Autonomy of Agencies at the Federal Level in Switzerland

This dissertation analyzes the autonomy and control of administrative bodies outside of the core administration from an economic perspective.  The analysis includes how the autonomy and control of these so-called agencies is structured and which factors influence it.
The paper is part of the EU Research Project entitled COST-Action IS601: Comparative Research into Public Sector Organizations (CRIPO).

Starting Point

The starting point for this dissertation is the more recent scientific literature that shows how in countries around the globe governmental tasks are increasingly no longer being performed by units of a central administration, but rather by partially autonomous organizations – so-called agencies.  One of the chief arguments for this “agency fever” is the expectation that the autonomization of federal authorities makes possible an increase in the effectiveness and efficiency of the administration.  In order, however, to be able to weigh the chances and risks of governmental task fulfillment by agencies, the operating methods of these organizations must first be understood.  This requires, first of all, a definition of the organizational form “agency”.  Secondly, the key concepts of autonomy and control must be conceptualized and measured.  Only then can one discover, in a third step, the influencing factors relevant for this. Up to now, such an analysis has not been available (or has been lacking) in Switzerland.

For this analysis, the autonomy of the Swiss administrative bodies in historical context can certainly be recognized as the distinguishing characteristic.  In addition to the executive, for a long time (formally, to some degree, into the 1960s), so-called “individual agencies” existed precisely on the cantonal level.  Cantonal master builders, cantonal forest officials, fire inspectors, etc., were for the most part directly elected by the parliament and, already at that time, allocated a type of global budget.  The role of the current departments was merely designed to have a support function for the members of government.  The strict departmental administrative hierarchy is therefore a relatively recent structural form.  Also on the level of the federal administration, which was developed on the model of the cantons, the autonomy of the office-holders vis-à-vis the government is still noticeable.  Thus, the Federal Assembly can certainly shift or re-group agencies, but the creation of new federal agencies is the responsibility of the parliament.

This dissertation project must be positioned between international research focusing on the comparison and measurement of autonomy and a national, legal treatment of the topic that focuses on individual cases.  This work, however, goes beyond merely applying the previous international agency studies to the Swiss context. 
Furthermore, the paper initially classifies the concepts using the existing conceptional principles.  Building upon this and using the relevant organizational theories, a Swiss-specific explanatory and impact model – of the conceptional reference framework – is developed for the actual autonomy of agencies on the Swiss federal level (see Fig. 1).

This reference framework is empirically studied by first evaluating the data from the written survey to all federal-level organizations (federal agencies, governmental commissions, FLAG agencies, public-legal organizations, and public corporations, as well as foundations and associations with public tasks).  The findings hereby obtained are then supplemented using the case studies of the three large federal infrastructure companies (the postal service, the SBB (Swiss Federal Railway), and Swisscom).  Furthermore, within the framework of the evaluation mandate of the Auditing Commission of the National Council for regulating the postal service, the SBB, and Swisscom, interviews could be conducted with all involved parties from the companies, the administration and the parliament.
From this emerges an empirically-based explanatory grid on the autonomy and control of federal-level agencies.  Using this precise reference framework, it is finally possible to show fields of action for structuring the autonomy and control of agencies and to formulate additional research questions.

Research Questions

To solve the discussed problem areas, the paper focuses on the following four questions guiding the research:

1. What is autonomy?
2. How can autonomy be measured?
3. Which factors influence the autonomy of federal-level agencies?
4. How can the autonomy of the agencies be explained?

Methodology

The paper uses a “mixed method” research design.  First, using a quantitative approach on a general-abstract level, the autonomy of agencies is measured, the theoretically deduced influence factors are analyzed, and general trends for explaining autonomy are derived from this.  Subsequently, the qualitative study of the controlling of the postal service, the SBB, and Swisscom on a concrete, individual level makes the confirmation of cause-effect relationships possible.  From this emerges a refined description of the influencing variables, and this provides, above all, for additional interpretations on the quantitative explanatory approaches.
The following data are available for the study:

• Written survey of all federal-level organizations on the perception of autonomy and control
• Transripts from the interviews on the evaluation of the controlling of the postal service, the SBB and Swisscom.

The research process is organized as follows (Figure 2):

Findings

This paper develops an empirically based explanatory grid on the autonomy and controlling of agencies at the Swiss federal level. The author achieves herewith two goals. First, testing the hypotheses on the creation of agency autonomy that are derived from organizational theories makes a contribution to science.  The study demonstrates that the strictly rational Public-Choice models, as well as the purely value-based and norm-based approaches (micro-institutionalism, culture theory), have too little explanatory power.  Conversely, the new institutional economy, which also accepts structural elements, e.g., hierarchies or legal forms,  as influencing factors on the behavior of actors, is confirmed empirically. On the other hand, practice-oriented fields of action for structuring the autonomy of agencies on the federal level are identified.  At several points, these fields of action highlight the potential for optimizing the operational controlling of agencies.

This dissertation was written by Etienne Huber under the supervision of Prof. Dr. Reto Steiner.