Impact of the Crisis on Total Early-Stage Entrepreneurship in High and Low Uncertainty Avoidance Developing Countries Contemporary Issues in International Business

(1)

Impact of the Crisis on Total Early-Stage

Entrepreneurship in High and Low Uncertainty

Avoidance Developing Countries

Contemporary Issues in International Business

by

Mariya Genova

S2538415

Supervisor: Drs. Huib Stek

Co-assessor: Dr. André van Hoorn

International Business and Management

Faculty of Economics and Business

(2)

Abstract

The purpose of this thesis is to explore whether the global economic crisis, operationalized by GDP per capita growth, impacts entrepreneurial activity within developing countries. Furthermore, the impact of the Hofstede`s uncertainty avoidance cultural dimension on entrepreneurship is being examined in this study. The hypotheses are tested using linear multiple regression analysis for a period of ten years and data on 38 countries for the pre-crisis period, and 25 countries for the crisis period, having participated in the GEM project. It is found that GDP growth has a significant positive impact on entrepreneurial activity within developing countries during the crisis, whereas it is not found that uncertainty avoidance moderates the relationship between GDP growth and entrepreneurial activity. Additionally, the study controls for the effect of three institutional factors – the effect of three of the Worldwide Governance Indicators indexes – government effectiveness, rule of law and control of corruption. In addition, robustness checks, using GLOBE uncertainty avoidance, are executed to support the results. The paper suggests that entrepreneurship as an engine of growth should be given strong support by government institutions of developing countries.

(3)

1. Introduction

There is no doubt that the concept of entrepreneurship has received through recent years an increasing interest from researchers of different spheres. Interestingly, there is some debate among them, regarding the existence of a consistent theory, related to entrepreneurship (Amit et al., 1993; Bull and Willard, 1993; Bygrave, 1989; Cooper and Artz, 1993). But leaving aside this scientific debate, entrepreneurship itself can be described as a complex phenomenon that involves a great variety of contexts and factors (Pinillos and Reyes, 2009).

Entrepreneurship can be ascribed as central to the functioning of developing economies. In doing so, entrepreneurs are able to ensure the adequate use of resources, as well as the expansion of the borders of economic activity. Wickham (2006) gives a good definition of an entrepreneurial individual who runs a small business and assumes all the risk and reward of a given business venture, idea, or good or service offered for sale. Blanchflower and Osvald (1990) see entrepreneurial people as risk-takers and innovators, who are able to reject the relative security of employment in large organizations to create wealth and accumulate capital. Tödling and Wanzenböcl (2003) imply the idea that the study of new firm creation has gained an important place due to its importance in economic development, renewal and employment generation. Entrepreneurship is seen as a main prerequisite for economic growth (Van Stel, 2005).

Notwithstanding, in order to contribute to the development of entrepreneurial activity in developing countries by knowing the vitality of entrepreneurship for functioning of developing economies, it is important to know which factors are directly related to it, the exact soundness of the existing relationship and which of the corresponding factors are worth strengthening by policy makers.

(6)

increasingly important pillars of the economy (Roudini and Osman, 2012). Furthermore, a global economic crisis, such as the most recent 2008 – 2009 crisis can affect the dynamics of structural change and hence maintain or even widen global disparities of entrepreneurial activity (Gries and Naude, 2011). A key channel is the behavior of entrepreneurs during and after a crisis (Gries and Naude, 2011). SMEs and entrepreneurs are playing a significant role in all economies and are the key generators of employment and income, and drivers of innovation and growth.

However, little attention has been given to the eventual effect that the crisis exerts on entrepreneurs, and especially start-up companies. Gries and Naude (2011) state that the relationship between entrepreneurship and the crisis has been widely neglected by researchers. The studies that have so far been involved in trying to find the effect that the crisis has on firms – the failure of existing firms and the new-firm creation – are few (Koellinger and Thurik, 2009; Naude and McGee, 2009; OECD, 2009). The authors imply the idea that even though the crisis is near its end, the rapid changes that affected the nature of surviving of new entrepreneurs can have long-term impacts. Furthermore, in most of the cases the changes in levels of entrepreneurship activity are studied within a single country, or a set of countries (Lingelbach et al., 2005; Smallbone et al., 2012; Wennekers et al., 2007). Though, in the academic literature little devotion has been given to a cross-country comparative analysis. Additionally, most research fails to address what exact influence the global economic crisis could exert on entrepreneurial activity across developing countries. Past academic studies only aim at explaining the different levels of entrepreneurial activity dependent on the economic conditions in different countries (Acs et al., 1994; Blanchflower and Oswald, 1994; Meager, 1992; Sternberg and Wennekers, 2005).

(7)

levels of business and market risk (Lingelbach et al., 2005). In addition, entrepreneurs in developing countries face a different set of circumstances than their counterparts in developed economies. Thus, it is expected that substantial differences in level of entrepreneurial activity are evident for developing economies. Notwithstanding, as the level of economic development is related to levels of entrepreneurship, some researchers find differences in entrepreneurial activity among countries with the same level of development1 (Van Stel et al., 2005). This is a main indication that reasons apart from economic ones may be able to explain differences in entrepreneurial activity, and this reasons may be rooted in cultural differences (Hofstede et al., 2004). Empirical evidence provides support to this concept and reveals that culture is an extremely important factor in explaining the differences in entrepreneurship in different countries (Noorderhaven et al., 2004; Wennekers et al., 2007).

A lot of research has been done in order to investigate the impact that culture could exert on entrepreneurship. There is plenty of research, regarding the cultural impact on a society and how it is able to develop, taking into account this cultural aspect (Sondergaard, 1994, Hayton et al., 2002). They, most of the time consider aspects in the field of culture and management and more specific ones, regarding the impact of culture on entrepreneurship. But still, the academic literature is scant on studies that aim at exploring the influence of culture on entrepreneurial processes in times of crisis. According to a study (Doğan and Özdemirc, 2012), the cultural values of business owners or leaders affect the perception of difference of crisis as opportunities or threats, and also they impact the strategic behavior in the condition of crisis. Doğan and Özdemirci (2012) prove that if the owners` or leaders` uncertainty avoidance is high, they will choose stability in the condition of crisis.

