inflation

Inflation is a sustained rise in the general price level of an economy, equivalently a fall in the purchasing power of money over time. It is a monetary phenomenon in the aggregate, even when its proximate triggers are goods-specific.

Definition and measurement

Inflation refers to the growth rate of a price index, not any single price. The distinction matters: relative-price shifts (oil up, electronics down) are not inflation; a rise in the level around which relative prices scatter is. Standard gauges—the CPI, the PCE deflator (the Federal Reserve's preferred measure since 2000), the GDP deflator, and the HICP in the euro area—diverge in scope, weighting, and substitution treatment. The 1996 Boskin Commission estimated U.S. CPI overstated inflation by ~1.1 percentage points annually, chiefly through substitution bias and unmeasured quality change, prompting geometric-mean formulas and the "chained CPI." Core measures (excluding food and energy) and trimmed-mean/median estimators (Cleveland Fed, Dallas Fed) attempt to extract the persistent signal from noisy relative-price shocks—an operational proxy for what theorists call "trend inflation."

Edge cases stress the concept: hedonic quality adjustment (how much of a computer's price fall is deflation vs. more computing?), owners' equivalent rent (imputed, non-transacted, and a large CPI weight), and the treatment of new goods (the "new-goods bias" Hausman emphasized for cellular phones).

The quantity theory and its limits

The oldest analytical frame is the quantity theory of money, MV = PY, traceable to Jean Bodin (1568) explaining the Spanish Price Revolution via New World silver, formalized by Irving Fisher (The Purchasing Power of Money, 1911). Milton Friedman's dictum—"inflation is always and everywhere a monetary phenomenon" (1963)—rests on the empirical stability of long-run money–price correlations. The theory works well cross-sectionally over long horizons and in high-inflation regimes (McCandless–Weber, 1995), but poorly at business-cycle frequencies in low-inflation economies, where velocity is unstable and monetary aggregates decoupled from prices—one reason central banks abandoned monetary targeting after the early-1980s Volcker experiment (1979–82) and the collapse of stable M1/M2 demand.

The Phillips curve and the expectations revolution

A. W. Phillips (1958) documented an inverse UK wage-inflation/unemployment relation over 1861–1957. Samuelson and Solow (1960) reframed it as a policy menu. Friedman (1968) and Phelps (1967) demolished the exploitable long-run tradeoff: rational agents adjust expectations, so only unanticipated inflation moves real activity, yielding a vertical long-run curve at the natural rate (NAIRU). The 1970s stagflation—simultaneous high inflation and unemployment following the 1973 OPEC embargo—vindicated them and buried naïve Phillips exploitation.

The modern New Keynesian Phillips Curve (NKPC), derived from Calvo (1983) staggered pricing, makes current inflation depend on the present discounted value of expected future marginal costs: πₜ = βEₜπₜ₊₁ + κ·mcₜ. This is forward-looking, expectations-centric, and structurally microfounded—but empirically fragile. Galí and Gertler (1999) needed a hybrid backward-looking term; Mavroeidis, Plagborg-Møller, and Stock (2014) showed the slope κ is weakly identified, and the curve has "flattened" since the 1990s, complicating inflation forecasting. The 2010s "missing disinflation" (inflation not collapsing despite the Great Recession's slack) and the 2021–22 surge (inflation exploding with apparently modest slack) both strain the framework.

Expectations: adaptive, rational, and beyond

The expectations term is the theoretical fulcrum. Muth's (1961) rational expectations, imported to macro by Lucas (1972, 1976), implies systematic monetary policy cannot fool agents—the Lucas critique invalidates reduced-form Phillips estimation. Sargent and Wallace (1975) derived policy ineffectiveness under rational expectations plus flexible prices. Yet actual expectations, measured by the Michigan Survey, the SPF, and market-implied TIPS breakevens, display bias, disagreement, and inattention. This spawned rational inattention (Sims, 2003), sticky-information models (Mankiw–Reis, 2002), and diagnostic/behavioral expectations (Bordalo et al.). A central open question: are inflation expectations "anchored" by credible policy, and through what mechanism do household vs. firm vs. financial-market expectations transmit to actual price-setting?

Costs of inflation

Fully anticipated inflation imposes shoe-leather costs (economizing on non-interest-bearing money—the Bailey–Friedman triangle) and menu costs. More damaging are unanticipated and variable inflation: arbitrary wealth redistribution from creditors to debtors, relative-price distortions when prices are staggered, and interaction with non-indexed tax systems (Feldstein). The Friedman rule (1969) prescribes deflation at the real interest rate to satiate money demand; it clashes with the modern consensus for small positive inflation (~2%), justified by downward nominal wage rigidity (Tobin's "greasing the wheels," Akerlof–Dickens–Perry 1996) and the zero lower bound. Blanchard, Dell'Ariccia, and Mauro (2010) provocatively asked whether the target should be 4%.

Hyperinflation and fiscal regimes

Extreme cases illuminate mechanisms. Weimar Germany (1923, prices doubling every ~3.7 days), Hungary (1946, the historical record), Zimbabwe (2008), and Venezuela (2010s) all trace to fiscal roots: monetization of unsustainable deficits. Cagan's (1956) seminal study modeled money demand under hyperinflation and the seigniorage-maximizing inflation rate. Sargent's "Ends of Four Big Inflations" (1982) argued stabilization requires a credible fiscal-monetary regime change, not merely gradual money-growth reduction—expectations shift discontinuously.

This connects to the Fiscal Theory of the Price Level (Leeper 1991; Sims 1994; Woodford 1995; Cochrane's 2023 synthesis), where the price level is determined by the government's intertemporal budget constraint—the present value of surpluses backing nominal debt—rather than by money supply alone. It remains contested (Buiter's critique) but re-entered mainstream debate as an interpretation of the 2021–22 inflation following large pandemic fiscal transfers.

The 2021–2023 episode and live debates

Post-pandemic inflation—U.S. CPI peaking at 9.1% in June 2022, euro-area HICP at 10.6%—reopened foundational disputes. Was it (a) supply-driven (global bottlenecks, Russia's 2022 invasion of Ukraine spiking energy/grain), (b) demand-driven (the ~$5 trillion U.S. fiscal response; Summers's early-2021 warning against Team Transitory), or (c) a shift in the fiscal/monetary regime? Bernanke and Blanchard (2023) decomposed it as initially shock-driven, then transmitting to wage-price dynamics. The relatively painless 2023–24 disinflation—inflation falling with limited unemployment rise—challenged NAIRU-based "sacrifice ratio" pessimism and revived interest in nonlinear Phillips curves (Benigno–Eggertsson 2023) where slope steepens in tight labor markets.

Open questions

• Slope and stability of the Phillips curve: structural flattening, nonlinearity, or identification failure?
• Expectation formation: which agents matter, and can central banks truly "anchor" them?
• Monetary vs. fiscal determination: is FTPL a genuine alternative regime or an accounting identity in search of a mechanism?
• Optimal target: does the low-r* / ZLB world justify raising the 2% norm?
• Measurement in a digital, service-heavy economy: quality change, imputed housing, and platform pricing.

Inflation thus sits at the intersection of monetary theory, fiscal policy, expectations, and measurement—a domain where the cleanest theorems (quantity theory, vertical long-run Phillips curve) coexist uneasily with the messiest empirics.