Master Thesis
The relationship between total fertility rates
and female labor participation rates with
top occupations held by women
By
(Martine) Alie Harmina Smeijers
s2394030 Folkingestraat 49A 9711JV Groningen martine_smeijers@hotmail.com
University of Groningen Faculty of Economics and Business
Thesis supervisor: Drs. A. Visscher
|
Co-assessor: Ms. Dr. M. J. KlasingAbstract
The total fertility rate (TFR) and the female labor participation rate (FLPR) have been analyzed by scholars in manifold. Scientific literature that analyzes the influence of the TFR and the FLPR on specific occupations lacks. This research tests what the relationship is between the TFR and the FLPR on top occupations held by women. This is a cross-country analysis. Three types of occupations are tested based on data from 77 countries; female top managers, women seated in national parliaments and women with ownership in a business. A fourth occupation, female entrepreneurship is tested based on a different sample consisting out of 71 countries. A quantitative method is applied by performing a multiple regression analysis. All hypotheses were expected to test significantly positive. Only the fourth occupation, entrepreneurship, is influenced significantly positive by the TFR and the FLPR. There was no relationship to the percentage of seats held by women in national parliaments. The FLPR is significantly positive related to business ownership and top managers. The TFR is significantly negative related to the percentage of top managers.
Acknowledgements
This master thesis is the final product of my formal academic career. I have walked a long road of formal education. A road that I have appreciated and enjoyed. Attending the University of Groningen was a privilege. The accumulated joy of all my previous diplomas will not compete with the satisfaction and pride of graduating for the MSc International Business & Management.
Writing this thesis contributed to my practical knowledge of statistics and scientific research. The process had the pace of a steam locomotive. The start was slow, but once rolling it is determined to arrive at its final station.
My family and friends have always been of great support and distraction. I’m thankful for both. I would like to thank my fellow student and little cousin Maik Bokhove. From now on I will now try to refer to him as ‘cousin’. Also, special thanks to Jan Hakvoort who was of great help for all kinds of questions at random times. Also, my appreciation to Lieke Noteboom, my academic antipode and best study companion.
Also, I would like to thank Drs. Ad Visscher for his guidance and constructive feedback. The style of guidance and the small talk on actualities are very much appreciated.
In advance I would like to thank my readers. Hopefully you will find this topic just as interesting as I do.
Enjoy reading.
TABLE OF CONTENTS Abstract Acknowledgements Table of contents 1. Introduction………...4 2. Background information………...7
2.1. The total fertility rate……….7
2.2. The female labor participation rate………....9
2.3. National policies………..12
3. Theoretical framework and hypotheses………..13
3.1. The total fertility rate and the female labor participation rate……….13
3.2. Education and higher professional positions………...14
3.3. Hypotheses ‘Women in Top Positions’………...16
3.4. Hypotheses ‘Female Entrepreneurship’………...17
4. Research methodology………19 4.1. Data sources……….19 4.2. Sample………..19 4.3. Outliers……….20 4.4. Variables………..22 4.5. Independent variables………..22 4.6. Dependent variables……….22 4.7. Control variables………..23 4.8. Data analysis………24 5. Descriptive results………...25
5.1. Women in top positions………...25
5.2. Female entrepreneurship………..26
5.3. Multiple modes………27
6. Testing assumptions………28
6.2.
Non-linearity………29
6.3. Homoscedasticity and heteroscedasticity………29
7. Multiple regression results………..32
8. Discussion………...37
9. Conclusion………..40
10. Limitations and future research………...42
References……….43
The Appendices……….53 Appendix 1. Outliers women in top positions
Appendix 2. Outliers entrepreneurship
Appendix 3. Descriptive results per continent ‘Sample data women in top positions’ Appendix 4. Descriptive results per continent ‘Female entrepreneurship’
Appendix 5. Checking assumptions on female top managers Appendix 6. Checking assumptions on female ownership in firms
Appendix 7. Checking assumptions on women seated in national parliaments Appendix 8. Checking assumptions on female entrepreneurship
Appendix 9. Regression tables female top managers Appendix 10. Regression tables female ownership in firms
1. Introduction
The difference between males and females has been introduced by the story of Adam and Eve in the first chapters of the oldest, best known and most disputed books in the world, the Quran and the Bible. Disparities influenced the task division of work from the very beginning. General history books describe the division of tasks thousands of years ago, where men were responsible for hunting and women would take care of the offspring and simultaneously gather foods (Salius, 2014). Task divisions sustained throughout history. It was not until the nineteenth century ere the first wave of feminism came into existence. In the nineteenth and early twentieth century so-called feminism came up. Feminism is a movement identified in many countries. Feminism tries to achieve gender equality, the right for women to vote and equal opportunities in education and employment (Freedman, 2003).
A milestone is the moment that women were permitted the right to vote. It was first introduced in New-Zealand in 1893. Australia followed in 1902 (Australian Government, 2010) and was the first country to include the right for women to be elected for parliament. The US followed in 1919 (US National Archives, 2014). Another milestone were the first women to attend and graduate from universities. A famous example is the Dutch Aletta Jacobs who studied at the University of Groningen and graduated in medicine in 1878 (Atria, 2013).
In the 1960s there was a great movement of feminism that started in the United States. A book ‘The Feminine Mystique’ by Betty Friedan (1963) influenced the thought of many. It triggered women to go beyond the life of taking care of their husband, children and household. During that period activists claimed full gender equality. They opted for equal labor payment and opportunities. Their statement was enhanced by women who started to wear men’s clothing including trousers, which became more than a fashion statement. Also women started to fulfill jobs that were physically extremely challenging.
Brasil, Christine Lagarde the CEO of the IMF or Estée Lauder who was a very successful entrepreneur. These are three women that have gained professional top positions. Be that as it may, women are still a minority among higher positions (White, 2013). Unequal payment is in some cases still present (Galor & Weil, 1996). Also female entrepreneurs spend less hours on their business compared to men (Verheul, Carree & Thurik, 2009). This is mainly due to family engaged obligations. Most gender inequalities among payment or fair opportunities to higher positions will likely be diminished in the future given the current attention to the topic. Equal rights and treatment cannot be completely generalized. Some gender differences will remain unchanged. Child bearing and giving birth is a major unchangeable difference between men and women that will continue to influence life and work. It is a gender specific, nontransferable role. (Schwartz, 1989) It is unlikely to become gender transferable through technological development ‘en masse’ in the near future.
The number of births is important to a population’s existence. It is usually measured by a country’s total fertility rate (TFR). Current global fertility rates have severely declined below a subsistence level of 2.1 children per woman in the last century. The world population will decrease when this continues (The Economist, 2009). Especially a great number of developed countries is not able to meet its subsistence level to secure long-term population stability (Eberstadt, 2010). A declining population influences the size of the labor force. It will present a range of social and economic challenges on the long term (Bloom, Nandakumar, & Bhawalkar, 2002). A fluctuating in the population causes harm to the economy. Imagine the regulation of the number teachers and classrooms.
researched by several scholars (Krishnan, 2009; Powell & Butterfield, 1994). Nonetheless, there is limited information available that relates the TFR to the positions held by women.
