Multi-dimensional living standards : a welfare measure based on preferences

We compute a distribution-adjusted welfare measure that aggregates outcomes in three dimensions of well-being, namely income, employment and longevity. Aggregation weights reflect preferences of people on these dimensions. The welfare measure is calculated for 26 OECD countries and selected emerging economies, and covers about three decades. Relying on a single theoretical model of a hypothetical representative agent, we combine life satisfaction regressions to capture the full welfare losses of unemployment with a calibration approach to capture the value of longevity. We test for robustness of results over a series of datasets and specifications and find that the resulting estimated shadow prices of (one percentage point of) unemployment and one year of longevity average 2% and 6% of income respectively. While we assume an identical utility function for all individuals, shadow prices of unemployment and longevity vary both across countries and within countries across income groups. We find that economic growth differs significantly from the growth of our welfare measure. The latter grew faster than GDP thanks to the gains that countries experienced on longevity, but was also more volatile due to changes in unemployment. Rising income inequality exerts a negative effect on our welfare measure. Gains in longevity have almost the same impact on welfare as income growth, while the long-term impact of employment was smaller.