Sectoral Spillovers in Inflation Dynamics: Empirical Evidence from Network Propagation

Distinguishing between sector-specific and aggregate shocks and assessing their contributions to inflation are vital for informed policy. This paper quantifies cross-sectoral spillovers in U.S. consumer price inflation using a factor-adjusted network approach that jointly models aggregate factors and sectoral network propagation. Using disaggregated personal consumption expenditure data across 26 sectors from 1959–2024, the model employs Lasso nuclear-norm regularization to estimate high-dimensional VARs while controlling for aggregate influences. Cross-sectoral spillovers account for roughly two-fifths of total price variation—more than twice the share attributable to aggregate factors—and are systematically mismeasured in conventional models: factor models understate spillovers by absorbing network transmission into common components, while VARs without factors overstate them by conflating comovement with propagation. The spillover structure is highly granular, dominated by large consumer-facing sectors such as food, furnishings, and services, with gasoline exerting more moderate but persistent effects. Spillovers propagate mainly through backward production linkages and scale with sector size, indicating that large downstream sectors play a disproportionate role in transmitting sector-specific shocks across the price network. The findings underscore the need for integrating sectoral networks and aggregate factors in modeling inflation dynamics and policy design.