Natural Rate of Unemployment

This post is the third in a series on how the GEM Project enhances mainstream models of labor-market behavior in highly specialized economies. Today I use generalized-exchange theory to take a fresh look at the natural rate of unemployment (NRU), broadly recognized as a central contribution to modern macroeconomics originally associated with Edmund Phelps and Milton Friedman.

Analyses of the natural rate of unemployment (U^N) almost always cite Friedman’s (1968) iconic definition: “… the level that would be ground out by the Walrasian system of general equilibrium equations provided there is embedded in them the actual structural characteristics of the labor and commodity markets, including market imperfections, stochastic variability in demands and supplies, the cost of gathering information about job vacancies and labor availabilities, the costs of mobility and so on.” Friedman further argued that U(t)=U^N implies price-inflation stability, making his natural rate equivalent to the more operational nonaccelerating inflation rate of unemployment (NAIRU). Later on came the strong-attractor hypothesis asserting that actual unemployment, once deviating from U^N, quickly returns to it.

The GEM Project enriches Friedman’s definition by dropping the restriction that NRU be “ground out by the Walrasian system of general equilibrium equations”, replacing it with the more powerful generalized-exchange framework. A substantial share of existing models in the mainstream literature, especially those dealing with labor pricing and allocation, are enhanced by the intuitive generalization of rational exchange from the marketplace to the large, highly specialized workplace.

Enriched NRU

Exchange generalization uniquely identifies a sub-class (denoted by U^T) of rational reservation-wage joblessness (U^R) that is associated with the two-way, inter-venue labor transfer required by the existence of continuous-equilibrium meaningful wage rigidity (MWR)  and the consequent rationing of rent-paying large-establishment venue (LEV) jobs. (Chapter 4) It was emphasized last week that the Harris-Todaro transfer mechanism occupies a central place in coherent, stabilization-relevant analysis. In it, the labor-market search decision to join the good-job-seeking queue (U^R) are governed by perceived LEV employment growth and wage-rent size, a process that is independent of match technology.

The generalization of exchange expands continuous-equilibrium search unemployment beyond the frictional joblessness (U^F) that results from the time required to match job vacancies and job seekers. Two-venue general-equilibrium (TVGE) macrodynamics assigns a critical role to reservation wages (W^R) in the timepath of non-frictional joblessness. A broad class of job seekers, denoted by U^R(t)=U(t)−Ụ^F(t), is restricted by Wⁿ≥W^R>W^m. The prevailing wage in LEV employment, jobs for which the unemployed worker is qualified but cannot obtain, is greater than or equal to the reservation wage, implying that U^R is involuntary despite W^R>W^m. Much the same point was made by Patinkin (1956, chapter 13).

In generalized-exchange modeling, U^R is usefully divided into four interrelated components. First, denoted by U^L (U^L<U^R), are workers laid-off with recall who are inhibited by their reservation wage from taking a small-establishment venue (SEV) job while waiting to return to their rent-paying LEV jobs. Second, U^P (U^P<U^R) represents permanent job losers who, reflecting their preferences for leisure over consumption or imperfect information about their true market opportunity costs, refuse low-wage employment. Third is the class (U^Y<U^R) that results from income-replacement programs triggered by job loss. U^Y is understood to be especially sensitive to support-program duration and generosity. In TVGE modeling, rational permanent job losers who exhaust income-replacement eligibility either join the LEV queue (U^T, the fourth class of non-frictional joblessness considered above) or exit unemployment by taking a SEV job or dropping out the labor force.

TVGE modeling, using labor homogeneity to eliminate the structural class of joblessness associated with skill mismatches, motivates an improved definition of the natural rate of unemployment: The compact natural rate of unemployment (U^N) is comprised of the frictionally jobless (U^F) and those who rationally participate in the LEV rationed-jobs queue (U^T).  Discretionary aggregate-demand intervention, which in the GEM model class rationally targets LEV employment, production, and profits, cannot effectively reduce either component of the compact natural rate (U^N=U^F+U^R). William Fellner (1976, pp. 51-52) got it right: “What is needed is a distinction between unemployment that can and unemployment that cannot be reduced by expansionary demand policies over a reasonable time horizon.” (Chapters 4 and 6)

The name “compact natural rate” is used to indicate that, in practical applications, U^N must also include structural mismatches, including the skill requirements of job vacancies relative to the human capital of unemployed workers, which in the introductory TVGE model have been suppressed by the assumption of point-of-hire labor homogeneity. More generally, GEM modeling with its continuous-equilibrium rationed jobs and hours in LEV firms implies a wholesale repeal of Keynes’s second classical postulate, pushing almost the entire labor force out labor-market, but not decision-rule, equilibrium.

Summary

Generalized-exchange modeling enriches the natural rate of unemployment with Harris-Todaro optimal labor transfer. U^N remains a notional variable, around which actual unemployment fluctuates, that is invariant with respect to shifts in aggregate nominal demand. Its operational role is the inform the trend real-side objective of stabilization authorities. Other functions assigned to NRU in the mainstream macro literature are much more problematic. In rebuilding the natural-rate model to accommodate MWR, the strong-attractor hypothesis is rejected, permitting U(t)>U^N(t) persistence that is characteristic of depression and stagnation. Further-more, it is now understood that the behavior of product-price inflation is loosely responsive to labor-market conditions, leading to rejection of the operational equivalence of NRU and NAIRU.

In the GEM model class, the central employment-stabilization role assigned to the management of nominal demand requires support by state-of-the-art modeling of the natural rate (U^N) as well as the various components of Okun’s law and nonstationary productivity growth. Such analysis must be a core competency of a responsible central bank, joining the careful analysis of price inflation relative to its established regime. The skill set associated with the real-side demand management has been deemphasized by mainstream market-centric theorists to the detriment of stabilization-authority effectiveness.

Blog Type: Wonkish San Miguel de Allende, Mexico