Exploring wage Phillips curves in advanced economies

We investigate the extent to which excess supply (demand) in labour markets contributes to a lower (higher) growth rate of average nominal wages for workers. Using panel methods on data from 10 advanced economies for 1992-2018, we produce reduced-form estimates of a wage Phillips curve specification that is consistent with a New Keynesian framework. We find comparable effects on nominal wage growth from several indicators of "slack" in the labour market: unemployment rates, unemployment rate gaps, the prime-age employment-to-population ratios, a composite labour market indicator constructed using a principal component for a wide range of labour force data, and unemployment rates separated by duration of unemployment. Our results provide evidence that while the slope of the wage Phillips curve seems to have become flatter following the global financial crisis in 2008, the relationship still appears to be highly significant. We find that the long-term unemployment rate (unemployment longer than six months) has had a larger effect on wage growth in the period since 2008. We also investigate the shape of the Phillips curve and find some evidence of a convex relationship between labour market slack and nominal wage growth, particularly for the pre-crisis period. Piecewise regressions suggest some mixed evidence on nominal rigidities in the aggregate data.