Taking into account the significance that entrepreneurship plays in developing economies for their wealth creation and the unexplored effect of the crisis and culture on entrepreneurship, the main aim of this study is to contribute to the academic literature by looking at the eventual effect that the global economic crisis exerts on levels of entrepreneurship in developing countries. Therefore, I formulate the following research question: What is the influence of the global

economic crisis on levels of entrepreneurship within developing countries?. In addition, second

1_{Countries such as India, Thailand, Russia are characterized by commensurate levels of economic development, but}

(8)

contribution of this study is addressing a more detailed second research question: What are the

differences between levels of entrepreneurship among high and low uncertainty avoidance countries affected by the global economic crisis?. The essential policy implications that could

be drawn from this study are also discussed. I consider this last question considerably important since entrepreneurs are considered to be a key driver for economic growth, innovation, employment and social integration (OECD, 2009). It is succinctly stated that policy makers need to go further than simply appreciating the role of entrepreneurs in economic growth. They must understand those key policy factors that help entrepreneurs thrive, on the one hand, or that could stamp them out (Roudini and Osman, 2012). Moreover, any global crisis badly affects entrepreneurial activity. Therefore, the need to develop entrepreneurship to overcome any global economic crisis remains the concern of developing countries in order to sustain economic growth (Roudini and Osman, 2012).

The effect of the global economic crisis is operationalized by comparing countries` national income growth rates. The sample constitutes 38 countries for the pre-crisis period and 25 countries for the crisis period, originating from developing economic environment, scoring differently on entrepreneurship (total early-stage entrepreneurship)2. This variable is consistently measured across variety of countries and appears to be a useful index in measuring the extent of entrepreneurship (Van Stel et al., 2005). I find evidence that GDP growth has a significant positive impact on entrepreneurial activity within developing countries during the crisis. Furthermore, the regressions results show no statistical significance for the importance of uncertainty avoidance during the global economic crisis and its moderation effect on the studied relationship. In order to see the effect of the crisis, the findings of a pre-crisis and a crisis period are compared (2003 – 2007; 2008 – 2012). Furthermore, I control for the effect of some of the World Bank Worldwide Governance Indicators (WGI)3 that represent the six dimensions of governance, such as government effectiveness, rule of law, and control of corruption. The results

2_{Total early-stage entrepreneurship - TEA rate represents the individuals, who are taking steps to start a business,}

i.e. these are the nascent entrepreneurs and owner-managers of business less than 3.5 years in existence, i.e. the new entrepreneurs (Hessel et al., 2013).

3_{The WGI project reports aggregate and individual governance indicators of 215 economies over the period 1996 –}

(9)

suggest that rule of law impacts entrepreneurship negatively, while control of corruption has a positive effect on entrepreneurship during the crisis.

The rest of the paper is organized as follows: Section 2 discusses relevant existing literature and presents the testable hypotheses. Section 3 introduces the methodology of the paper – the data and the used dependent and independent variables. Section 4 presents the results and the regression analysis. In the end, Section 5 gives the discussion, the implications, the directions for further research, and the concluding remarks.

2. Literature Review

In this section the different theoretical backgrounds are introduced, presenting two main issues. First, in the first sub-section, relevant literature is presented, discussing the different findings for the effect of GDP growth on entrepreneurship in regular times and during crisis. Hypotheses, based on this relationship are introduced. Next, this relationship between the dependent (TEA rate) and independent variable (GDP growth) is further discussed in terms of the moderation effect of uncertainty avoidance, relying on findings of Wennekers (2010), Wennberg (2013), Pinillos and Reyes (2009), etc. The main outcomes of this relationship, either positive or negative ones, are discussed and hypotheses are developed. In the end, the conceptual model is proposed, depicting the exact relations among the variables.

Development of entrepreneurship before and during the global economic crisis

Recent years have brought an increasing body of literature on the determinants of self-employment and entrepreneurship on a country level (Noorderhaven et al., 2004). However, most of the empirical work in this area is limited to mainly studying the influence of economic determinants (Blau, 1978; Carree and Thurik, 2002; OECD, 2000; Parker, 1996).

(10)

more developed (Schulz, 1990; Bregger, 1996). In most developing countries self-employment is typically a much larger proportion of the labor force than in developed countries (Blau, 1978). During regular times a low level of prosperity coincides with a low wage level, implying little pressure to increase efficiency or the average scale of enterprise (Noorderhaven et al., 2004). Thus, a major route for ambitious wage earners to increase their income is to set up a shop and become an entrepreneur. Consequently, economic development subsequently leads to a rise in wages which stimulates enterprises to work more capital intensively, to save on labor, and to reap economies of scale (Noorderhaven et al., 2004).

Following these lines of thought, Iyigun and Owen (1998) find that entrepreneurial human capital plays a relatively more important role in intermediate income countries, whereas professional human capital is relatively more abundant in richer economies. As an economy develops, individuals choose to invest in more time accumulating professional skills than accumulating entrepreneurial human capital (Iyigun and Owen, 1998). Furthermore, as per capita income grows and the payoff to being a professional increases, individuals are less willing to gamble on entrepreneurial ventures. This phenomenon occurs even though the expected value of entrepreneurship rises with per capita income (Iyigun and Owen, 1998).

This trend towards lower levels entrepreneurial activity may waken, or even be reversed at a still later stage of economic development when differentiation of consumer demand increases and services become more important, creating new opportunities for self-employment (Noorderhaven et al., 2004). Different studies have provided empirical evidence of the existence of a U-form relationship between the level of economic development and the level of entrepreneurship across countries (Carree and Thurik, 2002; Stenberg and Wennekers, 2005; Wennekers et al., 2007). This partly explains the indication of resurgence of entrepreneurial activity in some of the most highly developed economies (Noorderhaven et al., 2004).

(11)

create their own business. In that situation new market opportunities appear being a good precondition for potential entrepreneurs to start a business and obtain relevant profits (Martin et al., 2010). Furthermore, empirical results show that the initial business environment is important for subsequent new business registration. GDP per capita is significantly and positively related to entry density of new start-ups (Klapper and Love, 2010). Consequently, the higher the income levels of a country over time would lead to higher new business creation rates.

Based on the aforementioned, I argue that during regular economic times (i.e. during the period 2003 – 2007) the income levels of developing countries foster entrepreneurial activity. Thus, I propose the following hypothesis:

Hypothesis 1a: Country`s rate of GDP per capita growth has a significant positive effect on

entrepreneurial activity within developing countries before the advent of the global economic crisis.

The aforementioned relations between economic development and entrepreneurship are discussed in terms of regular economic times when the world economy is not affected by the direct hit of the global economic crisis. Hence, a rather broad and complex perspective is needed to see how an economic turmoil can affect entrepreneurship in developing economies.

The global economic crisis resulted in both financial and trade shock to the world economy. The global economic downturn is much deeper than expected, and eventual recovery is gradual and rather uncertain (Global Monitoring Report, 2009). During the second half of 2008, the global economy came to a halt: on an annualized basis, global GDP growth4 slowed to 2 percent after and average growth rate of 5 percent over 2003 – 2007. In many developed countries GDP growth after 2008 is even negative. Though originating in advanced countries, the crisis hit the developing countries hard (Qureshi, 2009). Moreover, economic growth in developing countries has declined sharply to the lowest rates in some decades, and per capita income fell in many countries. Average GDP growth in developing countries in 2009 is only about a quarter of what

4

(12)

was expected before the economic turmoil. The situation intensified into e full-blown crisis in the latter half of 2008 and a fifth of that achieved in the period of strong growth up to 2007 (Qureshi, 2009).

