Diffusion of International Law: A Textual Analysis of International Investment Agreements

What are the design, evolution, and effects of the international investment agreements currently in practice?

Project Summary

Investment is the field of international law with the highest number of intergovernmental treaties. While some model agreements have been highlighted, there is substantial variation of obligations across treaties and over time. Today, the available databases of International Investment Agreements lack comprehensiveness, which hampers the study of the design and effects of these agreements. Existing collections of texts, while useful, only provide partial information. Moreover, they are only available in the original language negotiated by parties and exist in formats that do not lend themselves easily for textual analyses.

This project was made up of two poles. First, the team created a new database of investment treaties in one single language and format. Second, the information of this database was used to push the research frontier by applying advanced statistical methods and computational linguistics to contribute to on-going debates in international law, international relations and international economics. Some delays have been encountered in compiling the database, mainly due to the difficulty in finding missing IIAs and some technical difficulties, which, in turn, delayed the use of the database for research. Some results have been put forward in the meanwhile, which can be found on the right-hand bar of this page. One example of such results is the textual analysis of African bilateral investment treaties and the power relationships shown by those. The paper, that can be found here, shows that while African nations, in general, are rule-takers in North-South relations, South-South relations are more interesting, as small island-states like Mauritius manage to set the terms of investment agreements, while powerful states like Egypt fail to dominate negotiations.

Academic Output

Executive Summary

First and most importantly the project aims to create a new database of investment treaties that complements existing endeavours, including integrated texts of BITs, and the relevant parts of PTAs, DTAs, and GATS commitments in one single language and format. For that purpose, besides the extraction of data from the treaties, advanced analytical techniques were used, including the digitization and machine translation of treaties, modelling and development of algorithms for the purpose of information extraction, with a ‘research friendly’ software interface. This new database is aimed to assist scholars and practitioners alike to study the similarities and differences of legal texts over time and across treaties. Second, with the information obtained from the database, the project attempts to push the research frontier by applying advanced statistical methods and computational linguistics digital tools to contribute to on-going debates in international law, international relations and international economics. We aim to analyse the degree of “boilerplating” that is occurring and how treaty texts diffuse. In addition, we aim to map the structure of the network of investment treaties and measure the distance between the current bilateral system and an eventual multilateral agreement. New treaty design measures is also considered to be examined under outcomes such as investment flows. Finally, it is envisaged to examine in more details the outcomes of investment treaty negotiations and studying the learning and diffusion process throughout the whole network of agreements

Working Paper

Diffusion of International Law: A Textual Analysis of International Investment Agreements (IIAs)

This project aims to create a new systematic database on international investment agreements, called EDIT (electronic database of investment treaties). Treaty texts are collected, digitized and non-English texts translated through machine translation software. A dictionary is being developed and the texts are prepared in a single format for research purposes. The database will be useful for researchers and practitioners interested in the design, evolution and effects of investment treaties.

Working Paper

Missing investment treaties

This article presents the situation of ‘Missing Investment Treaties’ (MITs), defined as those International Investment Agreements (IIAs) that have been concluded by States, but their text (and in some cases their existence) are not publicly available or are incomplete.

In order to determine the number of MITs, we examined the text and language availability of IIAs concluded by countries, that are available in public databases, and we complemented that information with country-specific search from international, governmental and private sources.

In turn, the article explores possible explanations to this State’s behaviour, using the following questions as guidelines: Why would countries sign agreements that are supposedly negotiated to promote, protect or liberalize foreign investment without making those texts available? Why would countries make available only some parts of IIAs? Is there any reason behind the language availability of an investment agreement? If a text is publicly available, does it correspond to the language of both contracting parties, only one of them, or of a third country? Is it possible to achieve IIAs’ objectives if the text of the treaty is not available, is partially available, or is available only in one language?

