Trusting female entrepreneurs

(1)

Bachelor’s Thesis

Trusting Female Entrepreneurs

Bsc Economics and Business

Specialization: Finance and Organization Economics FEB/UVA

Supervisor: T. Buser 28 June 2014

By: Frank Ooijevaar Student Number: 5744288

(2)

1. Introduction and Motivation

The last decades there is a rise in the willingness of the business-world to get women to work at higher functions (Eagly, 2007). Although this goal isn’t always achieved, it is evident that the world is changing and the diversity in management positions and other functions is growing. In entrepreneurship the women are taking a stronger position every year. Although the number of female entrepreneurs worldwide is slightly lower than that of men (GEM, 2013), in recent years the number of starting female entrepreneurs have become bigger than that of male (Minniti & Naudé, 2010). For developing countries female entrepreneurs could be seen as the new drivers to a better economic environment and the World Economic Forum identified women entrepreneurs as the way forward at their annual meeting in 2012 (WEF, 2012). This could be seen as a worldwide trend, but this trend is more evident in developing countries than it is in developed countries.

If an entrepreneur wants to start or expand their company, it is most likely that they will need investments to do so. These investments could be done by the government or by independent investors. Most investments are based on what the investor thinks of the company. Investors are more likely to invest if companies have a distinguished risk and reward profile and scalability. But is there a difference in the willingness to invest if the entrepreneur is male or female? Based on the above, the research question is formulated as follows: is there a difference between the trust in female and male entrepreneurs?

This question is also based on the gender gap that still remains evident for most countries (WEF, The Global Gender Report, 2013). Although there is progress on closing the gender gap in the past years, there are still differences in the rates of women working at higher functions compared to men (Bertrand and Hallock, 2000). Since entrepreneurship was a primarily male based type of business, it could be useful trying to find out if there is a difference here as well. And if a difference is found, to find out what causes this difference. It could be that there is a difference because the trust in male and female entrepreneurs differs for each investor, but it could also be possible the difference is created by investor’s preferences. These preferences could be a lower amount of risk taken with the investment, or a higher return made on the investment. These investor preferences and differences can be found by using a short survey. From this survey the preferences of investors, risk-loving or risk-averse, and the differences between preferring male

(4)

or female entrepreneurs could be assessed. Also preferences could become clear by using crowdfunding websites.

Crowdfunding is an upcoming solution for entrepreneurs to gather investors and/or the needed funds to start their own business, fund the costs of a new project or fund the costs to finish a project. (Larralde & Schwienbacher, 2010). Entrepreneurs seek external finance from the crowd, instead of using the traditional paths, such as banks, venture capitalists, private individuals, leasing, factoring suppliers/customers, partners/working shareholders (Cosh et al., 2009). The Internet helps these entrepreneurs to find the investors willingly to take part in their projects (Kleemann et al., 2008).

The crowdfunding websites can show if there is a difference between the investments made in male and female entrepreneurs. It can also show which type of projects are funded and if the return a project makes matters for the funded amount.

After this introduction the related literature will be discussed. The most related papers and/or articles will be discussed and it will be made clear what there is studied and what the outcomes were. Also there will be a comparison made between these papers and this thesis and how these papers can contribute in providing a better outcome for the survey done for this thesis. In the main section the hypotheses will be stated and explained where they come from. Also the survey will be explained and where the other data should be found. In the final section the results will be compared and discussed and there will be a conclusion where the outcomes will be assessed.

(5)

2. Related Literature

In the following section the related literature will be discussed. It will take a detailed look at the gender-gap situation, female leadership, and crowdfunding. All aspects are important for the research and could be directly or indirectly linked when answering the research-question.

There are studies done on the effects of the gender-gap that seems to remain evident. Not only is there an evident trend that female entrepreneurship is increasing and has the opportunity to become a driving force in economic reconstruction, but also the gender-gap remains (Minniti & Naudé, 2010). Meaning, women entrepreneurship increases, but the differences in influence, money earned and wages are still there. Although this paper mainly focusses on developing countries, it can still be used to overlook the gender-gap situation and maybe help on how to overcome this gap.

Other literature that could be used, involves studies on female leadership (A.H. Eagly, 2007), where the advantages and disadvantages of female leadership in the United States are assessed. Women increasingly have been praises for their leadership. They even show leadership styles associated with effective performance as leaders, more than men. Nevertheless, more people prefer male over female bosses, and it is more difficult for women to become leaders and to succeed in male-dominated leadership roles. This mix of apparent advantage and disadvantage that women leaders experience reflects the considerable progress toward gender equality that has taken place in both attitude and behavior, coupled with the lack of complete attainment of this goal. This study can be used as a starting point to find if this difference exists with entrepreneurs as well. As entrepreneurship used to be a very male-based type of business, it is interesting to find out if the differences found for the United States leadership-roles are also found in the willingness to invest in either male or female entrepreneurs.

Also literature assessing the trust of investors in entrepreneurs is important. The entrepreneur and the investor need to balance the level of control and trust building mechanisms so that the optimal level of confidence in partner co-operation can be achieved (Sheperd & Zacharakis, 2001). This article proposes that the entrepreneur can build trust with the investor by signaling commitment and consistency. Open and frequent communication acts as a catalyst for the other trust building mechanisms. This article can be used to define where the found differences are coming from.

(6)

Kleemann (2008) investigates the phenomena of crowdsourcing, or the outsourcing of tasks to the general Internet public. This phenomena is evidence of historically significant change in the relations between firms and their customers. Kleemann (2008) uses the definition: “Crowdsourcing takes place when a profit oriented firm outsources specific tasks essential for the making or sale of its product to the general public (the crowd) in the form of an open call over the Internet, with the intention of animating individuals to make a contribution to the firm’s production process for free or for significantly less than that contribution is worth to the firm.” Basically crowdfunding does the same thing. Only in this case the entrepreneur is asking for a financial contribution. This contribution is compensated with a promised return, a percentage of the profit if the project becomes profitable, or with a gift depending on the amount funded by the investor. These gifts differ per project. It can be a provided service, a membership in the projects outcome or part of the outcome of the project. This could be a factor that should be taken into account when creating the model from the data-sets attained from the crowdfunding-websites. In their article, Belleflamme, Lambert and Schwienbacher (2013) state that an entrepreneur raises financing from a large audience, the crowd, in which each individual provides a very small amount, instead of soliciting a small group of sophisticated investors. They compare two forms of crowdfunding. In the first form, entrepreneurs solicit individuals to pre-order the product, and in the second form, entrepreneurs solicit individuals to advance a fixed amount of money in exchange for a share of future profits or equity. In this research, the return doesn’t play a very large role, at first, but it could be taken into account later on (if this data could be obtained). Furthermore Belleflamme et al. (2013) assume that crowdfunders enjoy community benefits that increase their utility. They use a unified model in which they show that the entrepreneur prefers pre-ordering if the initial requirement is relatively small compared with market size and prefers profit sharing otherwise. From this model they conclude the entrepreneur needs to build a community of individuals with whom he or she must interact. The statement that a small requirement is preferred and that an entrepreneur needs to build a community could become evident from the model created in this paper as well. If an entrepreneur sets a smaller goal that is needed to make a project successful, this goal could be easier achieved and therefore preferred by the entrepreneur. The community in this case could be the number of investors in the project. In the case of profit sharing there need to be looked at the projects that actually give a return when profitable.

