Research Method
The projects draws from recent advances in computer technology to reveal hitherto undiscovered structures running through international economic law. In particular, this project will use state-of-the-art text-as-data analytics and network analysis to investigate patterns of convergence and divergence in international trade and investment law. Our research will focus on:
- the universe of free trade and investment protection agreements and
- the network of investment awards and trade panel/Appellate Body reports
Revealing the structures of convergence and divergence in both areas of international law will help policy-makers and negotiators, in particular in developing countries, to better understand the current structures of international economic law and to help evaluate its need for reform.
State-of-the-art text-as-data
Text-as-data approaches are derived from natural language processing and information retrieval. On one end of the spectrum, text as data may simply denote the use of computers to do what humans used to do: count terms following a human-made coding scheme. This automation of information retrieval drastically reduces the cost of doing research. Large bodies of texts can be assessed in minutes or even seconds. Moreover, once a code retrieving legal information is written, it can be easily adjusted to capture different elements or extended for application to new contexts. Its versatility and low costs thus replace, at least to some extent, the traditional human coder. On the other end of the spectrum, text as data also comprises the use of statistical tools to extract information without human supervision. Having computers run coding schemes accelerates traditional empirical research, but does not revolutionize it.
Social Network Analysis
Social network analysis is only now beginning to be applied to the study of law more generally. A small but growing number of academics have begun to study networks of legal operators, cases, statutes, and articles to describe those networks and examine their evolution and possible effects on doctrine. Our project examines case-specific tribunals as “nodes” and investigates outward and inward citations as “ties”. The information will allow us to determine the properties of the investment case law network: What is the “density” of this network, and how “centralized” is it? Are some cases more central in the network than others? Is there an evolution in the number of ties, structure/centrality of the network itself, or centrality of the specific nodes within the network?