The Learning State: Essays on Public Administration in Representation and Accountability

Arguing that cognition – especially acquiring knowledge – in representative government is underexamined compared to agency in an environment characterized by democratic deficits, de-centered governance, turbulence and social inequity, this dissertation reimagines public administration (PA) as the state’s ‘peripheral nervous system’ in a theoretical perspective of learning through administration from accountability processes initiated by members of the public. Existing accountability research emphasizes top-down, control perspectives, overlooking the potential for accountability’s deliberative and constructive roles, while learning accountability scholarship often focuses on high profile events. I offer a more systematic theorization and exploration of state learning from more mainstream challenges like complaints, appeals and litigation brought by citizens as affected individuals, which I term 'civic challenges' to emphasize them as a form of voice and citizen-state interaction. Drawing links between PA and democratic theories, a theoretical essay conceptualizes civic challenges as noisy democratic data, with administration occupying a unique position which may facilitate learning in representative government through civic challenge stewardship and interpretive responsibility. Three empirical essays follow which further explore and elaborate the perspective with qualitative and mixed methods applied in the context of English and Welsh policing. The first explores how civic challenges are interpreted in police forces. Through abductive analysis of 43 interviews with police insiders I contribute a multi-faceted framework which sheds light on plural, competing normative dispositions towards civic challenges, revealing learning as having ‘proprioceptive’ purposes of protection, stabilization and control through knowledge the public organization acquires about the public and itself in processing civic challenges. Interpretive activities included engaging with complainants and officers, sharing and consulting with internal and external stakeholders and emergent or established theming of data. A mixed-methods empirical essay explores and tests organizational conditions—especially workforce social diversity—favoring or constraining learning as understood by police. Lesson salience and integrity as viewed by frontline officials emerges as a novel learning condition. This highlights that the action-cognition separation inherent to ex-post accountability arrangements is a defining feature of learning accountability, offering a new way of understanding—and perhaps overcoming—the oft-stated but underexamined blame-learning tension in organizational learning theory. Testing bureaucratic-political hypotheses about diversity conditions for learning from discrimination and other complaint types with novel panel data finds greater association of ‘learning’ with higher workforce shares of minorities, attenuated with higher social group fractionalization. A fourth essay (with Valentina Mele and Sonia Ospina), uses interpretive narrative analysis of archival sources and over 50 interviews with civil society, policy experts and police to study an innovative social accountability arrangement. We find orchestration of safe intra- and inter-sector platforms and brokering roles of civil society and public organizations may help to promote collective identification of systemic, social equity issues in public policy and administration, leading to sustained attention if not collective accountability for addressing the issues. Overall, the dissertation contributes a novel perspective on PA’s role in long-growing democratic deficit concerns and offers early empirical insights into how PA may foster learning from civic challenges, an important, underexplored class of accountability pressure and citizen state interaction.