Inflation Explained: Mechanisms, Measurement, History and Effects — The Complete Guide

Definitions, the four official measures, stacked causes, a century of documented episodes and effects on debtors, savers and asset classes — the analytical reading of the inflation regime, in one place.

Inflation is not a monthly print to comment, it is a regime to diagnose. Identifying the regime — its drivers, its persistence, its breadth — always precedes asking how high it goes.

This guide assembles, in one entry point, what a reader of the Eco3min inflation sub-pillar needs to navigate the topic: the price mechanics, the measurement controversies, a century of documented episodes, the empirical record of asset classes, and the regime-by-regime reading framework that structures the Eco3min cluster on inflation. It opens onto the 49 satellite articles that develop each axis.

Contents

Part 1 — The mirage of the monthly print: what inflation actually is

A definition that hides three dimensions

The textbook definition — a sustained, generalized rise in the general price level — looks innocuous, but each adjective does precise analytical work. Sustained rules out one-off jumps; generalized rules out relative-price shifts confined to a handful of items; general price level implies an aggregate, not the cost of a particular good. A monthly print can mislead on all three counts at once. A rigorous reading assesses inflation across three distinct dimensions: magnitude (the rate), persistence (how long the rate stays above target) and breadth (the share of basket items posting sustained increases). The trimmed-mean and median CPI series published by the San Francisco Fed, and the ECB’s harmonized core inflation, exist precisely to filter the noise from the regime — a difference that weighs more than any monthly print. A rigorous definition of inflation requires distinguishing these three dimensions every time, and that is precisely what a hurried commentary on a monthly CPI never does.

The monthly print is an observation. The regime is a diagnosis. Conflating the two is the foundational analytical mistake in inflation macroeconomics.

A price increase is not inflation

A product becoming more expensive is not, in itself, inflation. When energy rises while wages, rents and services stay flat, the relative price of energy has shifted; this can be a precursor to inflation if diffusion follows, or a one-off shock that reverses. The 2008 oil spike — Brent at roughly 145 USD/bbl in July 2008 then near 40 USD by year-end (BIS commodity data) — illustrates a relative-price shock that did not produce sustained inflation in advanced economies. By contrast, the 2021-2024 cycle saw energy and food and services move together in the United States and the euro area for several quarters: that joint breadth is the empirical signature of an inflation regime, not an isolated shock. The core / headline inflation distinction technically reflects this diagnostic dichotomy.

Disinflation, deflation, reflation, stagflation, hyperinflation: a five-regime vocabulary

Vocabulary matters because policy responses differ radically across regimes. Disinflation is a slowdown in inflation — prices still rise, just less fast: the U.S. trajectory from 9.1% in June 2022 down to 3.0% in June 2023 (BLS) is the cleanest recent example. Deflation is a sustained fall in the general price level: Japan recorded a cumulative core CPI of approximately minus 2.3% over 1998-2012 (Statistics Bureau of Japan). Reflation denotes a deliberate policy push to lift prices toward target after a deflationary period — the explicit gamble of the BoJ since 2013 and the implicit objective of European central banks during 2015-2019. Stagflation couples elevated inflation with stagnant or falling output (the 1970s; see how stagflation reshapes financial markets). Hyperinflation, in Cagan’s classic 1956 threshold, is a monthly inflation rate above 50% — Weimar Germany breached that threshold in summer 1922, Hungary in 1946, Yugoslavia in 1994, Zimbabwe in 2008. The disinflation–deflation distinction is one of the most consequential in contemporary macro reading.

Why the monthly print misleads systematically

Three structural biases explain why a monthly CPI, taken alone, regularly misleads. First, base effects: 12-month inflation compares the current month to the same month a year earlier, which mechanically lifts or lowers the year-on-year reading depending on the base. Second, composition: a basket where one volatile component moves sharply (energy in 2022, consumer goods in 2021) shifts the headline without shifting the regime. Third, statistical noise: monthly price collection has irreducible variance, which the SF Fed estimates at 30-50 basis points on the annualized monthly reading. Inflation comes in waves precisely because these three biases compound non-linearly during regime transitions. This is why magnitude / persistence / breadth decomposition, rather than a monthly read, is the implicit standard of central-bank research departments.

Part 2 — The Eco3min framework of inflation regimes

A frequent error in macro reading consists of treating inflation as a continuous variable to forecast, when central-bank practice and the empirical literature show it is actually a family of discrete regimes, each with its own drivers, market behaviours and exit patterns. The framework proposed here synthesizes a century of data and doctrine into a five-regime reading grid. This diagnostic matrix serves as a compass: it positions observed conditions in one of five documented regimes before commenting on the headline.

The five-regime grid

Each regime corresponds to a distinct causal and behavioural configuration. Deflation and hyperinflation are the two extreme pathologies, separated by a continuum where the 2% target occupies a conventional equilibrium zone. The table below summarizes for each regime the observed quantitative threshold, the typical driver, the empirical median duration, the reference historical case, and the modal behaviour of major asset classes.

This grid does not predict — it diagnoses. Its analytical function is to force the reader to position observed conditions in one of the five regimes before commenting on the headline. The distinction between structural and cyclical inflation operates at a finer, complementary level: a regime can be structural (persistent equilibrium transition) or cyclical (oscillation around a stable mean). Regimes 1 and 5 are almost always structural; regimes 2 and 3 mix both; regime 4 is precisely the most actively debated empirical object since 2021.

