DO TAXES AFFECT CORPORATE DEBT POLICY? Evidence from U.S. Corporate Tax Return Data

This paper uses U.S. Statistics of Income (SOI) Corporate Income Tax Returns balance sheet data on all corporations, to estimate the effects of changes in corporate tax rates on the debt policies of firms of different sizes. Small firms face very different tax rates than larger firms, and relative tax rates have also changed frequently over time, providing substantial information to identify tax effects. The results suggest that taxes have had a strong and statistically significant effect on debt levels. For example, cutting the corporate tax rate by ten percentage points (e.g. from 46% to 36%), holding personal tax rates fixed, is forecast to reduce the fraction of assets financed with debt by around 3.5%. Since small firms normally rely much more heavily on debt finance yet face much lower tax incentives to use debt, the estimated effect of taxes would be strongly biased downwards without controls for firm size. Address. Corresponding author: Young Lee, address: Korea Development Institute, P.O. Box 113, Chongnyang, Seoul, Korea 130-012. Email: [email protected]. Telephone: 82-2-958-4071. DO TAXES AFFECT CORPORATE DEBT POLICY? Evidence from U.S. Corporate Tax Return Data Roger H. Gordon and Young Lee The realization that the tax deductibility of interest but not dividends creates an incentive for corporations to increase their use of debt finance dates back to Modigliani and Miller(1963). Surprisingly, however, economists have had great difficulty providing evidence that taxes in fact affect debt/asset ratios. To test for the effects of taxes on firms’ financial policy, one needs sufficient variation in tax incentives either over time or across firms. Most empirical studies, e.g. Auerbach (1985), Bradley, Jarrell, and Kim (1984), Graham (1996), Graham, Lemmon, and Schallheim (1998), Graham (1999), Gropp (1997), and MacKie-Mason (1990), have focused on cross-sectional variation in corporate tax rates across firms to test for tax effects. While virtually all publicly traded firms would face the same statutory tax rate in a given year if they earn anything approaching a normal rate of return, marginal tax rates can vary across firms when some firms have tax losses. Similarly, firms with unusually large deductions for depreciation are more likely to end up with tax losses in the future, for any given use of debt currently. These empirical studies then test to see if firms with tax loss carryforwards or large “nondebt tax shields” have less debt. Using cross–sectional variation in expected tax rates, however, identifies the effects of taxes only if the underlying reasons why some firms have tax losses or larger depreciation deductions do not themselves affect the firm’s choices for debt vs. equity. It is difficult to make this case. For example, recent investment not only generates large depreciation deductions but also provides good collateral for debt. Similarly, firms with tax losses often face cash pressures, leading them to borrow more to cover current operating expenses. It is not surprising, therefore, that the coefficient estimates for tax loss carryforwards and nondebt tax shields often have the wrong sign or are statistically insignificant. A more direct way to test for tax effects would be to look at financial policy over time, as tax rates vary, so that identification is based simply on statutory changes. Unfortunately, tax rates have not varied that much historically in the U.S. For example, from 1951 to 1986, the top U.S. corporate tax rate varied only from 46% to 52%, making it difficult We would like to thank Jim Poterba, the two referees, as well as seminar participants at the University of Maryland and Korea Development Institute, for comments on a previous draft. 1 In recent years, firms face a marginal tax rate much below the top marginal corporate tax rate only if their profits are under $75,000. If a firm with $350 million in assets, the average for Nasdaq SmallCap Market Companies in March 1998, were to earn less than $75,000 in a year, its rate of return would be only 0.021%, whereas a typical ratio of taxable income to assets would be around 2% in the 1990’s. 2 When a firm has tax losses that it cannot use to offset taxable profits during the previous three years, it must carry these losses forward in time, hoping to offset them against future profits. Current interest deductions then become less valuable, since the resulting tax savings occur only in the future, when and if the firm earns sufficient taxable profits. 3 Some of these studies simply include these indicator variables directly. Others calculate expected tax rates as a function of this information, and then test for effects of these variables indirectly, through their implications for expected tax rates.