Qualitative case studies suggest that the outcomes of tax treaty negotiations are determined by power politics and negotiating capability. In contrast, quantitative studies have tended to depart from a model that implies absolute gains, full rationality, and perfect information on the part of both treaty signatories. This paper bridges the gap by replicating two existing quantitative studies, introducing new, more sophisticated data. New fiscal data are drawn from the ICTD Government Revenue Dataset, while treaty content is measured using the ActionAid Tax Treaties Dataset. It finds that developing countries that raise more corporate income tax are more likely to sign tax treaties with wealthier countries, and more likely to negotiate higher withholding tax rates in those treaties, but not more likely to obtain a better negotiated result overall. In contrast, developing countries that raise more revenue in total are more likely to negotiate better outcomes in other clauses of the treaty that are more obscure and technically complex. There is also a strong learning effect, with better outcomes across the board as a developing country gains experience of signing tax treaties. Finally, greater asymmetries in investment stocks and material capabilities lead to worse outcomes for developing countries.
developing countries foreign direct investment corporate taxation double taxation treaties multinational corporations