The implementation of budgetary reform.

Abstract

The dissertation's emphasis is on developing a greater understanding of policy implementation. The importance of considering implementation within a policy process framework is emphasized and highlighted by using the policy process to better understand the case considered. That case is the implementation of the Haga Commission budgetary reforms, passed by the Norwegian parliament in March, 1985. An extensive review of the policy implementation literature finds the field split into two major schools, top-down and bottom-up. A model is developed to reconcile these two schools by concentrating on the theoretical significance of ambiguity and conflict for policy implementation. A number of factors crucial to the implementation process are identified as varying dependent upon a policy's ambiguity and conflict level. Four policy implementation paradigms are identified: low conflict-low ambiguity (administrative implementation), high conflict-low ambiguity (political implementation), high conflict-high ambiguity (symbolic implementation), and low conflict-high ambiguity (experimental implementation). For each paradigm the principle of central importance, variables which affect implementation under these conditions, and an example of a policy fitting the paradigm are identified. In addition, the various schools in the existing implementation literature are identified as being most relevant under one of these four set of conditions. The model is tested by studying the implementation of the Haga Commission budgetary reforms within the Norwegian public sector. Seven policy changes are identified in the reform package and each is defined in terms of its level of ambiguity and conflict potential. Based on that identification, predictions are made as to which variables will be of the greatest significance in explaining implementation outcomes. Data for statistically testing the hypotheses were collected at the micro-level from the 105 national agencies implementing the budgetary reforms. A multimethod approach was taken to data collection. Among the methods used were survey questionnaires answered by leaders at 97 agencies, elite interviews with budget directors at 52 agencies, and accounting data from all micro-level agencies. The data provide support for the theory developed. Discussions of future tests are presented.