Women are attending colleges and universities in increasing numbers. Moreover, women are also performing better within the universities (Lamberts, 2005; Lewin, 2006). Higher education can be a great contribution to women’s type of employment, because it provides better employment opportunities (Cameron, Dowling, & Worswick, 2001). Higher education contributes towards developing a more productive labor force that meets the demands of the global economy (Bardhan, Hicks, & Jaffee, 2013). It can be expected that higher educated women also perform jobs that require higher levels of knowledge and competences. Bloom et al. (2009) found that the years of female education is positively related with labor participation. However, it does not provide any data on the types of jobs performed by women and how it corresponds to fertility. This creates a research gap. This research could contribute towards understanding the influence of the TFR and the FLPR on different higher occupations. Subsequently, the following research question can be presented:
What is the relationship between total fertility rates and female labor participation rates with types of top occupations held by women?
2. Background information
2.1 The total fertility rate
The TFR of a country refers to the number of births per women in a lifetime. It is a factor that indicates potential long-term change in a population. An average of 2.1 children per woman is considered the subsistence level of a stabile population (United Nations, 2005). At the beginning of the previous century population growth was considered a threat to the economy. The Malthusian model (Malthus, 1872) predicts overpopulation with a formula for an exponentially growing population. Overpopulation decreases arable land to provide food. This results in a decrease of a country’s wellbeing. It’s nowadays even possible to import perishable foods. This makes the Malthusian theory no longer applicable to developed countries. However, the theory is still applicable to developing countries including those with the largest populations of the world like India, China, Pakistan and Brazil (World Bank, 2014).
India and the US are both expected to grow by a little over 1% annually (Eberstadt, 2010). Countries like Pakistan and Indonesia are also experiencing an increasing population, in the meantime Central and West African countries are expected to double or triple their population size by 2050 (United Nations, 2012). For years, organizations try to influence the TFR in African countries through provision of sexual education and contraceptives in order to gain control over the population size.
Contrary to the countries mentioned above (Pakistan, Indonesia and African countries), many others are experiencing a decline in their TFR and some face a decreasing population. Many developed economies are facing a population decline in the 21st century (Eberstadt, 2010). Japan is trying to overcome a major shrinkage. (Lee & Lee, 2014) Russia is also experiencing a major population decline and shrunk by nearly 7 million people from 1993-2010. If it was not for immigration, the total decline would have been greater (Eberstadt, 2010).
are worried about their population to decrease or increase. The literature appoints a whole range of possible causes for low TFRs in developed countries and high TFRs in developing countries.
An important milestone in the history of fertility was the introduction of the contraception method ‘the pill’. It was the first method that enabled women to, safely and successfully, control their fertility. The pill was introduced to developed countries such as the U.S. and the U.K. in the 1960’s. However, it was not until the 1970’s before doctors would prescribe it to unmarried women. During the 1960’s, 44% of US households were a stereotypical married couple with children. This declined to 24% in 1999 (Aldrich & Cliff, 2003). Society changed into what perhaps could be described as ‘more liberal’. Women’s control on family planning resulted in strong decreases of national TFRs (Eurostat, 2014). Women have less children and start to have them later in life (Kiernan, 2001). Among the so-called western-women today a child is rather a part of self-realization, as being part of having a good life. It is no longer the prescribed life by conservatives. (Nielsen & Rudberg, 2000) The extreme fertility decreases are applicable to the majority of the developed countries. Exceptions are France, Belgium, the U.S., and Scandinavia. The TFRs in these countries are above the 1.8 children per women. This is considered a significant difference to the common TFR of 1.4 among countries like Germany, Italy and the Czech Republic (Fehr & Ujhelyiova, 2012; Eurostat, 2014). The differences among these developed countries raises the question: what motivates women to have fewer and fewer children (Eberstadt, 2010)?
Another economical explanation that contributes to the current TFRs as it is, is given by Galor & Mountford (2008; 2006). Their theory states that the two factors, trade and population development are interrelated. The theory suggests that international trade amidst developed and developing countries asymmetrically affects the need for human capital. High trade persuades the need for human capital in developed countries. It contributes to a declining TFR. Parents invest in their offspring for the purpose of increasing their professional market value, i.e. human capital. This continues to conserve the contrast with developing countries, who lack a need for human capital. Especially if the offspring continues to contribute their salary to the family.
2.2 The female labor participation rate
Approximately half a country’s population is female. Women rarely make up fifty percent of a nation’s labor participation. From an economic perspective it is considered the underutilization of the total potential labor force. It has also implications for a country’s economic welfare and its economic growth. A greater labor force increases the general development and potential total product of a country (Psacharopoulos & Tzannatos, 1989). The underutilization of a labor force is of direct influence to a country’s income. It can be increased by a greater labor participation rate among women (Lewis, 1954). The potential economic growth triggers the interest for the factors that influence the female labor supply (Cristia, 2008). In 1954 Lewis wrote that “the transfer of women’s work from the household to commercial employment is one of the most notable features of economic development”. Women in developed countries heavily increased their labor participation during the 20th century. Despite this great movement the labor participation rate still lags behind (Aliaga, 2005). The economic benefit of female labor participation only applies to formal employment. Unfortunately, many women are still active in the informal labor market (Blunch, Canagarajah, & Raju, 2001).
outside their home are more likely to maintain social relations, gather information through social exchange, and are more likely to become politically active (Ross, 2008).
Social norms and values do not merely determine the probability that a women works, it also influences the type of work women exert. There are still relatively few women who fulfill the so called ‘blue collar’ jobs (Mammen & Paxson, 2000). Social norms and values do not solely influence which type of jobs women have. Women simply have different job preferences and associations with certain jobs compared to men. Quinlan and Shackelford wrote in 1980 that women were still extremely segregated into clerical, sales and service occupations. This still appears, to a certain extent, to be relevant in developed countries. Quinlan and Shackelford (1980) also observed that women tend to receive lower salaries than men, which is unfortunately still applicable (Harris, Gilbreath, & Sunday, 2002). The level of payment is an important factor. Many women will consider the potential gain when entering the labor market. The payment of formal employment should outweigh the price of staying home. It becomes less attractive to work if wages are relatively low compared to men (Galor & Weil, 1996). This applies especially for married women with a spouse who earns a good salary (Bloom et al., 2009).
Aside of the financial benefits of being employed, women also gain a lot of other benefits. Female employment contributes to the emancipation of women (Engels, 1978). Later Beegle, Frankenberg & Thomas (2001) found that women who earn their own income also have a greater influence within their household. Additionally, Ross (2008) found that female employment leads to a greater source of females who exert political influence.
