Two Models of Stagflation, One Useful, One Not

Mainstream FGME Analysis

Mainstream friction-augmented general-market-equilibrium theory cannot  explain the massive market failure that occurred during the stagflation decade. That episode featured simultaneous high unemployment and a powerful price-wage spiral that broke apart what had been a remarkably stable interindustry wage structure. We know that involuntary job loss was the engine of rising unemployment in the stagflation recessions of 1974-75, 1980, and 1981-82. But, given that forced job loss cannot coherently exist in FGME modeling, the consensus story ignores actual employment behavior that accompanied the persisting instability. It is the familiar macro conundrum produced by market-centric thinking. Wage recontracting must, but cannot, be rationally suppressed. Playing by the rules governing the New Neoclassical Synthesis, modern New Keynesian (NK) theorists have always had less capacity than discredited Early Keynesians (EK) to elucidate, albeit imperfectly, the unemployment part of the stagflation story.

The second central part of stagflation mechanics, the powerful price-wage-price spiral, is also out-of-reach for market-centric thinking that, by definition, pays no attention to information-challenged workplaces. If we are to understand stagflation, we must first understand how wages adjust for inflation. We must understand that rational workplace behavior mandates the use of catch-up to price change that has already occurred, providing a necessary tool to model the 1970s nominal spiral. Generalized-exchange mechanics, and the nominal persistence they enable, are not available to modern theorists who restrict wage adjustments for inflation to anticipated future price change. The hard fact is that exclusive reliance on rational expectations is irreconcilable with rational behavior.

Market-centricity denies employees the means to compel employers to increase wages in the circumstances of unemployment that exceeds its natural rate. Stagflation’s characteristic price-wage-price spiral, the nature of which was closely documented in my The Price of Industrial Labor (1984), cannot exist in FGME modeling. Mainstream theorists must disregard the fact that only half of all workers were able to defend their real wages against the 1970s adverse terms-of-trade shifts, that those workers were concentrated in information-challenged workplaces, and that the dispersion of interindustry wages – notably stable during the previous two decades – increased sharply. (See next week’s post.)

Simply put, FGME modeling is not adequate to the task of explaining stagflation. It does not come close. In the mainstream story, relative increases in energy prices cannot show up in higher wages; but they did. The transmission of damaged central-bank inflation credibility to higher wages cannot coherently produce involuntary job loss, but forced employment separation occurred in the millions. Rational expectations of price inflation, by their nature, cannot distort relative wages, yet the interindustry wage structure blew apart. The mainstream identification of stagflation’s causal role to be the lack of central-bank low-inflation credibility is more than wrong; the monetary-policymaking advice it provides it extraordinarily reckless.

In the real world, simply establishing and maintaining low inflation does not imply employment stabilization. The inadequacy of a single, low-inflation objective was most vividly illustrated in the virulent instability in 2008-09. In that crisis, effective action required direct policymaker focus on real-side objectives, the determinants of which were deteriorating dramatically. Working indirectly through an emphasis on stable inflation would have provided woefully insufficient information to be a useful guide to central-bank action.

GEM Project to the Rescue

The GEM Project’s generalized-exchange modeling easily explains the salient features of the stagflation decade, providing consistency with a broad range of evidence that is far beyond FGME thinking. The narrative begins, as it should, with the outsized commodity-price increases that greatly disturbed the early 1970s economy. The Workplace-Marketplace Synthesis, via its keystone meaningful wage rigidity (MWR), easily translates that shock into the virulent price-wage spiral that eventually produced two classes of involuntary job loss: temporary layoffs resulting from high-frequency nominal demand fluctuations and, later, permanent job downsizing associated with real wage-rent increases. (See next week.) Both effects work through weakening profits, over the business cycle for the former and via trend expectations for the latter.

In the GEM stagflation story, layoffs are the familiar outcome of interaction of rational MWR and slowdowns in total spending. Weakened demand centrally resulted from intermittent central-bank tightening of credit conditions, as it attempted to contain high inflation. The elevated incidence of temporary job loss was the principal cause of relatively high joblessness during the 1970s and early 1980s – a period that experienced three recessions, two of which were among the most severe postwar downturns. Meanwhile, in the time-intensive process required by nonstationary reference-wage recalibration, permanent job-downsizing was exacerbated by rising wage rents as workers in information-challenged workplaces resisted terms-of-trade shifts. To the extent that higher unit costs could not be passed on to purchasers of goods and services, expectations of nonstationary pure profit were damaged and eventually became unable to support existing levels of productive capacity in a range of industries. Vulnerable rent-receiving jobs were eventually and rationally downsized, feeding into long-duration joblessness.

The stagflation-decade continuous-equilibrium employment and wage paths predicted by generalized-exchange model fit a broad array of the relevant evidence. GEM modeling can, and does, rationally explain the breakdown in wage recontracting necessary to produce the many millions of involuntary job losers that actually occurred. Additionally, the two-venue theory can, and does, rationally explain the blowing apart of the interindustry wage structure that introduced a critical structural cost component into the trend adjustment process. Given that stagflation mechanics are rooted in information-challenged workplaces, consensus market-centric modeling lacks the capacity to explain consequent important phenomenon. Its policy conclusion is an empty slogan masquerading as a serious analytic conclusion.

Unlike its mainstream counterpart, the coherent generalized-exchange model has no need to suppress available evidence. If the generalized-exchange model had existed at the time of the early-1970s terms-of-trade shifts, the subsequent stagflation would have been easily predictable. Indeed, the two-venue analysis would have further demonstrated its power by also making sense out of the most under-appreciated part of the stagflation story. What happened next is considered in more detail next week.

Blog Type: Wonkish Saint Joseph, Michigan