**
The Phillips curve is a macroeconomic model that illustrates a historically observed inverse correlation between the rate of unemployment and the rate of wage inflation, which is often extended to price inflation. First identified in a 1958 paper by New Zealand‑born economist A. W. Phillips, the relationship was derived from weekly data on unemployment and nominal wage changes in the United Kingdom from 1861 to 1957. Although Phillips himself did not explicitly connect unemployment to price inflation, his statistical finding implied that tighter labor markets tend to generate faster wage growth, which can be transmitted into higher consumer prices.
The curve quickly became a cornerstone of post‑war Keynesian policy analysis, as policymakers interpreted it as a short‑run menu for choosing between lower unemployment and higher inflation. The idea was formalized by Paul Samuelson and Robert Solow in 1960, who explicitly linked Phillips’s wage data to price inflation and presented the curve as a stable, exploitable trade‑off. The subsequent theoretical debate, most notably the contributions of Milton Friedman and Edmund Phelps, introduced expectations‑augmented versions that distinguished between short‑run and long‑run dynamics, ultimately reshaping the role of the Phillips curve in modern macroeconomics.
Today, the Phillips curve remains a focal point for debates over monetary policy, inflation targeting, and the structural forces shaping labor markets. Empirical work since the 1990s has documented periods of flattening, resurgence, and even apparent breakdowns, prompting ongoing research into the roles of globalization, technology, and anchored inflation expectations.
History
Early Empirical Observation
In 1958, A. W. Phillips published “The Relation between Unemployment and the Rate of Change of Money Wage Rates in the United Kingdom, 1861‑1957” in Economica. Using a moving‑average technique, Phillips demonstrated that weeks with lower unemployment were associated with higher nominal wage growth. His dataset covered 96 years and 1,200 observations, providing a robust statistical foundation for the observed trade‑off.
Keynesian Adoption
The immediate impact of Phillips’s work was felt in the United States, where Paul Samuelson and Robert Solow (1960) extended the analysis to price inflation, arguing that the same inverse relationship applied to the broader economy. Their seminal article in the American Economic Review presented the “Phillips curve” as a policy tool: by tolerating a modest rise in inflation, governments could achieve lower unemployment, a notion that underpinned much of the 1960s macroeconomic policy.
Expectations‑Augmented Theory
The apparent stability of the curve was challenged in the late 1960s by rising inflation without corresponding drops in unemployment. Milton Friedman (1968) and Edmund Phelps (1967) independently introduced the concept of adaptive expectations, arguing that once workers and firms anticipate higher inflation, the trade‑off disappears. Their “expectations‑augmented Phillips curve” posited a vertical long‑run relationship at the natural rate of unemployment (NRU), implying that any attempt to push unemployment below the NRU would only generate accelerating inflation.
The Great Inflation and Policy Shift
During the 1970s, the United States experienced “the Great Inflation,” with unemployment remaining high while inflation surged—a phenomenon that seemed to contradict the original Phillips curve. The experience reinforced the expectations‑augmented view and led to a paradigm shift toward monetary targeting and inflation‑anchoring policies, epitomized by the Federal Reserve’s adoption of explicit inflation targets in the 1990s.
Theoretical Foundations
Original Empirical Curve
The original Phillips curve can be expressed as a simple linear regression:
\[
\pi_t = \alpha - \beta u_t + \varepsilon_t,
\]
where \(\pi_t\) is the inflation rate, \(u_t\) the unemployment rate, \(\beta > 0\) the slope coefficient, and \(\varepsilon_t\) an error term. Empirically, \(\beta\) was estimated to be roughly 0.5 in the United Kingdom data, indicating that a 1 percentage‑point rise in unemployment was associated with a 0.5 percentage‑point decline in wage inflation.
Expectations‑Augmented Model
Friedman and Phelps introduced expectations \(E_t[\pi_{t+1}]\) into the relationship:
\[
\pi_t = E_t[\pi_{t+1}] - \gamma (u_t - u^* ) + \varepsilon_t,
\]
where \(u^\) denotes the natural rate of unemployment and \(\gamma > 0\) the sensitivity of inflation to the unemployment gap. In the short run, if expectations are static, the curve retains a negative slope; in the long run, when expectations fully adjust (\(E_t[\pi_{t+1}] = \pi_t\)), the relationship becomes vertical at \(u = u^\).
New‑Keynesian Formulation
Modern macro models embed the Phillips curve within a New‑Keynesian framework, deriving it from price‑setting behavior under staggered contracts. The canonical “New‑Keynesian Phillips curve” (NKPC) is:
\[
\pi_t = \beta E_t[\pi_{t+1}] + \kappa x_t + \eta_t,
\]
where \(x_t\) is the output gap, \(\beta\) the discount factor, \(\kappa\) a parameter reflecting price rigidity, and \(\eta_t\) cost‑push shocks. The NKPC emphasizes forward‑looking expectations and the role of real marginal costs rather than unemployment per se, yet it retains the core insight that inflation dynamics are linked to real‑activity deviations.
