Institutional triggers that drive entrepreneurship - 20th of June, 2014 1

Hele tekst

(1)Institutional triggers that drive entrepreneurship - 20th of June, 2014. 1.

(2) Master thesis Jan Hakvoort - 20th of June, 2014. Institutional triggers that drive entrepreneurship A cross-country analysis Final version. Author Jan Hakvoort | S2402297 Master: International Business and Management Chopinweg 4, Eelde E: j.n.b.hakvoort@student.rug.nl | jan_hakvoort@live.nl T: + 31615875773. University University of Groningen Faculty of Economics and Business. Thesis supervisor. Co-assessor. Ms. Dr. M.J. Klasing. Ms. Dr. S.N. Ponsioen. Date 20th of June, 2014. 2.

(3) Institutional triggers that drive entrepreneurship - 20th of June, 2014. ABSTRACT Most of the research that tries to explain why some countries foster entrepreneurship better than others have either used a cultural or an institutional perspective to address this question. This paper is the first in its kind that tries to explain why these cross-country rates differ by exploring the cultural and institutional triggers that drive entrepreneurship. This is done, by using cross-country data from several databases over the time period 2009 - 2013. The sample consists of 64 countries, representing six continents across the globe. The findings show that attitudes toward entrepreneurship are positively associated with entrepreneurial activity and for that reason can be seen as the main triggers. Furthermore, the protection of property rights is only positively associated with entrepreneurial activity in developed economies. Keywords: attitudes, cross-country analysis, entrepreneurship, institutions. 3.

(4) Master thesis Jan Hakvoort - 20th of June, 2014. ACK OWLEDGEME TS Before I started to write my thesis, I thought I would never be able to do a quantitative research. I had no experience with this kind of research and I only had limited understanding of statistics. But after half a year of hard work and pooling all the data, findings and results, I can finally say that I succeeded. Writing this thesis gave me the opportunity to challenge myself by going beyond my own borders. As a result, I developed myself in various ways. In the first place, I learned how to conduct a quantitative research. In order to do this, I had to read a lot of statistical books and SPSS manuals. I learned how to interpret the results from the regression analysis and subsequently to draw conclusion from them. Secondly, I learned a lot about entrepreneurship and institutions in general and institutional triggers that drive entrepreneurship in specific. Lastly, I have developed myself in perseverance. Even though there were moments during my thesis that I was fed up, did not know how to continue, or encountered a problem, I moved on. Of course, I would not have been able to write my thesis without the support of my family and friends. Therefore, I would like to take the opportunity to thank several people. In particular, I would like to thank Lieke Noteboom, Marnix Roffel and Pim Wiegman for motivating me and giving me advice in times of difficulty. I also would like to thank my brother Wim Hakvoort, for helping me with creating the front page. Moreover, I would like to thank my thesis supervisor Mariko Klasing for her support, guidance, and feedback during the semester. Lastly, I would like to thank you, the reader, for your interest in the topic. Hopefully you will get inspired by the topic as much as I am.. Eelde, June 2014. Jan Hakvoort. 4.

(5) Institutional triggers that drive entrepreneurship - 20th of June, 2014. TABLE OF CO TE TS 1.. INTRODUCTION .............................................................................................................. 8. 2.. BACKGROUND .............................................................................................................. 10. 3.. 2.1. Entrepreneurship ........................................................................................................ 10. 2.2. Triggers to entrepreneurship ...................................................................................... 10. 2.2.1. Person-related variables ..................................................................................... 11. 2.2.2. Environmental-related variables ........................................................................ 11. 2.3. Barriers to entrepreneurship ...................................................................................... 12. 2.4. Conceptual model ...................................................................................................... 13. THEORIES AND HYPOTHESES .................................................................................. 14 3.1. Institutional perspective ............................................................................................. 14. 3.2. Formal institutions and their relationship with entrepreneurship .............................. 16. 3.2.1. Protection of property rights .............................................................................. 17. 3.2.2. Access to finance ................................................................................................ 17. 3.2.3. Freedom from corruption ................................................................................... 18. 3.2.4. Ease of starting a business .................................................................................. 19. 3.3. Informal institutions and their relationship with entrepreneurship ........................... 19. 3.3.1 4.. METHODOLOGY AND DATA ..................................................................................... 22 4.1. Data sources ............................................................................................................... 22. 4.2. Sample ....................................................................................................................... 23. 4.3. Variables and measures ............................................................................................. 23. 4.3.1. Dependent variable ............................................................................................. 23. 4.3.2. Independent variables ......................................................................................... 24. 4.3.3. Control variables ................................................................................................ 26. 4.4 5.. Attitudes toward entrepreneurship ..................................................................... 20. Data analysis .............................................................................................................. 26. DESCRIPTIVE RESULTS .............................................................................................. 27 5.

(6) Master thesis Jan Hakvoort - 20th of June, 2014 6.. TESTING HYPOTHESES ............................................................................................... 29 6.1. Principal component analysis .................................................................................... 29. 6.1.1. Creating a new variable: access to finance ......................................................... 29. 6.1.2. Creating a new variable: social desirability ....................................................... 30. 6.2. Correlation matrix...................................................................................................... 31. 6.3. Multiple regression analysis ...................................................................................... 33. 6.3.1. Evaluation of assumptions ................................................................................. 33. 6.3.2. Regression results ............................................................................................... 34. 7.. DISCUSSION .................................................................................................................. 40. 8.. CONCLUSION ................................................................................................................ 42 8.1. Limitations and future research ................................................................................. 43. REFERENCES ......................................................................................................................... 44 APPENDICES .......................................................................................................................... 51 Appendix 1. Detecting outliers ........................................................................................ 51. Appendix 2. Descriptive results per continent ................................................................. 52. Appendix 3. Checking for linearity and homoscedasticity .............................................. 55. Appendix 4. Regression tables ........................................................................................ 57. 6.

(7) Institutional triggers that drive entrepreneurship - 20th of June, 2014. LIST OF TABLES 1:. Dimensions of institutions. 16. 2:. Overview of descriptive results of all countries. 28. 3:. Correlation matrix of principal component analysis 1. 29. 4:. Communalities and component matrix of principal component analysis 1. 30. 5:. Correlation matrix of principal component analysis 2. 30. 6:. Communalities and component matrix of principal component analysis 2. 30. 7:. Correlation matrix. 32. 8:. Regression table 1 with TEA as dependent variable. 35. 9:. Regression table 2 with TEA as dependent variable. 36. 10:. Overview of descriptive results of Africa. 52. 11:. Overview of descriptive results of Asia. 52. 12:. Overview of descriptive results of Europe. 53. 13:. Overview of descriptive results of North America. 53. 14:. Overview of descriptive results of Oceania. 54. 15:. Overview of descriptive results of South America. 54. 16:. Regression table 3 with TEA as dependent variable. 57. 17:. Regression table 4 with TEA as dependent variable. 58. LIST OF FIGURES 1:. Conceptual model. 14. 2:. Summary of hypothesized institutional triggers on entrepreneurial activity. 21. 3:. Plot of ZRESID against ZPRED. 55. 4:. Histogram. 55. 5:. Normal probability plot. 56. 7.

