Participatory health system priority setting: Evidence from a budget experiment

Budget experiments can provide additional guidance to health system reform requiring the identification of a subset of programs and services that accrue the highest social value to 'communities'. Such experiments simulate a realistic budget resource allocation assessment among competitive programs, and position citizens as decision makers responsible for making 'collective sacrifices'. This paper explores the use of a participatory budget experiment (with 88 participants clustered in social groups) to model public health care reform, drawing from a set of realistic scenarios for potential health care users. We measure preferences by employing a contingent ranking alongside a budget allocation exercise (termed 'willingness to assign') before and after program cost information is revealed. Evidence suggests that the budget experiment method tested is cognitively feasible and incentive compatible. The main downside is the existence of ex-ante "cost estimation" bias. Additionally, we find that participants appeared to underestimate the net social gain of redistributive programs. Relative social value estimates can serve as a guide to aid priority setting at a health system level.