In this large multinational panel study of corporate and personal income tax changes, the authors show that leverage is strongly affected by changes in relative tax rates, except in countries where tax evasion is high. The 2003 dividend tax cut in the United States, which was not widely anticipated, is shown to have significantly reduced the use of leverage in companies with a higher proportion of individual investors (i.e., companies paid out more of earnings as dividends rather than as interest payments). An important finding is that tax changes significantly influence leverage when taking account of other alternative explanatory influences. Interestingly, changes over three-year spans indicate that leverage decreases quickly following tax cuts, whereas it increases relatively slowly after tax increases. In addition, tax changes in neighboring countries do not directly affect leverage in a focal country.

2016 is an election year in the United States, which means that there is potential for changes in the political parties of the leadership in Congress and the White House that might affect corporate and personal tax rates. The results of this research are meaningful in predicting macro trends in corporate leverage and in identifying companies that would be more likely to change their borrowing and dividend policies.

The results in the study are economically significant. For example, a 1% increase in corporate taxes increases leverage by 0.41%. Personal interest tax and dividend tax increases have much smaller effects on leverage, although the effects are still significant at –0.17% and +0.10%, respectively.

The increase in corporate taxes affects top taxpayers the most. The authors test unprofitable firms and non-tax-paying firms separately. Profitable (tax-paying) firms adjust their leverage much more than non-tax-paying firms; in fact, the adjustments go opposite ways in the two subgroups.

The study covers 29 OECD countries and the years 1981–2009, during which the authors identify 184 corporate tax changes, 99 personal tax changes, and 199 personal dividend tax changes. The authors also control for many well-documented variables that explain leverage and apply propensity score–matching models for robustness in addition to their multivariate regressions.

To facilitate comparisons among countries and accounting conventions, the authors define leverage strictly as interest-bearing debt and exclude financial leases (which are sometimes omitted from balance sheets) along with other certain liabilities. Moreover, they analyze leverage based on book values that are adjusted for certain accounting conventions, such as intangibles and accounting rules that vary across countries and time.

Prior research showed that taxes affect the corporate use of leverage. But there are lingering concerns about causality (e.g., do changes in taxes lead to changes in leverage, or is something else driving both?). By directly studying changes in tax rates, the authors arrive at a stronger causal determination. Specifically, they include firm-fixed effects to remove purely cross-sectional correlations between taxes and leverage, allowing them to focus on changes (shocks) to tax rates as drivers of changes in leverage. Moreover, prior studies focused on individual components of taxation, whereas this study includes all relevant corporate and personal taxes and their changes.

This article contributes a broader sample and an improved methodology to an established stream of research in the relationship between taxes and corporate borrowing. The study relies on changes in statutory tax rates to more clearly claim causality, and it is an updated and more robust empirical confirmation of the predictions from Miller’s 1977 tax equation (Journal of Finance). The usefulness to practitioners comes from the predictions, both macro and company specific, that accompany changes in tax laws. If the Republican Party gains a majority in Congress and wins the White House in the 2016 elections, the corresponding expected reductions in corporate taxes would have a macro deleveraging effect, which would be felt more strongly if dividend taxes fall more than interest taxes. There is also the possibility of complications from changes in tax law to encourage the repatriation of earnings currently parked abroad by US companies, which could have a considerable compounding effect on deleveraging.