New tests of the new-Keynesian Phillips curve

Abstract

Lagged dependent variables typically play an important role in empirical models of inflation. Do these lags reflect backward-looking inflation expectations, or do they proxy for rational forward-looking expectations, as in the new-Keynesian Phillips curve? Galí and Gertler [1999. Inflation dynamics: a structural econometric analysis. Journal of Monetary Economics 44, 195–222] attempt to answer this question using GMM to estimate specifications incorporating both lagged and future inflation. They report small coefficients on lagged inflation and conclude that the new-Keynesian model provides a good first approximation to inflation dynamics. We show that these tests have low power against alternative backward-looking specifications, and demonstrate that their results are also consistent with a backward-looking Phillips curve. Using an alternative approach, we find that the new-Keynesian pricing model cannot explain the importance of lagged inflation in standard inflation regressions, and find that forward-looking terms play a very limited role in explaining inflation dynamics.