(8) Master thesis Jan Hakvoort - 20th of June, 2014. 1.. I TRODUCTIO Many scholars throughout the world have turned their attention to the field of. entrepreneurship (Bruyat & Julien, 2000), since it is one of the major drivers of economic growth and development (Lumpkin & Dess, 1996). As a result, scholars in entrepreneurship have developed an extensive body of research, which reflects a vital and dynamic research field over the past decade (Wiklund et al., 2010). Through this development, research in the field of entrepreneurship has become one of the most widely cited in the management discipline, with leading journals and well recognized conferences dedicated in supporting its development (Bruton et al., 2008). One of the growing interests within this research field is cross-national entrepreneurship (Hayton et al., 2002). But what is entrepreneurship and who qualifies as an entrepreneur? An entrepreneur is broadly defined as "someone who perceives an opportunity and creates an organization to pursue it" (Bygrave & Hofer, 1991:14). Entrepreneurship on the other hand is defined as "the behaviors and actions undertaken by entrepreneurs to start a new business" (Ho & Wong, 2007:188). According to Reynolds et al. (2005), the entrepreneurial activity within a country is the aggregate result of individual decisions to become an entrepreneur. However, the rates of entrepreneurial activity differs strongly across countries (Van Stel, 2005). Therefore, a key question is why some countries foster entrepreneurship better than others? The understanding of why these rates of entrepreneurship vary cross-countries is however still limited (Aronson, 1991; Rondinelli & Kasarda, 1992). Some scholars argue that the cross-country variation in entrepreneurial activity is related to differences in levels of economic development and also to diverging cultural and institutional characteristics (e.g., Blanchflower 2000; Wennekers 2006). However, researchers have till now either used a cultural or an institutional perspective to explain the cross-country differences (Nguyen et al., 2009). From a cultural perspective, Hofstede's dimensions of national culture have frequently been applied (e.g., Hofstede, 1980; Shane, 1993; Davidsson, 1997; Hayton et al., 2002; Wennekers et al., 2007). However, Hofstede's dimensions of national culture, alone, do not adequately explain the cross-country differences in entrepreneurial activity (Baumol, 1990; Denzau & North, 1994; Busenitz et al., 2000). Another stream of scholars therefore solely focused on formal institutional factors to explain the cross-country differences in entrepreneurial activity (e.g., Sternberg & Litzenberger, 2004; Ho & Wong, 2007; McMullen et al., 2008).. 8.

(9) Institutional triggers that drive entrepreneurship - 20th of June, 2014 Few scholars have, however, used both approaches to explain the cross-country differences. Scholars have therefore called for an examination of both the formal, as well as, the informal institutional influences (Oliver, 1991; Scott, 2002). These influences encompass culture, as well as, other institutional arrangements such as, rules, regulations, norms, values and assumptions that constitute appropriate behavior (Scott, 2002). Therefore, a research that is focused on the cultural, as well as, the formal institutional triggers of entrepreneurship might add to the understanding of why these entrepreneurial rates differ. This research tries to fill this gap by exploring the institutional triggers (i.e., formal and informal) that drive entrepreneurial activity cross-countries. This is done, by using cross-country data from the Global Entrepreneurship Monitor, the Index of Economic Freedom (IEF), Transparency International, the World Bank and the International Monetary Fund over the time period 2009 - 2013. The sample consists of 64 countries, representing six continents across the globe.. To address the research gap, the following research question has been formulated: What are the institutional triggers that drive entrepreneurial activity cross-countries? To answer the research question, the remainder of this paper is structured as follows: section 2 presents the background of this paper, which provides general information on entrepreneurship and explores the triggers and barriers to entrepreneurship. Section 3 presents theory on formal and informal institutions and their influence on entrepreneurship. Moreover, this section presents the hypotheses that have been formulated according to the theory. Section 4 covers the methodology and the data, which is followed by section 5, the descriptive results. Section 6 provides the results of the various analyses that have been conducted, followed by the discussion which will be covered by section 7. Lastly, this research concludes with section 8 which consists of the conclusion, limitations, and suggestions for future research.. 9.

(10) Master thesis Jan Hakvoort - 20th of June, 2014. 2.. BACKGROU D. 2.1. Entrepreneurship Entrepreneurship is derived from the French word entreprendre in the Middle Ages,. which can be translated as "between-taker or go between" (Hisrich, 1990:209). From then on, entrepreneurship has received enormous attention in both scholarly and policy circles (Yeung, 2002). This is due to the fact that entrepreneurship makes several economic and noneconomic contributions (Acs & Storey, 2004), such as reducing unemployment and poverty, encouraging economic development and generating wealth creation (Westhead et al., 2011). Entrepreneurship can therefore be seen as an important phenomenon that drives innovation and promotes economic development (e.g., Schumpeter, 1934; Reynolds, 1997). Many policymakers therefore agree that entrepreneurs, and the new businesses they establish, play a critical role in the development and well-being of their societies (Xavier et al., 2012). The economic policies that these policymakers make, influence the level of entrepreneurial activity which, in turn, increases economic growth (Gohmann, 2010). Therefore, the propensity to become an entrepreneur is influenced by several triggers and barriers that enhance or constrain entrepreneurship, which will be discussed in section 2.2 and 2.3. But first, it needs to be clear what entrepreneurship exactly is and how it is measured. In this paper, entrepreneurship is defined as the "behaviors and actions undertaken by entrepreneurs to start a new business" (Ho & Wong, 2007:188). The level of entrepreneurship within a country is measured by using the Total early-stage Entrepreneurial Activity (TEA) rate, developed by the Global Entrepreneurship Monitor (GEM). This measure captures nascent and new entrepreneurs. Nascent entrepreneurs are "adults between 18 and 64 years of age who are trying to start a new business which they will partially own" (van der Zwan et al., 2013:23). New entrepreneurs are "adults between 18 and 64 years of age who currently own and manage a business for less than 3.5 years" (van der Zwan et al., 2013:23). In this paper any person engaged in any behavior related to new business creation, regardless of the outcome, qualifies as an entrepreneur. Therefore, any form of corporate entrepreneurship (i.e., entrepreneurship within an established company) falls out of the scope of this research.. 2.2. Triggers to entrepreneurship Scholars have found several triggers that shape an individual's propensity to become. an entrepreneur. These entrepreneurial triggers can be distinguished as either person-related variables (also called supply-side variables) or environmental-related variables (also called 10.

(11) Institutional triggers that drive entrepreneurship - 20th of June, 2014 demand-side variables) (e.g., Thornton, 1999; Verheul et al., 2002; Sternberg & Litzenberger, 2004). Person-related variables refer to the supply of potential entrepreneurs in a society, whereas environmental-related variables refer to the demand-side of entrepreneurs (Stephan & Uhlaner, 2010). 2.2.1 Person-related variables Individuals are influenced by person-related variables that trigger entrepreneurship. One of these person-related variables are traits. As identified by Korunka et al. (2003), entrepreneurs have several traits that increases the chance of the individual towards enterprising behavior (e.g., risk-taking propensity, need for achievement, need for autonomy, locus of control, self confidence and creativity). Therefore, the personality of an individual can be seen as a trigger to entrepreneurship. Personality is often loosely defined in terms of the "regularities in action, feeling and thoughts that are characteristic of the individual" (Westhead et al., 2011:59). Also an individual's age, education and employment history can shape expectations and access to resources, which can be leveraged to create, identify, and exploit business opportunities (Westhead et al., 2011). Thus, motivation, intention and skills shape the propensity of an individual to become an entrepreneur (Stephan & Uhlaner, 2010). Furthermore, a person's self-efficacy has been found to be a trigger to entrepreneurial activity, which is defined as "the strength of an individual's belief that he or she will be capable of successfully performing the roles and tasks of an entrepreneur" (Wennberg et al., 2013:759). Krueger and Dickson (1994) found that people who reported high levels of perceived self-efficacy, were more likely to start a new business and to explore new opportunities. 2.2.2 Environmental-related variables Individuals can also become an entrepreneur because of environmental-related variables. People can choose to start entrepreneurship out of necessity or because they see opportunities in the marketplace (e.g., Shane & Venkataraman, 2000; McMullen et al., 2008). Individuals who start entrepreneurship because of the existence of opportunities in the marketplace are called opportunity motivated entrepreneurs (OMEs). These individuals are more or less pulled into entrepreneurship by the identification of an entrepreneurial opportunity (Braunerhjelm & Henrekson, 2013). On the other hand, individuals are also pushed into entrepreneurship out of necessity. Necessity motivated entrepreneurs (NMEs) see entrepreneurship as a last resort because all other options for work are either absent or unsatisfactory (McMullen et al., 2008). 11.

