The Politicization of the Climate: How and Why Has the Annex I - Non-Annex I Division Affected Negotiations Under the Climate Change Regime?

Multilateral environmental agreements often institutionalize differentiated rules for groups of countries in order to account for diverse circumstances and contribution to the problem addressed. The United Nations Framework Convention on Climate Change (UNFCCC) implemented such differential treatment by rigidly dividing the world into a group of so-called Annex-I countries with emission reduction targets (OECD and transition economies) and a group of non-Annex I countries without such commitments (all other countries). In this article, we seek to shed more light on the argument that by establishing such differentiation the climate regime constructed new lines of confrontation above the substance-based disagreements between countries, rendering broad cooperation more difficult. Accordingly, the Annex I / non-Annex I dichotomy survived for over 20 years despite drastically changing economic and environmental realities. We aim (i) to measure to what extent the institutional split of UNFCCC parties into these two groups led to a deeper polarization in the climate negotiations, beyond the countries' actual preferences; and (ii), to elucidate whether interest-based or rather socialization-related causal mechanisms led to such division. We draw on a new dyadic dataset recording agreements and disagreements between country pairs at the negotiations over time, hand-coded from written summaries of the climate negotiations between 1995 and 2013 published in the Earth Negotiations Bulletins. We use this dyadic data to conceptualize the climate negotiations as a network of cooperative and conflictive negotiation interactions that continuously changes over time. After an explorative analysis of the network graphs over time, we use a Relational Events Model, which combines survival analysis with social network analysis, to assess the effect of group construction on cooperation and its causal mechanisms. Based on our dyadic dataset we investigate the effect of group membership on cooperative or conflictive behavior between all country pairs. An extensive set of exogenous controls enables us to disentangle the effect of group construction from the effect of countries' background characteristics. In addition, we control for structural characteristics of the negotiations and endogenous patterns of interaction between the negotiation partners by including selected network statistics. The results show that the institutionalized division of UNFCCC parties into Annex I and non-Annex I countries has indeed led to a polarization between these two groups in the negotiations, but that this group effect is stronger for non-Annex I (developing) countries. Parties are more deeply polarized during discussions about the mitigation targets that underlie the Annex I - non-Annex I division. We interpret this as supporting an interest-based explanation for the effect of differentiation. Our results only weakly support the idea that socialization might be driving the group effect.