Countries that greatly value female emancipation also have a strong preference for a high percentage of female labor participation. According to such governments, such as Sweden and the Netherlands, women who are willing to work are clearly a positive matter. Women have a great potential towards contributing to a country’s competitive position. Hence, many countries introduced special national policies. It should create and promote favorable incentives for women to work (Castellano & Rocca, 2014).
in increasing numbers. Perhaps it is needless to say, but there are great differences among the FLPRs in developed and developing economies. According to the theory of Mammen & Paxson (2000) are women living in undeveloped countries most likely to be working. Often on their family farm or in their small store. Fewer women are employed in developing countries. This situation occurs when the economy develops and men face more attractive employment opportunities. As a result the family businesses are often terminated. This is the point where the U-shape will be created. The improved employment opportunities benefit the family and the general wellbeing. The next phase is a growing economy and more men and women start to take part of the educational system at higher levels. These educated women have greater incentives towards labor participation. The opportunity costs of working versus staying home increases. Education could also be regarded as a consumer good. People will spend more on education when they have more to spend (Mammen & Paxson, 2000). Unfortunately, women often remain later in line with receiving education over their male siblings. Education is also an investment and enables people to yield a higher income later on. This higher generated income by male members of the household eventually increases chances of spending it on education. Hence, it might take a generation before it comes to the point that the female members are also send to primary, secondary and tertiary education in large numbers (Goldin, 1995). Poor women are at first less likely to be granted an education and subsequently to be formally employed. Women at the bottom of the social ladder are often lower or uneducated and tend to have more children. Considering their lack of education, it can be projected that their earnings will be also lower. This decreases the chance of being able to fund domestic help or formal childcare services (Abramo & Valenzuela, 2005). Formal childcare services, and to a lesser extent domestic help, are crucial factors that still determine a women’s possibility to work. To poor women it is a lot harder to organize these duties, specifically childcare which is partially due to financial constraints (Abramo & Valenzuela, 2005). At the end of the U in the developed countries, female participation rates are high as well as their average income. (Bloom et al., 2009) For these women the gain of working is notably greater compared to uneducated women.
2.3 National policies
The previous sections indicated that governmental policies can exert influence on the relationship between a country’s TFR and FLPR. Changes in legislation are globally utilized by governments as instruments to steer or change a country’s TFR or FLPR (Klerman, 1999; MacInnes, 2006). Former policies were designed to discourage fertility. It was just a century ago that women around the world had an average of seven children (Ahlburg & DeVita, 1992). Policies are nowadays steering towards a rate above subsistence level, while simultaneously trying to increase the FLPR. The value of policies to steer and balance the TFR and the FLPR are underlined by the Commission of the European Communities (2005). They are supporting legislation that promotes and supports families to combine formal employment and having children. Policies that positively affect women’s income tend to increase the FLPR (Khandker, 1987). According to Milligan (2002) and Gauthier (2007) policies designed to increase the TFRs did prove to be effective in Canada and Europe. The U.S., with a TFR slightly below replacement tries to promote fertility through the Family Medical Leave Act. This is a law that allows women to take maternity leave after birth (Cannonier, 2014). Other political interventions are the female quota. It requires certain companies and governmental
3. Theoretical framework and hypotheses
3.1 The total fertility rate and the female labor participation rate
Obviously child bearing and motherhood are closely connected to female labor participation. This is caused by the fact that raising children is time consuming. Also, women have to temporarily withdraw from the labor market around childbirth. Or as Psacharopoulos & Tzannatos (1989) put it, labor and children demand a great proportion of time simultaneously. There is less time for one factor if more time is spend on the other. Many scholars have researched the influence of the TFRs on the FLPRs and the other way around. They have tried to determine the direction and to find significance.
Women who pursue a career often postpone motherhood. Sometimes it results in fewer children or even no children, due to factors such as infertility caused by age and terminated relationships (Brewster & Rindfuss, 2000). When a child is born it decreases the number of years a women joins the labor market by two years during her reproductive life (Bloom et al., 2009). Having a child below the age of one decreases the probability of a woman still working full time by nearly 50%, however it hardly affects the hours worked by part-timers (Cristia, 2008). Cáceres-Delpiano (2012) finds that having multiple children reduces the percentage of women employed. However, this applies especially the jobs that are easier to terminate such as self-employment or volunteering jobs. It is not always a woman’s choice to reduce or terminate her labor participation. Mammen & Paxson (2009) state that in some countries women are simply forced to stop working upon marriage. In Japan, employers often consider women with forthcoming childbirth to be unable to fully commit to the job as their male counterparts (OECD Observer, 2014). In France women often withdraw from the labor market upon motherhood. The family roles are still old-fashionably divided. If women work, it us usually full-time (Castellano & Rocca, 2014).
before having children or while raising them. It could be identified as a part-time culture. However, if these country’s GDP per capita are taken into consideration it could just be a luxury position of not having to work. (Anxol, Fagan, Cebrian, & Moreno, 2007) The part-time option does not opt for women in Mediterranean countries where they generally apply an ‘exit or full-time’ choice to women. Having children therefore does appear to negatively influence the female labor participation rate in those countries (Vlasbom & Schippers, 2004). Among Latin American countries there is quiet some inconsistently, about the effects of having children on female labor participation (United Nations, ECLAC, 2005). According to Moschion (2013) Latin American countries are hardly effected.
There are scholars who find a negative relationship and scholars who find a positive relationship. Not surprisingly that some have additional thoughts to that. Mishra & Smyth, (2010) criticizes the findings and argues that there is no consensus within the current scientific literature. Therefore, no direct relationship between the TFR and the FLPR is proven. The different outcomes can be explained by the use of different measurements that resulted in different findings (Felmlee, 1993). Female labor force participation can effect fertility, but fertility can also affect female labor force participation. Panopoulou & Tsakloglou (1999) were at least not able to determine the direction. It could be a multidirectional relationship. If the relationship between the TFR and the LFPR is measured at national levels there are in fact significant relations noticed. Vere (2011) and Hirvonen (2008) found that there are some little differences concerning the influence of the TFR on the FLPR. It depended on the time period from which the data was retrieved.
Previous research has not given consensus on the relationship and direction between fertility and the female labor participation. A simple explanation is that many studies based their analysis on different input or data. Some studies use information based on questionnaires rather as the census. Other analyze employment instead of labor market participation, or focus on the influence of having a second child instead of a national fertility rate. This is a cross-country study.
