James Hollway, Assistant Professor of International Relations/Political Science, together with Karolina Milewicz and Duncan Snidal from the University of Oxford and Claire Peacock from Université Laval, have recently published a paper in the Journal of Conflict Resolution on the widening scope of the nontrade agenda in preferential trade agreements (PTAs). Using longitudinal network modelling, the authors find that endogenous cost considerations are the most significant factor explaining the inclusion of nontrade issues into PTAs. More details with James Hollway.
What gave you the idea to write this paper?
The idea for this paper was Karolina Milewicz’s, the first author, who had established the coding of the data on nontrade issues in trade agreements. I was brought on to the project for my methodological expertise in statistical network modelling, and once the team (together with Claire Peacock and Duncan Snidal) was assembled the paper began to take theoretical and methodological shape. Our observation was that trade agreements are increasingly incorporating nontrade issues, but our thinking was that there may be cost-based dependencies explaining why nontrade issues, once introduced into the trade regime, then diffused.
Can you describe the methods you use and your main findings?
The paper uses network analysis, especially a statistical network model called a stochastic actor-oriented model (SAOM). Our application is innovative within political science for applying a multiplex or multilevel SAOM to understand how trade-only and nontrade-issue-inclusive trade agreement networks “interlock” or depend on one another, and also how bilateral and plurilateral PTAs interlock. By modelling how these networks interlock or depend on one another, we can understand how the costs associated with nontrade issues are responsible for parties already committed to them demanding they be included in their successive PTAs.
Did you come across heterogeneous effects?
One particularly intriguing effect was that dyadic “power” asymmetries, however we measured them, were not statistically significant. “Power” mattered, though only as a monadic effect associated with powerful actors preferring bilateral and avoiding plurilateral trade agreements. We didn’t make too much of this at the time, since our argument was about cost-based mechanisms of issue diffusion, but I’m exploring power and networks more closely in a new project.
What are the policy implications of your work?
I think our research speaks to general developments in the network of PTAs. Our cost-based argument suggests that while nontrade issues are costly and difficult to introduce to the trade regime, once introduced, parties already committed to nontrade issues in one agreement will demand or entice similar commitments from successive partners. Another policy implication concerns our analysis of how the structures of bilateral and plurilateral trade agreements depend upon one another. We find that dependencies between trade-only and nontrade-issue-inclusive bilateral and plurilateral trade agreements show that parties with bilateral or plurilateral trade agreements without nontrade issues generally tend to move to the same type of trade agreements with nontrade issues, but that additionally there is a pathway from bilateral trade agreements without nontrade issues to plurilateral trade agreements that include nontrade issues. This clarifies how different types of trade agreements depend upon one another.
Is there potential for future research on this topic?
Absolutely. International cooperation is replete with the kind of dependencies we explore here. Empirically, Jean-Fréderic Morin (Laval), Joost Pauwelyn and I have published a paper that looks at the exploration and exploitation of environmental norms – a specific type of nontrade issue – in trade agreements, and there is more on the way from that project. Methodologically, I have developed together with Christoph Stadtfeld (ETH Zürich) a multilevel version of our dynamic network actor model (DyNAM) that allows more precise inference on network dependencies than SAOMs where the data on ties is time-stamped. Theoretically, as I said earlier, I am starting work on a new project that explores the relationship between power and networks more closely, prompted by observations in this and other papers.
* * *
Full citation of the article:
Milewicz, Karolina, James Hollway, Claire Peacock, and Duncan Snidal. “Beyond Trade: The Expanding Scope of the Nontrade Agenda in Trade Agreements.” Journal of Conflict Resolution 62, no. 4 (April 2018): 743–73. doi:10.1177/0022002716662687.
* * *
Interview by Nayantara Sarma, PhD candidate in Development Economics.
Illustration by gerasimov_foto_174/Shutterstock.com.