Finding Evidence for Country Specific Double Tax Treaties

We investigate the outcome of bilateral tax competition by analyzing Double Tax Treaties (DTT). DTTs are the crucial means of splitting a tax base between two countries. Yet, there are only few empirical results on the textual design of DTTs. We develop a novel approach to quantitatively analyze DTTs using Natural Language Processing (NLP) and study the heterogeneity of a country's DTTs, as well as the influence of the current OECD Model Convention of 2017 on the negotiation results. We evaluate bargaining outcomes and find that economically powerful countries do not always dominate their counterparts. Furthermore, the OECD Model Convention serves as a supportive framework to mitigate the influence from more powerful countries. Understanding the factors that determine the bargaining outcomes is especially important in the light of current initiatives on the (re-)allocation of taxation rights.