Time-varying impact of supply chain pressures on US inflation

ABSTRACTWe estimate a time-varying parameter VAR with stochastic volatility (TVP-VAR-SV) model using the global supply chain pressure index (GSCPI) to study the variation of the supply chain impact on US inflation over time. We find the increasing impact of the global supply chain shock on inflation, which has become more persistent during the COVID-19 pandemic. The findings suggest that supply chains have become a more influential in understanding economic variables, especially inflation.KEYWORDS: US inflationsupply chain disruptionTVP-VAR-SVJEL CLASSIFICATION: C32E31F62 AcknowledgementsThe authors appreciate Jinill Kim and the anonymous referees for their valuable comments and suggestion.Disclosure statementNo potential conflict of interest was reported by the author(s).Notes1 In this study, our interest lies in global supply chain. Global value chains (GVCs) and global supply chains are interchangeable in some context, but GVCs are typically considered to be a broader and more complex concept than supply chains. Supply chains primarily encompass production processes, whereas GVCs encompass not only production but also pre- and post-production processes where additional value has been added.2 A detailed description of the TVP-VAR-SV model is placed in the Appendix A.:3 We use the 3-month average for the quarterly values. The specific information about the data is given in Table B1 in Appendix B.4 They measure the real oil price by the US refiners’ acquisition cost for imported crude oil, which is then deflated by the US CPI.5 Since the Purchase Manager Index (PMI), which is a component of the GSCPI, can be influenced by both supply-side and demand-side factors, the demand-side factors are controlled when the GSCPI is constructed. See Benigno et al. (Citation2022).6 In our TVP-VAR setup, Δgt, represents the growth rate of the BDI, which can be viewed as its first moment, whereas the GSCPI quantifies standard deviations from the average value, reflecting its relationship with the second moment. This perspective highlights that Δgt and st do not contain duplicated information, despite the fact that the BDI is one of the components of the GSCPI. See Benigno et al. (Citation2022).7 This can be verified with the sample autocorrelation, sample paths, and posterior densities provided in Figure C1 in Appendix C as well.8 Di Giovanni et al. (Citation2022) find supply chain bottlenecks as the source of the recent high inflation.