Empirical Evidence
Post‑War United Kingdom and United States
Empirical studies in the 1960s confirmed a negative slope in many advanced economies. For example, Samuelson and Solow estimated \(\beta \approx 0.6\) for the United States using CPI data from 1948‑1960.
The 1970s Breakdown
During the 1970s, the relationship flattened dramatically. Studies by Friedman and Phelps showed that the estimated \(\beta\) approached zero when inflation expectations were incorporated, supporting the expectations‑augmented view.
The 1990s to Early 2000s
The 1990s witnessed a resurgence of a modestly negative slope in many OECD countries, coinciding with credible inflation targeting. Blanchard, Gali, and Leijonhufvud (1999) found a statistically significant NKPC coefficient (\(\kappa\) ≈ 0.2) for the United States, suggesting that output gaps still influenced inflation.
Recent Flattening (2010‑2022)
From the Global Financial Crisis through the COVID‑19 pandemic, the Phillips curve appeared unusually flat. Ball and Mazumder (2021) reported that a 1 percentage‑point change in the unemployment rate corresponded to only a 0.1 percentage‑point change in inflation in the United States. Proposed explanations include globalization, labor‑market slack not captured by headline unemployment, and anchored inflation expectations due to central‑bank credibility.
Emerging Evidence (2023‑2025)
Preliminary data from 2023‑2025 suggest a modest re‑steepening as wage growth accelerates in response to tight labor markets and supply‑chain disruptions. Early estimates from the International Monetary Fund (IMF) place the short‑run Phillips‑curve elasticity at around 0.3 for advanced economies, though considerable cross‑country variation remains.
Policy Implications
Monetary Policy Design
Central banks have traditionally used the Phillips curve to gauge the trade‑off between stabilizing inflation and supporting employment. The Taylor rule, for instance, incorporates an output gap term that implicitly relies on a Phillips‑curve relationship.
Inflation Targeting
The expectations‑augmented framework underpins modern inflation targeting. By committing to a clear inflation goal, policymakers aim to anchor expectations, thereby flattening the short‑run Phillips curve and reducing the inflation cost of lowering unemployment.
Fiscal Policy Considerations
Fiscal stimulus that reduces unemployment can, according to the Phillips curve, generate upward pressure on inflation. However, when the curve is flat, the inflationary cost may be negligible, allowing for more aggressive fiscal support without jeopardizing price stability.
Structural Reforms
If the natural rate of unemployment \(u^\) is high due to structural mismatches, the Phillips curve suggests that reducing \(u^\) through labor‑market reforms can shift the long‑run vertical line leftward, achieving lower unemployment without higher inflation.
Criticisms and Modern Developments
Empirical Instability
Critics argue that the Phillips curve’s slope has varied dramatically across decades, questioning its reliability as a policy guide. Romer (2000) highlighted that the curve’s predictive power fell below statistical significance in many post‑1990 samples.
Role of Globalization and Technology
Research by Autor, Dorn, and Hanson (2016) links the flattening to import competition and automation, which depress wage pressures even in tight domestic labor markets. This “global slack” hypothesis suggests that the traditional unemployment‑inflation link may be weakened by external factors.
Expectations Formation
The assumption of rational or adaptive expectations has been challenged by behavioral studies showing that agents may form expectations based on heuristics, media narratives, or anchoring to past inflation, complicating the simple forward‑looking NKPC.
Alternative Models
Some economists propose a “Phillips‑curve‑free” approach, focusing on financial stability and the zero‑lower‑bound as primary constraints on policy. Others advocate for a “core‑inflation” model that ties price dynamics to commodity prices and exchange rates rather than labor market conditions.
Ongoing Research
Current work explores non‑linear Phillips curves, where the trade‑off is steeper at very low unemployment levels but flattens elsewhere. Machine‑learning techniques are also being applied to detect hidden regimes and to incorporate high‑frequency data (e.g., job‑vacancy rates, wage‑growth expectations) into more responsive models.
INFOBOX:
- Name: Phillips curve
- Type: Economic model / macroeconomic relationship
- Date: First identified 1958 (Phillips’s paper)
- Location: United Kingdom (original data); globally applied
- Known For: Illustrating the inverse relationship between unemployment and inflation and influencing monetary‑policy frameworks
TAGS: macroeconomics, inflation, unemployment, monetary policy, expectations, New Keynesian economics, labor market, economic history
