Despite unparalleled financial and human capital investments, the quality, safety, and value of US health care remain suboptimal.1 There is general agreement on the scope and implications of these trends, but far less consensus on the fundamental drivers of health system underperformance. The disconnect stems from the fundamental difficulty in identifying the causes behind complex policy problems. In biomedical research and clinical investigation, problems and questions are identified, hypotheses are generated, and data are collected; causality is built into the investigative process Similar assumptions are often applied to policy analysis, but the reality is inherently less precise.