Document Type

Other

Publication Date

2026

Abstract

This paper examines the drivers of U.S. Consumer Price Index (CPI) inflation using a novel, integrated approach on high-frequency monthly data from 2003 to 2025. The analysis focuses on a comprehensive set of variables: money growth, a fiscal stance measure, long-term inflation expectations, labor market conditions, and real activity; to identify which forces consistently push inflation higher or lower. Our methodology begins with a Random Forest model to rank predictors, establishing lagged money growth and breakeven inflation expectations as dominant, while the fiscal stance shows weaker predictive power. This ranking guides the specification of a robust ARIMAX model that successfully accounts for inflation's complex seasonal and non-seasonal persistence, delivering cleaner residuals and better forecasts than statistical benchmarks. To study dynamic effects, we employ Local Projections (LP) and Structural Vector Autoregression (SVAR) to trace inflation's response to unexpected shocks. The results demonstrate a clear hierarchy of drivers: unexpected movements in long-term inflation expectations are the only shocks that generate a rapid, statistically significant pass-through to actual inflation. Policy rate and fiscal shocks are found to have limited or insignificant short-run impact. Taken together, the evidence consistently points to a conclusion: U.S. inflation is driven primarily by its own inertia and shifts in expectations, with monetary conditions serving as a medium-run trend anchor, and fiscal forces playing a limited and unstable role. This underscores that price stability is best achieved by managing expectations and ensuring monetary credibility.

Publisher

South Dakota State University

Rights

Copyright © 2026 Adrián Calderón

Download

Share

COinS