3.2 Education and higher professional positions
educational institutions (Castellano, Punzo & Rocca, 2012). Also, higher educated women have better employment and income perspectives (Cameron et al., 2001). In higher income countries with lots of higher educated women, there are also more women who work (Bloom et al., 2009). It is simply a matter of opportunity costs. Nuevo-Chiquero (2014) states that it is based on choices to maximize the household product. Motivating higher educated women through policies towards labor participation is a sum of the policy costs and the potential profit generated (in taxes) by the novices who are going to work. From an economic perspective, it is interesting to determine which countries utilize their human capital to the fullest. Since that also reveals approximately what percentage there is still to gain, i.e. the educated women who are not working. There are also women who did enjoy a tertiary education, but only work to a limited extent. They do not pursue a career. Hence, the women work below their level and are missing out on almost half of what could have been their salary (Fogli & Veldkamp, 2011). This also influences the potential taxes that can be received by the government. The reason to put the TFR into the picture is because it could be affected by full-time employment. It is of great economic value to have the TFR at least on a subsistence level. It is known that it is hard to combine a challenging job with a full house. Anyone who pursues a career in higher management levels or similar leading positions is required to be dedicated to the job and bound to work at least 40 hours a week (Schwartz, 1990). Business executives notice the ‘family price’ women have to pay. This was hardly applicable to men (Gochman, 1990). However, higher management positons or equivalents are accompanied by generous paychecks. This also increases the opportunity costs to work, staying home becomes expensive. The high income enables these women to outsource all kinds of other tasks, it simply comforts life. Housekeeping assistance and formal childcare are no issue to women in management or equivalent positions. According to Bratti (2003) the availability of these supporting services is a reason why higher educated women are able to have (more) children. Also, if women withdraw from the labor market too long, their market value will depreciate. (Rocha, & da Fuster, 2006). However, since having children is one of the basic elements of our existence, nature ensured a strong desire to do so. Hence, this paper argues that the decision to have children is made at some point prior to aspiring a professional top position. Many studies focused on the postponement of childbirth by higher educated women. It was found that the postponement of childbirth has no significant influence on the total number of children a higher educated woman gives birth to (Bloemen & Karwij, 2001; del Boca, Pasqua, & Pronzato, 2005; Moffitt, 1984).
found that in countries with a greater percentage of higher educated women the TFR was higher as well. Higher educated women make more hours than lower educated women when preparing for a birth event (del Boca & Locatelli, 2006). According to Fehr & Ujhelyiova (2013) lower educated women reduce their working hours by 31%, while higher educated women only reduce their labor supply by 4.6%.
As mentioned earlier influence favorable governmental policies the probability that a mother works. Countries like the Netherlands, Sweden and others which are known for their favorable policies to encourage female labor participation, simultaneously have relatively high TFRs and FLPRs. Higher educated women and labor participation are significantly related. However, no previous study found if these women are actually employed at higher positions. Based on the previous information in this section, women are able to outsource some of their obligations. Moreover, if the job outweighs the benefits of staying home, it can be assumed that they do fulfill a position similar to their educational level. Therefore, it is expected that a high TFR and a high TLPR can predict a high percentage of women in top positions.
3.3 Hypotheses ‘Women in Top Positions’
The theme ‘Women in Top Positions’ consists out of three different variables: (1) female top managers, (2) firms with female (partial) ownership and (3) seats held by women in national parliaments. These different variables represent different types of positions women may have in higher positions. Each variable will be tested for the influence of the TFR and of the FLPR separately. This resulted in the following hypotheses:
Hypothesis 1A: A high TFR positively influences the percentage of female top managers. Hypothesis 1B: A high TFR positively influences the percentage of firms with female
ownership.
Hypothesis 1C: A high TFR positively influences the percentage of seats held by women in national parliaments.
Hypothesis 2A: A high FLPR positively influences the percentage of female top managers. Hypothesis 2B: A high FLPR positively influences the percentage of firms with female
ownership.
Briefly summarized, a positive combination of a high percentage of the TFR and a high percentage of the FLPR is expected to predict different types of high percentages of women’s top positions. The three sub positions of women in top positions, top managers, firm ownership and women in national parliaments does not cover all types of positions yet. Another type of top position that is somewhat comparable, yet completely different is entrepreneurship. Entrepreneurship can be considered a type of top position as well. Entrepreneurship will be discussed in the next paragraph. The term ‘women in top positions’ will be used further throughout this research and refers to all the types of top positions mentioned here, top managers, firm ownership and seats in parliament. Entrepreneurship will be discussed separately due to the fact that it is a different type of top position of the ones mentioned above (does not require education etc.) and some data limitations.
3.4 Hypotheses ‘Female Entrepreneurship’
Running a business generally requires different competences as working on pay-roll. However, since personality and competences have a major influence on entrepreneurial success, entrepreneurship is within this research also considered a top position. Nonetheless, it differs from the type of top positions described in the previous paragraph. In contrary to the top positions described in the previous paragraph you do not need any official qualifications, such as diplomas, in order to be an entrepreneur. This makes entrepreneurship a different theme within this research. As a result entrepreneurship is separately described and analyzed.
which continuously needs to be aware of keeping its population balanced and economy growing.
However, contrary to men, who experience great financial incentives, the unlimited opportunities to increase wealth are generally not the motivation of women to start a business (Allen & Curington, 2014). Women rather consider it a replacement for unemployment or part-time work (Georgellis & Wall, 2005). Women face greater part-time constraints that need to be managed. Self-employment creates more flexibility in the working hours, and certain tasks such as administration can be conducted while looking after children (Aronson, 1991; Presser, 1992). Compared to employment on pay roll, self-employment creates lower opportunity costs for women who have children (Noseleit, 2014). Especially women with younger children, who would have to pay for childcare instead (Edwards & Field-Hendrey, 2002). Not surprisingly, women who start a business were usually unemployed or not participating in the labor market (Rosti & Chelli, 2005). It appears that women start a business whenever the time arises and often combine it with motherhood. Opportunism is usually the motivation of women for new business creation in developed countries (Brush & Cooper, 2012). Entrepreneurship in developing countries is often a type of forced self-employment due to a lack of formal employment options (Langowitz & Minniti, 2007). So, if solely developed countries are taken into account, female entrepreneurs are often mothers who were unemployed or did not participate in the market. This indicates that these women contribute to a greater percentage of labor participation and cause a relatively greater percentage of female entrepreneurs among those who participate in the labor force. In order to see if these assumptions are to be taken for real arguments, are the following hypotheses presented below.
Hypothesis 3A: A high TFR positively influences the percentage of female entrepreneurship. Hypothesis 3B: A high FLPR positively influences the percentage of female entrepreneurship.
4. Research methodology
This research focuses on the TFR and the FLPR and the relationship with types of higher occupations held by women. Throughout this research a deductive method is applied in order to work from background theories towards a theoretical framework and the hypotheses as listed in the previous chapter. Considering the type and subject of the research questions, a quantitative research approach will be most appropriate and therefore applied.
4.1 Data sources
In order to test the previous stated hypotheses a data set is required. This research uses of secondary data. Secondary data can be of superior of quality compared to data collected by just one scholar (Thomas, 2004). The data that will be used for the quantitative analysis are gathered from two main sources.
