Wage curve vs. Phillips curve: are there macroeconomic implications?

The standard derivation of the accelerationist Phillips curve relates expected real wage inflation to the unemployment rate and invokes a constant price markup and adaptive expectations to generate the accelerationist price inflation formula. Blanchflower and Oswald (1994) argue that microeconomic evidence of a low autoregression coefficient in real wage regressions invalidates the macroeconomic Phillips curve. This conclusion has been disputed by a number of authors on the grounds that the true autoregression coefficient is close to 1. This paper shows that given the assumption of a constant price markup, micro-level real wage dynamics have no observable implications for macro data on wage and price inflation.