Why five regimes and not three or seven

The five-regime granularity is neither arbitrary nor exhaustive — it reflects an empirical compromise. At three regimes (deflation / moderate inflation / strong inflation), the distinction between active disinflation and stable target inflation is lost, even though it weighs heavily on market behaviour. At seven regimes or more, boundaries blur and the grid loses its diagnostic function. Five regimes correspond to the number of empirical configurations that asset-class behaviours statistically distinguish significantly in the long Shiller series (Yale, 1871-2024) and in the BIS panel covering 60 economies (1960-2024).

The framework in action: the eco3min monthly score

To illustrate concretely how the five-regime grid applies to a real economy in real time, Eco3min publishes a monthly composite indicator that maps current US conditions onto the five-regime framework: the eco3min Inflation Regime Score (US). Built from three public FRED series — headline CPI (50% weight), core CPI (30%), and the 5-year-5-year forward breakeven (20%) — it produces a single bounded reading mapped to one of the five regimes. The dataset is fully reproducible from public data, updated monthly the day after each BLS CPI release, and freely downloadable as CSV or Excel.

As of March 2026, the score reads 2.85 (72.9th historical percentile since 2003), placing the US economy in the Target regime — though near the upper boundary, with the recent gasoline shock from Iran-related supply disruptions pushing headline CPI to 3.3% while core inflation remains contained at 2.6%. Over 23 years of data, the score has spent only about 34 months in the strict Target zone (1.5–3.0) — the “2% target” is more a theoretical objective than a state actually achieved in practice.

Inflation is not a number to forecast, it is a regime to identify — the diagnostic question always precedes the quantitative one.

Part 3 — How inflation is measured: basket, methodology, controversies

Four official measures, four verdicts

Four official indicators coexist in major economies — and they routinely disagree, sometimes by 100 basis points or more over the same period. This divergence is not a statistical anomaly but a deliberate methodological choice. The CPI (Consumer Price Index) tracks a basket of goods and services purchased by urban households: BLS series in the United States, INSEE’s IPC in France. The PCE deflator (Personal Consumption Expenditures), the Federal Reserve’s preferred indicator since 2000, is a chain-weighted index from National Accounts that captures consumption substitutions. The GDP deflator covers all domestically produced goods and services (including investment and government) but excludes imports. The HICP (Harmonized Index of Consumer Prices) is the Eurostat construct used by the ECB for monetary policy. A direct comparison of the four measures makes the methodological gap explicit.

Four indices, four baskets, four verdicts: measuring inflation is a methodological choice before it is an empirical finding.

Basket, weights, surveys: the invisible mechanics

Each monthly publication is the output of a heavy and largely invisible survey machine. INSEE collects roughly 200,000 prices per month across approximately 27,000 outlets in France, on a basket of around 1,000 product categories. The U.S. BLS surveys nearly 80,000 prices monthly.

Weights are revised annually based on effective household consumption measured in the national accounts. The HICP harmonization, in force since 1997, imposes common rules across euro-area members on basket scope: it notably excludes owner-occupied housing, which the U.S. CPI imputes via owners’ equivalent rent.

INSEE’s calculation methodology, basket choices and weight-revision protocol determine the headline as much as the underlying price changes do.

Hedonic adjustment: the hidden controversy

One technical choice deserves special attention: hedonic adjustment. When a 2024 laptop is twice as fast as the 2014 model at the same price, the BLS treats part of that performance gain as a quality-constant price decline. The method is statistically defensible — quality is improving and the price per unit of service rendered falls — but it pushes measured CPI mechanically lower. Boskin Commission estimates from 1996 suggested CPI overstated true inflation by approximately 1.1 percentage points per year, partly through inadequate hedonic adjustment; subsequent work has narrowed but not eliminated the dispute. The hedonic adjustment controversy is one of the most consequential — and least understood — drivers of the gap between perceived and measured inflation.

Headline vs core, services vs goods

Central banks watch core inflation — total inflation minus food and energy — because these two items are volatile and noisy, distorting the regime read. This distinction does not deny energy’s impact on the cost of living; it isolates the persistent component. The post-2020 cycle exposed a second decomposition that proved even more diagnostic: services vs goods. Goods inflation rolled over by mid-2023 while services inflation stayed sticky, particularly in shelter — a pattern visible across the United States, the United Kingdom and the euro area. Shelter alone accounts for roughly one-third of U.S. CPI weight, which means the trajectory of measured inflation is dominated by housing dynamics for many quarters and mechanically conditions the Federal Reserve’s strategy.

Perceived vs measured inflation: a structural gap

European Commission consumer surveys consistently show that perceived inflation runs several points above the official HICP — the gap reached 5 to 7 percentage points in the euro area during 2022 according to the ECB Economic Bulletin, March 2023. The drivers are well documented (Brachinger 2008): asymmetric salience of price increases, frequency-of-purchase weighting in memory, and narrow attention concentrated on daily-life goods (food, fuel). The perceived-versus-measured gap is structural, not a measurement error; it shapes wage demands, voting patterns and central-bank credibility.

Part 4 — What causes inflation: costs, demand, money, expectations

Cost-push vs demand-pull: the two foundational logics

The oldest analytical decomposition splits inflation into two distinct causal channels. Demand-pull inflation arises when aggregate demand outpaces productive capacity — wages rise, output gaps close, prices follow. Cost-push inflation arises when input costs (oil, wages, imports) rise and producers pass them on to selling prices. The 2021-2024 cycle stacked both: post-pandemic excess demand, supply-chain bottlenecks, the energy shock from the 2022 invasion of Ukraine, massive fiscal support, tight labour markets. The IMF’s October 2022 World Economic Outlook attributed roughly half of euro-area inflation to energy, the other half to demand and second-round effects. The two foundational logics of inflation rarely operate in isolation in modern episodes — which is precisely what makes diagnosis difficult in real time.

