Analyses of European integration processes are still dominated by the dichotomous debate between intergovernmentalism and suprana-tionalism. Sides are often taken in this debate based on very case-specific empirical findings. An analysis of two cases, European telecommunications and electricity policy, illustrates how misleading such generalisations can be. The former case is often used to support supranational claims, while the latter highlights the importance of intergovernmental vetos. However, further analysis reveals the very parallel supranational policy approach of the European Commission to both sectors and the resulting weakness of the dominant interpretation. An analysis using a multilevel governance approach to European integration can avoid these difficulties.

This article presents an alternative trajectory to policy paradigm change to that outlined by Peter A. Hall's social learning model, in which unsuccessful efforts by state officials to respond to policy failures and anomalies in the existing paradigm eventually trigger a broader, societal, political partisan debate about policy principles. From this society-wide contestation over policy goals, problems, and solutions, a new policy paradigm emerges. Drawing on the conceptual tools of policy feedback and policy networks, this article describes an alternative route to paradigm shift in which change is negotiated between state actors and group representatives. Discussions of change are largely confined to sectoral policy networks and the result is a more managed series of policy changes that culminate in a paradigm shift. This argument for a second, cumulative trajectory to paradigm shift is developed by examining agricultural policy change in three countries: the United States, Canada, and Australia.

In recent years, issues of environment and human development have been gathered together under the integrative framework of sustainability (sustainable development), creating unprecedented demands on policy. Major policy processes have been constructed and are producing new and supposedly far reaching agreements, policies and strategies. However palatable politically, the likelihood of these leading to significant positive changes in the environment or human use of it, however, is questionable. This situation is sketched, using the Australian setting as an example. The article then establishes in detail the problem attributes particular to sustainability, and argues the crucial proposition that this policy field is different from others in both kind and degree. Some implications of this for policy analysis and prescription are noted.