Wednesday, July 21, 2010

My Thoughts on Garrison

So I recently finished Garrison's Time and Money and I wanted to share a few reactions. First, it was a very good book and I recommend it. As far as I can tell, he's faithful to Hayek's vision on the macroeconomics of the capital structure, but I haven't read Hayek on that so I can't say for sure. Regardless, he certainly offers a well articulated vision of his version of the macroeconomics of the capital structure.

A major portion of his project is expressing Keynesianism and Monetarism in a Hayekian framework and comparing the three. Thinking through each in this way is very good I suppose, and I shouldn't criticize the effort (only the execution) - but I think many of my concerns about the execution are directly derived from the requirements of engaging in that sort of effort in the first place. To put Keynes in Hayek's framework, Garrison does away with a lot of what makes Keynes unique. How is this a proper analysis of Keynes? Some thoughts:

- Garrison assumes away a lot of the distinctiveness of Keynes, and then proceeds to blame Keynes for the shortcoming. For example, on page 158 after talking about an increase in uncertainty, Garrison writes "For the economy to avoid falling into the interior of the PPF, the funds released from the investment-goods sector would have to be absorbed in the economy's consumer-goods sector. This reallocation of resources, however, is already implicit in the movement along the unshifted supply of loanable funds: less saving; more consumption". The reason why he's able to make this assumption, of course, is that he completely omits any discussion of the relationship between uncertainty and liquidity preference until the next chapter. He talks about depressed investment demand in one chapter, without liquidity preference, and then he talks about liquidity preference in the next chapter, without depressed investment demand. It's no wonder that in the first chapter on Keynes the economy stays at full employment! It was liquidity preference that Keynes thought would drive the economy out of full employment!

- Separating the process of investor uncertainty from liquidity preference on pages 158-159 allows Garrison to scoff at the Keynesian notion that the interest rate will remain unchanged (i.e. - too high) in response to a declining demand for loanable funds. He critiques this and the depiction of the labor share of income - the "assumed structural fixity" of the two, as he calls it - without realizing (or at least without acknowledging) that it is precisely Hayek's assumed structural fixity (i.e. - no shift in the supply of loanable funds) that Garrison uses to pin Keynes to the PPF and completely nullify all of Keynes's insights. The only critique Garrison has of the stable labor share of income is directly related to this. His critique is that the labor share of income would not be stable because the interest rate changes. But the only reason why the interest rate changes in this chapter is because Garrison completely omits discussion of liquidity preference until the next one! Garrison is right that a baseline level of liquidity preference is a structural concern that can legitimately be treated separately. He's wrong to ignore the fact that additional liquidity preference goes hand in hand with the investor uncertainty that caused the leftward shift in the demand for loanable funds.

- In his chapter going over liquidity preference (the second chapter on Keynes), Garrison explains Keynesian policy recommendations in a way that sets him up, several chapters down the road, to declare that Milton "we're all Keynesians now" Friedman is closer to the Austrians than to Keynes. In Figure 9.2, I think Garrison captures Keynesian liquidity preference well. The fact that he expresses it in a loanable funds market is irksome (Keynes was strenuous about the fact that the interest rate was determined in the market for money and not the loanable funds market), but it works OK. Output is below the PPF for precisely the reasons Keynes highlighted. But for some reason, a couple pages later in figure 9.3 where he's explaining Keynesian policy, the policy shift isn't from below the PPF to the PPF and classical full employment (as Keynes said) - it's from the PPF beyond the PPF to an unsustainable production level. Where did that come from? I have no idea. Keynes, the man who wanted to use policy to re-create classical conditions (the reason why I think he bears a reasonably close resemblance to the German ordoliberals) becomes Keynes, the man who doesn't recognize the existence of a PPF and wants to spend, spend, spend.

- Which brings me to another concern... Garrison's PPF is not the usual PPF. It is not an impassable technological frontier, it is a "sustainable output" frontier. That can get tricky at times - watch out for it.

- Finally, when Garrison discusses Friedman's plucking model he insists that it is inconsistent with Keynes (who he thinks oscillates above and below the PPF, as I mentioned earlier), but that it is consistent with Austrianism. It's actually a very interesting section where Garrison makes the case that malinvestment does not imply overinvestment, and turns the traditional Austrian boom-bust cycle into a bust-boom cycle that is consistent with Friedman's plucking model. It was intriguing - I'll let other Austrians decide how convincing the argument is. Nevertheless - the idea that Keynes is not consistent with the plucking model follows from Garrison's distortions of Keynes that I mentioned above. I think it also (though less necessarily) follows from the assumption that the baseline from which the economy is "plucked" is a full employment baseline. I'm not sure why you would assume from the outset that that is the case. Garrison says "well, Keynes thinks we're always below full employment so we obviously can't be plucked from full employment". I would simply respond "I would have thought that the baseline level of liquidity preference is fairly steady, so what we're being 'plucked' from is a stable-growth, sub-optimal path that is below full employment". He doesn't even consider the prospect (i.e. - he assumes his own conclusions by assuming that what we're being plucked from is a full employment growth path). This was the only truly surprising part of the book. When he introduced the plucking model, I was sure he would say "Keynes is consistent with Friedman here, and they both ignore the Austrian insights which is why Austrianism contradicts this". Nope. He ended up saying "Keynes is inconsistent with Friedman on the plucking model, but Austrians are consistent because our boom-bust model is actually a bust-boom model". That approach genuinely surprised me.