(7)

In their article Bertrand and Hallock (2000) analyze the differences among top executives in a large set of US public corporations. It was the first detailed description of the relative position of female top executives in the 1990’s. The unexplained gender gap could be due to labor market discrimination, but it could also be attributed to some unobservable differences between men and women, such as a relative lack of long-term career commitment for women. But these differences are assumed to be minimized in the examined group of top executives, e.g. men and women in the sample tend to have the same level of high job motivation and high career ambitions. The data used is attained from ExecuComp and contains information on total compensation for the top five executives for all firms in the S&P 500, S&P Midcap 400 and S&P SmallCap 600 from 1992 up to 1997, and includes information on base salary, bonus and the value of granted stock options in the current year. This large sample size has multiple advantages. It provides sufficient statistical precision where female representation is small; it covers a variety of occupational categories among top managerial jobs and not only differences at the top; and the role of firm size and industrial specialization in the gender compensation can be assessed.

With their paper Bertrand and Hallock showed that although the number of women involved in the top-level management of US corporations is small, about 2,4% of the executives are women, it has increased substantially in recent years. Their results further indicate that the gender gap in compensation among top executives is at least 45%. An important fact is that female managers are under-represented in large corporations. Because the returns to firm size are very high among top executives this explains up to 15 percentage points of the gender gap. The scarcity of female CEO’s, Chairs, Vice-Chairs and Presidents accounts for as much as half of the unconditional gender compensation gap. There is no significant evidence of a concentration of women in the lower compensation occupations. The unexplained gender compensation gap for top executives is less than 5% after one accounts for all observable differences between men and women. Because top executives probably constitute a fairly homogeneous group with respect to job motivation, career commitments and human capital, finding an unexplained gender compensation gap in this sample could have reasonably been interpreted as evidence for taste discrimination against women.

Another way that is used by Bertrand and Hallock to consider wage gaps is the Oaxaca decomposition (1973). This method decomposes the overall gap into a portion that is due to

(8)

differences in observable skills between groups and a part that is still unexplained. This could be easily done by running separate regressions for men and women and then rewriting the overall wage gap in various ways. The outcomes of this method show that most of the total gap in compensation by gender for top managers (between 71% and 88%) is due to observable differences between men and women.

A major drawback of the ExecuComp data-set Bertrand and Hallock face, is that it doesn’t report age and tenure consistently for all observations. These two variables are available only for a subset of the observations in their sample. But since the return to age and experience, of the remaining sample, are large in the market of executives they expect that the relative youth and low seniority of the female executives is another important determinant of the gender gap. The relative youth of women cannot in itself fully explain the gender gap; 33% difference in compensation still exists between men and women after they have accounted for age and seniority. These findings, while imprecise, indicate that the gender compensation gap could be less than 5% after controlling for all observations.

In their article Croson and Gneezy (2009) review experimental evidence on preference differences between men and women. They focus on three factors that have been extensively studied: risk preferences, social preferences, and reaction to competition. From the reviewed experiments they have found that women are indeed more risk-averse than men. They find that the social preferences of women are more situational specific than those of men; women are neither more nor less socially oriented, but their social preferences are more malleable. Finally, they found that women are more averse to competition than are men.

When looking at risk preferences Croson and Gneezy (2009) reviewed objective probability lotteries, portfolio selection, and multiple explanations, like emotions, overconfidence and risk as challenge or threats, for the differences between men and women. All led to the same outcome, women are more risk-averse than men. Only exceptions to the rule are managers and professional populations, in which the differences in financial risk preferences are smaller than in the general population and often nonexistent. Major factors for the gender differences in risk taking are found in the affective reaction to risk. Men and women differ in their emotional reaction to uncertain situations and this differential emotional reaction results in differences in risk taking. Men are also more confident than women and, as a result, may have different perception of the probability

(9)

distribution underlying a given risk. Finally, men also tend to view risky situations as challenges, as opposed to threats, which leads to increased risk tolerance.

For the differences in social preferences Croson and Gneezy looked at how strongly social preferences manifest themselves in men and in women. This is studied by looking at when individuals exhibit a social preference, when others’ payoffs enter into their payoff function. To derive at a decision they reviewed multiple experiments using ultimatum games, dictator games, trust and reciprocity, prisoners’ dilemmas, social dilemmas, and public goods provisions. Identifying the outcomes of these experiments results in gender differences, however many of the found results are contradictory. They found that in some of the experiments women are more altruistic, inequality averse, reciprocal and cooperative than men, and in others they are less so. Croson and Gneezy believe that the cause of these conflicting results is that women are more sensitive to cues in the experimental context than are men.

Last section Croson and Gneezy looked at, is the gender difference in attitudes toward competition. Their recent findings suggest that women are more reluctant than men to engage in competitive interactions like tournaments, bargaining and auctions. Additionally, men’s performance, relative to women’s, is improved under competition. Thus as the competitiveness of an environment increases, the performance and participation of men increase relative to that of women. Croson and Gneezy used three components to examine the competition aspect. First they looked at the reaction to competition. Their findings suggest that men’s performance is more affected by the competitiveness of the environment than women’s performance. Second, they used self-selection of the incentive scheme. Their findings showed that more men than women chose the competitive environment and that women are less likely to choose to compete than men. Yet, women who choose competitive environments perform just as well as men in those settings. Last, they used the bargaining aspect, in which avoiding competition can have a strong impact. Croson and Gneezy found that in bargaining situations, women are less likely to exhibit competitive preferences than men, slightly in their reactions once in a negotiation, but more significantly in their propensity to engage in a negotiation at all.

