SOUnD Experimental Public Administration and Management

We address the growing reliance on experimental methods within public management research through three primary objectives. First, we propose the SOUnD typology to systematically classify experimental research based on four dimensions – Setting, Outcomes, Units, and Design – inte-grating and harmonizing existing frameworks to enhance consistency in design, transparency in reporting, and ultimately the validity of inference. Second, we present a systematic review of 315 randomized experiments from 229 public management studies, identifying common trends and areas for methodological advancement within each SOUnD dimension. Third, we explore challenges and opportunities for experimental research, ensuring that public management remains at the forefront of experimental social science. Specifically, we focus on three issues: theory-driven experimental designs, causal mediation, and non-random non-compliance. This work aims to advance experimental research methodology by introducing the SOUnD framework, providing a common foundation for public management scholars and other social scientists to contribute to and refine over time.