1 Introduction

How does an entrepreneur’s sex influence the ownership structure of entrepreneurial businesses? Previous research on women’s entrepreneurship has focused on how the activities of entrepreneurs are enabled or constrained by their sex (Bird and Brush 2002; de Bruin et al. 2007; Jennings and Brush 2013). Not only do females perceive limited opportunities to found their own businesses compared to males (Gupta et al. 2009; Hughes 2003), but once they launch firms, female-led businesses are more likely to suffer from a lack of available resources. Past studies have suggested that female entrepreneurs tend to open businesses with lower initial funding compared to their male counterpart (Alsos et al. 2006; Boden and Nucci 2000; Coleman 1988). Also, studies have shown that female entrepreneurs are less likely to appropriate financing provided by financial institutions than male counterparts (Carter et al. 2007). As a consequence, female-led businesses tend to be smaller in organizational size, lower in assets and profits, and weaker in performance on average (Fairlie and Robb 2009; Robb 2002; for mixed results on performance gap, however, see Robb and Watson 2012).

While past studies have emphasized the role of gender in the decision-making and practices of entrepreneurs, what has not been systematically examined until now is the gendered processes of how entrepreneurial businesses are organized at their formative stage. As Jennings and Brush have suggested (Jennings and Brush 2013:671), past studies have rarely explored whether female and male entrepreneurs differ in the organizing of their businesses. This is a serious oversight given that the survival and performance of businesses are heavily influenced by the initial structuring of organizations (Boden and Nucci 2000). To fill this lacuna, we examine the influence of the entrepreneur’s sex on the ownership structure of nascent firms that are in preparation for launch. We investigate whether and, if so, how the formation of entrepreneurial businesses is influenced by the gendered social system.

The ownership structure of firms under study is usually determined in the formative stage, even though it may change over time. In launching businesses, entrepreneurs should decide either to establish a solo enterprise or to form an entrepreneurial team with partners. While this decision should be made on the basis of strategic calculation in principle, the recruitment of co-founders is constrained by entrepreneurs’ capacity to do so. In other words, entrepreneurs’ decisions to find a co-founder are restricted by the preexisting social and cultural capital that they maintain (Aldrich et al. 1995; Bosma et al. 2004; Fairlie and Robb 2009; Ruef 2010; Ruef et al. 2003).

Regarding social and cultural capital, previous literature has suggested that significant differences exist between males and females (Aldrich 1989; Moore 1990). Gender-based differences exist both in terms of network size and composition (Campbell and Rosenfeld 1985; Lin 2000). In particular, past studies have reported that the social capital of female entrepreneurs tends to be smaller, less diverse, and more family-oriented compared to that of males (Cromie and Birley 1992; Greve and Salaff 2003; Klyver 2011; Renzulli et al. 2000). Moreover, recent literature has noted that female entrepreneurs are less likely to acquire cultural capital due to their lack in knowledge and skills, formal education and training, or prior work experience compared to their male counterpart (Carter and Williams 2003; Fairlie and Robb 2009; Fischer et al. 1993; Lerner et al. 1997).

Building upon the literature of gender, social and cultural capital, and entrepreneurship, we examine how female and male entrepreneurs differentially select their co-founders. Based on previous studies suggesting that the social and cultural capital of females tends to be limited compared to males, we hypothesize that female entrepreneurs are more likely to establish enterprises by themselves or with family members rather than with non-family members such as previous co-workers compared to male counterparts. Moreover, we explore the possibility that female-led businesses tend to display lower initial performance compared to male-led businesses. Using data from the Panel Study of Entrepreneurial Dynamics II (PSED II), a nationally representative sample of nascent entrepreneurs in the USA, our analysis yields results consistent with our theoretical expectation. The results indicate that female entrepreneurs are more likely to form either a solo or a family-only enterprise instead of a non-family enterprise compared to male entrepreneurs especially when they lack social or cultural capital. In particular, female owners are significantly less likely to recruit co-founders from their professional networks such as co-workers of previous workplaces than the males. Moreover, we find that solo or family businesses run by female entrepreneurs are less likely to show positive initial performance compared to males. We conclude with the implications of our study to further understand the effect of gender on the decisions, activities, and performance of entrepreneurs.

2 Theoretical backgrounds

2.1 The forms of capital and entrepreneurship

Individuals accumulate different forms of capital to achieve success or higher status in society. In addition to economic capital such as assets and properties, individuals may maintain other forms of capital such as social capital and cultural capital to fulfill their goals (Bourdieu 1986; Tatli et al. 2014). For entrepreneurs, social and cultural capital is convertible to economic capital; in other words, entrepreneurs can utilize their social and cultural capital to achieve economic success in their new businesses.

Social capital refers to the durable social relationships of individuals or organizations that serve as the “glue” and “goodwill” among actors (Adler and Kwon 2002; Payne et al. 2011; Putnam 2000; Welter and Smallbone 2006). At the individual level, social capital scholars have argued that individuals can utilize social capital both to promote job searches and to facilitate job promotion (Burt 1992; Fukuyama 1995; Granovetter 1985; Lin 2000). At the organizational level, social capital plays a critical role in enhancing the overall performance of firms (Reagans and Zuckerman 2001; Uzzi 1997). Inside organizations, social capital can create a collaborative environment by generating social solidarity and interpersonal trust among members (Adler and Kwon 2002; Bourdieu 1986; Coleman 1988; Davidsson and Honig 2003; Ruef et al. 2003; Yang and Aldrich 2014). Outside organizations, the social capital of organizational members can be used to appropriate external financial resources, to reduce transaction costs, and to recruit new members who possess non-redundant skills and knowledge (Burt 1992; Granovetter 1985; Lin et al. 2001).

