Individual Submission Summary
Share...

Direct link:

A New Technique for Determining the Authorship of Preferential Trade Agreements

Fri, October 1, 4:00 to 5:30pm PDT (4:00 to 5:30pm PDT), TBA

Abstract

Preferential trade agreements (PTAs) are the outcome of a complex negotiation process between multiple parties. Governments frequently claim that they played a dominant role in writing the agreement and “got what they wanted” in the negotiations. Yet, trade negotiations take place behind closed doors, which means that determining who actually wrote a given PTA, and thus whose interests it best represents, is difficult. Furthermore, commonly used empirical methods used in political science – such as hand coding and standard text analysis – treat these agreements as if they were single authored. Therefore, in this paper we utilize a new text-as-data method to assess patterns of authorship across PTAs. Our novel Supervised Author Topic Model for Treaty Text (SATT) leverages the previous PTAs signed by countries to infer the preferences governments hold over treaty language. From this we can estimate the proportion of a PTA written by each signatory.

We showcase the value of this empirical approach in two ways. First, we apply our technique to a corpus of more than 500 PTAs singed between 1960 and 2018. Regression analyses of authorship contributions demonstrate that economically powerful countries play a dominant role in the treaty-writing process. State power is the primary explanation for PTA content, even after considering and controlling for numerous other explanations. A focused analysis of US PTAs shows that this pattern may be accelerating over time. Furthermore, a survey of Indian PTAs reveals that relative power vis-à-vis other signatories matters greatly, as India writes some PTAs but has terms dictated by partners in others. Second, we apply our approach to the recently-signed Regional Comprehensive Economic Partnership (RCEP). We find considerable support for the proposition that China as well as the ASEAN countries dominated the agreement. Surprisingly, we find that the imprint of other notable signatories is comparatively low. Moreover, we show that the language in the agreement directly reflects Chinese preferences across many old and new trade issues.

These findings help us understand the evolution of trade cooperation and where it may be headed. They affirm that PTAs have become a vehicle through which powerful states compete to “write the rules” of the trade regime. The findings for RCEP also show that even regional powers may fail to have their preferences embedded in PTAs with major powers, whereas they likely would fare better in multilateral negotiations via the WTO. From a policy standpoint, countries like the US will fail to make their imprint on rules governing new trade issues unless they begin negotiating new PTAs soon. Finally, the fact that our findings apply consistently to PTAs across time and space gives us greater confidence that our approach can be generalized to other multi-authored political documents.

Authors