An estimation of the Phillips curve in Mexico using city-level data

We estimate the slope of the Phillips curve in Mexico between 2005 and 2020 using city level data. We overcome the endogeneity of unemployment and core inflation through a panel instrumental variable strategy. Time-period fixed effects account for aggregate supply, demand, and expectation variables, while possible endogeneity between unemployment and core inflation at the city-quarter level is addressed with local labor demand shock instruments. We find a statistically significant Phillips curve linking local unemployment to local core inflation in Mexico, but this relationship is relatively weak: an increase of 1 percentage point in city-level unemployment lowers year-on-year core inflation by approximately 0.18 percentage points. We analyze city-level characteristics that relate to steeper Phillips curves, and find that informality rates, cash transfers, and some demographic characteristics in cities strengthen the relationship between unemployment and inflation.