Monetary Cooperation during Global Inflation Surges

Q1. How does a demand reallocation shock generate an inflation-unemployment tradeoff?

A shift in preferences toward tradable goods (ω₀ > ω) reduces demand for nontradable goods, causing nontradable firms to fire workers; since nominal wages are rigid, the only way to sustain full employment is through a rise in the price of tradables (P^T), which operates through three distinct channels. First, higher P^T raises tradable sector firms’ real revenue per worker (nominal wages fixed), inducing them to hire more workers and expand output — the direct labor demand channel. Second, higher P^T generates income effects: as tradable output and income rise, households increase consumption of both tradable and nontradable goods, boosting nontradable employment through the income channel. Third, higher P^T generates expenditure switching away from tradables and toward nontradables (since nontradable goods become relatively cheaper), which also sustains nontradable employment. All three channels require letting P^T rise, which means tolerating inflation. In this sense, the demand reallocation shock acts as a cost-push shock — it shifts the Phillips curve upward, so that higher inflation is required to achieve any given level of employment. If the inflation cost is sufficiently low, the optimal response allows full employment; otherwise an interior solution trades off inflation against economic slack.

Q2. What is the optimal cooperative monetary policy, and how large are the quantitative tradeoffs?

Proposition 1: under international cooperation, the optimal response to ω₀ > ω entails a rise in inflation; if the full-employment inflation level P^fe satisfies χ’(P^fe) ≤ (1/ω₀)(α/(1−α) + 1 − ω₀), the cooperative optimum achieves full employment; otherwise the interior optimum sets χ’(P̄) equal to that expression, balancing marginal inflation cost against marginal employment benefit. The cooperative optimum is strictly superior to strict inflation targeting (P = 1) because the latter allows large unemployment without achieving any structural rebalancing. The global central bank internalizes the income effect from tradable expansion: as Y^T rises, households immediately spend the income on consumption of both goods, further boosting nontradable employment — an amplification mechanism that self-oriented national banks will not fully internalize. In the calibrated numerical example (α = 0.64, ω = 0.3, χ(P) = χ̄(P−1)²/2 with χ̄ = 299.25), a reallocation shock matching the US COVID-19 episode (ω₀ − ω ≈ 0.03) requires approximately 1.5 percentage points of inflation to maintain full employment; under strict inflation targeting (P = 1), unemployment rises by approximately 8 percentage points. These magnitudes are consistent with the observation that during the pandemic inflation cycle, central banks were willing to allow inflation rather than trigger a labor market collapse.

Q3. How does capital mobility change the inflation-unemployment tradeoff faced by individual countries?

Capital mobility reduces the domestic sacrifice ratio — the employment cost of disinflation — through two channels: trade deficits directly sustain nontradable demand (offsetting the fall in tradable sector employment), and they buffer tradable consumption from drops in domestic tradable output. When a single country contracts its monetary policy and P^T falls, domestic tradable output falls; but households react by borrowing internationally, so domestic consumption of tradables falls by less than one-for-one with output (formally: ∂C^T/∂Y^T = ω_{i,0}(1−β)/(ω_{i,0}(1−β)+β) < 1). Capital inflows thus sustain nontradable demand and nontradable employment, partially offsetting the contractionary effect on employment. From each country’s perspective, containing inflation “exports” part of the output loss abroad, making disinflation individually less costly than in a closed economy. This is precisely what creates the coordination failure in the global case: each country perceives a lower sacrifice ratio for disinflation because it does not internalize that this lower sacrifice ratio exists only if the rest of the world continues to produce and lend tradable goods.

Q4. How does the coordination failure arise in a global reallocation shock, and what is the precise mechanism of competitive appreciations?

Proposition 3: in a Nash equilibrium with a global symmetric shock, the full-employment inflation level P^fe coincides with the cooperative benchmark (since C^T_i = Y^T_i in symmetric equilibrium and capital flows net to zero), but if the inflation cost is high enough, self-oriented central banks impose a lower inflation ceiling (MP^u < MP^c) — resulting in lower inflation and higher unemployment than cooperation. Each national central bank individually seeks to reduce domestic inflation by hiking its policy rate to attract capital inflows (which ease the nontradable sector employment constraint through the open economy Phillips curve). But the individual strategy of hiking to attract inflows — which amounts to trying to appreciate the exchange rate (S_i = P^T_{i,t}/P^T_t) — is frustrated in a symmetric Nash equilibrium: when all countries hike simultaneously, capital flows net to zero globally, exchange rates remain unchanged, and only the synchronized monetary contraction remains. This is the mechanism of competitive appreciations: countries try to fight domestic inflation by appreciating their currencies, but appreciate against each other, leaving only a global slump. In the numerical example at ω₀ = 0.35, the uncooperative equilibrium reduces inflation by approximately 1 percentage point relative to cooperation but pushes unemployment to approximately 7 percent (vs. full employment under cooperation at that shock size).

