Tuesday, April 5, 2011

Ryan Murphy on my 1920-21 Paper

Ryan Murphy writes:

"Daniel Kuehn, frequent commentator on the blog Coordination Problem, argues that the 1921 depression is not a counterexample to aggregate demand explanations of the Great Depression. I have sometimes used the 1921 argument since I came across it, but I’m not particularly invested in it. It wouldn’t surprise me, given the caliber of the historians Kuehn opposes, that the argument is wrong. However, I have many problems with his conclusions. Broadly speaking, there are two ways to interpret Keynes. One is the textbook version which we see in schools, known academically as the neoclassical synthesis. The other is the interpretation of the Post-Keynesians. From the purely historical standpoint, I think the Post-Keynesians have a better claim to be truly representative of what Keynes meant. Kuehn emphasizes the role of interest rates in Keynes’s thought, but I think that is both too far but not far enough. It is not far enough in the sense that, when you get down to it, Keynes of The General Theory had only secondary interest in using interest rate policy to get the economy to full employment. What he really wanted to do, since animal spirits would subvert interest rate manipulations, was to nationalize the level (thought not the allocation) of investment. Where Kuehn goes too far is that he misses the entire point of the neoclassical synthesis. The IS-LM model that we see today more or less shows that the effects Keynes was talking about are SPECIAL cases, not general ones, which require assumptions like sticky wages or the liquidity trap for the policies Keynes endorsed (i.e. deficit spending) to be meaningful. Kuehn does briefly address whether Keynes believed de-stickying wages could fix things, but it’s neither relevant for modern policy debates (MICROFOUNDATIONS!!) nor the Post-Keynesian interpretation of Keynes. As such, he flies past both the Keynesian analysis of both old Keynesians like Samuelson and the Lucas Critique which really gave teeth to those assumptions. I’m just repeating the mainstream interpretation of history of thought perspective here, and Kuehn doesn’t adequately address it, despite his assertions and citations of historians who agree with him. Perhaps Kuehn is completely correct in his historical analysis regarding the crash of 1921. That is not my area of expertise, although neither is history of thought. But his supposedly conciliatory efforts between Keynesians and Austrians (betrayed by his refusal to give any substantial criticism of Keynesians) hinge on his interpretation of modern Keynesianism, which I have reason to suspect. I saw this link from Robert Murphy (no relation), who was a referee on the paper. Why Robert, did you allow this to go through?"

A few thoughts: First, I reject his juxtaposition of two ways to interpret Keynes: textbook of Post-Keynesian. I see Keynes as a Hicksian where the interest rate clears the money market but not necessarily the loanable funds market. For that reason, I don't have much problem talking about Keynes with respect to "textbook Keynesianism". Of course it's different, but not in a way that really talking about them all as "Keynesianism". I also don't see how the fact that the Post-Keynesians have latched on to distribution and uncertainty questions gives them any primacy in interepreting Keynes. Uncertainty was indeed central - income distribution was not. But uncertainty was central in its role in liquidity preference, and that is something that "textbook Keynesianism" covers, so I see no real reason to privelege Post-Keynesians here. Anyway -the point being that I think all this waxing poetic about "Keynesian economics vs. the economics of Keynes", and whether Post-Keynesians carry the true mantle of Keynes is all pointless politicking with no real bearing for understanding the economics.

Murphy also thinks that interest rate policy was secondary to Keynes. If by this he means "monetary policy", I would say "not secondary, but not exclusive either". But even with fiscal policy and the socialization of investment - the whole point was to prevent investment from being hamstrung by artificially high interest rates. Interest rates are central whether we're talking about a monetary or a fiscal response. Murphy also gets the neoclassical synthesis and IS-LM exactly backwards. The special case is full employment. There's no reason to think that that particular point is hit on. Now, just because we're not at full employment doesn't mean we're in a depression, but there's a reason why people say "full employment by accident", not "underemployment by accident". The liquidity trap is a special case, it is true, but the neo-classical synthesis does not say that a Keynesian downturn is a special case. In the paper, I noted that monetary policy was sufficient and perfectly "Keynesian" as long as we're outside a liquidity trap. I still think this. But just because monetary policy is sufficient (and probably wise to stick with), that doesn't mean that fiscal policy is "meaningless" outside a liquidity trap, as Murphy suggests. Fiscal policy works as well inside as it does outside a liquidity trap. The thing is, in a liquidity trap monetary policy becomes less effective (not completely ineffective, but less effective), which is why we associate fiscal policy with liquidity traps.

He says I "fly past the Lucas Critique". It's true I don't mention it, but I can't think of how I'm being inconsistent with it. It seems to me I'm affirming the Lucas Critique in some ways by noting that Powell, Woods, and Murphy should not cut-and-paste lessons from historical cases where the policy regime and economic conditions were completely different. He says I don't adequately address the span of the history of thought. OK. What do you expect me to do? I had to cut stuff relevant to 1920-21 for length... I'm not going to wade into the squabbles of the 50s, 60s, 70s, 80s, etc.. I think I identified a perspective of Keynes that is fairly consistent throughout the Keynesian tradition, from Keynes himself to the present - perhaps excluding the Post-Keynesians (which doesn't particularly concern me). That doesn't cover all the details and theoretical innovations of intervening generations, but I think it's still a pretty useful exercise. One of my reviewers suggested that I may be describing Keynes accurately, but not modern Keynesians like Krugman. I responded to the reviewer that I think Krugman would approach 1920-21 precisely how I did and how Keynes did: it's not the same sort of recession, it doesn't require the same response, and it's more akin to 1981 than 1929 or 2008. I think recent posts on Krugman's blog indicate that I called that one right and my reviewer was simply wrong. My paper doesn't address every single variant of Keynesianism, it's true. But it holds up remarkably well in describing all Keynesians through the neo-classical synthesis, and all New Keynesians that aren't completely dismissive of "Old Keynesianism". This whole "Keynesian economics vs. the economics of Keynes" narrative is vastly overblown.

This I had to just laugh at: "But his supposedly conciliatory efforts between Keynesians and Austrians (betrayed by his refusal to give any substantial criticism of Keynesians) hinge on his interpretation of modern Keynesianism, which I have reason to suspect." I am conciliatory in two senses: (1.) I think elongation of the capital structure in response to interest rates makes sense and oughta be taken seriously as a macroeconomic process, and (2.) I don't blithely dismiss ABCT on the basis of evidence that doesn't actually dismiss it (and I expect Powell, Woods, and Murphy to extend the same courtesy for Keynesianism). Murphy will have to live with the fact that I am, in fact, a Keynesian. If that's not conciliatory, then I guess I'm not conciliatory after all.

I know I have a lot of criticisms of this review, but I do want to extend my appreciation to Ryan Murphy for taking the time to look over my paper, think about it, and write about it.


All anonymous comments will be deleted. Consistent pseudonyms are fine.