It is Not as Bad as it Looks: Comparing Countries’ Cooperation on Climate Change Inside and Outside the UNFCCC Negotiations

Traditionally, the United Nations Framework Convention on Climate Change (UNFCCC) has been the main forum to debate global agreements on climate change, with state negotiation delegations as its most visible actors. In recent years, partly due to the inability of the multilateral process to reach solutions, and to growing recognition of the complexity of the challenge, climate change has begun to be addressed across multiple channels outside the official negotiations. So far, there is scant research into the potential interrelationships between both spheres of governance. This study takes a first step in systematically analyzing such interrelationships. We seek to gain a more systematic understanding of the interlinkages between the way countries interact in the UNFCCC negotiations and the way they cooperate in the real world, for example by establishing other bi- or multilateral agreements, exchanging resources, or sharing information. Given that the UNFCCC negotiations are more visible and politicized, we expect to generally see more cooperation in interactions outside this forum. Furthermore, outside cooperation may help to pave the way for more cooperative negotiations. Therefore, we should see that more cooperative UNFCCC meetings tend to follow a period of increased cooperation in the real world, or that groups of countries that cooperate in the real world also increase their cooperation within the negotiations. To explore these patterns, we rely on two novel datasets. The first one records agreements and disagreements between country dyads at the negotiations, hand-coded from summaries of the negotiations published between 1995 and 2013 in the Earth Negotiations Bulletin. The second one is based on automated text coding of press releases, and contains information on cooperative and conflictive interactions between country dyads outside the negotiations between 1995 and 2015. We use a combination of descriptive statistics and dynamic social network analysis to analyze the data.