Entrepreneurship research has increasingly paid attention to the effect of social capital on the activities of entrepreneurs (Baron and Tang 2008; Ireland et al. 2003; Maurer and Ebers 2006). Empirical studies on entrepreneurs have suggested that the extent of the social capital of entrepreneurs increases their motivation to open a new business (Liao and Welsch 2005). Also, they can use their social relationships with bankers, venture capitalists, and experienced discussion partners to attain greater access to financial resources and, consequently, to enhance their chance of survival in their formative stage (Boden and Nucci 2000; Gopalakrishnan et al. 2008).

Cultural capital is another critical aspect to understand entrepreneurial activities and outcomes. Cultural capital is accumulated through the acquisition of knowledge and skills, talent and experience, and the completion of formal education or professional training (Bourdieu 1986). Cultural capital can be utilized to either attain higher social status or to reproduce one’s social position. For entrepreneurs, the possession of cultural capital—including personal traits such as self-confidence or leadership skills that are developed from prior experience of receiving professional training or leading teams in workplace—is influential in determining the outcome of their nascent businesses (Davidsson and Honig 2003). Moreover, cultural capital may be translated into social capital. Individuals with ample training and education in a specialized area or previous start-up experience can more easily recruit partners when they start their own businesses.

More recently, research on entrepreneurship has extended its focus to the formation of entrepreneurial teams (Ruef 2010; Ruef et al. 2003; West 2007). In opening businesses, entrepreneurs should decide either to establish a solo enterprise or to form an entrepreneurial team with partners. In an entrepreneurial team, each additional partner may divide responsibilities and labor, and she/he can also provide a unique skill set to the team. Also, founders of the team should utilize his/her leadership skills to coordinate the division of labor among team members. Herein, the extent of social and cultural capital that entrepreneurs possess can influence their decision about whether to form an entrepreneurial team or not. The decision to seek co-founders is inevitably restricted by the preexisting network ties that entrepreneurs maintain (Ruef et al. 2003; Yang and Aldrich 2014). Also, cultural capital entrepreneurs possess such as leadership traits or self-confidence in opening businesses may influence the likelihood that they succeed in recruiting collaborators with whom to establish new businesses together. While the entrepreneur’s decision to either open a solo business or to recruit co-founders from preexisting social ties should be made based on a strategic calculation in principle, this decision is heavily influenced by the preexisting social and cultural capital entrepreneurs maintain in reality.

2.2 Gender and entrepreneurship

Women’s entrepreneurship research has focused on the role of gender in the formation and maintenance of entrepreneurial businesses (Bird and Brush 2002; Jennings and Brush 2013). Past studies have suggested that females are less likely to be self-employed and to found their own businesses compared to males (Gupta et al. 2013; Gupta et al. 2009; Hughes 2003). After launching their businesses, female entrepreneurs tend to maintain businesses with lower levels of financial resources (Alsos et al. 2006; Boden and Nucci 2000; Coleman 1988). For example, female entrepreneurs are significantly less likely to receive external financing from venture capital due to the gendered stereotypes on entrepreneurship and/or the lack of social capital (Brush et al. 2014; Greene et al. 2001). This limited access to economic resources partially explains why female-led businesses underperform compared to male-led businesses in terms of size growth, profit-making, and organizational survival (Fairlie and Robb 2009; Robb 2002).

In addition to economic capital, the aspect of social and cultural capital is critical to fully understanding the gendered formation of entrepreneurship. Social network studies proposed that males tend to have larger and more heterogeneous networks than females (Campbell and Rosenfeld 1985). This has been explained by males’ higher participation than females in public affairs, given that females are forced to be housekeepers; since females are less likely than males to work in larger and formal organizations, they are used to be marginalized in the social networks dominated by males (Marsden 1988; McPherson and Smith-Lovin 1982). Under the patriarchal social system, females are also less likely to have prior work experience or experience in supervising employees.

Due to the drastic increase of female participation in economic activities, females’ cultural capital has dramatically increased than before. Females have increasingly accumulated cultural capital such as self-confidence, leadership skills, and work ethic through the completion of formal education and training and professional work experience. Research has also suggested that women with prior work experience are more likely to maintain larger social capital than those without it (Renzulli et al. 2000). But still, research on entrepreneurship indicates that female entrepreneurs not only have significantly less years of industry experience or professional training (Carter and Williams 2003), leading to smaller and less diverse networks compared to males (Greve and Salaff 2003; Klyver 2011). When female entrepreneurs have limited experience in their past work organization (Fischer et al. 1993), their limited interaction with professional others leads to lack of both social and cultural capital (Aldrich and Cliff 2003; Cromie and Birley 1992). Even females who occupied upper-level managerial positions are less integrated into the core “old boy” networks that hold power and resources within and outside the organization (Burke and McKeen 1994; Davidson and Cooper 1992; Ibarra 1992; Linehan 2000; Scase and Goffee 1989).