(12) Master thesis Jan Hakvoort - 20th of June, 2014 Furthermore, the formation of business clusters has been found to be positively associated with entrepreneurial activity in a region (Sternberg & Litzenberger, 2004). Lastly, entrepreneurs are both constrained and enabled by the institutions in their environment (Bruton & Ahlstrom, 2003; Scott, 2007). The institutional environment defines and limits entrepreneurial opportunities, and thus affects entrepreneurial activity (Aldrich, 1990; Gnyawali & Fogel, 1994; Hwang & Powell, 2005). Institutions specifically aimed at supporting entrepreneurship include government policies and regulations, quality of research and development, physical infrastructure and other formal support for new and high-growth firms (Levie & Autio, 2008). Efficient institutions and their policies, therefore clearly play a fundamental role in the development of entrepreneurial opportunities (e.g., Baumol, 1996; Levie. & Autio, 2008; Verheul et al., 2002). However, the understanding of which institutional factor drives entrepreneurial activity within a country is still an under researched subject.. 2.3. Barriers to entrepreneurship Next to the fact that there are triggers to entrepreneurship, there are also barriers to. become an entrepreneur. One of such a barrier is an institutional void, which has been found to have a deterrent effect on entrepreneurial activity (e.g., Khanna & Palepu, 2000; Mair & Marti, 2008). Institutional voids arise from the interplay between the existing power structure, legacy institutions, and institutional practices such as microfinance (Mair & Marti, 2008). In other words, institutional voids occur when the formal institutions of a society do not function well. According to Morck and Yeung (2003), countries that are characterized by an institutional void, have significant government corruption, poor law enforcement characteristics, an overburdened legal system, unreliable market mechanisms, impoverished education systems and an inadequate infrastructure for communication, transportation, and health. If a country lacks formal institutions and has poor enforcement characteristics, entrepreneurs become discouraged from starting new ventures (Bruton et al., 2010). Entrepreneurs operating in such environments, are faced with several challenges such as connecting to informal, trust-based sources of power and resources (Covin & Miller, 2013). Furthermore, fear of failure has been found to be a potent factor inhibiting entrepreneurial activity. Fear of failure is defined as "a self evaluative framework that influences how a person defines, orients to, and experiences failure in achievement situations" (Wennberg et al., 2013:759). Fear of failure has been found to play a role in an individuals’ achievement motivation and also their occupational aspirations, including decisions to exploit a business opportunity or not (Burnstein, 1963; Welpe et al. 2012). 12.

(13) Institutional triggers that drive entrepreneurship - 20th of June, 2014 Also capital requirements have a deterrent effect on the entrepreneurial propensity within a country. Scholars found that capital is an essential precursor to entrepreneurial activity (e.g., Schumpeter, 1934; Kauermann et al., 2005) since it influences the ability of firms to enter into markets and, moreover, it influences their performance post-entry (Ho & Wong, 2007). Capital requirements can therefore be seen as a deterrent effect on entrepreneurial activity since certain complicated production processes need large amounts of capital (Bain, 1959; Koch, 1974). Moreover, entrepreneurs that have limited access to capital are deterred to start their own enterprise (Van Auken, 1999). Lastly, regulation pertaining to entry is recognized as a barrier to entry, since excessive regulations make it extremely difficult for an entrepreneur to open and operate a business (Heckelman, 2000). Porter (1980) agrees with this view and argues that government regulation may impose entry barriers to new entrants. The more that government imposes regulations on business, the harder it is for entrepreneurs to create them (McMullen et al., 2008). Regulatory and procedural requirements, such as excessive licensing requirements, registration difficulties and bureaucratic corruption, entail business costs that potential entrants must incur, in terms of financial outlay (e.g., Ho & Wong, 2007; McMullen et al., 2008).. 2.4. Conceptual model The conceptual model of this paper can be found in figure 1, which visualizes the. triggers and barriers to entrepreneurial activity. As mentioned earlier on, institutional voids, the fear of failure, capital requirements and regulatory business costs have been identified as barriers to entrepreneurial activity within a country. On the other hand, people are triggered to become an entrepreneur because of person-related variables (motivation, intentions, skills, personality and self-efficacy) and environmental-related variables (opportunities, necessity, clusters and institutions). Since this paper is specifically focused on identifying which institutional triggers (i.e. formal and informal) drive entrepreneurial activity, all other aforementioned triggers and barriers to entrepreneurial activity are left aside. However, as can be seen in the methodology section, a set of variables will be used as control variables.. 13.

(14) Master thesis Jan Hakvoort - 20th of June, 2014 Figure 1: Conceptual model. Source: Author. 3.. THEORIES A D HYPOTHESES. 3.1. Institutional perspective According to Bruton et al. (2010), the application of institutional theory is a popular. and helpful theoretical foundation to research in the field of entrepreneurship. The application of this theory plays a major role in helping to explain the drivers of entrepreneurial success, apart from industry conditions and organizational resources (Ahlstrom & Bruton, 2002; Peng, 2006). Historically, the resource-based view of the firm (Barney, 1991) has been one of the key theories in entrepreneurship because access to resources is central to the success of a new venture (Bhide, 2000). While resources are certainly vital, the resource-based view can be criticized for ignoring the institutional environment (Peng et al., 2008). According to Baumol et al. (2009) issues such as culture, legal environment, tradition and history in an industry all can impact an industry and, in turn, entrepreneurial success. This means that economic activities, such as entrepreneurship, cannot be analyzed when ignoring the formal and informal institutions (e.g., Baumol, 1990; Denzau & North, 1994). Institutional theory views individuals and organizations as embedded in institutional arrangements (Busenitz et al., 2000). The main idea behind the institutional theory is that it is concerned with how various groups and organizations better secure their positions and. 14.

(15) Institutional triggers that drive entrepreneurship - 20th of June, 2014 legitimacy by conforming to the rules and norms of the institutional environment (Scott, 2007). But what are institutions exactly? According to Scott (2002), institutions have an impact on the behavior of individuals and organizations in subtle but pervasive ways (Scott, 2002). Institutions signal which behavior is acceptable and determine which norms and values are socialized into a given society (Peng & Heath, 1996; Ahlstrom & Bruton, 2002). This is derived from rules such as regulatory structures, governmental agencies, laws, courts, professions, and other societal and cultural practices that exert conformance pressures (DiMaggio & Powell, 1991). Institutions therefore signal which behavior is appropriate and thus also render other actions unacceptable or even beyond consideration (DiMaggio & Powell, 1991). This means that institutions have an impact on the cognitive and ethical considerations that shape human judgment and behavior (Scott, 1995). Moreover, they affect organizational behavior by constraining and defining which actions are acceptable and supportable both within and between organizations (Aldrich & Fiol, 1994). The institutional forces are identified by Scott (1995:33), who sees institutions as a set of three pillars and defines it as follows: "institutions consist of cognitive, normative, and regulative structures and activities that provide stability and meaning to social behavior. Institutions are transported by various carries (cultures, structures, and routines) and they operate at multiple levels of jurisdiction." According to North (1990:3) institutions are "the rules of the game" under which individuals and organizations act and compete. They are the "humanly devised constraints that structure human interaction" (North, 1990:3). North divides institutions into formal and informal ones, which can be seen as complementary to Scott's pillars (see table 1). North's formal institution is comparable with Scott's regulative pillar and is measurable (e.g., rules, laws and constitutions), whereas North's informal institution is comparable with the normative/cognitive pillar, which is in the heads of people (e.g., norms of behavior, conventions, self-imposed codes of conduct). This means that culture is a "substratum of institutional arrangements" (Hofstede et al., 2002:800) and can therefore be seen as a part of informal institutions that "underpin formal institutions" (Redding, 2005:123). Entrepreneurs will adapt their activities and strategies molded to fit the opportunities and limitations provided through the formal and informal institutional framework (Aidis et al., 2008). So why do entrepreneurship rates differ cross-countries? According to institutional theory, new venture creation rates will be higher in societies where new ventures have more legitimacy (Nguyen et al., 2009). In other words, entrepreneurship will flourish in countries in which government regulations and societal norms support new ventures, and in which knowledge of creating new ventures is widely available (Busenitz et al., 2000). Taken that in 15.

