AI Technologies and the Reconfiguration of Discretion in Street-level Bureaucracy

In this chapter, we review scholarly research on AI technologies in street-level bureaucracy and introduce the concept of ‘algorithmic discretion’ to define a new technology-supported paradigm in street-level discretion. We suggest that algorithmic discretion drives three major trends in public administration which affect key values and operating principles of street-level bureaucracy negatively: (1) the de-skilling of street-level bureaucrats, (2) the prevalence of “automated inequality”, and finally (3) challenges related to rendering algorithmic decisions accountable, transparent and explainable in a technological context where the complexity of models and the opacity of their in-built biases and flaws are such that explanations for decisions are ever harder to give. Through a critical examination of these issues, this chapter offers insights into how algorithmic discretion has reconfigured street-level work during the last two decades and discusses the implications of these changes for future street-level bureaucracy research.