Last year when Jordi Galí published “The State of New Keynesian Economics: A Partial Assessment” in the Journal of Economic Perspectives (2018), I paid attention. Galí is among the elite of New Keynesian theorists, making him a bellwether on the condition of mainstream macro theory ten years after its spectacular failure to be useful during the Great Recession. Most embarrassingly, the stabilization policymakers were not shy about how inadequate they believed modern macroeconomics to be.
Galí’s concludes that there has been great progress from 2008-09: “New Keynesian economics is alive and well. The New Keynesian model has proved to be quite flexible, with a growing number of extensions being developed by researchers in order to incorporate new assumptions or account for new phenomena.” I looked forward to reading his rationale for that optimism. I was disappointed. Especially relative to the fundamental problem of usefully explaining the extreme instability that characterized the perilous Great Recession, New Keynesian economics remains pretty much in the same sorry state it was a decade ago. Discussing three problems captures Galí’s misguided assessment.
First problem. Galí gets the gets the central problem wrong: “Ten years later, tons of ammunition has been fired against modern macroeconomics in general, and against dynamic stochastic general equilibrium models that build on the New Keynesian framework in particular. The criticisms have focused on the failure of these models to predict the crisis, a weakness often attributed to their lack of a financial block in the model that could account for the key factors behind the crisis, whose origin was largely financial. Other aspects of the New Keynesian model and its extensions that have been the target of criticism include the assumptions of rational expectations, perfect information, and an infinitely lived representative household.”
The GEM Project gets the central problem right. Market-centric rational-behavior modeling is inherently not up the task of explaining macro instability in highly specialized economies and must be fundamentally reworked. Failing to microfound MWR which we now know requires the generalization of rational exchange from the marketplace to information-challenged workplaces, Galí is forced to rely on the badly misleading neoclassical description of the Keynesian micro-macro nexus tasked to reverse (in important circumstances) real-to-nominal causality.
Second problem. He gets the solution wrong: “Much recent research, for instance, has been devoted to extending the basic model to incorporate financial frictions…. In addition, the New Keynesian model has been the framework of choice in much of the work aimed at evaluating alternative proposals to stimulate the economy in the face of the unusual circumstances triggered by the crisis, including the use of fiscal policy and unconventional monetary policies.”
NK market-centric general-equilibrium modeling, unable to rationally suppress wage recontracting, severely restricts the propagation of macro shocks by nominal demand. Galí and his NK colleagues are then forced to focus their extreme-instability analysis on the nature and prevention of originating macro shocks. In particular, they have identified the prevention of future financial shocks as the stabilization-relevant object of NK macro theory. How to be polite here? Doesn’t everybody know, at least deep down, that the prevention of harmful financial shocks is not possible short of erasing risk from the macro economy. The continuing focus of Galí and his friends on the particular problems of 2008-09 illustrate the impossibility of preventing all originating financial shocks that come in huge array of permutations, especially as rational actors adjust to new regulations. Moreover, who decided that only financial shocks are disruptive? If prevention is all we can come up with, macro stabilization is doomed. Fortunately, the GEM Project has a feasible, therefore better, idea: Prevent nonstationary nominal demand contractions from propagating inevitable financial shocks.
Third problem. It is not surprising that Galí also gets the useful research agenda wrong, citing “recent work on two very active research programs: the implications of the zero lower bound on nominal interest rates and the interaction of monetary policy and household heterogeneity.” The GEM Project Blog has devoted significant space to the zero lower bound debate. The generalized-exchange analysis is definitive. In a nutshell, rational investment spending is intuitively shown to be dominated by expectations of pure profit; interest rates play a minor role. In the extreme-instability circumstances that characterize the Great Recession, the conclusion is even stronger. Interest rates played no noteworthy role in the collapse of investment spending. With respect to the other critical component of weakening aggregate demand, consumption spending is overwhelmingly driven by expected income and wealth – a relationship that is microfounded in the GEM Project but, to Barro’s regret, must be ignored in market-centric general-equilibrium modeling. The zero lower bound problem is an artifact of the misleading NK market-centric macro model, playing no significant role in the extreme instability of 2008-09.
It is also interesting that doomed NK market-centric research on adequately modeling extreme instability has drifted off to the difficult problem of household heterogeneity. Generalized-exchange modeling points to firm heterogeneity as the truly powerful rational-behavior innovation here.
Fundamental message. The mainstream market-centric general-equilibrium model is badly flawed and consequently provides badly misleading guidance to NK analysis of actual macro instability. That problem has been featured in the GEM Project Blog, making clear that there is no easy fix. If the centrality of rational behavior, optimization, and equilibrium are to be preserved, the NK reliance on market-centric analysis must be scrapped. Mainstream macro modeling must be reconstructed to accommodate the essential existence of two venues of rational price-mediated exchange – markets and workplaces restricted by costly, asymmetric information. We must stop deluding ourselves that future macro shocks can be prevented and that effective stabilization policy is not rooted in the management of total-demand contractions that propagate the shocks which eventually occur.
Blog Type: New Keynesians Chicago, Illinois