Moreover, scholars indicate that females tend to have family-oriented networks compared to males (Aldrich 1989; Greve and Salaff 2003). Due to gender stereotypes, women are expected to take primary responsibility for the domestic sphere, even when they are employed (Bradley 2007; Yang and Aldrich 2014). Women are increasingly surrounded by a family-oriented environment after marriage and in child rearing (Aldrich and Cliff 2003; Wellman 1985). The gendered role in housekeeping and childcare often limits women’s competency to form heterogeneous and formal social networks (Munch et al. 1997; Wellman 1985). Regarding entrepreneurship, female entrepreneurs nominate their spouses as the first source for business advice, followed by their friends and professional experts (Orhan 2001). The order is different for male entrepreneurs who nominate professional experts as their first source and their spouses as second. Past research has also indicated that females are more likely to seek financial resources from informal sources such as family loans rather than from external equity (Carter et al. 2007; Chaganti et al. 1995; Coleman and Robb 2009; Powell and Eddleston 2013).

3 Research hypotheses

In the formation of nascent businesses, entrepreneurs make the decision to either open a solo business or an entrepreneurial team. Also, when founding a team, entrepreneurs may recruit new partners from two groups of individuals: family members and non-family members (Aldrich et al. 1995; Granovetter 1985; Ruef 2010; Ruef et al. 2003; Yang and Aldrich 2014). Family members such as spouses, parents, and siblings share common identities and maintain a high level of interpersonal trust (Bourdieu 1973; Dyer 2006). Moreover, family members have frequent opportunities to discuss ideas for founding a new business, and they also serve as a potential pool of reliable partners (Randerson et al. 2015). On the other hand, non-family members such as friends and co-workers can also develop high levels of trust with entrepreneurs; this strong trust generated with friends or co-workers is often translated into collaboration with them in founding an entrepreneurial team. In particular, non-family members can provide a skill set that the lead entrepreneur does not possess, thus facilitating the division of labor between co-founders.

Based on this distinction, we introduce a new typology of ownership with four possible and mutually-exclusive outcomes: solo, family, non-family, and mixed enterprises. Solo enterprises are defined as businesses that are owned only by the respondent. Family enterprises are businesses that are co-owned by the respondent and at least one of his/her family members such as a spouse, a significant other, or a relative. Non-family enterprises mean businesses that are co-owned by the respondent and his/her non-family members. Finally, we conceptualize mixed enterprises as entrepreneurial teams that are co-founded by at least one family member and at least one non-family member. Thus, to form a mixed business, a founder should recruit co-founders from both family and non-family ties.

The social and cultural capital that entrepreneurs possess plays an important role during the formation process of nascent businesses. Entrepreneurs who have frequent interactions with their family members may tend to rely on family ties when opening a business. Entrepreneurs who have diverse connections with friends and co-workers, on the other hand, may recruit one of them because of their specialized skills and knowledge. Entrepreneurs whose social capital is limited, on the other hand, will tend to establish a business by themselves. Since female entrepreneurs tend to possess limited and family-oriented social capital, they may struggle to find non-family partners who can collaborate with them in establishing a business. In addition, female entrepreneurs are also less likely to have prior employment experience at the managerial level compared to male counterparts, and as a consequence, their limited cultural capital would restrict them from successfully recruiting co-founders to their nascent businesses. Thus, in terms of business formation, we hypothesize as follows:

  • Hypothesis 1a: Female entrepreneurs will be more likely to found either a solo or a family-only enterprise rather than a non-family or a mixed enterprise compared to their male counterparts.

  • Hypothesis 1b: The tendency of female entrepreneurs to found either a solo or a family-only enterprise will be stronger when they lack social or cultural capital.

Furthermore, past studies have suggested that female entrepreneurs are relatively more embedded in their family relationships compared to male entrepreneurs (Aldrich 1989; Greve and Salaff 2003; Orhan 2001). Since females are more likely to maintain family-oriented networks, they tend to rely on family members in establishing an enterprise than males. Also, the stronger motivation of females to attain a better balance between work and family compared to males will lead to a greater likelihood that females start a family enterprise with family members such as spouses. In addition, females tend to have less professional and managerial experience than males, and therefore, female entrepreneurs may have limited professional networks with co-workers from their previous workplace. Accordingly, we expect as follows:

  • Hypothesis 2a: Female entrepreneurs will be more likely to recruit their co-founder(s) from their family members compared to their male counterparts.

  • Hypothesis 2b: Female entrepreneurs will be less likely to recruit their co-founder(s) from their co-workers compared to their male counterparts.

Finally, we extend our analysis by further examining the consequence of the gendered formation of entrepreneurial businesses. Past empirical studies have suggested that businesses owned by females are more likely to be smaller in size and lower in initial financing, leading to weaker performance as a result (Alsos et al. 2006; Fairlie and Robb 2009; Robb 2002). To illuminate another causal link through which entrepreneurs’ gender influences their firm performance, we investigate how a combination of an entrepreneurs’ sex and their ownership type influences the performance of their enterprises. More specifically, we explore the possibility that female entrepreneurs’ tendency to form either a solo or a family enterprise has a negative influence on the performance of their businesses. To this end, we hypothesize the following:

  • Hypothesis 3: Female-led solo or family enterprises are less likely to display a positive firm performance compared to male-led enterprises.

