In this paper, I develop a new typology of international bureaucracies that aims to capture the potential influence of these institutions on IGO policy-making. Applying this typology to 16 empirical cases, evidence is presented to explain why the different cases under study fall into specific categories of the typology. In order to achieve this objective, the paper consists of four parts. First, the concept of bureaucratic autonomy, which scholars commonly use to study executive agencies within national political systems, is modified in order to fit the working context and internal structure of IGOs. The main argument is that if one wants to capture the potential autonomous behavior of international bureaucracies it is useful to distinguish between their ability to develop autonomous preferences (autonomy of will) on the one hand and the capacity to transform these preferences into action (autonomy of action) on the other. In a second step, (ideal) types of international bureaucracies are created by simply these two components of autonomy. This results in the following four types of bureaucracies. 1) Autonomous bureaucracies (characterized by both high autonomy of will and high autonomy of action), 2) innovation hubs (high autonomy of will/ low autonomy of action), 3) politicized bureaucracies (low autonomy of will/ high autonomy of action) and 4) managers of the status quo (low autonomy of will/ low autonomy of action). In a third step, suitable indicators are developed in order to measure each of the two components and locate the secretariats of 16 IGO within the four-fold typology. Finally, I present empirical evidence that can account for the location of the empirical cases within the two-dimensional autonomy space. By presenting empirical insights into the bureaucratic autonomy of IGO secretariats, this paper intends to contribute to the topic of category 1 of the workshop program.
