Stagflation, 1970s

Stagflation refers to the persistent co-occurrence of high inflation, high unemployment, and weak or negative growth in the industrial economies through the 1970s. Two oil shocks (1973, 1979), the collapse of Bretton Woods (1971–1973), and the political pressure on central banks to maintain low unemployment combined to produce conditions the postwar Phillips-Curve framework had described as theoretically scarce. The Volcker disinflation (1979–1982) ended the episode at the cost of the deepest postwar US recession to that point.

The event is the central acid test of competing macroeconomic frameworks in the postwar era. No tradition emerges from the decade with its reputation entirely intact, but the differential survival is consequential: Old Keynesian reputations broke against it; Monetarist and New Classical reputations were made by it; Austrian accounts received their fairest postwar hearing; Post-Keynesian cost-push and conflict-inflation frameworks gained ground that the prior two decades had denied them.

What happened

US headline CPI inflation rose from below 2 percent in 1965 to 6 percent by 1970, briefly fell under Nixon’s 1971–1974 wage-and-price controls, then accelerated to 12 percent by 1974, fell to 5–6 percent in the late 1970s, and rose again to over 13 percent by early 1980. Unemployment, which had averaged 4–5 percent through the 1960s, rose to 5.5–6 percent in the early 1970s and to 9 percent in 1975 — and rose again to 10.8 percent by late 1982 during the Volcker disinflation. Real GDP growth slowed from a 4 percent postwar trend to under 3 percent.

Two large supply shocks structured the decade. The Yom Kippur War in October 1973 triggered an OPEC oil embargo and a fourfold rise in the dollar oil price; the Iranian revolution in 1979 produced a second roughly threefold rise. Each shock raised headline inflation directly through energy prices and indirectly through wage-price feedback in indexed labour contracts. Both occurred against a background of fiscal deficit pressure from Vietnam and Great Society spending, and against a Federal Reserve under Arthur Burns (1970–1978) widely judged ex post to have been too accommodative.

The Bretton Woods system of fixed exchange rates, which had constrained US monetary policy through the 1960s, ended in stages between 1971 (Nixon closes the gold window) and 1973 (final transition to floating rates). The constraint that ended was a real one; the inflation that followed was not solely caused by its absence, but its absence permitted a degree of monetary accommodation that the prior regime had not.

The Old Keynesian reading

The dominant framework of the 1960s read the Phillips Curve as a stable empirical relationship between inflation and unemployment — a “policy menu” from which a society chose a position based on its preferences. Samuelson and Solow’s 1960 paper had explicitly framed it this way. On this reading, the 1970s presented a deep puzzle: inflation and unemployment rose together, which the menu framework said was approximately impossible.

The Old Keynesian response was to attribute the joint movement to supply shocks (the oil shocks) and to argue that the Phillips Curve had been temporarily shifted by supply-side factors but remained intact in its underlying structure. James Tobin’s 1972 AEA presidential address is the most articulate version of this defence. In its sophisticated form, the Old Keynesian framework remained capable of accommodating the data — but its capacity to do so came at the cost of the framework’s predictive sharpness, and it failed to anticipate the persistence of inflation through periods when supply pressures had eased.

By the late 1970s, no graduate program teaching from the unmodified postwar synthesis was producing students who could explain the contemporaneous data. The framework’s reputation broke not because it was straightforwardly falsified but because it had become unable to predict in advance the events it was reduced to retrospectively rationalising.

The Monetarist reading

Friedman and Schwartz’s 1963 Monetary History and Friedman’s 1968 AEA address had pre-committed monetarism to a position that read the 1970s nearly to specification. The Phillips Curve as the postwar mainstream had drawn it was short-run and expectational; sustained attempts to hold unemployment below the natural rate would produce accelerating inflation; the Fed’s accommodative monetary stance under Burns was effectively such an attempt; the joint rise in inflation and unemployment was therefore not anomalous but predicted.

Monetarism’s credit on the 1970s comes principally from the ex ante nature of this prediction. The 1968 address was published with US unemployment at 3.6 percent, the lowest in a generation; the framework anticipated the failure of that combination to be sustainable. The Volcker disinflation of 1979–1982 was conducted under self-described monetarist operating procedures (non-borrowed-reserve targeting), and the inflation collapse it produced was in the broad terms Friedman framed a successful real-time test.

The harder questions for monetarism are (i) whether the disinflation cost in lost output was within the range Friedman’s framework had publicly suggested (it was higher), and (ii) whether the post-1982 disconnect between standard monetary aggregates and nominal output retroactively undermines the 1970s reading. Neither question dislodges the central monetarist credit on this episode.

The New Classical reading

Lucas, Sargent, Wallace, and Barro extended Friedman’s natural-rate framework through the rational-expectations apparatus. Where Friedman’s expectations were adaptive (agents revise upward in response to inflation surprises), New Classical expectations were rational (agents anticipate the inflationary consequences of accommodative monetary policy and front-run them). On this reading, the postwar Phillips Curve was an artefact of an unusually stable monetary regime that had produced unusually well-anchored expectations; once the regime shifted, the Curve dissolved.

