Thesis

sence of malaise, which is captured through a 9-item survey on negative emotions and physical responses. For clarity, we invert the latter scale, where zero indicates malaise and nine indicates no signs of malaise, and refer to this as emotional satisfaction. We argue that these five measures are meaningful indicators for individual’s satisfaction in multiple life domains.3 All measures are standardized to a mean of zero and a standard deviation of one. Control variables We also include two control variables. The most important potential confounder is intelligence,4 for which we use a proxy measured at age 10. Intelligence quotients (IQ) are widely used to assess intelligence but require expensive and time-consuming tests. In the BCS70, British researchers developed the British Ability Scales (BAS), which provide a meaningful profile of general cognitive abilities. Our analysis includes the BAS Matrices Scale, a non-verbal reasoning test similar to the Raven Progressive Matrices, which proxies intelligence as a component of Spearman’s g for general intelligence. Cohort members completed this test at age 10, alongside assessments of their reading and math skills. We also control for sex in the regressions. 3Although we make every effort to provide an elaborate overview of individuals’ satisfaction in life, the measures that we select have their limitations. The job satisfaction measure, for instance, is only reported by individuals that have a job. This measure might, therefore, not capture the fact that certain individuals are not satisfied because they do not have a job, while early reading and math skills might relate to the possibility of getting a job. We examine the extent to which our selected domain satisfaction is reflected in the measure on life satisfaction in Figure A.5 in the appendix. 4The literature describes the relationship between intelligence and factors such as academic success in young adulthood and outcomes later in life (e.g., Borghans et al., 2016; Gale et al., 2009; Schneider et al., 2014). 21

RkJQdWJsaXNoZXIy MjY0ODMw