Political scientists and
We evaluate the TPP in a new way by systematically comparing the recently-released text of the TPP to the language in previous trade agreements that its members have signed. Our text-as-data analyses reveal that the language in the TPP comes disproportionately from US trade agreements. The ten PTAs that most closely resemble the TPP are all US agreements, and the contents of the most controversial chapters in the TPP draw heavily from US PTAs....
...
... we adopt a methodology that computes textual similarity for any two documents – in our case the TPP and each of the 74 existing PTAs – by identifying perfectly-matching text in common word sequences greater than six words. Relying on common 6-grams or greater allows us to preserve important information contained in word order and ensures that simple, everyday phrases do not constitute a match. This approach, and these same metrics, have been used in other social-science applications (e.g. Corley et al., 2011; Ehsbaugh-Soha, 2013). For each PTA–TPP pairing we generate the number of perfectly matching words, which we then divide by the total words in the PTA to produce a “percentage copied” number that is comparable across agreements. We also aggregate these agreement percentages at the country level, to see which of the 12 members was most successful in getting its preferred contents in the TPP.
...
The six most-copied chapters in the TPP (investment, financial services, general services, telecommunications, and safeguards) draw particularly heavily upon past US agreement language. This includes some TPP chapters in which two-thirds or more of an earlier US PTA chapter is copied verbatim. ...
...
The most striking result in Table 2, however, concerns the agreement language on investment. As noted earlier, selected parts of the TPP’s investment chapter have attracted substantial attention and controversy. Investment is widely considered among the most important chapters of the TPP and long has been “a high priority for the USA in its FTA negotiations” (Fergusson et al., 2015: 36). We consider the contents of the entire chapter, not just the subset on ISDS, and find that a staggering 80% of the language from past US PTA chapters on investment is included in the TPP’s investment chapter, as depicted in the left side of Figure 4 (and also Table 2). This is by far the highest percentage for the USA or any other TPP member across any chapter. Indeed, more than 80% of seven US investment chapters are copied verbatim into the TPP’s investment chapter, as shown in the right side of Figure 4. One of the text illustrations in the Appendix, on minimum standards of treatment for investment, also shows this dominance. Furthermore, the amount of text being copied is significant: more than 7500 words from past US investment chapters are written directly into the TPP’s investment chapter – a final piece of evidence that the US “got what it wanted” in this controversial area to an extent even greater than is realized.
Of course, the substantial similarity with U.S. FTAs doesn't preclude there being some TPP text that is adverse to U.S. interests; a trade agreement can say a lot in a very small space. But nevertheless, the U.S. was clearly very influential in writing the TPP.
And yet despite all that influence, the TPP is at least as controversial in the U.S. as it is in other TPP countries. The investment chapter, which the authors note was substantially drawn from U.S. FTAs, is particularly controversial here.
If the U.S. wrote the TPP, why is it so contentious in the U.S.? Because the U.S. is not a single entity, and there is much internal disagreement as to what should be in the text of a trade agreement. Perhaps what we need is a way to compare the actual TPP text with the text of proposals submitted by various interest groups, to help explain who in the U.S. was most influential in writing the TPP and why some groups in the U.S. now object to it.