The first database is the World Bank (2014), which provides data for the majority of the variables. The World Bank is a large organization and employs over 10,000 people in 120 offices around the world. The main targets of the World Bank are to fight extreme poverty and to increase the income of the bottom 40% in every country. In order to support these targets, the World Bank provides an extended range of datasets and international statistics. The comprehensive datasets equip people with over 1200 different indicators from 1980 to 2014, in order to anticipate on the world currents pitfalls and to acknowledge where development is needed most. The type of research conducted requires the most recent data available. The data used is the most recent available and not older than 2010.
The second database that is of great relevance to this research is the database of GEM Consortium. GEM represents the ‘Global Entrepreneurship Monitor’ and includes 75% of the world population and 89% of the GDP. Each year GEM analyzes the entrepreneurial activity, aspirations and attitudes of people around the world. GEM collects its data through an annual survey among at least 2000 participants per country (Amorós & Bosma, 2014). The data used are the most recent available and not older than 2010. Some countries only measure or publicize information every other year. The reference list provide an overview of the locations where each variable is extracted from.
4.2 Sample
reliable, a sample should produce the same results every time it is tested. If those results also can be generalized it is considered to be externally validated (Thomas, 2014). The validity of a study can be defined as “the extent to which the results of a research study are well-founded and correspond accurately to the real world” (Brains, Willnat, Manheim & Rich, 2011).
In this case the real world is literally is the whole world. Hence, not every country in the world publicizes all its data. Therefore, all countries who lack data are excluded. There is extremely little overlap among the countries that contain data for the “female entrepreneurship” theme and the “women on top positions’ theme. Therefore, it is chosen to use data from different countries for the different themes. This results in the use of two different samples.
4.3 Outliers
After correcting the data for missing data entries the data were checked for outliers. All countries that were not within the boundaries determined by the formula of Tukey (1977) are outliers and eliminated from the sample. An explanation of the formula and the calculation of the boundaries for outliers are available in appendix 1. The list of countries that remain for the theme of ‘Women in top positions ’are presented in table 1. The sample includes 77 countries and covers all continents.
Table 1. Sample list Women on top positions
1 Afghanistan 41 Kyrgyz Republic
2 Albania 42 Laos PDR
3 Angola 43 Latvia
4 Argentina 44 Lebanon
5 Armenia 45 Lithuania
6 Azerbaijan 46 Macedonia, FYR
7 The Bahamas 47 Mexico
8 Bangladesh 48 Moldova
9 Barbados 49 Mongolia
10 Belarus 50 Montenegro
11 Belize 51 Myanmar
12 Bolivia 52 Nepal
13 Bosnia and Herzegovina 53 Nicaragua
14 Botswana 54 Panama
15 Bulgaria 55 Paraguay
16 Central African Republic 56 Peru
17 Chile 57 Poland
18 China 58 Romania
19 Colombia 59 Russian Federation
20 Congo, Dem. Rep. 60 Serbia
21 Costa Rica 61 Slovak Republic
22 Croatia 62 Slovenia
24 Dominican Republic 64 St. Lucia
25 Ecuador 65 Surinam
26 El Salvador 66 Tajikistan
27 Estonia 67 Tanzania
28 Ethiopia 68 Trinidad and Tobago
29 Georgia 69 Turkey 30 Ghana 70 Uganda 31 Guatemala 71 Ukraine 32 Guyana 72 Uruguay 33 Honduras 73 Uzbekistan 34 Hungary 74 Venezuela 35 Iraq 75 Yemen 36 Jamaica 76 Zambia 37 Jordan 77 Zimbabwe 38 Kazakhstan 39 Kenya
40 St. Vincent and Grenada
Source: Author
The sample for the ‘Entrepreneurship’ theme is presented below in table 2. The sample is corrected for outliers and the calculation and outliers are available in appendix 2. The final size of the sample is 71 countries and it includes countries from all continents.
Table 2. Sample list Female entrepreneurship
1 Argentina 38 Libya 2 Australia 39 Lithuania 3 Austria 40 Macedonia 4 Bangladesh 41 Malaysia 5 Barbados 42 Mexico 6 Belgium 43 Namibia
7 Bosnia and Herzegovina 44 Netherlands
8 Botswana 45 Pakistan 9 Brazil 46 Panama 10 Canada 47 Peru 11 Chile 48 Philippines 12 China 49 Poland 13 Colombia 50 Portugal
14 Costa Rica 51 Puerto Rico
15 Croatia 52 Romania
16 Czech Republic 53 Russian Federation
17 Denmark 54 Singapore
18 Ecuador 55 Slovak Republic
19 Egypt 56 Slovenia
20 El Salvador 57 South Africa
21 Estonia 58 Spain
22 Finland 59 Surinam
24 Germany 61 Switzerland
25 Ghana 62 Thailand
26 Greece 63 Trinidad and Tobago
27 Guatemala 64 Tunisia
28 Hungary 65 Turkey
29 India 66 United Arab Emirates
30 Indonesia 67 United Kingdom
31 Ireland 68 United States
32 Israel 69 Uruguay
33 Italy 70 Venezuela
34 Jamaica 71 Vietnam
35 Japan
36 Korea, Dem. Rep 37 Latvia
Source: Author
4.4 Variables
This study involves two independent and six dependent variables. All variables concern data retrieved from the mentioned databases previously, World Bank and GEM, and date from the most recent data available no later than 2010.
4.5 Independent variables
Total fertility rate (TFR) is an independent variable that can be defined as the total number of children a woman would give birth to if she would live at least to the end of her fertile years. The TFR is a commonly used variable by scholars to determine relationships and influences, such as the years of education (Da Rocha, 2006; Mishra & Smyth, 2010; Fehr, 2012; Bloom et al., 2009). The TFR is retrieved from the World Bank who publishes the TFR of 199 countries annually. The World Bank collects the input from other sources such as data provided by the United Nations, Eurostat and national census bureaus.
Female labor participation rate (FLPR) refers to the part of the female population who is older than 15 years and contributes to society through labor, i.e. the people who are economically active. The FLPR is available for 185 countries. The data are retrieved from the World Bank who collected them from a selection of sources including the United Nations, Census reports, Eurostat and the U.S. census bureau.
4.6 Dependent variables
can be defined in a variety of ways. In this study it concerns women who fulfill a professional job within a public or governmental organization at a higher management level or comparable.
Female top managers refers to the percentage of firms within a nation who report to have female top managers. This variable will be one of the variables that is used to test the topic of ‘Women in top positions’. The percentage of women in top management positions is interesting because it is known that those type of positions require usually over 40 hours of work per week. Therefore, it should be harder to combine with children. The data are collected from World Bank who conducted a survey.
Seats taken by females within national parliaments refers to the percentage of seats in a single or lower national parliament that are occupied by women. Such governmental positions held by women is of great importance for a country, because it enables women to influence policies. This variable therefore also contributes to the topic of ‘Women in top positions’. The data that are gathered from the World Development Indicator by the World Bank, who retrieved this information from the Inter-Parliamentary Union.
