Identities and change in public administrations: which interaction over time?

In the name of their modernization, public administrations are confronted to series of change processes, under varied forms and addressed to several aspects of the administration.
One can reasonably suppose that these upheavals cause effects on civil servants, on their relation to the organisation and on the way they see it. Within such a large question, our preoccupation is to question the existence of links between a major planned change within an administration (such as a NPM reform) and its organizational identities. Three hypotheses can be tested:
• the impact: public reforms involve transformation of identities;
• the independence: identities persist whatever the reforms may be;
• and finally the co-structuring: reforms modify partly identities, but these ones shape also the reforms process.
The position we adopt in this paper tends to be inscribed in this third hypothesis: we assume indeed that the introduction of a NPM reform leads to a hybridization of identity logics within public administration, but in other respects the reform process can be eventually influenced by the identity logics.

We illustrate this assumption with empirical data collected in a Belgian Ministry. According to a first qualitative analysis we distinguished 6 identity profiles, characterised by their reference to either one or the other identity logic (public service or public managerialism) and their adopted positioning (fidelity or reality) (see EGPA 2005). 12 to 18 months later, we conducted a second set of interviews with the same sample of people we met before. This leads us to give these first results a longitudinal dimension and to propose an analysis of identity dynamics in a temporal perspective. We thus explain how identity profiles evolve over time. A concomitant analysis of the administration context and the identity changes allow us to reveal the links between identities and the continuous change context in which they are inscribed.

Furthermore, we also proceed to the statistical exploitation of questionnaires sent to the whole Ministry personnel, in order to broadening our empirical basis. Doing this, we attempt to estimate the relative importance of the identity profiles as they are represented within this organization.
Through our case study, our hypothesis of organizational identities and reform process co-structuring appears as valid, as shows the hybridization of identity logics within the administration we studied. Our results allow us to assume a modification of identity points of reference following a reform, as well as a reinterpretation of this reform through the identity filter. Our longitudinal approach reveals the identity dynamics constructed over time, by identifying notably the shift processes from one position to another.