Is Female Entrepreneurship Only Empowering for Single Women? Evidence from France and Germany

Entrepreneurship has been suggested as an alternative career model for women to gain economic empowerment while maintaining caring obligations. In this study, we investigate how gender and living situation affect entrepreneurs’ engagement in their business, home, well-being and business success in both France and Germany. Data from the European Social Survey were used, which included 470 French and 622 German self-employed people. For the French, women reported more working hours when living alone but there were no gender differences for the other living situations. For the Germans, there were no gender differences when the self-employed person lived alone; for the other living situations, men reported more working hours. Women reported working more household hours than men in both countries. There were no gender differences in life satisfaction for German self-employed people regardless of living situation; for the French, gender differences varied by living situation. Men reported more business success than women in both countries. Results suggest that self-employed people in Germany follow a traditional breadwinner model, whereas in France, self-employed women do more paid and unpaid work at the same time. In sum, entrepreneurship may only be empowering for self-employed women living alone.


Introduction
Women constitute an increasing proportion of the labor force in most industrialized countries.This trend holds true for entrepreneurship (Drew and Humbert 2012), as there has recently been a large increase in the number of female entrepreneurs (Huarng et al. 2012).As a rising number of women start new businesses, considerable efforts have been made to analyze the impact of gender on entrepreneurship (Braches and Elliott 2016;Henry et al. 2016).Entrepreneurship has been suggested as an alternative career model for women; self-employment is seen as a pathway for women to gain economic empowerment while maintaining family or other caring obligations.Although entrepreneurship is portrayed as meritocratic and a gender-equal career path with no formal entry requirements, a gender bias within the entrepreneurial discourse has been demonstrated (Henry and Marlow 2014).Hagqvist et al. (2018) suggested that reproducing traditional gender norms seems to be rather pronounced among self-employed people.They stipulated that work is still highly gendered and time demands pose the biggest threat to experiencing work-to-life interference, which affects well-being negatively.The idea that women enter entrepreneurship to accommodate family obligations affects female entrepreneurs negatively.Many women-operated firms reflect feminized working patterns, such as working part-time or basing the business within the home, which, as a result, reinforces the perception of women as mothers and care-takers first, and undermines the credibility of female entrepreneurship (Braches and Elliott 2016).In this study, we want to shed light on the merits of female Despite the claims of female entrepreneurship as an empowering career trajectory, existing evidence indicates little support for this idea.In this study, we investigate how gender and living situation affect entrepreneurs' engagement in their business (i.e., work hours), home (i.e., household hours), well-being (i.e., life satisfaction), and business success (i.e., the number of employees, household contribution).This will shed light on the question whether entrepreneurship really is a way to female empowerment, and whether success for women and men might differ.We study these factors in two specific countries, France and Germany, which have a unique combination of social norms and governmental policies.Before looking at the individual associations, we first introduce the context of the two countries of study.

Context: Germany
We chose to study the comparison of Germany and France as both countries provide a unique combination of social norms and governmental policies.In Germany, attempts have been made by the government to increase the number of women in top positions.However, only 3% of management board seats are occupied by women and 12% in supervisory boards of the top-200 companies (Holst et al. 2012).In Germany, 75% of women between the ages of 20-64 worked in 2017 (Statistisches Bundesamt destatis).This is the 3rd highest rate in the EU.At the same time, the birth rate is decreasing (Statistisches Bundesamt destatis).Germany places fourth when it comes to entrepreneurship out of opportunity instead of necessity, which speaks to the economic climate in the country (Sternberg et al. 2017).A strong economy with low unemployment and a steady increase in individuals' economic situation has led to relatively high opportunity costs for starting a business (Bosma and Kelley 2019).Women account for only 4.3% of the population that is self-employed without employees whereas men come up to 5.6%.The gap increases when we look at those self-employed with employees, where men account for 5.7% and women only for 2.3%.However, the gap is equally large in France with 2.2% and 5.8% men in the group of self-employed people with employees (OECD 2019a(OECD , 2019b)).
Traditionally, German social policies emphasize the relevance of marriage, and in many families in Germany (former East and former West), the norm is that men are the main breadwinner and women are the primary caregivers.This trend is evident as data show that German women are more likely to interrupt their careers for taking care of a child.Where 82-85% of fathers work continuously until their child reaches legal maturity, only 32% of mothers with children under three years worked and the percentage stays low with 67% working mothers until the child turns nine (Wirtschafts-und Sozialwissenschaftliches Institut 2018).Parental leave benefits are relatively extensive in Germany, and it is not uncommon for women to take three years off.Parental leave policies in Germany are modelled following the Scandinavian example; if both parents take at least two months' parental leave each, the overall paid leave for the family increases from 12 to 14 months.Those "couple months" were introduced together with other changes (e.g., making the parental leave payments dependent on the income of the adult claiming the benefits) in 2007.The numbers indicate that fathers indeed increased their parental leave from 3.5% in 2006 to 35.7% in 2015.However, the vast majority (79.9%) of male parents only took the "mandatory" two months in 2014 (Geis-Thöne 2018).Particularly the introduction of the so called "parental leave benefit plus", which was introduced in 2015 and allows parents to stay at home for longer (28 instead of 14 months) with reduced benefits, indicates a move towards traditional ideas of long stay-at-home periods combined with part-time employment (Statistisches Bundesamt destatis).This new form of parental leave benefits was chosen by one-fifth of mothers but only 8.2% of fathers in 2016 (Statistisches Bundesamt destatis).Childcare reforms in 2013 introduced an entitlement of subsidized childcare for children younger than three years, even though the demand could not be met by some states, causing conflict, temporary solutions with individualized daycare solutions, and an encouragement directed at mothers to stay home.Despite the political efforts to increase men's uptake of childcare, Germany can be considered a conservative country with a strong societal pressure to conform to traditional gender roles (Bennhold 2010).The 2015 introduced and subsequently unconstitutional declared cash-for-care benefits for parents abstaining from using child-care facilities are another indicator (Blum et al. 2018).