A global economic crisis such as the most recent 2008-2009 crisis affects the dynamics of structural change and hence maintain or even widen global disparities (Gries and Naude, 2011). Thus, a key point is the behavior of entrepreneurs during and after a crisis (Gries and Naude, 2011). It is evident that during times of economic slowdown or recession, the purchasing power declines and people remain reluctant to invest, thus affecting entrepreneurship adversely (Nayab, 2011). The responses of advanced and some developing economies to the crisis have been unprecedented (Gries and Naude, 2011). The reduction of financing, accompanying the global economic crisis, found crucial for start-up companies, lead to a reduction in new firm start-ups, higher rates of firm failure, and slower growth, less investment and employment and productivity changes for existing entrepreneurs (Gries and Naude, 2011).

In developing countries, the effects of the global economic crisis on entrepreneurial activity are much more pronounced than in developed ones. Growth and public revenue in developing countries tend to be more export-driven or dependent than in developed countries, which possess larger internal markets. Thus, a global shock such as the current crisis leads to a disproportionately negative impact on entrepreneurship and entrepreneurial innovation in developing countries (Gries and Naude, 2011). Moreover, the advanced economies are able to provide a quicker relief in the form of expansionary monetary and fiscal policies and thus, entrepreneurs in such economies are able to recover more quickly with less permanent effects than in developing countries where the constraints may bind for a much longer period of time (Gries and Naude, 2011). The world in its globalizing posture reflects on mobility of capital, labor and goods across national boundaries in entrepreneurial effort of developing countries

(Roudini and Osman, 2012). Roudini and Osman (2012) imply the idea that the global economic crisis in an era of government-led privatization and promotion of entrepreneurial spirit of small- and medium-sized enterprises (SMEs) may throw most developing countries in economic and financial stagnation faster due to their vulnerability.

(13)

business fail as revenues fall or disappear altogether (Hipple, 2010). At the same time, measures such as unemployment have always increased during recessions. Hence, Hipple (2010) argues that a competing counter-cyclical effect results in a rise in self-employment if laid-off wage and salary workers start businesses for themselves. For instance, the total number of self-employed workers in a non-agricultural industries declined (on net, by about 760,000 from 15.0 million in the fourth quarter of 2007 to 14.2 million in the second quarter of 2009).

It is important to note here that the conditions of an economy affect the availability of entrepreneurial opportunities (Emenike, 2013). One such condition is the level of economic growth of a single country. In case that a country experiences sustained economic growth and stable macro-economic conditions, there is a likelihood that entrepreneurial activity is of a great value to a country (Bosma and Hardig, 2006). In times of harsh economic conditions people tend to consider engaging in entrepreneurship often leading to initial upsurge in new business creation (Emenike, 2013). Later on this upsurge is often followed by higher business failure rates and decline in funding source leading to a decline in new business start-up rate.

Based on the argument above that in case of deteriorated economic growth entrepreneurship activity is negatively affected, I argue that the recent economic turmoil impacts negatively entrepreneurial activity in developing countries due to their higher vulnerability during crisis. Therefore, I formulate the following hypothesis:

Hypothesis 1b: Country`s decreased rate of GDP per capita growth has a significant

negative effect on entrepreneurial activity within developing countries during the global economic crisis.

The moderation effect of uncertainty avoidance

(14)

Entrepreneurship is both influenced and shaped by the society in which the very entrepreneurial activity is placed (Coppola, 2011). As a matter of fact, the way a particular society is structured and the way it develops create the groundwork for the support of entrepreneurship. Coppola (2011) also adds that in the same way as the society frames the opportunities and the maintenance to entrepreneurship, the culture of a society, meant as the entire set of values, norms and beliefs, affects the way in which entrepreneurship is formed and managed.

Some of the cultural aspects, such as individualism and uncertainty avoidance, are notably related to entrepreneurial activity (Hayton et al., 2002). Furthermore, the concept of culture is further researched and developed by Hofstede5. His major study (1980) recorded differences in cultural patterns across 40 countries and sought to understand the institutional mechanisms that let these patterns continue stable in time and across successive generations. Hofstede (1980) defines culture as patterned ways of thinking and feeling that constitute the mental programming, which enables distinguishing people from one another, belonging to different groups. This mental programming constitutes patterns of ideas and notably, their attached values, being preserved and passed down from one generation to another.

One of the Hofstede`s dimensions that is considered to have a great influence on entrepreneurial decisions is uncertainty avoidance. Wennekers et al. (2007) state the fact that up to date uncertainty avoidance has received only scant attention in academic literature. De Luque and Javidan (2004) refer uncertainty avoidance to the extent to which individuals in a society feel threatened in ambiguous situations. It also relates to the level of order and rule-based reduction of uncertainty and the way individuals are willing to, practically, accept uncertainty. There is agreement that entrepreneurs make judgmental decisions in the face of uncertainty. They also reap the potential rewards of perceiving and utilizing opportunities. In this very process they also run the risk of losing their money and reputation as well (Wennekers et al., 2007). Uncertainty avoidance raises the question whether lower or higher levels of it hamper or

5

(15)

enhance total early-stage entrepreneurship. Gifford (2003) stresses the fact that there is also some consensus that people, involved in entrepreneurship, are less averse to risk.

Alternative views hold the idea that entrepreneurs are immanently more optimistic rather than less risk averse or dispose of relevant information reducing uncertainty and risk. In one of their research work, Mueller and Thomas (2000) investigate the case whether entrepreneurial traits vary systematically across countries. Their study shows the interesting results that cultures with a high score in individualism and a low score in uncertainty avoidance rate highest on a measure of entrepreneurial orientation6. Following these lines of thought, another study (Wennberg et al., 2013) by using data from Global Entrepreneurship Monitor and the Global Leadership and Organizational Behavior Effectiveness study for 42 countries investigates how the effects of individual`s self-efficacy and fear of failure on entrepreneurial entry are contingent on national cultural practices.

Wennberg et al. (2013) finds that in societies with high-uncertainty avoidance, individuals exhibiting a high fear of failure are more likely to display such towards entrepreneurship. In addition Wennekers et al. (2010) also find that in the group of high uncertainty avoidance countries a strongly negative relationship between GDP per capita and the level of business ownership is found. In a group of low uncertainty avoidance countries no such influence of GDP per capita is found. De Luque and Javidan (2004) grasp the tendency that societies characterized by higher levels of uncertainty avoidance are prone to take more moderate calculated risk. Moreover, executives from cultural backgrounds low in uncertainty avoidance are likely to engage in greater entrepreneurial activity (Schneider, 1989).