Klapper and Parker (2010), show that policymakers increasingly are exploring ways of promoting economic activity and growth among women in developing countries. It is becoming apparent that female entrepreneurship represents a potentially valuable tool for promoting growth and

(10)

reducing poverty. Yet relatively little is known about the role of female entrepreneurship, especially in developing countries, or about the opportunities and barriers that women entrepreneurs face in practice.

In their paper, Klapper and Parker (2010) address the recurring question of whether individual choices or the business environment are primarily responsible for lower rates of female entrepreneurship and their lower average financial performance. They assess this by using multiple factors, like women’s engagement in entrepreneurship, women’s performance in entrepreneurship, and the challenges in the business environment and policy implications.

First, women’s engagement in entrepreneurship. Klapper and Parker state that the basic facts about women’s rates of participation in entrepreneurship are stark. Regardless of whether entrepreneurship is defined in terms of new venture creation, business ownership, or self-employment, a higher proportion of men than women engage in this activity in industrialized economies. Women entrepreneurs in industrialized countries are also more likely to be part-time workers. Another important gender difference in entrepreneurial participation is the industries in which businesses are established. Women entrepreneurs remain heavily over-represented in a few industry sectors, especially sales, retail, and services. Industry concentration is important because it has implications for the preference of women-owned ventures. The sectors that women cluster in are typically characterized by smaller scale, more intense competition and lower average returns.

Second, Klapper and Parker discuss the findings in women’s performance in entrepreneurship. They have found that there is a broad agreement among researchers that, in both industrialized and developing countries, women entrepreneurs earn less income then male entrepreneurs. A similar story applies if performance is measured in terms of turnover, employment creation performance, growth rates, or survival prospects. Klapper and Parker explain these findings again with the fact that women entrepreneurs tend to be concentrated in industries with lower capital intensities and lower average returns to capital, as well as the fact that women on average operate smaller businesses, utilizing less capital and finance from banks and other lenders than men do (Aronson, 1991; Carter, Williams and Reynolds, 1997; Watson, 2002). Also, it might be thought that greater risk aversion among women than men might explain their lower average returns and

(11)

growth performance in entrepreneurship, leading women to choose positions further down the expected profit-risk frontier than their male counterparts (Watson and Robinson, 2003).

Last, Klapper and Parker look at the challenges in the business environment and policy implications. They review possible institutional barriers to female entrepreneurship, including credit constraints, property rules, and adverse social norms. Evidence from cross-country studies show that women receive a lower share of available external funding than men for business and other purposes. Yet the evidence suggests that this might be less attributable to explicit discrimination than to weaknesses in the business environment that make lending to women a higher credit risk.

None of the above mentioned articles have researched if there is a difference in the trust of investors in different gender entrepreneurs. By taking multiple factors into account, there could be assessed if there really is a gender preference or that most investment decisions are made based on risk and return. By using a survey the general preferences will become clear, and by creating a model, there could be assessed what the outcome of the actual data will be.

(12)

3. Implementation

3.1 Hypothesis

To solve the research question, and find if there is a difference between the trust in female and male entrepreneurs, a hypothesis needs to be stated. Based on the found assumptions in the related literature the following hypothesis is chosen: There is a difference in trust between male and female entrepreneurs.

To examine if this hypothesis is true or not, multiple sources should be used. These sources are the Global Entrepreneurship Monitor (GEM), a small survey and a model created from a data-sets which is obtained from multiple crowdfunding websites. In the following section the use of these sources will be explained.

There is some data to be found at GEM on female entrepreneurship. GEM annually assesses the entrepreneurial activity, aspirations and attitudes of individuals across a wide range of countries (GEM) However, these data aren’t very specified and can’t provide a clear overview of the trust or investments made in different entrepreneurs.

By taking a small survey under students, businesspeople, investors and entrepreneurs there could be assessed if people are more likely to invest in a male or female entrepreneur. Not only the investment preference but also their confidence and/or beliefs in different gender entrepreneurs could be assessed. These differences can be found by sketching multiple entrepreneur profiles and letting the respondents choose between two profiles. These profiles will differ for each question. The differences will be in gender, risk-taking and return. From these differences the preferences of the respondents between gender or riskiness will come clear.

Also crowdfunding websites will be used to find out if there is a difference in investments made in projects of male and female entrepreneurs. From these websites multiple data-sets will be used to create a model which will access if there are differences to begin with, and if these differences will become bigger or smaller if other variables are added. The starting point for this model will be the differences between male and female entrepreneurs when looked at the reached funded percentage. This percentage is found by dividing the total funded amount by the total needed amount. Next, different variables will be added to see if the difference becomes smaller or bigger.

(13)

These variables are: total needed amount, number of investors and the days of funding. The same regression will be done on the variables for goal amount and funded amount as well.

3.2 Data collection

GEM (=Global Entrepreneurship Monitor): annually assesses the entrepreneurial activity, aspirations and attitudes of individuals across a wide range of countries. GEM is the largest ongoing study of entrepreneurial dynamics in the world. It explores the role of entrepreneurs in national economic growth, unveiling detailed national features and characteristics associated with entrepreneurial activity. The program has three main objectives: measure differences in the level of entrepreneurial activity between countries, uncover factors leading to appropriate levels of entrepreneurship, and suggest policies that may enhance the level of entrepreneurial activity. (GEM)

Make use of crowdfunding websites to look for differences in investments made in projects of male and female entrepreneurs. The data from multiple crowdfunding websites will be used to create a model. These websites are chosen from an article in the Dutch newspaper De Telegraaf (24-12-2013), in which the eight best, Dutch, crowdfunding websites are given. From these eight websites, only six had clear data available. The created data-sets consist of the common information like name of the project, name of the entrepreneur, the amount needed for the project, the amount funded at the end of the funding period or funded so far, the number of investors, the number of days remaining to fund and, if given, the return.

A survey will assess if there is a difference between the preferences of “investors” (the respondent gets the role of investor) between male and female entrepreneurs. Multiple profiles will be sketched and the respondent will be asked to choose one of two profiles per question. The profiles will be different types of entrepreneurs (high risk versus low risk), but also different gender entrepreneurs. From this survey there will become a trend evident if the respondent is just risk-loving or risk-averse or if there is a clear difference in the preferences between male and female entrepreneurs. The survey is created by using the website www.thesistools.com

(http://www.thesistools.com/web/?id=421702) and can be found in the Appendix, under Appendix 1.