Context: France
The percentage of women working in France in 2017 was 66.7%, which places the country 17th in a European comparison (Eurostat 2017).A total of 30% of entrepreneurs are women, and only 13% of the total of French owner-managers with more than ten employees are women-owned businesses (INSEE 2012).Female entrepreneurs in France are also more prevalent in sectors in which they traditionally dominate as employees (Vial and Richomme-Huet 2017).According to the National Entrepreneurship Context Index (NECI) created by the GEM, France ranks 10th out of 54, suggesting that the environment is relatively conducive to entrepreneurship (Bosma and Kelley 2019).Germany is ranked 19th.
In France, the birth rate, on the other hand, is the second-highest in Europe (Vial and Richomme-Huet 2017).Governmental policies aim to increase gender equality, family-supportive policies and wide participation in employment.With only 16 weeks of maternity leave (generally 6 before the birth and 10 after) and 11 days of paternity leave, France's parental leave policy certainly suggests that parents are encouraged to return to work quickly (CAF 2018).Those 16 weeks are rather short compared to the potential 14 months of paid parental leave Germany is offering.In France, there has been a public discussion to increase the paternal leave to at least four weeks of mandatory leave to allow the other parent to take a more active role in childrearing (Willsher 2017).In France, the rate of mothers looking for work or returning to work before the child's first birthday is at 80% (Ananian 2010).This is partly enabled by the widespread use of governmental support through social contribution exemptions and tax deductions of household services.This sector of household services employs roughly 1.2 million people (Morel 2012).Given the widespread part-time and multiple-employer solutions in this industry, the number of households served is significantly higher.In 2005, 2.6 million households declared legal expenses for domestic services (Farvaque et al. 2013).The French system promotes a model of the family in which the woman occupies the role of the working mother.In sum, Germany and France take opposite approaches when it comes to parental leave incentives and policies (within the European Union), with Germany aiming to facilitate gender equality by following the Scandinavian model with extended parental leave options for both partners, yet based in a traditionally gender-stereotypical culture (at least in the West), and with France aiming to facilitate gender equality by having mothers come back to work quickly.
In the following, we will go through theory, empirical evidence and our propositions of the effect of gender, country and living situation on work hours, household hours, life satisfaction and business success.

