How can we conceptualize and empirically study autonomy across cases and over time? In this paper, we demonstrate how the concept of bureaucratic autonomy, which is used regularly in national contexts to describe the varying potential impact of bureaucratic organization in the interaction between politicians and bureaucratic elites, can be fruitfully applied to the empirical study of international public administrations (IPAs). In a first step, we define bureaucratic autonomy and specify its internal conceptual dimensions. The concept distinguishes between (1) an IPA’s ability to develop autonomous preferences (autonomy of will) and (2) the capacity to transform these preferences into action (autonomy of action). In a second step, we suggest structural indicators that allow for a comparison across organizations and present first empirical results for twenty IPAs. In a final step, we discuss the merits of selected indicators to also capture variation of autonomy over time.
