Working Paper: NBER ID: w18263
Authors: Florian Heider; Alexander Ljungqvist
Abstract: We use a natural experiment in the form of 121 staggered changes in corporate income tax rates across U.S. states to show that tax considerations are a first-order determinant of firms' capital structure choices. Over the period 1990-2011, firms increase long-term leverage by 104 basis points on average (or $32.5 million in extra debt) in response to an average tax increase of 131 basis points. Contrary to static trade-off theory, the tax sensitivity of leverage is asymmetric: firms do not reduce leverage in response to tax cuts. Using treatment reversals, we find this to be true even within-firm: tax increases that are later reversed nonetheless lead to permanent increases in a firm's leverage - an unexpected and novel form of hysteresis. Our findings are robust to various confounds such as unobserved variation in local business conditions, union power, or unemployment risk. Treatment effects are heterogeneous and confirm the tax channel: tax sensitivity is greater among profitable and investment-grade firms which respectively have a greater marginal tax benefit and lower marginal cost of issuing debt.
Keywords: tax sensitivity; leverage; capital structure; corporate income tax; natural experiment
JEL Codes: G0; G32
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Increase in state corporate income tax rates (H79) | Increase in long-term leverage (G32) |
| Tax cuts (H29) | No reduction in leverage (G32) |
| Within-firm tax increases that are reversed (H32) | Permanent increases in leverage (G32) |
| Increase in state corporate income tax rates (H79) | Heterogeneous treatment effects among profitable and investment-grade firms (D29) |
| Tax considerations (H24) | Significant determinant of capital structure (G32) |