Gender and Living Situation on Work and Household Hours
The idea that time allocation is different between men and women can be explained with (Eagly and Wood 2012)'s social role theory.Social role theory suggests that men and women occupy different social roles because of physical sex differences, which produce a division of labor.As a result, people infer typically male and female dispositions and associated gender roles, which are shared beliefs and expectations about men and women.Typical male behavior then refers to activities outside of the house, such as employment, being in authority positions, and being the main breadwinner.Typical female behavior are activities to fill caretaking roles at home, such as domestic tasks and childrearing.Agentic traits, associated with being male, consist of qualities of assertiveness, mastery, dominance, and independence, while communion, associated with being female, refers to qualities such as warmth, friendliness, and concern with others.Gender roles affect what is expected of people, what is internalized as norm, and also affect career behavior (Eagly and Wood 2012).Social role theory also takes into consideration social changes, such as the increased occupational participation of women, and the higher educational level of women.Past societal changes have led to an increased gender similarity, such that women have attained more masculine roles.However, men have not undergone transitions to the same extent.There are some suggestions that men increased their uptake of tasks related to childcare and domestic work, yet still few men enter caring professions or hold more communally demanding jobs (Eagly and Wood 2012).Self-reported time allocation is a common way of measuring engagement in paid work and in unpaid work (Greenhaus et al. 2003;Konietzko 2015;Sappleton and Lourenço 2016).Self-reported work hours are a predictor of career success (Ng et al. 2005).
The case of self-employment is particularly interesting in this regard.The stereotype of an entrepreneur is masculine, assertive, competitive, hardworking, and innovative, therefore aligning more with agentic characteristics.In an effort to find a comprehensive list of predictors for firm performance of high-technology firms, (Soriano 2010) found positive relations with private loans instead of family loans, a high need for achievement, dynamism and the pursuit of external growth, all of which are more likely to be found among men rather than women.Among Dutch entrepreneurs, women work fewer hours than men (Verheul et al. 2009).Using data from European self-employed people, men in male-dominated sectors worked the longest hours, and women in female-dominated sectors worked the least amount of hours (Sappleton and Lourenço 2016).In a study on different German datasets, it was found that time allocation was more gendered for self-employed than for employed individuals; self-employed women reported almost the same hours spent on household work compared to employed women, whereas self-employed men spent less time on household work than employed men (Konietzko 2015).Among Irish entrepreneurs (Drew and Humbert 2012), female entrepreneurs found flexible work more important than men, and fathers relied more on their partner to care for children than did mothers.It was also reported that fathers spent around 5 hours and mothers around 14 hours on household work.In a study on French entrepreneurs (Vial and Richomme-Huet 2017), although 57% of all entrepreneurs in their dataset had children, only 18% of the women had children, suggesting children may be more of a hindering factor for entrepreneurship for women than for men.Therefore, empirical evidence on a variety of different European self-employed people seems to suggest that traditional gender role behavior is upheld rather than revolutionized through entrepreneurship.
Based on social role theory, we expect that women work fewer paid work hours than men and that women work more household hours than men.Regarding the effect of country, we expect that French self-employed people work more paid work hours than German self-employed people.We expect that the traditional breadwinner model is stronger in Germany, whereas the empowering nature of the family policies is stronger in France, suggesting that only women with family (partner or children) in Germany drastically reduce work hours.The strong traditional role modelling in Germany would also suggest that the expected household hours for women are higher in Germany than in France.Additionally, France offers tax deductions to people using household service providers, assisting with child minding, caring for the elderly, cleaning or cooking.Those 1.2 million people (mainly women themselves) should reduce the number of household hours worked by French women (Morel 2012).Regarding the impact of family, the need for household work should increase with children, yet the time available to do household work should be reduced with children.As we expect women in France to work more paid hours than women in Germany, we expect German women to use their available time to do more household work when having children, whereas French women will be unable to do more household work when having children.The presence of children will potentially not influence men's household hours in either country.

Gender and Living Situation on Life Satisfaction
The work stress literature, and here specifically, the effort-recovery model (Meijman and Mulder 1998) proposes that recovery is important to sustain effective cognitive, psychological and physiological functioning because the effort to meet the demands of work and other life spheres can take a toll on the bodily systems.One of the commonly mentioned demands of being self-employed is long work hours (Cardon and Patel 2015).Among entrepreneurs, research does not find an association between long work hours and job satisfaction (Sappleton and Lourenço 2016), exhaustion or physical health complaints (Taris et al. 2008).However, Sappleton and Lourenço (2016) found that self-employed women were more satisfied with their work than were men and (Taris et al. 2008) found that the inability to withdraw from work is negatively related to the well-being of self-employed people.Having a partner and/or child can be another demand where self-employed women need to exert effort, but, on the other hand, family gives the opportunity to not think about work, which can be positive for recovery.In addition, a partner can be a source for instrumental and emotional support, facilitating recovery further.(Drew and Humbert 2012) found that among Irish entrepreneurs, mothers and fathers reported spending a similar number of hours on leisure time.Also, Drew and Humbert (2012) found no difference in perceived personal sacrifices made by fathers/mothers (e.g., delaying children) and no differences in perceived sacrificed time with a partner or children.However, as argued in the previous sections, factual time commitment to work and household are likely to be different between genders and depend on family status and country culture.While well-being is a complex concept, we follow the ideas from Diener that well-being refers to people's evaluation of their lives, whereby life satisfaction is a person's global judgment of their life (Diener 2000).
Life-course studies on the general population, not self-employed people specifically, revealed that childless individuals have equal or higher well-being compared to parents (Umberson et al. 2010).
A study using European data showed that married individuals with children report the highest levels of life satisfaction, but this was due to socioeconomic advantages (Vignoli et al. 2014).European data on the general population in both Germany and France showed that there was no significant difference in life satisfaction between individuals who are 'married with children' compared to 'married no children', 'cohabiting with children', 'cohabiting without children', but a significant difference to the two categories of 'no partner with children' and 'no partner no children' (Ainsaar and Rootalu 2016).That is, the different studies on life satisfaction levels among the general population seem to suggest that children do not necessarily increase levels of life satisfaction (but they decrease when being without partner), but having a partner increases life satisfaction levels.In addition, the studies suggest that women have higher life satisfaction levels than men (Ainsaar and Rootalu 2016).However, these studies did not take into consideration that self-employed individuals may be different.
Self-employment is often considered a way for women to work and take care of their family responsibilities (Annink and Dulk 2012).However, for those women without family, levels of life satisfaction may be even higher, because they did not enter entrepreneurship out of necessity, and they will likely have more time to devote to their business.We argued in the previous section that the French system of family policies encourages women to return to work quickly after giving birth.This is likely engrained in the culture as to also affect self-employed women to go back to work quickly and with as many hours as before having a child.We also predicted that women in France with a partner and children will continue to work similar hours and, because of the traditional gender roles, also do more household work.Therefore, we expect that self-employed women in France with children will have lower levels of life satisfaction compared to men, because they will have increased demands of managing children, partner, and work as before having children.Because we predicted women in Germany to lower their work hours drastically when having a partner and children, we would not expect large differences in life satisfaction within German women.

