The effect of the financial crisis on
female entrepreneurial activity
Do they require more support in times of financial crisis
compared to males?
Master Thesis Anne Remmert Student number: 5970504 University of Amsterdam
Faculty of Economics and Business
MSc Business Economics | Organization Economics Supervisor: T. Buser
Abstract: The aim of this paper is to examine the effect of the financial crisis on female entrepreneurship. In order to examine this effect, the focus has been set on early stage female entrepreneurship and data has been obtained from the Global Entrepreneurship Monitor (GEM) dataset. Next to it, a comparison with males has been made to determine whether females were affected differently by the financial crisis than males. For the research of this thesis, different ordinary least squares regressions have been made and several explanatory variables have been included in order to see whether they can explain the effect of the crisis on early stage
entrepreneurial activity. Evidence has been found for the negative impact of the financial crisis on female entrepreneurial activity. Nevertheless, the difference between females and males tend to be not significant.
This paper contributes to existing literature on female entrepreneurship and by examining the effects of the financial crisis on the perceptions of females towards entrepreneurship. Herewith, a better understanding is created for governments in order to encourage and support female entrepreneurship by setting the right policies. Keywords: females, entrepreneurship, financial crisis, female entrepreneurial activity
Table of Content
1. Introduction ... 3
2. Literature Review ... 5
2.1 Entrepreneurship ... 5
2.1.1 Female entrepreneurship ... 6
2.1.2 Necessity and opportunity driven entrepreneurship ... 7
2.2 Gender differences ... 8
2.3 Explanatory variables on the decision to become an entrepreneur ... 10
2.3.1 Fear of Failure ... 10
2.3.2 Knowing an entrepreneur ... 11
2.3.3 Perceived opportunities ... 13
2.3.4 Perceived capabilities ... 14
2.4 Financial crisis ... 14
3. Methodology and Data ... 16
4. Results and Discussion ... 20
4.1 Gender difference statistics ... 20
4.2 Summary statistics ... 20
4.3 Correlations ... 25
4.4 Ordinary Least Squares (OLS) Regressions ... 26
4.3.1 Results of first set of OLS regressions ... 27
4.3.2 Results of second set of OLS regressions ... 29
5. Conclusion ... 32
6. Limitations and Outlook ... 34
References ... 35
1. Introduction
Entrepreneurs play an important role in the economic value of countries. For instance, van Praag and Versloot (2007) examine the role of entrepreneurs in their contribution to job creation, innovation, productivity and economic growth. Most existing literature explains that entrepreneurs contribute to these economic values of countries. Therefore, governments have specific policies for subsidizing and encouraging start-up entrepreneurs to stimulate economic growth and development.
The number of literature based upon entrepeneurship is growing vastly and, interestingly, there is a significant increase of research conducted with respect to female entrepeneurship. Although the number of female entrepreneurs and their contribution to the economy is significantly growing, female entrepreneurs are globally underrepresented (Minniti, 2010; Ascher, 2012). And even though there is a large increase in female entrepreneurship it is understudied and a much clearer understanding of female entrepreneurship is needed.
There are a wide variety of factors that explain the decision of an individual to become an entrepreneur or not. Financial constraints play an important role in the individual’s decision to (not) become an entrepreneur. Financial constraints are, for example, the accessibility of bank loans. In the financial crisis, capital funding from banks had become more difficult for new business start-ups (Block and Sandner, 2009). It would be interesting to see what happens with the perception towards entrepreneurship in the times of the financial crisis and what effect it has on females compared to males. Yet, little has been written about how the financial crisis affected female entrepreneurship. In order to extend the understanding of female entrepreneurship and what variables could explain the effects of the financial crisis on the decision of females to become an entrepreneur, the following research question has been formulated:
What is the effect of the financial crisis on female entrepreneurship? With greater understanding of these effects, governments would be more able to set the right policies in order to encourage and support female entrepreneurship and, thus, increase the economic value of their country.
By conducting this research, a contribution is made to the small number of existing female entrepreneurs. Since female entrepreneurs are significantly contributing to the value of economies, and this number is increasing, this topic is interesting for governments. If the governments would have more insight they would be more able to set policies in order to support female entrepreneurs in times of financial and/or economic crisis. The financial crisis is an interesting time frame since it faces financial and economic difficulties. Governments need to understand what happens with the perceptions of females in times of financial and economic difficulties in order to encourage them to become a successful entrepreneur. With this understanding, they will know how to react when there are financial/economic difficulties in the future. Naturally, a comparison with males is being made in order to see whether there are differences between males and females.
The findings of the research of this thesis suggest that the financial crisis had a negative impact on the decision of females to become involved in early stage entrepreneurial activity. Nevertheless, females were not affected significantly different by means of the financial crisis when compared to males. Explanatory variables have been added to the OLS regression in order to determine whether they can explain the effect of the financial crisis on early stage entrepreneurship. These variables are perceived opportunities, the belief of having the required knowledge, experience and skills, fear of failure and whether they know an entrepreneur or not. These variables and their impact are widely discussed in this thesis.
This thesis is structured as follows. Chapter two contains the literature review including the discussion of the explanatory variables and the financial crisis. In chapter three the focus is on the methodology and data. Chapter four contains the results and in chapter five the conclusion has been presented. The last chapter discusses the limitations and recommendation for further research.
2. Literature Review
2.1 Entrepreneurship
A great number of existing literature states that entrepreneurs play an important role in the economic development and growth of many different countries. Van Praag and Versloot (2007) examined the contributions of entrepreneurs to the economy compared to non-entrepreneurs. In their research, 57 high quality and relevant articles have been studied and analyzed. This study confirmed the many similar results of different studies on the contributions of entrepreneurs to countries’ economies. They stated that entrepreneurs have a greater impact on job creation than non-entrepreneurs. This is also confirmed by the study of Acs and Mueller (2008) who found that start-up companies with 20 to 500 employees have a great employment effect. Acs and Mueller (2008) obtained data from metropolitan statistical areas urbanized areas with at least 50000 citizens in the United States. Next to it, they used microdata of Longitudinal Establishment and Enterprise to observe employment dynamics. They found that individuals who initiated a business with at least 20, but no more than 500 employees, created the greatest employment effect. Acs and Mueller (2008) used data of 320 metropolitan statistical areas of the total of 370 in the United States. This means they cover not all the regions in the United States which could lead to biased results. In addition, the effects in different regions are not controlled for time variant factors. It is, for example, important to control for time variant factors because it could be that different economic shocks can cause changes in the employment dynamics in some areas.
