Kartik Athreya is an economist at the Richmond Fed. He must be an interesting colleague. His recent book, Big Ideas in Macroeconomics (2013), identifies and elaborates upon four rules of engagement in modern macroeconomics. Those rules matter, defining what makes a model acceptable to gatekeepers in the academy. Nonconforming models are targeted for exclusion from mainstream discussion and dissemination.
First and foremost, a theory must be coherent. Models lacking consistent microfoundations are today rejected out of hand. From Athreya (p.16): “Macroeconomic accounts should be forced to be internally consistent.… The only permissible form of disagreement about any explanation for a given set of facts (data) among macroeconomists should be disagreement on the appropriateness of premises, not conclusions given those premises.” Second, agents should optimize well-specified objective functions, the pursuit of which is organized around the modern decision-rule concept of equilibrium. Third, the model can be used to advise stabilization authorities only if its assumptions sufficiently correspond to relevant evidence. If not, the model may still be admitted to active debate and research; but it should not be used for policy guidance. Fourth, trading arrangements that provide the context for rational price-mediated interactions should be clearly specified.
Looking at that list, I would prefer some attention to the particular worthiness of assumptions that are axiomatic. Friedman aside, should there not be bonus points for theorists who construct models on primitives that are widely understood to be true? But, overall, Athreya’s rules look good to me. The development of responsible macroeconomics would benefit from their broad application by gatekeepers in the academy. Closer to this blog’s purposes, the acceptability criteria can also be used to contrast the two prototypical models featured in the GEM Project: mainstream coherent market-centric dynamic stochastic general equilibrium (DSGE) and upstart coherent two-venue dynamic stochastic general equilibrium (TVGE). In the first, rational exchange is restricted to the marketplace; in the second, such exchange is generalized from the marketplace to large-establishment workplaces.
With respect to the first rule, both the single-venue DSGE and the TVGE analytic frameworks are coherent. In the second, each model class is populated by agents who are motivated by clearly specified objective functions and are in continuous decision-rule equilibrium.
With the third rule, the performance of the two model classes begins to diverge. Most critically, mainstream market-centric DSGE analysis cannot coherently suppress wage recontracting, implying that its unemployment is inherently voluntary. (Chapter 1) By contrast, TVGE analysis rationally suppresses large-establishment wage recontracting, enabling construction of meaningful wage rigidity (MWR) that causally links nominal disturbances to involuntary job loss. (Chapter 2) The assumptions and outcomes of coherent generalized-exchange model are broadly consistent with relevant evidence and can be taken seriously by stabilization authorities. Meanwhile, the assumptions and outcomes of coherent market-centric modeling are consistent with remarkably little of the available evidence, starting with 6 million involuntarily lost jobs in the recent Great Recession. By Athreya’s rules, policymakers cannot look to it for guidance. That a lesson was well understood by Ben Bernanke in his successful management of the 2008-09 extreme instability.
Fourth, market-centric theorists fail to adequately deal with modern trading arrangements. Despite the Nobel-honored contributions of Coase, Simon, and Williamson as well as the ubiquity of easily observable and documented corporate behavior, the dogged mainstream pursuit of Walrasian-class model-building has arbitrarily restricted price-mediated exchange to the marketplace. Consensus gatekeepers’ insistence on market-centric trading arrangements has been greatly damaging, most recently reducing mainstream theorists to irrelevancy, frequently accompanied by scorn, during the 2007-09 Great Recession. By contrast, TVGE modeling takes what has been learned about trading arrangements over the past 100 years seriously, plausibly melding price-mediated exchange that occurs in the marketplace and workplace.
In practice, consensus gatekeepers of what is acceptable in the academy are, for the rules other than coherence, surprisingly hit and miss in aligning their behavior with Athreya’s rules of engagement. In particular, mainstream theorists frequently and conveniently ignore the requirement that policy advice be grounded in models the premises of which are broadly consistent with available evidence. Perhaps if more economists objected to proliferating policy theorems that are out of compliance with the third rule, gatekeepers would be pushed to become more responsible managers of the development of stabilization-relevant macroeconomics. I, of course, believe that a useful starting place would be recognition that the two-venue model class produced in the GEM Project uniquely conjoins analytic coherence and stabilization-relevance. Think about it. More than a hundred years of frustration from failed mainstream market-centric efforts to engineer that crucial combination makes a strong case that exchange generalization is today, and will be for some time to come, the biggest idea in macroeconomics.
Blog Type: Wonkish Chicago, IL