Embedding as a pitfall for survey-based welfare indicators: evidence from an experiment

Abstract

"Welfare measurement using multiple indicators requires knowledge about how individuals weight different aspects of wellbeing. The better life index provides a way of resolving this individual level weighting issue. It invites people to weight 11 dimensions, embedding measurable indicators. Ideally, the specific embedding of the same indicators should not change their weights in a welfare function. Our experiment shows, however, that varying the embedding structure decisively affects people's weightings, i.e. we observe strong embedding effects. The better life index hence fails to measure citizen's true preferences. Embedding thus opens a gateway to manipulating survey-based welfare measures." (Author's abstract, Published by arrangement with John Wiley & Sons) ((en))

Cite article

Hetschko, C., Reumont, L. & Schöb, R. (2019): Embedding as a pitfall for survey-based welfare indicators: evidence from an experiment. In: Journal of the Royal Statistical Society. Series A, Statistics in Society, Vol. 182, No. 2, p. 517-539. DOI:10.1111/rssa.12410