Paul Samuelson, a former colleague of mine at MIT, is recognized as a giant among economic theorists. Less known, he was also once a poster child for macroeconomists’ on-going struggle between stabilization relevance and model coherence. In a 1986 interview, he recalled his initial resistance to Keynes’s General Theory. In particular, he remembered getting stuck on its assertion of equilibrium involuntary unemployment. He understood his unease to be ultimately rooted in labor-price rigidity. Why are employers reluctant to cut nominal wages? Would such cuts help eliminate high joblessness? Does labor-supply elasticity, associating income loss with lowered wages, matter? Are workers even on their supply curve?
The evidence has long been strong that large firms are reluctant to cut wages in lieu of temporary layoffs but that they do eventually obtain wage givebacks given sufficient permanent job destruction. Samuelson’s problem was that nobody (including Keynes) could figure out how to wedge meaningful wage rigidity (MWR) into market-centric continuous-equilibrium modeling. It was a huge stumbling block that no mainstream theorist could produce, absent MWR, involuntary job loss. Free parameters enable stabilization relevancy at the cost of problematic model incoherence. Samuelson’s (1986, pp.159-161) personal solution was characteristically practical: “The way I finally convinced myself was to just stop worrying about it…. I assumed a disequilibrium system, in which people could not get on the supply-of-labor curve.” The powerful Neoclassical Synthesis, a golden-age of stabilization-relevant macroeconomics, and the subsequent 30-year macro-methodology war all took root.
The greatest contribution of the postwar anti-Keynesian insurgents, the winners of the enduring theorist war, has been making their rejection of Samuelson’s free-parameter solution stick. They properly insist on coherence in economic modeling. Incompatible models for, on the one hand, household and firm behavior and, on the other, fluctuations and growth are unacceptable for good reasons. Meanwhile, their greatest failure has been their longstanding inability to explain the most critical stabilization evidence, starting with the unshakable fact of U.S. recessions that most of the higher unemployment results from involuntary job loss.
The grand struggle between macro-instability facts and analytic coherence set the stage for an interesting, misunderstood contribution to modern thinking. The acknowledged leader of the anti-Keynesian revolution convinced many good theorists to simply ignore involuntary unemployment. Robert Lucas’s (1981, p.243) message was that the writ of serious market-centric research does not extend to Samuelson’s central concern: “Involuntary unemployment is not a fact or a phenomenon which it is the task of theorists to explain…. It does not appear possible, even in principle, to classify individual unemployed people as either voluntarily or involuntarily unemployed depending on the characteristics of the decision problem they face. One cannot, even conceptually, arrive at a usable definition of full employment as a state in which no involuntary unemployment exists.” I found the emotion aroused by Lucas’s statement a bit surprising. It seems to me an accurate assessment of some of the consequences of working within a market-centric general-equilibrium model. (See De Vroey’s (2004) and Shapiro and Stiglitz (1985) for insightful elaborations on forced joblessness.)
The GEM Project breaks away from arbitrarily restricting rational exchange to the marketplace, resolving Lucas’s reservations and easily incorporating stabilization-relevant involuntary job loss into coherent macroeconomics. (Chapters 2, 3) Making Lucas’s assessment irrelevant reflects substantial progress in developing useful macroeconomics. Three obvious questions, heretofore unanswerable, no longer must be ignored. First, who are the involuntary job losers? In generalized-exchange modeling, that class of employees is identified by their critically suppressed decision problem. They were not offered, in lieu of losing their rent-paying jobs, a wage reduction that did not violate their market-opportunity costs.
Second, why don’t job losers immediately accept their best alternative job? Once again, the key is that forced job loss inherently involves chronic wage rents, a phenomenon that is out of reach of coherent market-centric theorists. Labor rents induce the rational rejection of immediately available market-wage jobs. The existence of rationed rent-paying jobs, for which job losers are qualified, would accept, but cannot obtain, can motivate joining the continuous-equilibrium Harris-Todaro unemployment queue (ȖⱤ). (Chapter 5) In a related phenomenon, job losers remain rationally unemployed during the period of price-discovery required to calibrate the rent received in their previous employment. Costly information is necessary in the determination of true market-opportunity costs and the fully microfounded decision to accept a lower-paying job. (Chapter 5)
Third is the jackpot question. Why doesn’t the employer offer a wage cut that would prevent forced job loss? Generalized-exchange analysis has derived, from axiomatic model primitives, continuous-equilibrium MWR, featuring cyclical downward nominal rigidity and chronic rents, that optimally suppresses large-establishment wage recontracting and uniquely microfounds involuntary job loss and unemployment. (Chapters 2, 3) The hard fact is, absent the MWR channel, the crucial, real-world phenomenon of forced job loss cannot exist in coherent macro modeling. The longstanding effort of mainstream theorists to construct macro theories that are both internally consistent and stabilization relevant while ignoring MWR is today, has been, and always will be an analytic dead-end.
In the end, the Project rejects both approaches, Samuelson’s to stop worrying about model coherence and Lucas’s to ignore involuntary unemployment. That happy outcome is made possible by the development of generalized-exchange macroeconomics.
Blog Type: Wonkish Saint Joseph, Michigan