By neoclassical, I mean modeling rooted the tenets of optimization and general equilibrium. This and next week’s post look the last half century of the neoclassical study of business cycles, here measured by unemployment fluctuations. It will be short. The literature has only two distinct neoclassical theories of continuous-equilibrium rational-choice unemployment. The first was pioneered by Robert Lucas in the 1970s and will be outlined immediately below. The second is much more recent, culminating in GEM Project research that has been featured in this Blog. The newer version of continuous-equilibrium unemployment will be summarized next week.
Lucas’s market-centric equilibrium. A half century ago, Robert Lucas figured out how to model business cycles as a dynamic equilibrium phenomena, changing the course of macroeconomics. His 1972 paper “Expectations and the Neutrality of Money” laid out the core theoretical apparatus, featuring rational expectations, that explains optimizing production/consumption choices in the context of imperfect information. Lucas’s 1975 paper, “An Equilibrium Model of the Business Cycle,” is here understood as an extension of the earlier analysis, explicitly rooting it in the business-cycle literature.
According to Lucas, optimizing workers in specialized economies make intertemporal choices that are inherently restricted by signal-extraction problems. Upon observing a change in demand for labor services, they must decide whether the shock is permanent or transient. If the former, they engage in intertemporal work-leisure substitution. If the latter, they stand pat.
From Axel Leijonhufvud (1983): “By the early thirties, business cycle theorists had come to realize that the use of the equilibrium toolbox could be strictly justified only for stationary and perfect-foresight processes. This pretty much excluded business cycles – and there was no other toolbox. Keynes’s new method successfully evaded this dilemma. Lucas’s new method attempts to solve it.” In hindsight, the most remarkable aspect of his neoclassical revolution was its speed. In a few years, a small number of theorists, attached to a few universities, deposed mainstream Early Keynesianism and fundamentally altered the nature of macroeconomics.
Lucas’s solution, however, had troubling implications, especially for macroeconomists who need their modeling to be policy relevant. In the new neoclassical analysis, market disequilibrium, the focus of their evidence-consistent work, is eliminated. So is involuntary unemployment, the raison d’etre of Keynes’s General Theory. To the degree that Lucas produces real fluctuations, they are associated with voluntary adjustments of hours at work. Third, business cycles are always benign, reflecting optimizing choices that improve, not damage, collective welfare. The detailed apparatus of identifying and managing aggregate nominal demand – the principal achievement of Keynes and the Keynesians – can be scrapped
New Keynesians. With the subsequent emergence of New Keynesians (NK), the friction-augmented general-market-equilibrium (FGME) model class became dominant in stabilization research, driving the market-disequilibrium Early Keynesian Neoclassical Synthesis out of graduate-school curriculums and mainstream journals.
NK theorists were eager to reassert the neoclassical tenets of optimization and equilibrium in macro modeling. But they generally found Lucas’s intertemporal-substitution mechanism unappealing, largely because of its substantial misalignment with the available evidence. The most troubling – indeed, downright embarrassing – inconsistency with readily observable facts is Lucas’s insistent rejection of the existence of involuntary job loss. NK theorists have invested huge resources and thousands of top-tier journal pages to the investigation of rational market frictions in the attempt to identify a optimizing super friction that suppresses wage recontracting sufficiently to microfound causality from nominal demand disturbances to evidence-consistent movements in unemployment. Most of the effort has focused on elaborating on the venerable market-centric labor search/match theory long used to explain frictional joblessness, hoping that a punched up approach would also explain involuntary joblessness. They failed. (There is more next week about this research-focus debacle and its extraordinary damage to stabilization-relevant macroeconomics.)
In their continuing pursuit of FGME research, NK theorists have quietly accepted Lucas’s conclusion that forced job loss and involuntary unemployment cannot exist in general market equilibrium. They now accommodate layoffs in recession by assuming the most inoffensive wage rigidity available, being careful to discourage any new MWR research and hoping nobody who matters will make a big deal about their violation of the fundamental neoclassical tenets of optimization and equilibrium. In other words, nobody will object to their quiet return to the ignominiously exiled Early Keynesianism and its keystone Neoclassical Synthesis. It should be made clear that, unlike the New Keynesians, Lucas – ever the rigorous theorist – has never been embarrassed about his rejection of forced joblessness. He correctly argues that the concept cannot exist in general-market-equilibrium modeling. His advice is to accept that fact, stop obsessing about cyclical unemployment, and continue to perfect FGME modeling absent involuntary job loss.
Blog Type: New Keynesians Saint Joseph, Michigan