Working Paper

Rule-takers or rule-makers? A new look at African bilateral investment treaty practice

Who are the rule-takers and rule-makers in the African BIT universe? Using computational measures of textual similarity this paper provides a nuanced empirical answer to this question. First, we find that African states tend to be rule-takers in North-South relations, yet enjoy greater agency in negotiations of South-South BITs. Only few African countries, however, use their greater say in intra-African negotiations to include public policy exceptions in BITs. Indeed, recent North-South BITs contain more policy space than their Southern counterparts in Africa. Second, rule-makers and rule-takers also exist within the African South-South BIT landscape. Yet, in contrast to North-South relations, negotiation outcomes seem to be shaped more by expert knowledge than by power asymmetries. Powerful states like Egypt fail to dominate negotiations, while small-island-state Mauritius with its strategic investment policy agenda succeeds in setting the terms of investment agreements. This paper thus provides a more nuanced view of the African treaty landscape, points to new areas of research and highlights the importance of technical expertise in achieving coherent treaty networks.

 

Article

Uncovering the secrets of the investment treaty universe

One could think that the universe of international investment agreements (IIAs) has few surprises left to offer. After hundreds of investment tribunals have interpreted its obligations and thousands of scholarly contributions have analyzed its provisions the contours and content of the IIA universe should be well understood. Yet, have you heard about the revision of the Finnish model bilateral investment treaty (BIT) in the late 1990s or Japan’s decision to radically redesign its investment treaty template in 2002? Did you know that the United Kingdom’s treaty network is twice as consistent as that of Egypt or Pakistan? Have you noticed that Israel copied from British treaties to design its own BIT program or that 81 per cent of the Trans-Pacific Partnership’s (TPP) investment chapter is the same as the investment chapter in the United States–Colombia free trade agreement (FTA), concluded ten years before, in 2006? This is only a selection of the IIA universe’s surprises that we discovered as part of our mappinginvestmenttreaties.com project. In that project, we treat investment treaty texts as data following an approach similar to what is employed in plagiarism detection software. Through our analyjsis we seek1 t0o equip policy-makesrs, practitionerfs and resea1r1chers with a more sophisticated understanding of the IIA universe and better open-access web-based tools to analyze it. In this note we explain the main concepts behind the research, outline our method and showcase the key findings.

Article

A requiem for the Trans-pacific partnership: something new, something old, and something borrowed?

On 4 February 2016, after almost seven years of negotiations, the Trans Pacific Partnership Agreement (‘TPP’) was signed by 12 negotiating countries. The TPP was then labelled by all signatory countries as a ‘new’, ‘high standard’, and ‘21st century agreement’. However, the ratification process of the agreement was stalled and most likely in a definitive way, after the United States decided to withdraw from the TPP in January 2017. Before regretting this development, looking back to the halt of the ratification process of the TPP one can ask how much innovation this treaty really had and the usefulness of mourning the failure of having a TPP agreement, either in terms of future usage of TPP text, or in terms of political relevancy. This article aims to describe the level of novelty of the TPP, specifically in comparison with existing trade and investment agreements between TPP signatory countries, notably the United States. For that purpose, we have focused on the core disciplines of the agreement that were highlighted as novelty parts of the TPP, or that generated debate during the negotiation of the treaty. As a benchmark, we have compared the texts of the previous treaties concluded between TPP signatory states, with the TPP chapters on investment, government procurement, regulatory coherence, sustainable development, intellectual property, cross-border trade in services, telecommunications, electronic commerce, competition, state-owned enterprises, small and medium-sized enterprises (‘SMEs’), transparency and anti-corruption. The article concludes that the TPP was largely ‘Made in America’ — the same country that triggered its demise — as the structure and content of the treaty clearly follow the texts of previous agreements concluded by the United States. However, the influence of other TPP signatories is also perceived in the final text, notably Australia, Canada, Chile and Peru. We also conclude that some parts of the TPP were not particularly novel for signatory countries, as the treaty built on existing trade and investment agreements, offering a consolidation of commitments already present in treaties in force between TPP signatories. However, the TPP also delivered innovation, by including certain disciplines that have not been traditionally established in preferential trade agreements (like regulatory coherence and e-commerce) and others that have benefited from a larger development compared to existing agreements (like intellectual property and sustainable development). Both consolidation and innovation features can be useful for a TPP 11 or for future preferential trade agreements.