Money, the quantity theory, and Friedman’s enduring claim

In his 1968 American Economic Association presidential address, Milton Friedman argued that inflation is everywhere and always a monetary phenomenon. The quantity theory of money — formalized by Irving Fisher (1911) as MV = PY — provides the algebraic backbone. Empirically, the relationship is loose at short horizons and tight at long ones: cross-country regressions over 30-year windows in BIS Working Papers consistently link sustained inflation differences to differences in monetary growth. But the post-2009 quantitative-easing era showed that massive base-money expansion can coexist with disinflation when velocity collapses. The Fisher-Friedman framework and its limits, and the more precise question of whether money creation mechanically generates inflation, remain at the heart of contemporary analytical debate — particularly after the post-pandemic experience.

The Cantillon effect and distributional money creation

Richard Cantillon (1755) noted that new money does not enter the economy uniformly: the first holders gain, the last holders pay.

Modern QE illustrates the mechanism. The Federal Reserve expanded its balance sheet from approximately 4.2 trillion USD in January 2020 to nearly 8.97 trillion in April 2022 (Federal Reserve H.4.1 release). This liquidity was injected first through asset markets, lifting equity and real-estate valuations before goods inflation arrived.

The Cantillon effect documents who benefits first from money creation, and the answer is not symmetric across asset classes or income deciles. The mechanism conditions the liquidity cycles observed on crypto-assets as much as on real estate or growth equities.

Expectations: the channel that closes the system

If households and firms expect 5% inflation, wages are set at 5% and contracts indexed accordingly — the expectation becomes self-fulfilling. Edmund Phelps (1967) and Milton Friedman (1968) introduced the role of expectations into the Phillips curve. Today, central banks track survey-based measures (University of Michigan, ECB Consumer Expectations Survey) and market-based measures (TIPS-Treasury breakeven inflation rates). The post-2022 episode validated the doctrine: U.S. five-year-five-year-forward breakevens never moved durably above 2.7% during the 2022 surge — a sign markets believed in eventual mean reversion. Inflation expectations and how they form ultimately determine whether a shock becomes a regime or fades.

No single school owns inflation: it is monetary in stable regimes, fiscal under stress, behavioural always.

Imported inflation, greedflation, structural shifts

Open economies inherit inflation through the exchange rate: a 10% depreciation of the euro against the dollar mechanically lifts dollar-denominated import prices. Imported inflation and exchange-rate transmission dominate small open economies and explain part of the 2022 euro-area / United States gap. The post-2021 cycle also revived the greedflation debate — the idea that corporate profit margins amplified the price surge.

The structural backdrop also matters: deglobalization, demographic ageing and the energy transition tilt the multi-decade inflation regime upward relative to the 1990-2019 disinflationary template. Goodhart and Pradhan (2020) made this the reference thesis. Wage-price spirals remain the classic amplification mechanism, but their triggering conditions are more restrictive than commonly assumed: 2022-2024 saw rapid nominal-wage growth without a true spiral, because expectations remained anchored and productivity growth contained unit costs.

↑ Back to contents

Part 5 — A century of inflation episodes: the eight-case grid

The history of inflation since 1920 is not a sequence of idiosyncratic accidents but a series of causal configurations that recur. Eight episodes structure the academic and operational literature. The table below summarizes for each one the period, the peak reached, the dominant mechanism, the initial regime in which the episode unfolded, the exit mode, and the empirical lesson. This comparative grid — gathering in one place what textbooks treat in separate chapters — is one of the most robust differentiating angles of the Eco3min cluster on inflation.

Weimar 1921-1923: the textbook hyperinflation

The German hyperinflation remains the canonical case. The mark traded at roughly 4.2 per dollar in 1914 and at 4.2 trillion per dollar in November 1923, according to Reichsbank archives cited in Bresciani-Turroni (1937). Cagan’s 1956 monthly series places peak inflation at approximately 29,500% in October 1923. The drivers combined Versailles reparations denominated in gold currencies, monetization of the deficit by the Reichsbank, and a final collapse of money demand once anchoring failed. The episode ended in November 1923 with the Rentenmark, a currency reform backed by mortgage claims on land. An anatomy of the Weimar collapse shows how a fiscal-monetary feedback loop, once unanchored, accelerates non-linearly — precisely the signature of a regime transition, not a linear drift.

The 1930s deflation: the recession machine

The Great Depression delivered the opposite shock. U.S. CPI fell roughly 25% cumulatively between 1929 and 1933, according to the BLS historical series. Falling prices increased the real burden of nominal debts, triggering the cascade described by Irving Fisher (1933) — debt-deflation: liquidations, bankruptcies, bank runs, further deflation. The 1930s deflation as a recession machine remains the founding case explaining why central banks fear deflation more than moderate inflation — a doctrinal asymmetry that still shapes 21st-century reaction functions.