4 Data and method

4.1 Data

We use data from the Panel Study of Entrepreneurial Dynamics II (PSED II) to test our hypotheses on the gendered formation of entrepreneurial businesses and its consequence. PSED II is a nationally representative sample of nascent entrepreneurs who have engaged in starting new businesses over the past 12 months (Reynolds et al. 2004; Reynolds and Curtin 2007). About 1200 individuals out of 34,000 were identified as nascent entrepreneurs in the screening process, and then, detailed information on the respondent as well as the co-owners of the respondent was collected from the respondent-owner. With its rich detail on the backgrounds and characteristics of nascent entrepreneurs, the data is useful in examining the formation process of entrepreneurial firms. To capture the formative stage of these firms, we use its first wave conducted between 2005 and 2006. Among nascent entrepreneurs in the first wave, 50.4% have already received revenues from their business activities at the time of the interview.

PSED II is particularly advantageous over other datasets for three reasons. First, it asks questions to entrepreneurs who are actively engaging in the creation of businesses at the time of the survey. Thus, we can avoid measurement errors that are often involved in retrospective surveys. Second, the survey includes nascent entrepreneurs who do not attain successful outcomes. Therefore, we can overcome the success bias presented in studies that limit their focus to entrepreneurs who went through the nascent period. Finally, the data includes detailed information on the ownership structure of entrepreneurial firms. Hence, the data provides a unique opportunity to investigate the relationship between entrepreneurs’ socio-demographic backgrounds such as sex and the recruitment of co-founders such as family members, co-workers, and friends.

4.2 Dependent variables

We are interested in the factors that explain the distinct ownership structures of entrepreneurial businesses. PSED II data is appropriate to address our question since it asks entrepreneurs about the ownership structure of their nascent businesses. Specifically, it collects information on whether their businesses are owned by themselves or by multiple individuals. In the case of entrepreneurial teams where multiple founders are involved, it further asks respondents about their relationships with other co-founders. These relationships are categorized as spousal relationship, relationship with significant others, kinship with relatives, non-work-related friendship, professional relationship with co-workers, or other non-family relationships. Regarding the ownership structure of nascent businesses, we classify the structure into four main types: solo enterprises, family enterprises, mixed enterprises, and non-family enterprises. Among 1151 respondents who are identified as nascent entrepreneurs in the data, 594 entrepreneurs (51.6%) are opening a solo enterprise, 343 entrepreneurs (29.8%) are establishing a family enterprise, 170 respondents (14.8%) are launching a non-family enterprise, and 44 entrepreneurs (3.8%) are founding a mixed enterprise.

In addition to our main dependent variable, the three types of ownership structures, we examine the social relationships between the respondents and their co-owners within entrepreneurial teams. Here, we first make a distinction between family and non-family ties, and then we distinguish non-family ties into friendship ties (e.g., co-owners who were friends) and professional ties (e.g., co-owners who were co-workers). A single entrepreneur can recruit multiple co-owners from different types of ties to found a mixed team. Among 535 entrepreneurial teams, there are 389 respondents (72.7%) who reach out family ties,Footnote 1 111 respondents (20.8%) who utilize friendship ties, and 75 respondents (14.0%) who rely on co-worker ties.

To investigate how the different ownership structures of female-led and male-led businesses influence firm performance, we limit our analysis to nascent businesses that have already received any kind of revenue (N = 570).Footnote 2 Using this sub-sample, we use a variable that asks whether the enterprise’s monthly revenue exceeded the monthly expenses at the timing of the interview. This variable can serve as a proxy for initial performance, especially given that detailed information on the size of sales or profits is not available for the nascent businesses under study in PSED II.Footnote 3 We identify high-performing businesses as those the monthly revenue of which exceeded the monthly expenses (i.e., producing profits). Among 570 enterprises, there are 276 businesses (48.4%) that produced profits at the time of interview.

4.3 Independent variables

This study primarily focuses on the effect of entrepreneurs’ sex on the ownership structure of new businesses. We examine whether female entrepreneurs tend to establish a solo or a family business instead of a non-family or a mixed business in comparison to male entrepreneurs. Thus, the key independent variable of our analysis is the respondents’ sex. In this binary variable, females are coded as 1, while males are coded as 0.

In addition, we include a series of independent and control variables in all of our models. First of all, we focus on the role of social and cultural capital in the formation of entrepreneurial teams. Past studies have also suggested that the entrepreneurs’ preexisting forms of capital matter in launching and running new businesses (Aldrich et al. 1995; Bosma et al. 2004; Bourdieu 1986; Fairlie and Robb 2009; Ruef 2010; Tatli et al. 2014). In this study, we examine whether entrepreneurs who maintain social and cultural capital are more likely to open non-family or mixed enterprises with non-family partners who have unique skill sets. In addition to examining the direct effect of social and cultural capital, we test whether the relationship between entrepreneurs’ sex and the business ownership structure is moderated by their possession of social and cultural capital.

