The Death of the Phillips Curve?
Abstract: Are inflation dynamics well captured by Phillips Curve models, or has this framework become less relevant over time? The evidence for the U.S. suggests that the slopes of the price and wage Phillips Curves– the short-run inflation-unemployment trade-offs – are low and have got a little flatter. For example, the recursive estimate of the unemployment coefficient in the core PCE Phillips Curve has fallen a little from -0.09 to -0.07 since the Great Recession. However, the decline is not statistically significant. Dynamic forecasts from the wage and price Phillips Curves estimated using data ending in 2007q4, almost 10 years ago, are pretty close to inflation today. This suggests that (i) low current inflation is not that surprising, and (ii) factors such as increased globalization, increased e-commerce activity, changes in concentration, the aging of the U.S. population and mismeasurement of the NAIRU are not that important (or offset each other). The Phillips Curve is still a useful, albeit imprecise, framework for understanding inflation.