(14)

3.3 Creating a model from the crowdfunding data-set

With the data attained from the crowdfunding websites a model can be created by using Stata. In this model there will be tested if there are any differences between male and female entrepreneurs. This difference will be based upon the difference found by comparing the different genders. To find differences the crowdfunding data-set will be put into Stata. The most important variables that are going to be used for the gender comparison, are the amount that was set by the entrepreneur, which is named Goal, the amount that is funded by the investors, named Funded, and the percentage of this goal amount that is actually reached by crowdfunding, named % reached. By creating dummy variables for male and female and comparing these with the other variables, differences could become evident. Ordinary Least Squares (OLS) regression will be used to find the different parameters for the model. And with a t-test the different sets of data (male and female) can be tested to see if the hypothesis is or isn’t supported.

(15)

4. Results

4.1 Survey Results

4.1.1 General Statistics and Overall Results

With the survey a total of 101 were surveyed. This group consisted of 59 males and 42 females. The outcomes of the survey are presented in Appendix 2 by frequency and by percentage. The general statistics of the respondents are presented in Appendix 2, Table 1. As shown in the table the respondents are split into four groups, based on occupation. These four groups consist of students (49), entrepreneurs (14), investors (16) and others (22). The survey was done online, so only the results of the respondents are shown and it wasn’t possible to find out who responded and who didn’t. The respondents were collected by sending out an email to multiple investing firms and to multiple starting as well as established entrepreneurs. Also students were asked to participate, and a group that didn’t fall into one of the other three groups, which consisted mostly of working men/women, with other professions than investor or entrepreneur. The preferences of each of those groups will be discussed in the next section and the tables can be found in the Appendix.

In Appendix 2, Table 2 the overall results are presented. These results show the preferences of the respondents per question. As shown in the first part of the table, the gender preferences for male are male, and for female are female. In the second part of the table the risk preferences are shown, and these risk preferences remain evident throughout the rest of the overall results. Males aren’t loving nor averse, but females show a clear preference for low risk, meaning risk-averse. The combined preferences remain evident for the choices of sketched entrepreneur profiles that have to be made in the following questions of the survey. For every sketched profile the male respondents have chosen for the male entrepreneurs. Some male respondents are indifferent if the risk between the profiles is the same, but most chose the male entrepreneur. For the female respondents it’s the other way round. They don’t stick with their gender preferences, but will choose for their risk-preference, being risk-averse. This is clearly shown in the first project preference question where a high risk project of a female entrepreneur is compared with a low risk project of a male entrepreneur. Although the female respondents showed a general preference for their own gender, most of them chose for the low risk project of

(16)

the male entrepreneur. This shows that the female respondents will rather choose for their risk preference than their gender preference.

The outcomes of a Pearson’s Chi-squared test and a Fisher’s exact test can be used to show the significance of the results. First the Chi-squared test is used. With these results, the p-values of each part of the outcomes can be calculated by using the Fisher’s exact test and then adding up all those p-values. This adds up to a total of 0.906, so the difference is significant.

4.1.2 Student Results

There were almost the same number of male and female student respondents (24 vs 25 respectively). The male students preferred the male entrepreneurs and the female students were almost indifferent between male or female entrepreneurs (11 vs 12 respectively and 2 being indifferent). In risk preferences, both male and female show a clear preference for low risk. When comparing the outcomes for the different sketched entrepreneur profiles, again, there is found that the male respondents stick with their gender preferences and will always choose the male entrepreneurs. For the female respondents they stick with their risk preference and prefer the low risk projects. Only exception for female students, compared to the overall results, is that they prefer the high risk project of a male entrepreneur over the high risk project of a female entrepreneur.

4.1.3 Entrepreneur Results

Unfortunately, there was only one female entrepreneur respondent. So the outcome of this section is mostly based on the male results. Gender preferences are the same again. The male entrepreneur respondents prefer male and the female respondent prefers female. For the risk preferences, both male and female prefer the high risk projects. The results of the different projects is the also the same. Male respondents prefer the projects of the male entrepreneur and the female respondent stays with her risk preferences and chooses for the high risk projects. When the risk of the projects of male and female entrepreneurs is the same, both male and female respondents will choose for their own gender.

4.1.4 Investor Results

The majority of the male respondents were indifferent when looking at gender preferences. Three male respondents preferred male and the rest was indifferent. For the female respondents there is

(17)

a clear preferences for female. The male respondents prefer the high rick projects, and the female respondents are indifferent; three preferred male, three preferred female, and three were indifferent. Comparing the different projects, male investors prefer the high risk projects over the low risk projects and when the risk is the same, they will choose for the male entrepreneur. Female investors have an overall preference for the female entrepreneur for all different projects.

4.1.5 Other Results

The last group surveyed was the group defined as “Other”. This is the group of respondents which aren’t a student, entrepreneur or investor. Both male and female respondents have a gender preference for male. Also both prefer the low risk projects. The preference of most respondents is for the low risk projects when the risk of the projects isn’t the same, and when the risk is the same they will choose the projects computed by their own gender.

(The outcomes of sections 4.1.2 to 4.1.5 are shown in Appendix 2: Table 3 – 6) 4.2 Crowdfunding data-set Results

For the model the data from six crowdfunding websites is used. On these websites 215 projects have been found and put into excel. Of these 215 projects, the majority are started by male entrepreneurs; they account for 130 projects. Of the remaining projects, 49 are of female entrepreneurs and the remaining 36 are started by a combination of both male and female entrepreneurs. The variables used in excel consist of name of the project, gender of the entrepreneur, goal amount, funded amount, percentage reached, number of investors, days to go for the funding period, type of project, name and age of the entrepreneurs, and the return. Most of this information can be found on these websites, but there are some missing values. For the variable age, the same problem as Bertrand and Hallock faced quickly became evident. Only a few of the 215 projects were able to provide information about the age of the entrepreneur, so therefore this variable is left out of the regression.

This excel data is then imported into Stata and will be used for regression to find the difference between male and female entrepreneurs. To find this difference a dummy for female has to be created. This dummy will give the female entrepreneurs a value of 1, and the male entrepreneurs and the projects managed by multiple entrepreneurs of different gender a value of 0. When using a regression including the dummy, the difference can be found. For this regression the variables

(18)

goal amount, funded amount, percentage reached, number of investors, days to go for the funding period and the dummy variable are used. (The outcomes of the regressions are given in Appendix 3).