Entrepreneurs also tend to contribute to the economic value of a country because of high productivity growth rates. Robbins et al. (2000) investigated the value added on economic activity at the state level of the United States. 48 states in the United States have been observed over ten years, from 1986 until 1995. Their results show that states with a higher percentage of very small businesses (less than 20 employees) significantly contribute more to higher levels of GDP growth in comparison to states with lower percentages of very small businesses. This can be linked to early stage entrepreneurs, since they most often start off a business with no more than 20 employees. Furthermore, entrepreneurs contribute to the economic value of countries by their high quality of innovation (van Praag and Versloot, 2007).
Overall, it can be stated that entrepreneurs significantly contribute to the economic development of countries.
This thesis focuses on female entrepreneurs and will include a comparison with male entrepreneurs in order to see the differences between both and whether females were affected differently by the financial crisis than males. In addition, the focus is on early stage entrepreneurs because this thesis investigates what effect the financial crisis has on an individual’s decision to become an entrepreneur. Therefore, the definition that will be used in this thesis is as follows: An early stage entrepreneur is someone who is either a nascent entrepreneur or owner/manager of a new business. This definition is made by the Global Entrepreneurship Monitor dataset, which is discussed extensively in chapter three. A nascent entrepreneur is a person that undertakes action to set up a business that they will own or co-own and did not receive salary or other payments for more than three months. An owner/manager of a new business is owning and managing a business where he or she received salary or other payments from for more than three months but less than 42 months. The entrepreneur is categorized in the age range of 18 until 64.
2.1.1 Female entrepreneurship
There is an exponential increase of literature based upon female entrepreneurship. Researchers (Minniti, 2010; Ascher, 2012; Brush and Cooper; 2012) noted the importance of female entrepreneurship and their contributions to the economic development of countries. Although there is an increase in number of female entrepreneurs, they are still underrepresented and understudied worldwide. An increasing understanding, especially from governments, of the impact of female entrepreneurship is needed as female entrepreneurs are of growing importance for the economic value of countries (Ascher, 2012).
Muravyev et al. (2009) used data from the Business Environment and Enterprise Performance survey of 2005 to examine entrepreneurial gender discrimination with respect to external financial accesses. The results of their basic econometric analysis show that female entrepreneurs face difficulties in obtaining bank loans and pay on average higher interest rates compared to male entrepreneurs. Next to the financial constraints, subjective factors as fear of failure and perceptions towards their own skills have a different impact on females. Most of the subjective factors have the same direction of impact on male and female entrepreneurial activity
(Verheul et al., 2006). Nevertheless, the intensity of this impact differs sometimes. However, this has not been examined yet in times of financial crisis. Females in the early stage of entrepreneurship have higher fear of failure rates and the belief in their own skills and knowledge is lower compared to males (Minniti, 2010). Since most factors have the same direction of impact on male and female entrepreneurs, the following hypothesis is made.
Hypothesis 1: The financial crisis has a negative effect on early stage entrepreneurial activity for both males and females
The gender gap with respect to entrepreneurship is widely explained in existing literature. In section 2 of the literature review of this thesis, the gender differences will be discussed in more detail. Nevertheless, the effects of the financial crisis have not been examined yet. The understanding of the motivations and perceptions of females of becoming an entrepreneur or not is essential for governments in order to set the right policies and encourage female entrepreneurship (Ascher, 2012).
2.1.2 Necessity and opportunity driven entrepreneurship
The dataset that is being used for the research of this thesis is obtained from the Global Entrepreneurship Monitor (GEM). An essential asset of GEM for this study is that it makes a distinction between necessity and opportunity driven entrepreneurs. In turn, this allows access to information about both groups of entrepreneurs. The definition of a necessity driven entrepreneur is the individual’s decision to set up a business because he or she believes there is no other option, whereas an opportunity driven entrepreneur starts a business because he or she sees an opportunity to, for example, increase their income. Therefore, it can be stated that the purpose to start a business is explained by the different motivations of individuals (Block and Sandner, 2009). It is interesting to take these different motivations for starting a business into account as the financial crisis could cause a shift from opportunity driven entrepreneurship towards necessity driven entrepreneurship due to the higher rates of unemployment. This could have an impact on the motivation to become an entrepreneur or not. Which leads to the following hypotheses.
Hypothesis 2a: The number of necessity driven entrepreneurs increases in times of financial crisis.
Hypothesis 2b: The number of opportunity driven entrepreneurs decreases in times of financial crisis.
Verheul et al. (2006) examined the effect of different variables on female and male entrepreneurship across 29 countries. By performing regression analyses on the GEM data they found that high unemployment rates increased the level of females that start a business out of necessity since they have no other means of income. Nevertheless, the impact of unemployment on entrepreneurial activity is smaller for females than for males. However, the reasons for unemployment have not been taken into account. For instance, an individual could be unemployed because he or she was dismissed or because he or she quit voluntarily. An extension of this result has been given by Brush and Cooper (2012). They explained that less developed countries, often paired with high unemployment rates, show a higher number of necessity and opportunity driven female entrepreneurs. Also, Minniti (2010) confirmed that females often started a business out of necessity because they required flexibility in their time and locations. Minniti (2010) used GEM data that covered individuals from 34 different countries. The results were estimated by an equalized bootstrap which is a non-parametric method. Nevertheless, the difference between the number of males and females is bigger when analyzing entrepreneurs who start a business out of an opportunity.
The effect of the financial crisis on female early stage entrepreneurship is being investigated in the research of this thesis. The differences between males and females are examined in order to see whether females were affected differently than males. Explanatory variables have been used to determine whether they can explain the effect of the crisis on early stage entrepreneurship. Additionally, a comparison between the number of necessity and opportunity driven entrepreneurship have been made in order to see whether hypotheses 2a and 2b are correct.
2.2 Gender differences
Even though the subject female entrepreneurship is understudied, several researchers (Minniti, 2010; Estrin and Mickiewicz, 2011; Muravyev et al., 2009) showed a significant difference between male and female entrepreneurs. The causes of the gender differences can be explained by subjective factors, such as perceptions and characteristics, and objective factors, such as education and work status (Minniti, 2010).
According to Minniti (2010), males are more likely to become an entrepreneur than females. Minniti (2010) discussed this gender gap by describing the differences in perceptions. For instance, the perceptions of having the knowledge and skills or the attitude towards fear of failure. In these studies the effect of the financial crisis on these attitudes and behaviors has not been taken into account. Whether these attitudes and behaviors can explain the effect of the financial crisis on early stage entrepreneurial activity has been examined in this thesis. Furthermore, as discussed in section 2.1.1., Muravyev et al. (2009) stated that the gender of entrepreneurs matters for the access to financials. They stated that females are less likely to obtain a bank loan.
