Predicting Public Sector Accountability: From Agency Drift to Forum Drift

TLDR

Principal‑agent theory has dominated public sector accountability research, framing the field around the concept of drifting agents such as independent agencies, opaque institutions, or street‑level bureaucrats. The article aims to expose the limits of principal‑agent theory as a predictive model of accountability evolution and to lay groundwork for more accurate theories. The authors explain discrepancies by linking descriptive studies of accountability to classical principal‑agent concerns, offering theoretical suggestions to reconcile observed forum drift. Empirical evidence reveals that accountability research frequently encounters drifting principals or forums that fail to hold agents accountable, rather than the expected drifting agents.

Abstract

Principal-agent theory has been the dominant theory at the heart of public sector accountability research. The notion of the potentially drifting agent—such as independent public agencies, opaque transnational institutions, or recalcitrant street-level bureaucrats—has been the guiding paradigm in empirical accountability research. The aim of this article is first of all to signal the limits of principal-agent theory as a predictive model of how accountability evolves. A string of findings in accountability research shows that we are in practice often not dealing with the envisaged problem of drifting agents (or actors in accountability terminology). Unexpectedly, we encounter recurring reports of drifting principals, or more accurately forums, which mysteriously choose not to hold their agents accountable. The article puts forward possible reasons for the observed discrepancies encountered in public accountability research, by identifying why such situations are at odds with the model's assumptions, as well as theoretical suggestions on fruitful ways to go forward. In the end, this study seeks to provide building blocks for theories of public sector accountability with an improved predictive capacity. This is done by connecting descriptive studies of the multifaceted character of accountability to "classical" principal-agent theory concerns about agency and control.