Government spending multipliers and public indebtedness: an investigation for the US.

This paper examines the state-dependent effects of government debt on government spending mul-tipliers. In the first part of the paper, I estimate the spending multipliers conditional on the level of government debt using US historical data and the two-state direct projection method. For the estimation, a new quarterly US historical government debt series is constructed. Two important empirical results are obtained. First, in contrast to the conventional prediction, the estimated short-run multipliers in a high debt state are larger than those in a low debt state. Furthermore, I do not find strong evidence to support a sharp increase in interest rate and levying heavier future distortionary taxes, potential candidates which offset the stimulus effects of government spending, in response to a rise in government spending in a high debt state. Instead, I find evidence to sup-port that the government may try to stabilize debt levels by limiting debt-financed spending and subsequent spending cuts, so-called spending reversals, in a high debt state. In the second part, I construct a New Keynesian model to explain the large short-run multipliers in a high debt state. The empirical result reveals that spending reversals are observed only in a high debt state. Based on this empirical finding, I introduce a state-dependent government spending rule in the standard New Keynesian model. Specifically, spending reversals are assumed to take place only when the debt-to-GDP ratio exceeds a threshold. The model suggests that the state-dependent spending rule and the interaction between spending reversals and monetary policy could be a potential channel to understand the large short-run multipliers in a high debt state.