Constant-sum paired comparisons for eliciting stated preferences: a tutorial.

There is growing recognition of the importance of formally including public preferences and values in societal decision-making processes. Constant-sum paired comparison (CSPC), sometimes known as a 'budget pie' task, is a stated preference method than can be used to elicit and measure these preferences and values. It requires respondents to allocate resources between two alternatives, and the relative allocation of this resource is assumed to reflect the importance or priority that respondents attach to the attribute levels in each alternative. CSPC is useful in addressing questions over preferences for the distribution of resources, and allows for an explicit linkage of budget constraints, opportunity costs, outcomes and group characteristics. A key property of CSPC is the ability to allocate some resources to the less preferred alternative, forcing respondents to reflect on the relative value of both alternatives, and possibly giving it an advantage in contexts such as healthcare where respondents may find it ethically difficult or objectionable to make all-or-nothing allocations. This tutorial will outline the theory underlying CSPC, and will work through a detailed example of administering and interpreting a CSPC elicitation, including defining attributes and levels, constructing experimental design, task presentation, and analysis and interpretation.