Gender and Living Situation on 'Business Success'
In addition to studying differences among self-employed people regarding work hours, household hours and life satisfaction, we also wanted to connect differences to 'business success', which is vital to the purpose of owning and managing a business.The difficulty is the definition and measurement of business success.The most common criteria in the entrepreneurial literature is financial success (Wach et al. 2016).A recent study suggested a measure of subjective business success involving the evaluation of aspects such as firm performance, financial reward, satisfying work relationships with employees, personal fulfillment, etc. (Wach et al. 2016).We measure business success with two indicators: the number of employees and household contribution.To classify entrepreneurial ventures into micro, small-or medium-sized businesses, it is common in Europe to use the number of employees as a criterion besides financial indicators; that is, the number of employees symbolizes size of the venture, and thereby, success.In addition, we chose to include household contribution to study the impact the business venture makes for the family income depending on gender.Specifically, Bertrand et al. (Bertrand et al. 2015) made the claim that women rarely earn more than men within the same household, that there is a sharp drop in share of household income at 50%.It is argued that as part of the traditional gender roles, men are supposed to earn more than their female partners.Bertrand et al. argued that as soon as women reach about half of the income their male partners earn, women are making adjustments to not earn more than half of the household income, by for instance working part-time or doing more household hours to compensate the gender norm violation.Because self-employment is often pursued to get greater freedom and increase financial gain (Annink and Dulk 2012), the idea of looking at household income is interesting from a female entrepreneurship perspective.
Concerning differences among self-employed people, (Marshall and Flaig 2014) found that marriage and family was negatively associated with the earnings of self-employed women.Among Irish entrepreneurs, mothers felt that children and housework interfered with their business more than fathers (Drew and Humbert 2012).Among Dutch entrepreneurs, women were less productive than men, were less risk-prone, had less industry experience, invested less financial capital, and had less contact with other entrepreneurs (Verheul et al. 2009).However, among French self-employed people, women experienced less difficulty being alone as entrepreneurs, reported less difficulty to hire qualified labor, to open a bank account, and to find the necessary funds (Vial and Richomme-Huet 2017).
Based on the gender norms in both countries, we propose that women are less likely to have employees and are less likely to earn more than half of the household income.Because we expected that French self-employed women would work more hours than Germans, they potentially increase their potential for business success.

Sample
We used data from the European Social Survey (ESS), a survey based on face-to-face interviews across European nations.The ESS survey achieves high response rates of at least 70%, and its sampling procedure ensures representativeness.Based on the accompanying recommendations, we applied post-stratification weights in combination with population size weights.
Selecting all self-employed people from France and Germany, the total sample (after applying the weighting procedure) was 1092 self-employed people, 470 French and 622 Germans, 720 men and 372 women.In France, 177 self-employed people were female (38%), and in Germany, 196 were women (32%).