Firms with female ownership refers to the percentage of firms within a nation that have women among the principle owners of the firm. The information is gathered through a survey by the World Bank and the percentages are published on their website. The sample includes data from 86 countries. The variable is used to test the hypotheses on ‘Women in top positions’. Ownership could also refer to entrepreneurship, but in this case it only includes women who are part of a management team or a board of directors, and who therefore are also employees rather than entrepreneurs.
Female entrepreneurship refers to the percentage of women within the working age of a country who are a nascent entrepreneur or already own and manage a new business. The data are gathered and published by GEM Consortium (2014). GEM, Global Entrepreneurship Monitor is considered to conduct the greatest study on entrepreneurship worldwide. The database includes data on 81 countries of which the most recent data is entered and no older than 2010. The working age of a population is considered to be 18-64 years in this database. Nascent entrepreneurs are those entrepreneurs who considered those people who are attempting to set up a business, which they also (partially) own.
4.7 Control variables
population. However, the average should be considered extremely generalized. There are great income differences among people, but it is useful when comparing countries. GDP per capita is in close relation to labor force participation and therefore used as a control variable. The latest data published by the World Bank is used.
Education, refers to the school life expectancy of females on average. This is the average years of schooling women have received through primary, secondary and tertiary education. The data are extracted from the World Factbook published by the Central Intelligence Agency of the United States. Education is often linked to the number of children per women. Moreover, education influences the attractiveness of labor participation.
4.8 Data analysis
5. Descriptive results
This chapter will discuss the descriptive results of the data that are used. This research uses two different samples, (1) for ‘Women in Top Positions which contains three variables and (2) for the fourth variable ‘Female Entrepreneurship’. The samples will be described in the same order.
5.1 Women in top positions
The sample for the theme Women in Top Positions consists out of 77 countries. The sample covers five major continents in the world. The greatest share is taken by Latin American countries which also includes countries from Central America such as Mexico and Costa Rica. The second greatest share is taken by countries from Europe. Hence, it must be noted that the sample lacks data from West European countries. Africa concerns in this case merely sub-Saharan countries. Countries from the Middle East include typical countries such as Yemen, Israel and Afghanistan. In addition it also includes the countries who officially belong to the middle-east, but have some historical ties with Russia and the former Soviet Union, such as Georgia, Kazakhstan and Uzbekistan.
Russia is politically considered to belong to Europe, however Russia its greatest land share is in Asia. Russia is an extremely large and independent country, especially if its cultural and institutional characteristics should be labeled into Europe or Asia. Based on its size, geographic dispersion and independent institutional and cultural characteristics it is chosen to not label Russia as European nor Asian. It will only blur the characteristics per continent. The percentage share per continent is presented below in table 3.
Table 3. Division of continents among the sample of women in top positions
Frequency Percent Valid Russia 1 1,3 Sub-Saharan Africa 12 15,6 Asia 8 10,4 Europe 18 23,4 Latin America 26 33,8 Middle East 12 15,6 Total 77 100,0
the ‘Women in top positions’ is given below in table 4. Separate overviews per continent are available in appendix 3.
Table 4. Descriptive Statistics dependent variables ‘Women in Top Positions’
N=77 FLPR TFR Firms with female ownership Firms with female top manager Women in national parliaments Education GDP per capita Mean 53,18 2,603 35,343 18,73 20,03 12,73 6299,90 Mode 56 1,5a 43,5 19 13 13 572 Std. Deviation 15,106 1,2338 13,7374 8,862 9,532 2,850 6064,32 Minimum 15 1,3 4,1 0 0 6 39 Maximum 88 6,0 76,0 39 40 18 22729
a. Multiple modes exist. The smallest value is shown.
5.2 Female entrepreneurship
The sample for female entrepreneurship differs in countries from the previous described database. The database for the topic on female entrepreneurship consists out of 71 countries and covers all continents. The sample consist for 39,4% out of European countries. Hence, in contrary to the sample of ‘Women in top positions’, are the West European countries represented within this sample. Again Russia is labeled individually. This sample includes relatively little countries from the Middle East. However, it does include three North African countries, which are countries that share some characteristics with countries from the Middle-East.
Table 5. Division of continents among the sample of Entrepreneurship
The mean, standard deviation, minimum and maximum values of the sample are presented below in table 6. The descriptive results per continent are available in appendix 4.
Table 6. Descriptive statistics dependent variable ‘Female Entrepreneurship’
N=71 FLPR TFR Entrepreneurship GDP per capita Education Mean 51,21 1,973 9,908 19479,59 14,68 Mode 56 1,8 6,2 97a 16a Std. Deviation 10,922 ,5887 7,0812 19643,493 2,666 Minimum 24 1,3 1,2 97 7 Maximum 73 3,9 34,4 80477 20
a. Multiple modes exist. The smallest value is shown.
5.3 Multiple modes
6. Testing Assumptions
This chapter will test several assumptions in order to check if the data of both samples are suitable for the contemplated multiple regression analysis. The data will be tested for multicollinearity. This test will also provide clarity if the TFR and the FLPR are related. Subsequently the data of both samples are tested for non-linearity and heteroscedasticity.
6.1 Multicollinearity
Before conducting the regression analysis, the correlation matrices need to be created first. A correlation matrix is used to detect multicollinearity. Multicollinearity is a significant correlation among variables (Field, 2009). The correlation matrix for the sample of Women in Top Positions is presented in table 7. A different sample is used for the theme ‘Female Entrepreneurship’. The corresponding correlation matrix is created and presented in table 8. The correlation matrices are conducted with a one tailed analysis since it merely tests for positive hypotheses.
The Pearson’s R is used to measure the strength of the linear relationship between two variables. In the case of R=1 there is a perfect positive linear relationship among the tested variables. If X increases, Y increases simultaneously in the same direction. If R= -1, the opposite occurs. If X increases, then Y will decrease. Finally, there is also the possibility that there is no relationship at all. In this case the R=0 (Field, 2009).
A correlation is considered to be significant if it is < 0.7 (Biostatistics Resource Channel, 2013). This could indicate that there is a case of multicollinearity and indicates a potential threat of outcome bias. Values mentioned by others are <0.5 (Cohen, 1988) or <0.9 by Field (2009). However, the most commonly used number is <0.7. Therefore, this will be the number that is applied in this research. Examining the matrix of female entrepreneurship it appears that there are no correlations exceeding the 0.7. The outcomes that appear to correlate below the <0.7, but above the >0.5 are highlighted in Bold in the matrices.
6.2 Non-linearity
After multicollinearity, non-linearity needs to be tested too. Non-linearity refers to a chaotic or random case. In the case of non-linearity there are limited options available of generalizing data (Field, 2009). Therefore, the data is tested for linearity and normality. This can be conducted through creating a histogram and a plot.
