Gender differences in self-employment: The contribution of credit constraints and risk aversion to self-employment entry, duration, and earnings in Canada

The self-employment rate of Canadian women grew tremendously over 1982-1998. Nevertheless, significant gender differences remain in entry, duration, and earnings. Lin, Picot, and Compton (2000), controlling for observable characteristics, find that women are less likely to enter and more likely to exit self-employment than men. Kuhn and Schuetze (2001) note that decreasing exit rates account for most of the rise in female self-employment. Devine (1994a) finds the female-male ratio of self-employed earnings is 0.46 in 1990 in the United States. Hundley (2000) and Clain (2000) demonstrate that some of this earnings gap is due to household responsibility diversely affecting the sexes. Additional research by Coate and Tennyson (1992) and Evans and Jovanovic (1989), illustrates how differential credit constraints can impact self-employment. In particular, Evans and Jovanovic (1989) show that if credit constraints exist, self-employment outcomes will be positively correlated with wealth. They do not consider gender.;For women, wealth's effect is positive and three times larger than for men. I calculate that if everyone had high investment income the gap in entry rates would fall by fourteen percent. Moreover, a woman with high investment income has a 24 percent greater probability of survival. The impact on men is negative. Conversely, investment income positively affects male earnings but negatively affects those of women. Whether these results debunk gender differentiated liquidity constraints or not, wealth remains an important factor in self-employment disparity. Likewise, children positively impact women's entry, but each added child decreases the survival and earnings capacity of the female entrepreneur. The same is not true for men. These results are consistent with women using self-employment to accommodate household production.;I use Canada's Survey of Labour and Income Dynamics for 1993-1998 to investigate whether gender differences in self-employment are driven by wealth. Although I focus on wealth, I also consider household responsibility as a source of divergence in labour market behaviour. Using regular and random effects probits, proportional hazards models, and selection corrected earnings regressions, I find that wealth plays a dramatically different role for women than for men.