I just read Olivier Blanchard’s PIIE Policy Brief, “Do DSGE Models Have a Future?” (August 2016). Reading modern Blanchard is disheartening. He sensibly critiques mainstream DSGE macro modeling, building credibility among economists who are dismayed by the stabilization-irrelevance of consensus thinking. Then, in practice, he squanders that credibility by carelessness in identifying, or at least recognizing, new research that makes fundamental progress on macro theory that is both micro-coherent and stabilization-relevant by venturing outside the comfortable mainstream DSGE box.
Blanchard’s major flaw in assessing the state of DSGE modeling is a stubborn insistence on ignoring the obvious. Over the past decade, he has become increasingly determined to avoid nominal wage inflexibility capable of rationally suppressing recontracting. (The GEM Project has named such labor pricing “meaningful wage rigidity”.) Ignoring MWR is nothing new. Its absence has become an enduring characteristic of New Keynesian construction, and assessment, of DSGE models. Why have Blanchard and other NK theorists decided to disappear the keystone concept in Early Keynesian modeling that established macro as a separate, policy-relevant field of study? The question is separable into two parts. Is MWR necessary to micro-coherent, stabilization-relevant macro theory? If so, why is the phenomenon ubiquitously excluded from mainstream DSGE analysis?
The two questions inform the dilemma that is diminishing Blanchard during the elder-statesman phase of his career, muddling his contribution to stabilization-relevant macroeconomics. In his Policy Brief, Blanchard reiterates two central NK DSGE model requirements: fully microfounded rational behavior and the central role for “essential distortions” that emphasize “nominal rigidities and a role for aggregate demand”. His dilemma centers on how to comply with both mandates.
On the one hand, he can ignore the Early Keynesians who assigned MWR the central role in reversing the real-to-nominal causality that is at the core of market-centric Walrasian general-equilibrium analysis. Ignoring MWR, however, is destructive to the stabilization relevance of any macro model. The Early Keynesians understood that MWR uniquely suppresses labor-price contracting and that such suppression is necessary for the existence of involuntary job loss, especially the layoffs that result from adverse nominal demand disturbances. Stabilization-relevant theory cannot ignore layoffs that account for the lion’s share of rising unemployment in recession.
On the other hand, he can become stabilization-relevant by simply assuming MWR, motivating a central role for nominal demand. He can return to the 1960s consensus and be excoriated by the modern mainstream for violating the central rallying cry of the New Classical macro revolution that effectively banished Early Keynesians from graduate-school instruction and cutting-edge research: that “the behavior of consumers, firms, and financial intermediaries, when present, is formally derived from microfoundations.” Moreover, why violate a requirement that is eminently sensible? The real problem burdening Blanchard and other New Keynesians is much less complicated. They have never been able to figure out how to microfound MWR.
Return to our two questions. First, if the rational existence of IJL induced by adverse nominal demand disturbances matters in the construction of micro-coherent, stabilization-relevant macro theory, and it does, MWR must be a crucial part of any legitimate research agenda. It cannot be ignored. Second, MWR is ignored because mainstream theorists are badly frustrated by decades of failure in figuring out how to microfound it and have decided to push it off center stage.
O.K. Microfounding MWR turned out to be a really tough problem. Discouragingly slow, faltering progress on it is forgivable. What is not forgivable, especially for the Robert M Solow Professor of Economics Emeritus at MIT, is that New Keynesian gatekeepers have given up trying. They are no longer willing to work at an obviously crucial macro frontier and, what’s worse, they are dealing with their failure by delegitimizing the problem, making it difficult to disseminate new research on microfounding MWR by theorists who have not given up.
Blanchard and other NK gatekeepers make it especially difficult to work on MWR by enforcing restrictive mandates arbitrarily attached to DSGE modeling. The requirement that all rational exchange be restricted to the marketplace is most interesting to those of us investigating the intuitive generalization of rational exchange from the marketplace to workplaces inherently constricted by costly, asymmetric information and routinized jobs. Relaxing market centricity opens research to practitioner-recognized labor pricing that rationally suppresses downward wage recontracting over stationary business cycles. In stabilization-relevant macro theory that microfounds the NK “essential” role for nominal demand, MWR must be assigned a central role explaining countercyclical involuntary job loss and procyclical movement of employment, output, labor income, and profit. Moreover, with respect to a chronic DSGE-modeling problem that Blanchard emphasizes in his Policy Brief, the rational existence of MWR also microfounds the once-famous Barro-Grossman’s Keynesian analysis of the primacy of income in the determination of consumption. (See Chapter 6.)
Blanchard concludes his summary with his belief that things are pretty good in the DSGE-model world. There are problems, but nothing that cannot be fixed. There is need to look for big changes that depreciate mainstream human capital and embarrassingly falsify what has been taught to graduate students for more than a generation. So what if stabilization authorities in the aftermath of the extreme 2008-09 instability, generating more than 6 million involuntarily lost jobs and trillions of dollars in welfare cost, bitterly complain about the uselessness of the mainstream DSGE modeling. Relentless assertion of the established market-centric model is already smothering such complaints.
What is already happening is not an acceptable outcome. What I cannot figure out is why Blanchard does not follow his mentor Robert Solow’s example and simply admit that he has not been able to microfound MWR, IJL, and the centrality of nominal demand in the instability experienced in highly specialized economies. But, and Solow was strong on this, the absence of personal success does not diminish the criticality of suppressing downward wage recontracting in the construction of useful stabilization-relevant, micro-coherent models. From that perspective, Blanchard might be better able to accept that the future of DSGE modeling, absent one essential change, is pretty bleak. The requirement restricting rational exchange to the marketplace must be eliminated, substituting general decision-rule equilibrium for general market equilibrium in the DSGE methodology. The exchange generalization investigated in the GEM Project makes micro-founding MWR, IJL, and the central stabilization role for nominal demand pretty straightforward.
Blog Type: Wonkish San Miguel de Allende, Mexico