First the variable reached is used as dependent variable on which the other variables are regressed, second the funded amount is used, and finally the goal amount is used as the dependent variable. From the outcomes, shown in Appendix 3 Table 7 column 1, there is found that for regressing female on the dependent variable, percentage reached, the coefficient is negative. This means that there indeed is a difference between male and female entrepreneurs when looking at the percentage reached. In the table it is shown that this negative effect is even more than ten percent. When more variables are added to the model, the outcomes of the coefficient for female remains negative, but it does become smaller. Eventually, when all variables are added, shown in Appendix 3 Table 7 column 4, the coefficient for female remains negative at -0.0854858, meaning around minus 8.5 percent.

Looking at the regression where the funded amount is the dependent variable, again a negative coefficient is found. As shown in Appendix 3 Table 8 column 1, this negative coefficient for the female dummy again is very large. But, when adding more variables to the model this coefficient shrinks again. When all variables are added, shown in Appendix 3 Table 8 column 4, the coefficient for female is -7772.733, which is much smaller than the coefficient of -23170.36 it started with. But again there can be concluded that there indeed is a difference between the male and female entrepreneurs.

Appendix 3 Table 9 shows the outcomes of the regression where the goal amount was used as the dependent variable. Columns 1 to 5 represent the outcomes where all variables are used for regression. The outcomes for the coefficient for female, again, is negative, but becomes less negative as more independent variables are added. Columns 1, 2, 6 and 7 represent the outcomes when the independent variable percentage reached is left out of the regression. Again the coefficient for female is negative, but it is in range with the outcomes found in the total regression (-6395.88 in total vs. -8944.865 when reached is left out). When the amount funded is left out of the regression, shown in columns 1, 8, 9 and 10 the coefficient is also negative, but is much bigger than the outcomes of the total regression (-20421.35). From these regressions, there can be concluded that there indeed is a difference between men and women when looking at the

(19)

goal amount. This means that on average female entrepreneurs ask for a lower goal amount than male entrepreneurs.

5. Discussion and Conclusion

5.1 Limitations

Although most data was available on the crowdfunding websites, there were still some missing variables. The most important variables used, like the goal amount to be funded and the percentage of goal amount reached, are therefore missing some variables. This could influence the overall outcome and therefore, possibly, create an incorrect outcome. Also it is very difficult to implement what the influences of the type of project are. It could be possible that different types of projects attract more (or maybe less) investors. Another missing variable that could be important is the age of an entrepreneur. The age of an entrepreneur could influence the decision of an investor to invest in a project. So it would be useful to find out how these different types could be implemented into a model to monitor if they influence outcomes.

Also, to create the best possible model to test the outcomes, it would be best for all crowdfunding websites to keep track of all their projects. With all data available, the testing of the different variables would become easier to do, and should give a better outcome as all the data is now available. To show the overall differences between male and female, crowdfunding websites from all over the world should be taken into account as well. For this research only the eight best crowdfunding websites from the Netherlands have been used (of which only six had data available). A good website to take into account would be www.kickstarter.com, as this websites alone has more data than all the Dutch websites combined (KickStarter, 2014). Through KickStarter 153234 projects have been launched of which 64185 ended up being successful. Due to the lack of time and that a clear overview of all data wasn’t available for this website, it wasn’t possible to take this website into account as well.

Limitations for the survey were the amount of respondents for some groups. In the survey the respondents were asked what their occupation is: student, entrepreneur, investor or other. Most important for this survey would be the general preferences of the investors and entrepreneurs. But as the outcomes of the survey revealed, most respondents were students. As everybody was asked

(20)

to answer the questions from an investors’ perspective, it should give a somewhat accurate outcome, but not the best outcome possible. A better way to find the overall preferences is to only survey investors. This would provide a more clear and accurate outcome of the overall preferences of the investors.

5.2 Conclusion

Looking at the outcomes of both the survey and the created model it is clear that there is a difference between male and female entrepreneurs. Looking at the stated hypothesis: “there is a difference in trust between male and female entrepreneurs”, it can be concluded from the outcomes that there indeed is a difference in trust between male and female entrepreneurs. For each regression done with Stata, the coefficient for female will result in a negative coefficient compared to male. So it can be stated that the hypothesis is true. The regression on the funded amount and the regression on the goal amount, showed a negative coefficient for female entrepreneurs. This shows that female entrepreneurs attain a lower funded and goal amount, than male entrepreneurs. But this could also mean that female entrepreneurs simply ask for a lower goal amount, and therefore attain a lower coefficient. This could be due to the heavily over-represented women entrepreneurs in a few industry sectors, which are characterized by smaller scale, more intense competition and lower average returns (Klapper and Parker, 2010). But as the regression with percentage reached as dependent variable only gives a negative outcome, it could be concluded that the overall performance of female entrepreneurs is lower than that of men and there indeed is a difference.

This also answers the research question: “is there a difference between the trust in female and male entrepreneurs?” There can be concluded that there is a difference in trust, but it is very hard to exactly point out where this difference comes from. As shown in the results of the survey, the male respondents have an overall preference for male entrepreneurs and are indifferent when looking at the risk of a project. For male respondents gender preference appeared more important as they will always choose for the male entrepreneur, independent of the risk of a project. The outcomes provided by GEM, presented in Appendix 4, also doesn’t give a clear outcome. It shows that the overall percentages of the worldwide female population from 18 up to 64 years in

(21)

age who are entrepreneur is lower than for the same population group of men. But this overview doesn’t include specific results on the performance of these population groups.

The female respondents have an overall preference for female entrepreneurs, and prefer the low risk projects. This risk averseness of female entrepreneurs was also concluded by Croson and Gneezy (2009). They state that the is a clear difference in the affective reaction to risk. Men and women differ in their emotional reaction to uncertain situations and this differential emotional reaction results in difference in risk taking. Only difference between the survey and the findings of Croson and Gneezy (2009), is that managers and professional populations would be an exception to the rule, which doesn’t appear from the survey. Looking at the overall results, risk preferences are more important than gender preferences, for the female respondents, as they will always choose for the low risk project, even if this project is conducted by a male entrepreneur. When the projects have the same risk, either high or low, both male as female respondents will prefer their own gender over the other.