To measure the social and cultural capital entrepreneurs maintain, we first include a social capital variable that asks whether the respondent-entrepreneur have a mentor in the start-up process. In this binary variable, we identify entrepreneurs with social capital as those who choose mentor’s influence as one of the two main opportunities that prompted them to launch their businesses. Secondly, we add a cultural capital variable that measures the number of people the respondents have previously supervised in workplaces. By asking their previous experience of coordinating others in workplace, this variable captures the leadership traits that entrepreneurs have previously developed.Footnote 4 Also, we include another cultural capital variable that captures the experience, education, talent, or expertise that the entrepreneurs self-reported in the survey. In this variable, we identify respondents who are self-confident with their cultural capital as one and zero otherwise. These respondents are entrepreneurs who choose prior work experience, formal training or education, or talent/expertise in the field as one of the two main reasons why they start their new businesses.

Additionally, we include eight control variables in all models: the respondents’ marital status, household size, race, age, educational backgrounds, financial resources they maintain, industrial experience, and managerial experience. First of all, entrepreneurs will have fewer opportunities to interact with non-family members when they get married or live in a large household (Wellman 1985). Thus, we expect that entrepreneurs who are currently married or who live in a large household are more likely to open family-owned businesses rather than non-family businesses. To address this possibility, we include a binary variable on the respondents’ marital status as well as a continuous variable on the number of people who regularly live in their household in our models. As to entrepreneurs’ race, we add the following binary variables: Whites, Blacks, Hispanics, and others. The variable for “others” is set as the reference category in our models. For the respondents’ age, we add both their ages and the squared ages under the rationale that the effect of age may possibly show a U-shape. To examine the effect of the respondents’ basic educational backgrounds, we use three binary variables as follows: having less than a college education, having a college education, and having graduate education. The variable for having less than a college education is set as the reference category. Also, we generate a financial resource variable by using a survey item that asks whether the respondents perceive the acquisition of capital as one of the two problems they face in initiating their businesses. In this binary variable, we identify those who do not feel the acquisition of financial resources as one of the main challenges in opening businesses as 1. Additionally, to specifically control for the entrepreneurs’ basic professional backgrounds, we include (1) the respondents’ previous work experience in the industry where they open new businesses and (2) their previous experience as managers in our analysis.

Finally, we control for the industrial type of nascent businesses since the industrial environment may differentially influence the formation of ownership structure. In particular, solo and family enterprises may be found less frequently in capital-intensive industries such as manufacturing and construction or knowledge-intensive industries such as finance, health, education, and business sectors in contrast to service industries such as retails and restaurants. Using information from PSED II, we include eight industrial type variables in all models: services (retail stores, restaurants, bars, or nightclubs), professional services (health, education, social, and business), finance (including insurance and real estate), manufacturing, construction, agriculture/mining, transportation/utilities/communication, and other industries.Footnote 5

The descriptive statistics of our dependent and independent variables are presented in Table 1. The correlation among the independent variables is reported in Appendix 1 (Table 6).

Table 1 Description of dependent and independent variables
Full size table

4.4 Analytic strategy

We employ a multinomial logistic regression model to explain the different ownership structures of businesses. Since the dependent variable is categorical, we estimate the likelihood that an entrepreneur will open a solo enterprise (p1), a family enterprise (p2), or a mixed enterprise (p3), relative to the reference category of opening a non-family enterprise (p4). The multinomial logistic regression model is formulated as

$$ \log\ \left(\frac{P_j}{P_4}\right)={\alpha}_j+{\beta}_j\cdot {X}_i\left(j=1,2,3;{p}_1+{p}_2+{p}_3+{p}_4=1\right) $$

where αj is a constant, βj is a vector of coefficients, and Xi is our set of independent variables. Each logits, log(p1/p4), log(p2/p4), and log(p3/p4), have their own constants (αj) and regression coefficients (βj) for each independent variables (Long and Freese 2005). The coefficients estimate the effect of independent variables, including the entrepreneur’s sex, on the log odds of falling into one of the ownership categories while controlling for other variables. In interpreting the results, we also provide the relative risk ratios (RRR) of the coefficients, which present the likelihood that the respondent falls into a solo or a family enterprise rather than a non-family one. We include the sample weight provided by PSEDII to correct for differential sampling probabilities when analyzing full sample. Also, we checked for the multicollinearity issue by computing variance inflation factors (VIFs), and we cannot detect any problem of multicollinearity.Footnote 6

In an additional analysis, we examine the differential recruitment of co-founders within entrepreneurial teams. In forming teams, entrepreneurs can recruit a family member, a friend, and/or a co-worker from previous workplace; since the dependent variables are binary, we utilize a series of binomial logistic regression models. We include the same set of independent variables that we use in the multinomial logistic regression analysis. While the regression coefficients are the log odds of recruiting the particular type of members (e.g., a family member, a friend, a co-worker), we also present the odds ratios in the main text to facilitate interpretation.

We also extend our analysis to investigate whether female-led and male-led entrepreneurs with different ownership structures display dissimilar initial performance. Since the dependent variable is binary, we use binomial logistic regression models. We first examine the direct effect of entrepreneurs’ sex on initial performance and then investigate whether the effect of gender on firm performance is moderated by specific ownership structures of new businesses. We include the same set of independent variables that we use in previous models to explain firm performance.