A histogram can be used to check the sample data for normality. It shows the frequencies of all measured observations. A normal distribution can be checked by imagining a line in the middle of the histogram. If the left and right from the middle are symmetric the data are perfectly normal distributed. However, data are usually not perfectly symmetric. The data should be bell-shaped. The bell-shape is a good indication of normally distributed data. The shape shows that most data is centered on the exact middle.
The P-P plot is normal when the diagonal line is followed by the dots. Considering the sample size, not extremely big, yet sufficient, it could be possible that the P-P plot appears to be somewhat wavy. This can be due to a limited sample size. The P-P plot can be checked in combination with the histogram. If the histogram looks abnormal and the P-P plot looks very wavy, the data may be non-normal. If the P-P plot is wavy, but the histogram appears to be normal. This is caused by the limited sample size and does not violate the assumptions.
Per dependent variable a histogram and a plot is created and presented in appendices 7 - 10. The graphs and plots show normal and linear data. The data clusters in the center of the histogram. In the case of the plot, all dots should cluster along the line. The dots are called residuals and these observations are the predicted points. Dots clustering along the line meet the assumption of normality. All four dependent variables meet the assumption of normality.
6.3 Heteroscedasticity and homoscedasticity
Table 7. 1 tailed Pearson’s R Correlation matrix of the sample ‘Women in top positions’ N=77 Mean Std. Deviation FLPR TFR Firms with female ownership Firms with female top manager Women in national parliaments Education GDP per capita FLPR 53.18 15.11 1 TFR 2.6 1.23 ,264 * ,010 1
Firms with female Ownership 35.34 13.74 ,289** ,005 -,019 ,436 1
Firms with female top managers 18.73 8.86 ,318** ,002 -,296** ,004 ,539** ,000 1 Women in national parliaments 20.03 9.53 ,144 ,105 ,072 ,267 ,145 ,105 ,094 ,208 1 Education 12.73 2.85 -,132 ,127 -,659** ,000 ,058 ,309 ,276** ,008 ,021 ,428 1 GDP per capita 6299.9 6064.3 -,055 ,316 -,507** ,000 ,110 ,171 ,128 ,133 ,019 ,436 ,624** ,000 1
Table 8. 1 tailed Correlation matrix of the sample ‘Female Entrepreneurship’
N=71
Mean Std. Deviation FLPR TFR Female entrepreneurship GDP per capita Education FLPR 51,21 10,922 1 TFR 1,97 ,5887 -,089 ,230 1 Female entrepreneurship 9,91 7,0812 ,393** ,000 ,457** ,000 1 GDP per capita 19479 19643 ,226 ,029 -,379** ,001 -,325** ,003 1 Education 14,68 2,67 ,111 ,178 -,482** ,000 -,328** ,003 ,462** ,000 1
7. Multiple regression results
The hypotheses of this research are tested by means of a multiple regression analyses. The data will be entered and calculated through the SPSS program. A multiple regression analysis is useful to determine to what extent a combination of independent variables can predict the course a dependent variables takes (Baarda, de Goede, & van Dijkum, 2004). Since both TFR and FLPR are considered to be independent variables, they are expected to influence the dependent variables, firms with female ownership, firms with top managers, women seated in national parliaments and female entrepreneurship. For each dependent variable 15 combinations are created to test different combinations and identify the influence of the independent variables and control variables on the dependent variables. These are all available in the appendices: appendix 9 provides an overview of female top managers, appendix 10 of females with participation in ownership, appendix 11 of seats held by women in national parliaments and appendix 12 of female entrepreneurship. The tables 9, 10 and 11 that will follow in this chapter are concise overviews of relevant findings this chapter refers to. All models display the β, standardized regression coefficients. Below the β, the standard error is displayed within ( ), brackets. The R² is inserted in order to indicate how much variance the model can explain compared the total that could be explained (Field, 2009). The F-statistic is included to present the comparison of the amount of systematic variances among the data.
The magnitude of the relationship between the TFR and female top managers can provide an example of the influence of the TFR on female top managers. The β-value of the TFR for female top managers should be multiplied with the standard deviation of female top managers. The outcome shows the change to the percentage of female top managers for every standard deviation the TFR increases. For instance, if the TFR of Uruguay is 2.1 If it is increased with one standard deviation it becomes 3.33 (2.1 + 1.2338)1. The percentage of female top managers in Uruguay is 19. The increase of one standard deviation of the TFR will decrease the percentage of female top managers with 3,19% (β-value -0.360*Std. deviation 8.862). To Uruguay this will decrease their percentage of top managers to 15.81% (19 – 3.19). The TFR is negatively influencing the percentage of female top managers.
The FLPR is tested to be significant in every combination at a significance level of (p < 0.01). If tested individually it is significantly positive at ,318 (p < 0.01), combined with all other variables it results in ,429 (p<0.01). Hypothesis 1A predicted a positive influence of the TFR on the percentage of female top managers within countries. It turns out, that the TFR in fact has a significantly negative influence. Therefore, hypothesis 1A cannot be supported. Hypothesis 2A also predicted a positive relationship, yet for the FLPR. All combinations of variables show significance at a (p < 0.01), meaning that hypothesis 2A is supported. However it must be noted that the R² is 0,101 and for that matter only explains 10%, which is a very limited percentage.
Hypothesis 1B links a high TFR to the percentage of firms with female ownership and expects this relationship to be positive. Hypothesis 2B expects that a high FLPR positively influences the percentage of firms with female ownership. The outcomes discussed in this paragraph are all presented in table 10. All other statistic results are available in attachment 10. Despite the expectation that the TFR and FLPR could predict female ownership of firms within a country, only the FLPR appears to be influential. At ,289 (p < 0.05) FLPR is tested independently and positively. When tests are combined with the other research variables as presented in table 10, the FLPR has a significance of ,307 at (p < 0.01). The positive relationship is significant throughout all combinations of variables tested that apply to the dependent variable ‘firms with female ownership’. The TFR shows no significance in any model alike as the control variables. Moreover, the R² of ,014 when the TFR, GDP per capita and education are combined supports the insignificance of the influence of these variables of female firm ownership. Based on these findings it can be stated that hypothesis 1B, that assumed a positive
link between TFR and female firm ownership is not supported. This is contrary to the outcome of hypothesis 2B that is supported and therefore confirms that there is a positive relation between a nation’s FLPR and female firm ownership.
Hypothesis 1C states that a high TFR positively influences the percentage of seats held by women in national parliaments. Hypothesis 2C posits that a high FLPR positively influences the percentage of seats held by women in national parliaments. Both hypotheses are tested and all results came out insignificant. There is nothing that tests significant on the seats held by women in national parliaments. Therefore the results are placed in the appendices and can be found under appendix 11. There is not even the smallest sign of a significant relationship with the independent nor the dependent variables. The lack of significance indicates that all high government positions are fulfilled politically correct. Meaning that there is little difference between countries with a high fertility rate or high female participation rate, compared to countries with low rates and the number of women seated in their parliaments. Obviously, hypotheses 1C and 2C are not supported.