Buttner and Rosen (1988) investigated the perceptions of bank loan officers in the United States towards successful entrepreneurs. The nine personality characteristics that are associated to become a successful entrepreneur are autonomy, leadership, readiness for change, the ability of taking risk, endurance, low need for support, lack of emotionalism, low conformity and persuasiveness. In this study they examined the differences of these characteristics between males and females. The 106 questionnaires that were returned showed a significant negative perception of a bank loan officer towards female entrepreneurs when compared to male entrepreneurs. The first six personality characteristics mentioned above were statistically seen significantly lower for females compared to males. This implies that bank loan officers believe that the personality characteristics needed to become a successful entrepreneur are much weaker for females than for males. However, they did not take the characteristics of the bank loan officers into account. For instance, if a bank loan officer would be acquainted with a female that started a business and who is running it successfully, it would likely have a positive influence on the perceptions of the bank loan officer.
In the next section the explanatory variables on the decision to become an entrepreneur will be discussed. Together with this, the different effects on males and females will be taken into account. According to existing literature about gender differences in entrepreneurship, the following hypothesis has been made.
Hypothesis 3: The financial crisis has a stronger effect on females’ decision to get involved in early stage entrepreneurial activity compared to males.
2.3 Explanatory variables on the decision to become an entrepreneur
As discussed previously, there are different variables that explain the individual’s decision to become an entrepreneur. The following subsections discuss the variables that are being used for the research of this thesis. These variables include the individual’s motivation and characteristics.
2.3.1 Fear of Failure
Fear of failure can be linked to the fear to lose or risk aversion. An individual is more risk averse when he or she is facing a high fear of failure rate and is therefore less willing to take risk (Lerner and Keltner, 2001). It is widely known that females are on average more risk averse than males. For example, Jianakoplos and Bernasek (1998) state that females are significantly more risk averse than males in financial decision making; especially when they are single. The tobit regression results, based on the data retrieved from 3143 households of the United States in 1989, show that this risk aversion increases by the number of children and decreases when personal wealth increases. In addition, Charness and Gneezy (2012) analyzed 15 sets of experiments from different researchers in order to determine whether there are also differences between males and females in financial risk taking. These experiments were all based on similar designs of an investment game and were therefore comparable. The basis of the investment game is that individuals receive money and can choose to invest this money in a risky asset. Interestingly, 14 of the experiments showed the same results: which is that females invest less than males in risky assets. With respect to financial decision making, Charness and Gneezy (2012) concluded that females are more risk averse than males. Nevertheless, most participants were students. It would be interesting to see the differences in risk aversion between males and females that are involved in entrepreneurial activity or have the intention to start a business. Another study concerning differences between males and females with respect to risk taking is that of Levin et al. (1988). In their experiment, 110 students were evaluated during gambling games. Again, the students could invest five or ten dollars to win different numbers of money with different probabilities of winning that certain number. Results showed that females were less willing to invest which emphasized that females are more afraid to lose than males. Nevertheless, no real money was used in their experiment. In order to create a simulation of the reality, incentives such as money are needed to see what people would really do.
Arenius and Minniti (2005) investigated the effect of fear of failure on entrepreneurs. For their research, the GEM data of the year 2002 was used. Their analysis of the impact covered 28 countries, which resulted in 51.721 individuals for the sample. Individuals were asked if their fear to fail would prevent them from starting a business. The binominal logistic regression showed a significant negative effect of the fear of failure on being a nascent entrepreneur. Also Minniti (2010) showed the negative impact of fear of failure on the individual’s decision to become an entrepreneur. Furthermore, females appeared to have higher fear of failure rates compared to males. This could possibly be explained by the different sectors in which female entrepreneurs start their business. These sectors can include more competition making it harder to survive. As discussed previously, the research of Levin et al. (1988) showed that females were more afraid to lose than males. This fear of losing could still play a role when females are considering starting a business. It would be interesting to see whether fear of failure could explain the effect of the financial crisis on early stage entrepreneurship and whether it explains a difference between males and females.
Hypothesis 1 states that the financial crisis has a negative effect on the total early stage entrepreneurial activity. In times of financial crisis there is more uncertainty. Therefore, it is expected that the fear of failure partly explains the effect of the financial crisis on early stage entrepreneurship. According to the existing literature, females in the early stage of entrepreneurship have higher fear of failure rates compared to males (Minniti, 2010). This leads to the following hypotheses. Hypothesis 4a: Fear of failure has a significant negative effect on female early stage entrepreneurial activity.
Hypothesis 4b: Fear of failure partly explains the effect of the financial crisis on female early stage entrepreneurial activity.
2.3.2 Knowing an entrepreneur
Knowing an entrepreneur has a positive significant influence on becoming an entrepreneur (Arenius and Minniti, 2005). Information can be shared and successful entrepreneurs could stimulate and motivate individuals to enter entrepreneurial activities. Analyses of the GEM data for 34 countries showed that on average females knew fewer entrepreneurs than males (Minniti, 2010). Ascher (2012) examined the impact of female entrepreneurship by exploring several different studies and
comparing the results. In his study, the fact that females know fewer entrepreneurs is explained by the smaller social network, and therefore human capital, of females. Females appear to lack some knowledge with regards to setting up a successful business. Another study with respect to the social capital of female entrepreneurs is that of Carter et al. (2003). In this study, data was obtained from a telephone questionnaire of 235 females from the United States that owned a business. The data was used to examine the impact of social and human capital on the use and access of financial capital. In order to measure the social capital of females, network diversity and network tie strength were scrutinized. The logistic regression analysis showed that social capital did not have a significant impact on the way they made use of external financials such as, for example, bank loans or equity financing. The network diversity and network tie strength had solely a significant impact on the use of personal sources in order to finance their business. Nevertheless, this research merely points out the importance of social capital for female entrepreneurs, but does not include the effect of social capital on the decision to become an entrepreneur or not. Davidsson and Honig (2003) investigated this impact. They stated that social capital had a strong positive impact on the decision to become an entrepreneur. This social capital refers to knowing family or friends that are running a business. This result was based on a randomly selected sample from Sweden with 380 early stage entrepreneurs and 608 individuals in the control group. Results from the logistic regression showed that parents involved in entrepreneurship, and friends and neighbours as well, positively influenced an individual’s decision in order to become an entrepreneur. However, gender differences were not taken into account.
In general, social capital tends to have an impact on the decision to become an entrepreneur. Social capital in this thesis is described as knowing an entrepreneur. Since the number of entrepreneurs in times of the financial crisis has decreased (Klapper and Love, 2011), it is expected that the number of individuals that know an entrepreneur is decreasing as well. To conclude, the following hypotheses have been made.
Hypothesis 5a: Knowing an entrepreneur has a significant positive effect on female early stage entrepreneurial activity.
