Academic and research sources
The Relation between Unemployment and the Rate of Change of Money Wage Rates
Phillips' original wage-inflation evidence.
Phillips documented a negative relationship between unemployment and nominal wage inflation in U.K. data. Accessed 2026-05-23.
Economica · 1958
The Role of Monetary Policy
Friedman's natural-rate and expectations critique.
Friedman argued that no permanent inflation-unemployment trade-off remains once expectations adjust. Accessed 2026-08-20.
American Economic Review · 1968
Money-Wage Dynamics and Labor-Market Equilibrium
Phelps' expectations-augmented labor-market analysis.
Phelps independently developed an expectations-based natural-rate account of wage and employment dynamics. Accessed 2026-08-20.
Journal of Political Economy · 1968
What Determines the Sacrifice Ratio?
Ball's measurement of the output cost of disinflations.
NBER and University of Chicago Press (Monetary Policy, ed. Mankiw) · 1994
The Slope of the Phillips Curve: Evidence from U.S. States
Hazell and coauthors estimate a small Phillips-curve slope even in the early 1980s and only a modest decline afterward. Accessed 2026-08-20.
Hazell, Herreño, Nakamura, Steinsson · 2022
Research footing
Evidence and data
Read unemployment gaps together with expected inflation, wages, inflation breadth, and cost-push evidence. Estimating the slope requires separating labor-demand variation from supply and expectation shifts.
Calibration or measurement
Within this equation, kappa sets the slope, expected inflation and the cost-push wedge shift the intercept one-for-one, and the natural rate locates the zero-gap point. Empirical slope estimates can vary across samples and identification designs.
Boundaries
- The graph is a conditional snapshot for given expectations, not an expectations-update model.
- The graph does not trace a disinflation path or estimate a sacrifice ratio.
- Supply shocks and uncertain natural-rate estimates can confound a slack comparison.
Use Guidance
When sufficient
Current comparative statics at a supplied expected-inflation anchor and cost-push wedge. It shows how the conditional inflation reading changes with unemployment slack.
When sketch only
Do not use the graph alone to infer acceleration, expectation anchoring, policy credibility, or a long-run vertical curve. Those results require an explicit expectations process and policy-demand block.
When to switch
Switch to NKPC (theoretical:new-keynesian-phillips) when expectations and forward-looking pricing are central. Switch to a sticky-information Phillips curve when information frictions explain the slope better than sticky prices.
Falsification signal
Holding expected inflation, the cost-push wedge, and the natural-rate estimate fixed, a sustained conditional pattern in which higher unemployment comes with higher inflation would contradict the negative-slope restriction used on this page.