Becoming (un)employed and life satisfaction: asymmetric effects and potential omitted variable bias in empirical happiness studies

Abstract

Becoming unemployed has negative effects on life satisfaction; a transition from unemployment to employment, however, has stronger positive effects in absolute terms. The asymmetry of the nonpecuniary effect indicates a potential ‘omitted variable bias’ in previous empirical happiness studies.

Keywords:

• happiness
• life satisfaction
• asymmetric effect
• labour status
• employment
• unemployment

JEL Classification:

• I31
• J00

Notes

¹ Cf. Blanchflower and Oswald (2004); Powdthavee (2009) and Stutzer (2004).

² The question is: ‘In conclusion, we would like to ask you about your satisfaction with your life in general. Please answer according to following scale: 0 means “completely dissatisfied”, 10 means “completely satisfied”. How satisfied are you with your life, all things considered?’, http://panel.gsoep.de/soepinfo2009/.

³ See: http://www.diw.de/en/diw_02.c.238121.en/changes_in_the_soep_data_set.html.

⁴ The analysis is for the period 1994 to 2008, because this is the only period where all necessary variables are available.

⁵ For both points in time, therefore, people without jobs, part-time workers and the self-employed have been excluded.

⁶ Concerning changes in household income, the equivalence-weighted monthly net household income in t is compared to that in t+1, and the growth rate is calculated.

⁷ For the asymmetric effects of the OLS models A, B and C, see the significant f-tests between Employed → Unemployed and Unemployed → Employed (see Table 1).

⁸ The f-tests in the OLS models B and C between the income associated with the transition to unemployment and the transition to employment show no significant results.