Gender Differences in Entrepreneurial Propensity
PHILIPP KOELLINGER*, MARIA MINNITI†, and CHRISTIAN SCHADE‡
* Department of Applied Economics, Erasmus University Rotterdam, P.O. Box 1738, 3000 DR
Rotterdam, The Netherlands. (e-mail: koellinger@ese.eur.nl)
†Department of Strategy and Entrepreneurship, Southern Methodist University, P.O. Box
750333, Dallas, TX, 75275 USA
(e-mail: mminniti@cox.smu.edu)
‡Institute for Entrepreneurship and Innovation Management, School of Business and Economics,
Humboldt-Universität zu Berlin, Unter den Linden 6, 10099 Berlin, Germany
(e-mail:cds@wiwi.hu-berlin.de)
Abstract
Using data from representative population surveys in 17 countries, we find that the lower rate of
female business ownership is primarily due to women’s lower propensity to start businesses
rather than to differences in survival rates across genders. We show that women are less
confident in their entrepreneurial skills, have different social networks and exhibit higher fear of
failure than men. After controlling for endogeneity, we find that these variables explain a
substantial part of the gender gap in entrepreneurial activity. Although, of course, their relative
importance varies significantly across countries, these factors appear to have a universal effect.
JEL classification numbers: L26, J24, J16
Keywords: Entrepreneurship, gender, perceptions, female entrepreneurship
2
I. Introduction
The number of self-employed women in the U.S. has increased significantly (Devine, 1994a,
1994b; Fairlie, 2004). Similar trends are shown in many other developed and developing
countries (Brush 2006), although female self-employment rates vary considerably across them
(Cowling 2000). Yet, across countries, women own significantly fewer businesses than men
(Blanchflower, 2004; Minniti and Nardone 2007). Interestingly, evidence shows that, after
correcting for various factors such as size and sectoral distribution, women’s failure rates are not
significantly different from those of men (Kepler and Shane, 2007; Perry, 2002). Thus, at least a
portion of the difference between genders must be due to the fact that fewer women than men
start businesses.
Ample data exist about self-employed individuals (for example, see Blanchflower, 2004).
Such data allow the analysis of people’s actual employment situations. They do not allow,
however, the important distinction between entry decisions and survival, and do not yield
accurate explanations of possible differences in startup propensity. Thus, previous studies on
gender differences in entrepreneurial propensity failed to consider the proper population of
individuals involved in the actual process of starting a business. Using a large cross-country data
set particularly well-suited to study startup propensity we fill this gap in the literature.
Our study substantiates the role of perceptions in explaining the gender gap using an
econometric approach that controls simultaneously for unobserved heterogeneity and for the
endogeneity of perceptions in the decision to start a business. Although we recognize the
existence and importance of country specific differences influencing the size and causes of the
3
observed gender gap, we show that the lack of confidence among women in their own
entrepreneurial skills is a major reason for the gender gap in all countries in our sample. We see
two possible explanations for this finding. Either, men and women perceive entrepreneurial
opportunities differently. Or, they have objectively different skills and circumstances. Our results
suggest that both explanations are correct to some extent. In addition, we find some evidence that
men and women have different social networks and different attitudes towards failure, factors
that also explain a substantial part of the gender gap in business start-ups.
II. Theoretical background and related literature
Using a large sample of 30 developed and developing OECD countries, Blanchflower (2004)
showed the decision to start a business or become self-employed, for both men and women, to
correlate to several variables such as age, education, work status, and household income.
However, even after correcting for differences in the distribution of these characteristics, as well
as country characteristics, the rates of self-employment have been shown to differ significantly
across gender for a sample of 37 countries in various stages of development (Minniti and
Nardone 2007). Thus, other factors are at play.
A sizeable amount of literature suggests that, when considering new business creation
and self-employment, women face higher opportunity costs then men, primarily because of the
role they play in the family. Several works provide evidence for the United States. Devine
(1994a, 1994b), for example, found that being married with a spouse present and being covered
by someone else’s health insurance increases the likelihood of self-employment for women.
4
Similarly, Lombard (2001) found that a woman is more likely to choose self-employment if her
husband has health insurance, and the greater her relative earnings potential as self-employed
and her demand for flexibility are. Finally, Edwards and Field-Hendrey (2002) found that self-
employment is a more likely choice for women whose fixed costs of work are high, such as
women who have small children.
Several studies have confirmed that the gender asymmetry in opportunity costs of self-
employment is not a U.S. specific phenomenon. Using U.K. data, Cowling and Taylor (2001),
among others, suggested that decisions about employment and marriage, household production
and child-rearing are interdependent. Rosti and Chelli (2005) showed that, in Italy, women are
more likely to enter self-employment from inactivity or unemployment whereas men tend to
enter to improve their long term career options. Georgellis and Wall (2005) found that, for
German women, self-employment is a closer substitute for part-time work and labor-market
inactivity than it is for men, and attribute such differences to the different labor market
opportunities and occupational strategies of women.
Other works have suggested that gender differences in self-employment stem from
discrimination and cultural factors. Clain (2000), for example, found that U.S. women who
choose self-employment have personal characteristics that are less valued in the market place
than women who work full-time in wage and salary employment, while the reverse is true for
men. Kuhn and Schuetze (2001) found the increase in the number of self-employed Canadian
women to result from their relative disadvantage in paid employment as opposed to men whose
increase in self-employment resulted from a secular deterioration of the labor market. Honig
(1998) showed that Jamaican women receive substantially lower returns to self-employment than
5
men. And, using a panel of Dutch entrepreneurs, Verheul and Thurik (2001) showed that an
important component of the gender difference in self-employment may stem from women’s
ability to obtain financing.
1
Although there is evidence that all these factors are somewhat important within and
across countries, their combined power in explaining gender differences with respect to self-
employment remains unclear (Parker 2009). In addition to these economic explanations,
behavioral differences related to gender-specific perceptions and preferences may be important.
Gneezy et al. (2003) run experiments in Israel and provided evidence that men and women have
different attitudes toward competition. Using a sample of Swedish students, Bengtsson et al.
(2005) provided evidence that gender differences exist in self-confidence and optimism. Dohmen
et al. (2010) found different risk attitudes among German men and women. Finally, and
importantly, Croson and Buchan (1999) provided evidence that some gender differences in
preferences hold even across cultures.
2
As Casson (1982, 14) put it, ‘The essence of entrepreneurship is being different – being
different because one has a different perception of the situation.’ In other words, an important
difference between an individual starting a business and one who is not doing so is that the
former perceives a business opportunity where the latter does not. A potential implication is that
1
The evidence on financing is inconclusive. Considering the U.S., for example, Buttner and Rosen (1988) found
evidence suggesting that women face modest penalties in commercial financing, whereas Cole and Wolken (1995)
found that the gender asymmetry disappears when sectoral distribution, age, and size of the business are corrected
for.
2
Croson and Gneezy (2009) provide a comprehensive overview of gender differences with respect to preferences
and, in particular, report evidence showing that men tend to be more self confident than women in a variety of
domains.