(16) Master thesis Jan Hakvoort - 20th of June, 2014 mind, this paper argues that cross-country differences in institutions are seen as the root cause of differing entrepreneurship rates. Or as Braunerhjelm and Henrekson (2013:108) put it: "cross-country differences in long-term economic performance are ultimately caused by differences in the rules of the game in society or the broader institutional setup." Accordingly, this paper uses the following overarching proposition, from which hypotheses are formulated: Proposition 1: Institutional triggers will be positively associated with entrepreneurial activity. Table 1: Dimensions of institutions. Source: Peng et al. (2009). 3.2. Formal institutions and their relationship with entrepreneurship Formal institutions signal which behavior by individuals and organizations is. appropriate (Peng & Heath, 1996). Governments or other authoritative body can regulate individual and organizational action by giving incentives and/or imposing sanctions (Scott, 2001). According to Tonoyan et al. (2010:850) "formal institutions are those written or formally accepted rules and regulations which have been implemented to make up the economic and legal set-up of a given country." Formal institutions include laws, regulations, and government policies that protect private property, provide support for opening new businesses, reduce the risks for individuals creating new businesses, and create an environment in which resources are widely accessible (e.g., access to finance) (Nguyen et al., 2009). Lastly, corruption may have a serious negative impact on the development of entrepreneurship within a country (Akimova, 2001). Conversely, freedom from corruption is therefore assumed to have a positive effect on entrepreneurial activity. An absence of effective institutions that protect private property, fair competition, and financial discipline, makes the risk and costs of doing business excessively high (Broadman et al., 2004) and therefore are barriers to entrepreneurship. Entrepreneurial activity within a country is therefore directly related to the country's regulations and policies governing the allocation of rewards (Baumol et al., 2009). Thus, the 16.

(17) Institutional triggers that drive entrepreneurship - 20th of June, 2014 protection of property rights, access to finance, freedom from corruption and business freedom all have an impact on the level of entrepreneurship within a country and will therefore be examined further in the text. 3.2.1 Protection of property rights Vital for the functioning of a free-market economy is the ability to accumulate private property (de Soto, 2000). Property rights refer to the degree to which government creates the right to private property and enforces the laws written to protect those rights (Beach & O’Driscoll, 2003). Firms want to protect their valuable intellectual property from leaking to competitors by writing. and executing a reliable contract (Oxley, 1999). This requires. adequate specification of property rights, monitoring, and enforcement of contractual terms (Oxley, 1999). When the protection of property rights is absent in a country, individuals are less likely to invest in improving their assets and are required to allocate effort away from productive activities in the effort to secure legally unprotected property (de Soto, 2000). Therefore, a problem is the enforcement of a contract when a violation is detected. The enforcement of intellectual property rights relies on the general enforcement powers of the court and can therefore vary considerably among countries (Oxley, 1999). The duration of a patent, a widely used instrument for the protection of intellectual property, varies considerably between countries. Kondo (1994) has found that the duration of patent protection ranges from 5 years in several Latin American countries to close to 20 years in most European countries. When governments fail to ensure that private property is protected, individuals have less confidence that they need to initiate commercial activities, save income, and make long-term plans (Beach & O’Driscoll. 2003). Hence, the following hypothesis has been formulated: Hypothesis 1: The protection of property rights within a country will be positively associated with entrepreneurial activity. 3.2.2 Access to finance For starting and running a business, entrepreneurs need financial capital, which is needed to purchase or rent the premises, finance market research and advertising, invest in equipment, and purchase raw materials (Wennekers et al., 2007). Entrepreneurs can use selffinancing (e.g., savings, gifts) for their business start-ups and/or third party financiers such as informal investors, mortgage loans and venture capital (Bygrave & Hunt 2005).. 17.

(18) Master thesis Jan Hakvoort - 20th of June, 2014 Therefore, financial capital is an important determinant of firm formation, because it not only influences the ability of firms to enter into markets, but also their performance postentry (Ho & Wong, 2007). Moreover, a well-functioning formal financial system within a country, which is determined by the financial freedom, ensures the availability of diversified savings, credit, payment, and investment services to individuals (Miller et al., 2014). This means that when individuals have better access to capital within a country, this raises the likelihood of new business startups (Van Gelderen et al., 2005). Hence, the following hypothesis has been formulated: Hypothesis 2: Access to financial capital within a country will be positively associated with entrepreneurial activity. 3.2.3 Freedom from corruption In essence, corruption involves the behavior that violates the trust individuals and organizations have in public officials (Anokhin & Schulze, 2009). However, the success of new business ventures is mostly determined by relying on these public officials with whom entrepreneurs only have indirect contact (Anokhin & Schulze, 2009). Entrepreneurs can therefore try to influence the actions of public officials themselves by paying bribes. State capture is a common form of bribing where entrepreneurs try to influence the formulation of laws and other government policies to their own advantage through illicit or non-transparent means (Fries et al., 2003). According to McMullen et al. (2008:884) "corruption weakens the rule of law, gradually replacing it with the rule of man." The rule of law is predictable and facilitates an environment in which entrepreneurial opportunities arise. On the other hand, the rule of man suffers from poor enforcement of laws, rights, and contracts. As a result, the rule of man undermines the stability and reliability of these institutions, which individuals need to start their own business (McMullen et al., 2008). As corruption increases, bureaucrats are more likely to increase regulations and impose licensing restrictions (de Soto, 1989). Corruption may therefore have a serious negative impact on the development of entrepreneurship within a country (Akimova, 2001). Conversely, freedom from corruption is assumed to have a positive effect on entrepreneurial activity. Hence, the following hypothesis has been formulated: Hypothesis 3: Freedom from corruption within a country will be positively associated with entrepreneurial activity.. 18.

(19) Institutional triggers that drive entrepreneurship - 20th of June, 2014 3.2.4 Ease of starting a business When governments maintain policies in which entrepreneurs are forced to comply with many rules and procedural requirements, are expected to report to an array of institutions, and have to spend substantial time and money in fulfilling documentation requirements, then this will discourage them from starting their own enterprise (de Soto, 2000). These regulatory and procedural requirements entail business costs that potential entrants must incur and which may deter potential entrepreneurs from starting their own enterprise (Ho & Wong, 2007). For example, in Russia it will take an entrepreneur 97 days at significant cost and a navigation through a sea of red tape to start a new business (de Soto, 2000). Moreover, even when the enterprise is founded, government regulation may intensify in some countries (Beach & O’Driscoll, 2003). A more business-favorable institutional environment, however, will ease such barriers and encourage entrepreneurs to start their own business (Baumol et al., 2009). Therefore, licensing a new company can involve completion of a single form and take as little as a few hours in some countries (e.g., Hong Kong) (World Bank, 2005). Moreover, it has been shown by Kshetri & Dholakia (2011), that reducing the time taken to start a business by ten days, increases a country's GDP by a 0.4 percentage point. Thus excessive regulation in terms of licensing requirements, registration difficulties, bureaucratic corruption are expected to discourage entrepreneurs from starting their own business. Conversely, the following hypothesis has been formulated: Hypothesis 4: The ease of opening and operating a business within a country will be positively associated with entrepreneurial activity.. 3.3. Informal institutions and their relationship with entrepreneurship Informal institutions are "traditions, customs, societal norms, shared mental models,. unwritten codes of conduct, ideologies, and templates that have never been consciously designed but are still in everyone’s interest to keep" (North, 1994:360). Informal institutions are similar to Scott's cognitive and normative pillar. The cognitive pillar consists of the knowledge and skills possessed by the people in a country pertaining to establishing and operating a new business (Busenitz et al., 2000). The cognitive pillar is increasingly important to entrepreneurship research in terms of how societies accept entrepreneurs, inculcate values, and even create a cultural milieu whereby entrepreneurship is accepted and encouraged (Bosma, et al., 2009). The normative pillar, on the other hand, measures the degree to which a country's residents admire entrepreneurial activity and value creative and innovative thinking 19.