Although there is a clear difference found between male and female entrepreneurs, it is still not evident where this difference exactly comes from. To further conduct this research, more and different variables should be taken into account. By using more variables it could become clear where these differences actually come from. If it is only based upon risk and/or gender preferences, or that the type of project, or the age of the entrepreneur also play a role in investing in a project or not.

(22)

6. References

• Aronson, R. (1991). Self-employment: A Labour Market Perspective. Ithaca, NY: ILR Press.

• Belleflamme, P., T. Lambert and A. Schwienbacher (2013). Crowdfunding: Tapping the right crowd. Journal of Business Venturing, 9 July 2013.

• Bertrand, M. and K. F. Hallock (2000). The Gender Gap in Top Corporate Jobs. National Bureau of Economic Research, October 2000.

• Brush, C.G. and S.Y. Cooper (2012). Female entrepreneurship and economic

development: An international perspective. Entrepreneurship & Regional Development: An International Journal; Special Issue: Female entrepreneurship and economic

development: an international perspective, Volume 24, Issue 1-2, p. 1-6.

• Carter, N., M. Williams, and P. Reynolds (1997). Discontinuance among New Firms in Retail: The Inﬂuence of Initial Resources, Strategy, and Gender. Journal of Business Venturing, Volume 12 (2), p. 125-45.

• Cosh, A., D. Cummings and A. Hughes (2009). Outside Entrepreneurial Capital. The Economic Journal, 119, October 2009, p.1494-1533.

• Croson, R. and U. Gneezy (2009). Gender Differences in Preferences. Journal of Economic Literature, Volume 47, No. 2, June 2009, p. 448-474.

• Eagly, A. H. (2007). Female Leadership advantage and disadvantage: resolving the contradictions. Psychology of Women Quarterly, Volume 31, Issue 1, p. 1-12. • Global Entrepreneurship Monitor (2014). Data; Key Indicators

(http://www.gemconsortium.org/key-indicators), consulted on 3 June.

• KickStarter (2014). Stats (https://www.kickstarter.com/help/stats?ref=footer), consulted on 25 June.

• Klapper, L. F. and S. C. Parker (2010). Gender and the Business Environment for New Firm Creation. The World Bank Research Observer, Volume 26, No. 2, August 2011, p. 237-257.

• Kleemann, F., G. Günter Voß and K. Rieder (2008).Un(der)paid Innovators: The

Commercial Utilization of Consumer Work through Crowdsourcing. Science, Technology & Innovation Studies, Volume 4, No. 1, July 2008, p. 5-26.

(23)

• Larralde, B., and A. Schwienbacher (2010). Crowdfunding of Small Entrepreneurial Ventures. Book chapter for “Entrepreneurial Finance” (Ed. D.J. Cumming), forthcoming at Oxford University Press.

• McGowan, P., C L. Redeker, S.Y. Cooper and K. Greenan (2012). Female

entrepreneurship and the management of business and domestic roles: Motivations, expectations and realities. Entrepreneurship & Regional Development: An International Journal; Special Issue: Female entrepreneurship and economic development: an

international perspective, Volume 24, Issue 1-2, p. 53-72.

• Minniti, M. and W.A. Naudé (2010). What Do We Know About The Patterns and Determinants of Female Entrepreneurship Across Countries? European Journal of Development Research, 13 May 2010, p. 1-17.

• Oaxaca, R. (1973). Male-Female Wage Differences in Urban Labor Markets. International Economic Review, Volume 14, October 1973, p. 693-703.

• Sheperd, D.A. and A. Zacharakis (2001). The venture capitalist-entrepreneur relationship: Control, trust and confidence in co-operative behavior. Venture Capital: An International Journal of Entrepreneurial Finance, Volume 3, Issue 2, p. 129-149.

• De Telegraaf (2013). Crowdfunding: de 8 beste sites op een rij – door Bas van Essen (http://www.telegraaf.nl/mijnbedrijf/22163418/__Crowdfunding__de_8_beste_sites_op_e en_rij__.html), 24 December 2013, consulted on 14 June.

• Watson, J. (2002). Comparing the Performance of Male- and Female-controlled

Businesses: Relating Outputs to Inputs. Entrepreneurship Theory and Practice, Volume 26 (3), p. 91-100.

• Watson, J., and S. Robinson (2003). Adjusting for Risk in Comparing the Performances of Male- and Female-controlled SMEs. Journal of Business Venturing, Volume 18 (6), p. 773-88.

• World Economic Forum (2014). The Global Gender Report 2012 & The Global Gender Report 2013 (http://www3.weforum.org/docs/WEF_GenderGap_Report_2012.pdf & http://www3.weforum.org/docs/WEF_GenderGap_Report_2013.pdf), consulted on 3 June.

(24)

Appendix

Appendix 1: Survey

Page: 1

Entrepreneurship Thesis

Hi,

I'm writing my bachelorthesis about entrepreneurship. The specific topic is the

difference between the trust in female and male entrepreneurs? This question will be looked at from an investors perspective.

This survey will take about 2 minutes. Thanks for participating.

Start Page: 2

Entrepreneurship Thesis

1. Age: * 2. Gender: Male Female 3. 24

(25)

Occupation:

4.

Gender Preferences:

Would you rather invest in a male or female entrepreneur? Male

Female

5.

Risk vs Return:

Two entrepreneurs are going to start a project. After one year the world can be either in state A or state B. Entrepreneur 1 starts a project with high risk. Entrepreneur 2 starts a project with low risk. If the world is in state A after one year, the project of entrepreneur 1 will result in a high return (30%); but is the world is in state B the project return will be low (2%). The return of the project of

entrepreneur 2 will be the same in either state A or B (12%). Which project do you prefer?

The high risk project The low risk project

6.

Different profiles:

Now a number of different entrepreneur profiles will be given, and you’ll be asked to choose one of them. (The returns for the high and low risk projects are the same as the previous question; high-risk: 2% or 30% depending on the state after one year; low-risk: 12%)

Which project do you prefer?

A high risk project of a female entrepreneur A low risk project of a male entrepreneur

7.

(26)

A high risk project of a male entrepreneur A low risk project of a female entrepreneur

8.

A high risk project of a male entrepreneur A high risk project of a female entrepreneur

9.

A low risk project of a male entrepreneur A low risk project of a female entrepreneur

Submit

Page: 3

Thank you for participating in my survey.

Thanks to your contribution I'm one step closer to graduating.

