Choice Experiments: identifying preferences or production functions?

This paper presents an alternative perspective on the process by which respondents consider options within choice experiments. Building on the “new” model of consumer demand by Stigler and Becker (1977), it suggests that the attributes within choice experiments are not valued directly, but are used to generate higher level “constructs” (i.e. improvement in the environment) which are then valued. The implication is that what are currently viewed as marginal utilities of attributes are in fact marginal utilities of an environmental outcome mixed with (subjective) marginal productivity of the attribute to achieving the environmental outcome. It is suggested the Hierarchical Information Integration methods may allow one to separately identify the utility and production functions, and identify individual heterogeneity therein