By Nathan Tankus, a student and research assistant at the University of Ottawa. He is currently a Visiting Researcher at the Fields Institute. You can follow him on Twitter at @NathanTankus
Normally, I’m a harsh critic of neoclassical economics and neoclassical economists. However, sometimes the most frustrating things about neoclassical economists is their lack of familiarity with neoclassical models (especially older ones) and current neoclassical research. Monday provided a rather extraordinary example of this trend: Paul Krugman is apparently not familiar with Paul Krugman’s research! Before I explain why, I need to provide some context.
The major policy issue around the Eurozone has historically been seen as its existence as a currency area. As a result, neoclassical economists (especially American ones, including Krugman) have tended to analyze it in terms of Optimal Currency Area theory (I briefly discussed Optimal Currency Area theory in another piece about the Euro here). The originator of this literature, Robert Mundell, often called the “founder of the euro”, argued that the important element determining optimal currency areas was a high degree of labor mobility. Thus when Krugman started writing about the Eurozone, he inevitably pointed out that Euro didn’t live up to Mundell’s standards for labor mobility. Monday, Krugman claims to have realized that labor mobility within the Eurozone has been hurting the Eurozone:
And while I didn’t think of it until now, there’s even a case to be made that labor mobility within Europe is actually worsening the problem, making the euro less sustainable. Via FTAlphaville, Frances Coppola documents the extraordinary rates of emigration among young people in Europe’s disaster economies — not really a surprise when you consider the incredible levels of youth unemployment. But as she says, once those young people are gone, who will pay the taxes to support retirees?
The reason that this is so amazing is that Coppola’s reasoning comes straight out of his (admittedly older) research on Economic Geography. To quote Krugman himself:
Now suppose that some resources. say, workers are mobile. If one of the locations offers a larger market, they will have an incentive to move to that location. But the movement of workers itself tends to increase the size of the market wherever they go, decrease it where they come from; so one immediately arrives at the possibility that a small asymmetry between locations, perhaps arising from some small chance event, will prove self-reinforcing.
This type of locational analysis was used by Krugman to give some provisional explanations for the emergence of cities. It of course, is not limited to Krugman (indeed these ideas go back centuries) but he was the first person to formalize it mathematically among mainstream economists. It should be obvious that this logic is easily applicable to the Eurozone, where different “locations” are replaced by countries and the triggering event is large differences in each country’s level of unemployment (not to mention public services). The interesting question is why did this obvious connection go unnoticed by Krugman? The answer it seems (as is often the case) is the model he was applying.
As is often the case when something happens in the real world that contradicts a model, Krugman reached into the Optimal Currency Area literature looking for a reason that reality didn’t live up to the assumptions. Krugman discovered a reason in the writings of his colleague Peter Kenen on “fiscal integration” (see here). Since fiscal issues were obvious by that point, this was clearly an attempt by Krugman to rescue the perceived usefulness of Optimal Currency Area theory. However, Kenen’s work in this area clearly wasn’t a good guide to the Euro, especially since it convinced Kenen that the Euro was worthwhile. In 2003 he went so far as to state: “Hence, Britain should join [the Euro] soon. It should not stay outside, waiting for perfection, but should get inside, working for improvement.”
The truth is that in fact OCA theory wasn’t and isn’t a useful guide to the Eurozone crisis. While some followers of Optimal Currency Area theory were against the Euro, neither proponents or detractors had an inkling such a crisis was likely or even possible. Peter Kenen himself said as the crisis set in: “There was, I agree, excessive reliance by U.S. economists on the theory of optimum currency areas.”
Thus, Krugman followed many other neoclasssical economists (even after the crisis) in thinking about monetary integration in terms of this one (narrow) literature rather than other neoclassical models, including even his own. However, the truth would seem to be worse then this. Not only did his penchant for thinking in terms of mathematical (neoclassical) models lead him to ignore the destabilizing possibilities of inter-regional labor mobility within the Eurozone, but his models have traditionally been built on excluding these dynamics from the analysis of international economics and trade! In 1995 he co-wrote a paper called “Trade policy and the Third World metropolis” in which he states:
Immobility of labor also changes results in important ways. Simple geography models like Krugman  respond in a monotone way to declining transport costs: when these costs fall below a critical level, industry concentrates in one region. Here, because labor is immobile (and thus wage differentials between regions emerge), continuing reductions in transportation costs eventually lead to a reindustrialization of the low-wage region. We believe that this is not just an artifact of the model: it represents a real distinction between interregional and international economics because labor is in fact far less mobile between than within nations.
To translate this to English, when the real world comes to look more like traditional neoclassical models of comparative advantage and trade, poor and rich countries come closer to converging at least in terms of industrial concentration. As far as I can tell Krugman (nor the literature) never produced a formal model showing that lower international labor mobility has the same implications as no international labor mobility. As long as sustained net emigration and immigration is possible (which of course it is and does indeed happen) there is no reason why a cumulatively causal dynamic similar to his closed economy model isn’t possible (at least, I haven’t come across any mathematical defense of this proposition).
This has serious implications for how we understand the evolution of, and more importantly the formulation and application, of economic ideas. In his Development, Geography and Economic Theory. Krugman argues that, “the influence of ideas that have not been embalmed in models soon decays.” However, the discussion above shows that vague assertions that aren’t modeled but are associated with particular models (like low international mobility being analytically the same as no international mobility) have been repeated as axioms for at least two centuries.
Krugman’s explanation for why some models prosper is not historically accurate. International labor mobility was a common assumption among thinkers labeled by Adam Smith as mercantilists (such as Josiah Tucker and James Steuart) that was simply ignored and rooted out of political economy by a series of classical liberal economists starting with David Hume. The same writing by David Hume Krugman praises as “arguably the first example of modern economic reasoning.” This is arguable, of course, among people who aren’t intellectual historians. When I read the condemnation of the economic effects of emigration in Tucker’s writing and the arguments that rising wages would induce net immigration, I can’t help but think he’s a much more useful guide to the Eurozone and modern affairs then most modern economists (including Krugman). As should be obvious, the victory of the classical liberals wasn’t because the classical liberals were using mathematics or that their “model” was more clear – it was because the “Mercantilists” were attacked polemically and forgotten.
As Schumpeter noted in his great History of Economic Analysis:
Yet, as we know, facts and ideas that were familiar to the projectors of the seventeenth century and, in a purified form, to the scientific economists of the first half of the eighteenth, such as Boisguillebert, Cantillon, and Verri, might have set the writers of 1800–1850 on the track of what I believe to be a more adequate analysis. But these Facts and ideas were practically forgotten by 1800.”
Further, if in economics it is true, “that to be taken seriously an idea has to be something you can model”, should not economists be desperate to adopt new mathematical techniques for modeling as they are invented?
Yet the history of the last 60 years (at least) show this not to be case. It has been almost 60 years since systems dynamics was invented and mathematics that specifically eschews general equilibrium thinking has a large pedigree. Yet these modeling techniques have seen little use and, more importantly, the assumptions most models use are the same or similar to models developed as far back as the 19th century (specifically, Knut Wicksell’s work). To become a useful discipline mainstream economics has to either abandon its fetish with formal modeling or build better models. Preferably both.