MNC Dividends, Tax Holidays and the Burden of the Repatriation Tax: Recent Evidence

We address two issues: 1. Do dividends from foreign subsidiaries depend on the residual home country tax, and can this be reconciled with existing models? The evidence seems to be inconsistent with both the Hartman-Sinn ‘New View’ and the Weichenreider and Altshuler-Grubert repatriation avoidance models. 2. Does the huge inflow of dividends in response to the 2005 repatriation tax holiday suggest that the burden of the repatriation tax in a worldwide-credit system is very significant? We review the evidence on the negative relationship between dividends and repatriation taxes including new results for the relationship between total foreign dividends and average foreign tax rates at the parent level. The explanation for the negative impact of the repatriation tax seems to be that tax avoidance strategies are not costless, as was assumed by the earlier models, and that the marginal costs rise as the pool of accumulated financial assets grows relative to the subsidiary’s real assets. Subsidiaries in low-tax locations refrain from repatriating longer as the marginal cost of additional deferrals rises to equal the repatriation tax. A recent paper by Grubert and Altshuler suggests that the impact of tax differences on repatriations declines over time and disappears after 25 years. The ‘immature’ stage seems to last a long time. Analysis of 2004 repatriations at the subsidiary level indicates that the parent’s average foreign tax rate is most important to its decision, not the subsidiary’s own effective tax rate or the average effective tax rate in its country of incorporation. Tax planning has made the country of incorporation less significant. The burden of the repatriation tax is a particularly significant issue because it bears on the comparison of exemption versus worldwide credit systems. Past estimates of the burden, including both actual payments and the ‘implicit’ cost of avoiding repatriations, have been modest. Furthermore, it is difficult to identify any effect of the potential repatriation tax on companies’ investment decisions. But this insignificant importance of the repatriation tax has been called into question by the huge repatriations (of almost $400 billion) under the 2005 tax holiday in which companies could repatriate and pay a 5.25 percent tax net of a scaled down foreign tax credit. The paper therefore examines the Treasury company level data for companies’ participation in the tax holiday. There is, however, no necessary conceptual link between participation in the tax holiday and the burden of the dividend tax. The measure of the tax relevant for real investment decisions is the present value of the direct and implicit taxes relative to the returns. Even if that burden is low a mature company with large accumulations may well choose to pay the tax holiday price because of the rising costs of deferrals. Even in a Sinn steady state ‘new view’ equilibrium, a repatriation tax holiday would trigger asset liquidations and large repatriations. A company will repatriate to where the marginal cost of further accumulations is below the 5.25 percent tax price. The reason is the ‘fresh start’ which permits it to save costs on future deferrals. Some of the participants in the tax holiday had very low current repatriation avoidance costs as evidenced by the fact that many had substantial accumulations of ‘previously taxed income’ (PTI) under the CFC rules that they could have chosen to repatriate tax free. As expected, a company’s tax holiday repatriations are a positive function of its accumulated untaxed income and foreign profit margin, and a negative function of its average foreign tax rate, the ratio of its real capital to sales and its accumulated PTI.