Finally, we replicate our main analysis on the relationship between entrepreneurs’ sex and the ownership structure of their businesses by analyzing a dataset with limited information on entrepreneurs but a larger number of observations: the Survey of Business Owners and Self-Employed Persons (SBO) provided by the US Census. Using this data, we run multinomial logistic regression models to investigate the robustness of our findings on the relationship between entrepreneurs’ sex on the ownership types of their businesses. Results from this supplementary analysis are presented in Appendix 2.

5 Results

We present our findings on the relationship between entrepreneurs’ sex and the business ownership type in Table 2.

Table 2 Multinomial logistic regression on the relationship between entrepreneurs’ sex and business type
Full size table

The results from the multinomial logistic regression analysis using PSED II suggest that female entrepreneurs are significantly more likely to found either a solo or a family enterprise in comparison to a non-family enterprise. Being a female is associated with a 2.212 times increase (b = 0.794, p < 0.001) in opening a solo enterprise relative to opening a non-family enterprise, even after controlling for the entrepreneur’s marital status and household size, various capital covariates, and socio-demographic characteristics. In addition, female entrepreneurs are associated with a 2.246 times increase (b = 0.809, p < 0.001) in opening a family enterprise compared to opening a non-family enterprise. These results provide support for hypothesis 1a that female entrepreneurs will be more likely to found a solo enterprise than male counterparts. When females form an entrepreneurial team, our results also support our expectation in hypothesis 1a that female entrepreneurs will be more likely to recruit family members rather than non-family members compared to male entrepreneurs. Different from hypothesis 1, however, the results show that being a female is related to a 2.445 times increase (b = 0.894, p < 0.05) in founding a mixed enterprise compared to a non-family enterprise. The results suggest that female entrepreneurs tend to launch solo, family, or mixed enterprises rather than non-family enterprises.

Among the control variables, the results indicate that married entrepreneurs are more likely to open a family-only business than a non-family-only one. Being married is associated with a 4.011 times increase (b = 1.389, p < 0.001) in founding a family enterprise relative to forming a non-family enterprise. Additionally, the entrepreneur’s previous work experience in the industry where one opens a business is positively related to founding a mixed enterprise rather than a non-family enterprise by 0.044 times (b = − 0.045, p < 0.05). Finally, the entrepreneur’s age matters for opening a solo enterprise. Our results support an inverse-U shape on the effect of age, such that entrepreneurs are more likely to open a solo business as they age but less likely to found a business alone after reaching a certain age.

We further investigate the moderating effect of the different forms of capital on the relationship between entrepreneurs’ sex and ownership type. The results are reported in Table 3.Footnote 7

Table 3 Multinomial logistic regression on the relationship among entrepreneurs’ sex, forms of capital, and business type
Full size table

Our findings suggest that female entrepreneurs who do not possess cultural and social capital are associated with a 3.600 times increase (b = 1.281, p < 0.001) in the odds of establishing a solo enterprise relative to forming a non-family team. This is a dramatic increase from the effect of being a female in general on founding a solo business, which was 2.212 times in Table 2. In addition, being a female entrepreneur is associated with a 3.435 times increase (b = 1.234, p < 0.001) in opening a family business rather than a non-family business. This is also an obvious increase from 2.246 times for female entrepreneurs in general in Table 2. The likelihood of female entrepreneurs to open a mixed team also increases from 2.445 times to 3.228 times (1.172, p < 0.05) when they do not have cultural or social capital. These results provide support for hypothesis 1b that the tendency of female entrepreneurs to found either a solo or a family enterprise will be stronger in the absence of social or cultural capital.

When female entrepreneurs are self-confident in their prior experience, talent, or expertise, on the other hand, their tendency to establish a solo business is offset; female entrepreneurs with cultural capital are related to a 0.057 times decrease (b = 1.281–1.340 = − 0.059) in founding a solo enterprise rather than a non-family firm. While the focus of our paper is on females who found their businesses, the results on the male counterpart are also noteworthy. As to male entrepreneurs, those who are self-confident in their experience, education, talent, or expertise are 2.243 times (b = 0.808, p < 0.01) more likely to form a solo business compared to males with limited cultural capital. These findings are opposite to the case of female entrepreneurs; while females with self-confidence tend to find a non-family partner in opening businesses, males with self-confidence are rather more likely to form a business alone.

In terms of social capital, the results show that the tendency of female entrepreneurs to open a solo business becomes weaker when they rely on a mentor in launching their businesses. For female entrepreneurs who maintain a relationship with a mentor, the likelihood of opening a solo business drops from 3.600 times to 1.383 times (b = 1.281–0.957 = − 0.059, p < 0.05 one-tailed). The relationship between being a female and opening a family business also decreases when entrepreneurs have developed leadership traits, which is a form of cultural capital. While female entrepreneurs are related to a 3.435 times increase in founding a family team rather than a non-family one, one unit increase in the leadership variable is associated with a 1.4% decrease (b = − 0.014, p < 0.05) in the case of females.

The significant moderating effects of social and cultural capital can be further expressed through marginal effects (Long and Freese 2005; Williams 2017). Figure 1 provides a visualization of the moderating relationship among entrepreneurs’ sex, the social and cultural capital they possess, and the ownership structure of their businesses. The two figures on the top illustrate the predicted margins of founding solo businesses, while the figure at the bottom shows the predicted margins of opening family businesses.