The 1970s great inflation and the Volcker shock

Three forces converged in the 1970s: the breakdown of Bretton Woods in 1971, two oil shocks (1973 and 1979), and an accommodative Federal Reserve under Arthur Burns. U.S. CPI peaked at 14.8% YoY in March 1980 (BLS); UK RPI peaked near 24% in August 1975 (ONS); French CPI peaked around 14% in 1981 (INSEE). The 1970s stagflation became the founding trauma of modern central banking. Paul Volcker, sworn in as Fed chair in August 1979, raised the federal funds rate to roughly 20% by June 1981. The cost was a deep recession in 1981-1982 and unemployment reaching 10.8% in November-December 1982 (BLS). But headline inflation collapsed from 14.8% to 3.7% by April 1983. The Volcker shock established that breaking entrenched inflation requires credible, sustained tightening — politically costly, economically necessary.

Hyperinflation does not arrive through linear acceleration — it arrives when demand for the currency itself collapses.

Japan: the 30-year deflationary trap

After the 1989-1990 asset-bubble collapse, Japan’s CPI hovered around zero or in negative territory for two decades. Cumulative core CPI was approximately −2.3% over 1998-2012 (Statistics Bureau of Japan). The Bank of Japan held the policy rate at zero from 1999 with brief interruptions, deployed the world’s first quantitative easing in 2001, and adopted yield-curve control in 2016. Japan’s 30-year deflationary trap remains the cautionary tale of demographic-driven, expectations-anchored stagnation — and the main reason the Bank of Japan has long preferred the risk of overheating over the risk of deflation.

Chronic regimes: Argentina, Turkey, Zimbabwe

Some countries live with inflation. Argentina’s INDEC reported 211.4% YoY in December 2023, accelerating in early 2024 — the culmination of eight decades of recurrent fiscal-monetary indiscipline (Argentina’s 80-year inflation history).

Turkey’s TÜİK posted 85.5% YoY in October 2022 after the central bank, under presidential pressure, cut rates as inflation rose — an anatomy of modern monetary drift.

Zimbabwe’s November 2008 monthly inflation, estimated by Hanke and Kwok (2009) at approximately 7.96 × 10²² percent, ranks as the second-worst hyperinflation in recorded history — the last documented modern hyperinflation.

The 2021-2024 burst and its resolution

For the 2021-2024 cycle: U.S. CPI peaked at 9.1% in June 2022 (BLS), euro-area HICP at 10.6% in October 2022 (Eurostat), French HICP at 7.3% in February 2023 (INSEE), UK CPI at 11.1% in October 2022 (ONS).

A close reading of 2021-2024 documents that the disinflation that followed was faster than most forecasts — confirming the recurring failure of central-bank inflation forecasts in both directions. The contrast with the Great Moderation 1990-2007 measures the magnitude of the regime shift.

A complete panorama of U.S. inflation over 113 years is documented in the Eco3min dataset on 113 years of American inflation, which assembles the chronology by regime.

↑ Back to contents

Part 6 — The silent transfer: who wins, who pays

Inflation is not neutral — it is a massive and silent redistribution mechanism between categories of economic agents. This redistribution is not a side-effect of the phenomenon: it is its first structural property. A century of data documents that gains and losses are systematically asymmetric depending on each agent’s balance-sheet position, pricing power, contractual indexation and time horizon.

The debtor silently captures what the saver loses overnight: inflation is the tax that crosses no parliament.

Debtors win, savers pay — Fisher’s silent tax

An unanticipated rise in inflation transfers wealth from creditors to debtors: a fixed-rate borrower repays the same nominal amount in cheaper currency. Irving Fisher’s 1930 distinction between nominal and real interest rates makes the mechanism precise. The U.S. household debt-to-GDP ratio fell from approximately 99% in 2008 to around 73% by 2024 (BIS), a deleveraging amplified during 2021-2022 by the inflation episode itself. Conversely, savers in cash and short-duration deposits absorbed cumulative real losses on the order of 7% to 10% over 2021-2023 in major economies. The silent erosion of debt and the finding that cash during high inflation is statistically the worst placement document the same dynamic from two sides of the balance sheet. Purchasing-power erosion is the everyday translation of the mechanism.

The regressive distribution: why inflation hits low incomes harder

Inflation is not neutral across income deciles. ECB Occasional Paper No. 339 (October 2024) documents that the lowest-income decile in the euro area faced effective inflation roughly 1.5 percentage points above the household average during 2022, because food and energy weigh more heavily in their consumption basket. The regressive nature of inflation is a structural feature, not an accident of the 2022 episode. The mechanism compounds with a second asymmetry: wealthier households hold more real and financial assets that can revalue, while modest households mostly hold cash and deposits with negative real returns.

The state: the great silent beneficiary

Sovereign issuers gain twice during an inflationary episode. First, inflation erodes the real value of outstanding nominal debt: U.S. federal debt-to-GDP fell from 134% in 2020 Q2 to 119% by 2023, partly through inflation-driven nominal-GDP growth (Treasury, BEA). Second, bracket creep — fixed nominal tax thresholds — lifts effective tax rates without legislative action. Add the taxation of nominal capital gains that may be real or illusory depending on inflation. The state’s structural advantage during inflation and the dual bracket creep + nominal-gains mechanism shape the political economy of monetary policy — and explain why governments rarely resist, in practice, periods of moderate inflation.

Companies: dispersion through pricing power

Inflation reshuffles corporate margins along one central variable: pricing power. Firms with strong brands, oligopolistic positioning or contractual indexation pass costs through; firms competing on price and exposed to commodity inputs absorb the squeeze. Pricing power and margin dynamics explain why inflation regimes produce sharp sectoral dispersion in equity returns — a phenomenon documented in the 100 years of equity data mobilized in the next paragraph. The demographic dimension adds a layer: inflation and demographics exert growing pressure on pension systems, particularly when indexation is partial or lagged. For salaried households, whether inflation makes them poorer despite nominal wage growth depends precisely on the gap between wage indexation and effective inflation. The true cost of cash inaction is the individual version of the aggregate saver loss. Finally, exchange-rate transmission is one of the central channels through which domestic inflation propagates to trading partners.