Measures
Demographics.Gender was coded 1 (male) and 2 (female).Country was coded 0 (France) and 1 (Germany).Living situation was coded 0 (not living with a partner, not with children), 1 (not living with a partner, living with children), 2 (living with a partner, not living with children), 3 (living with a partner, living with children).The living situation variable was derived as a combination of the question 'Respondent lives with husband/wife/partner' and 'Respondent has children living at home'.Because one of the living situation categorizations had a very low frequency, we report analyses without this categorization (not living with a partner but living with children).Table 1 shows the frequencies of this categorization for the overall sample size, and depending on gender and country.Additionally, we included age as the control variable as we expected that age and the chances of having children at home would be interrelated. 11 The ESS survey did not include information on the age of the business or the experience of the self-employed in running a business.Household hours.Household hours was assessed with the following question 'I would now like to ask you about housework.By housework, I mean things done around the home such as cooking, washing, cleaning, care of clothes, shopping, maintenance of property, but not including childcare or leisure activities.About how many hours a week, in total, do you personally spend on housework?' Life satisfaction.Life satisfaction was assessed with the question 'All things considered, how satisfied are you with your life as a whole nowadays?' on a scale from 0 (extremely dissatisfied) to 10 (extremely satisfied).
Number of employees.The question 'How many employees do you have?' was used to measure the number of employees self-employed people had.
Household contribution.This was measured with the question 'Around how large a proportion of the household income do you provide yourself?'.Answer options were 'none', 'very small', 'under a half', 'about a half', 'over a half', 'very large', 'all', the answers 'refused' and 'don't know' were coded as missing.

Descriptive Statistics
Table 2 depicts means and standard deviations (in parentheses) separately for men and women, living alone, with a partner, and with a partner and children for self-employed people from France and Germany, respectively.In Germany, the mean age was 54.37 years (SD = 14.80 years), with men on average 55.20 years (SD = 16.87 years, range 21-94 years) and women 50.59 years (SD = 13.38 years, range 21-77 years).The average age of respondents from France was 56.76 years (SD = 15.68 years), with men 57.12 years (SD = 17.02 years, range 22-90 years) and women 59.34 (SD = 16.41 years, range 27-91 years).

Prediction of Work Hours
An ANOVA was conducted to examine the main and interactive effects of gender, country and living situation categorization on reported working hours.The three-way interaction between gender, country and living situation was marginally significant (F(2) = 2.90, p = 0.055.Despite the marginally significant result, we display the interaction in Figure 1.In the German sample, there was no significant gender difference when self-employed people lived alone (F(1, 125) = 0.20, p = 0.657, M male = 39.92,SD male = 16.56;M female = 38.41,SD female = 20.92).There was a significant gender difference when self-employed people lived with a partner (F(1, 248) = 26.66,p < 0.001, M male = 51.82,SD male = 18.26;M female = 39.40,SD female = 15.25), and an even stronger gender effect when self-employed people lived with a partner and children (F(1, 175) = 60.35,p < 0.001, M male = 51.11,SD male = 15.47;M female = 30.10,SD female = 19.89),such that men worked longer work hours than women.In the French sample, there was only a significant gender difference when self-employed people lived alone (F(1, 89) = 6.66, p < 0.05, M male = 54.02,SD male = 23.95;M female = 66.56,SD female = 20.69),indicating that women worked longer work hours than men.However, there was no longer a significant gender difference when self-employed people lived with a partner (F(1, 173) = 1.73, p = 0.190, M male = 57.10,SD male = 15.70;M female = 53.13,SD female = 23.86) and with a partner and children (F(1, 173) = 0.01, p = 0.930, M male = 55.04,SD male = 21.23;M female = 55.34,SD female = 21.49).In addition, the main effect of country was significant (F(1) = 107.95,p < 0.001), with Germans working on average 44.73 (SD = 19.04)and French 56.21 (SD = 20.81)hours.

Prediction of Work Hours
An ANOVA was conducted to examine the main and interactive effects of gender, country and living situation categorization on reported working hours.The three-way interaction between gender, country and living situation was marginally significant (F(2) = 2.90, p = 0.055.Despite the marginally significant result, we display the interaction in Figure 1

Prediction of Household Hours
The question regarding hours dedicated to housework was only asked to people who live with their partners.The three-way interaction between gender, country and living situation was significant (F(1) = 5.01, p = 0.025) and is displayed in Figure 2. In the German sample, there was a significant gender difference when self-employed people lived with a partner (F(1, 247) = 59.01, p < Figure 1.Three-way interaction between gender, country and home for working hours.

