The idea of path-dependency is applied to an examination of health policy reform in Germany, France, Great Britain and the United States. In the path-dependent model, actors are hemmed in by existing institutions and structures that channel them along established policy paths. Therefore, in any system, big (non-incremental) change is unlikely. However, sometimes we do observe big change. Why? By developing the interplay of structure with conjuncture, the occasional accomplishment of big change - in spite of path dependency - can be systematically understood.

This paper represents a case study in welfare state expansion. It takes an actor-centered point of view and reconstructs the long process which has ultimately led to a compromise solution to the problem of providing long-term care, especially for the elderly. It describes the previously means-tested arrangement of long-term care provision and its shortcomings. Furthermore, it sketches the different stages through which the political debate on this issue has passed during the last twenty years: from the initial phase of defining the “social problem”, to the final stage when the approval of the compromise package became a question of “all or nothing”. The analysis of the politics of long-term care insurance reveals that the difficulties of reaching an agreement mainly resulted from the complex decision-making situation. It was in essence a problem of decision-making under conditions of general uncertainty.

Using Atkinson and Coleman's typology of policy networks, this article shows how many of the differences in policy outcomes can be traced to the structure of the policy environment in each nation. French and Germany policy makers adopted a strategy of investing in high speed passenger transport to revitalize their declining railway sectors. The French TGV was developed in a state-directed policy network which insisted on cost containment and commercial viability. In Germany a corporatist style of policymaking in the rail sector led to delays and higher costs for the ICE train. A separate clientele pluralist network led by the Research Ministry developed the Transrapid maglev option, but in order to finance and deploy an operational system, the Chancellor and cabinet had to create a concertation network. The policy network approach provides a useful framework for conducting comparative analysis. In addition, these detailed cases suggest that it is useful to add a dynamic, cross-temporal dimension to the static typology.

When the European Monetary System was negotiated in 1978, governments in France, Britain, and Italy took very different approaches to this new international institution for coordinating exchange rate policies. The French government actively supported the creation of the European Monetary System, the Italian government entered the system but on weaker terms than the French, and the British government refused to enter the system, preferring to allow the pound to float. To explain these different policy choices, I analyze the impact of domestic politics and institutions on exchange rate policy, paying particular attention to how the organization of bank-industry relations and government instability shape policymakers' policy preferences and their abilities to implement these preferences.