The decade-long stagflation crisis that began in the early 1970s was a very difficult time. Its economic loss was, in the 20th century, second only to the 1930s Great Depression. Stagflation refers to high product-price inflation occurring simultaneously with high unemployment. Today, mainstream theorists cite the extraordinary episode as evidence supporting their assertion that low trend inflation should be the primary objective of stabilization policymakers. This post examines why that advice has always been mistaken.
Mainstream analysis. Roger Farmer (2010, p.60) concisely summarized the modern consensus understanding of stagflation: “During the 1970s, the U.S. economy experienced high inflation and high unemployment at the same time and the data did not lie anywhere near the Phillips curve…. The Phillips curve broke down because firms and workers began to increase wages and prices in an inflationary spiral. Wages went up because workers believed that prices would rise. Prices went up because higher wages were passed on to consumers.” The narrative is motivated by a collapse in the credibility of the central bank’s commitment to low trend inflation. Macro theorists have long argued that a convincing Fed commitment to its nominal objective would have prevented the huge welfare loss of the stagflation decade.
That conclusion is consistent with consensus market-centric, general-equilibrium theory. Decades ago, combining innovations from Friedman, Phelps, and Lucas, the macro mainstream settled on a reduced-form wage equation, named the rational-expectations Phillips curve, that they asserted had been derived from Walrasian equilibrium:
w(t) = ao+a1 (UN-U(t))+Etp(t+1),
where w denotes nominal aggregate wage change, UN is the natural rate of unemployment, U is actual joblessness, p represents the rate of change in the price index, and E denotes expectations rationally constructed on the cost-effective use of available information.
Unfortunately, as readers of the GEM Blog know, the mainstream market-centric model class cannot rationally suppress wage recontracting, cannot rationally accommodate involuntary job loss, and cannot rationally motivate causality from nominal demand disturbances to recognizable movements in output, employment, and income. As a result, the consensus Phillips Curve is of little use in explaining actual macro instability. It has unsurprisingly misled investigations of the causes and consequences of the stagflation decade.
Stagflation facts. An extraordinary concentration of labor-adverse terms-of-trade shocks occurred in the United States in the early 1970s:
- The quadrupling of oil prices associated with the OPEC embargo in 1973 was the largest single shock. By the end of the decade, the cartel helped engineer more than a fifteen-fold petroleum price increase. Since energy costs directly accounted for about a tenth of the total consumer price index during this period, many nominal wages jumped as the higher fuel prices fed through the economy. Real wages, in terms of the prices of the goods and services that most workers produced, increased significantly in large or unionized establishments.
- Food prices also jumped in the early 1970s, putting further upward pressure on wages. The 1972 Russian crop failure put substantial pressure on world grain markets. Meanwhile, there was a mysterious collapse in the anchovy catch off Peru; and meat prices jumped as animal-feed (produced from grain or fishmeal) costs rose sharply.
- The over-valued dollar of the 1960s became unsustainable, and the subsequent depreciation and higher import prices further fed the price-wage spiral. The gold window was closed in mid-1971 as a prelude to the Smithsonian agreement, which realigned dollar exchange rates. That change was followed, in 1973, by two additional dollar devaluations.
- Wage-price controls were imposed in the United States 1971, altering the time distribution of inflation and intensifying the price-wage spiral.
At the beginning of the 1970s, the U.S. interindustry wage structure was relatively compact; but that quickly changed as roughly half of all employees received cost-of-living adjustments that protected them from the adverse terms-of-trade shifts. A careful look at the relevant evidence makes clear that wages, on average, went up initially because of the early-1970s price shocks and continued, as higher labor costs more broadly pushed up product prices. The U.S. experienced a full-blown price-wage-price spiral that market-centric general-equilibrium theory couldn’t begin to explain. Its failure was no secret. Everybody familiar with the evidence, including policymakers at the Fed, knew that the mainstream stagflation story was bogus. Wages did not rise because “workers believed that prices would rise”. The wage-price spiral occurred because prices actually rose, causing a deterioration in living standards against which roughly half of all workers were able to protect themselves.