Prediction of Household Hours
The question regarding hours dedicated to housework was only asked to people who live with their partners.The three-way interaction between gender, country and living situation was significant (F(1) = 5.01, p = 0.025) and is displayed in Figure 2. In the German sample, there was a significant gender difference when self-employed people lived with a partner (F(1, 247) = 59.01, p < 0.001, M male = 6.16,SD male = 6.44;M female = 15.13,SD female = 11.89) and an even stronger one when they lived with a partner and children (F(1, 183) = 130.34, p < 0.001, M male = 5.57, SD male = 6.08;M female = 21.29,SD female = 12.67), with women working expectedly more hours than men.In the French sample, there was a significant gender difference when self-employed people lived with a partner (F(1, 181) = 36.25,p < 0.001, M male = 8.95, SD male = 9.90; M femal = 21.04,SD female = 17.13) and when they lived with a partner and children (F(1, 166) = 58.00,p < 0.001, M male = 5.01, SD male = 6.77;M female = 16.95,SD female = 13.56).Here, the strongest main effect was that of gender (F(1) = 288.46,p < 0.001), with women reporting working on average 18.41 (SD = 13.99) and men 6.44 (SD = 7.51) hours on household chores.The main effect of country was not significant.
Here, the strongest main effect was that of gender (F(1) = 288.46,p < 0.001), with women reporting working on average 18.41 (SD = 13.99) and men 6.44 (SD = 7.51) hours on household chores.The main effect of country was not significant.

Number of Employees
For the whole sample, 63.6% of all self-employed people had zero employees, 10% employed one person and the highest number of employees employed was 60.A logistic regression revealed a non-significant three-way interaction between gender, country and living situation.However, the two-way interaction between gender and living situation was significant (Wald = 9.14, p = 0.010).There was a non-significant gender difference within self-employed people living alone (χ 2 = 1.719, p = 0.190, 31% of men had at least one employee vs. 23% of women).There was a significant gender difference within self-employed people living with a partner (χ 2 = 4.089, p = 0.043, 48% of men had at least one employee vs. 38% of women), and self-employed people living with a partner and children (χ 2 = 31.786,p < 0.001, 42% of men had at least one employee vs. 14% of women), with women having

Number of Employees
For the whole sample, 63.6% of all self-employed people had zero employees, 10% employed one person and the highest number of employees employed was 60.A logistic regression revealed a non-significant three-way interaction between gender, country and living situation.However, the two-way interaction between gender and living situation was significant (Wald = 9.14, p = 0.010).There was a non-significant gender difference within self-employed people living alone (χ 2 = 1.719, p = 0.190, 31% of men had at least one employee vs. 23% of women).There was a significant gender difference within self-employed people living with a partner (χ 2 = 4.089, p = 0.043, 48% of men had at least one employee vs. 38% of women), and self-employed people living with a partner and children (χ 2 = 31.786,p < 0.001, 42% of men had at least one employee vs. 14% of women), with women having a lower likelihood of having at least one employee.

Household Contribution
For this analysis, we were only able to test our predictions for those self-employed people who lived with a partner and those with a partner and children.We dichotomized the household contribution question (0 = 'none, 'very small', 'under a half', 'about half', 1 = 'over a half', 'very large', 'all').A logistic regression revealed a non-significant three-way interaction.The two-way interaction between country and gender was not significant.The two-way interaction between gender and living situation was significant (Wald = 6.99, p = 0.008).Results revealed that within self-employed people who live with their partners, 64% men reported contributing at least more than half the overall household vs. 8% of women.Within self-employed people who live with a partner and children, 64% of the men reported contributing to the household income with more than half vs. 22% of the women.As a note, this latter percentage can be attributed to French women, who reported contributing more than half with 33% compared to 10% of German women.

Discussion
This study set out to investigate the status of female entrepreneurship by looking at the number of paid work hours, unpaid household hours, life satisfaction, number of employees, and income share in both France and Germany.We were interested in answering the question of whether entrepreneurship is indeed a way to female empowerment or simply another avenue for the manifestation of traditional gender roles.Using ESS data, we compared the case of German and French self-employed people, looking additionally at differences between genders and the living situation.