Thus, based on the aforementioned, I predict that before the crisis, in developing countries, scoring high on uncertainty avoidance, income growth rates are either positively or negatively related to levels of entrepreneurial activity. Therefore, I formulate the following two hypotheses:

Hypothesis 2a: Entrepreneurial activity is positively associated with GDP per capita growth

in high uncertainty avoidance developing countries before the global economic crisis.

6_{Mueller and Thomas (2000) use as a measure of entrepreneurial orientation innovativeness and internal locus of}

(16)

Hypothesis 2b: Entrepreneurial activity is negatively associated with GDP per capita

growth in high uncertainty avoidance developing countries before the global economic crisis.

Something that is noteworthy is that none of these pieces of research have explored the influence of culture on entrepreneurial processes during crisis. Doğan and Özdemirci (2012) are researchers who prove that the cultural values of business owners or leaders affect the perception of difference of crisis as opportunities or threats, and also they impact the strategic behavior in the condition of crisis. Doğan and Özdemirci (2012) use the cultural value of uncertainty avoidance as one of their independent variables, the authors prove that if the owner`s or leader`s uncertainty avoidance is high, they will choose stability in the condition of crisis. Furthermore, uncertainty avoidance is found to be related to managerial resistance to change (Geletkanycz, 1997) and to propensity to take risks (Bontempo et al., 1997) with the degree of risk in turn being a major criterion in decisions of how to strategically react to a crisis (Sternad, 2012).

Meissner (2013) argues that there is great potential for self-employment everywhere, but actual rates remain at a lower level. Furthermore, there are two main obstacles that prevent people from self-employment (Meissner, 2013). On the one hand, it is the lack of starting capital due to the uncertain economic situation caused by the global economic crisis, and especially fear of failure with its multiple facets such as financial ramifications, legal responsibilities, or disappointment of family. On the other hand, other factors that hamper entrepreneurship are legislation bodies and educational institutions remaining challenged in improving the conditions and regulations affecting self-employment. In countries directly involved by the economic crisis more than every second respondent views the crisis as a severe obstacle (Meissner et al., 2013).

(17)

Thus, based on the aforementioned, I predict that during the crisis in developing countries that both score low and high on uncertainty avoidance, entrepreneurial activity is negatively associated with decreased income growth rates. Therefore, I formulate the following hypothesis:

Hypothesis 2c: Entrepreneurial activity is negatively associated with decreased GDP per

capita growth in both low and high uncertainty avoidance developing countries during the global economic crisis.

Conceptual model

On the basis of the literature review, the conceptual model for this research project is shaped. Before putting focus on the very conceptual model7, some theoretical information is presented.

Any model is based on theoretical assumptions. Hence, it becomes relevant to understand what theory is and the role it plays in constructing a model in every research design (Jonker and Pennink, 2010).

In every research the process of operationalization plays a significant role. In other words, it is the process how the theoretical framework and the empirical reality are related when they are translated into observable constructs. According to de Groot (1969), operationalization is the process of changing a theoretical construct into constructs that can be seen in the empirical world, and the very translation process is guided by theory from literature. The model should be a truthful representation of the phenomenon, being studied (Jonker and Pennink, 2010). Furthermore, the model will be used to simplify the problem by reducing the number of properties that have to be included. Thus, making it easier to focus on the essentials.

In the current case, there are many other factors, including other cultural ones, which have an impact on entrepreneurship, but the focus itself is put exclusively on GDP growth. The theoretical rationale is that in times of crisis the GDP growth of a country grows very slowly or in some cases it even turns to be negative. Thus, predicting that GDP growth exerts influence on entrepreneurial activity.

7_{One very general definition of a conceptual model is – a conceptual model is nothing more than an abstraction of}

(18)

Furthermore, this relationship is investigated in terms of the moderation effect of uncertainty avoidance. The causal assumption here is that the counties` low or high scores in uncertainty avoidance cause difference in the entrepreneurial activity per country, all these seen through the prism of the crisis. The cultural concept is operationalized by the Hofstede`s index of uncertainty avoidance, the concept of economic development is operationalized by GDP growth, entrepreneurial activity is operationalized by the TEA (total early-stage entrepreneurship) rate. The effect of the crisis is seen in terms of the time period, being chosen (2003 – 2007 and 2008 – 2012). This relationship is graphically displayed in the following scheme:

Figure 1.Conceptual mode.

3. Methodology

(19)

Research design

According to Jonker and Pennink (2010), a design describes a flexible set of assumptions and considerations, leading to specific contextualized guidelines that connect theoretical notion and elements to dedicated strategy of inquiry supported by methods and techniques for collecting empirical material. Still, the essence of sound research remains making clear choices that structure the research (Jonker and Pennink, 2010). In addition, a sound design should link between the nature of the research question, its connection with the chosen theory, the methodology used and taking into account the context in which this takes place (Jonker and Pennink, 2010).

Research question

The linear multiple regression research method (Allison, 1977) is chosen for this thesis, since it fits well with the objective identified in the introduction. Taking into account the nature of the proposed research question: What is the influence of the global economic crisis on levels of

entrepreneurship within developing countries?, and trying to answer the second research

question: What are the differences between levels of entrepreneurship among high and low

uncertainty avoidant countries affected by the global economic crisis?, I include in the

statistical analysis one moderator variable. In the current case this is the Hofstede`s uncertainty avoidance index (UAI). Moderation implies that the causal relation between the two variables changes as a function of the moderator variable and the statistical analysis must measure and test the differential effect of the independent variable on the dependent variable as a function of the moderator (Baron and Kenny, 1986). Using such moderator variable intends to answer the second research question.

Unit of analysis

(20)

the countries, scoring differently on total early-stage entrepreneurship. The relationship that is explored here concerns the influence that the country`s GDP growth has on total early-stage entrepreneurship. This relationship is examined before and during the global economic crisis in order to see if the crisis impacts negatively entrepreneurial activity across countries.

The approach

A deductive (top-down) research approach has been adopted for the current study. After the identification of topic of interest, the relevant theoretical backgrounds were studied and a relevant research question was defined. The initial gathering of data allowed to look into the scientific claims that are relevant to the chosen topic and to discuss if the results predicted by the literature review fit with the empirical findings.