The Lucas critique (1976) gave this argument its sharpest methodological form: any econometric model estimated over a period of stable policy could not be used to evaluate a change in policy, because agents’ decision rules — and therefore the model’s parameters — would shift with the policy regime. Old Keynesian Phillips-Curve estimates were not just wrong about the 1970s; they were wrong in principle, because they had treated as structural a relationship that was itself a function of policy.

The Volcker disinflation tested the New Classical strong claim that a credible regime change could disinflate without significant output cost. The output cost turned out to be significant; the strong claim was disconfirmed; the methodological apparatus survived and was absorbed into the New Keynesian successor program.

The Austrian reading

The Austrian framework reads stagflation as the predictable consequence of credit expansion. The 1960s pumping of the money supply through Vietnam-era deficits produced a malinvestment boom; the 1970s inflation and the recessions that punctuated the decade were the unwinding of that boom. The oil shocks were proximate triggers but not fundamental causes. The decade’s outcome — high inflation requiring a sharp recession to break — was, on this reading, what credit expansion always eventually produces.

The Austrian view was substantially excluded from postwar mainstream macroeconomics, partly on theoretical grounds (its capital-theoretic apparatus did not translate into the formal models the field had adopted) and partly on cultural ones. The 1970s gave it its fairest postwar hearing — Hayek’s 1974 Nobel was awarded in this period — and the broadly market-liberal turn of the late 1970s and early 1980s drew on Austrian rhetoric even where it adopted Friedman’s substantive framework rather than Mises’s or Hayek’s.

The Post-Keynesian reading

Post-Keynesians read the 1970s through cost-push and distributional-conflict frameworks rather than monetary ones. Wage-price spirals driven by labour-management conflict, the OPEC supply shock, and the breakdown of postwar incomes-policy norms produced an inflation that was not fundamentally about money supply and was not responsive to demand contraction in the way the monetarist framework predicted. The Volcker disinflation worked, on this reading, principally by breaking labour bargaining power through unemployment — an effect monetarist accounting underweights and Post-Keynesian accounting foregrounds.

The framework’s strongest empirical claim is that inflation in the 1970s tracked oil prices and labour-market conflict more closely than it tracked any standard monetary aggregate. The framework’s weakest is that it provides little forward-looking guidance: it identifies many proximate sources of inflationary pressure but no clear mechanism for resolving them short of recession. The 1980s and 1990s, in which inflation was substantially controlled without the labour-conflict dynamics returning, were not a vindication of the Post-Keynesian framework as a positive account of inflation.

The Volcker disinflation

Paul Volcker’s appointment as Federal Reserve Chair in August 1979 and the October 6, 1979 announcement of a switch to non-borrowed-reserve targeting marked the operational beginning of the disinflation. The Federal Funds rate rose from approximately 11 percent to a peak of 19 percent by mid-1981; unemployment rose from 5.6 percent in 1979 to 10.8 percent by late 1982; CPI inflation fell from 13.5 percent in 1980 to 3.2 percent by 1983.

The episode is the single most consequential test of competing macroeconomic frameworks in the postwar era. The disinflation worked: it succeeded in reducing inflation expectations sufficiently that subsequent decades operated against a much lower inflation backdrop. It also cost more in output than New Classical strong-claim accounts had suggested, but no more than Old Keynesian sticky-expectations accounts predicted. It produced no clean victory for any tradition but a substantial reweighting of the field around the natural-rate framework, regime credibility, and central-bank independence — themes that would structure the Great Moderation of 1985–2007.

Retrospectives

By the 1990s, the consensus reading of the 1970s had stabilised around something like the New Keynesian synthesis: the Phillips Curve is short-run and expectational; the natural rate is real but harder to estimate ex ante than monetarism had suggested; the central bank can disinflate but at a cost that depends on credibility; rule-based, independent central banks do this work better than discretionary ones. This synthesis is, in the broad strokes, a Friedman-Phelps natural-rate framework with rational-expectations methodology and Old Keynesian sticky prices — i.e., a settlement between the three traditions most engaged in the original episode.

The synthesis held through the Great Moderation and was substantially shaken by the 2008 crisis, which revealed financial-system mechanics the framework had abstracted from. The 2021–2023 inflation episode, in which inflation rose and fell in patterns that did not match standard New Keynesian Phillips-Curve estimates, produced renewed engagement with supply-side and Post-Keynesian conflict-inflation frameworks. Whether the 1970s synthesis survives that engagement is open.

Unresolved questions

• How much of 1970s inflation was monetary vs supply-driven? The decompositions vary; the monetarist and Post-Keynesian frameworks weight them differently and continue to disagree.
• What was the cost of disinflation? Old Keynesian estimates of the sacrifice ratio in 1979–1982 turned out closer to the realised cost than New Classical estimates. The methodological lesson — that credibility cannot eliminate the cost of disinflation — is broadly accepted.
• Was the Burns Fed’s accommodation ideological or institutional? Memoir and minutes evidence cuts both ways. The institutional reading (Burns faced political pressure structurally similar to what any Fed Chair would face) underwrites the modern case for central-bank independence; the ideological reading (Burns personally believed in a higher tolerable inflation rate than the consensus) underwrites the case for personnel.
• How much do the 2021–2023 inflation episode and the 1970s share, and how much was the recent episode genuinely new? This is contested in real time and will be one of the major macroeconomic debates of the 2020s.