10° 40' N, 61° 30' W

Friday, May 09, 2003

While I was out . . .

[this post has been edited--DCS]

While writing my macroeconomics assignment over the long weekend I missed out on the Jane Galt-started brouhaha on economics as a science. It was really about positivism, as Stephen Karlson correctly notes. Not wanting to restart that (not that I can; I don't get that much traffic) focus on this comment of Galt's:

Economics is the least scientific where it has the most trouble finding testable data, which is to say macroeconomic policy, which is why there's so much truth-shading by politics-minded economists in areas like tax policy; assertions are often arguable, but rarely falsifiable.

Galt's partly right about this. Economic data only goes back for a century and a half in the most advanced countries, and most of the time this is too small a sample on which to make reasonable estimates. Most econometric models of the busines cycle, say, use just over 20 years of data, the definition of which may have changed over time. Forming a model based on this small sample would you to conclude that the future would be exactly like the last 20 years. Not a good basis to work from.

The reason way I say this in partly right has less to do with data and more to do with the nature of macroeconomics itself. Among other things I've been doing while I should be studying, I've been reading "Modern Economics and its Critics, 1", by the Cambridge professor Sir Partha Dasgupta. He has this to say about macro:

Many of my non-economist friends think macroeconomics is economics, and are surprised when I tell them it is not so. Macroeconomics is among the most problematic of fields within economics, because it is very, very hard. Lord only knows why you should expect to be able to get things "right" if you have to describe an entire economy in terms of nine or ten variables.27 But then, Lord only knows why you should expect to say very much of practical moment if you were to work with models comprising thousands of commodities and millions of households. So macroeconomic reasoning is essential.

The "frontier" fields in economics today are mostly in micro--behavioural economics, auction theory, institutional economics, the economics of crime--to name four. It took years of persistence, but have finally began to take notice of the innovative measures proposed by microeconomists, like tradeable emissions permits and spectrum auctions. In contrast, since the New Classical-Monetarist-New Keynesian debate ended in empirical (though not theoretical) failure for all sides, macroeconomics (well, the macroeconomics of stabilisation) has lost a lot of its daring. Macoeconomists hav huge difficulties even explaining the world as it is, and while econometrics has come far, it still is no subsitute for a real economy. Since the early 1980s few policymakers are willing, like Margaret Thatcher in Britain or Roger Douglas in New Zealand, to risk pain or financial-market repudiation by instituting even a moderately unconventional macroeconomic policy; in macroeconomics, as in economics as a whole, a live "experiment" is the mostly the only one there is.

While Robert J. Samuelson writes that "unfamiliar problems may require unfamiliar responses" and Edward Hugh looks for an "unconventional toolbox" to combat deflation, there is little sign of it today, especially in the places that really need it. "Truth-shading", as Galt says, has some to do with this. More likely, though, is a simple status quo bias among macroeconomic policymakers--in uncertain times (which is, to say, most of the time) they, and the people that follow what they do, stick with the devil they know.