Boettke on the Lamp Post Problem
By Arnold Kling
I recently came across Where Did Economics Go Wrong? Modern Economics as a Flight From Reality, published by Pete Boettke in Critical Review. Looking at its themes, you can see issues that I have dwelt on considerably on this blog. However, given that it appeared in 1997, it won’t be Boettke who gets accused of plagiarism. For example, he writes,
we can usefully distinguish the older use of the equilibrium model as an ideal type from its use by free-market Chicago-school economists as a description of reality, as well as its use by interventionist Neo-Keynesians as a critical standard with which reality could be indicted when it failed to measure up. In the latter two uses of equilibrium, it constitutes a static ideal, and the question is whether reality does or does not match it. In the ideal-type use, by contrast, the question is how departures from the ideal type–denied by the Chicago school; equated with market “failure” by Neo-Keynesians–may constitute forms of incomplete success…disequilibrium is not necessarily a market failure; something less than perfection may yet be better than any attainable alternative.
What both schools overlooked was the fact that equilibrium…could not possibly represent a dynamic world of time, ignorance, and uncertainty; but that the divergence between ideal and reality can highlight the ways reality may have institutionalized error-correcting properties that can, in fact, be seen as propelling the world in a direction reminiscent of general equilibrium. But like any ideal type, equilibrium is a postulate that is not necessarily effected in the real world.
The way I put it is “Markets Fail. Use Markets.”
Another overlap with my concerns is what I call the lamp post problem. Boettke sees economists’ use of formal models as blinding them to the “world of time, ignorance, and uncertainty.”
As an example, consider the conventional treatment of the Great Depression as a market failure. The conventional economist looks at the 1930s and sees an economy that is not on its production possibility curve. But which production possibility curve should it be on? The production possibility curve of 1950, the next peace-time period of full employment, was not feasible in the 1930s. The physical capital was not available for the boom in suburbia that fueled post-war growth, nor was the human capital available for the clerical work that emerged.
On the other hand, the economy could have stayed on the production possibility curve of 1929. But that would not have been efficient. As Alexander Field pointed out in A Great Leap Forward (both Tyler Cowen and I put this among the best economics books of 2011), the production possibility curve shifted out in the 1930s–more than in any other decade, according to Field.
The PSST approach sees the unemployment of the 1930s and the outward shift of the production possibility curve as both reflecting the groping of entrepreneurs toward new patterns of sustainable specialization and trade. The conventional macroeconomist treats the higher production possibility curves as if they appear by a process of immaculate conception and sees the unemployment as a market failure.
The PSST approach is not easy to embed in a formal model. One can compare formal models to the lamp post in the old joke about a drunk who loses his watch in a dark place but decides to search for it somewhere else where there is a lamp post, because the light is better there.
For the economist, is it better to grope in the dark, meaning to discard formal models? Or is the better approach to strain to bring the lamp post to the likely place to find the watch? That is, should we attempt to formalize the issues of “time, ignorance, and uncertainty.”
Boettke points out that attempts to use formal models in the “economics of information” wound up treating information as something that a consumer or entrepreneur can simply “pull off the shelf,” perhaps by paying some cost. This serves to re-cast the problem into familiar equilibrium terms, But it does a poor job of representing the process of discovery, which I think is at the heart of the PSST approach to economic fluctuations.
So is there a way to make formal modeling a help, rather than a hindrance? Can the lamp post be moved to where the watch can be found?