What If the Fire Alarms Are Rigged? The Case of MDB Accountability Offices

• In Event: Democracy, Accountability, and International Organizations

Abstract

Following the Asian Financial Crisis, there were calls for greater accountability at the international development agencies, most notably the World Bank and International Monetary Fund. A variety of ex ante and ex post institutional changes were implemented at both of these organizations and at numerous other regional multilateral development banks (MDBs). This paper will focus on a particularly interesting institutional change: the establishment of internal accountability offices (IAOs) through which civil society groups in recipient countries can bring claims against MDBs. The first of these accountability offices was the World Bank’s Inspection Panel, established in 1993. Over the last twenty years, IAOs have been established at numerous other regional MDBs. Substantively, IAOs represent the first accountability mechanism by which those negatively impacted by MDBs’ development loans gone awry can air their complaints and potentially receive some remedy, possibly impacting the design or implementation of the loan program midstream. This goes to the heart of the many of the complaints concerning both the mottled impact of MDB lending and the broader democratic deficit critiques of international organizations. This paper is concerned with how these IAOs function, and what influences their activities.
Accountability offices are internal to the MDBs and have no enforcement power. Nonetheless, scholars have hopefully pointed to MDB IAOs as promising and potentially successful attempts by state principals to gain the upper hand in their attempt to constrain IO agents (Keohane and Nye 2003; Nielson and Tierney 2003; Grant and Keohane 2005; Buntaine 2015). MDB IAOs are described as fire alarm mechanisms. Communities and individuals to bring complaints when MDB loan projects cause harm and violate MDB policies. The accountability mechanisms can assess these complaints, deem certain loan projects “non-compliant,” and make recommendations for compensation, program changes or institutional changes to address the source of this non-compliance with principal preferences.
While this theoretical story is appealing, it is supported by very limited empirical work (see Buntaine 2015 for an exception). Accountability mechanisms are still often intertwined in important ways with the parent MDBs they monitor, despite institutional design features that are supposed to ensure their independence. This article asks: How independent are the MDBs’ independent accountability mechanisms? Do they ultimately serve their parent MDB’s bureaucratic interests or serve as fire alarms for state principals? Theoretically, the article offers an in-depth analysis of a specific type of fire alarm mechanism and its independence from the agent it monitors. Empirically, the article investigates ways that the MDB can breach the accountability mechanism’s independence, and documents how accountability mechanism activities—decisions about eligibility and about action plans for non-compliant loan agreements—may vary based on this MDB intrusion. These empirics offer a way to observe IO agents, in this case MDBs, actively protecting their bureaucratic interests in less crude ways than simple observations of staff or budget increases (a common observable measure of bureaucratic interest; see for example, Vaubel, Dreher and Soylu 2007).
The paper relies largely on a new dataset of all complaints filed through the nine MDB accountability offices gather by Accountability Counsel (an international non-governmental organization), as well as original coding of the reasons given for ineligibility. The paper will start by providing background information on MDB accountability mechanisms, including how they were designed to serve as fire alarm mechanisms and the institutional design features intended to ensure independence. Then, I discuss four ways that MDB accountability mechanisms may be compromised, and provide broader data on institutional design differences. Next, I present descriptive evidence on MDB IAM activities, both their eligibility decisions and action plans, in order to assess the degree to which their activities are influenced by MDB interests. Finally, I discuss attempts by principals to review and recontract with IAMs in order to empower them vis-à-vis their parent MDBs, before concluding.
The paper offers a complex portrait of the tug-of-war between state principals and international organizational agents, arguing that MDB accountability mechanisms are not fully compromised, but often succumb to pressure from their parent MDBs. This caving can be empirically observed through patterns of their activity, particularly eligibility decisions and action plans. When their independence is breached, MDB accountability mechanisms are not able to fulfill their role as fire alarm mechanisms, including informing principals of agent deviations from principal preferences.

Author

• Erica R Gould, Stanford University