Hypothesis 5b: Knowing an entrepreneur partially explains the effect of the financial crisis on female early stage entrepreneurial activity.
2.3.3 Perceived opportunities
The perceived opportunities are measured by the GEM survey. The individuals were asked if they believe that the area where they live offers good opportunities for starting a business in the next six months. Ardichvili et al. (2003) described this belief as the entrepreneurial opportunity identification. In order to become a successful entrepreneur, it is important to be able to identify the right opportunities. These opportunities or entrepreneurial identification are partly based on the knowledge, social network and personality traits of individuals (Ardichvili et al., 2003). The social network or knowing an entrepreneur was discussed in the previous section of this thesis. The knowledge needed in order to be able to identify entrepreneurial opportunities will be discussed in the next section. An example of a personality trait is optimism and self-efficacy. During the financial crisis, females faced difficulties with, for instance, obtaining bank loans which, in turn, could have a negative influence on the optimism of females. Therefore, it is expected that the perceived opportunities partially explain the effect of the financial crisis on the decision to become an entrepreneur. The following hypotheses have been made with respect to perceived opportunities.
Hypothesis 6a: Perceived opportunities have a significant positive effect on female early stage entrepreneurial activity.
Hypothesis 6b: Perceived opportunities partly explain the effect of the financial crisis on female early stage entrepreneurial activity.
Governments also have a significant influence on the perceived opportunities of individuals. Estrin and Mickiewicz (2011) examined the impact of institutions on the individuals’ decision to start a business, especially on females. Data was obtained from the GEM and covered information from 55 countries over the years 2001 until 2006. Next to it, they used data obtained from the following sources: the World Bank, the Economist Intelligence Unit, Polity IV and the Heritage Foundation. For their research they made use of probit models and showed that an increase in the size of the state caused a decrease in early stage entrepreneurs. This negative effect had a greater influence on females compared to males. Next to it, they determined that policies of governments are fundamental in order to perceive opportunities in entrepreneurship. For instance, high taxes reduce the incentives for individuals that would have liked to start a business out of an opportunity. Because the influence of governments is
important, a clear understanding of the perceptions of females towards entrepreneurship is highly sought after.
2.3.4 Perceived capabilities
The belief of having the knowledge, skills and experience to set up a business is called perceived capabilities. These three factors belong to an individual’s human capital and all positively contribute to the probability and process of starting a business. As discussed previously, Davidsson and Honig (2003) investigated the effect of human and social capital on being a nascent entrepreneur by making use of a logistic regression. One component of human capital is having the knowledge. This is namely based on the type or level of education. According to Davidsson and Honig (2003), an extra year of education had a positive, but weak impact on the probability of being a nascent entrepreneur. For instance, by having the experience, it was much easier for individuals to obtain a bank loan. The experience positively contributed to the perceptions of bank loan officers towards successful entrepreneurs (Buttner and Rosen, 1988). Furthermore, the experience had the strongest impact on being a nascent entrepreneur when the experience was gained from previous experiences in starting up a company (Davidsson and Honig, 2003).
Several researchers have studied the relationship between human capital and entrepreneurship. Overall, it can be stated that individuals with a significant number or level of human capital are more capable of seeing the opportunities for starting a business. The following hypotheses have been made.
Hypothesis 7a: Perceived capabilities have a significant positive effect on female early stage entrepreneurial activity.
Hypothesis 7b: Perceived capabilities partly explain the effect of the financial crisis on female early stage entrepreneurial activity.
2.4 Financial crisis
The financial crisis started in 2007 in the United States because of the fall of the subprime market. This crisis became global in mid-2008 and is now known as the global economic and financial crisis (Acharya et al., 2009; Dolenc et al., 2012). Due to the crisis, the unemployment rate increased. At the end of 2007 the unemployment rate in the United States was 5%. In almost two years this rate increased to 10.1% (Jagannathan et al., 2013).
As discussed earlier in this thesis, entrepreneurs, especially female entrepreneurs, play an important role for the economies of countries. Nevertheless, the effect of the financial crisis on female entrepreneurship has not been examined yet.
Parker and van Praag (2006) examined the performance of small business ventures once they were started. They state that, on average, capital constraints had a negative effect on the performance of entrepreneurs. Nevertheless, they did not test the effect on males or females separately and the effects during the financial crisis. According to Block and Sandner (2009) the venture capital market faced financial constraints during the financial crisis. Finding investors, which are often banks, became harder because these investors were negatively affected by the crisis and reduced their investments. The venture capital market is important for early stage entrepreneurs because it is an important source for funding start-up companies. However, during the financial crisis the number of venture capital funds had decreased.
There is a gap in the existing literature about the effect of the financial crisis on entrepreneurship, especially female entrepreneurship. This thesis provides a clear understanding of the changes in perceptions of females towards entrepreneurship.
As discussed before, Verheul et al. (2006) stated that there is a difference in the magnitude of the effects of the perceptual variables between males and females. Females tend to be more sensitive towards these variables. Additionally, in times of financial crisis there is more uncertainty. These aspects have been taken into account by formulating hypotheses 1 and 3.
3. Methodology and Data
In this section the dataset that has been used for the research of this thesis is discussed. Next to it, the regression model is explained and the hypotheses are formulated.
Data is obtained from the GEM dataset. GEM is the largest dataset of entrepreneurial dynamics and contains data on activity, attitudes and aspirations of individuals towards entrepreneurship. These data are obtained from surveys. The Adult Population Survey data, which is used for the research of this thesis, is based on individual level and is obtained from surveys via random sampling. Individuals have been asked to fill out the survey only once. The surveys which were not fully completed have been removed. The total sample consists of 9065 individuals, of which 3280 individuals were surveyed before the financial crisis and the remainder, 5785 individuals, during the financial crisis (table 1). The age of the individuals is between 18 and 64 years old. This sample is used when comparing females with males. Since the research of this thesis focuses on females, the sample size has been divided. Therefore, the sample size for females consists of 4508 females, where 1537 females have been measured before the financial crisis and 2971 females during the financial crisis.Because the results for females are compared with those of males, the sample size of males is also reported. This sample consists of 4557 males, where 1743 observations of males are before the financial crisis and 2814 during the financial crisis.
Table 1: Overview of the different sample sizes that have been used for the research of this thesis.
Sample Before crisis During crisis Total
Total sample size 3280 5785 9065
Sample size females 1537 2971 4508
Sample size males 1743 2814 4557
For this study, the GEM data from the United States will be used. Other variables that could have an influence on the results, such as the economy of countries or different government policies, are therefore excluded. For the financial crisis dataset, data of the United States will be collected for the years 2008, 2009 and 2010. In order to determine the effects of the financial crisis, the results are compared with
the years 2004, 2005 and 2006 as the financial crisis started in 2007. Since the financial crisis started in the United States during 2007, this year will not be taken into account to exclude biased results. Since the individuals in this total sample have not participated more than once in this survey, this dataset can not be seen as panel data.
