Conceptualizing and Measuring International Bureaucratic Autonomy

Based on a synthesis of relational and sociological aspects, we suggest conceiving of the bureaucratic autonomy of international secretariats as the combination of the (in our terms, ‘sociologically’ grounded) capacity to develop independent preferences (‘autonomy of will’) and the (‘relationally’ grounded) ability to translate these preferences into action (‘autonomy of action’) (see also Caughey et al. 2009; Maggetti 2007).

In order to develop autonomy of will, an administration first requires the ‘administrative cohesion’ to overcome obstacles to collective action and to interact with political actors as a unified organizational entity (see Mayntz 1978, 68). Cohesion enables the bureaucracy to develop a ‘single set of corporate goals’, which allows its members to work toward the same cause (Caughey et al. 2009, 3). If such an ability is missing, ‘pockets of autonomy’ within the lower echelons of the bureaucracy—at the unit or departmental level—are likely to emerge (Cortell and Peterson 2006, 263; Trondal et al. 2012). Such ‘pockets’ restrict the administration’s ability to construct and maintain a common identity and to function as a unified entity working toward the fulfillment of its mandate (see Selznick 1949). The development of an autonomous will also require what we call ‘administrative differentiation’, which refers to the bureaucratic capacity to develop preferences that can potentially differ from those of the political principals.

‘Autonomy of action’ refers to the ability of an administration to translate these preferences into action. Within the nation state, this aspect of autonomy is highest if an administration has ‘a monopoly jurisdiction (that is, they have few or no bureaucratic rivals and a minimum of political constraints imposed on them by superiors)’ (Wilson 1989,

The dimensions of structural bureaucratic autonomy. Source

Fig. 2.1. The dimensions of structural bureaucratic autonomy. Source: Authors’ compilation based on Carpenter (2001), Caughey et al. (2009), and Verhoest et al. (2004)

182). Even though IOs compete with other IOs over competences and resources (Busch 2007), the main power cleavage at the international level is not so much a matter of bureaucratic rivalries but of conflict between the secretariat and the member states. Thus, autonomy of action crucially depends on statutory powers (i.e., formal secretarial competences vis-a-vis political principals throughout the policy cycle) and independent administrative resources (Brown 2010; Hooghe and Marks 2015). Figure 2.1 provides an overview of our conceptualization of bureaucratic autonomy.

This specification of autonomy as an at least partly relational concept raises an important question: from whom is the international secretariat considered to be autonomous? We are interested in determining the potential impact of international bureaucracies on policy-making. Thus, we study the bureaucracy’s autonomy from political actors and, more generally, from politics, throughout the policy-making process. The political actors in question are the member states of the organization and their representatives within the different political bodies of the IOs. In order to study bureaucratic autonomy empirically, we focus on the secretariats of the following 20 IOs:

  • 1. Association of Southeast Asian Nations (ASEAN)
  • 2. European Central Bank (ECB)
  • 3. European Union (EU)
  • 4. International Atomic Energy Agency (IAEA)
  • 5. Inter-American Development Bank (IDB)
  • 6. International Labour Organization (ILO)
  • 7. International Monetary Fund (IMF)
  • 8. International Maritime Organization (IMO)
  • 9. Organization for Economic Co-operation and Development (OECD)
  • 10. Organization for Security and Co-operation in Europe (OSCE)
  • 11. United Nations (UN)
  • 12. United Nations Educational, Scientific and Cultural Organization (UNESCO)
  • 13. World Bank Group (WB/IBRD)
  • 14. World Health Organization (WHO)
  • 15. World Intellectual Property Organization (WIPO)
  • 16. World Trade Organization (WTO)
  • 17. Food and Agriculture Organization of the United Nations (FAO)
  • 18. International Organization for Migration (IOM)
  • 19. World Meteorological Organization (WMO)
  • 20. UN Refugee Agency (UNHCR)

Next, we propose a number of indicators that enable the systematic collection of information regarding the levels and intensities of autonomy. While for ‘autonomy of action’ the indicators can rely on a well-developed body of literature, operationalizing ‘autonomous will’ is more challenging. This is perhaps no coincidence, as the factors focused on here have been more prominent in qualitative works and the empirical basis of ‘cohesion’ and ‘differentiation’ is arguably—unlike competencies and resources— more difficult to observe. We therefore take as our point of departure the observation that administrative structures allow bureaucrats to operate jointly as unified actors to varying degrees, and draw on characteristics of the international secretariat’s structure and staff in order to operationalize ‘autonomy of will’.