(27)

Appendix 2: Tables representing the Survey Results

Male Female Total

Respondents 59 (58.42%) 42 (41.58%) 101 (100%) Average Age 29,84746 27 28,66336634 Occupation Student 24 (23.76%) 25 (24.75%) 49 (48.51%) Entrepreneur 13 (12.87%) 1 (0.99%) 14 (13.86%) Investor 7 (6.93%) 9 (8.91%) 16 (15.84%) Other 15 (14.85%) 7 (6.93%) 22 (21.78%) Total 59 (58.42%) 42 (41.58%) 101 (100%) Table 1: General Statistics Survey

Gender Preferences Male Female No preference Total

Male 42 (41.58%) 9 (8.91%) 8 (7.92%) 59 (58.42%)

Female 15 (14.85%) 22 (21.78%) 5 (4.95%) 42 (41.58%) Total 57 (56.44%) 31 (30.69%) 13 (12.87%) 101 (100%) Pearson’s Chi-2 = 16.5406 Pr = 0.000; Fisher’s exact = 0.000

Risk Preferences High Low No preference Total

Male 27 (26.73%) 26 (25.74%) 6 (5.94%) 59 (58.42%)

Female 15 (14.85%) 24 (23.76%) 3 (2.97%) 42 (41.58%) Total 42 (41.58%) 50 (49.50%) 9 (8.91%) 101 (100%) Pearson’s Chi-2 = 1.6952 Pr = 0.428; Fisher’s exact = 0.433

Project Preferences High Risk Female Low Risk Male No preference Total

Male 21 (20.79%) 38 (37.62%) 0 (0.00%) 59 (58.42%) Female 15 (14.85%) 25 (24.75%) 2 (1.98%) 42 (41.58%) Total 36 (35.64%) 63 (62.38%) 2 (1.98%) 101 (100%) Pearson’s Chi-2 = 2.9034 Pr = 0.234; Fisher’s exact = 0.308

Project Preferences High Risk Male Low Risk Female No preference Total

(28)

Project Preferences High Risk Male High Risk Female No preference Total

Project Preferences Low Risk Male Low Risk Female No preference Total

Table 2: Overall Survey Results

Male 17 (34.69%) 5 (10.20%) 2 (4.08%) 24 (48.98%)

Female 11 (22.45%) 12 (24.49%) 2 (4.08%) 25 (51.02%) Total 28 (57.14%) 17 (34.69%) 4 (8.16%) 49 (100.00%)

Male 10 (20.41%) 13 (26.53%) 1 (2.04%) 24 (48.98%)

Female 11 (22.45%) 14 (28.57%) 0 (0.00%) 25 (51.02%) Total 21 (42.86%) 27 (55.10%) 1 (2.04%) 49 (100.00%)

Male 9 (18.37%) 15 (30.61%) 0 (0.00%) 24 (48.98%)

Female 7 (14.29%) 16 (32.65%) 2 (4.08%) 25 (51.02%) Total 16 (32.65%) 31 (63.27%) 2 (4.08%) 49 (100.00%)

Male 12 (24.49%) 11 (22.45%) 1 (2.04%) 24 (48.98%) Female 9 (18.37%) 15 (30.61%) 1 (2.04%) 25 (51.02%) Total 21 (42.86%) 26 (53.06%) 2 (4.08%) 49 (100.00%)

Male 13 (26.53%) 8 (16.33%) 3 (6.12%) 24 (48.98%)

Female 15 (30.61%) 8 (16.33%) 2 (4.08%) 25 (51.02%) Total 28 (57.14%) 16 (32.65%) 5 (10.20%) 49 (100.00%)

(29)

Male 13 (26.53%) 9 (18.37%) 2 (4.08%) 24 (48.98%)

Female 8 (16.33%) 14 (28.57%) 3 (6.12%) 25 (51.02%) Total 21 (42.86%) 23 (46.94%) 5 (10.20%) 49 (100.00%) Table 3: Survey Results Student

Male 9 (64.29%) 3 (21.43%) 1 (7.14%) 13 (92.86%)

Female 0 (0.00%) 1 (7.14%) 0 (0.00%) 1 (7.14%)

Total 9 (64.29%) 4 (28.57%) 1 (7.14%) 14 (100.00%)

Male 7 (50.00%) 3 (21.43%) 3 (21.43%) 13 (92.86%)

Female 1 (7.14%) 0 (0.00%) 0 (0.00%) 1 (7.14%)

Total 8 (57.14%) 3 (21.43%) 3 (21.43%) 14 (100.00%)

Male 5 (35.71%) 8 (57.14%) 0 (0.00%) 13 (92.86%)

Female 1 (7.14%) 0 (0.00%) 0 (0.00%) 1 (7.14%)

Total 6 (42.86%) 8 (57.14%) 0 (0.00%) 14 (100.00%)

Male 10 (71.43%) 3 (21.43%) 0 (0.00%) 13 (92.86%)

Female 1 (7.14%) 0 (0.00%) 0 (0.00%) 1 (7.14%)

Total 11 (78.57%) 3 (21.43%) 0 (0.00%) 14 (100.00%)

Male 9 (64.29%) 3 (21.43%) 1 (7.14%) 13 (92.86%)

Female 0 (0.00%) 1 (7.14%) 0 (0.00%) 1 (7.14%)

Total 9 (64.29%) 4 (28.57%) 1 (7.14%) 14 (100.00%)

Male 6 (42.86%) 6 (42.86%) 1 (7.14%) 13 (92.86%)

Female 0 (0.00%) 1 (7.14%) 0 (0.00%) 1 (7.14%)

Total 6 (42.86%) 7 (50.00%) 1 (7.14%) 14 (100.00%)

Table 4: Survey Results Entrepreneur

(30)

Male 3 (18.75%) 0 (0.00%) 4 (25.00%) 7 (43.75%)

Female 0 (0.00%) 6 (37.50%) 3 (18.75%) 9 (56.25%)

Total 3 (18.75%) 6 (37.50%) 7 (43.75%) 16 (100.00%)

Male 4 (25.00%) 2 (12.50%) 1 (6.25%) 7 (43.75%)

Female 3 (18.75%) 3 (18.75%) 3 (18.75%) 9 (56.25%) Total 7 (43.75%) 5 (31.25%) 4 (25.00%) 16 (100.00%)

Male 4 (25.00%) 3 (18.75%) 0 (0.00%) 7 (43.75%)