In order to determine the effect of the financial crisis on female early stage entrepreneurial activity, two sets of Ordinary Least Squares regressions (OLS) have been made. The first set of regressions show the effect of the financial crisis on female early stage entrepreneurial activity. Explanatory variables have been used to determine whether these variables can explain the effect of the crisis on female early stage entrepreneurial activity. The second set of regression shows the differences between males and females before and during the financial crisis in order to see whether females were affected by the crisis differently than males. As discussed in the literature review, the explanatory variables that will be used are fear of failure, knowing an entrepreneur, perceived capabilities and perceived opportunities. Additionally, a comparison will be made between opportunity driven entrepreneurs; those who see an opportunity in the market, and necessity driven entrepreneurs; those who have no other option for employment. It is expected that during financial crisis the decision to become an entrepreneur is greater due to necessity factors, since more people lose their jobs and have no other option.
In order to examine the real effect of the female perception towards entrepreneurship, the focus is on early stage entrepreneurial activity. This includes nascent entrepreneurs and owners of new businesses. Therefore, the dependent variable for the OLS regressions is total early stage entrepreneurial activity (TEA).
In the literature review, the hypotheses for the different variables which are applied to the research of this thesis have been discussed. For a clear and total overview, those hypotheses are:
Hypothesis 1: The financial crisis has a negative effect on early stage entrepreneurial activity for both males and females.
Hypothesis 2a: The number of necessity driven entrepreneurs increases in times of financial crisis.
Hypothesis 2b: The number of opportunity driven entrepreneurs decreases in times of financial crisis.
Hypothesis 3: The financial crisis has a stronger effect on a female’s decision to become involved in early stage entrepreneurial activity compared to a male’s decision.
Hypothesis 4a: Fear of failure has a significant negative effect on female early stage entrepreneurial activity.
Hypothesis 4b: Fear of failure partly explains the effect of the financial crisis on female early stage entrepreneurial activity.
Hypothesis 5a: Knowing an entrepreneur has a significant positive effect on female early stage entrepreneurial activity.
Hypothesis 5b: Knowing an entrepreneur partly explains the effect of the financial crisis on female early stage entrepreneurial activity.
Hypothesis 6a: Perceived opportunities have a significant positive effect on female early stage entrepreneurial activity.
Hypothesis 6b: Perceived opportunities partly explain the effect of the financial crisis on female early stage entrepreneurial activity.
Hypothesis 7a: Perceived capabilities have a significant positive effect on female early stage entrepreneurial activity.
Hypothesis 7b: Perceived capabilities partly explain the effect of the financial crisis on female early stage entrepreneurial activity.
For a clear understanding of the dependent and independent variables, a summary of the variables and their description are shown in table 2. These descriptions are available on the website of the Global Monitor Entrepreneurship.
Table 2: Summary of the descriptions of the variables that have been used for the research of this thesis.
Variable Description
TEA Total early stage entrepreneurial activity: Dummy variable is 1 if
individual is involved in early stage entrepreneurial activity
TEAopp Opportunity driven early stage entrepreneurial activity: Dummy
variable is 1 if individual is involved in opportunity early stage entrepreneurial activity
TEAnec Necessity driven early stage entrepreneurial activity: Dummy
variable is 1 if individual is involved in necessity early stage entrepreneurial activity
Knowent Knowing an entrepreneur: Dummy variable is 1 if individual
knows someone personally who started a business in the past 2 years
Opport Perceived opportunities: Dummy variable is 1 if individual
believes that there will be good opportunities in the next 6 months to start a business in the area where they live
Cap Perceived capabilities: Dummy variable is 1 if individual believes he or she has the required knowledge, skills and experience to start a new business
Fearfail Fear of failure: Dummy variable is 1 if fear of failure would
prevent him or her from starting a business
Gender Gender: Dummy variable is 1 if individual is a female and 0 if
individual is a male
Age The age of the individuals of the sample is between 18 and 64 years
4. Results and Discussion
4.1 Gender difference statistics
Upon comparing the early stage entrepreneurial activity for both males and females before and during the crisis, a large difference between males and females can be seen (figure 1). An average decrease in early stage entrepreneurial activity for both males and females is shown during the timeline. Remarkably, the downward slope of the percentage of males that are involved in early stage entrepreneurial activity during the financial crisis is greater when compared to females. The last part is not in line with the expectation that females are more affected by the financial crisis with respect to the decision to become an entrepreneur. In order to determine whether hypothesis 1 can be confirmed or rejected, OLS regressions have been made and are discussed in section 4.4. Of the total sample of 9065 individuals, 1229 individuals (13.65%) are involved in early stage entrepreneurial activity (table B of the appendix).
Figure 1: The percentage of males and females who are involved in early stage entrepreneurial activity before and during the crisis.
4.2 Summary statistics
The statistics summary contains the number of observations, the mean, standard deviation and the minimum and maximum values that the variables can have. Since all variables are dummy variables, these values can only be 0 or 1. The summary
0% 5% 10% 15% 20% 25% 30% 35% Before crisis 2004 Before crisis 2005 Before crisis 2006 During crisis 2008 During crisis 2009 During crisis 2010 P er ce n ta ge
Percentage individuals involved in early stage entrepreneurial activity
statistics of the total sample of male and female early stage entrepreneurs are shown in appendix table A. Since this data contains observations from 2004 until 2010, the effect of the financial crisis is not visible. Therefore, a distinction between the time before the financial crisis and the time during the financial crisis has been made. Upon comparing the mean of individuals that are involved in TEA before the financial crisis with the TEA mean during the financial crisis, a decline is observed (tables 3a and 3b). In this comparison, the number of males and females are almost equally divided (Appendix table C). It can be concluded that during the financial crisis a smaller percentage of individuals are early stage entrepreneurs. Nevertheless, it cannot be determined yet whether this decline is justified by the financial crisis. In order to test the effect of the financial crisis on TEA, some regressions have been calculated, which are discussed in section 4.4. Furthermore, as expected, there is a decline of opportunity driven entrepreneurs and a slight increase of necessity driven entrepreneurs in times of the financial crisis compared to the times before the financial crisis.
Table 3a: Summary statistics of the variables of the total sample size: before the financial crisis (2004-2006).