Table 2.1 provides an overview of the 10 indicators and how they are related to the different dimensions and sub-dimensions of bureaucratic autonomy (including the assumed link between theoretical dimensions and observations).4

In order to allow for comparison across dimensions and sub-dimensions, the values of each indicator range from 0 to 1. Because an additive approach would increase the weight of those dimensions that consist of several sublevels, we used averaged values to combine sub-level scores.



Name of the indicator and description

1. Autonomy of will 1.1. Administrative cohesion (to overcome problems of collective action)

Organizational centralization

More centralization indicates greater administrative cohesion.

Homogeneity of personnel Greater homogeneity in the national origin of staff indicates greater administrative cohesion. Mandatory internal mobility of personnel High personnel mobility indicates weaker administrative cohesion.

Length of employment

Longer terms of employment indicate greater administrative cohesion.

staffJhq ratio: Ratio of staff (to total staff) working at IО headquarters.

staffjhomogen: Ratio of ten largest nationalities (in terms of staff) to total organizational personnel.

staff_mobility: Degree to which organizational rules enforce internal staff mobility.

  • 1: no mobility rules;
  • 0.5: mobility is voluntary, but explicitly encouraged; 0: mobility is mandatory. staff_permratio: Ratio of staff with open-ended contracts to total number of staff.

1.2. Administrative differentiation (potential to develop distinct preferences)

Independence of administrative leadership

Independent administrative leadership indicates greater administrative differentiation from political principals.

Capacity to provide, collect and process independent information

The capacity to access and process information that does not come from member states indicates greater administrative differentiation.

sg_internal: Share of heads of administration recruited from within the organization. Only the last five SGs are considered.

research: Centrality of research bodies at different hierarchical levels:

  • 1: existence of a research body at the department level (directly below the SG);
  • 0.66: existence of two or more research bodies at the division level (two hierarchical levels below the SG); 0.33: existence of one research body at the division level (two hierarchical levels below the SG);
  • 0: no research body at division level or above.
  • 2. Autonomy of notion
  • 2.1. Statutory powers (to realize autonomous preferences)

Agenda-setting competences of the SG

More agenda-setting power for the SG indicates greater potential to realize autonomous administrative preferences.

Sanctioning competence

More competence for the organization regarding the sanctioning of non-compliant member states indicates greater potential to realize autonomous administrative preferences.

sg_agenda: Degree to which the administrative head is involved in setting the (provisional) agenda for legislative meetings.

  • 1: SG is responsible for preparation of the draft agenda and items cannot be removed prior to the actual legislative meeting;
  • 0.66: SG is responsible for preparation of the draft agenda, but items can be removed (through objection by a single member state, or decision of the executive body) prior to the actual meeting;
  • 0.33: the executive body, not the SG, is responsible for preparation of the draft agenda and items cannot be removed prior to the actual meeting;
  • 0: the executive body, not the SG, is responsible for preparation of the draft agenda and items can be removed prior to the actual meeting, sanctioning: Sanctioning powers of the organization vis a vis its members (see Brown 2010 for a similar measure):
  • 1: autonomous capacity to impose sanctions;
  • 0.66: power to call for sanctions against noncompliant members;
  • 0.33: denial of membership benefits (e.g., voting rights and IO services);
  • 0: only ‘naming and shaming’ by issuing reports or admonitions, (sanctions related to the failure of member states to pay mandatory contributions are not included).



Marne of the indicator and description

2.2. Administrative resources (to realize autonomous preferences)

Size of human resources

More administrative personnel indicates greater potential to realize autonomous administrative preferences.

Independence of financial resources

More independent sources of revenue indicate greater potential to realize autonomous administrative preferences.

staff_perpolicy: Number of total secretarial staff per policy field.

  • 1: the organization employs 1500 staff or more per policy field;
  • 0.66: the organization employs between 1000 and 1499 staff per policy field;
  • 0.33: the organization employs between 500 and 999 staff per policy field;
  • 0: the organization employs less than 500 staff per policy field.

Income: Degree to which the organization can rely on independent sources of income.

  • 1: self-financing;
  • 0.5: mandatory contributions;
  • 0: voluntary contributions.

In case an organization relies on several financial resources, we use the source with the highest share of the budget.

< Prev   CONTENTS   Source   Next >