Private sector debt overhang and government spending multipliers: Not all debts are alike

Abstract

This study empirically examines whether government spending multipliers vary depending on the level of debt in the private sector by estimating an interacted panel vector autoregressive model (IPVAR) with unbalanced panel data from 29 OECD countries. Results reveal that (i) government spending multipliers become larger in private debt overhang; (ii) and not all debts are alike: empirical evidence on the larger multipliers in household debt overhang are mixed, while the larger multipliers in firm debt overhang are clear; (iii) in firm debt overhang, positive government spending strongly crowds in both private consumption and investment compared comparing with low firm debt periods, and the multipliers are larger than one; and (iv) in household debt overhang, those crowding-in effects, particularly the effects on private consumption, are not as strong. I investigate various private debt-dependent transmission channels of government spending, including the labor and financial markets. Differences in the behavior of the labor market and investment decision across firm debt overhang and household debt overhang may be important for understanding the empirical results.