Summary of Results
Work hours are an important indicator of engagement and can predict career success (Ng et al. 2005).We found no gender differences among German self-employed people living alone, among French self-employed people living with a partner and with a partner and children.Surprisingly, French single self-employed women worked more than their male counterparts.We found the expected gender difference of men working more than women only for German self-employed people living with a partner, and living with a partner and children.From the perspective of work hours, the French self-employed people seem to be more equal.
The question about household hours was only asked to self-employed people living with a partner and living with a partner and children.Results are similar across countries, indicating that women worked drastically more hours in the household than men.Combined (for those who indicated living with a partner or with a partner and children), this also means that in France, women work more hours overall, whereas in Germany, men work more hours overall, as German women reduced their paid work hours when living with a partner and with a partner and children.Both those results confirm our assumptions following social role theory.Women work fewer paid hours but more household hours.Also, our expectations regarding country differences have been confirmed: the French work more hours overall.
Life satisfaction is often seen as an overall indicator of well-being (Diener et al. 1999).In the group of German self-employed people, there were no differences between genders.This may indicate that self-employed people were happy with their work situation, and those living with a partner or with a partner and children were happy with their role of the men (on average) working more paid hours and the women (on average) working more unpaid household hours.In this way, self-employed people in Germany seem to mimic the traditional breadwinner model, with the men working more paid hours and the women working more unpaid hours, as soon as they live with a partner.In the group of French self-employed people, women were more satisfied with their lives than men when single.There was no gender difference when living with a partner.French men reported higher levels of life satisfaction when living with a partner and children.We see (Sappleton and Lourenço 2016)'s results as dependent on country (with its policies, culture, and norms) and the living situation.While they found that self-employed women were more satisfied than men, in this study we could only confirm this for French self-employed women living alone.It seems that for French women, life satisfaction takes a dip when living with a partner or a partner and children.In France, due to the culturally engrained pressure to go back to work quickly and with as many hours as before, women may perceive increased demands to manage household, children, partner and work, which could explain the lower life satisfaction.The traditional breadwinner model, which we see more closely resembled in the German data, seems to lead to similar levels of life satisfaction for both genders.
Regarding the number of employees, there were no country differences.Gender differences existed, such that men were more likely to have employees when living with a partner, and living with a partner and children.However, when self-employed people lived alone, there were no gender differences.From an investor standpoint, this result is troublesome as this result could be read as investing in women is more lucrative when they live alone.Additionally, it seems that the more work hours of French women returning to work quickly with children does not make them more successful.There were no country differences for the two success indicators, suggesting that the different policies within Germany and France do not create differences in business success.Rather, life satisfaction is higher among Germans.
Lastly, we also investigated household contribution, but this was only asked to people living with a partner or living with a partner and children.Similar to the number of employees, men were more likely to contribute more than half than women.These results are in line with the claim that women do not out-earn their male partner and if they do, they adjust their share of household hours to compensate for the gender violation (Bertrand et al. 2015).This would indicate that the suggestion that entrepreneurship is pursued for greater freedom and financial gains (Annink and Dulk 2012), is dependent on gender and applies more to men.Women seem to embrace the possibility for greater freedom by, for example, reducing paid work hours but we do not see women living with a partner or with a partner and children pursuing entrepreneurship for financial gain.Single women (living without partner or children), on the other hand, seem to enter entrepreneurship not out of necessity to accommodate a family, enabling them to invest more hours.

