Since the 1990s, the US tax treaty network has expanded to include most large developing countries. However, there remains a glaring exception: The US only has two tax treaties in Latin America (Mexico and Venezuela), and one pending tax treaty (Chile). The traditional explanation for why the US has no treaty with, for example, Argentina or Brazil is the US refusal since 1957 to grant tax sparing credits to developing countries. Before the Tax Cuts and Jobs Act of 2017 (TCJA), this explanation was wrong, because the combination of deferral and cross-crediting meant that tax holidays in a source country would not lead to a shift in revenues to a residence country even without tax sparing. This traditional view needs however to be updated given TCJA. On the one hand, there is now an exemption for direct dividends, so that the traditional rationale for not entering into treaties is gone. On the other hand, GILTI means that deferral is abolished and to the extent the income of a CFC exceeds the GILTI threshold there can in fact be a transfer of revenue to the US. Nevertheless, I believe that Latin American countries should enter into treaties with the US, for three reasons. First, cross-crediting still means that even with GILTI there may not be a revenue shift. Second, there are good reasons to enter into treaties even with a revenue shift, such as attracting FDI and limiting tax evasion. Third, Latin American countries are increasingly capital exporters, and the absence of a treaty hurts their multinationals. Finally, now is an opportunity, because the entire US treaty network needs to be updated to take account of TCJA.
Avi-Yonah, Reuven S. "If Not Now, When? US Tax Treaties with Latin America After TCJA." International Tax Journal 45, no. 4 (2019): 51-53.