Deadlines and Commitments: The Effectiveness of and Compliance with Agreements on Air Pollution in Europe

Abstract

Since the 1970s, a group of mainly European states have sought to reduce emissions of air pollutants through cooperation under the Convention on Long-range Transboundary Air Pollution (CLRTAP). Focusing on these efforts, this PhD dissertation aims to answer two overarching questions: 1) What explains states’ (non)compliance with international agreements on air pollution? 2) How might states best design institutions and treaties targeted at reducing such pollution? While my theoretical point of departure is the two main theories in the international cooperation literature, the management school and the enforcement school, I also develop and test hypotheses derived from other theories and perspectives. I rely on a variety of mutually reinforcing analytical tools and data-collection methods to answer my research questions. One of my four articles is an in-depth case study; another uses a multimethods approach; and two use statistical tools only. My main findings may be summarized in five points: First, the management school cannot explain much of the rather widespread noncompliance with CLRTAP protocols. My evidence lends no support to the propositions that treaty ambiguity and lack of state capacity explain noncompliance. However, unexpected changes between commitment and deadline may explain some of the noncompliance with the 1999 Gothenburg Protocol. Second, neither does my evidence lend much support to the enforcement school’s explanation of noncompliance. Specifically, my in-depth study of Norway’s noncompliance with its 2010 target for nitrogen oxides (NOx) under the Gothenburg Protocol shows that Norway from 2007 onwards implemented costly policies that did not maximize Norway’s net private benefit. Third, the explanatory power increases substantially if we allow some synthesizing of the management and enforcement schools. Building my case study of Norway’s noncompliance with its NOx target under the Gothenburg Protocol, I develop the “deadline-pressure hypothesis,” which can be seen as a hybrid of the management and enforcement schools. The evidence suggests that a logic of consequences was the more important driver of Norway’s NOx policies as long as the protocol’s deadline was distant; in contrast, a logic of appropriateness grew stronger as the deadline was approaching. Fourth, other perspectives contribute to explaining compliance with CLRTAP protocols. Arguing that the preferences of office-holding politicians determine policy strength, the “office-incumbent” hypothesis is consistent with much of the evidence from the case study of Norway’s NOx noncompliance. It was not until an environmentalist party gained considerable influence over NOx policies that stricter NOx policies were enacted. Cross-national emissions data seem, however, to contradict the office-incumbent hypothesis: After 2005, NOx emissions reductions were even stronger in all the 10 other Northern and Western European Gothenburg parties than they were in Norway. Moreover, state capacity may explain (non)compliance with CLRTAP protocols, but not in the (positive) way hypothesized by managerialists. Using several measures of state capacity, I find a negative effect on compliance with five protocols under the CLRTAP. I argue that such a negative effect is less counterintuitive than it may seem; indeed, it may actually be expected among reluctant states that pursue policy goals correlating negatively with compliance. Finally, the influence of European Union (EU) law on air pollution is substantially stronger than that of CLRTAP regulations. My assessments show that the 1988 Sofia Protocol had little, if any, impact on NOx emissions. In contrast, a series of EU directives caused considerable emissions reductions in EU member states in the post-agreement period (1989– 1996). Because the EU relies on both managerial and enforcement measures to enhance the effectiveness of its legislation, this finding is consistent with the “hybrid” theory.