Towards a distribution-sensitive better life index : design, data and implementation

The Better Life Index was introduced by the OECD as a tool to chart the multi-dimensional well-being of OECD member countries, Brazil and the Russian Federation. However, the Better Life Index relies only on aggregate country-level indicators, and hence is insensitive to how multi-dimensional well-being outcomes are distributed within countries. This paper discusses how a distribution-sensitive Better Life Index could be designed and implemented. Based on five concrete recommendations for the design of the index, a family of indices is suggested. These indices are shown to be decomposable in interpretable building blocks. While a rich and comprehensive micro-level data set is necessary to implement the distribution-sensitive Better Life Index, no such data set is currently available for all OECD member countries. The paper proposes a ‘synthetic’ data set that relies on information about macro-level indicators and micro-level data from the Gallup World Poll. The implementation of the distribution-sensitive Better Life Index is illustrated with this synthetic data set. While the small sample size and other survey features of the Gallup World Poll imply a number of potential biases, illustrative calculations based on this synthetic data set indicates that, when taking distribution into account, Nordic countries are top-ranked whereas Greece, the Russian Federation and Turkey occupy the bottom positions. The results indicate sizeable losses due to multi-dimensional inequality for OECD member countries. Moreover, there are large differences in the level and composition of multi-dimensional inequality.