Data collection

The data that are used in this master`s thesis project are provided by five different information sources. These are the Global Entrepreneurship Monitor (GEM)8 project, the data base of the World Bank, the Worldwide Governance Indicators (WGI)9 project, indexes of Hofstede, and the Global Leadership and Organizational Effectiveness (GLOBE) research program. These independent sources provide respectively data on total early-stage entrepreneurship rate (TEA), GDP growth, GDP per capita (GDPPC), the control variables (effective governance, rule of law, control of corruption), and on levels of uncertainty avoidance (UAI, Hofstede, GLOBE).

8_{The Global Entrepreneurship Monitor (GEM) project is an annual assessment of the entrepreneurial activity,}

aspirations and attitudes of individuals across a wide range of countries. The project was initiated in 1999 as a partnership between London Business School and Babson College. The first study covered almost ten countries, but since then nearly 100 national teams from every corner of the world have participated, and what is more this number continues to grow annually. The GEM project is without any doubt the largest ongoing study of entrepreneurial dynamics in the world(source: the GEM project).

9_{The Worldwide Governance Indicators (WGI) project reports aggregate and individual governance indicators for}

(21)

GEM explores the role of entrepreneurship in national economic growth and reveals some national features and characteristics of the different countries that are directly linked to entrepreneurship. The quality of the data is absolutely guaranteed. The data is harmonized by a central team of experts. The three main objectives of the project are as follows: measuring differences in the level of entrepreneurship between countries, revealing factors that lead to appropriate levels of entrepreneurship, and last but not least, suggesting policies that may enhance the national level of entrepreneurship. The data collected are unique because, unlike any other data on entrepreneurship, the GEM experts study, at the grassroots level, the behavior of individuals with respect to starting and managing their own firm. This used approach, evidently, provides a more detailed picture of entrepreneurship than is found in official national registry data sets (source: the GEM project).

The focus of this thesis is put on developing countries. According to the classification of the World Bank, countries that have GDP per capita income less than 1,035 US dollars can be described as the low-income countries, these ones with income between 1,036 US dollars and 4,085 US dollars form the group of the lower middle income countries, these with income between 4,086 and 12,615 are the upper middle income countries. These three groups are referred to as developing economies. This term is used for convenience. It is not intended to imply that all economies in the group are experiencing similar development or that other economies have reached a preferred or final stage of development (World Bank, 2014).

In 2013, the number of countries, taking part in the project has risen to 72. Of these 72 countries that have participated in at least one of the GEM editions. Due to lack of information on GDP per capita for Taiwan, this country is excluded from the list of countries. In order to test the proposed hypotheses, I use the largest possible number of developing countries. In this case, this thesis considers a case of 38 developing countries for the pre-crisis period and 25 for the crisis period. After 2007, 13 of the countries record levels of GDP per capita above 12,615 US dollars, thus, they are excluded10 from the analysis for the crisis period (2008 – 2012). I exclude these countries, since according to the classification of the World Bank they already belong to

10

(22)

the group of the developed countries. Each of the countries, have participated at least ones in the GEM project for the two periods 2003 – 2007, 2008 – 2012, and data, respectively, on their uncertainty avoidance index, GDP growth rate and GDP growth is available as well. For the purpose of this study, the average value of the TEA rate of each country is used to measure entrepreneurship.

Appendix A presents the countries that are part of this analysis, as well as the countries` GDP per capita value and their corresponding average TEA rate11.

The variables

In the current study three main variables are used in order to test the hypotheses: total early-stage entrepreneurship (TEA), GDP growth (GDP_growth) and uncertainty avoidance index (UAI). The operationalization of the variables is made in the consecutive paragraphs.

Total early-stage entrepreneurial activity. The TEA rate represents the individuals, who

are taking steps to start a business. These are the nascent entrepreneurs and owner-managers of business less than 3.5 years in existence, i.e. the new entrepreneurs (Hessels et al., 2013). Nascent entrepreneurs are those adults between 18 and 64 years of age and they are, in particular, trying to start a new business which they will partially own. These adults should be actively involved in this start-up activity. Examples of such start-up activity could be: developing a specific business plan, searching for a location from where the future business will be active, and/or involvement in the organization of a start-up team. New entrepreneurs are adults 18 and 64 of age who currently own and manage a business for less than 3.5 years. The individual could be both the owner-manager of the new business and simultaneously involved in start-up activities for the launch of a new business. Such an individual is counted as one active person in the calculation of the TEA rates (Hessels et al., 2013).

Since GEM does not survey each country every year, I follow the approach of Pinillos and Reyes (2009) and Anokhin and Schulze (2009), and use the average of the TEA estimates over the two time periods (2003 – 2007 and 2008 – 2012) of this study as my measure. The variable

11_{For three countries, however, data on TEA were not available for the periods being researched. Thus, following}

(23)

TEA measures the average total early-stage entrepreneurship rate for each of the 38 developing countries that have participated in at least one of the GEM project editions from 2003 to 2007 (pre-crisis period) and for the 25 developing countries from 2008 to 2012 (crisis period) and for which also information on their uncertainty avoidance level and GDP growth is accessible. The year 2007 is included in the pre-crisis period12.

GDP growth. The eventual impact of the crisis is operationalized by this variable. GDP

growth is also measured as an average value before the crisis for the period 2003 – 2007 and after the crisis for the period 2008 – 2012. The GDP growth rate is an annual percent rate. It is the annual percentage growth rate of GDP at market prices based on constant local currency. Aggregates are based on constant 2005 U.S. dollars. Moreover, GDP is the sum of gross value added by all resident producers in the economy plus any product taxes and minus any subsidies excluded from the products` value. It is calculated without making deductions for depreciation of fabricated assets or for depletion and degradation of natural resources (World Bank, 2014).

Uncertainty avoidance index. The values of this index are between 0 and 100, and the

higher the value, the higher the level of uncertainty avoidance of the country, i.e. countries that score the highest on this index are most inclined to avoid risk situations. The highest values of this index are for countries such as Uruguay and Guatemala (98), El Salvador (94), Argentina, Chile and Costa Rica (86). The lowest scores on this index are for Vietnam (30), India (40), Malaysia (36), Philippines (44), South Africa (49), (Hofstede, 2001). The variable UAI measures, accordingly, the uncertainty avoidance score on each of the 38 countries, participating in the GEM project. For the robustness checks, the GLOBE uncertainty avoidance index is used, which is a 7-scale index. Thus, countries that score between 0 – 4 are low uncertainty avoidant, those scoring from betweem 4 to 7 are high uncertainty avoidant.

Uncertainty avoidance is used as a moderator variable. The one that is altering the relationship between the other two variables (Wuensch, 2014). Moderation implies that the causal relation between the two variables changes as a function of the moderator variable and the

12

(24)

statistical analysis must measure and test the differential effect of the independent variable on the dependent variable as a function of the moderator (Baron and Kenny, 1986). Using such moderator variable intends to answer the second research question: What are the differences

between levels of entrepreneurship among high and low uncertainty avoidance countries affected by the global economic crisis?.

