Comparative ethnography : 3 observant participations on the reshaping of government.

Our investigations target different processes of government reshaping in France (agencification, democratisation and public-private partnerships), within three policy venues (a regional health agency, a regional council, an European think tank) and the research settings differ significantly vis-à-vis financial dependence and research goals. Comparing contrasted ethnographic practices will control the subjectivist bias as it highlights common epistemological and methodological issues. Participant observation ensures researchers depart from conventional wisdom and ready-made concepts (Becker, 1998). Moreover, ethnographers have access to unattainable data which is particularly relevant for policy analysis. Beyond textual analysis of policy outcomes, researchers can follow the ongoing development of policy-taking (grey literature, backstage discussions/interactions, power struggles…). More, ethnography observing what makes sense for natives leads to unexpected research directions (Evans-Pritchard, 1973). However, we contend that a detached objectivist stance fails to grasp informative elements that the social actors do not want to express for various reasons (Favret-Saada, 1981). A degree of indigenisation helps to be trusted in the social milieu we worked on and in. Compared to quantitative analyses or expert-to-expert interviews, the added value of ethnography rests on the “embodied experience of the investigator” (Cefaï, 2010). Therefore, we explore “narrative ethnography” (Tedlock, 1991). Being simultaneously observers and actors (trainee in a PR unit, associated research in an operational department, expert in a think tank), we address the shift from participant observation to observant participation, which requires to study the heuristic disturbance we create (Devereux, 1967) and the transformation of our Self gone native (Dalby, 1983; Wacquant, 2004).