Fig. 1
figure 1

Predicted margins of opening solo or family businesses

Full size image

The top left figure shows that the predicted margin of opening a solo business is 5.9 percentage points higher for females without self-confidence than females with self-confidence. For males, on the other hand, the predicted margin of opening a solo enterprise is 14.0 percentage points lower when they lack self-confidence. We suspect that, while females without confidence are forced to found businesses alone, males with self-confidence may voluntarily choose to establish businesses by themselves. Further studies are needed, however, to provide a solid explanation to these distinct patterns between males and females.

Next, the top left figure illustrates that the predicted margin dramatically drops for female entrepreneurs when they possess social capital. While the predicted margin is 0.573 for females who do not maintain a relationship with a mentor, it is 0.446 for females with a mentor. As to males, on the other hand, the predicted margin of opening a solo business is not significantly influenced by their possession of social capital.

Finally, the figure at the bottom shows that being a female without any leadership experience in prior workplace is associated with a predicted margin of 31.4% of opening a family business, while the margin drops below 1% for females who have supervised 600 or more individuals. The results suggest that females with leadership traits are less likely to open a family business compared to females without those traits. A similar but less pronounced pattern is found for male entrepreneurs. The predicted margin of opening a family enterprise is 26.5% for males without leadership experience, but it decreases below 15% when they have supervised 1200 or more individuals.

Next, we examine the effect of entrepreneurs’ sex on the recruitment of different team members using binomial logistic regression models. These results are presented in Table 4.

Table 4 Binomial logistic regression on the relationship between entrepreneurs’ sex and the recruitment of team members
Full size table

Our findings suggest that female entrepreneurs do rely on family ties in founding entrepreneurial teams compared to their male counterparts. Being a female is associated with a 1.781 times increase (b = 0.577, p < 0.05) in recruiting a family member such as a spouse as a co-founder in establishing a team. The results provide support for our expectation in hypothesis 2a that females are more likely to found a family enterprise than males.Footnote 8 Additionally, being a female entrepreneur is significantly related to a 56.3% decrease (b = − 0.827, p < 0.05) in recruiting a previous co-worker as a co-founder in one’s team. In the case of recruiting a friend as a co-founder, there is no significant difference between males and females. The results support hypothesis 2b that female entrepreneurs will be less likely to recruit co-workers as their co-founders in comparison to male counterparts.

Our results also indicate that the marital status of an entrepreneur has a significant influence on the formation of entrepreneurial teams. The results show that being married is associated with a 3.525 times increase (b = 1.260, p < 0.001) in founding an entrepreneurial team with one’s family member, but at the same time, it is related to a 69.9% decrease (b = − 1.199, p < 0.001) in recruiting a friend and a 54.5% decrease (b = − 0.787, p < 0.05) in recruiting a past co-worker as a co-founder in one’s team. Additionally, the entrepreneur’s previous work experience in the industry where one opens a business is positively related to founding a business with a previous co-worker by 1.026 times (b = 0.026, p < 0.05). An entrepreneur’s age is also related to a 0.273 times decrease (b = − 0.319, p < 0.05) in recruiting a friend as a co-founder in one’s team. In other words, the results show that entrepreneurs are less likely to form a team with a friend as they age.

We move on to our analysis on how entrepreneurs’ sex and their ownership type restrict the initial performance of their businesses (Table 5).

Table 5 Binomial logistic regression on the relationship between entrepreneurs’ sex and initial firm performance
Full size table

In the first model, the results suggest that being a female is associated with a 0.498 times decrease (b = − 0.689, p < 0.001) in displaying a positive initial performance even after controlling for the entrepreneurs’ social, cultural, and economic capital as well as other socio-demographic characteristics. On the other hand, no direct relationship is found between the ownership type of nascent businesses and their initial performance.

In the interaction model, we further examine how the relationship between entrepreneurs’ sex and their initial firm performance is moderated by specific ownership type. Supporting our expectation in hypothesis 3, the results clearly show that being a female is associated with less promising performance after they launch either a solo or a family business rather than male-led non-family business. Females who found solo businesses are negatively related to showing promising initial performance by 78.3% (b = − 1.529, p < 0.05) compared to the reference group, which is males who establish non-family enterprises. Also, females engaging in family businesses are associated with a 87.5% (b = − 2.079, p < 0.01) decrease in displaying favorable performance than males founding non-family firms. The results on female-led family businesses are in a clear contrast to our findings on male-led family businesses; male-led family enterprises are related to a 2.855 times (b = 1.049, p < 0.01) increase in their odds of showing promising performance compared to male-led non-family ones. Additionally, the initial performance of female entrepreneurs who launch non-family businesses is not significantly different from male-led non-family firms (b = 0.974).

In sum, our analyses support our hypotheses on the gendered formation of entrepreneurial businesses. Our findings suggest that entrepreneurs’ sex has a significant influence on the ownership type of nascent businesses. The tendency of female entrepreneurs to found a solo business becomes more salient when they lack social or cultural capital such as mentorship or self-confidence. Also, the odds of females to found a family business increases when they lack cultural capital such as leadership traits. Additionally, females tend to form an entrepreneurial team by recruiting family members rather than non-family members such as co-workers in the previous workplace. Finally, female-led solo or family businesses are negatively associated with promising performance at the initial stage than other business types. Our findings support our theoretical expectation that the limited nature of the social and cultural capital of females structures the manner in which female entrepreneurs open and operate their businesses.

