Stagflation is a period of weak or stagnant economic growth combined with high inflation and high or rising unemployment. It creates a difficult setting for households, businesses and policymakers: purchasing power is under pressure while production and job opportunities are weak. For a quant researcher, the term describes a combination of conditions; it does not by itself identify their cause or predict asset returns.
The RBA's supply-shock explainer emphasises this combination. There is no universal numerical cutoff for “weak” growth or “high” inflation. An ECB analysis published in 2022 explicitly notes the absence of a unique definition and sets a persistence rule for its own analysis. Such a study-specific rule is not a general classification standard.
Three indicators, one difficult year
Consider a fictional economy in two consecutive calendar years. Real GDP is an annual volume index; consumer prices are an annual-average price index; unemployment is the annual-average share of the labour force unemployed. Coverage and methods stay unchanged. Each index is independently rebased to 100 in Year 0. These annual observations have no seasonal adjustment or annualisation applied.
| Indicator | Year 0 | Year 1 | Matched annual comparison |
|---|---|---|---|
| Real GDP index | 100 | 99 | −1% real growth |
| Consumer-price index | 100 | 106 | +6% inflation |
| Unemployment rate | 5% | 7% | +2 percentage points |
The real-output and price changes use their respective starting index levels:
The unemployment-rate increase is percentage points, not a 2% relative increase. The price calculation compares annual averages; it is not December-to-December inflation.
Fictional consecutive calendar years. Each panel has its own labelled scale; output and prices are independently rebased indices, while unemployment is a percentage of the labour force. Lines join annual observations only.
This is an illustration of a stagflationary mix, with 6% assumed high for this teaching scenario. It is not a universal stagflation test, an official recession determination or evidence about any current economy. Weak growth can also mean slow positive growth; an outright fall in real GDP is not required by every use of the term. Two price-index levels establish an inflation rate, but cannot establish that inflation is accelerating or persistent. That requires a longer history.
A possible mechanism is not a diagnosis
An adverse supply shock can make important inputs more expensive or less available. Firms may then produce less while charging higher prices. The RBA describes how a sufficiently large or persistent shock can reduce both current production and potential output, while poorly anchored inflation expectations can help entrench inflation. A contraction in activity therefore need not remove price pressure immediately.
But the three readings above do not prove a supply shock occurred. They contain no information about energy supply, productivity, demand, policy or expectations. Calling the outcome stagflation is a description; attributing it to one mechanism needs additional evidence and an identification strategy.
The historical analogy also needs care. Federal Reserve History's account of the Great Inflation discusses energy disruptions alongside monetary policy, expectations and mistaken real-time estimates of economic capacity. The episode cannot be reduced to “oil went up, so every later oil increase must produce the same regime.”
Why policy faces a trade-off
Tighter monetary policy can restrain demand and inflation pressure, but may further weaken activity and employment. Easing to support demand may add to inflation pressure. The Cleveland Fed's explanation sets out this tension. An interest-rate change also cannot directly replace a missing physical input.
The term alone cannot determine the appropriate response. That depends on the shock, its expected persistence, inflation expectations, economic slack and the policymaker's mandate. A negative output gap is an estimate relative to sustainable capacity; falling measured output does not establish its size if capacity has also changed. Nor does the policy trade-off imply a stable long-run bargain in which a chosen amount of inflation permanently buys lower unemployment; the Fed's historical account explains the failure of that assumption.
Turning the label into a research variable
A regime study needs an explicit rule: which economy, which real-growth measure, which price index, which unemployment series, what horizons and what persistence requirement? Using headline rather than core inflation, or annual-average rather than year-end changes, can change the resulting labels. Record the rule rather than treating “stagflation” as a self-explanatory data field.
For a backtest, retain publication dates and the data vintages available at each decision. A full calendar year's average is unavailable before the year ends and the necessary releases arrive. Later revisions and retrospective labels must not leak into earlier signals through look-ahead bias. A measured historical association with returns still needs separate testing; the label supplies no automatic trading rule.
Sources
- Reserve Bank of Australia — Causes of Inflation: cost-push mechanisms, expectations and “Supply Shocks and Stagflation.”
- ECB — Does the private sector foresee a stagflation episode?, Economic Bulletin 4/2022: definition and an explicitly study-specific persistence convention; no 2022 forecast is used as a current observation here.
- Federal Reserve Bank of Cleveland — What Is Stagflation?, May 2026: the combination of conditions and the monetary-policy trade-off.
- Federal Reserve History — The Great Inflation: historical policy, expectations, energy shocks and real-time capacity estimates.