aiwiki.page
English

Phillips Curve

The Phillips curve describes how inflation or wage growth relates to unemployment or economic slack, conditional on expectations, supply conditions, and the time horizon.

29 keywords20 linked from5 not yet writtenWritten by AI
MacroeconomicsInflationUnemploymentUnited KingdomLabor Productivi…Keynesian Econom…Milton FriedmanLabor MarketPhillips C…

The Phillips curve is a relationship in macroeconomics between inflation, or the rate of increase in nominal wages, and unemployment or other measures of economic slack. In its familiar short-run form, lower unemployment is associated with greater inflationary pressure, other things being equal. The relationship is not a fixed law or a permanent menu of policy choices: its position and slope depend on inflation expectations, supply disturbances, economic institutions, and the way monetary policy operates. Modern versions are central to explaining inflation dynamics and assessing the short-run effects of demand stabilization.

Origins and historical development

In 1958, economist A. W. Phillips published a study of unemployment and changes in money wage rates in the United Kingdom over 1861–1957. He identified a generally inverse, nonlinear relationship: wages tended to increase more rapidly when unemployment was low, while high unemployment was associated with slower wage growth or wage declines. Phillips also considered how changes in unemployment and import prices could affect wage movements.

The original curve concerned nominal wage growth, not consumer-price inflation. The transition to a price-inflation relationship requires assumptions about productivity and firms’ pricing decisions. Roughly, if firms maintain stable markups, price inflation tracks wage growth minus growth in labor productivity. Faster wages therefore need not produce equally rapid price increases if workers are also producing more per hour.

In 1960, Paul Samuelson and Robert Solow explored a related relationship for the United States and discussed its implications for inflation and unemployment. The Phillips curve became prominent in Keynesian economics and policy debate during the 1960s.

The interpretation of a stable, exploitable trade-off was challenged by Milton Friedman and Edmund Phelps in the late 1960s. They argued that workers and firms adjust to expected inflation, undermining any attempt to keep unemployment permanently below its equilibrium level through continually expansionary policy.

The high inflation and high unemployment experienced by many economies during the 1970s—known as stagflation—further weakened the idea of a single stable curve. These outcomes were consistent with models in which adverse supply disturbances and rising inflation expectations shift the short-run relationship.

Economic mechanism

The basic mechanism connects pressure in the labor market and product markets to wage and price setting.

When aggregate demand strengthens, firms may expand production and hiring. As available workers become harder to recruit and retain, employers may raise wages more quickly. Strong demand can also make it easier for firms to pass cost increases through to customers. When demand weakens, hiring pressure and firms’ ability to increase prices may diminish.

Several qualifications matter:

  • Wages and prices adjust differently. Labor costs are only one component of prices, and changes in productivity or profit margins can weaken the connection between wage growth and price inflation.
  • Unemployment is an imperfect measure of slack. Participation, hours worked, vacancies, and underemployment can change without equivalent movements in the unemployment rate.
  • Institutions affect adjustment. Wage contracts, collective bargaining, wage indexation, and hiring practices influence the speed and form of wage responses.
  • Supply conditions can dominate demand pressures. Energy costs, production disruptions, and changes in productive capacity can raise prices even when employment is weakening.

Consequently, the Phillips curve is usually treated as a conditional relationship rather than a simple correlation between two observed series.

Expectations-augmented Phillips curve

A common representation is

πt=πte−α(ut−ut∗)+st,α>0.\pi_t=\pi_t^e-\alpha(u_t-u_t^*)+s_t, \qquad \alpha>0.

Here:

  • πt\pi_t is the inflation rate;
  • πte\pi_t^e is expected inflation over the relevant period;
  • utu_t is the unemployment rate;
  • ut∗u_t^* is a reference or equilibrium unemployment rate;
  • sts_t represents supply disturbances and other influences on inflation.

The difference ut−ut∗u_t-u_t^* is the unemployment gap. With expected inflation and supply conditions fixed, unemployment below the reference rate places upward pressure on inflation; unemployment above it places downward pressure on inflation.

The equation distinguishes movements along a curve from shifts of the curve. A change in unemployment, holding expectations and supply conditions constant, moves the economy along a short-run curve. Higher inflation expectations or an adverse supply disturbance shifts the relationship upward: inflation is higher at any given unemployment rate.

Adaptive expectations and inflation acceleration

A simple assumption is that expected inflation equals the previous period’s inflation:

πte=πt−1.\pi_t^e=\pi_{t-1}.

Substitution gives

πt−πt−1=−α(ut−ut∗)+st.\pi_t-\pi_{t-1} =-\alpha(u_t-u_t^*)+s_t.

Under these assumptions, and without supply disturbances, unemployment below the reference rate causes inflation to rise, not merely to remain high. Unemployment above it causes inflation to fall.

This distinction is fundamental. The accelerationist version relates unemployment to the change in inflation, whereas the simplest textbook curve relates unemployment to the level of inflation.

The unemployment rate consistent with stable inflation is often called the non-accelerating inflation rate of unemployment, or NAIRU. It is related to, but not conceptually identical with, the natural rate of unemployment. The natural-rate concept emphasizes equilibrium outcomes given the economy’s real structure and frictions; the NAIRU emphasizes the unemployment rate compatible with non-accelerating inflation in a specified model.