The consensus expectations narrative used to explain the stagflation decade is irreducibly inconsistent with the sharply increased inter-industry wage dispersion that actually occurred. Think about it. The altered wage structure implies that about half of all employees experienced sharply higher wages while half did not. The evidence also shows that the lucky workers were concentrated in the large-establishment venue. (See Annable (1984).) Why would large-establishment, but not small-establishment, employees expect higher inflation. There never has been, and never can be, a market-centric explanation for that critical evidence, uncovered at the Federal Reserve in the 1970s, closely analyzed in the early 1980s, and conveniently ignored by the macro academy ever since.
The GEM Project comes to the rescue. The generalization of rational exchange from the marketplace to bureaucratic workplaces enables the construction of a stabilization-relevant micro-coherent Phillips Curve. It is critically shown to require a set of assumptions: the terms of trade remain unchanged, as do small-firm productivity growth, relative sector size, the natural rate of unemployment, and government wage regulation. The simplified (inherently short-term) aggregate wage equation yields the Phillips curve produced by the Project’s marketplace-workplace model class:
where Ł is the optimal price-inflation lag structure (t-k to t) and pM denotes the central bank’s inflation objective.
The more powerful model features inflation catch-up (familiar in the practitioner literature but absent in mainstream economic analysis), while reducing the role of rational expectations to anticipated changes in the central bank’s inflation target. This version of the Phillips relation is fully microfounded, while its rational-expectations predecessor has been shown in the GEM Blog to be inconsistent with optimizing behavior. The Website’s E-book, Chapters 4 and 5, usefully elaborates on the nature and use of the stabilization-relevant version of the Phillips relation.
Policy implications. In the stagflation-decade circumstances, the complex task of stabilization authorities was to contain inflation without setting off a collapse in investor-lender credibility of its real-side objective, i.e., high trend employment and output. Modeling the mechanics of how contracting demand motivates extreme instability and depression is an important innovation of the Project, most recently summarized in the GEM Blog four weeks ago. Absent an unchecked total-spending collapse, the end phase of navigating difficult stagflation macrodynamics results when rising large-establishment labor rents push down expected trend profits sufficiently to trigger firm downsizing. Associated job destruction induces rational recalibration of employee reference standards and consequent wage givebacks. Substantial job downsizing was characteristic of the 1980s. (The E-book’s Chapter 5 provides elaboration on the predictable period of downsizing that still remains a one-off mystery to mainstream theorists and their market-centric models.)
Here is the central policymaker problem during the stagflation decade. If the central bank responded to 1970s chronic shocks by preventing the powerful price-wage-price spiral from translating into some higher aggregate price inflation, the result would have been a huge jump in involuntary job loss as output and income drastically contracted. It is a good bet that such an unprecedented collapse would have generated sufficient loss of central-bank credibility with respect to its real-side objective to induce widespread inaction among investors and lenders, rationally waiting for a credible bottom to collapsing asset prices, employment, and output. In generalized-exchange modeling as well as the real world, such unchecked inaction feeds downward spiraling total spending and, ultimately, the unimaginable welfare loss of a 1930s-class depression.
Stabilization authorities confronting stagflation had contain inflation without setting off a collapse in their real-side credibility. The much maligned Arthur Burns, deeply schooled in the 1930s depression, recognized that essential task. Given my job heading up stagflation analysis at the Fed, I was privy to his thinking. He convinced the Federal Reserve’s governing monetary-policy body to have the patience required to accomplish its task without generating a depression. We got to stagflation’s downsizing end-phase absent an unchecked collapse in total spending, and Burns’ greatest fear was avoided. It discredits mainstream macro theorists that their Ptolemaic defense of market-centric modeling has caused them to bury the important lessons of the stagflation crisis.
Blog Type: Wonkish Chicago, Illinois