Much of the rest was very good - my biggest concern was the two chapters on Keynes (and even those were quite good overall). I provided other thoughts on Garrison in this post.

I know it's probably hard to get much out of this as a simple blog post, but it's a way for me to get my thoughts written down, and perhaps people can refer back to this if they read Time and Money. Next on my list, I think, is the first of Joseph Dorfman's three volume series on "The Economic Mind in American Civilization". This one covers the period from 1606 to the founding. Joseph Dorfman was Murray Rothbard's advisor. If someone has a better suggestion before I come home from work and start reading that, let me know :)


  1. "It's no wonder that in the first chapter on Keynes the economy stays at full employment, and in the second chapter on Keynes!"

    This sentence is confusing.

  2. Sorry - I wrote up two posts pretty quickly this morning, in this one just jotting down reactions that have been stewing for awhile.

    I'm going to read through the whole thing and edit it without explicitly writing an update - not to dodge criticism, but just for the sake of smooth reading.

    Thanks Natalie.

  3. There - I hope that's better I expanded other points for clarification too.

  4. Malinvestment does not imply overinvestment. You didn't know this? After all the conversations we've had, and after all the assurances you've given that "you know ABCT," I'm genuinely shocked you found this surprising.

    Maybe you should take my advice and sit down with Mises/Hayek directly.

    Some works dealing specifically with ABCT: Theory of Money and Credit; Prices and Production; Money, Bank Credit, and Economic Cycles; and Human Action (specifically the relevant chapters - but, like I always suggest, the whole book).

  5. The concept itself isn't surprising - his emphasis on it was. I'm well aware it's the distortion of the structure that drives ABCT, rather than any broader overinvestment. That is obvious. That overinvestment would be strenuously repudiated as even an outcome was a little surprising.

    But I think your point is a little too strong. Frank Shostak seems to think in terms of over-investment as being interchangable with malinvestment:

    And if Shostak is to be trusted, Rothbard thought in those terms too.

    That was just an early result in a very quick search.

    I personally don't think the idea of over-investment is all that silly. If interest rates are intertemporal coordinators, why couldn't artificially low interest rates pull more economic activity into an earlier time period, leading to a boom or an over-investment in addition to its distributional effects? Could you explain to me why that's so implausible?

    Is is it necessary? No. I understand that. But it sounds like a very plausible part of the story - not one to simply be tossed aside in an effort to conform with Friedman's plucking model.

  6. I didn't say there wasn't overinvestment; I said that malinvestment does not imply it. That means, while there may be overinvestment in addition to malinvestment, malinvestment is not a problem because of overinvestment. This is the position Krugman took in the late 90s when he "addressed" the Austrian theory. He accused us of presenting an overinvestment story.

    Now, in the course of a business cycle, is there overinvestment? Absolutely. But that's not the worrisome part of the cycle. It's overinvestment in areas that ought not to have investment. So it's more complicated than a simple overinvestment story (which you agree with).

    Shostak oversimplifies the theory. The article is focusing on the concept of originary interest and he then brings it in line with intertemporal coordination (which is fine).

    But this claim: "An overinvestment in capital goods results in a boom, while the liquidation of this overinvestment produces a bust. Hence, the boom-bust economic cycle. " is wrong. It is not the overinvestment of capital goods that produces the boom. Rothbard says something similar, but I believe this to be a case of contextonomy. Rothbard knows it is not a simple case of over-or-under investment that causes cycles.

  7. OK - well we sound decently close then.

    Garrison, however, was downplaying the very prospect of overinvestment period (or at least overinvestment that wasn't compensated through underconsumption). He was downplaying the very prospect of a bubble economy in ABCT, in an effort to fit the plucking model. That was the effort that surprised me - not so much the point you're making here.

  8. To be honest, I'm completely ignorant on Friedman's plucking model so I can't really comment.

    Let me correct the previous statement, though, that overinvestment accompanies overconsumption. There is no trade-off because they work together.

  9. The plucking model is more of an empirical observation of Friedman's that a lot of the mid-twentieth century recessions have looked like somebody plucked a string down and it bounced back up to a steady growth path. To put it more formally, he found that the magnitude of upswings in output were far more correlated with the bust preceding the upswing than with the bust following the upswing.

    This suggests a bust-boom cycle rather than a boom-bust cycle. It's one of the grounds on which Friedman dismissed the Austrian theory and other theories that he perceived to be boom-bust.


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