The Conquest of Inflation: An Introduction to Sargent's Analysis
📄 Summarized from the full manuscript · Human-reviewed for faithfulness before publication
In brief
Have central bankers actually learned to control inflation, or did the world just get lucky? This short essay introduces Thomas Sargent's "The Conquest of American Inflation" to a central-bank audience. It lays out two familiar accounts of postwar inflation's rise and fall -- a deliberate anti-inflation reorientation, or an unchanged policy riding a falling natural rate of unemployment -- then previews Sargent's third: the Fed continually re-estimates the inflation-unemployment tradeoff from recent data, and because policy shapes that estimate, low inflation is self-reinforcing only while the tradeoff looks unfavorable. Fresh disturbances could revive inflation, a warning the essay argues fits the U.S., with its weaker anti-inflation mandate, better than Europe.
What this paper finds — and why it matters
Introducing Thomas Sargent’s monograph “The Conquest of American Inflation” to a central-bank audience, this short essay frames the postwar rise and fall of OECD inflation as a test case for whether “central bankers have learned to control inflation, or… the current situation of low and stable inflation [is] simply a result of a series of favorable but temporary factors.” It lays out the standard Kydland-Prescott/Barro-Gordon time-inconsistency account of why inflation rose – a central bank tempted to exploit an inflation-unemployment tradeoff, rational private expectations that anticipate this temptation, and the impossibility of binding commitment, yielding an inefficiently high-inflation “Nash outcome” – and the two “traditional” competing explanations for why it later fell: either a deliberate institutional reorientation toward low-inflation mandates (the “Ramsey outcome,” more plausible for Europe’s newly independent, price-stability-mandated central banks) or, per Peter Ireland (1999), an unchanged policy simply riding a series of favorable shocks that lowered the natural rate of unemployment (offered as the more plausible account for the U.S., where institutional reforms were more limited). The essay then previews Sargent’s own, third account: the central bank behaves under similar assumptions to Kydland-Prescott and Barro-Gordon but has incomplete knowledge of the true Phillips curve, continually updating its beliefs from recent data and thereby making its own beliefs partly self-fulfilling, since the estimated tradeoff itself shifts as policy shifts. On this reading, the Fed reduced inflation not because it concluded unemployment could never be affected, but because the recently estimated tradeoff had become “unfavorable” enough that exploiting it was not worthwhile – leaving open the possibility that new disturbances could make the tradeoff look favorable again and revive inflation. The essay stresses Sargent does not claim to have identified the correct account of U.S. policy, only that both the reorientation and adaptive-learning stories are consistent with theory and evidence, and argues the warning implicit in Sargent’s analysis is more relevant to the U.S., given its comparatively weak formal anti-inflation mandate, than to Europe.
Summary of a classic paper, AI-assisted and human-reviewed. See the linked original for the authoritative claims and full conditions.
Questions & answers
Q1. What is the central question this introductory essay poses about the fall in OECD inflation since the 1980s?
Whether the shift to low, stable inflation reflects that “central bankers have learned to control inflation,” or is “simply a result of a series of favorable but temporary factors” – a question the essay says “our expectations regarding the future depend in a fundamental way” on how we answer (p. 5). It notes inflation in Sweden and the U.K. followed “much the same development as in the United States” (Figures 1-2), suggesting Sargent’s U.S.-focused analysis “may be relevant also to other countries.”
Q2. What are the three assumptions behind the Kydland-Prescott/Barro-Gordon explanation of rising 1960s-70s inflation?
First, that central banks wanted to keep unemployment below its natural rate, exploiting nominal wage rigidities to do so; second, that wages and contracts are set under rational expectations of monetary policy, so the public comes to anticipate the temptation to inflate; third, that the central bank cannot bindingly commit to low inflation (p. 7). Together these three assumptions yield what the essay calls, following Sargent, the “Nash outcome”: “inflation on average will be inefficiently high, without any gains in the form of lower unemployment” (p. 7) – a story the essay notes “fits well with the experiences of the 1960’s and 1970’s,” since U.S. inflation began rising “soon after the Federal Reserve had been recommended to lower unemployment by accepting higher inflation” (p. 7-8).
Q3. What two “traditional” explanations does the essay offer for why inflation later fell, and which does it favor for Europe versus the U.S.?