(20) Master thesis Jan Hakvoort - 20th of June, 2014 (Busenitz et al., 2000). A country's culture, beliefs, norms, and values influence the entrepreneurial orientation of its residents (e.g., Busenitz & Lau, 1996; Knight, 1997; Tiessen, 1997). When measuring the influence of culture on entrepreneurial activity, the values approach is by far the dominant approach (Shteynberg et al., 2009). However, this approach presents contradictory results on which factors drive entrepreneurial activity. Some of these studies find individualism, low power distance and low uncertainty avoidance to be associated with higher entrepreneurship rates, whereas others report the opposite pattern (e.g., Hayton et al., 2002; Hofstede et al., 2004; Wennekers et al., 2007). Therefore, this paper favors measuring culture based on the attitudes toward entrepreneurship within a country. 3.3.1 Attitudes toward entrepreneurship Shapero and Sokol (1982) argue that levels of entrepreneurship could be explained by culture, because cultures carry different beliefs about the desirability and feasibility of starting a new enterprise. This finding is in line with Amorós and Bosma (2014), who argue that individual's evolving attitudes and perceptions toward entrepreneurship could affect the level of entrepreneurship. The overarching theory for this is the social desirability of entrepreneurship within a country. The social desirability of entrepreneurship "are the subjective norms or commonly held perceptions regarding the status and rewards of entrepreneurship in a given population" (Stephan & Uhlaner, 2010:1349). Culture influences the level of entrepreneurship within a country, through the social desirability of entrepreneurship (e.g., Levie & Autio, 2008; Reynolds et al., 2004). But how is social desirability measured? Social desirability can be measured according to three attitudinal measures, which assess societal impressions about entrepreneurship as a desirable career choice and whether entrepreneurs are afforded high status and receive positive media attention (Amorós & Bosma, 2014). Measuring such attitudes toward entrepreneurship is important, because it contains information about the image of entrepreneurship within a country (Van der Zwan et al., 2012). Positive views on these measures may indicate 'legitimation' or 'moral approval' of entrepreneurship within a culture, which may influence the decision to engage in entrepreneurship (Freytag and Thurik, 2007). Hence, the following hypotheses have been formulated: Hypothesis 5a: Entrepreneurship seen as a desirable career choice within a country will be positively associated with entrepreneurial activity.. 20.

(21) Institutional triggers that drive entrepreneurship - 20th of June, 2014 Hypothesis 5b: High status of entrepreneurship within a country will be positively associated with entrepreneurial activity. Hypothesis 5c: Positive media attention of entrepreneurship within a country will be positively associated with entrepreneurial activity. Hypothesis 5d: Social desirability of entrepreneurship within a country will be positively associated with entrepreneurial activity Figure 2: Summary of hypothesized institutional triggers on entrepreneurial activity. Source: Author. 21.

(22) Master thesis Jan Hakvoort - 20th of June, 2014. 4.. METHODOLOGY A D DATA. 4.1. Data sources To test the hypotheses, data is extracted from various databases. The main data source. in this paper is the Global Entrepreneurship Monitor (GEM) database, which provides internationally comparable data on entrepreneurial activity, aspirations and attitudes of individuals (GEM, 2014a). The main measure of the GEM database is the individual involvement in venture creation (GEM, 2014a). The database differentiates itself from other databases, which mostly record firm-level data. The research program started in 1999 and the first study covered 10 countries. Since then, the GEM program has expanded every year. Nowadays, more than 100 countries are involved in the research program and the GEM program is therefore considered the largest study of entrepreneurial activity in the world (Van der Zwan et al., 2012). The data from the GEM database comes from the adult population survey, which tracks the entrepreneurial activity, and attitudes of individuals (Amorós & Bosma, 2014). The adult population survey is a comprehensive questionnaire, administered to a minimum of 2000 adults in each GEM country (Amorós & Bosma, 2014). The GEM database provides aggregated country scores for several variables from 2001 till 2013. Another database that will be used in this paper is the Index of Economic Freedom (IEF) from the Heritage Foundation and the Wall Street Journal. The IEF database contains data of economic freedom covering 186 countries. According to the IEF economic freedom is defined as: "the fundamental right of every human to control his or her own labor and property (IEF, 2014)." Economic freedom is measured based on 10 quantitative and qualitative factors, which are property rights, freedom from corruption, fiscal freedom, government spending, business freedom, labor freedom, monetary freedom, trade freedom, investment freedom, and financial freedom (IEF, 2014). The IEF database provides country scores for every factor from 1995 - 2014. The IEF and its data is widely used among researchers (e.g., McMullen et al., 2008; Gohmann, 2010) and will therefore also be used in this research. Furthermore, the Transparency International database will be used for extracting data on corruption. The measure used is the Corruption Perception Index (CPI) which ranks countries based on how corrupt their public sector is perceived to be (Transparency International, 2014). The CPI is a widely used measure to assess the level of corruption within a country and previous research has shown it to be a reliable proxy (e.g., Friedman et al., 2000; Gohmann, 2010). 22.

(23) Institutional triggers that drive entrepreneurship - 20th of June, 2014 Also, the database of the World Bank will be used, which is one of the most comprehensive databases of national statistics about people, the economy, the environment and states and markets (WorldBank, 2014). This database provides data of more than 200 countries and 1,200 indicators from 1980 - 2013. Lastly, the database of the International Monetary Fund (IMF) will be used which provides data for several fiscal indicators for 189 countries over a time period from 1980 2013, including projections for the next two years (IMF, 2014).. 4.2. Sample The IEF, Transparency International, the World Bank and the IMF all provide data for. more than 180 countries, whereas the GEM database is limited to around 100 countries, each of which was included in the other aforementioned databases. After eliminating outliers (see appendix 1) and countries with missing data on one or more variables, a total dataset of 64 countries remained. The sample covers six continents: Africa (8 countries), Asia (19 countries), Europe (20 countries), North America (8 countries), Oceania (1 country) and South America (8 countries). The sample consists of the following countries: Argentina, Australia, Bangladesh, Barbados, Belgium, Botswana, Brazil, Chile, China, Colombia, Costa Rica, Croatia, Ecuador, Egypt, El Salvador, Estonia, Finland, France, Germany, Ghana, Greece, Iceland, India, Indonesia, Iran, Ireland, Israel, Italy, Jamaica, Japan, Jordan, Luxembourg, Malaysia, Mexico, Morocco, Namibia, Pakistan, Panama, Peru, Portugal, Romania, Russia, Saudi Arabia, Serbia, Singapore, Slovak Republic, Slovenia, South Africa, South Korea, Spain, Switzerland, Thailand, The Netherlands, The Philippines, Trinidad & Tobago, Tunisia, Turkey, Uganda, United Arab Emirates, United Kingdom, United States of America, Uruguay, Venezuela and Vietnam.. 4.3. Variables and measures All variables used in this paper have been collected from the aforementioned. databases. The country scores are averaged from the years 2009 - 2013 to increase the reliability of the scores. 4.3.1 Dependent variable Entrepreneurial activity. National entrepreneurial activity is the dependent variable in this paper. The Total early-stage Entrepreneurial Activity (TEA) is a widely used measure for entrepreneurial activity and previous research has shown it to be a reliable proxy (e.g., Ho & Wong, 2007; McMullen et al., 2008; Stephan & Uhlaner, 2010; Wennberg et al., 2013). The 23.

