The thesis that runs through the GEM Project is that macroeconomics, consistent with the neoclassical tenets of optimization and equilibrium, can powerfully explain instability if rational price-mediated exchange is extended from the marketplace to information-challenged workplaces. In that context, Neil Irwin’s know-nothing approach to macro problems is shown to be an artifact of poorly constructed theory, not an inherent outcome of an incomprehensible macroeconomy.
The Irwin approach is reckless and requires careful rebuttal. Four posts, including today’s, will be devoted to outlining rigorous macro theory that easily explains the instances of instability about which Times reporters have recently been writing. In particular, the Project’s generalization of exchange provides an intuitive platform for the melding of inconsistent labor pricing rooted in industry heterogeneity, which turns out to be fundamental to understanding modern, highly specialized macroeconomies. What follows briefly summarizes the relevant macro literature on wage determination, which is key to understanding inflation and unemployment. The review sets the stage for the subsequent three posts.
Labor-pricing inconsistency, which turns out to be a condition of downward wage rigidity and stabilization-relevant macroeconomics, is an empirical fact of highly specialized economies. Since before the 1930s depression and the establishment of macro theory as a distinct field of study, aggregate labor modeling has been conducted within the general-market-equilibrium (GME) framework that problematically suppresses wage inconsistency. Working uneasily within that market-centric model class, Modigliani (1944) focused postwar macro thinking on nominal labor-price rigidity, which he simply assumed. Early Keynesians (EK) subsequently detached their analysis from that crude restriction with three innovations. Market frictions were posited that slowed the adjustment of nominal wages to changes in unemployment. Phillips’ 1958 equation relating money wage change and joblessness was augmented with catch-up to product-price inflation that had already occurred. The Phelps-Friedman treatment of the natural rate of unemployment was incorporated into the mainstream model.
The inflation- and friction-augmented Phillips curve, providing a good fit to the available data, became the keystone for EK macro theory:
where lower-case variables denote rates of change, U is the jobless rate, UN is unemployment’s natural rate, and p the product-price inflation over the catch-up period (ķ) at time t. That Phillips equation is foundational to Samuelson’s Neoclassical Synthesis, in which slowly adjusting nominal wages in the short-term arbitrarily morph into GME labor pricing at some point in the longer-run. Paul Samuelson later called the Neoclassical Synthesis, which he organized, “Model-T Keynesianism”.
Once attached to mainstream market-centric modeling, the equation soon became the single most controversial relationship in macro analysis. Theorists, who relied on the EK Phillips curve to provide some stabilization-relevance, had surrendered the justifiably prized micro-coherence of longstanding neoclassical macro analysis. They became an inviting target for reeducation during the New Neoclassical counter-revolution that gathered irresistible momentum during the stagflation decade that began in the early 1970s.
The neoclassical reconstruction of nominal wage dynamics culminated in 1980s with the broad acceptance of the rational-expectations Phillips curve, rooted in the seminal work of Lucas (1972). That version, strongly implying the centrality of expectations in price inflation, has endured as a central building block of mainstream macro theory:
Et is the expectations operator conditional on agents’ cost-effective use of information available at time t.
Subsequent posts demonstrate, from the vantage point of generalized rational exchange, that the expectations-augmented Phillips Curve is unacceptably flawed. The analysis has two parts. The first investigates the periodic adjustment of wages for inflation, a misunderstood process. Much of the early Phillips-curve controversy resulted from the inclusion of product prices as an explanatory variable, the sluggish specification of which was used to generate same-direction movement from nominal disturbances to employment and output. Careful modeling of wage-price dynamics recalibrates the significance of the Lucas revolution. In particular, it rejects his rejection of the rationality of wage catch-up to product-price inflation that has already occurred. It turns out that the use of expectations in the explanation of periodic wage changes is irrational. The second part looks closely at the Phillips curve itself. That analysis, relying on meaningful wage rigidity (MWR) and pure wage rent (PWR), microfounded in the GEM Project, moves the stabilization-relevant focus away from the relationship between wages and inflation to wages and unemployment.
Indeed, the original Phillips problem was the nature of the relation between wage change and unemployment, the cyclical behavior of which is largely concentrated in involuntary job loss. Resolution of fundamental questions about that relationship informs our understanding of the discretionary management of total spending. 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.”
Many mainstream theorists are exhausted after the 30-year microfoundations war that the Lucas army won. As a prominent macroeconomist confidently, and incorrectly, concluded, “… the debate [over expectations versus catch-up] has to move to the data and can only be settled through empirical work.” (Correspondence (2007) from Olivier Blanchard to the author.) Perhaps reflecting the general exhaustion, David Romer (2001, p.251) famously offered a free-parameter version of nominal wage dynamics as the “natural compromise” between the catch-up and expectations:
w(t)=ao+a1(UN–U(t))+(1–ψ)pķ(t)+ψEtp(t+1), s.t. ψ ∈ [0,1].
ψ governs the relative contributions of expectations and catch-up to nominal wage change. Romer illustrates exhaustion by relegating ψ to theoretical indeterminacy. It cannot be surprising that his guesswork differs the GEM Phillips curve (see below), which is rooted in rational behavior. It is noteworthy here that Rudd and Whelan (2005) tested the “natural compromise” hybrid approach, finding little support for an important role played by forward-looking adjustments. (Recall that Rudd is featured in the most recent of Neil Irwin’s know-nothing articles.)
It is surprising that relying on empirical research to specify ψ (and, therefore, the policy-relevant nature of wage-price dynamics) became acceptable in the aftermath of the vigorous rejection of EK free parameters used to suppress wage recontracting. After the 1970s stagflation dramatically demonstrated the sensitivity of contemporaneous Phillips-curve estimates to particular time periods (indicating model misspecification), that rejection became a rallying cry for the neoclassical market-centric reconstruction of macroeconomics.
Reviving the quest for microfoundations, the GEM Project has sorted out the optimal roles of catch-up and expectations, making a fundamental contribution to the theoretic infrastructure supporting macroeconometric wage modeling. The findings will be embedded in the continuous-equilibrium Phillips relation constructed in the next two posts. They must be embedded in any policy-relevant model of wage-price dynamics that purports to be consistent with rational behavior.
Blog Type: Policy/Topical Chicago, Illinois