As researchers (Wennekers, 2010) have observed a U-form relationship between entrepreneurial activity and level of economic development (measured by income per capita), the current thesis`s main aim is to provide empirical evidence whether different levels of uncertainty avoidance impact the relationship between GDP growth and levels of entrepreneurship in developing countries. The use of a moderator variable derives from the nature of the research question that has been posed in this master`s thesis. Lindley and Walker (1993) state that a moderator variable may reduce or enhance the direction of the relationship between a predictor variable and a dependent variable. They also imply that this variable may even change the very direction of the examined relationship between the two variables, diverging it from positive to negative and vice versa.

Moreover, different authors (Baron and Kenny, 1986; Lindley and Walker, 1993) imply the fact that also a moderator variable can be considered when the examined relationship between the variable is strong. In addition, they state that most of the time a moderator variable is considered when there is unexpectedly weak or unsteady relationship between the variables. Since it is expected that entrepreneurship will vary depending on levels of uncertainty avoidance, moderation in the regressions is essential. Uncertainty avoidance is presented as a dummy variable. Two dummy variables are introduced respectively for the two periods (2003 – 2007 and 2008 – 2012). They score 0 for low uncertainty avoidance countries and 1 for high uncertainty avoidance countries.

Control variables

(25)

Governance effectiveness. It reflects perceptions of the quality of public services, the

quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government's commitment to such policies (Worldwide Governance Indicators, 2014). The estimate of governance ranges from approximately -2.5 (weak) to 2.5 (strong) governance performance. This variable is measured as an average value before the crisis for the period 2003 – 2007 and during the crisis for the period 2008 – 2012 both for the sample of countries with Hofstede`s uncertainty avoidance index and GLOBE uncertainty avoidance index.

Conventional wisdom suggests that good governance fosters entrepreneurship. Trust in government effectiveness, political stability, rule of law, and voice in government affairs should be related to citizens` willingness to take risks, associated with investing, starting and managing new businesses (Friedman, 2011). The logic is that economic, social and self-actualization benefits of starting up and managing new businesses must outweigh the risks and burdens in order for entrepreneurship to occur.

In addition, government policies mold institutional structures for entrepreneurial action, encouraging some activities and discouraging others (Minniti, 2008). It is therefore clear that government policy has the power to influence entrepreneurial activity. It is also clear that such influence is not necessarily desirable as it may steer entrepreneurs toward actions that have negative socioeconomic externalities (Minniti, 2008). Regulation of entry is seen as a government procedure that can foster or hamper potential entrepreneurs. Puia and Munnis (2007) suggested that the regulation of entry that is the set of procedures governing the entry of new or international ventures is particularly important for entrepreneurship, as countries with more entrepreneurship tend to be associated with lower levels of entry regulations.

(26)

Rule of law. Rule of law reflects perceptions of the extent to which agents have confidence

in and abide by the rules of society, and in particular the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence (Worldwide Governance Indicators, 2014). The estimate of governance ranges from approximately -2.5 (weak) to 2.5 (strong) governance performance. This variable is measured as an average value before the crisis for the period 2003 – 2007 and during the crisis for the period 2008 – 2012 both for the sample of countries with Hofstede`s uncertainty avoidance index and GLOBE uncertainty avoidance index.

It seems clear that in order for entrepreneurship to function effectively, property rights need to be clearly defined and enforced. There has to be a mechanism for contracts to be agreed and a legal system in place to enforce such contracts (Hartog et al., 2010). Nystrӧm (2008) provides empirical evidence for the existence of a powerful link between legal structure and the security of property rights on the one hand, and entrepreneurship on the other. Aidis et al. (2009) provide evidence for the positive impact of rule of law. They find that strong protection of property rights plays a pivotal role in the institutional environment conductive to entrepreneurial activity, especially at lower levels of development. Aidis et al. (2009) use a sample of both developing and middle-income countries. Notwithstanding, when they include in the sample a set of highly developed countries, the impact of rule of law disappears.

In similar lines, Estrin et al. (2009) find evidence that a strong property rights system is important for high-growth aspiration entrepreneurship, but has less pronounced effects for entrepreneurial entry. In the same vein, Troilo (2011) by building on prior research, finds that property rights matter more for high technology entrepreneurs and rule of law matters for high job growth entrepreneurs, while both are significant for Schumpeterian entrepreneurs13.

Other researchers state that regulations matter as well. Levie and Autio (2010) argue that burdensome regulation of entry may have a disproportionate effect on potential entrepreneurs, i.e. entrepreneurs may need the protection that registration offers, and therefore must pay the price of regulation or decline to enter. The way regulations are enforced matters as well. This is

13_{Schumpeter emphasized the role of the entrepreneur as a prime cause of economic development. The role of the}

(27)

true because the rule of law determines the potential for expropriation of entrepreneurs` property rights. If rule of law does not prevail, regulations are simply ignored (Levie and Autio, 2010). By analyzing a six-year panel of 54 countries, Levie and Autio (2010) find evidence that regulations reduce strategic entrepreneurial entry in countries in which the rule of law is strong. In other words, rule of law, while not having a direct effect on entry, moderates the influence of regulation on strategic entrepreneurship entry rates. Consequently, I control for the effect of rule of law on country`s entrepreneurial activity.

Control of corruption. Reflects perceptions of the extent to which public power is

exercised for private gain, including both petty and grand forms of corruption, as well as "capture" of the state by elites and private interests (Worldwide Governance Indicators, 2014). The estimate of governance ranges from approximately -2.5 (weak) to 2.5 (strong) governance performance. This variable is measured as an average value before the crisis for the period 2003– 2007 and during the crisis for the period 2008 – 2012 both for the sample of countries with Hofstede`s uncertainty avoidance index and GLOBE uncertainty avoidance index.

Corruption might be a means for achieving certain benefits that make work in the official economy easier, such as winning a contract from a public authority, getting a license (e.g. for operating taxis or providing other services, or getting permission to convert land into `construction ready` land). A common complain of entrepreneurs is that they are subjected to a great deal of harassment (particularly long delays during critical times by bank officials processing loan applications, and other public officials overseeing entrepreneurs` operations). Thus, it is often essential for entrepreneurs to make unrecorded payments to these officials in order to minimize harassment (Taslim, 1994). By providing a framework analysis of the harmful consequences of corruption, Taslim (1994) concludes that any corrupt act raises the effective cost of inputs and reduces the number of entrepreneurs who can operate profitably.

(28)

corruption and the independent measures of entrepreneurial and innovative activity across countries.