Part 7 — Inflation and asset classes: 100 years of empirical record

Bonds: the asset that suffers first

Nominal fixed-income instruments lose purchasing power one-for-one with unanticipated inflation. The 2022 episode delivered one of the worst documented modern performances for the U.S. Treasury: the Bloomberg U.S. Treasury Index fell roughly 12.5% over the calendar year, the largest annual decline in decades (Bloomberg). Bonds are structurally first in the line of inflation casualties, especially long-duration nominal sovereigns. Sensitivity scales mechanically with duration: a 30-year Treasury lost more than a 2-year at comparable rate increase.

Inflation-linked bonds: the protection mechanism and its limits

TIPS (Treasury Inflation-Protected Securities) and French OATi index principal and coupon to CPI/HICP. The mechanism protects against unanticipated inflation but not against rising real yields: 2022 saw TIPS fall roughly 11.85% (Bloomberg US TIPS Index) because real yields rose faster than breakevens. The mechanics of TIPS, OATi and inflation-linked bonds, including the duration trap, are routinely misunderstood in mainstream financial press. The distinction between protection against anticipated inflation (already priced into the breakeven) and protection against surprise inflation (the real service rendered by TIPS) is central to understanding when and why these instruments perform.

Equities: 100 years of empirical record

Robert Shiller’s long-run U.S. data (Yale, online) show that real S&P 500 returns are positive over decades but turn negative across many inflation regimes.

Real total returns ran near zero or negative during 1965-1982. 2022 saw real returns of approximately −23% (S&P 500 nominal −18% with CPI at +6.5%). Inflation compresses equity multiples primarily through the discount-rate channel: the empirical relationship between real interest rates and CAPE is documented in the dedicated Eco3min dataset as one of the most robust in historical equity-valuation research.

A century of equity-inflation data documents that inflation hurts stocks on impact but typically less severely than bonds over long horizons. The money supply / equity returns relationship further allows reading the Cantillon effect from markets.

Gold, commodities, bitcoin: the contested hedges

Gold rose from 35 USD/oz under Bretton Woods to a January 1980 peak of roughly 850 USD. In real (CPI-adjusted) terms, that 1980 peak was not exceeded until the 2010s. Gold has historically hedged better against monetary regime shifts than against year-by-year inflation prints — the empirical safe-haven claim is more nuanced than the marketing narrative.

Broad commodity indices outperformed during 2021-2022 then gave back gains in 2023-2024. Commodities act as hedge or amplifier depending on the regime — documented in the broader analysis of how commodities, inflation and monetary policy interact.

Bitcoin, marketed as “digital gold”, fell roughly 65% in 2022 even as CPI surged, breaking the store-of-value narrative for that cycle (the empirical test of bitcoin as inflation hedge).

Real estate: the conditional hedge

Housing tracks inflation over long horizons but is sensitive to financing conditions. The U.S. Case-Shiller index rose roughly 18% nominally during 2021, then fell 5% during 2022-2023 as 30-year mortgage rates moved from 3% to over 7% (FRED).

Real estate’s inflation-protection paradox rests on a simple empirical fact: it works in nominal terms over decades, but the credit channel can dominate short-term. The real cost of housing can decouple from goods inflation, modifying overall macro dynamics.

Value vs growth, sectoral rotation

Inflation typically penalizes long-duration cash flows more than near-term ones, mechanically favouring value over growth. The 2022 rotation was textbook: U.S. value outperformed growth by roughly 22 percentage points (Russell indices). The literature on inflation-driven style rotation documents the mechanism.

For broader context: how markets react during stagflation and the underestimated soft-stagflation scenario are useful analytical complements. The Eco3min dataset on the US dollar in global crises 1973-2023 further supports international transmission analysis.

No asset class has historically hedged across every inflation regime — dispersion across regimes dominates the empirical average.

↑ Back to contents

Part 8 — Monetary policy and inflation: the transmission mechanics

The 2% target and its history

New Zealand pioneered explicit inflation targeting in 1990 at 0-2%. Canada followed in 1991, the Bank of England in 1992, the ECB at close to but below 2% in 1998 (revised to a symmetric 2% in July 2021), the Federal Reserve at 2% in January 2012. The 2% target is a policy convention, not a scientific finding — its origin is more pragmatic than theoretical. Why central banks raise rates against inflation follows from the credibility imperative attached to that target.

The Phillips curve and its flat decade

A. W. Phillips (1958) documented an empirical inverse relationship between unemployment and wage inflation in the United Kingdom over 1861-1957. The Phelps-Friedman critique (1967-1968) added expectations: no permanent trade-off between unemployment and inflation, only a short-run one. The 2010-2019 decade challenged the short-run version too: U.S. unemployment fell from 10% to 3.5% with no acceleration in core inflation. The Phillips curve and its empirical limits remain a live academic debate — particularly after the 2021-2024 episode revived the hypothesis of a non-linear curve with local flattening and re-steepening at the extremes.