(24) Master thesis Jan Hakvoort - 20th of June, 2014 TEA is defined as the percentage of the adult population (18 - 64 years of age) that currently are in the process of starting a business and those who are currently owning and managing a business for less than 3.5 years (e.g., Van der Zwan et al., 2013; Amorós & Bosma, 2014). People that currently are in the process of starting a business are called nascent entrepreneurs and those who are currently owning and managing a business for less than 3.5 years are called new entrepreneurs. This means that the total entrepreneurial activity within a country is measured by the combination of nascent and new entrepreneurs (Van der Zwan et al., 2013). The GEM database is used to collect TEA rates from multiple countries and multiple years. 4.3.2 Independent variables Protection of property rights. The protection of property rights is measured using data from the IEF database. This database measures the degree to which a country's laws protect private property rights, the extent to which those laws are respected and the likelihood that private property will be expropriated by the state (Miller et al., 2014). Moreover, the IEF also assesses the independence of the judiciary and the enforcement characteristics of a country so that individuals and businesses are able to enforce contracts (Miller et al., 2014). A country's protection of property rights is graded by using a scale from 0 to 100. A low score is given to countries which fail to protect private property or countries with less independent judiciary. On the other hand, high scores are given to countries that have effective legal protection of property rights (Miller et al., 2014). Access to finance. Since there is not one single measure that directly measures the access to finance within a country, a combination of three variables in this paper is used. The first variable is the financial freedom which comes from the IEF database. The financial freedom rate measures the efficiency of banks, as well as, the independence of government control and interference in the financial sector (Miller et al., 2014). Moreover, the financial freedom rate assesses the openness to foreign competition and the extent of financial and capital market development (Miller et al., 2014). The financial freedom of a country is graded by using a scale from 0 to 100. A score of 0 is given to countries in which supervision and regulation are designed to prevent private financial institutions, whereas a score of 100 represents a country that has negligible government interference (Miller et al., 2014). The second variable is the private credit by deposit money banks to GDP (%) which is extracted from the Global Financial Development database from the World Bank. This variable measures the financial resources provided to the private sector by domestic money banks as a share of GDP (World Bank, 2014). 24.

(25) Institutional triggers that drive entrepreneurship - 20th of June, 2014 Lastly, the stock market capitalization to GDP (%) is used. This variable is also extracted from the Global Financial Development database from the World Bank and measures the total value of all listed shares in a stock market as a percentage of GDP (World Bank, 2014). To combine these three variables into one new (access to finance) variable, a principal component analysis will be performed. Freedom from corruption. Since corruption increases insecurity and uncertainty into economic relations (Miller et al., 2014), it may have serious negative impacts on the entrepreneurial activity rate within a country (Akimova, 2001). Conversely, freedom from corruption is expected to be positively associated with entrepreneurial activity. Freedom from corruption is measured by using data from the Transparency International database. Transparency International assesses how corrupt a country is perceived to be, by using the CPI measure (Transparency International, 2014). In total, Transparency International provides corruption rates for 177 countries. Corruption levels from the Transparency International database are provided on a scale from 0 (highly corrupt) to 100 (very little corruption) (Miller et al., 2014). Business freedom. Business freedom is measured using data from the IEF database and measures the ease of starting, operating, and closing a business (Miller et al., 2014). Business freedom indicates the overall efficiency of government regulation of business (Miller et al., 2014). Business freedom is graded by using a scale from 0 to 100, in which 100 indicates a free business environment (Miller et al., 2014). The scores are based on 10 factors, which are all weighted equally (e.g., number of procedures to start a business, number of days to start a business, and number of procedures to obtain a license). Social desirability. Social desirability is measured by using a combination of three variables (i.e., entrepreneurship seen as a desirable career choice, entrepreneurship is given high status, and entrepreneurs receive positive media attention). These three variables are extracted from the GEM database which defines the variables as follows: entrepreneurs seen as a desirable career choice is the percentage of the adult population (18 - 64 years of age) who agree with the statement that in their country, most people consider starting a business as a desirable career choice (GEM, 2014b). Secondly, entrepreneurship is given high status is defined as the percentage of the adult population (18 - 64 years of age) who agree with the statement that in their country, successful entrepreneurs receive high status (GEM, 2014b). Lastly, the GEM defines entrepreneurs receive positive media attention as the percentage of 25.

(26) Master thesis Jan Hakvoort - 20th of June, 2014 the adult population (18 - 64 years of age) who agree with the statement that in their country, often stories about successful new businesses are seen in the public media (GEM, 2014b). To combine these three variables into one new (social desirability) variable, a principal component analysis will be performed. 4.3.3 Control variables Gross Domestic Product per capita PPP adjusted. Gross Domestic Product per capita has been found to negatively influence entrepreneurship rates (e.g., Wennekers et al., 2005; Stephan & Uhlaner, 2010) and will therefore be used as a control variable. This research uses the GDP per capita adjusted for Purchasing-Power-Parity (GDP per capita PPP adjusted), which is extracted from the IMF database. This means that an international Dollar has the same purchasing power over GDP as the U.S. Dollar has in the United States (World Bank, 2014). Average years of schooling (15+). Education has been found to be associated with entrepreneurial activity (Vinogradov & Kolvereid, 2007) and will therefore be used as a control variable. Data on education is extracted from the World Bank education database by using the average years of total schooling 15+. Total years of schooling is the average years of education completed (primary, secondary and tertiary) among people aged 15 and over (World Bank, 2014). Unemployment rate. Wilderman et al. (1998) argue that less developed countries exhibit higher rates of unemployment compared to developed countries. Furthermore, countries that maintain high levels of unemployment have higher rates of self-employment (Wilderman et al., 1998). Therefore, the unemployment rate will be used as a control variable in this research. Data on unemployment is extracted from the World Development database from the World Bank by using the total unemployment rate as a percentage of the total labor force.. 4.4. Data analysis The main statistical method used in this research to test the hypotheses is the multiple. regression analysis, which is performed by using SPSS software. A multiple regression analysis is used to estimate the relation between a set of individual variables on a dependent variable. In this case multiple regression analyses are performed on a dataset of 64 countries in order to predict whether the independent variables would lead to higher entrepreneurial activity. In addition, GDP per capita PPP adjusted, average years of schooling (15+) and 26.

(27) Institutional triggers that drive entrepreneurship - 20th of June, 2014 unemployment rates are used as control variables, which might influence the dependent variable (i.e., entrepreneurial activity). Before these multiple regression are performed, two principal component analyses are conducted. A principal component analyses has the benefit that it can combine several variables into one new variable. Also a correlation matrix is conducted in order to test the correlations between the variables independently. Before regressions are performed, several assumptions are checked, which are: multicollinearity, non-linearity and heteroscedasticity. The multiple regression analysis is tested on multicollinearity by using the correlation matrix. Moreover, the data is tested on non-linearity and heteroscedasticity by using histograms, normal probability plots of residuals and a plot in which ZRESID (standardized residuals) will be plotted against ZPRED (standardized predicted values of the dependent variable based on the model).. 5.. DESCRIPTIVE RESULTS The sample of this research consists of 64 countries and covers six continents: Africa. (8 countries, 12.5%), Asia (19 countries, 29.69%), Europe (20 countries, 31.25%), North America (8 countries, 12.5%), Oceania (1 country, 1.56%) and South America (8 countries, 12.5%). On average the TEA is 12.15% (see table 2), which is more or less the same as the TEA in Barbados (12.55%). The lowest TEA rate (3.45%) can be found in Europe (Italy) and the highest TEA rate (32.1%) in Africa (Ghana) (see appendix 2 tables 10-15). South America is however the continent with the highest average TEA, which is 19.23%. Furthermore, when examining the formal institutional factors, it becomes clear that the average score of the protection of property rights, freedom from corruption, business freedom and access to finance is 54.58, 51.65, 72.74 and 0.00 respectively. These scores are more or less the same as the scores in Malaysia (52), Turkey (49.5), Slovak Republic (72.28) and Trinidad & Tobago (-0.015) respectively. Moreover, when examining the attitudinal factors, it becomes clear that Africa is the continent with the highest average score on all three measures, which is 80.2 for career choice, 84.22 for high status and 75.83 for media attention (see appendix 2 table 10, Africa). The average scores for all countries taken together is 67.14 for career choice, 72.77 for high status and 62.19 for media attention. These average scores are more or less the same as the scores in Portugal (67), South Africa (72.6) and El Salvador (62) respectively. The average level of the social desirability of entrepreneurship, however, is the highest in South America 27.

