Building subjective wellbeing indicators at the subnational level : a preliminary assessment in OECD regions

This paper provides, for the first time, estimates of subjective well-being variables in 373 OECD subnational regions, allowing comparison of various measures of how people experience and evaluate their lives within and across all 34 OECD countries. Different weighting strategies as well as several robustness checks have been carried out to ensure regional representativeness and to provide reliable indicators. The results show that it is possible to obtain robust regional estimates of subjective well-being through the Gallup World Poll for the variables satisfaction with life and social support network. These estimates could be included in the OECD Regional Well-Being Database to provide two additional well-being dimensions measured uniquely with subjective indicators. In addition to these two variables, the paper explores the feasibility of other subjective indicators, either from Gallup or the European Statistics on Income and Living Standards (EU SILC), which are particularly meaningful when measured at the city or regional levels, such as satisfaction with the living environment, satisfaction with commuting time, trust in the political system, and feeling of safety in the community. Finally, a regression analysis is performed to give an insight of the explanatory power of both individual and regional specific characteristics to self-reported life satisfaction. The results show that regional fixed effects capture around 10 percentage points of the variation in life satisfaction, of which 30% can be associated to observable regional characteristics such as mortality rate and air pollution. Furthermore, life satisfaction tends to be negatively affected not only by the individual unemployment status but also by the level of unemployment of the region; in the OECD area, everything else equal, being unemployed is associated with 7 percentage points less of life satisfaction compared to someone that is employed; whereas, an increase of 1 percentage point in the unemployment rate at the regional level is related with 0.4 percentage points less of satisfaction with life as a whole.