Long and variable lags, persistent forecast failure

Friedman’s phrase from the 1960s — long and variable lags — fully applies today. Recent BIS estimates place monetary-policy transmission lags between 12 and 30 months. Federal Reserve and ECB forecasts have systematically underestimated 2021-2022 inflation by hundreds of basis points and the speed of disinflation in 2023-2024. The systemic reasons for forecast failures are structural, not the failure of one institution. In the heterodox debate, Modern Monetary Theory offers a different reading of the constraints (what MMT actually says about inflation); the orthodox response remains skeptical, but the post-pandemic episode revived the role of fiscal policy as an inflation driver.

The central bank cannot forecast the inflation it commits to fight — which is precisely why anchoring expectations is the policy, more than the rates themselves.

Part 9 — Why inflation diverges across countries

France vs euro area vs United States

Even within the euro area, inflation diverged sharply during 2022: HICP peaked at 7.3% in France (February 2023, INSEE) versus 11.6% in Germany (October 2022, Destatis) and over 20% in the Baltics. The drivers: energy-mix differences, regulated tariffs (the French bouclier tarifaire), heterogeneous fiscal responses, labour-market structures. The France-eurozone-U.S. divergence is more diagnostic of national policy and structure than of common shocks. Monetary transmission itself differs by structure: the same ECB tightening produces contrasting effects depending on whether the economy is dependent on variable-rate mortgage credit (Spain, Netherlands) or long-term fixed-rate (France).

Germany and the Weimar trauma

The Bundesbank’s institutional culture, transmitted to the ECB through the European Monetary Institute and the Maastricht framework, treats price stability as the absolute priority — a direct legacy of the 1923 hyperinflation. The Weimar trauma still shapes ECB doctrine a century later, visible in the asymmetric reaction functions during 2010-2014 (when deflation risk was arguably underweighted) and 2022 (when tightening came faster than for the Federal Reserve despite weaker domestic demand pressures).

The hyperinflation threshold: rare but not extinct

Cagan’s 50%-monthly threshold is rare but not extinct: it has occurred 56 times in recorded history per Hanke and Krus (2013). The quantitative threshold, mechanism and triggering conditions always combine fiscal dominance, loss of monetary anchoring, and a flight from the local currency. No advanced economy has crossed it in the modern era — but the conditions are not metaphysical: they require a combination of simultaneous credibility losses across monetary, fiscal and political anchors.

↑ Back to contents

Reasoned glossary of misunderstood concepts

Fifteen conceptual distinctions on which the financial press and public debate regularly slip. Each entry states the typical confusion before posing the useful distinction.

Inflation vs price increase

A one-off rise in a single good (gasoline, bread) is not inflation. Inflation requires a sustained and generalized rise in the general price level. Conflating the two leads to interpreting every supply shock as an inflationary regime.

Disinflation vs deflation

Disinflation is a slowdown in inflation (prices still rise, just less fast). Deflation is a sustained fall in prices. Politically, one is celebrated, the other feared — they call for very different responses.

Headline vs core inflation

Headline inflation includes food and energy; core excludes them. Central banks steer on core because it filters volatility — not because they deny the cost-of-living impact of energy.

CPI vs PCE vs GDP deflator vs HICP

Four measures, four baskets, four verdicts. CPI tracks urban households, PCE tracks all expenditures (Fed’s preferred), the GDP deflator covers domestic production, the HICP harmonizes the euro area for the ECB.

Perceived vs measured inflation

The structural gap — not a statistical defect. Asymmetric salience of price increases, frequency-of-purchase weighting in memory, and narrow attention on daily-life goods explain why perceived inflation regularly exceeds the official HICP by several points.

Cost-push vs demand-pull inflation

Cost-push: input costs rise and are passed through. Demand-pull: aggregate demand exceeds capacity. The diagnosis is not neutral: monetary tightening responds well to the second, less well to the first.

Structural vs cyclical inflation

Cyclical oscillates around a stable mean with the economic cycle. Structural reflects a persistent equilibrium transition (demographics, deglobalization, energy transition). Policy tools differ radically.

Nominal vs real interest rates

The nominal rate is the headline rate. The real rate is the nominal rate adjusted for inflation (anticipated or realized). A 5% nominal rate with 4% inflation economically yields 1% real — this is the magnitude that drives savings, investment and asset valuation.

Stagflation vs ordinary recession

An ordinary recession combines falling output and disinflation (or deflation). Stagflation combines falling output and persistent inflation. Assets do not behave the same way: the first favours bonds, the second penalizes them.

Hyperinflation vs strong inflation

Cagan (1956) places the quantitative threshold at 50% monthly. Below, one speaks of strong or very strong inflation. The distinction is not pedantic: hyperinflation entails a collapse of money demand, a non-linear dynamic absent at 20% annual inflation.

Cantillon effect vs quantity theory

The quantity theory aggregates everything (MV = PY). The Cantillon effect emphasizes that new money enters through certain channels first (asset markets, regulated sectors, certain agents) — hence distributional effects even before any general price effect.

Adaptive vs rational vs anchored expectations

Adaptive: extrapolate from the past. Rational: integrate all available information (Lucas, 1970s). Anchored: believe the central bank’s target and weight shocks accordingly. The three coexist in the population, and the “anchored” share determines the regime’s resilience.

Greedflation vs profit-led inflation

“Greedflation” is a polemical term. “Profit-led inflation” is its analytical version: the Bernanke-Blanchard 2023 GVA decomposition that splits unit-profit contribution from price growth. The second is testable; the first is rhetorical posture.

Imported vs domestic inflation

Imported inflation comes from import prices transmitted via exchange rate. Domestic inflation comes from unit labour costs and domestic margins. Policy tools differ: the first responds to FX, the second to interest rates.

