14th journées Louis-André Gérard-Varet

Font Size: 
Eliciting preference-based weights for a multidimensional well-being index
Verity Watson, Koen Decancq

Last modified: 2015-03-13


Measuring well-being in a multidimensional framework is more complicated than doing so in a one-dimensional setting. For example, the outcomes in the different dimensions need to be weighted to reflect the relative importance of each dimension to an individual's overall well-being. In this paper, we test the feasibility using a stated preference method - the discrete choice experiment (DCE) - to elicit individuals' preference-based weights for a multidimensional well-being index. We take the Human Development Index (HDI) as an example of a multidimensional well-being index and elicit individual's trade-offs between the three HDI dimensions for two countries: Belgium and Colombia. We use four tests of data quality to assess the feasibility of the DCE method: transitivity of choices, use of compensatory decision making, sensitivity to task framing and plausibility of preference heterogeneity. We find respondents were engaged in the task and express well-defined preferences: almost all pass the transitivity tests and make compensatory choices. We find little evidence of framing effects in responses. We find evidence of plausible preference variation within and across the samples.


Multidimensional well-being index; stated preferences