Abstract

In the context of multidimensional measures of well-being, a key question for policy is whether particular groups have differing priorities and are therefore likely to react differently to given economic or social shocks. We explore this issue by presenting the results of two related analyses that suggest positive answers on both counts. First, we apply reference class weights to unique data on adult capabilities in the UK and show that relative weights vary across some groupings. Furthermore, in some cases, deprivation rankings of groups vary depending on which weights are used. Second, we explore possible behavioural consequences of different weights by examining the extent to which groups respond differently to three economic and social shocks (unemployment, widowhood and ill health). In this case, we find that weights and responses vary noticeably with age and region and sometimes with gender. We conclude that whilst equal weighting may be practically unavoidable when constructing indices of welfare in the absence of information on weights, their estimation from survey or experimental data is likely to be justified and may change views about policy needs or efficacy