European Policy Analysis
-liation of advocacy coalition members ( in those studies in which members are identified ), we found that interest groups and government agencies are the dominating actor types in Swedish policy subsystems .
We found no empirical evidence of policy professionals or think tanks in any of the 25 cases . However , we also found that many coalition members are business representatives who participate regularly in the policy process . In summary , our findings provide tentative evidence that policy actors increasingly engage in networking as a strategy for coordinating and obtaining influence in the policy process . At the same time , our findings regarding the organizational affiliation of coalition members would seem to be at odds with a trend toward more open policy process .
Third , our results suggest that many policy subsystems contain interactions between national and local levels of government , which speaks to the notion that policy subsystems are nested across actors operating at different levels of government — yet , they show less evidence of Europeanization . We interpret this finding primarily as a result from the bias in the sample toward issues related to environmental and energy policy and natural resource management , which tend to feed tensions between national regulations and local interests . While this might explain regular participation by local-level politicians in many cases , the lack of focus on Europeanization is noteworthy and something that calls for more research .
In addition to these country-specific patterns , we also ask if and how applications of the ACF in Sweden might suggest refinement and specification of the ACF ? We found that approximately one-third of the applications explicitly acknowledge limitations to the ACF , primarily related to the framework ’ s descriptive validity . Some applications question whether the notion of advocacy coalition is useful to describe actor constellations in the policy process and some offer concrete suggestions for modification of the ACF . It can be noted though that most suggestions for modification involve specifications of ACF ’ s core concepts ( learning , advocacy coalitions , and belief systems ), while no study addresses the utility of the notion of “ coalition opportunity structure ” as a means to support the application of the ACF in corporatist settings . In fact , we found only three studies dealing superficially with opportunity structures but these studies did not empirically examine the role of opportunity structures — including the level of openness and consensus needed for policy change — in shaping advocacy coalition formation , strategies , or policy change . Thus , while the notion of coalition opportunity structures was introduced by ACF theorists to ease application of the framework outside the United States , this specification has received virtually no empirical attention in Swedish applications .
About one-third of the applications reviewed here test hypotheses derived from the ACF . These studies largely support the explanatory validity of the framework when applied in Sweden . Specifically , studies confirm the importance of policy core beliefs as an explanation of advocacy coalition formation , the role of learning in the policy process , and external shocks as a driver of major policy change . Nevertheless , our review suggests that the compilation of prior applications of the ACF in Swedish policymaking is insufficient as a basis for thoroughly assessing the applicability of the framework in Swe-
35