Saturday, February 15, 2014

Where has Tobin Gone?

In 1937, was born the spawn of Satan, and he was named Robert E. Lucas Jr. On his scalp was etched the dreaded "666." In the 1960s, Robert attended the Academy of the Satanists, known to the public as the University of Chicago, where he was schooled in the ways of Lucifer (Milton Friedman). In the 1970s, Robert, the antichrist, decreed that there should be no more worship in the Church of the Keynesian Cross. A secret society was formed, and in the 1990s they didst make a pact with Michael Woodford. In return for his life, Woodford was sworn to use the Tools of Satan - the functional equation, the contraction mapping theorem, the dreaded Euler equation. Further, Woodford, his henchmen and henchwomen, were to appoint Satan's ministers to the key positions of power in the economics profession - journal editorships - where they were to enforce the use of the Tools of Satan. Publication of even one paper making use of the tools of the Church of the Keynesian Cross would mean banishment to the underworld forever.

Which seems to capture how Paul Krugman thinks about the economics profession. Here's the part of Krugman's post I want to focus on:
Let me offer an example of how this ended up impoverishing macroeconomic analysis: the strange disappearance of James Tobin. In the 1960s Tobin developed and elaborated a sophisticated view (pdf) of financial markets that offered insights into things like the role of intermediaries, the effects of endogenous inside money, and more. I’ve found myself using Tobinesque analysis a lot since the financial crisis hit, because it offers a sophisticated way to think about the role of finance in economic fluctuations.

But Tobin, as far as I can tell, disappeared from graduate macro over the course of the 80s, because his models, while loosely grounded in some notion of rational behavior, weren’t explicitly and rigorously derived from microfoundations. And for good reason, by the way: it’s pretty hard to derive portfolio preferences rigorously in that sense. But even so, Tobin-type models conveyed important insights — which were effectively lost.

Then came the financial crisis, and many economists apologetically admitted that they had erred by not incorporating finance into their models, and announced plans to try to do that in the future.
Tobin has indeed disappeared from the curriculum in any PhD programs in economics that I know about. I work in the field of money and banking, go to conferences, publish papers, edit papers, referee papers, and I have never come across a paper in the mode of Tobin (1969) in my whole professional career - I got out of graduate school in 1984. I have certainly read Tobin's stuff. There was a set of papers that I read while in graduate school that I thought were interesting, including the 1969 paper that Krugman mentions.

To understand why Tobin's work has disappeared from modern discourse in macroeconomics, a good starting point is Sargent's 1982 paper on "Beyond Demand and Supply Curves in Macroeconomics." In 1982, there was a budding literature on the foundations of monetary exchange, and explicit models of financial intermediation. Sargent discusses Tobin's work directly, points out what the insights were from that work, and goes on to explain how the new literature deepened our understanding of Tobin's insights, and gave us new insights. In the 1980s, there was extensive work on financial intermediation models - Diamond and Dybvig's work (banking panics) and Diamond's delegated monitoring work. Yours truly, Bruce Smith, and Bernanke and Gertler, among others, thought about how to integrate some of those ideas into general equilibrium frameworks. There has been a wide-ranging program in monetary economics for the last 35 years or more, including work by Kiyotaki and Wright, Rocheteau, Lagos, yours truly (again), and many others. As one example of what you can do with those types of models, you can look at some of my papers (this one, or this one). As well you can look up some of the work of Randy Wright (Wisconsin), Ricardo Lagos (NYU), Guillaume Rocheteau (UC Irvine), Mark Gertler (NYU), Nobu Kiyotaki (Princeton), or Markus Brunnermeir (Princeton).

Why did Tobin's work disappear from discussion? It was superceded. We now have better models, that are much richer in their implications. Tobin was fine for his time, but why drive on the freeway in a golfcart, when you can drive a Ferrari (as one of my friends once said)? Krugman should read this stuff. I think he would like it. I, my colleagues, coauthors, and students, have not been "apologetically admitting that we erred by not incorporating finance into our models." That's what we do, and some of us have been doing it for a long time. As Krugman says, "it’s pretty hard to derive portfolio preferences rigorously..." So, I guess we must be pretty smart, as we can do some of that.


  1. Although I'm probably a bit of an old Keynesian, I'm very interested in thinking about the micro-foundations for asset allocation. One of the areas where I have previously used Tobin's framework for insight is in relation to preferred habit explanations for the term structure. Casual observation suggests to me that the different spending horizons that people have is an important factor here (pension funds account for a lot of term debt holdings). To develop some micro-foundations for this, I would imagine that you would need a model with agents with different tolerance of return risk at different maturities so, at the very least, some kind of OLG framework. I have had a brief look through the linked papers of your own (and I intend to work through these in more detail), but neither appears to incorporate this (apologies if I missed this). Is there any work you can refer me to that specifically looks at this? Thanks.

    1. In the paper on QE I wrote, I take the approach that I'm not going to think about preferred habitat or market segmentation, but focus instead on liquidity premia and collateral to get an effect from QE. That's not saying that market segmentation is not important - it could be. In any case, Tobin would have said that we start at the level of preferences over assets, which isn't very helpful if we want to think about QE. A Tobin framework would say that QE works, and the effect depends on how substitutable long bonds are for T-bills, which doesn't get you any further down the road. So, my work just looks at one aspect of the problem, and takes this a step deeper. It says that the effect depends on (i) how scarce collateral is in the aggregate (in a well-defined sense); (ii) how long-maturity bonds are valued as collateral relative to short-maturity bonds. So, my argument is that that's much more informative, but of course it's only a piece of the puzzle.

      For market segmentation, there is a whole literature on asset market segmentation and the effects of open market operations (Grossman and Weiss, Rotemberg, Lucas, Alvarez, worked on this). That idea could be used to think about QE. OG works, and Lagos-Wright shares some OG features, with some other benefits I think.

    2. Actually, I had read enough of that paper to appreciate the aspects you were focussing on, because I looked at it when you linked to it last year in your "deflationary QE" posts.

      Anyway, thanks for the references. I'll have a look through those.

  2. Tobin is a better economist than all the listed ones together as he had the betterintuition. Fancy models are useless when they do not match the facts or help policy (macroeconomists as engineers).

    People who use models that deal with financial frictions but do not understand the grave implications (Stiglitz and Greenwald) and remain laissez-faire Chicago boys (even ditching their Friedman and whining about too expansionary monetary policy) simply do not get it. Oh, perhaps they get the details but not the bigger picture.