Vad händer i själva verket?: Om styrning och handlingsutrymme i Skolverket under åren 1991–2014
Eva Maria Magnusson har studerat hur olika former av styrning påverkar handlingsutrymmet för olika nivåer i en myndighet – Skolverket.
Eva Maria Magnusson
Anders Forssell, Uppsala universitet Docent Anders Ivarsson Westerberg, Södertörns högskola
Docent Magnus Frostenson, Handelshögskolan Örebro universitet
Uppsala univeristet
2018-06-04
Vad händer i själva verket?: Om styrning och handlingsutrymme i Skolverket under åren 1991–2014
Företagsekonomiska institutionen
Abstract in English
This thesis deals with questions of governance, control and discretion in state agencies. It is grounded on a case study of the Swedish National Agency for Education (Skolverket), from its prelude in the late 1980s, to its start in 1991, through several governments, seven ministers and four direktor generals, up tp 2014.
The empirical part of the thesis follows the periods of the four director generals. The study is based on a vast number of documents, supplemented with interviews. For the analysis, a model of governance and control was condensed from the literature. This model attempts to cover both the political governing of agencies and the internal organizational control of agencies. This model was furthermore developed into three ideal type control styles: Old public management/ Agency, New public management/ Actor, and New public governance/ Arena.
Analysis showed that Skolverket started out as a new type of agency, resembling the ideal type of NPG/ Arena, but over time, with an increasing pressure from the political level, it gradually turned into a more “normal” state agency, at one period resembling the ideal type of NPM/ Actor but finally becoming more similar to the ideal type of OPM/ Agent, however with still strong internal traits of an Arena.
Further analysis showed that the actual governance and control featured several crucial characteristics that proved to be of importance for the amount of discretion left for the directors and officials of the agency. These characteristics were the degree of 1) unified or plural value systems, 2) partial or comprehensive use of control “tools”, and 3) high or low degree of specificity. These three characteristics were combined into eight different types of roles, each with a specific type and level of discretion, from the most restricted (“Authoritarian”) to the most unrestricted (“Laissez-faire”) and six other roles in between. It is proposed that this scheme of roles form a novel way of analysing the level and distribution of discretion in agencies.