Article

Building a Corpus of Multi-lingual and Multi-format International Investment Agreements

In this paper, we present an on-going research project whose aim is to develop a new database of international investment agreements that complements existing endeavors. In particular, this paper describes our efforts to build a standard- ized corpus of multi-lingual and multi-format agreement texts in order to enable re- searchers in the fields of international law and economics systematically investigate investment treaties.

Other Output

G20 International Investment Agreements and Guiding Principles for Global Investment Policymaking

The objective of this research is to map the international investment agreements (IIAs) concluded by Group of Twenty (G20) countries, examining whether there is a common thread between those countries with respect to investment protection and especially the G20 Guiding Principles for Global Investment Policymaking. For that purpose, the most important provisions on investment protection were examined in all available IIAs across all G20 countries. These include the scope and definition of investment and investors; the existence of establishment or pre-establishment commitments; standards of treatment and their limitations; and the main standards of protection, like expropriation, transfers, and umbrella clauses. Finally, we examined the presence of investor-state dispute settlement (ISDS) provisions and their limitations. At the same time, we mapped the presence of the Guiding Principles for Global Investment Policymaking in existing IIAs concluded by G20 countries.

Other Output

Automatic Labeling of Articles in International Investment Agreements: Using Semi-Supervised Learning and Word Embeddings

International investment agreements (IIAs) are international commitments amongst contracting parties to protect and promote investment. Although each treaty has a distinctive structure in terms of placement and organization of information, IIAs as instruments of international law share underlying textual and legal structures. Treaty articles are important components in IIAs, with some articles titled and others untitled. In order to understand and analyze the treaty structure thoroughly, assignment of titles to each article is crucial for content analysis. In this master thesis, we attempt to automatically assign titles to untitled international investment treaty articles using semi-supervised learning. Various titles have been assigned to similar texts due to the variability of negotiating partners, languages, traditions, etc. Hence, in order to have a condensed representation of various article titles, we firstly cluster 34,524 titled articles into ten topics by expanding word and document semantics with embeddings. We then use these ten classes as the labels in our classification task where titles are assigned to 10,074 untitled articles. The classification task is performed with supervised classifiers (k-nearest neighbors (KNN), support vector machine (SVM), multi-layer perceptron (MLP), stochastic gradient descent (SGD), convolutional neural network (CNN)) and partially supervised k-means clustering. Expert annotations of 100 untitled articles are used as the gold standard in our evaluation set. K-means clustering with the retrained word embeddings tailored to our corpus has brought about an increase of 30% in accuracy compared to a simple CNN classifier, which has scored the highest amongst all supervised classifiers. The comparison between these two machine learning paradigms (supervised and semisupervised learning) leads to the conclusion that word embeddings can effectively expand the semantic features for words and documents, which allows us to perform accurate categorization of texts from closely related sub-fields of one research area, for instance, to categorize the ten topics in the study of IIAs.

Research Team

Manfred Elsig
Coordinator
University of Berne

Eddy Bekkers
Co-Coordinator
University of Berne

Rodrigo Polanco Lazo
Co-Coordinator
Universidad de Chile

Todd Allee
Principal Member
University of Maryland College Park

Wolfgang Alschner
Principal Member
Graduate Institute Geneva

 

Rios Annette
Principal Member
University of Zurich

Peter Egger
Principal Member
Eidgenössische Technische Hochschule Zürich

Mascarell Laura
Principal Member
University of Zurich

Douglas Nelson
Principal Member
Tulane University

Magdalena Plamada
Principal Member
University of Zurich

Dmitriy Skougarevskiy
Principal Member
Graduate Institute Geneva

Kyoko Sugisaki
Principal Member
University of Zurich

Martin Volk
Principal Member
University of Zurich

Roberto Echandi
Associated Member
World Bank Group

Andrew Newcombe
Associated Member
University of Victoria

Status

completed

Disciplines

SDGs

Policy domains

Regions

Countries

All countries

Host Institution

Coordinator

Co-Coordinator

Year