Abstract:
This paper investigates the effects of oil supply, oil-specific consumption demand, oil inventory demand shocks, and global economic activity shocks on state-level tradable and non-tradable inflation in the US. We use oil shock data following the work of Baumeister and Hamilton (2019) and estimate both linear and non-linear impulse responses using a lag-augmented local projections model in a panel context. Our results from a linear model show that both supply and demand-side oil shocks have a statistically significant impact on both types of inflation. While supply, global economic activity, and demand shocks have a greater impact on tradable inflation, non-tradable inflation responds more strongly to inventory shocks. Further, the non-linear model results provide evidence of heterogeneity in the magnitude and persistence of impact between high- and low-oil dependence regimes. Non-tradable inflation is more sensitive to nearly all components of oil price shocks in the high-oil dependence regime.
