Of words and deeds: Latin American declaratory regionalism, 1994–2014

We would also like to indicate that the statistical results we show can be replicated using the coding and database available at the following Harvard Dataverse link: https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi%3A10.7910%2FDVN%2FUCWKGC.

Abstract

The idea of an integrated Latin American region goes back to the early post-independence period, and yet, in substance, Latin American regionalism has remained far behind its stated aims. The perceived implementation gap has raised the question why policymakers continued to talk about something they appeared to avoid in practice. This article contributes to the debate on Latin America’s integration gap by exploring the phenomenon of declaratory regionalism - the practice of referring to the region and its institutions in political speeches. Based on quantitative text analysis of the speeches presidents delivered annually at the UN’s General Assembly between 1994 and 2014, we show that this practice has not been uniform. Presidents distinguish between different forms of regionalism, integration and cooperation, and frame the geographical region they refer to accordingly. In motivating presidents to speak about integration as opposed to cooperation, ideology and democratic performance stand out as crucial factors.

Acknowledgements

The authors would like to thank the Millennium Nucleus for the Study of Stateness and Democracy in Latin America (RS130002), supported by the Millennium Scientific Initiative of the Ministry of Economy, Development and Tourism of Chile, as well as the Kellogg Institute for International Studies (University of Notre Dame) for their generous support. The Robert Schuman Centre for Advanced Studies (European University Institute) also contributed greatly to this project through the publication of an initial working paper. In addition, the authors would like to acknowledge Olivier Dabène, Andrés Malamud, Detlef Nolte, José Antonio Sanahuja, and Carsten-Andreas Schulz, as well as four patient reviewers, for critical comments that helped to sharpen the arguments put forward here. All remaining errors are our own.

Notes

We would also like to indicate that the statistical results we show can be replicated using the coding and database available at the following Harvard Dataverse link: https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi%3A10.7910%2FDVN%2FUCWKGC.

¹ These include Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Cuba, Dominican Republic, Ecuador, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Panama, Paraguay, Peru, Uruguay and Venezuela.

² For contending views see Sanahuja (2007, 31) and an early manifestation from Nye (1968, 856).

³ On the use of the data see Voeten, Strezhnev, and Bailey (2009); for a well-known study see Kim and Russett (1996).

⁴ The speeches were retrieved from the United Nations Bibliographic Information System (http://unbisnet.un.org/). All words were identified and counted using WordStat. We reviewed all references to ‘integration’ and ‘cooperation’ manually to include only those referring to regional cooperation and integration. Thus, phrases such as ‘The system of human rights protection should be revised, particularly in the field of family, children, education and international cooperation’ (from the Uruguayan speech, 1999) were not included.

⁵ The year 2000 needs to be seen as an exception given that all speeches revolved around the establishment of the Millennium Development Goals by the UN.

⁶ We thank one of the reviewers for pointing this out.

⁷ Few IR scholars have used count models. The most frequently used statistical model, linear regression, makes the problematic assumption that underlying continuous processes generate observations that are also continuous. The true relationship in our dependent variables is not linear, and a linear approximation would not, in most cases, even be a reasonable working assumption (King 1988, 845).