Table 9. Regression table of the dependent variable Female Top Managers
Model N=77 Dependent variable ‘Female Top Managers’
Constant 7,794* (4,500) 24,257*** (2,279) 8,814** (3,549) -5,330 (5,675) 14,312*** (3,946) 7,388 (7,606) 6,015 (7,688) Control variables GDP per capita -,075 (,000) -,147 (,000) Education ,276** (,345) ,324*** (,325) ,112 (,416) ,187 (,464) Independent variables TFR -,296*** (,792) -,448*** (,872) -,333** (,988) -,360** (1,002) FLPR ,318*** (,064) ,360*** (,061) ,432*** (,062) ,420*** (0,61) ,429*** (,061) R² ,076 ,088 ,101 ,204 ,260 ,263 ,276 F-statistic 6,196** 7,193*** 8,421*** 9,479*** 8,551*** 8,685*** 6,858***
* Correlation is significant at < 0.1 level ** Correlation is significant at < 0.05 level *** Correlation is significant at < 0.01 level
Table 10. Regression table Female with ownership in a business
Model N=77 Dependent variable ‘Female participation in firm ownership’.
Constant 21,370*** (5,555) 30,996** (13,110) 19,975 (13,279) Control variables GDP per capita ,129 (,000) ,100 (,000) Education ,014 (,833) ,006 (,801) Independent variables TFR ,056 (1,744) -,045 (1,730) FLPR ,289** (,101) ,307*** (,106) R² ,083 ,014 ,101 F-statistic 6,831** ,343 2,020
Table 11. Regression table of the dependent variable Female Entrepreneurship
Model N=71 Dependent variable ‘Female Entrepreneurship’.
Constant -,937 (2,651) -3,143 (3,757) -16,381*** (4,067) -9,693 (6,301) Control variables GDP per capita -,260** (,000) Education -,090 (,289) Independent variables TFR 0,457*** (1,288) ,496*** (,061) ,359*** (1,253) FLPR ,393*** (,072) ,437*** (1,136) ,494*** (,060) R² ,209 ,155 ,398 ,472 F-statistic 18,208*** 12,610*** 22,523*** 14,768***
8. Discussion
This research tried to determine the influence of the TFR and the FLPR on specific higher professional positions held by women. This chapter will discuss the results of all four occupations one by one. The outcomes of this research contribute to current literature. It exposes outcomes that were not published previously from a cross-country perspective. The relationship between the TFR and the FLPR was discussed earlier in this research. The literature showed no consensus on this relationship. The multicollinearity analyses of this research also showed different results, yet not significant. The two different samples used in this research, 1) ‘Women in Top Positions’, 2) ‘Female Entrepreneurship’ have a minimal overlap in countries. This indicated that the set of countries influences the outcome of the results. The fact that it differs among samples implicates that in future research it is important to check the relationship of the specific dataset.
family dilemmas). In that case the data should not be cross-country based like this research, but they should be gathered through surveys.
Female ownership of firms, turned out to be only significantly positive related to the FLPR. This implies that if more women are working, their involvement through business ownership also increases. The relationship with the TFR is slightly negative, but insignificant. Therefore, again the arguments of comfort created by the high position job do not hold. Alike the top management positions, running a company takes great amounts of time. It seems that the time factor overrules the argument that the job can be combined based on the privileges it provides. The percentage of women in national parliaments appears not to be influenced by the TFR or the FLPR. The national policies that try to promote or discourage women to work and have children, differ heavily per country. However, this is not related to the number of women in their parliaments. The most logical argument for the insignificant results is that the number of women in national parliaments is subject to all kinds of other, more political instead of economic, factors. Another factor that could bias the insignificance is the extremely limited size of people within national parliaments, especially compared to the other variables that focus on all women in a country. Perhaps a measurement technique that includes a percentage of all women in higher political positions results in a different outcome.
The fourth type of occupation tested by this research is the percentage of female entrepreneurs. Entrepreneurship can be conducted on many different levels. Some only provide services from time to time, while others run a large business producing products and work over 60 hours a week. The flexibility and self-influence on the intensity of this occupation was expected to be a determining factor (Aronson, 1991; Presser, 1992; Noseleit, 2014) It enables women to combine motherhood and generate an income (Georgellis & Wall, 2005). It should be taken into consideration that in some cases women did not become an entrepreneur completely voluntarily. These women are considered so-called necessity entrepreneurs (Shane & Venktaraman, 2000). The other group of entrepreneurs are those who act out of opportunism. However, current data did not provide the ability to distinguish them. The TFR and the FLPR were tested significantly positive and therefore confirm hypotheses 3A and 3B. Entrepreneurship contributes to a country’s innovation and economic development (Schumpeter, 1934). The results of this research, along with the provided arguments, indicate that entrepreneurship is combined with motherhood by many women. The results provide additional arguments to encourage more women to combine motherhood and entrepreneurship.
clearly differs from employment by the state (seats in parliament). Several hypotheses tested positive significantly. However, the levels of significance are all <0.7. Also, the R²s are not extremely high. All in all, a limited percentage up to 40% can be explained by the TFR and the FLPR.
9. Conclusion
The relationship between the TFR and the FLPR and their relationship to different factors are overly analyzed. However, the type of top occupations held by women is not one of these factors. This research was designed based on that research gap which was identified through literature research. Women have always contributed less to a country’s national product compared to men. This is still relevant despite several feministic revolutions. Scientific research could contribute by explaining whether the TFR and the FLPR are related to the percentage of women in top positions. The research question is therefore formulated as follows:
“What is the relationship between the total fertility rates and the female labor participation rates on types of top occupations held by women?”
The hypotheses of this research tested the influence of the TFR and the FLPR on four different types of occupations held by women. Three occupations tested in this research were classified as so called top positions; (1) female top managers, (2) women who have ownership in a business, and (3) women who are seated in national parliaments. A fourth top occupation tested is (4) female entrepreneurship. An entrepreneur may or may not be in a top position. It depends on the size and impact of her business. The importance and positive economic contribution of entrepreneurship is acknowledged internationally. Women in top positions are important to a country from several perspectives, such as education brought into practice and female influence, power and perspectives. Higher positions simply increase the economy by larger amounts of money per capita compared to positions that are far from being considered a top position.
arguments that women could combine top positions with having children. In fact, top positions provide higher payments and privileges. Also higher payments increase the opportunity costs not to work. Also, withdrawal from the labor market for a few years will decrease a person’s professional market value. Nonetheless, the results of this research show that these arguments do not outweigh the benefits of staying home.
The outcome on positions held in national parliaments did not show any significance. It appears that the TFR and the FLPR have simply no relationship or influence on this matter. This research also expected the TFR and the FLPR to positively influence female entrepreneurship within countries. This was confirmed by significantly positive results. The main argument is that entrepreneurship offers the benefits of flexibility in time and intensity. This appeals to women and enables them to combine work and children. This factor generally does not apply to the other top positions tested in this research.