Saturday, July 10, 2010

Roger Garrison Reading Update

So I'm a good deal through Garrison's Time and Money and I plan on getting a good deal more through it this weekend, if not finish it. I have a few thoughts.

First, he has an interesting rendition of Keynes. A lot of the stuff he's spot-on with. He's not under the false "Keynes is a big-government guy"/"circular flow Keynesianism" kind of simplified understanding. He brings out the nuance and it's obvious Garrison has a great deal of respect for Keynes. But sometimes he's misleading or too dismissive. He says, for example, that Keynes believed any reduction in consumption demand would result in lower consumption in subsequent periods. I don't recall that position. He uses a Keynesian cross analysis that starts with the labor market and thereby claims that Keynes is a "labor-based macroeconomic theory" (which Garrison contrasts with the Austrain "capital-based macroeconomic theory"). I thought this was especailly odd. But then, I take the Hicks position and consider liquidity preference to be the heart of Keynesianism - so Keynesianism for me is driven by the interaction of the market for money and the market for loan able funds, making it a capital-based theory like Austrianism (just in a different way - Austrianism is driven by the capital structure, Keynesianism is driven by liquidity preference).

Garrison declares that Hicks misinterpreted Keynes without explaining why he thinks this and completely forgoes an IS-LM approach, or anything like it, or any discussion of liquidity preference at all. The Keynesian cross is a fine way of expressing some Keynesian ideas, but by focusing exclusively on it, I think Garrison presents a very lop-sided picture of Keynes. He's still discussing Keynes now, so maybe liquidity preference will come in but it doesn't seem like it (and flipping ahead I don't see where it's going to come in). Hicks's reconsideration of the issues in 1976 is widely cited by Austrians as being somehow meaningful as a verdict on how most Keynesianism has been done after Keynes, but I think that's a little naive. What about Modigliani? What about Hansen? What about Klein? Hicks wasn't the only one to see Keynes in that light, so Hicks's reconsideration isn't the death knell for this way of thinking about Keynes (and my understanding is that even in 1976 he didn't really abandon the money market/loanable funds market distinction that is really the main point of IS-LM).

Anyway - that's Keynes.

I like a lot of what Garrison has to say on ABCT too. His analysis is mostly clear, although the fact that it is entirely geometric makes it a little stylized and informal sometimes. I especially like his discussion of deficit finance, regulation, risk, etc. in the context of his three-panel model. Often, Austrians use Hayek's views on the capital structure are used to discuss monetary policy, and then use a vague libertarian political philosophy and microeconomic insights to discuss fiscal and regulatory policy. Garrison does it all with Hayek's capital structure model, which is a very interesting exercise. I think his discussion of deficit finance left a little to be desired (many of his conclusions were entirely dependent on what amounted to his assumptions about the multiplier - rather than on anything having to do with the capital structure), but it was overall a good, thoughtful analysis.

One thing I want to communicate clearly is that nothing I've read so far has convinced me that an Austrian-Keynesian synthesis of some sort isn't viable. In fact, for me it's reinforced the viability of such a synthesis. Both schools of thought are very important for their innovations, and I think both are also wasted by many of their adherents. Many Austrians use Austrian economics as a nothing more than a theoretical crutch for libertarianism. Likewise, many Keynesians use Keynesian economics as nothing more than a theoretical crutch for left-liberalism. For me, this is a waste of the economic insights of both schools. This is how I see it - a lot of thinking on macroeconomics has a representative agent flavor to it. People assume that what happens at the level of individual interactions and markets can be writ large to say something about the course of the economy as a whole. Austrianism and Keynesianism both depart from this naive macroeconomics, but in different ways:

1. Austrianism notes that microeconomic interactions can propogate distortions that destabilize macroeconomic realtionships. So that smooth microeconomic functioning can result in a choppy macroeconomy.

2. Keynesianism notes that microeconomic actions are not necessarily cumulative in the sense that optimized microeconomic behavior may lead to sub-optimized macroeconomic behavior. In other words, you can't assume that because everything is good with an agent, everything is good for the macroeconomy.

Both insights are good and they can be taken together. Then it becomes an empirical question of in which cases (1.) or (2.) are more important (my sense is (1.) is important, but probably not all the time). I don't think this sort of thing is accomplished by Garrison. He presents Keynes in Austrian terms to compare them, and in the process loses the whole point of Keynesianism.

1 comment:

  1. My recent post on ET.net might shed some light on why no Austro-Keynesian synthesis has taken place (not an insignificant reason would be the disparate policy prescriptions). Also, I think that Mises and the Austrian school take into account all of Keynes' contributions - albeit, not in the emphasis you guys do.

    The only item we seemed to come to agree on on my earlier post http://www.economicthought.net/2010/06/unequivocal-keynesian-contributions/ was that Keynes advanced his theory of liquidity preference, which gives host to liquidity trap and so forth. From my understanding of Human Action and Money and Credit, Mises does talk a decent amount about the demand to hold cash and the implications thereof.

    Maybe the other contributions would simply be irreconcilable given the methodology.

    ReplyDelete

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