1. Embedding as a pitfall for survey-based welfare indicators
- Author
-
Clemens Hetschko, Louisa von Reumont, and Ronnie Schöb
- Subjects
Statistics and Probability ,Economics and Econometrics ,Index (economics) ,media_common.quotation_subject ,050109 social psychology ,OECD Better Life Index ,300 Sozialwissenschaften::330 Wirtschaft::336 Öffentliche Finanzen ,0502 economics and business ,welfare measurement ,Econometrics ,Economics ,ddc:330 ,0501 psychology and cognitive sciences ,Beyond GDP ,I31 ,050207 economics ,Function (engineering) ,media_common ,Structure (mathematical logic) ,Public economics ,Data manipulation language ,05 social sciences ,Individual level ,survey-based welfare indicators ,Weighting ,C83 ,B41 ,embedding effect ,Embedding ,Statistics, Probability and Uncertainty ,300 Sozialwissenschaften::330 Wirtschaft ,Welfare ,C43 ,Social Sciences (miscellaneous) - Abstract
How can we assess the welfare of a society, its evolution over time and predict its change due to particular policy interventions? One way is to use survey-based welfare indicators such as the OECD Better Life Index. It invites people to weight a variety of quality of life indicators according to their individual preferences. 11 broad dimensions aggregate these indicators. Our experiment shows that people do not provide consistent ratings across differently labelled dimensions that embed the same indicators. They also do not adjust the rating of equally named dimensions changing sets of indicators. These results show that survey-based measures might suffer from strong embedding effects and, as a result, may fail to measure citizens’ true preferences for the indicators.
- Published
- 2017