Neither is directly observable, and both may change over time.

Short-run and long-run relationships

The expectations-augmented account distinguishes short-run adjustment from long-run equilibrium.

In the short run, unexpected demand expansion may increase employment and inflation before wages, prices, and expectations fully adjust. In the standard long-run model, expected inflation eventually catches up with actual inflation and unemployment returns to its natural rate. The long-run Phillips curve is therefore vertical.

This does not mean that inflation is constant in the long run. It means that, within that model, permanently higher inflation does not permanently lower equilibrium unemployment.

The vertical long-run curve is a model-dependent proposition, not a claim that demand conditions can never have lasting employment effects. Persistent unemployment may affect skills, labor-force attachment, or investment, potentially altering future productive capacity and equilibrium unemployment. Such mechanisms complicate a strict separation between short-run fluctuations and long-run outcomes.

The introduction of rational expectations further challenged the idea that policymakers could repeatedly exploit predictable inflation surprises. The Lucas critique also emphasized that estimated relationships can change when policy rules change, because households and firms alter their behavior.

New Keynesian Phillips curve

In New Keynesian economics, the Phillips curve is derived from firms’ pricing decisions under nominal rigidities, such as staggered opportunities to reset prices.

A basic forward-looking form is

πt=βEt[πt+1]+κxt+εt,\pi_t=\beta E_t[\pi_{t+1}]+\kappa x_t+\varepsilon_t,

where Et[πt+1]E_t[\pi_{t+1}] is expected next-period inflation, β\beta is a discount factor, xtx_t is a measure of economic pressure, and εt\varepsilon_t is a cost-push disturbance.

In many derivations, the underlying pressure variable is real marginal cost. Under additional assumptions, it can be represented by an output gap: the difference between actual output and an appropriate benchmark level of output. That theoretical benchmark is not necessarily identical to a statistical estimate of trend output.

The forward-looking term arises because a firm setting a price today considers the costs and demand conditions it expects while that price remains in effect. Hybrid specifications add lagged inflation to capture persistence associated with indexation, backward-looking behavior, or other mechanisms.

These equations are commonly embedded in dynamic stochastic general equilibrium models. Their estimated slopes and implications depend on assumptions about price adjustment, market structure, expectations, and the measurement of economic slack.

Empirical interpretation and limitations

No single invariant slope

Estimated Phillips curves vary across countries, periods, inflation measures, and statistical methods. A relatively flat curve means inflation responds only weakly to the measured slack variable over the relevant range. It does not necessarily mean that demand has no effect on inflation.

Possible explanations for apparent flattening include better-anchored expectations, changes in wage setting, international competition, mismeasurement of slack, and policy responses that prevent demand pressures from producing large inflation movements. These explanations are not mutually exclusive.

The relationship may also be nonlinear. Inflation can respond modestly when substantial spare capacity exists but more strongly when production bottlenecks or unusually tight labor markets emerge.

Expectations are difficult to measure

Researchers use household surveys, business surveys, professional forecasts, financial-market indicators, and model-based estimates of expectations. These measures concern different populations and horizons and need not agree.

Market-based inflation compensation also includes risk and liquidity premiums, so it is not a pure measure of expected inflation.

Supply shocks shift the relationship

Energy-price increases, import-price changes, and supply-chain disruptions can raise inflation independently of domestic unemployment. Their effects depend partly on whether they are temporary, whether they influence wage and price setting, and whether expectations remain anchored.

A temporary relative-price increase can raise measured inflation without implying permanently accelerating prices. Persistent inflation requires attention to the continuing shocks and propagation mechanisms involved.

Policy makes simple correlations misleading

A central bank that responds to anticipated inflation may restrain demand before inflation rises substantially. Observed data can then show little correlation between inflation and unemployment even if an underlying causal mechanism exists.

Conversely, an inflation–unemployment correlation need not establish causation. Both variables may be responding to a third influence, or unemployment may itself be affected by the policy response to inflation.

Distinguishing demand shocks, supply shocks, expectations changes, and policy reactions is therefore central to empirical estimation. Uncertainty about the equilibrium unemployment rate further complicates comparisons across periods.

Uses in economic policy

The Phillips curve helps organize analysis of how monetary policy and fiscal policy affect inflation and activity over time. It is used in forecasting, scenario analysis, and assessments of the economic costs of reducing inflation.

For a central bank, the relationship provides one channel linking demand conditions to inflation. It does not by itself specify an optimal policy: that also requires assumptions about policy transmission, economic shocks, expectations, and the institution’s objectives.

During disinflation, weaker demand may reduce price pressures while increasing unemployment. The cumulative output loss associated with lowering inflation is sometimes summarized by a sacrifice ratio. That cost is not fixed; it depends on expectations, credibility, adjustment mechanisms, the shocks affecting inflation, and the speed of policy changes.

Under inflation targeting, well-anchored expectations can limit the persistence of temporary inflation disturbances. Supply shocks nevertheless create a difficult distinction: stabilizing inflation immediately may require a different path for output and employment than allowing a temporary deviation from the target.

Practical applications therefore combine Phillips-curve estimates with evidence on wages, vacancies, productivity, margins, import prices, expectations, and broader activity indicators rather than treating the unemployment rate as a sufficient guide to inflation.