Purpose
– The purpose of this paper is to test whether the usual positive effect of self-employment on job satisfaction remains once the greater autonomy and flexibility afforded by self-employment have been factored in, as well as the existence of differences in unobserved characteristics across individuals.
Design/methodology/approach
– It may be thought that the probability of being self-employed and the declared job satisfaction are not independent from each other due to differences in unobserved characteristics – as psychological or personality traits – across individuals. Therefore, self-employment should be treated as an endogenous variable when it is introduced as an explanatory variable in a job satisfaction equation. Given this, the paper proposes the estimation of a treatment effect model in which self-employment and job satisfaction equations are estimated jointly.
Findings
– The results suggest that the usual positive effect of self-employment on job satisfaction is due to the greater work autonomy afforded by self-employment, and not to the greater willingness of the self-employed to report higher levels of satisfaction. Thus, the paper finds that once flexibility and autonomy are considered, the usual positive effect of self-employment on job satisfaction disappears and becomes negative.
Research limitations/implications
– It would be useful further empirical analysis using other data, especially panel data, to test the robustness of the results.
Originality/value
– The paper proposes an alternative way to analyse the relation between self-employment and job satisfaction by taking into account both the greater autonomy and flexibility afforded by self-employment, as well as psychological or personality traits.
The Corrections of Errors in the Estimations on the Nonstandard Workers Scale by an Empirical Analysis of Bogus Self-employment in Korea
