Ptolemaic Woes: Misusing Search Theory

Print/Save PDF

Search theory has long provided an appealing, albeit incomplete, explanation of the natural rate of unemployment. However, when extended to employment/unemployment fluctuations at cyclical amplitudes, an ambitious expansion of the search-model agenda pioneered by Mortensen, Pissarides, and Diamond, the theory breaks down, defeated by problems that reduce it to a Ptolemaic exercise. While the much-used approach is consistent with the mainstream market-centric general-equilibrium model class, it is irrelevant to actual macro instability and the concerns of stabilization policymakers.

Voluntary unemployment. This is the central problem. Modern search/match (S/M) theory is inherently restricted to voluntary (frictional) unemployment. The market-centric model class has no access to meaningful wage rigidity and therefore cannot rationally suppress wage recontracting, cannot rationally generate involuntary job loss, cannot rationally explain recognizable movement in output and  employment that result from aggregate-demand disturbances, and cannot rationally produce chronic wage rent that forces the (cyclical and trend) rationing of good jobs.

Relying on the intuitive fact that frictional joblessness is decreasing in recruitment efficiency, the time required in matching vacancies and applicants is the raw material with which mainstream New Keynesian theorists work in their attempt to inject some, always tiny, cyclicality in voluntary joblessness in general-market-equilibrium modeling. Having generated an embarrassingly huge literature, modern search-match modeling has become the bedrock for many macro theorists’ careers. They simply ignore the common knowledge that their base theory is inconsistent with the most important evidence.

Job separation. Coherent S/M job separation is inherently restricted to voluntary quits, typically posited to occur exogenously at a constant rate. That model convenience is extraordinarily debilitating, assuming away the crucial part of cyclical dynamics that is associated with involuntary job loss. Outside of the macro academy, layoffs are understood to be the central characteristic of business cycles. Inside the academy, respected theorists get away with pretending that forced job loss does not exist.

Substantial match capital. The majority of new hires are for nonsupervisory, routinized jobs, the human capital for which is largely employer-specific and provided via on-the-job training. Substantial idiosyncratic match capital, an essential feature of S/M modeling, mischaracterizes most employees and the needs of their employers. Intensive information gathering in the hiring process is not cost-effective, especially when ubiquitous probationary periods more efficiently provide information about absenteeism, other work habits, and productivity. Intensive information gathering is no more than a Ptolemaic assumption, unsupported by the facts, that is used to protect the mainstream market-centric general-equilibrium model class and the many reputations closely tied to it.

Silly causation. S/M theorists conveniently posit that causation runs from the number of recruiters (R), mediated by their productivity (e), to new hires (H): R(t)=H(t)/e(t). Practitioners consider such causation laughable nonsense, pointing out that the level of job vacancies drives recruitment activity and the volume of new hires. If the unit cost of recruiting new hires for nonsupervisory, routinized jobs is low (reflecting the absence of significant match capital), reasonable levels of e reduce S/M modeling to the market-centric baseline. Positing e* to be aggregate recruiter efficiency, A to be labor productivity, and N to be production workers, the hardline search-theoretic set-up is X=AN*, N*=N+R, e*R=N. It follows that W/P=(1‒1/e*)A, where W and P respectively denote the nominal wage and product price. Reasonable recruitment efficiency implies the outcome is the textbook market-centric result (W/P=A). Recruitment importance goes away.

Nash bargain. S/M theorists frequently assign Nash bargaining a central place in their analyses, a model-building choice that is wholly rooted Ptolemaic convenience. Strong assumptions are required for Nash bargaining to yield a determinant wage; especially problematic is that labor pricing must be downward flexible at cyclical frequencies. Nash cannot suppress wage recontracting and, by varying the hire-wage for similar jobs, creates an unmanageable wage structure. (Chapter 8) No large firm uses such wage setting, making clear instead their reluctance to cut nominal wages.

Here’s a good question. Why are mainstream New Keynesians so pleased about replacing the Early Keynesian rigid-wage labor market, which has stabilization relevance, with the modern version of search theory that cannot accommodate recognizable cyclical behavior in employment and output?

Evidence. Modern search theory does not come close to fitting stabilization-relevant evidence. Economists know that. How can the available data be explained by the behavior of voluntary unemployment when most cyclical variation is obviously caused by involuntary job loss? Why do our most influential macro theorists assert that forced layoffs, the most critical business-cycle phenomenon, do not exist? The answer: The Ptolemaic urge to protect reputations and to avoid confronting the decades-long misleading, badly misleading, of graduate students is very powerful.

Blog Type: New Keynesians Chicago, Illinois

Write a Comment

Your email address will not be published.