Essay on the use of subjective well-being data in economic analysis : an empirical study using developed and developing countries data
This thesis studies the determinants of subjective well-being, with the main focus on the data relating to developing countries. Chapters 2, 3, and 4 use new South African integrated household data to study the determinants of perceived quality of life at the cross-section. The fifth chapter compares the cross-sectional and over time structures of subjective economic well-being for Indonesia, whilst the sixth chapter uses reported well-being data from the British household panel survey to test an 'old' economic hypothesis in a new way. Chapter 2 tests whether the determinants of subjective well-being are the same when comparing poor and rich nations. Using South Africa as a case study, we find from the full sample analysis that in most comparable cases, the coefficient signs of the usual socio-economic factors in the life satisfaction regression equations for South Africa in 1993 are typically similar to that which would have been expected from data in the more-developed countries. However, our subpopulation regressions reveal very distinct life satisfaction patterns by race and region prior to the end of apartheid in South Africa. Chapter 3 analyses the labour market phenomenon in South Africa. We test whether unemployment hurts less in terms of life satisfaction when there is more of it around. After controlling for the relevant socio-economic factors, we find the unemployed's well-being to be significantly and positively correlated with the levels of others' unemployment in the region. Using the South African data set of 1997, the fourth chapter explores the contemporaneous relationship between measures of criminal victimization and reported well-being. We find crime victims to report significantly lower well-being than the non-victims, ceteris paribus. Reported life satisfaction is lower for nonvictimized respondents currently living in higher crime areas. However, we find some evidence that criminal victimization hurts less in areas of higher crime rates. Chapter 5 examines the cross-sectional and longitudinal relationships between objective and subjective economic ladder for Indonesia. It finds that individuals' perceptions of economic rank in the economy are more dependent on his or her socio-economic characteristics (i.e. health, education, marital status), as well as attitudes towards future economic ladder, than the current spending behaviour would normally reveal. The correlation between objective and subjective economic ladder is also weakened considerably when an individual's inborn predispositions are controlled for in the regression. Chapter 6 tests whether one's partner's happiness increases one's own happiness in a marriage. After using "residual" self-rated health to provide an instrument for the partner's life satisfaction and allowing controls on individual fixed-effects, we find strong evidence of an interdependent relationship in the reported life satisfaction between married partners, which is not present for those whom are merely Cohabiting.