Plague of Otiose Modeling

Print/Save PDF

“Otiose” is a more polite synonym for “useless”. It came to mind during my recent rereading of “Labor Markets and Monetary Policy: A New Keynesian Model with Unemployment,” American Economic Journal: Macroeconomics (April 2010) by Olivier Blanchard and Jordi Galí. I remembered the article being something of a poster child for the otiose research that has for decades dominated mainstream New Keynesian (NK) stabilization analysis. The cumulating reputational damage, added to the embarrassment of being irrelevant to public and private policymaking, has dug a deep hole out of which the macro academy has been unable to climb.

Otiose modeling. The most striking aspect of the AEJM paper is how genuinely unaware B&G appear with respect to its practical use. They tout a more realistic labor market and imply stabilization-policy relevance. If “realistic” is measured by correspondence with actual evidence and policy relevance is rooted in useful advice, they come close to delivering neither. “We extend the NK model by introducing a more realistic labor market, with frictions similar to those found in the Diamond-Mortensen-Pissarides search and matching model of unemployment (the DMP model, henceforth). This extension allows us to characterize the effects of productivity shocks on unemployment and inflation, and to show how these effects depend both on monetary policy and on the nature of labor market frictions. It also allows us to derive optimal monetary policy, and characterize its dependence on labor market frictions….” Note that the article’s otiosity is telegraphed by the unjustified reliance on productivity shocks to induce macro instability. Here is a rule of useful stabilization modeling rooted in long practical experience: Policy-relevant instability modeling requires business cycles to be primarily driven by aggregate nominal demand.

B&G additionally claim to have introduced unemployment into NK instability modeling: “Inefficient unemployment fluctuations arise when we introduce real-wage rigidities  As a result, in the presence of staggered price setting by firms, the central bank faces a trade-off between inflation and unemployment stabilization, which depends on labor market characteristics. We draw the implications for optimal monetary policy.”

The innovation that NK theorists have sought since they signed on to the New Neoclassical Synthesis (NNS) is to ground meaningful wage rigidities in the non-otiose tenets of optimization and equilibrium. Satisfying that goal would have placed the B&G model among the all-time most useful contributions to the stabilization literature. Instead, after a lot of hand-waving, the authors abandon any such ambition and instead default to the typical NK practice of assuming arbitrarily staggered wage setting of the sort associated with Calvo (1983). Needless to say, that is a deeply disappointing outcome given the B&G promises to do better. Calvo rotating wage-rigidity assumption is inherently irrational, is not fact-based, and is explicitly prohibited by the NNS. Resorting to that familiar cop-out badly misleads any attempt to describe actual labor pricing and use, providing no useful guidance to useful model-building.

Non-otiose modeling. The GEM Project constructs policy-useful macro theory on optimizing, price-mediated exchange organized by continuous general decision-rule equilibrium. The case arguing for the use of rational behavior to motivate model behavior is powerful. Think through the alternatives. Irrationality, the replacement with sufficient scope, is essentially random and surely not a viable option. Positive virtues of rationality include its clarity and persuasive power, as well as the capacity to systematically interpret evidence and to distinguish among competing theories.

Those virtues are commendable, but I believe that by far the most important benefit from rational behavior is the robust guidance it provides in the construction of non-otiose economic models, helping to better explain and extrapolate important economic phenomena. That conclusion is vividly illustrated by the checkered history of wage-rigidity analysis in mainstream macro theory. As noted, the default NK approach has for some time been relying on convenient, albeit irrational, assumptions that suppress wage recontracting. Most popular is Guillermo Calvo (1983) positing that a rotating fixed fraction of wages to remain unchanged in each period, permitting the remaining wages to be recalibrated to changing market conditions. The probability that a subsector labor price cannot be adjusted in the given period is pure convenience, implying strong restrictions on rational behavior. Calvo’s staggered pricing helps facilitate aggregation and appears to be, despite its dependence on arbitrary free parameters, the labor-pricing mechanism of choice in modern NK modeling. Early Keynesian (EK) theorists made clear that their use of an irrational time-separation assumption to suppress wage recontracting and produce involuntary layoffs was an interim choice, awaiting guidance from microfounded wage rigidity. But NK theorists believe that rational-behavior wage rigidity is no longer a requirement of adequate stabilization theory.  For B&G and their NK brethren to keep betting the ranch on the faith-based argument that a super friction is still hiding out there waiting to be discovered has become as unacceptable today as it was for EK theorists decades ago.  

The GEM Project comes to the rescue by generalizing rational exchange from the marketplace to workplaces restricted by costly asymmetric employer-employee information and routinized jobs. Both irrational Calvo and rational GEM labor pricing provide linkage between adverse demand disturbances and involuntary job loss. But the important message here is that the rational-behavior GEM provides much more powerful theorist guidance for how highly specialized economies actually work than does the way-off-base Calvo assumption.

Generalized-exchange modeling allows macroeconomists to get serious about realistic depictions of labor markets in highly specialized economies:

  • Rational-behavior two-venue modeling identifies LEV workplaces as the almost exclusive home for rational wage rigidities and involuntary job loss in the aftermath of contracting nominal demand. Calvo’s assumption locate both phenomena in his arbitrarily specified, ill-defined rotating subsector, badly misleading anyone interested in the true nature of labor pricing and cyclical layoffs.
  • Rational-behavior two-venue modeling identifies the circumstances when downward wage rigidity is loosened, permitting optimizing wage givebacks, providing a rational foundation for the important modeling of the role of wage adjustment speed in the circumstances of chronic mass unemployment that is today helping to usefully refine Keynesian analysis.   
  • Along similar lines, rational-behavior two-venue modeling associates temporary mass involuntary job loss (layoffs) with stationary total-spending cyclicality and permanent mass involuntary job loss (job downsizing) with nonstationary demand collapse. The analysis permits the rational modeling of 1930s-class depressions, a crucial phenomenon well beyond the reach of Calvo’s irrational rotating wage rigidity.
  • Rational-behavior two-venue modeling permits the effective specification of wage rigidity, extending the concept to chronic wage rent consequentially paid by highly specialized firms restricted by costly, asymmetric workplace information and routinized jobs.

Practitioners scornfully reject the NK market-centric descriptions of their profit-seeking management of labor pricing, acquisition, and use. How intensively to recruit new hires doesn’t come close to being their most important labor concern. They believe the assertion that they negotiate labor pricing individually with workers is ridiculous, typically refusing to believe that reputable economists claim such a foolish notion is true. They cannot believe that reputable macro theorists do not know that wages are typically posted. They also know that layoffs are involuntary and occur when product demand weakens. Absent a noncyclical collapse in demand they know that cutting wages is irrationally inconsistent with profit-seeking. They additionally know that in the circumstances of collapsing demand wage givebacks occur relatively slowly.

Blog Type: New Keynesians

Write a Comment

Your email address will not be published.