Variable Observations Mean Std deviation Minimum Maximum
TEA 3280 0.167378 0.37337 0 1 TEAopp 3280 0.1405488 0.3476085 0 1 TEAnec 3280 0.0204268 0.1414768 0 1 Knowent 3280 0.4152439 0.4928392 0 1 Opport 3280 0.345122 0.4754805 0 1 Cap 3280 0.5960366 0.4907651 0 1 Fearfail 3280 0.2131098 0.409567 0 1 Gender 3280 0.4685976 0.499089 0 1 Age 3280 42.78933 12.77264 18 64
Table 3b: Summary statistics of the variables of the total sample size during the financial crisis (2008-2010).
Variable Observations Mean Std deviation Minimum Maximum
TEA 5785 0.1175454 0.3220969 0 1 TEAopp 5785 0.0893691 0.2853004 0 1 TEAnec 5785 0.0235091 0.1515268 0 1 Knowent 5785 0.351599 0.4775108 0 1 Opport 5785 0.3439931 0.4750798 0 1 Cap 5785 0.6240277 0.4844148 0 1 Fearfail 5785 0.3031979 0.4596797 0 1 Gender 5785 0.5135696 0.499859 0 1 Age 5785 45.84978 12.37294 18 64
Since the focus of this research is on females, summary statistics that are based on the sample with only females are made as well. By dividing these data into two tables, effects of the financial crisis become clearer. The results of the summary statistics show a decrease of 29.48% in the mean of TEA for females during the financial crisis (tables 4a and 4b). Thus during the financial crisis, a smaller percentage of females was involved in early stage entrepreneurial activity. Next to it, the standard deviation decreased as well while the number of observations increased. To test if the ratio of females that are involved in early stage entrepreneurial activity has changed significantly, a t-test has been applied (1).
(1) 𝑡 − 𝑡𝑒𝑠𝑡 = !"#!"#$%"! !"#!"#$%& !"#!"#$%"
! !
!"#!"#$%& !
𝑇𝐸𝐴!"#$%" and 𝑇𝐸𝐴!"#$%& are the means of early stage female entrepreneurs before and during the financial crisis, respectively. Var is the variance and N is the total observations of the sample. This formula has as result t = 3.9309 and therefore it can be stated that the difference between the early stage entrepreneurial activity of females before and during the financial crisis is significant at a 1% significance level.
Table 4a: Summary statistics of the variables of the female entrepreneurs sample before the financial crisis (2004-2006).
Variable Observations Mean Std deviation Minimum Maximum
TEA 1537 0.137931 0.3449398 0 1 TEAopp 1537 0.1164606 0.3208809 0 1 TEAnec 1537 0.0143136 0.1188188 0 1 Knowent 1537 0.3715029 0.4833637 0 1 Opport 1537 0.2615485 0.439621 0 1 Cap 1537 0.5068315 0.500116 0 1 Fearfail 1537 0.2277163 0.419495 0 1 Age 1537 43.61158 12.546661 18 64
Table 4b: Summary statistics of the variables of the female entrepreneurs sample during the financial crisis (2008-2010).
Variable Observations Mean Std deviation Minimum Maximum
TEA 2971 0.0972736 0.2963799 0 1 TEAopp 2971 0.0723662 0.2591367 0 1 TEAnec 2971 0.0205318 0.1418345 0 1 Knowent 2971 0.3234601 0.4678754 0 1 Opport 2971 0.3066308 0.4611723 0 1 Cap 2971 0.5432514 0.4982097 0 1 Fearfail 2971 0.3261528 0.4688829 0 1 Age 2971 45.89095 12.1482 18 64
Another result of the summary statistics is a decrease in the mean of females that started a business out of an opportunity during the financial crisis, whereas it shows an increase in the mean of females that started a business out of necessity during the financial crisis (figure 2). Thus the number of necessity driven female early stage entrepreneurs increased, while the number of opportunity driven female early stage entrepreneurs decreased during the financial crisis.
Figure 2: The percentage of opportunity versus necessity driven female entrepreneurship based on the sample of females that are involved in early stage entrepreneurial activity.
As indication, the percentage of opportunity versus necessity drive entrepreneurs for males is shown in graph A of the appendix. Here, it is clearly visible that the expectations are confirmed for the years 2009 and 2010. Nevertheless, the year 2008 shows the opposite of what has been expected. Next to it, the summary statistics for males are provided by tables D and E in the appendix.
In order to see whether the fear of failure of females would prevent them from starting a business, the explanatory variable Fearfail can be evaluated from tables 4a and 4b. Before the financial crisis the mean of the explanatory variable Fearfail is 0.2277163, which implies that 22.77% of the females think that fear of failure would prevent them from starting a business. During the financial crisis this percentage increased as the mean is 0.3261528 resulting in 32.62% of the females thinking that fear of failure would prevent them from starting a business. All variables can be evaluated in the same manner as Fearfail.
Although the focus of this study is on the effect of the financial crisis on female early stage entrepreneurship, it is interesting to see what the differences are between the explanatory variables before and during the financial crisis (figure 3). Remarkable from the graphs below is that explanatory variables change. The variables have been included in the first and second set of regression in order to determine whether they can explain the effect of the financial crisis on TEA.
0% 20% 40% 60% 80% 100% Before crisis 2004 Before crisis 2005 Before crisis 2006 During crisis 2008 During crisis 2009 During crisis 2010 P er ce n ta ge
Opportunity versus Necessity-‐driven female early stage entrepreneurship
Figure 3: Differences between the explanatory variables before and during the financial crisis
In order to determine whether these variables could explain the effect of the financial crisis on early stage female entrepreneurship, regressions have been made and are discussed in section 4.3.
4.3 Correlations
According to the assumptions of the OLS regression, explanatory variables should not correlate with each other at a high level. This is also known as multicollinearity. When multicollinearity occurs, the variables are not consistent and could therefore lead to biased results. Again, the total sample size and the sample size with only females have been examined separately. For both sample sizes it holds that no correlation is that high that it causes multicollinearity (tables 5 and 6). Furthermore, almost all variables are correlated with each other with a significance level of 1% or 5%. Only fear of failure is not significantly correlated with knowing an entrepreneur and perceived opportunities in the sample size of only females. As there is no multicollinearity, the OLS regressions are deemed to reflect appropriate models.
0% 10% 20% 30% 40% 50% 60%
Knowent Opport Cap Fearfail
P er ce n ta ge
Changes in explanatory variables
Table 5: Correlations of explanatory variables of the total sample size.
Knowent Opport Cap Fearfail Gender
Knowent 1.000 Opport 0.2482*** (0.0000) 1.000 Cap 0.2571*** (0.0000) 0.2086*** (0.0000) 1.000 Fearfail -0.0267** (0.0111) -0.0553*** (0.000) -0.1081*** (0.000) 1.000 Gender -0.0715*** (0.000) -0.1112*** (0.000) -0.1697*** (0.0000) 0.0492*** (0.000) 1.000
P-values are reported in parentheses. ***) Significant at 1% level **) Significant at 5% level *) Significant at 10% level
Table 6: Correlations of explanatory variables of the female total sample size.
