Almost all countries have historically allowed businesses to write off interest expenses against taxable income. Critics argue that the tax-favored status of debt has created a corporate debt pile-up, thereby exacerbating economic downturns. This argument, which gained more attention after the 2008 global financial crisis, implicitly assumes that the tax incentives have led to a large increase in the use of debt. However, despite extensive efforts by researchers, it is an open question whether the tax incentives are indeed a primary determinant of corporate debt policy. This is mainly because isolating the impact of interest deductions from other tax effects is empirically challenging. The Tax Cuts and Jobs Act of 2017 (TCJA) provides a unique opportunity to overcome the empirical difficulties and directly estimate the effects of the tax benefits of debt on firms. In my paper, I exploit this setting and document that the tax advantage of debt has first-order effects on firms’ financing, investments, and employment decisions.
Before the TCJA, businesses could generally deduct interest expense from their taxable income. As part of a comprehensive tax reform, the TCJA limits firms’ interest deductions to 30 percent of adjusted taxable income plus interest income for tax years beginning in 2018. This rule reduces the tax advantage of debt for “high-interest” firms, for which interest payments exceed the specified limit. However, to support small businesses, the act makes an exception for firms with average annual sales below $25 million. This exception creates a discontinuity in the tax benefits of debt for firms on the two sides of the sales threshold, providing the opportunity to do a clean comparison of financial decisions in otherwise similar firms.
I employ a regression discontinuity design, which compares high-interest firms in a narrow bandwidth on both sides of the $25 million threshold. Although limiting interest deductions had been discussed since the first tax reform proposal in 2016, the details and the exception rule were not known until a few weeks before the new tax code went into effect. So, any detected difference across the two groups of firms is likely caused by the differences in the tax incentives.
One of my main findings is an almost dollar-for-dollar relation between corporate debt and the present value of the tax benefits of debt. By construction, my estimates are based on a sample of small and high-leverage firms. For example, an average firm in my sample has around $46 million in capital stock and $38 million in long-term debt. I estimate that, sample firms that are affected by the limitation on interest deductions on average lose $8.1 million in the present value of the tax benefits of debt. In response to this change, my most conservative estimate suggests that the affected firms decrease long-term debt by an average $7.71 million compared with the control group. This means that per one dollar of reduction in the present value of the tax benefits of debt, firms decrease debt by $0.95. However, equity financing is not affected, suggesting that firms do not simply replace debt with equity as the tax benefits of debt shrink.
The reduction in total financing could affect the firms’ real outcomes. My estimates suggest that losing the tax benefits of debt negatively affects firms’ investments and hiring: Treated firms decrease their investment rate by 6.8 percent and their hiring rate by 17.3 percent relative to the control group. This translates into $4.8 million less in investments and 41.52 fewer workers in an average firm that is affected by the limitation on interest deductions.
I consider two potential explanations for the main results. One is that the results could be driven by changes in firm characteristics and not by the reduction in the tax benefits of debt. I rule out this possibility by showing that the treatment has no effect on key firm characteristics that are important for financial and real policies. In contrast, the main findings are consistent with a cost of capital explanation. While limiting interest deductions reduces the incentive to issue debt and thus leads to lower leverage, it also increases the cost of external financing for affected firms. This increase in the cost of capital leads to a decline in new investments and hiring. Also, because equity is generally more expensive than debt, affected firms do not raise equity to replace the decline in debt, which is consistent with a decline in demand for total new financing.
It is important to evaluate whether the main findings could be generalized to large public firms. Given the specific design of the tests that is dictated by the details of the TCJA, it is not possible to conduct the same tests in a sample of large firms. However, to shed light on this issue, I compare the pre-reform magnitude of debt tax shields relative to taxable income in firms of different sizes. The relative size of the debt tax shield monotonically increases in firm size and is nearly four times larger in the largest firms (top 30 percent) relative to the smallest firms (bottom 30 percent). This suggests that the tax benefits of debt are at least as important in the budget of large firms. So, the estimated treatment effects on small firms likely provide a lower bound for the effects in large companies.
Overall, these results inform policy discussions as more countries evaluate the removal of debt subsidies from the corporate tax system. My findings suggest that the policy change is ultimately a trade-off between the negative effects on firm outcomes and the business cycle benefits of having less aggregate corporate debt. The findings also contribute to the field of corporate finance by improving our understanding of the main determinants of corporate financial policies. The lack of clear evidence of tax effects in the previous literature has led many prominent scholars to be skeptical of the trade-off theory of capital structure. My results document a first-order role for tax incentives in shaping corporate policies.
This post comes to us from Professor Ali Sanati at American University. It is based on his recent article, “How Does Removing the Tax Benefit of Debt Affect Firms? Evidence From the 2017 US Tax Reform,” available here.