A sizable part of global trade and investment flows are governed by rules enshrined in free trade agreements and custom unions. These rules, however, can be surprisingly hard to find and complex to navigate. Databases on preferential trade agreements (PTAs) typically only have limited regional coverage or do not provide easy access to treaty full texts. This is despite the fact that PTAs are regulating an ever-larger share of trade and encompass novel disciplines such as e-commerce and investment. From investors and traders to negotiators and policymakers, the stakeholders of PTAs thus struggle to find and comprehend the rules that apply to them. A collaborative project between researchers from the University of Ottawa and the European University at St. Petersburg, the Inter-American Development Bank (IDB), the International Center for Trade and Sustainable Development (ICTSD), and the United Nations Conference on Trade and Development (UNCTAD) is now about to change that.
The RTA Exchange Agreement Text Search Tool provides access to close to 450 PTA texts and allows negotiators and analysts to quickly identify treaties of interest, browse their content and embark on a full text search. The collaboration also creates new opportunities for leveraging artificial intelligence to build new analytics that will make it easier for users to navigate the increasingly complex jungle of trade rules in PTAs.
Preferential trade agreements are becoming increasingly important
While the World Trade Organization (WTO) provides the ground rules of global economic governance, most novel rulemaking today takes place in PTAs. The WTO rulebook has remained largely unchanged since the creation of the organization in 1994. As a result, many policy areas vital for 21st century trade, from e-commerce and regulatory coherence to investment protection, currently fall outside the ambit of the WTO. These so-called “WTO-extra” issues are instead being addressed in hundreds of bilateral or regional preferential trade agreements. And not only the number, but also the length of trade agreements is increasing year by year.
Due to their decentralized nature, coming to grips with these PTA rules is more challenging. There is no central repository for the text of PTAs. Nor do these rules follow common templates. The disciplines on e-commerce or investment found in U.S. trade agreements can thus be very different from those enshrined in European treaties. Those working with or affected by trade agreements thus face a dilemma. On the one hand, rules in PTAs are increasingly important as they regulate novel areas of economic activities. On the other hand, these rules are also more decentralized and diverse, making them harder to find and to understand.
The new RTA Exchange tool makes PTA texts easier to navigate
The new RTA Exchange tool seeks to address this challenge by providing better access to PTA rules. The interactive website allows users to browse through the texts of trade agreements and to zoom-in on specific chapters and articles. A full text search furthermore facilitates the swift identification of relevant terms and concepts. As trade agreements proliferate and become longer and more complex, the tool thus makes it easier to access and review the rules they set.
The RTA Exchange Agreement Text Tool is based on a collaboration between IDB, UNCTAD and university researchers in Ottawa and St. Petersburg. It provides searchable access to the ToTA (Text of Trade Agreements) Dataset – a machine-readable and standardized full text corpus of 448 PTAs notified to the WTO, which is made available on a Github repository. ToTA was created as a shared resource to improve access to PTA full texts and to enable text-as-data research. Early studies using ToTA have found that the PTA universe is based on clusters of similarly-worded agreements that are often, but not always, home to specific regions. These studies also suggest that there are considerable differences amongst issue areas: whereas chapters on transparency or government procurement tend be relatively uniform across agreements, chapters on e-commerce or competition tend to be more diverse.
Towards better analytics of trade rules
The RTA Exchange Agreement Text Tool as well as ToTA are based on the shared insight that technology can make trade agreements easier to navigate for all stakeholders. This is particularly useful for small developing countries that have limited resources for preparing trade negotiations. Providing treaty texts in digital format, however, is only the first step down that road. Now that this machine-readable infrastructure exists we can employ machine learning and artificial intelligence to mine it. It is costly and difficult, if not impossible, for a human to digest the tens of thousands of rules found in trade agreements. But computers do not face the same constraints. By leveraging modern computer science, we can thus produce new analytics, as in the studies mentioned above, that enable human stakeholders of trade agreements to make more informed decisions.
Providing better analytics of trade rules is thus the necessary next step. How can we identify different normative approaches or distill best practices from diverse treaties? How can we track trends and novelties in treatymaking to keep abreast of change? How can we know what treaty design is most trade enhancing and whether agreements can be improved by combining elements today found in distinct instruments? By leveraging artificial intelligence, we can answer all these complex questions quicker and more comprehensively. The RTA Exchange Agreement Text Tool is an important first step on that road towards better trade rule analytics.
Wolfgang Alschner is an empirical legal scholar specialized in international economic law and the computational analysis of law.