Historical Origins and Core Concept
A. William Phillips's 1958 empirical study of British wage and unemployment data established a striking pattern: periods of low unemployment coincided with rising wages, whilst high unemployment brought wage declines. This observation challenged earlier economic thinking and prompted researchers to examine whether similar trade-offs existed in other industrialised economies.
The relationship reflects a fundamental economic mechanism: tight labour markets push up wages as employers compete for scarce workers, whilst slack labour markets suppress wage growth. Over decades, economists broadened this insight to encompass inflation itself, creating a framework linking cyclical unemployment to price-level changes. The Phillips curve became central to post-war macroeconomic policy because it suggested policymakers could choose between lower unemployment (accepting higher inflation) or price stability (accepting higher unemployment).
Three Versions of the Phillips Curve
Traditional Phillips Curve: This earliest formulation directly links money wage growth to the unemployment rate. It captures the empirical observation that wage pressure emerges when joblessness falls, without explicitly modelling inflation expectations or supply shocks. Policymakers initially treated the trade-off as stable and exploitable.
New Classical Phillips Curve: Building on the Lucas critique and rational expectations theory, this version incorporates the natural rate of unemployment (NAIRU) and expected inflation. The equation recognises that workers and firms anticipate future price changes when negotiating wages. Only unexpected inflation—or deviations from the natural rate—move unemployment in the short run. This model explains why sustained inflation fails to lower unemployment permanently.
New Keynesian Phillips Curve (NKPC): Introduced in 1995, this framework grounds inflation dynamics in firms' pricing decisions under sticky prices. The NKPC shows inflation depending on expected future inflation, the output gap (deviation from potential output), and supply shocks. This version underpins modern central bank models because it better captures how monetary policy transmission works through firms' real marginal costs and forward-looking behaviour.
Mathematical Formulations
Each Phillips curve version has a distinct mathematical structure reflecting its theoretical assumptions:
Traditional:
π = πᵉ − b(U − Uₙ) + v
New Classical:
π = πᵉ − b(U − Uₙ) + v
New Keynesian:
π = βEₜ(πₜ₊₁) + κỹ + v
where κ = [λ((1−α)σ + φ + α)] / [(1−α) + αε]
and λ = [(1−θ)(1+θβ)] / θ
π— Current inflation rateπᵉ or Eₜ(πₜ₊₁)— Expected inflation (current or forward-looking)U— Unemployment rateUₙ— Natural rate of unemployment (NAIRU)b— Sensitivity of inflation to unemployment gapỹ— Output gap (actual minus potential output)κ— Slope parameter linking inflation to output gapβ— Households' discount factorθ— Price stickiness (duration of fixed prices)σ— Intertemporal elasticity of substitutionφ— Inverse labour supply elasticityα— Labour's share of outputε— Firms' price elasticity of demandv— Unexpected supply shocks
Key Considerations When Using the Calculator
The Phillips curve remains a powerful tool, but several assumptions and real-world complications can affect its predictive accuracy.
- Expectations matter enormously — In modern economies, inflation expectations anchor behaviour more than actual unemployment rates. If central banks lose credibility, expected inflation can rise even when joblessness is high, flattening the Phillips curve. Always check whether your expected inflation assumptions align with survey data or market expectations.
- The natural rate is unobservable — Economists debate the true NAIRU or natural rate constantly, and it shifts over time due to demographics, technology, and labour market institutions. Small changes in your assumed natural rate can dramatically alter the unemployment gap and predicted inflation, so sensitivity analysis is essential.
- Supply shocks create large deviations — Oil price spikes, pandemic disruptions, and wage-cost pressures can push inflation far above or below what the unemployment rate alone would predict. The shocks term (v) captures these, but in real time identifying their magnitude is difficult, making short-term forecasts volatile.
- The long-run Phillips curve is vertical — All three versions predict that permanently lower unemployment requires ever-accelerating inflation—a trade-off that cannot be sustained. In the long run, inflation returns to expectations and unemployment to its natural rate, so the Phillips curve offers no escape from the inflation-unemployment dilemma.
NAIRU versus Natural Rate of Unemployment
These terms are often used interchangeably, but they have distinct meanings. The natural rate of unemployment reflects structural labour market features—skills mismatches, job search frictions, sectoral shifts, and geographic immobility. It is a feature of the economy's long-run equilibrium independent of inflation.
NAIRU (non-accelerating inflation rate of unemployment) is the specific unemployment level consistent with stable, non-accelerating inflation. In low-inflation equilibrium, the two concepts align, but during periods of changing inflation trends or credibility shifts, they can diverge. Central bankers focus on NAIRU because breaching it risks triggering upward spirals in wage and price growth.