Female 6 (37.50%) 3 (18.75%) 0 (0.00%) 9 (56.25%)

Total 10 (62.50%) 6 (37.50%) 0 (0.00%) 16 (100.00%)

Male 5 (31.25%) 2 (12.50%) 0 (0.00%) 7 (43.75%)

Female 4 (25.00%) 5 (31.25%) 0 (0.00%) 9 (56.25%)

Total 9 (56.25%) 7 (43.75%) 0 (0.00%) 16 (100.00%)

Male 4 (25.00%) 0 (0.00%) 3 (18.75%) 7 (43.75%)

Female 0 (0.00%) 7 (43.75%) 2 (12.50%) 9 (56.25%)

Total 4 (25.00%) 7 (43.75%) 5 (31.25%) 16 (100.00%)

Male 4 (25.00%) 0 (0.00%) 3 (18.75%) 7 (43.75%)

Female 0 (0.00%) 7 (43.75%) 2 (12.50%) 9 (56.25%)

Total 4 (25.00%) 7 (43.75%) 5 (31.25%) 16 (100.00%) Table 5: Survey Results Investor

(31)

Male 13 (59.09%) 1 (4.55%) 1 (4.55%) 15 (68.18%)

Female 4 (18.18%) 3 (13.64%) 0 (0.00%) 7 (31.82%)

Total 17 (77.27%) 4 (18.18%) 1 (4.55%) 22 (100.00%)

Male 6 (27.27%) 9 (40.91%) 0 (0.00%) 15 (68.18%)

Female 0 (0.00%) 7 (31.82%) 0 (0.00%) 7 (31.82%)

Total 6 (27.27%) 16 (72.73%) 0 (0.00%) 22 (100.00%)

Male 3 (13.64%) 12 (54.55%) 0 (0.00%) 15 (68.18%)

Female 1 (4.55%) 6 (27.27%) 0 (0.00%) 7 (31.82%)

Total 4 (18.18%) 18 (81.82%) 0 (0.00%) 22 (100.00%)

Male 4 (18.18%) 11 (50.00%) 0 (0.00%) 15 (68.18%)

Female 0 (0.00%) 7 (31.82%) 0 (0.00%) 7 (31.82%)

Total 4 (18.18%) 18 (81.82%) 0 (0.00%) 22 (100.00%)

Male 11 (50.00%) 3 (13.64%) 1 (4.55%) 15 (68.18%)

Female 3 (13.64%) 4 (18.18%) 0 (0.00%) 7 (31.82%)

Total 14 (63.64%) 7 (31.82%) 1 (4.55%) 22 (100.00%)

Male 11 (50.00%) 4 (18.18%) 0 (0.00%) 15 (68.18%)

Female 4 (18.18%) 3 (13.64%) 0 (0.00%) 7 (31.82%)

Total 15 (68.18%) 7 (31.82%) 0 (0.00%) 22 (100.00%) Table 6: Survey Results Other

(32)

Appendix 3: Tables representing the Stata Output Reached

Dependent variable: reached

Independent variables: (1) (2) (3) (4) female -0.1024689 (0.1231579) (0.1260857) -0.1085897 -0.0901701 (0.118843) (0.1217497) -0.0854858 Goal -0.000000268 (0.0000011) -0.00000193 (0.0000011) 0.000000211 -(0.00000119) investors 0.00104033 (0.0003105) (0.0003164) 0.0010964 Daystogo -0.0097022 (0.0027697) constant 0.9454359 (0.0668773) (0.0881408) 0.9593068 (0.0845761) 0.8865378 (0.0852838) 0.9432674 R² 0.0049 0.1227 0.1843 N 156 156 156 148

Table 7: Regression outcomes reached

Funded

Dependent variable: Funded

Independent variables: (1) (2) (3) (4) female -23170.36 (10625.66) (9269.567) -9727.025 (8772.296) -8414.178 -7772.733 (8994.14) Goal 0.8161283 (0.0811084) (0.0813683) 0.6977501 (0.0875674) 0.8245615 investors 100.0188 (22.9224) 77.36242 (23.373) Daystogo -661.4069 (204.6088) constant 43016.53 (5559.396) (6479.933) 7124.646 (6242.913) 1938.054 (6300.255) 5786.78 R² 0.0262 0.4205 0.4850 0.5301 N 179 156 156 148

Table 8: Regression outcomes Funded

(33)

Goal

Dependent variable: Goal

Independent variables: (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) female -22876.38 (9023.202) (7156.186) -9014.791 (5550.191) -5579.753 -5412.417 (5513.12) (5572.933) -6395.88 (7163.926) -8935.566 -8944.865 (6724.28) -23014.61 (9071.22) (8526.668) -19638.33 (8431.769) -20421.35 Funded 0.4879435 (0.0484928) (0.0467732) 0.7715214 (0.0495509) 0.741177 (0.049539) 0.6940093 0.4672802 (0.054492) (0.049294) 0.464167 reached -45044.54 (4431.904) -45552.45 (4411.1) (4657.522) -38078.97 (5922.011) -1436.764 (5850.364) -10235.16 (5898.161) -1051.448 investors 26.89292 (15.28188) (14.91995) 32.67681 (19.85327) 16.56723 (18.02528) 29.71338 (22.49519) 106.4379 (21.42883) 107.4406 Daystogo 432.0814 (127.7246) (143.8482) 810.7429 (192.1113) 805.6996 constant 51822.27 (4899.788) (4503.852) 27708.95 56281.75 (4478.8) (4489.147) 55217.47 (5130.811) 44739.01 (4623.053) 26854.84 (4543.404) 16192.91 (7450.035) 53180.64 (7022.613) 49451.16 (7757.65) 31569.79 R² 0.0400 0.4223 0.6561 0.6630 0.6922 0.4250 0.5474 0.0404 0.1636 0.2669 N 156 156 156 156 1 156 148 156 156 148

Table 9: Regression outcomes Goal

Appendix 4: Table representing the GEM percentages of the 18-64 population, who are either a nascent entrepreneur or owner-manager of a new business

2008 2009 2010 2011 2012 2013

Overall 10,49% 10,70% 11,86% 10,86% 13,08% 13,84% Female 8,25% 7,98% 9,79% 8,32% 10,66% 11,70% Male 12,94% 13,14% 13,84% 13,32% 15,51% 15,97% Table 10: Average percentages of GEM outcomes per year

Trusting female entrepreneurs