(28) Master thesis Jan Hakvoort - 20th of June, 2014 (0.48). The lowest social desirability score can be found in Asia (-2.18 for Japan), whereas the highest score can be found in Africa (1.94 for Ghana). Lastly, when examining the descriptive results of the control variables, the following has been found: first of all, Africa is the continent with the lowest GDP per capita PPP adjusted ($1,389.13 in Uganda), whereas Europe has the highest level ($77,246.33 in Luxembourg). In this sample, the average GDP per capita PPP adjusted is $20,470.29 which is more or less the same as what the Estonians on average earn ($20,442.05). Secondly, the average years of total schooling (15+) is the lowest in Africa, which is 5.01 years in Morocco. North America has the highest score, which is 13.09 years in the United States of America. Thirdly, the unemployment rate goes as high as 24.53% (South Africa) to as low as 0.97% (Thailand). The average unemployment rate in this sample is 8.53% which is more or less the same as the unemployment rate in Italy (8.82%). Table 2 Overview of descriptive results of all countries Variables. Minimum. Maximum. Mean. Standard Deviation. TEA. 3.45. 32.1. 12.15. 6.7. Social desirability. -2.18. 1.94. 0E-7. 1.0. Career choice. 28.6. 89.6. 67.14. 12.81. High status. 46.2. 100. 72.77. 10.34. Media attention. 32.8. 87. 62.19. 13.26. Protection of property rights. 4. 91. 54.58. 24.15. Freedom from corruption. 19.5. 89.5. 51.65. 18.99. Business freedom. 40.08. 97.8. 72.74. 13.61. Access to finance. -1.9. 2.48. 0E-7. 1.0. GDP per capita PPP. 1,389.13. 77,246.33. 20,470.29. 15,348.38. Average years of schooling. 5.01. 13.09. 9.1. 1.91. .97. 24.53. 8.53. 4.94. (15+) Unemployment N = 64. 28.

(29) Institutional triggers that drive entrepreneurship - 20th of June, 2014. 6.. TESTI G HYPOTHESES. 6.1. Principal component analysis Most of the time when conducting research, researchers try to measure things that. cannot directly be measured, the so called latent variables. This is the case in this research, where the relationship between access to finance and entrepreneurial activity is tested. Since there is not one single measure that directly measures the access to finance within a country, a combination of three variables in this paper is used. In order to combine three variables into one new variable, a principal component analysis needs to be conducted. The principal component analysis is a statistical technique for identifying groups or clusters of variables (Field, 2009). A principal component analysis decomposes the original data (or variables) into a set of linear variates (Field, 2009). 6.1.1 Creating a new variable: access to finance A principal component analysis was conducted on three variables (financial freedom, private credit by deposit money banks as a percentage of GDP and stock market capitalization as a percentage of GDP), N = 64. An oblique rotation technique (oblimin direct) was used since the variables are allowed to correlate. The Kaiser-Meyer-Olkin measure (KMO) verified the sampling adequacy for the analysis, KMO = 0.675 (mediocre according to Field, 2009). Moreover, Bartlett's test of sphericity χ2 (3) = 37.443, p < .001, indicates that correlations between the variables are sufficiently large for a principal component analysis. Furthermore, the correlation matrix indicates that all three variables are highly significant and correlate positively with each other p < .001 (see table 3). Only one component was extracted and had an eigenvalue over Kaiser's criterion of 1 (1.935) which explains 64.52% of total variance explained. Since only one component was extracted, the solutions cannot be rotated. Table 3 Correlation matrix of principal component analysis 1. Correlation Financial freedom Private credit by money banks to GDP (%) Stock market capitalization to GDP (%). Financial freedom. Private credit by deposit money banks to GDP (%). Stock market capitalization to GDP (%). 1.000 .523***. .523*** 1.000. .450*** .429***. .450***. .429***. 1.000. *** p < 0.01 (one-tailed), determinant = .511. 29.

(30) Master thesis Jan Hakvoort - 20th of June, 2014 Table 4 Communalities and component matrix* of principal component analysis 1 Communalities Initial 1.000 1.000. Financial freedom Private credit by deposit money banks to GDP (%) Stock market 1.000 capitalization to GDP (%) Extraction method: Principal component analysis * One component extracted. Component Extraction .681 .662. .825 .814. .593. .770. 6.1.2 Creating a new variable: social desirability Social desirability can be measured according to three attitudinal measures, which assess societal impressions about entrepreneurship as a desirable career choice and whether entrepreneurs are afforded high status and receive positive media attention (GEM, 2013). Therefore, a principal component analysis was conducted on these three variables, N = 64. An oblique rotation technique (oblimin direct) was used since the variables are allowed to correlate. The KMO verified the sampling adequacy for the analysis, KMO = 0.688 (mediocre according to Field, 2009). Moreover, Bartlett's test of sphericity χ2 (3) = 40.953, p < .001, indicates that correlations between the variables are sufficiently large for a principal component analysis. Furthermore, the correlation matrix indicates that all three variables are highly significant and correlate positively with each other p < .001 (see table 5). Only one component was extracted and had an eigenvalue over Kaiser's criterion of 1 (1.982) which explains 66.1% of total variance explained. Since only one component was extracted, the solutions cannot be rotated. Table 5 Correlation Matrix of principal component analysis 2 Career choice Correlation Career choice 1.000 High status .514*** Media attention .494*** *** p < 0.01 (one-tailed), determinant = .511. High status. Media attention. .514*** 1.000 .465***. .494*** .465*** 1.000. Table 6 Communalities and component matrix* of principal component analysis 2 Communalities Initial Career choice 1.000 High status 1.000 Media attention 1.000 Extraction method: Principal component analysis * One component extracted. Component Extraction .683 .659 .640. .827 .812 .800. 30.

(31) Institutional triggers that drive entrepreneurship - 20th of June, 2014. 6.2. Correlation matrix Prior to performing the regressions, a correlation matrix is carried out. Table 7. presents the correlation matrix, which includes all variables used in this research and their individual correlations with each other. In this research the Pearson's correlation coefficient (also called Pearson's r) is used. Pearson's r measures the extent to which a linear relationship exists between two variables (Baarda, 2004). The values for r range from -1 to +1, where r = 1 indicates a perfect linear positive relationship between the variables. Conversely, r = -1 indicates that the two variables are perfectly negatively correlated with each other. A Pearson's r of zero indicates no linear relationship at all (Field, 2009). Since every variable is perfectly correlated with itself, the correlation matrix presents r = 1 along the diagonal. Since this research uses directional hypotheses (i.e., independent variable positively associated with dependent variable), onetailed tests are performed. When examining the correlation matrix, it becomes clear that some of the variables correlate highly with each other (i.e., r > .7). This is a sign of multicollinearity, which might bias the outcome of the regressions. The variables that might be multicollinear, and thus problematic for regressions, are presented in bold in the correlation matrix (see table 7).. 31.

