The empirical literature on the stability of the Phillips curve has largely ignored the bias that endogenous monetary policy imparts on estimated Phillips curve coefficients. We argue that this omission has important implications. When policy is endogenous, estimation using aggregate data can be uninformative regarding the existence of a stable Phillips curve relationship. But we also argue that regional data can be used to identify the structural relationship between unemployment and inflation. Using city and state-level data from 1977-2017, we show that the reduced form and the structural parameters of the Phillips curve are, to a substantial degree, quite stable.