An empirical nail in the coffin of Calvo pricing

I have never been shy about the fact that I am no fan of Calvo pricing as commonly practiced in New-Keynesian model. I have presented on this blog at several occasions theoretical evidence against it, as well as empirical evidence that price are either not rigid enough to matter, or that the way they change is not consistent with Calvo pricing.

Sascha Becker pushes the argument further. There is ample evidence that the relationship between inflation and price dispersion is U-shaped: At very low inflation, the is maximal dispersion. It decreases and increases with higher inflation. Monetary models can explain this, but very differently. Money search models require market power for this to happen. New-Keynesian models with Calvo pricing need sticky prices. Using monthly data from price level indices for 38 sectors in 12 countries over 13 years, Becker looks for sectors where there is more or less competition and more or less price stickiness. The U-shape is systematically found where there is more market power, but not where there is price stickiness...