(32) Master thesis Jan Hakvoort - 20th of June, 2014 Table 7 Correlation matrix of all variables Mean. S.D.. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 1. TEA. 12.146. 6.697. 1. 2. Social desirability 3. Career choice. 0.000. 1.000. 1. 67.144. 12.815. .623** .000 .574** .000. .827** .000. 1. 4. High status. 72.775. 10.338. .378** .001. .812** .000. .514** .000. 1. 5. Media attention. 62.189. 13.261. 54.58. 24.178. .800** .000 -.464** .000. .494** .000 -.541** .000. .465** .000 -.284* .011. 1. 6. Property rights. .589** .000 -.388** .001. -.302** .008. 1. 7. Corruption freedom. 51.648. 18.997. -.394** .001. -.476** .000. -.591** .000. -.317** .005. -.247* .024. .931** .000. 1. 8. Business freedom. 72.736. 13.614. -.489** .000. -.369** .001. -.464** .000. -.183 .074. -.249* .024. .712** .000. .673** .000. 1. 9. Access to finance. 0.000. 1.000. -.389** .001. -.370** .001. -.430** .000. -.239* .028. -.229* .035. .781** .000. .708** .000. .573** .000. 1. 10. GDPPC PPP. 20470.285. 15348.384. -.517** .000. -.567** .000. -.645** .000. -.329** .004. -.404** .000. .762** .000. .801** .000. .654** .000. .707** .000. 1. 11. Schooling. 9.098. 1.911. -.484** .000. -.574** .000. -.531** .000. -.471** .000. -.395** .001. .619** .000. .605** .000. .553** .000. .505** .000. .671 ** .000. 1. 12. Unemployment. 8.526. 4.944. -.147 .123. -.053 .339. .081 .263. -.084 .254. -.129 .156. .002 .493. -.030 .408. .095 .228. .027 .415. -.106 .201. .062 .314. 12. 1. ** Correlation is significant at the 0.01 level (one-tailed), * Correlation is significant at the 0.05 level (one-tailed), N = 64.. 32.

(33) Institutional triggers that drive entrepreneurship - 20th of June, 2014. 6.3. Multiple regression analysis. 6.3.1 Evaluation of assumptions Lastly, multiple regression analyses are performed, which test the correlation between different independent variables simultaneously. By performing multiple regression analyses, it is possible to calculate to what extend a combination of independent variables predict the dependent variable (Baarda, 2004). In this research, a total of 42 unique multiple regression analyses are performed. But before conclusions can be drawn based on multiple regressions analysis, several assumptions must be met (Berry, 1993). These assumptions are: no perfect multicollinearity, linearity and homoscedasticity Multicollinearity is tested by using the correlation matrix (see table 7). According to Field (2009) multicollinearity may exist when there is substantial correlation among variables (r > .9). When examining the correlation matrix, two variables have a correlation above this threshold, namely the protection of property rights and freedom from corruption (r = .931, p (one-tailed) < .05, N = 64). However, variables might still be multicollinear even when Pearson's r is < .9. When these variables are combined in the same regression, they might bias the outcome. In order to tackle this problem, a total of 42 unique regressions are performed. The variables that might be multicollinear, and thus problematic for regressions, are presented in bold in the correlation matrix (see table 7). Moreover, the data is tested on homoscedasticity. This means that the residuals should have the same variance at each level of the predictors (Field, 2009). When this assumption is not true, then heteroscedasticity is present, which means that the variances are very unequal. Heteroscedasticity also biases the outcome of a regression since it biases the standard errors, which in turn lead to biases in the test statistics (Field, 2009). To detect heteroscedasticity a plot will be used in which ZRESID (standardized residuals) is plotted against ZPRED (standardized predicted values of the dependent variable based on the model). When examining figure 3 (appendix 3) it becomes clear that no heteroscedasticity exists, since the plot shows an random array of dots dispersed around zero. Lastly, the data is tested on linearity. This is done by examining a histogram and a normal probability plot. The relationship that is tested in this research needs to be a linear one, in which the predictors lie along a straight line. If non-linearity exists, then this limits the generalizability of the findings (Field, 2009). When examining figure 4 and figure 5 (appendix 3) it becomes clear that the residuals are normally distributed and thus linearity exists.. 33.

(34) Master thesis Jan Hakvoort - 20th of June, 2014 6.3.2 Regression results In this research, a total of 42 unique regressions were performed, which results in 42 unique regression models. Table 8 and 9 present the main models used in this research to test the hypotheses (the remaining regression tables can be found in appendix 4). In each model the standardized regression coefficients (β) are presented along with its standard error in brackets. The standardized regression coefficients are presented, because this makes comparison between variables easier since there is a standard unit of analysis. The base model in this research is model 1, which includes the three control variables (i.e., GDP per capita PPP adjusted, average years of total schooling (15+), and the unemployment rate). The models 2 - 9 measure the relationship of the independent institutional and cultural variables against the aforementioned control variables individually. Furthermore, model 10 presents the three attitudinal measures (i.e., career choice, high status, and media attention) against the three control variables, whereas model 11 presents the four institutional measures (i.e., protection of property rights, freedom from corruption, business freedom, and access to finance) against the control variables. Moreover, model 12 includes all variables except social desirability, whereas model 13 includes all variables except the three attitudinal measures. Lastly, model 14 and 15 capture the less developed economies, whereas model 16 and 17 capture the developed economies.1 Hypothesis 1 links the protection of property rights with entrepreneurial activity and can be tested by using model 6, 11, 12 and 13 (see table 8). In none of these models, the relationship between the protection of property rights and entrepreneurial activity becomes significant. Since the aforementioned models include both GDP per capita PPP adjusted and freedom from corruption, the outcome might be biased, due to high correlation. Therefore, other models have been made in which the relationship between the protection of property rights and entrepreneurial activity is tested without using these variables (see appendix 4 regression table 3 and 4). Also in none of these models, the relationship between the protection of property rights and entrepreneurial activity has been found significant. However, the protection of property rights becomes significant when making a distinction between developed economies and less developed economies (see table 9). Model 16 and 17 both capture the developed economies and show a highly positive significant relationship between the protection of property rights and entrepreneurial activity. As a result, hypothesis 1 is partially supported. 1. The distinction between developed and less developed economies has been made by calculating the median of the GDP per capita PPP adjusted, which is $15,379.09.. 34.

(35) Institutional triggers that drive entrepreneurship - 20th of June, 2014 Table 8 Regression table 1 with TEA as dependent variable Model Constant Control variables GDP per capita PPP Schooling Unemployment. Model 1. Model 2. Model 3. Model 4. Model 5. Model 6. Model 7. Model 8. Model 9. Model 10. Model 11. Model 12. Model 13. 24.333*** (3.927). 6.554 (7.026). 13.474 (8.353). 7.170 (5.714). 17.622*** (3.987). 24.018*** (3.975). 23.992*** (4.087). 28.710*** (5.191). 24.351*** (4.162). -.371 (8.333). 29.051*** (5.798). 3.247 (8.970). 22.565*** (5.378). -.397*** (.000) -.207 (.507) -.176 (.146). -.195 (.000) -.133 (.485) -.191* (.138). -.389*** (.000) -.131 (.535) -.165 (.145). -.276** (.000) -.130 (.463) -.115 (.134). -.236* (.000) -.053 (.482) -.144 (.133). -.468** (.000) -.227 (.521) -.248* (.148). -.446** (.000) -.213 (.514) -.180 (.148). -.294* (.000) -.175 (.512) -.150 (.148). -.398** (.000) -.207 (.512) -.176 (.149). -.154 (.000) -.106 (.479) -.137 (.132). -.383* (.000) -.209 (.526) -.147 (.151). -.032 (.000) -.104 (.488) -.088 (.134). -.163 (.000) -.041 (.488) -.101 (.134). Independent variables Career choice. .393*** (.069) .175 (.077). High status. .411*** (.054). Media attention. .258* (.073) -.022 (.078) .388*** (.061). .460*** (.823). Social desirability Property rights Corruption freedom Business freedom Access to finance N Adjusted R2 F-statistic. .277* (.072) -.038 (.077) .348*** (.058). .111 (.047) .064 (.060) -.185 (.071) .002 (1.020) 64 .298 9.904***. 64 .379 10.612***. 64 .311 8.111***. 64 .428 12.785***. 64 .423 12.532***. 64 .291 7.464***. 64 .287 7.346***. 64 .305 7.917***. 64 .286 7.305***. 64 .448 9.510***. .255 (.097) .032 (.114) -.268* (.079) -.077 (1.225). .482 (.088) -.169 (.107) -.305** (.069) -.199 (1.079). .494*** (.815) .319 (.086) .015 (.101) -.323** (.070) -.142 (1.090). 64 .291 4.695***. 64 .471 6.600***. 64 .444 7.293***. * p < 0.10, ** p < 0.05, *** p < 0.01. 35.

No results found