Discussion

Valerie Ramey opened the general discussion by praising the paper as a good illustration of how theory and empiricism can inform each other. In particular, identification strategies rooted in theory are essential to good empirical work, she argued. Ramey discussed two identification issues when it comes to estimating Phillips curves. First, cost-push shocks are correlatedwith the explanatory variable, that is, output. Second,monetary policy is endogenous. The paper carefully addresses the second issue by exploiting variations at the regional level, according to Ramey. However, the first issue persists and the paper’s estimates should be interpreted as lower bounds on the slope of the Phillips curve, she argued. Ramey encouraged the authors to use existing state-level data sets on demand shocks, as in Emi Nakamura and Jón Steinsson (“Fiscal Stimulus in aMonetary Union: Evidence fromUSRegions,”American Economic Review 104, no. 3 [2014]: 753–92), to address the first issue. The authorswere very much in agreement with Ramey’s comment. Frederic Mishkin seconded the authors’ conclusions by referring to recentworkof his. PeterHooper, Frederic S.Mishkin, andAmir Sufi (“Prospects for Inflation in a High Pressure Economy: Is the Phillips Curve Dead or Is It Just Hibernating?” [Working Paper no. 25792, NBER, Cambridge, MA, May 2019]) found that the Phillips curve is flatter for periods during which monetary policy is more effective at stabilizing inflation. Indeed, the objective and the control variable should be uncorrelated when monetary policy is set optimally, he argued. Hooper et al. (“Prospects for Inflation in a High Pressure Economy”) exploit regional variations—as in the authors’ paper—to identify the relationship between unemployment and inflation. The identification assumption is that monetary policy is exogenous to regional shocks. Regional variations have also been exploited in the context of fiscal policy, Mishkin noted, citing again the work of Nakamura and Steinsson (“Fiscal Stimulus in a