Q5. How do competitive appreciations differ from competitive depreciations, and what are the scope conditions?

Competitive appreciations are the mirror image of competitive depreciations (which characterized the Great Depression and the aftermath of the 2008 GFC): in both cases each country uses its monetary policy to shift costs abroad, but the direction differs — depreciations arise during periods of weak global demand when countries try to steal demand from neighbors; appreciations arise during periods of global tradable goods scarcity and high inflation when countries try to export inflation. The structural difference is the initial state: competitive depreciations occur when global aggregate demand is deficient and the zero lower bound binds — each country wants to depreciate to boost exports; competitive appreciations occur when global demand for tradables is strong relative to supply (ω₀ > ω) and inflation is high — each country wants to appreciate to attract capital inflows that buffer domestic employment from disinflation. The key asymmetry is the direction of the international spillover: in the depreciation case, countries export demand; in the appreciation case, countries export inflation costs. The gains from cooperation in both cases arise for the same reason — the Nash equilibrium involves globally excessive monetary tightening or loosening relative to the cooperative benchmark — but the policy recommendation is opposite in sign.

Q6. What do the model’s predictions imply for the COVID-19 episode and the 1980s disinflation, and when do gains from cooperation materialize?

Gains from monetary cooperation arise only when condition (28) is violated — when central banks are willing to sacrifice full employment to fight inflation; for the COVID-19 episode, gains were likely small (labor markets remained strong throughout); for the 1980s synchronized tightening, the model implies positive gains that would have been achievable through international cooperation. For the COVID-19 episode: throughout the 2021–2023 inflation cycle, unemployment rates in advanced economies remained low and fiscal support maintained aggregate demand, suggesting monetary policy did not sacrifice employment — the model implies condition (28) did not bind and the cooperative optimum was approximately achieved. The world “escaped competitive appreciations this time.” For the 1980s disinflation: the synchronized monetary tightening under Volcker (US), Bundesbank (Germany), and others was accompanied by a deep global recession and explicitly prioritized inflation reduction over employment — precisely the conditions under which condition (28) binds and competitive appreciations generate a suboptimal outcome. These dynamics motivated the heated international cooperation debates of the period, culminating in the Plaza Accord of 1985 (Sachs 1985; Frankel 2015). The model also applies to negative tradable supply shocks (supply chain disruptions, tariffs) in Supplemental Appendix E, so its predictions about cooperation gains extend to protectionist-driven scarcity.

Key concepts

demand reallocation shock : a shift in the preference weight on tradable goods (ω₀ > ω) that reduces nontradable demand relative to tradable demand; in the model it corresponds to a structural demand shift toward durables and goods (as observed during the COVID-19 recovery), creating simultaneous inflationary pressure in tradables and deflationary pressure in nontradables, and generating an inflation-unemployment tradeoff absent in standard cost-push formulations.

convex tradable supply : the feature of the tradable sector (parameterized by α > 0) whereby supply is upward-sloping due to capacity constraints — a 1% rise in the tradable goods price P^T raises tradable output by (1−α)/α percent; calibrated to α = 0.64 (implying a 0.57 price-output elasticity) following Boehm and Pandalai-Nayar (2022) for sectors at high capacity utilization; without this feature, tradable supply would be perfectly elastic and the inflation-unemployment tradeoff would disappear.

competitive appreciations : the Nash equilibrium coordination failure in which each national central bank hikes its policy rate to attract capital inflows (reducing domestic disinflation costs), generating nominal exchange rate appreciation; since all countries do this simultaneously, appreciations cancel out in equilibrium, leaving only a globally excessive monetary contraction with lower-than-cooperative inflation and higher-than-cooperative unemployment; mirror image of competitive depreciations but arising from global scarcity (not deficiency) of tradables.

sacrifice ratio : the employment cost per unit of disinflation; reduced in open economies relative to closed economies because capital inflows buffer domestic tradable consumption from drops in domestic tradable output, and sustain nontradable demand; self-oriented central banks perceive a lower sacrifice ratio than a global central bank, which is the source of the competitive appreciation externality.

nominal wage rigidity : the short-run friction that makes demand reallocation shocks costly: with flexible wages, reallocation from nontradable to tradable employment would occur through real wage adjustment alone; with rigid nominal wages, real wages fall only if tradable goods prices rise (inflation), so monetary accommodation is required for structural reallocation without unemployment.