Happiness and environmental quality

Subjective wellbeing — happiness — is of increasing interest to economists, including environmental economists. There are several reasons for thinking that environmental quality (EQ), defined as high levels of environmental goods and low levels of environmental ‘bads’, will be positively related to happiness. Quantitative evidence on this remains limited, however. Some papers use cross-sectional data aggregated at country level, but it is open to doubt whether these aggregated measures reflect individuals’ real EQ exposures. Other papers use individual-level data, but in general have spatial data at very coarse resolution, and consider a limited range of EQ variables, exclusively around individuals’ homes. This thesis reports two related strands of work. The first designs, implements and analyses data from two new cross-sectional surveys. It builds on earlier work by using spatial data at very high resolution, and advanced Geographical Information Systems (GIS) techniques; by simultaneously considering multiple EQ characteristics, around both homes and workplaces; and by investigating the sensitivity of results to the choice of happiness indicator. The second strand develops and implements a new methodology focused on individuals’ momentary experiences of the environment. It extends a protocol known by psychologists as the Experience Sampling Method (ESM) to incorporate satellite (GPS) location data. Using an app for participants’ own smartphones, called Mappiness, it collects a panel data set comprising millions of geo-located responses from thousands of volunteers. EQ indicators are again joined to this data set using GIS. Results of the first strand of work are mixed, but support some links between happiness and the accessibility of natural environments, providing quantitative (including monetary) estimates of their strength. The second strand demonstrates that individuals are significantly and substantially happier outdoors in natural environments than continuous urban ones. It introduces a valuable new line of evidence on this question, which has great potential for future development