Anahita Assadi: A Silent Revolution: How digital technologies are transforming public administration through the backdoor

Abstract

In this thesis, I examine how the adoption of digital technologies affects the internal organization of public administration, a cornerstone of democratic governance. The empirical studies address the impact across three key components of administrative organization: core values, the role of public officials, and internal hierarchies. The thesis revolves around a central tension: the political nature of public administration and the often apolitical approach to implementing digital reforms. I argue that while public administration is inherently political, digital transformation is unfolding largely without sufficient political debate or explicit decisions about its broader societal implications.

Through three empirical studies in Sweden, I investigate the introduction and use of AI and digital platforms. The first study examines the tension between efficiency and the rule of law in AI adoption, drawing on frame analysis of parliamentary debates and government documents. It finds that policymakers frame AI primarily as a tool for enhancing efficiency, often overlooking its implications for the rule of law. The second study focuses on the Swedish Social Insurance Agency (SIA) and its use of AI in public decision-making. Interviews and document analysis reveal the rise of ‘agency AI experts’ who drive AI implementation, invoking values such as accountability, transparency, and legal certainty. However, these values are defined narrowly, and efficiency is downplayed, contrasting with previous research. The third study examines how digital technologies reshape professionalism, focusing on an educational platform in Stockholm. Interviews with teachers and principals identify a shift toward ‘platformized professionalism’, characterized by architectural control and trust based in transparency.

Collectively, the studies show that digital technologies deeply reshape core administrative components. Moreover, the adoption of these technologies shifts influence within public administration toward system architects. These developments raise questions about who should make decisions shaping policy implementation and the relationship between the state and its citizens. I argue that policymakers do not sufficiently address these transformations. If democratic governments do not engage with the broader implications of digital transformation, they risk ceding decision-making power to unelected actors like public officials and tech firms, potentially leading to unforeseen political and societal consequences.