Investing in All the Single Women
Success in entrepreneurship is traditionally defined in terms of financial gains, the number of employees, and long work hours seem to be expected (Cardon and Patel 2015).This privileges people with no caring or other responsibilities, which traditionally applies to single people and people with a spouse at home (which is true more often for men than for women).This narrow definition of success does not consider the actual value created by the company (social and environmentally conscious entrepreneurship could potentially have more value than more exploitative forms) or the societal value of activities outside of paid employment.Given the declining birth rates in more developed countries, giving birth to children could potentially be seen as crucial contribution to society.
Following the results above, one might draw the troubling conclusion that investors should only invest in single women and men (irrespective of their living situation) when financing entrepreneurship as they are likely to employ more people.However, we know from microloan projects aimed at women in developing countries that they have higher rates of success with their businesses and consequently raise their communities' living standard (Pitt et al. 2003).A more appropriate conclusion would be the question of how structures can be changed so that women have the same opportunities to succeed in their entrepreneurship efforts.
Given the results, one might also draw the conclusion to only invest in older self-employed women if the intention was to maximize the hours worked.Self-employed women in France were on average 75 years old in the dataset.They worked the longest hours among the whole sample, and among French self-employed people, they had one of the highest levels of life satisfaction.To draw such a conclusion would of course be shortsighted.There is no information available on why older self-employed women worked so many hours.A potential reason could be poverty in old-age which disproportionately affects women due to lower life earnings.We are not aware of much research into older self-employed people for an exception, see (Zissimopoulos and Karoly 2007), let alone older, single, female self-employed people.Our results highlight a need to research more on older self-employed people.
Furthermore, it should be investigated how men can become more active members of their household firstly, to enable women to pursue business success in the traditional sense and secondly, to be more involved in the home sphere.One reason to study Germany and France as context was the different parental leave policies and accompanying culture.It is difficult to argue that any one system of parental leave would be better; it may depend on whether the goal is to maximize economic gain, life satisfaction, or something else entirely.In the German system with women reducing their investment in paid labor, the results clearly indicated greater life satisfaction.However, the high levels of life satisfaction may potentially only transpire when women are being content with this traditional division of labor.With an increasing number of women that are better educated than men, this contentment might vanish.There is also an economic argument to be made for the underutilization of a highly skilled labor force.The government first invests in the education of women (which is essentially free in Germany) to then miss out on a full utilization of said part of the labor force.
However, the French model with four months of well-paid parental leave and no guaranteed, publicly financed childcare up until the age of three (Boyer and Fagnani 2018), we see women in high status jobs returning to full-time work more quickly.Furthermore, despite working more hours than their German counterparts, French women still did not reach men's level of success as measured in the number of employees and household contribution.Even though French women work as much or more than their male counterparts, they still did not contribute more than 50% to the household income.We can conclude that the French parental leave policy is not beneficial for French women as they work as much as men, work more in the household, have equally low life satisfaction as French men (even lower life satisfaction with children), and still contribute less to the household income.While work hours per se tend not to be associated with life satisfaction (Baslevent and Kirmanolu 2014), it may be more frustrating to work long hours, work additional long hours at home, share responsibilities for children, and not be as successful as male entrepreneurs.
When success is defined by the number of employees and contribution to household income, as done in this study, it seems entrepreneurship is not beneficial to women.However, we see comparable life satisfaction for men and women when the opportunity for a reduction in working hours (i.e., Germany) is used.Entrepreneurship is generally lauded for allowing more freedom, which could contribute to greater life satisfaction by enabling people to balance work, home, and childcare demands through flexible hours.This could be particularly relevant in countries with limited early childhood education and care entitlements.
In 2012, the GEM monitor report found that female entrepreneurs in developed countries in Europe had surprisingly low growth aspirations (Kelley et al. 2012).The authors attributed some of it to the high levels of fear of failure, but also to the decision by many women to start a business alone without co-founders, making it less likely to be a successful venture.In the 2017 GEM monitor report, it is described that in addition to stereotypes and household responsibilities, women are disadvantaged from the start as they are less likely to have an entrepreneurial network, role models, mentorship opportunities and professional connections (Herrington and Kew 2017).Recently, a gender bias in funding decisions was revealed (Kanze et al. 2018), such that women are asked questions to prevent failure whereas men are asked growth-oriented questions.However, a recent study painted an interesting perspective (Johnson et al. 2018); they suggested that in crowdfunding, women have an advantage with the result of getting more funding because women are stereotypically attributed a higher trustworthiness.This suggests that while not all gender stereotypes are to the detriment of women, there are gender stereotypes that affect entrepreneurial decisions and success.In addition, it may be that there are still gender biases in both Germany and France that contribute to some of the results found in this study.Further research is needed to examine how gender affects entrepreneurship in developing countries and countries with higher levels of gender equality.In addition, in-depth analyses of household dynamics and successful female entrepreneurs are needed.Specifically, a qualitative approach may be useful to study the negotiations in dual-earner couples or dual-entrepreneurship couples on how to allocate time and distribute tasks.Furthermore, interviewing highly successful female entrepreneurs, their view on stereotypes, barriers, time allocation in different life spheres, and taking into consideration the background of different gender cultures would be highly valuable.We also suggest research into the idea of "successful" entrepreneurship focusing particularly on how different individuals define success, and if those definitions are gendered or dependent on national culture.Qualitative methods could be a promising avenue to gain a deeper understanding of people's reasoning, particularly around questions of gender and success.We would also encourage research taking a longitudinal approach following dual-career couples and investigating how their perspective on success and the labor needed to maintain the relationship changes over time.
A change in the metric of success may also help create better conditions for more female entrepreneurship.It is not only parental leave policies that may facilitate more and more successful female entrepreneurship.In knowledge economies with declining birth rates, a shift in attitudes towards a greater appreciation of childcare could certainly lead to a more equal distribution of said work, allowing people to reach their full potential in business and contribute equally to the maintenance of the household and childcare, independent of their gender.

Figure 1 .
Figure1.Three-way interaction between gender, country and home for working hours.

Figure 1 .
Figure 1.Three-way interaction between gender, country and home for working hours.

Figure 2 .
Figure 2. Three-way interaction between gender, country and home for household hours.

Figure 2 .
Figure 2. Three-way interaction between gender, country and home for household hours.

Figure 3 .
Figure 3. Three-way interaction between gender, country and home for life satisfaction.

Figure 3 .
Figure 3. Three-way interaction between gender, country and home for life satisfaction.

Table 1 .
Frequency of categorization living with a partner/living with children in overall sample and depending on gender and country.Work hours.Work hours was assessed with the question 'Regardless of your basic or contracted hours, how many hours do you normally work a week (in your main job), including any paid or unpaid overtime?'.

Table 2 .
Summary of descriptive statistics (means and standard deviations in brackets).
Note.Table depicts means and standard deviations (in parentheses) separately for men and women, living with partner or children, with a partner, and with a partner and children for self-employed people from France and Germany, respectively.