Bracket creep vs explicit taxation

Bracket creep is the rise in effective tax rate produced by the absence of indexation of nominal thresholds. It is a tax increase without legislation — hence the state’s structural advantage during inflation.

Reasoned bibliography: who says what, and why it is debated

Fifteen foundational or structuring references for the analytical reading of inflation. For each: the central contribution in one sentence, and the documented debate in another. Reasoned, not exhaustive.

Cantillon, Richard (1755) — Essai sur la nature du commerce en général

First to formulate that new money does not enter the economy uniformly. Foundation of any modern debate on the distributional effects of QE.

Fisher, Irving (1911) — The Purchasing Power of Money

Algebraic formalization of the quantity theory (MV = PY). Documented limit: velocity V is not constant — the post-2009 experience confirmed it.

Fisher, Irving (1930, 1933)

Nominal/real interest-rate distinction (1930) and debt-deflation theory (1933). Founding case explaining why central banks fear deflation more than moderate inflation.

Cagan, Phillip (1956) — The Monetary Dynamics of Hyperinflation

Sets the hyperinflation quantitative threshold at 50% monthly. Reference for distinguishing hyperinflation from very strong inflation. Weimar data reused by all subsequent literature.

Phillips, A. W. (1958)

Empirical inverse relationship between unemployment and wage inflation in UK 1861-1957. Criticized for absence of expectations (Phelps 1967, Friedman 1968); flattening debate active since 2010.

Phelps, Edmund (1967) — Phillips Curves and Expectations of Inflation

Independently of Friedman, introduces expectations into the Phillips curve. Co-founds the critique of the permanent unemployment-inflation trade-off.

Friedman, Milton (1968) — The Role of Monetary Policy

“Inflation is everywhere and always a monetary phenomenon.” Dominant doctrine 1980-2010; relativized since 2009 by the QE experience and the instability of velocity.

Bresciani-Turroni, Costantino (1937) — The Economics of Inflation

Foundational empirical study of the Weimar hyperinflation. Primary source for monetary and exchange-rate data 1921-1923.

Boskin, Michael et al. (1996) — Final Report

Estimates that the U.S. CPI overstates true inflation by approximately 1.1 percentage points per year. Modified BLS practice without closing the measurement debate.

Hanke, Steve & Kwok, Alex (2009)

Rigorous estimate of Zimbabwe’s November 2008 monthly inflation (≈ 7.96 × 10²² %). Reference for the 2nd most extreme documented hyperinflationary episode.

Hanke, Steve & Krus, Nicholas (2013) — World Hyperinflations

Systematic compilation of the 56 historical hyperinflations crossing the Cagan threshold. Reference database for comparative analysis.

Goodhart, Charles & Pradhan, Manoj (2020) — The Great Demographic Reversal

Thesis on deglobalization and ageing as upward structural drivers of inflation. Challenges the 2000s-2010s disinflationary consensus.

Bernanke, Ben & Blanchard, Olivier (2023)

Causal decomposition of the 2021-2024 shock. Methodological framework adopted by ECB and IMF. Central reference for the “profit-led inflation” debate.

Brachinger, Hans Wolfgang (2008)

Theoretical framework of perceived inflation. Explains why perception regularly exceeds official measure. Adopted by ECB and CES surveys.

Reinhart, Carmen & Rogoff, Kenneth (2009) — This Time Is Different

Historical database of financial and inflationary crises across eight centuries. Empirical reference for regime-by-regime comparison across eras.

↑ Back to contents

FAQ — Frequently asked questions on inflation

Eight structured questions that recur in Google searches and conversations with sophisticated readers. Condensed answers drawing on the detailed parts of the guide.

What exactly is inflation?

Inflation is a sustained, generalized and persistent rise in the general price level. Three adjectives, three distinct requirements.

A one-off rise in a single good is not inflation. An isolated month of high CPI does not establish a regime. A rigorous reading evaluates magnitude, persistence and breadth jointly — see Part 1.

Three adjectives — sustained, generalized, persistent — separate inflation from a mere price shock. Any commentary that omits one regularly produces directional errors.

What is the difference between inflation and deflation?

Inflation is the sustained rise in the general price level; deflation is its sustained fall.

One is managed via monetary tightening. The other, rarer and harder to halt, can produce a debt-deflation cascade (Fisher 1933). Japan offers the textbook case over 1998-2012 (cumulative core CPI of approximately −2.3%). See the glossary.

How is inflation measured in the United States and the euro area?

The BLS publishes the CPI: a basket of approximately 80,000 prices collected monthly across thousands of urban outlets. The Federal Reserve targets the PCE deflator (chain-weighted, broader coverage).

Eurostat publishes the harmonized HICP that serves as the ECB target (symmetric 2% since July 2021). The HICP excludes owner-occupied housing — a major methodological difference from the U.S. CPI, which imputes it via owners’ equivalent rent. Details in Part 3.

Why is perceived inflation higher than official inflation?

Three documented mechanisms (Brachinger 2008, ECB Bulletin March 2023): asymmetric salience of price increases in memory, frequency-of-purchase weighting, attention concentrated on daily-life goods (food, fuel). The gap reached 5 to 7 points in the euro area in 2022. Not a measurement defect but a structural cognitive phenomenon.

How does hyperinflation get triggered?

Through the combination of three documented conditions: fiscal dominance (deficit structurally monetized), loss of expectations anchoring, and flight from the local currency.

Cagan’s quantitative threshold (1956) is 50% monthly — crossed 56 times in recorded history per Hanke and Krus (2013). No advanced economy has crossed it in the modern era. See the dedicated satellite.

