why some people prefer to say “unit root”.

Anyway, enough of that. At the end of the comment, Tyler recommends his own book “Risk and Business Cycles”. I dug up a few reviews of it (this one in the Quarterly Journal of Austrian Economics is quite good), and it seems clear to me what the problem is. Basically, Tyler’s got a view of the macroeconomy not too dissimilar from my own.

Both Tyler and myself are quite a long way outside the mainstream of neoclassical economics. He’s basically an Austrian, I’m a Post-Keynesian. And in fact, Tyler’s particular brand of “New” Austrianism is very close to Post-Keynesianism indeed. Specifically, he rejects the key Austrian premis that recessions and malinvestments are always caused by gaps between the “natural” and”money” rates of interest opening up (as a result of Big Bad Government, natch), encouraging investors to make mistakes about the time-preferences of consumers and invest in production technologies with the wrong returns period. Tyler takes from rational expectations macroeconomics the idea that it doesn’t make sense to assume that policy-makers can systematically fool the rest of the economy, and from modern portfolio theory and financial economics, the idea that one of the real determinants of investment is the equity risk premium (a concept I discussed here). It’s a “Risk-based Business Cycle” theory in which the business cycle is driven in an Austrian manner by cycles of malinvestment and liquidation, but these cycles do not have a monetary origin. To cut a long story short, his model of the business cycle is one which is more or less entirely driven by animal spirits on the part of entrepreneurs. That’s why he thinks that all these monetary factors are irrelevant.

It all makes sense now. Or at least it doesn’t but it fits into place a lot better. The problem is that some people are good at translating their ideas for the layman (like Paul Krugman) and some aren’t. Tyler’s made what I consider to be a big mistake; he’s decided that he wants to put over his view of the macroeconomy, but he doesn’t want to get bogged down in thousand word explanations of the minutiae of why he doesn’t believe in monetary theories of the business cycle (contrast my own approach to similar questions in the posts I’ve linked in this discussion; I love getting bogged down in these discussions), so he ends up trying to have a fast way with mainstream theory, and in my opinion oversimplifies mightily.

So it’s basically the fault of Volokh for using software (unlike our own Movable Type) which doesn’t allow extended posts. A long Tyler Cowen post on monetary economics might be really good, but it would make the rest of the Volokh conspiracy more or less impossible to read. And the dumbed-down short one … ends up being pretty bad. So I apologise for any negative impression I might have given about Tyler Cowen as an economist, while standing by substantially all of my comments, including the harsh ones, about the actual piece from yesterday. So the Volokh heavy mob can stop sending Crooked Timber those death threats now, please.