Knowent Opport Cap Fearfail
Knowent 1.000 Opport 0.2162*** (0.0000) 1.000 Cap 0.2400*** (0.0000) 0.1810*** (0.0000) 1.000 Fearfail 0.0049 (0.7427) -0.0131 (0.3806) -0.0901*** (0.0000) 1.000
P-values are reported in parentheses. ***) Significant at 1% level **) Significant at 5% level *) Significant at 10% level
4.4 Ordinary Least Squares (OLS) Regressions
The effect of the financial crisis on the decision of females to become involved in early stage entrepreneurial activity has been examined by the first set of OLS regressions. The second set of OLS regressions displays the differences between females and males prior to and during the crisis. First of all, the results of the first set of OLS regressions have been discussed. This is followed by the results of the second set of the OLS regressions.
4.3.1 Results of first set of OLS regressions
The basis regression equation for the first set of OLS regression is as follows. 2 𝐹𝑇𝐸𝐴 =β0 + β1*Crisis
Where FTEA is the number of females that are involved in early stage entrepreneurial activity, β0 is the intercept and β1 is the coefficient of the dummy variable Crisis. The
dummy variable Crisis is 0 for the years 2004, 2005 and 2006. For the years 2008, 2009 and 2010 the dummy variable Crisis is 1. In order to observe whether the different explanatory variable explain the effect of the financial crisis on FTEA, the variables have been included in the regression one by one. As a final point, all variables are added at the same time in the regression. The equations are as follows:
3 𝐹𝑇𝐸𝐴 =β0 + β1*Crisis + β2*Opport
(4) 𝐹𝑇𝐸𝐴 =β0 + β1*Crisis + β3*Knowent
(5) 𝐹𝑇𝐸𝐴 =β0 + β1*Crisis + β4*Fearfail
(6) 𝐹𝑇𝐸𝐴 =β0 + β1*Crisis + β5*Cap
7 𝐹𝑇𝐸𝐴 =β0 + β1*Crisis + β2*Opport + β3*Knowent + β4*Fearfail + β5*Cap
The βs related to each variable show the effect of the explanatory variable on FTEA. It can be concluded that all variables have a significant effect on FTEA (table 7). Therefore, hypotheses 4a to 7a are confirmed. Nevertheless, differences in the explanatory variable do not explain the effect of the financial crisis on FTEA. For instance, the perceived opportunities and perceived capabilities have improved during the crisis (figure 3), but regression (2) and (5) show an actual increase of the effect of the crisis on FTEA instead of a decrease. The same holds for knowing an entrepreneur and fear of failure. Therefore, changes in the explanatory variables cannot explain the effect of the financial crisis on FTEA. Hypotheses 4b to 7b are not confirmed.
Regression (1) conveys that during the financial crisis, females are 4.07 percentage points less likely to be involved in early stage entrepreneurial activity than before the financial crisis. The change of the dummy variable Crisis in these regressions can be explained as follows. When controlling for perceived opportunities, the effect of the crisis on the amount of females that are involved in
early stage entrepreneurial activity increases with almost 16% compared to not controlling for perceived opportunities. Regression (6) controls for all explanatory variables and shows that females are 4.42 percentage points less likely to be involved in early stage entrepreneurial activity. Although regression (4) presents the significant effect of fear of failure on FTEA, when controlling for all explanatory variables Fearfail is not significant any more. Therefore, hypothesis 4a is not confirmed with respect to regression (6). Nevertheless, by solely considering regression (4) it can be stated that fear of failure has a negative significant impact on the female early stage entrepreneurial activity. Therefore, hypothesis 4a is confirmed by this regression.
Table 7: Regressions of female TEA on the Crisis dummy and explanatory variables. The dependent variable is the binary variable FTEA, which equals 1 if the individual is involved in early stage entrepreneurial activity and zero otherwise. Standard errors are reported in parentheses. These regressions are based on the sample size of females. Definitions of the variables are given in table 2.
Variable (1) (2) (3) (4) (5) (6) Crisis -0.0407*** (0.0099) -0.0471*** (0.0097) -0.0341*** (0.0097) -0.0386*** (0.0099) -0.0461*** (0.0096) -0.0442*** (0.0094) Opport 0.1421*** (0.0101) 0.0996*** (0.0101) Knowent 0.1375*** (0.0097) 0.0878*** (0.0098) Fearfail -0.0212** (0.0103) -0.0085 (0.0098) Cap 0.1503*** (0.0091) 0.1131*** (0.0093) Intercept 0.1379*** (0.0080) 0.1008*** (0.0083) 0.0868*** (0.0086) 0.1427*** (0.0083) 0.0617*** (0.0090) 0.0239** (0.0096) N 4508 4508 4508 4508 4508 4508 Adj R2 0.0035 0.0454 0.0462 0.0042 0.0602 0.1028 ***) Significant at 1% level **) Significant at 5% level *) Significant at 10% level
The adjusted R2 gives information about how well the data fits the regression by taking into account the effect of every explanatory variable on FTEA. Comparing regression (1) with regression (6), the coefficient of determination increased which implies that adding all variables increases the strength of the model. Overall, controlling for perceived opportunities, knowing an entrepreneur, fear of failure and perceived capabilities, the crisis significantly decreases the number of females that are involved in early stage entrepreneurial activity by 4.4 percentage points.
4.3.2 Results of second set of OLS regressions
In the second set of the OLS regressions, the differences between males and females have been examined. For these regressions, the total sample size has been used. In order to determine whether females were affected differently by the financial crisis than males, the following basic regression equation has been constructed.
8 TEA =β0 + β1*Crisis + β2*Gender + β3*Crisis*Gender
Where TEA is the number of individuals that are involved in early stage entrepreneurial activity, β0 is the intercept, β1 is the coefficient of the dummy variable
Crisis, β2 is the coefficient of the dummy variable Gender and β3 is the coefficient of
the interaction term of Crisis times Gender. For the following regressions, explanatory variables have been added one by one to determine whether they can explain the effect of the crisis on TEA and the differences between males and females. Additionally, the last regression contains all explanatory variables.
