We examine the impact of potential repatriation taxes on U.S. multinationals' pre-2018 tax planning actions and capital structure policies. We construct a proxy for exogenous changes to repatriation tax rates and find that repatriation-tax-sensitive multinationals are more likely to pursue tax-free repatriation techniques, with examples provided. We also find a positive relationship between repatriation tax increases and both bond issuance and debt ratios, but only for repatriation-tax-sensitive multinationals. These results are concentrated in firms with lower costs of debt capital, while no increased likelihood of inversions is found. Our findings have important implications for current and proposed U.S. tax policies.

This study shows that domestic mergers and acquisitions (M&A) were inhibited by the U.S.’s worldwide tax policy on foreign-earned income. Double Irish structures, a complex web of subsidiaries that reduce foreign tax rates and therefore increase potential repatriation tax rates, are associated with lower levels of domestic M&A by U.S. firms. These results do not reflect a continuation of prior trends or declines in worldwide acquisitiveness and are robust to several econometric approaches. We suggest that the Double Irish variable mitigates a confounding effect in an alternative tax variable that clouds inferences in tests of repatriation taxes and domestic M&A.

Many studies find a negative relationship between acquirers’ stock returns and their size, past acquisitiveness, and performance. This counter-intuitively suggests that large, well-performing, and acquisitive firms are worse-than-average acquirers. We hypothesize that these findings stem from bias related to an omitted variable: the predictability of acquisitive behavior. We theoretically model the direction of this bias for each variable and test these predictions using Heckman's two-stage procedure with a relevant and plausibly excludable variable related to tax avoidance. By mitigating this bias, our approach generates several new insights about acquisition value that are obscured by OLS regressions of deal announcement returns.