Supply chain linkages amplify inflation shocks and persistence
An ECB working paper demonstrates that incorporating production network linkages into Bayesian vector autoregressions reveals stronger and more persistent inflation spillovers from sectoral shocks, with supply chain structures significantly amplifying energy and agricultural price pass-through.
From supply chain webs to headline CPI
Research by Francesca Monti and Leïla Van Keirsbilck integrates Bureau of Economic Analysis Input-Output tables for 35 U.S. production sectors into a 42-variable Bayesian Vector Autoregressions framework.
Estimating data from 2004 to 2024, the study shows that accounting for intermediate input linkages alters shock transmission significantly.
A 5 percent oil supply shock generates larger, longer-lasting increases in consumer prices than models assuming independent sectoral trends.
Cross-sectional regression confirms that peak price responses scale with sectoral energy intensity, yielding a slope of 0.592 and an R-squared of 0.495. Even narrow sectoral disturbances, such as a 5 percent cereal price shock, spread to broader wholesale and retail sectors, generating aggregate inflation effects.
The cost of neutralizing energy spikes
The model utilizes the Prior for the Long Run framework, imposing economic long-run cointegration structures without forcing rigid short-run restrictions.
In pseudo-out-of-sample evaluations against the Survey of Professional Forecasters from 2019 to 2024, the production-network prior consistently improves forecast accuracy over standard Minnesota priors.
Furthermore, counterfactual simulations reveal that using contractionary monetary policy to fully offset the two-year CPI inflation impact of an oil shock demands steeper interest rate increases and causes deeper output losses in industrial production.
A necessary upgrade for inflation models
This empirical framework demonstrates that standard macro models undercount pipeline pressures by ignoring input-output networks.
Yet, using static historical matrices risks underestimating how fast supply chains adapt during real-time crises.
Central banks must incorporate network priors to avoid misjudging sectoral inflation persistence.