9 𝑇𝐸𝐴 =β0 + β1*Crisis + β2*Gender + β3*Crisis*Gender + β4*Opport
10 𝑇𝐸𝐴 =β0 + β1*Crisis + β2*Gender + β3*Crisis*Gender + β5*Knowent
11 𝑇𝐸𝐴 =β0 + β1*Crisis + β2*Gender + β3*Crisis*Gender + β6*Fearfail
12 𝑇𝐸𝐴 =β0 + β1*Crisis + β2*Gender + β3*Crisis*Gender + β7*Cap
13 𝑇𝐸𝐴 =β0 + β1*Crisis + β2*Gender + β3*Crisis*Gender + β4*Opport +
β5*Knowent + β6*Fearfail + β7*Cap
Once more, the βs related to each variable show the effect of the explanatory variable on TEA. Regression (1) shows that over all periods being involved in early stage entrepreneurial activity is 5.5 percentage points lower for females compared to males (table 8). The effect of the crisis for females is - 4.07 percentage points, whereas the effect of the crisis for males is a decrease in TEA of 5.54 percentage points. Therefore, it can be concluded that the effect of the financial crisis is slightly weaker for females than for males. Overall, hypothesis 1 can be confirmed since these effects are significant. Nevertheless, these differences between males and females is indicated by the β of the interaction term Crisis*Gender and appears to not be significant. Therefore, hypothesis 3 is not confirmed.
Table 8: Regressions of TEA on Crisis dummy variable, Gender dummy variable, an
interaction term and explanatory variables. The dependent variable is the binary variable TEA which equals 1 if the individual is involved in early stage entrepreneurial activity and zero otherwise. Standard errors are reported in parentheses. These regressions are based on the whole sample size. Definitions of the variables are given in table 2.
Variable (1) (2) (3) (4) (5) (6) Crisis -0.0544*** (0.0104) -0.0494*** (0.1019) -0.0433*** (0.0102) -0.0515*** (0.0104) -0.0602*** (0.0101) -0.0469*** (0.0099) Gender -0.0554*** (0.0119) -0.0331*** (0.0117) -0.0428*** (0.0117) -0.0544*** (0.0119) -0.0272** (0.0117) -0.0115 (0.0114) Crisis*Gender 0.0137 (0.0149) 0.0023 (0.0146) 0.0100 (0.0146) 0.0145 (0.0149) 0.0134 (0.0145) 0.0040 (0.0142) Opport 0.1422*** (0.0074) 0.0922*** (0.0075) Knowent 0.1533*** (0.0073) 0.1002*** (0.0074) Fearfail -0.0374*** (0.0081) -0.0163** (0.0077) Cap 0.1679*** (0.0073) 0.1233*** (0.0074) Intercept 0.1933*** (0.0082) 0.1338*** (0.0086) 0.1238*** (0.0086) 0.2008*** (0.0083) 0.0801*** (0.0093) 0.0293*** (0.0097) N 9065 9065 9065 9065 9065 9065 Adj R2 0.0093 0.0476 0.0558 0.0115 0.0645 0.1059 ***) Significant at 1% level **) Significant at 5% level *) Significant at 10% level
The adjusted R2 is 0.0093 and indicates that the data weakly fits the regression. In order to increase the adjusted R2 possible explanatory variables need to be added to the regression. Adding the explanatory variables one by one shows an increase of the adjusted R2. Regressions (2) until (5) show that all explanatory variables have a significant effect on TEA at a 1% level. Thus, having the perceived opportunities has a significant positive effect on TEA of 0.1422 and having the fear of failure significantly decreases TEA by 0.0374.
In regression (6) all explanatory variables are included and show a significant effect on TEA. Since the adjusted R2 in the 6th regression is higher than all other regressions, namely 0.1059, this regression fits the data the best. Therefore, it can be concluded that the effect of the crisis on male early stage entrepreneurial activity is a decrease of 4.69 percentage points, controlling for all explanatory variables. Compared to regression (1) this is a decrease of 13.79%. For female early stage entrepreneurs, controlling for all explanatory variables, the effect of the crisis is a decrease of 4.29 percentage points. Compared to regression (1) this is a decrease of
5.41%. Interestingly, gender differences are explained for a large part by the explanatory variables. Controlling for all explanatory variables reduces the effect of gender on TEA from - 5,5 percentage points to - 1,2 percentage points. Nevertheless, the difference of 0.4 percentage points between males and females, by taking the effect of the crisis into account, is not significant and thus hypothesis 3 cannot be confirmed. In conclusion, it cannot be stated that females were affected significantly different by the financial crisis compared to males.
5. Conclusion
The regression formulations demonstrated the impact of the financial crisis on the decision of females to become involved in early stage entrepreneurial activity. By conducting this research, a contribution has been made to the existing, understudied subject female entrepreneurship. Next to it, the differences between males and females have been examined in order to see whether females have been affected differently by the financial crisis than males.
Based on existing literature and the information provided by the GEM dataset, the focus has been on four explanatory variables which possibly could explain the effect of the financial crisis on the females that are involved in early stage entrepreneurship. The four variables are based on the subjective perceptions and include: perceived opportunities, perceived capabilities, knowing an entrepreneur and fear of failure.
In general, during the financial crisis, the number of total early stage entrepreneurs has decreased. Next to it, for both males and females, the average number of necessity driven entrepreneurs increased while the number of opportunity driven entrepreneurs decreased. Nevertheless, in order to determine whether the financial crisis had a significant effect on the individual’s decision to become involved in early stage entrepreneurial activity, twelve OLS regressions were made.
Results from the first set of six OLS regressions show that all explanatory variables have a significant impact on the female early stage entrepreneurial activity when they are added separately to the basic regression with only the dummy variable Crisis. Nevertheless, the final regression, which includes all explanatory variables, shows that fear of failure does not have a significant impact on female early stage entrepreneurial activity. Overall, controlling for perceived opportunities, knowing an entrepreneur, fear of failure and perceived capabilities, the crisis significantly decreases the number of females whom are involved in early stage entrepreneurial activity by 4.4 percentage points.
The second set of OLS regressions presents the differences between females and males prior to and during the crisis. Again, all explanatory variables appear to have a significant influence on the early stage entrepreneurial activity of males and females when they are added separately to the regression. Additionally, in the last regression, which includes all explanatory variables, the variables still have a
significant impact on early stage entrepreneurial activity. Controlling for all explanatory variables, the effect of the financial crisis on a female’s decision to become involved in early stage entrepreneurship is a decrease of 4.29 percentage points. Compared to males, this is 0.4 percentage points lower. Therefore, the effect of the financial crisis is slightly weaker for female early stage entrepreneurship than for male early stage entrepreneurship. Nevertheless, the regression results convey that this difference is not significant. Therefore, it can be concluded that females have not been affected differently by the financial crisis than males.
Entrepreneurs tend to contribute to the economic growth and development of many countries. The number and importance of female entrepreneurs is growing and therefore a clear understanding of the gender differences with respect to entrepreneurship is needed. With a clear understanding of the impact of the financial crisis, governments are more able to set the right policies in order to support and encourage entrepreneurship in times of economic financial difficulties in the future.
