Saturday, July 13, 2013

A suggestion for modeling tight money

Paul Krugman suggests that tight money advocates need to offer up better models if they want to be taken seriously (HT - Brad DeLong). I think it's a good point. He's mainly talking here about tight-money-oriented monetarists and New Keynesians.

Another group that isn't monolithic in its call for tight money, but is still a major contributor to that position is the Austrians, of course. It's an Austrian suggestion that I have to make, to any enterprising graduate students that are listening. Some Austrians - following the po-mo fork in the road that hit that school of thought a couple decades ago - rely on a more impressionistic, narrative case against monetary (and fiscal) stimulus, of course. But before all that came along Hayek offered a clear, neoclassical model that provided an argument for being concerned about the real distortions of expansionary monetary policy. The Hayekian model has had some innovations. Roger Garrison most notably has put the argument in modern language and added a production frontier to it and (in some versions) even a Keynesian cross.

The shortcoming of the Garrison version of Hayek's model is that it doesn't really catch up to the standards of modern macro in terms of formality, microfoundations, or rational expectations (even of a weak sort). A lot of Austrians like to think their ideas aren't given much credit because all those mean mainstream economists are just looking for political glory. This is stupid. And it's going to keep you in backwaters if you convince yourself of this and nobody is going to like you personally if you accuse them of this (to quote Tyler Cowen or Peter Boettke... I forget which... you are imposing a huge "lunch tax" if you talk like this).

I've had a thought on how to put together a modern Hayekian model for a while now that I thought I'd share. I think it would have a lot of credibility with mainstream economists, I just haven't personally been able to prioritize nailing it down. So feel free to give it a shot, just mention me in the acknowledgements as part of the inspiration. Let me know if you decide to work on it, though, so that if I've actually started to do something with it we don't duplicate efforts and insights.

1. Start with a Romer model of increasing product variety. The basic structure of this model is that there is a final goods sector and an intermediate goods sector with an increasing variety of products in the intermediate goods sector. Greater intermediate good product variety increases productivity in the final goods sector, which perfectly captures the Hayekian point about roundabout production... if...

2. ...if those intermediate goods have time component (which currently they do not). But adding this element should be relatively trivial - just index them for time. Now instead of a variety of products you have a time structure of production.

3. Monetary policy comes in in the profit maximization by the intermediate good producer. These are time-indexed goods now so costs need to include the interest rate (think of the intermediate good as a good-in-process that you don't get paid for until the final good is sold). An intermediate good with a higher time index is going to have higher interest costs associated with it. All this is pretty easy so far in the standard set up of the model... I imagine in solving it things will get trickier because of the added interest rates and that's where the work will come in.

4. This alone packs in an awful lot of Hayek. Of course once you've added a time dimension you can bring expectations in too if you want.

If you introduce a monetary shock into this I can't imagine it's going to have different results from the standard Hayekian model. The only difference is that this time it will be microfounded and a modification of a widely used model.

Of course, then you have to make the case that monetary policy is "too loose" right now. That's the hard part of the argument - they think money is way too loose right now while all of us think it's too tight. This is the point that I'm going to make strongly in a Critical Review article coming out later this year - that the Austrian argument is actually quite reasonable in a lot of ways, it's only that the idea that money is too loose right now isn't credible.


  1. I'm glad you bring up Roger Garrison. I really like the way he explains the Hayekian triangle, interest rate connection. The concept of a boom being a situation where we are "beyond the production possibilities frontier" makes a lot of sense to me.

    I'm guessing that many believe (if put into a Roger G model) that we are below the PPF and need to push ourselves out to that level, perhaps using NGDP as a guide. I believe that the NGDP trend could be overstated due to problems such as the multi-decade increase in the Freddie-Fannie portfolios and associated guarantees. This kept interest rates artificially low in relation to the amount of debt buildup that was occurring.

  2. BTW, your model sounds interesting, but it sounds like a lot of work.

    As for loose/tight money, my guess is that most Austrians don't see money as being loose right now in the hands of consumers. They see it as loose in the hands of financial institutions and other "early recipients" of loose money. Many have argued that that is why we've seen the price increases in stocks, housing, and bonds. I realize one could argue that housing and stocks were bouncing off lows and that bond price increases are an artefact of a depressed economy, but I think there is merit to the loose finance money story.

    Do you believe that money is currently tight for the financial industry?

    1. Do they think that money in the hands of consumers is just right?

    2. Well, one presidential candidate made the comment that if Bernanke insists on dropping money out of helicopters, he should drop it on the people instead of the banks.

      Personally, I think money is tight for consumers. Policies that inflate home values contribute to that problem in the long run by increasing housing costs, thereby reducing people's disposable income.

    3. Thanks, Ken. :)

      I also think that money is too tight for consumers and too loose for finance. IIUC, one problem is current law, which restricts the options for the Fed. But I also think that the whole monetary setup favors creditors too much.

    4. Due to practical considerations, I think it's pretty difficult to get away from favoring the creditor side with monetary policy. I also question whether helping people obtain more debt is the answer to the problem of people having excess debt. Unfortunately, I think we have to let the deleveraging run it's course.

      I'm not a fan of Fed intervention, but I do believe they are trying to get money in the hands of consumers via the mortgage securities purchases. Increasing home values increases potential HELOCs. I don't like the side effects, but I do believe some people will actually improve their debt situation (paying off credit cards for example).

      #3 in Daniel's post is pretty good. That is where the monetary effect should be seen. The Austrian view is that a distortion would occur due to the rate drop lowering "high time index" costs more than "low time index" costs. In other words, while everything becomes cheaper, long term projects become particularly so. The question is how significant that distortion is. The model he describes might actually be able to quantify that.

  3. Interesting post, Daniel Kuehn.

    But how does Roger Garrison's depiction of a Mises-Hayek ABCT-style situation deal with conditions under uncertainty or ambiguity?

    I feel that there is a kernel of truth to argue that a boom and bust can occur due to blunders by a monetary authority, but I don't really get the sense of whether ABCT describes how conditions of uncertainty or ambiguity gradually dissipate and a recovery is achieved.

    Also, here's a review of Roger Garrison's book.

  4. Here's a YouTube video where he explains ABCT and Capital Based Macroeconomic Theory.


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