Does gold really protect against inflation?

Gold has historically hedged better against monetary regime shifts (end of Bretton Woods, expectations unanchoring) than against year-by-year inflation prints. The January 1980 peak (≈ 850 USD/oz) was not exceeded in real terms until the 2010s. The empirical safe-haven thesis is more nuanced than the marketing narrative — details in the gold and inflation satellite.

Why do central banks target 2% rather than 0%?

The 2% target is a pragmatic convention, not a scientific finding. Three reasons: it leaves a safety margin above 0% to avoid deflation risk; it eases relative-price adjustments (notably nominally rigid wages downward); it leaves room for policy-rate cuts in a recession (“zero lower bound”). New Zealand pioneered it in 1990, the Federal Reserve adopted it in January 2012. See the dedicated satellite.

Why do central-bank forecasts miss so often?

Three documented systemic reasons:

First, macro-econometric models calibrated on past regimes mis-estimate transitions. Second, monetary-policy transmission lags are long and variable — 12 to 30 months per the BIS. Third, communication itself is a system variable: announcing a forecast modifies expectations that modify inflation.

The Fed and ECB underestimated 2021-2022 inflation by hundreds of basis points, and the speed of 2023-2024 disinflation. See the dedicated satellite.

The complete inflation cluster: 49 satellite articles

Foundations and measurement

The first sub-cluster lays out the conceptual foundations and the measurement architecture. What is inflation? A rigorous definition; the four official measures compared; how INSEE calculates inflation; the perceived-versus-measured gap; services-versus-goods divergence; shelter as the dominant CPI component; hedonic adjustment; disinflation versus deflation; and reflation and market implications.

Causes, theories and central banks

The second sub-cluster covers the causal mechanisms and the institutional framework. Cost-push versus demand-pull; the Fisher-Friedman quantity theory; the Cantillon effect; wage-price spirals; inflation expectations; imported inflation; greedflation and corporate margins; deglobalization as structural driver; the Phillips curve and its limits; and why central banks miss inflation forecasts.

A century of episodes

The historical sub-cluster anchors the analytical claims. Weimar 1921-1923; Zimbabwe 2007-2009; Argentina’s eight decades of inflation; the Volcker shock; the 1970s stagflation; the Great Moderation; the 2021-2024 burst; Japan’s 30-year deflationary trap; the 1930s deflation; and Turkey 2021-2024.

Effects on agents and markets

The fourth sub-cluster maps differential impacts on each economic actor. Debt erosion benefiting debtors; the regressive distributional effect; bracket creep and nominal capital gains; the true cost of cash inaction; exchange-rate transmission; corporate pricing power; demographic and pension pressure; the state as silent beneficiary; a century of equity data; and why bonds suffer first.

Asset classes, geography and complements

The fifth sub-cluster covers inflation-linked instruments, geographic divergences and remaining concepts. TIPS, OATi and inflation-linked bonds; gold as inflation hedge; bitcoin as store-of-value test; commodities as hedge or amplifier; value-versus-growth rotation; MMT and inflation; fiscal policy’s forgotten role; France-eurozone-U.S. divergences; the German Weimar trauma and the ECB; and the hyperinflation threshold.

Existing Eco3min anchors on inflation and real rates

The cluster connects to the existing Eco3min library: structural versus cyclical inflation; inflation’s impact on savings and investments; the long history of real interest rates; and the Eco3min dataset on US net liquidity, a pivot for understanding the monetary-transmission channel beyond policy rates.

Conclusion — How Eco3min reads inflation

Inflation regimes are not forecast, they are recognized — and the costliest mistake is applying the previous regime’s playbook to the regime now forming.

Inflation is the variable through which monetary, fiscal, structural and behavioural forces converge — which is precisely why no single school of thought has ever explained all documented episodes. Rather than a summary, here are the five methodological reflexes that constitute the Eco3min reading framework, distilled from the eight parts of this guide:

1. Identify the regime before debating the print. Five empirically distinct regimes (deflation, disinflation, target inflation, elevated inflation, hyperinflation). Diagnosis precedes forecast. This is the object of the Eco3min framework in Part 2.
2. Decompose magnitude, persistence and breadth. Three dimensions, never one. The SF Fed publishes median and trimmed-mean for this reason; hurried commentators ignore them at the cost of their directional errors.
3. Choose the measure suited to the question. CPI for urban households, PCE for the Fed, GDP deflator for domestic production, HICP for the euro area. Four different verdicts for four different questions.
4. Read expectations as the closing variable of the system. A shock becomes a regime when expectations unanchor. As long as they remain anchored, the shock fades. This is the implicit central-bank doctrine since Volcker.
5. Accept that effects are asymmetric by construction. Debtors and the state silently capture, savers and nominal creditors pay first, firms sort themselves by pricing power. These asymmetries are features of the system, not accidents of one cycle.

The 49 satellite articles linked above develop each axis with the empirical density that any one section of this guide could only sketch. For the raw datasets underlying the regressions and regime-by-regime comparisons, the Eco3min real rates × CAPE dataset, the yield-curve inversion history dataset, the 113-year U.S. inflation history and the eco3min Inflation Regime Score (US) — the cluster’s monthly-updated diagnostic indicator — form the cluster’s reference numerical foundation.

To report a factual error or suggest an update, contact the Eco3min editorial team via the contact page. Corrections are integrated quickly and flagged at the bottom of the page.

Last updated — 5 May 2026