As I see it, Paul Romer believes that George Stigler laid down the methodological principal that one should always assume perfect competition in one’s microfoundations, and in so doing Stigler was acting as an ideologue rather than a technocrat, and that this is harmful.
It seems to me that Paul is more right than wrong.
It seems likely to me that Paul Romer has properly classified George Stigler in his beastiary. As I see it, George Stigler thought we did not need to evaluate whether Chamberlinian and other models of imperfect competition were true because we already knew that they were intellectually dangerous: not good to think with. Why? Because they gave an intellectual opening to “planners”–who were completely unaware of thie magnitudes of potential government failure–to argue for their interventionist “planning”.
A sound analysis, I think Stigler thought, would take account of the magnitudes of potential government failures. It would lead to the correct conclusion that, even if Chamberlinian features were true, and even if they were included in the model, anything other then letting the market rip would be a bad policy mistake. However, Chamberlinian models that ignored government failure could be used to argue for all kinds of destructive interventionist policies–antitrust policies, Keynesian policies, you name it. In this context, requiring that models assume perfect competition in their microfoundations as a methodological principle was a “Noble Lie”, a proper esoteric teaching. It led those who could not grasp the true esoteric teaching about the magnitude of government failure to the right and true conclusions. And one could teach those who could handle it the esoteric teaching–including the point that perfect competition was the proper microeconomic foundation to assume in the course of teaching gentlemen, and was a proper part of the methodology of positive and normative economics.
The problem is that Paul Romer wants to analyze issues in which perfect competition is not an assumption that leads one to anything like the True Knowledge about appropriate policies for economic growth. And yet Lucas and company are insisting that he make it. And it is to this, I think, that Paul Romer objects–most strongly.
And I think Noah misses this. Paul objects not to simplified microfoundations, and not to math, but rather to “mathiness”, which is restricting your microfoundations in advance to guarantee a particular political result and hiding what you are doing in a blizzard of irrelevant and ungrounded algebra.
Paul Romer on Mathiness: “Paul Romer… comes down harshly ‘mathiness’ in growth theory…:
…us[ing] math in a sloppy way, to support their preferred theories. Romer warns direly that the culture of econ theory has become a lot more tolerant of mathiness…. Romer has now joined the chorus of old famous guys–Krugman, Solow, Stiglitz, Farmer–who are very vocally mad about the way mainstream economics theory is done…. Interestingly, although he’s talking only about growth theory… most of the people he’s mad at are the same guys that the Keynesians are mad at – Robert Lucas, Ed Prescott, and David K. Levine…. Prescott and McGrattan…. Boldrin and Levine…. Moll….
I think Romer is on to something about the culture of econ theory… [where] people seem to view math more as a tool for stylized description of ideas than as a tool for quantitative prediction of observables…. But… [even in] earlier models, I don’t really see much more tight connection of variables to observables…. [The] Solow model… tie[s] capital K to observable things like structures and machines and vehicles. But you’ll be left with a big residual, A. Then you can break A down and extract another term for human capital, H. Can you really measure human capital?… I don’t really see a huge difference between that and the ‘location’ used by Prescott and McGrattan (2010)…. Mathiness isn’t anything new, it’s just the way these econ fields work…
What Noah misses, I think, is what Bob Solow wrote the Solow model down do. The first point of the Solow model is that you cannot account for the coming of modern economic growth via any increase in the savings rate–in the accumulation of buildings and machines–as long as you think factors of production are paid their marginal product. The second point of the Solow model is that, properly calibrated, if factors are paid their marginal products, accumulation accounts for only a small part of per-capita growth at the frontier over any time period as long as a decade or so. The third point is that with the ancillary assumption that ideas can be learned by looking at the successful one would expect swift convergence to the frontier. The fourth point is that human capital–number of years of schooling–does not do much to resolve the modern economic growth puzzle.
Romer’s point is that good theory seeks strong correlations between observables that it then tries to explain as emergent properties resulting from some usefully-modeled market-structure foundations. Good theory does not theorize about unobservables. From this perspective, the Solow model is (mostly) a demonstration of approaches that do not work, and is useful because it rules things out–and it is thus the exact opposite of explaining modern economic growth via unobservables.