A “complete reorientation of monetary policy” – politicians and central bankers learning to target the natural rate directly or appointing a conservative, independent central banker (citing Rogoff 1985) – versus Peter Ireland’s (1999) alternative that “the fall in U.S. inflation… can be explained without any reorientation of monetary policy, if one assumes that the economy has been hit by favorable shocks which have lowered the natural rate of unemployment” (p. 8-9). The essay judges reorientation “a reasonable interpretation of the developments, at least in Europe, where a number of institutional changes have been carried through,” but notes “in the U.S. such reforms have not been made to the same extent,” and for the U.S. “the issue seems to be less clear” (p. 8-9).
Q4. What is distinctive about Sargent’s own explanation, as previewed here?
Sargent keeps the Kydland-Prescott/Barro-Gordon behavioral assumptions but adds that the central bank has incomplete knowledge of the true Phillips curve and updates its beliefs adaptively, weighting recent data more heavily, so that “by its very behavior, the central bank makes these beliefs self-fulfilling: the relationship between inflation and unemployment depends on the conduct of monetary policy, and as policy changes, so does the estimated Phillips curve” (p. 9). On this account, disinflation happened “not because of a conviction that unemployment can not be affected, but rather because the trade-off between inflation and unemployment has recently been so ‘unfavorable’… that the Fed has abstained from trying to lower unemployment” (p. 9-10) – a reading the essay stresses is explicitly consistent with, not a refutation of, Ireland’s shock-driven story, since both operate within similar Kydland-Prescott/Barro-Gordon foundations.
Q5. Why does the essay treat Sargent’s account as a live warning rather than settled history?
Because the self-fulfilling nature of the estimated tradeoff means it can turn “unfavorable” or “favorable” again as new disturbances hit the economy: “as the economy is hit by new disturbances and the Phillips curve seems to change, this view will be revised, and inflation may well increase again” (p. 9-10). The essay is careful to note “Sargent does not claim that his theory is the correct description of Federal Reserve behavior,” only that both the reorientation and adaptive-learning stories “are consistent with economic theory, and both seem realistic” – and argues the warning is “more relevant for the U.S. than for Europe,” since explicit inflation targets and clear policy mandates have been “more heavily emphasized in Europe… than in the United States” (p. 10).
Key terms in this paper
Definitions below follow the paper's own usage.
- The "Nash outcome" (time-inconsistent inflation)
- The paper's label (via Sargent) for the equilibrium implied by the Kydland-Prescott/Barro-Gordon time-inconsistency story: because a central bank cannot bindingly commit to low inflation, "the only credible inflation target is the rate of inflation at which the central bank has no incentive to create further inflation," so "inflation on average will be inefficiently high, without any gains in the form of lower unemployment."
- The "Ramsey outcome" (reoriented, credible low-inflation policy)
- The paper's label for the alternative equilibrium reached if policymakers abandon the belief that monetary policy can affect unemployment in the long run and reorient policy accordingly (e.g., by targeting the natural rate, or appointing a conservative, independent central banker per Rogoff 1985): "politicians desert the erroneous belief that monetary policy can affect unemployment in the long run, and therefore reorient policy ('the triumph of the natural rate')," in principle reaching lower inflation "at no cost in terms of higher unemployment."
- Self-fulfilling, adaptively estimated Phillips curve (Sargent's account)
- The essay's preview of Sargent's own account: the central bank operates under Kydland-Prescott/Barro-Gordon-type assumptions but has incomplete, evolving knowledge of the true Phillips curve, updating its beliefs from incoming data while weighting recent observations more heavily; "by its very behavior, the central bank makes these beliefs self-fulfilling: the relationship between inflation and unemployment depends on the conduct of monetary policy, and as policy changes, so does the estimated Phillips curve" -- so disinflation reflects a recently "unfavorable" perceived tradeoff, not a permanent reorientation.
- Falling natural rate of unemployment as an alternative explanation
- Peter Ireland's (1999) explanation, contrasted with the reorientation story: "the fall in U.S. inflation during the 1980's and 1990's can be explained without any reorientation of monetary policy, if one assumes that the economy has been hit by favorable shocks which have lowered the natural rate of unemployment" -- so that the basic Kydland-Prescott/Barro-Gordon logic still holds and policy itself never changed, only the underlying labor-market environment.