Defense and Non-defense vs Debt: How does defense and non-defense government spending impact the dynamics of federal government debt in the United States?

Abstract

This paper explores the dynamic relationship among defense and non-defense government spending, the government debt, and the output gap in the United States. We estimate structural vector auto-regression (SVAR) models for the full sample (1947:Q1 to 2021:Q1), as well as two sub-samples: (1947:Q1-1980:Q1), during which debt-to-GDP ratio was falling, and (1981:Q1- 2021:Q1), during which the debt-to-GDP ratio was rising. The impulse responses (IRF) and forecast error variance decomposition (FEVD) are computed to analyze the dynamics objectively and systematically. The impulse responses for the full sample and the second sub-sample indicate that non-defense spending responds to a shock to the output gap in a counter-cyclical fashion. Moreover, we find significant evidence of the impact of debt-to-GDP ratio on both defense and non-defense spending. A shock to a debt-to-GDP ratio causes non-defense spending to rise on impact and fall over the forecast horizon in all three time intervals. Defense spending, however, responds differently to this shock. While it is not significantly impacted in the full sample, it decreases in the first sub-sample and increases in the second sub-sample. The results from variance decomposition also show that a shock to debt-to-GDP ratio can explain most of variations in non-defense spending and part of the variations in defense spending. These results confirm that debt to GDP ratio is an important determinant of both defense and non-defense spending especially in the most recent years, as both categories tend to decline in response to a shock to debt to GDP ratio. This reinforces the idea that policy makers should focus mostly on reducing debt and therefore reducing interest payments associated with debt accumulation to have greater flexibility in spending on different categories.