On the contrary, other studies find evidence for the positive effect of corruption on entrepreneurial activity. Blackburn and Wang (2008) argue that organized corruption eliminates uncertainty and can make entrepreneurship attractive to more households. This is the case with many East Asian countries, pointed as the `East Asian Paradox`. Developing countries such as China, India, Indonesia, South Korea, and Thailand have enjoyed considerable growth with high levels of corruption (Blackburn and Wang, 2008). In addition, Meon and Weill (2010) observe that corruption is consistently detrimental in countries where institutions are effective. However, it may be positively associated with efficiency in countries where institutions are ineffective. Along the same vein, Dreher and Gassebner (2007) prove empirically that corruption facilitates firm entry in highly regulated economies. They also conclude that despite the overall impact of corruption on growth being negative, it may still promote entrepreneurial activity which has been suppressed by rigid regulations.

Consequently, as the effect of corruption on entrepreneurship seems to be rather debated, and depends on different country factors, I control for the effect of control of corruption in this study.

The research model

The advantage of using GEM data is that it provides a direct measure of entrepreneurial activity. Notwithstanding, the use of this data creates some challenges, since there are missing values due to coverage and method used by GEM (some counties included in the sample are not surveyed every year).

(29)

Furthermore, the following methodology is adopted, estimating linear multiple regressions14 with total early-stage entrepreneurship as the dependent variable:

TEA = βо + β1GDP_growth + β2UAI + β3DUM_UAI + β4Gov_eff + β5Rule_law + β5Con_corr + ɛ

I test seven models, which include different set of predictor variables and control variables. These seven regression models are tested both for the pre-crisis period (2003 – 2007) and for the crisis period (2008 – 2012) as well. The equations for the models being tested are as follows:

TEA = βо + β1GDP_growth + ɛ (1) TEA = βо + β1GDP_growth + β2Gov_eff + ɛ (2) TEA = βо + β1GDP_growth + β2Rule_law + ɛ (3) TEA = βо + β1GDP_growth + β2Con_corr + ɛ (4)

TEA = βо + β1GDP_growth + β2Gov_eff + β3Rule_law + β4Con_corr + ɛ (5) TEA = βо + β1GDP_growth + β2UAI + β3DUM_UAI (6)

,and

TEA = βо + β1GDP_growth + β2UAI + β3DUM_UAI + β4Gov_eff + β5Rule_law + β5Con_corr + ɛ (7)

14_{In this thesis project a regression analysis is used as a method. According to Brooks (2008), multiple regression}

(30)

Robustness checks

In this master`s thesis seven successive research models are tested for two successive periods, respectively. In order to see whether the results on the impact of uncertainty avoidance are reliable, robustness checks are made. Therefore, I execute the same regressions of model 6 and 7 for the two periods, respectively. However, one of the independent variables – Hofstede`s uncertainty avoidance – is replaced by the uncertainty avoidance index of the GLOBE15 project.

The dimension that is used in this master project is uncertainty avoidance. As defined by GLOBE (House et al., 2004), uncertainty avoidance refers to the extent to which members of collectives seek orderliness, consistency, structure, formalized procedures, and laws to cover situations in their daily lives. The GLOBE uncertainty avoidance index is measured by using four items, and the index is defined at societal level (House et al., 2004). The questions (in the samples) address the extent to which life is structured, predictable, orderly and consistent, and rules and regulations to regulate societal practices are emphasized. House et al. (2004) examine uncertainty avoidance as an aspect of practices and values at both societal and organizational levels. The implications of this cultural dimension for organizational leadership are explored.

There are two forms of questions for each dimension. The first of these two forms measure managerial reports of actual practices in their organization and managerial reports of what should be (values) in their organization (House et al., 2004). The second form measure managerial reports of practices and values in their societies. Despite the Hofstede`s uncertainty avoidance index, which scores from 0 to 100, the GLOBE uncertainty avoidance index scores from 0 to 7. Similar to Hofstede, higher scores indicate greater uncertainty avoidance. Due to some data synchronization issues, the number of countries from the GLOBE sample used in the analysis is 21 for the pre-crisis period and 17 for the crisis period (see Appendix E). This arouses from the fact that the GLOBE project encompasses smaller number of developing countries. In addition

15_{The Global Leadership and Organizational Behavior Effectiveness (GLOBE) Research Program was conceived in}

(31)

for some of them there is no available data for total early-stage entrepreneurship. South Africa (South Africa white and South Africa black) possesses double scores on AUI. Due to facilitation purposes, the average scores are calculated and taken into account. The results of the testing are analyzed in the consecutive section.

4. Analysis and Results

Descriptive statistics and correlations

Table2 and Table3, respectively, report the descriptive statistics for the whole set of variables (see Appendix B). In addition, I test for normal distribution of the data being used. For this purpose I use the Kolmogorov – Smirnov and the Shapiro – Wilk tests16. I conduct the test for the data of the two periods – the pre-crisis and the crisis period (see Appendix C). For the first period 2003 – 2007 (see Table 4, Appendix C) the values of all the variables of the K-S test, except this one for uncertainty avoidance UAI(D) = 0.146, p < 0, are not statistically significant which is a main indication that the distribution is normal. The TEA(D) = 0.151, p > 0.05, the GDP_growth(D) = 0.102, p > 0.05, Gov_eff(D) = 0.099, p > 0.05, Rule_law(D) = 0.118, p > 0.05, Con_corr(D) = 0.097, p > 0.05, are significantly normal.

For the second period 2008 – 2012 (see Table 5., Appendix C) the values of all the variables included are not statistically significant as well. Hence, the distribution is normal. The TEA(D) = 0.099, p > 0.05, the GDP_growth(D) = 0.132, p > 0.05, Gov_eff(D) = 0.110, p > 0.05, Rule_law(D) = 0.110, p > 0.05, Con_corr(D) = 0.142, p > 0.05, UAI(D) = 0.164, p > 0.05, are significantly normal.

As the sampling distribution is normally distributed, this enables me to present the Pearson`s correlations for the two samples (Field, 2009). Table 6 reports the correlations for the pre-crisis period and Table 7 reports the correlations for the crisis period, respectively.

16_{The Kolmogorov – Smirnov test and the Shapiro – Wilk test see whether distribution as a whole deviates from a}

(32)

For the pre-crisis period it is evident that GDP growth is negatively correlated with total early stage entrepreneurship. However, the correlation is not statistically significant, hence this result cannot be used as a preliminary support for the first hypothesis. All the three control variables are significantly negatively correlated with the dependent variable. Uncertainty avoidance and the dummy variable for uncertainty avoidance are negatively correlated with the dependent variable as well. Two of the control variables – rule of law and control of corruption – are significantly positively correlated with the third control variable government effectiveness. Uncertainty avoidance is negatively correlated with the dependent variable and significantly negatively correlated with GDP growth. Uncertainty avoidance is also negatively correlated with two of the control variables – government effectiveness and rule of law, and positively correlated with the third control variable – control of corruption.

