Household Leverage and Monetary Policy Transmission: Evidence from U.S. States

Abstract

I investigate whether pre-existing household leverage amplifies the transmission of monetary policy shocks across U.S. states. I combine high-frequency monetary policy surprises constructed from the U.S. Monetary Policy Event-Study Database (Acosta et al., 2025) with pre-crisis state-level household debt-to-income ratios averaged across the 2003-2007 period. Using a panel of 50 states over 2003Q1 to 2025Q2, I estimate Local Projections (Jordà, 2005) with state and time fixed effects to produce impulse response functions across leverage levels, and supplement these with a panel Bayesian Vector Autoregression. States with higher pre-crisis leverage levels show amplified responses in unemployment, house prices, and personal income growth following a contractionary monetary policy surprise. The housing market emerges as the primary amplification channel in both frameworks, with its effects propagating into labor markets and income at longer horizons. The ordering of peak responses across outcome variables is consistent with balance sheet amplification operating through collateral constraints, with house prices peaking at quarter two, unemployment at quarter eight, and income growth at quarter nine. Household leverage constitutes a macroeconomic vulnerability with direct implications for the Federal Reserve and macroprudential policy design.