6 Conclusion

This study illuminates the influence of entrepreneurs’ sex on the formation of ownership structure in nascent businesses. Challenging the gender-neutral assumption that male and female entrepreneurs share a common social environment, we find that females are constrained to establishing either a solo business in comparison to males especially when they lack social or cultural capital. Also, when they establish an entrepreneurial team, female entrepreneurs tend to utilize family ties rather than professional ones compared to their male counterpart. The performance of nascent female-led businesses is also restrained when females launch a solo or a family enterprise. Our study provides important implications to the literature of gender and entrepreneurship.

First of all, our study expands our understanding of the different social environment that female entrepreneurs encounter. Despite the dramatic growth of women’s participation in entrepreneurship, females are still less likely to form entrepreneurial firms compared to men. In addition to the gendered stereotype linking entrepreneurship to masculinity (Gupta et al. 2013; Gupta et al. 2009), our study reveals an additional challenge faced by females, which is their lack of social and cultural capital (Aldrich and Cliff 2003; Carter and Williams 2003; Cromie and Birley 1992; Fairlie and Robb 2009; Greve and Salaff 2003; Klyver 2011). While establishing an entrepreneurial team can be beneficial in certain industries, females who are limited in their social and cultural capital are less likely to recruit non-family partners who can bring unique skill sets to their team. Due to their limited resources, females are forced to launch their firms either by themselves or with family members such as spouses. Moreover, we find that female-led solo and family businesses are associated with less promising performance in their initial stage. Entrepreneurship research can benefit from extending its focus to the different forms of capital that females possess to launch and to successfully operate their businesses.

Our study also contributes to the literature on family and entrepreneurship. Recent studies have increasingly emphasized the embedded nature of entrepreneurial activities within family relationships (Aldrich and Cliff 2003; Brannon et al. 2013; Greve and Salaff 2003; Orhan 2001). Until now, however, few studies have shed light on how female and male entrepreneurs differentially recruit their family members in their new firms (Ruef et al. 2003). This current study shows that females—who tend to have limited social and cultural capital than males—can utilize family networks as a substitute for their limited professional networks. Our results imply that female entrepreneurs can recruit emotionally-intimate family members to their nascent firms as a way to both add a complementary skill-set to the firm and attain a better balance between work and family. At the same time, however, our study also suggests that their reliance on family ties can lead to weaker firm performance.

Additionally, our study illuminates the connection among entrepreneurs’ sex, the ownership structure of their businesses, and their firm performance. Previous studies have suggested that female entrepreneurs tend to launch businesses that are smaller in size and lower in initial funding, which often leads to negative entrepreneurial outcomes (Alsos et al. 2006; Boden and Nucci 2000; Fairlie and Robb 2009; Robb 2002). Our research offers a new approach to furthering our understanding of the gender gap in entrepreneurial performance. Our findings infer that the performance gap between males and females already emerges among nascent businesses at their very early stages. We show that the initial performance of entrepreneurial businesses is possibly influenced and constrained by a combination of entrepreneurs’ sex and the initial ownership structuring of their businesses.

Notwithstanding the robust findings supporting the relationship between entrepreneurs’ sex and the ownership structure of businesses, we acknowledge that rooms remain for future studies to deepen our understanding of the gendered formation of entrepreneurial businesses and its consequence. First of all, we were not able to directly control for the preexisting social relationships of entrepreneurs who prepare to launch a firm. Data with complete information on the social networks of female entrepreneurs will allow researchers to further illuminate the exact role of social capital in the launching of entrepreneurial businesses. Also, the role of gendered stereotypes on entrepreneurial activities is not fully examined in our study of the formation process of entrepreneurship. While we indicate the lack of social and cultural capital among females as the main mechanism through which females are more likely than males to open either a solo or a family enterprise instead of a non-family enterprise, we are not certain about how the stereotypes linking entrepreneurship with independence and masculinity influence the decisions by females to launch a firm by themselves rather than to form a team with partners. Future studies are needed to disentangle the relationship among gender stereotype, the formation of entrepreneurial firms, and their future performance. In addition, our analysis on business performance is limited due to our exclusive focus on nascent businesses—many of which are yet to earn any revenues. While the data on nascent businesses we use is appropriate to study the initial structuring of female and male entrepreneurs, a longitudinal study including businesses that either survived or died during the initial stage would be needed to fully examine the gendered outcomes of firm performance. A more rigorous longitudinal analysis can be conducted to further reveal the causal relationship among entrepreneurs’ sex, the ownership structure of their businesses, and their entrepreneurial success.

In this study, we challenge the individualistic assumption that the decision of economic actors is exclusively based on their rational calculation to maximize their utility as atomic actors. Our research suggests that the decisions of entrepreneurs to recruit a co-founder in their firms are embedded in the gendered social system to which they belong. As resources such as social and cultural capital are unequally allocated to females and males, we conclude that the decision of entrepreneurs to organize and operate a business is heavily structured by entrepreneurs’ sex.