Table 5. Correlation matrix (Pearson correlation, period 2003 - 2007)

Variable 1 2 3 4 5 6 7 1 TEA 1 2 GDP_growth -0.010 1 3 Gov_eff -0.396* 0.049 1 4 Rule_law -0.375* 0.057 0.925** 1 5 Con_corr -0.285* -0.126 0.859** 0.895** 1 6 UAI -0.128 -0.416** -0.141 -0.068 0.170 1 7 DUM_UAI -0.047 -0.200 -0.072 0.013 0.153 0.765** 1

** Correlation is significant at the 0.01 level (1-tailed) * Correlation is significant at the 0.05 level (1-tailed)

(33)

correlated with GDP growth. Two of the control variables – rule of law and control of corruption– are significantly positively correlated with the third control variable – government effectiveness. Again, uncertainty avoidance is negatively correlated with government effectiveness and rule of law, but positively correlated with control of corruption. The dummy variable of uncertainty avoidance is significantly negatively correlated with GDP growth and government effectiveness, negatively correlated with the other two control variables, and highly positively correlated with uncertainty avoidance.

Table 6. Correlation matrix (Pearson correlation, period 2008 - 2012)

Variable 1 2 3 4 5 6 7 1 TEA 1 2 GDP_growth 0.415* 1 3 Gov_eff -0.194 0.058 1 4 Rule_law -0.345 0.050 0.810** 1 5 Con_corr -0.074 -0.164 0.733** 0.797** 1 6 UAI -0.102 -0.530** -0.313 -0.219 0.120 1 7 DUM_UAI 0.013 -0.408* -0.461* -0.326 -0.052 0.835** 1 ** Correlation is significant at the 0.01 level (1-tailed)

* Correlation is significant at the 0.05 level (1-tailed)

Regression analysis

The results of the multiple regressions analysis for the pre-crisis period (2003 – 2007) are reported in table 7. The results of the regression analysis for the crisis period (2008 – 2012) are reported in table 8. Table 9 and 10 in the appendix (see Appendix D) report the regression results for the two periods, respectively, and introduce the collinearity diagnostics by reporting the various inflation factors (VIF) for each variable in the subsequent equations.

(34)

control variables. Equations 6 and 7 include the uncertainty avoidance variable and dummy variable for high and low uncertainty avoidance countries. Equation 6 tests only for the influence of the uncertainty avoidance dimension, whereas equation 7 not only tests for the impact of this dimension, but controls for the effect of the three control variables as well. The number of observations, the R square, and the adjusted R square are also included. The F-ratio17 is a measure of how much the model has improved the prediction of the outcome compared to the level of inaccuracy of the model (Field, 2009), or i.e. it tells us whether the model is good. For most of the models in table 7 and 8 the values of the F-ratios as above 1, which implies the idea that the models are reliable. The R square of all the regressions is generally low. However, the values of the R square raise considerably when the three control variables are included in the model (both for the pre- and crisis period).

Table 7 and 8 report the coefficients estimates for the 7 models being proposed, and for the two periods, respectively. Model 1 for the pre-crisis period indicates that GDP growth is negatively related to total early-stage entrepreneurship. However the relationship is not statistically significant (β = -0.038, p > 0.05). Thus, there is no support for hypothesis 1b. In model 2 and 3 the GDP growth variable is again not statistically significant. However, the two control variables – government effectiveness and rule of law – are significantly negatively related to entrepreneurship (β = -4.962, p < 0.05, and β = -4.160, p < 0.05). The third control variable – control of corruption – however is not significantly related to entrepreneurship. In model 5, where I put all the control variables together, neither one is statistically significantly related to levels of entrepreneurship.

The results regarding the impact of government effectiveness are consistent with Friedman (2011). In line with the obtained results, government effectiveness is perceived as significantly negatively related to entrepreneurship (Friedman, 2011). Regarding the impact of rule of law, Levie and Autio (2010) argue that burdensome regulation of entry may have a disproportionate effect on potential entrepreneurs, i.e. entrepreneurs may need the protection that registration

17

(35)

offers, and therefore must pay the price of regulation or decline to enter. Other scholars look into rule of law and find out that it is positively related to entrepreneurship (Aidis et al., 2009).

The picture gets quite different when I observe the results for the crisis period. This time in model 1 GDP per capita growth has a significant impact on total early-stage entrepreneurship. The coefficient presents a positive relationship, opposite to expectations (β = 1.051, p < 0.05). This implies the idea that the higher GDP per capita growth, the more entrepreneurs a developing country have during the global economic crisis. Thus, hypothesis 1b is rejected. In all of the seven models the positive relationship between GDP growth and entrepreneurship is statistically significant, as in model 5 and 7 this relationship is very statistically significant (p-values are

below 0.01). Comparing model 2 with this one from the pre-crisis period it is evident that

government effectiveness is again negatively related to entrepreneurship, but the p-value is above 0.05 (β = -2.986, p > 0.05). Similar to model 3 from the pre-crisis period rule of law is statistically significantly related to entrepreneurial activity, and the relationship is again negative (β = -4.695, p < 0.05). Something interesting to observe is the significance of the third control variable – control of corruption – in model 5. In model 5 both rule of law and control of corruption are statistically significant, however the coefficient of control of corruption is positive (β = -13.864, p < 0.01, and β = 13,434, p < 0.01, respectively). This is seen as a main indication that the lower the levels of corruption within a developing country (or i.e. the higher the control of corruption), the more people choose entrepreneurship as a career. These results are in line with the findings of Anokin and Schulze (2009) and Avnimelech and Zelekna (2011).

Model 6 and 7 test for the moderation effect of uncertainty avoidance on the examined relationship. For the pre-crisis period in model 6 the coefficient estimate for uncertainty avoidance is negative, but not statistically significant (β = -0.100, p > 0.05). The dummy variable is not statistically significant as well (β = 2.694, p > 0.05). In model 7 where all the predictor variables are included, uncertainty avoidance tends to have statistically significant negative effect on entrepreneurship (β = -0.214, p < 0.05), but the dummy variable and GDP growth are not statistically significant (β = 4.121, p > 0.05, and β = -0.155, p > 0.05,

respectively). The negative slope of the uncertainty avoidance coefficient implies that the higher

Impact of the Crisis on Total Early-Stage Entrepreneurship in High and Low Uncertainty Avoidance Developing Countries Contemporary Issues in International Business