The impact of government support on innovation
activity in developing countries
Master Thesis Business Administration-International Management Track
Student: Georgia Desinioti Student number: 11385650
1st Supervisor: Dr. Mashiho Mihalache 2nd Reader: Dr. Vittoria Scalera
Date: 22/06/2017
Faculty: Economics & Business
2
Statement of originality
This document is written by Student Georgia Desinioti who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
Table of Contents
1. Introduction ... 5 2. Literature Review ... 8 2.1 Innovation ... 8 2.2 Innovation Systems ... 10 2.3 Developing countries ... 112.4 The institution of Government ... 13
2.5 Supportive governmental policies on innovation ... 14
2.6 Government and innovation in developing countries ... 17
3. Theoretical Framework ... 19
3.1 Government Support and Innovation Activity ... 20
3.2 The moderating effect of economic freedom ... 22
3.3 The moderating effect of the type of government system ... 24
3.4 Conceptual Framework ... 25 4. Methodology ... 26 4.1 Description of Sample ... 26 4.2 Data Collection ... 28 4.2.1 Dependent Variable ... 28 4.2.2 Independent Variable ... 28 4.2.3 Moderators ... 29 4.2.4 Control Variables... 29 4.3 Method... 31 5. Results ... 32 5.1 Descriptive Statistics ... 32 5.2 Correlation Analysis ... 34 5.3 Regression Analysis ... 35 5.3.1 Assumptions ... 35
5.3.2 Hierarchical Multiple Linear Regression ... 38
6. Discussion ... 40
6.1 Findings ... 41
6.2 Managerial Implications ... 44
6.3 Limitations and Future Research ... 45
7. Conclusion ... 47
8. References ... 48
4
Abstract
Innovation is a long-discussed term in the literature. Significant scholars such as Schumpeter, who is one of the introducers to the topic, have developed a great number of theories concerning innovation and its application in the business world as well as in society. Nonetheless, there is a relevant topic that has not received much attention and needs to be addressed. This topic refers to the stimulation of innovation activity through the implementation of supportive governmental policies in developing countries. Previous studies have approached similar topics but there is no investigation of such a relationship between governmental support and innovation regarding developing countries. Therefore, this thesis attempts to fill this gap in the literature. Moreover, an examination of whether the aforementioned relationship is moderated by the degree of economic freedom and the type of government system complements the focal research.
Using a sample of 30 countries and a study period of 5 years, it is inferred that neither the main relationship is confirmed nor the moderating effects on it. Nevertheless, the study has been a useful channel directing to further future research on that promising topic.
Key words: innovation activity, government support, economic freedom, type of government system
5
1.
Introduction
In today’s competitive and continuously changing business environment, both small and large, national and international enterprises have to be in a state of vigilance for searching the components that will guarantee to them superiority against their competitors. Due to the rising phenomenon of globalization, the business landscape requires the achievement of a durable competitive advantage (Porter, 1990; Brookfield, 2003). This implies that businesses’ ultimate purpose should not be the pursuit of profit but of competitive advantage, because only competitive advantage can secure long-term profitability and consequently business survival (Barney, 1991; Peteraf, 1993).
The achievement of a sustainable competitive advantage mainly emanates from innovation (Pavitt, 1981; Nidumolu, Prahalad & Rangaswami, 2009). Innovation is a long-discussed term in the literature, which provokes contradicting opinions among scholars and public. Equally contradicting are the definitions given by scholars for the content of innovation. Nevertheless, it has been accepted that innovation refers to new products, new processes, new technologies etc. that can affect in a positive manner not only the business environment but also the society (Ahlstrom, 2010). Generally, it is defined as “a term free of values and comprehensive covering the whole spectrum of activities from discovery to first time practical application of new knowledge of any kind which aims at the fulfillment of requirements and meeting the goals of recipients in a new fashion and way there risk and uncertainty is inherent at any stage” (Kotsemir &Meissner, 2013, p. 3).
As already mentioned, the promotion of innovation is important not only for the business world but also for public welfare (Ahlstrom, 2010). Thus, its
6
enforcement should also be pursued by national institutions. Among the institutions that are able to lead to innovation enhancement is national government. The national government of a country is responsible for the implementation of supportive policies toward different economic sectors, like innovation (Todtling & Trippl, 2005). Policies such as public spending on R&D, fostering of learning process, tax incentives, protection of intellectual property, facilitation of trade and investments constitute measures of support on behalf of government (Porter & Stern, 2001).
Many years ago, most of developed countries have successfully established such policies pursuing the achievement of this objective i.e. the stimulation of innovation activity. However, the recent years, numerous developing economies have started engaging in innovation promotion and enhancement (Li & Kozhikode, 2009). Businesses originating in countries like China and India, have shifted their processes into new, innovative-ones and thus, have managed to compete large global companies long-established in the market place (Luo & Tung, 2007). Factors associated with the poor institutional environment of less developed countries, deterred them from engaging in innovation activity earlier. However, factors such as global integration and technological improvements have contributed to innovation forwarding in developing economies (Aubert, 2005).
Needless to say, extensive references can be found in the existing literature regarding the phenomenon of emerging markets, i.e. the emergence of firms originating in developing countries, “catching up” in innovation. Large body of the existing literature is dedicated to the explanation of how innovation from emerging countries is affected by culture, as well as innovation patterns, sustainability, the case of imitation etc. (Helpman, 1992; Etzkowitz & Zhou, 2006; Khazanchi, Lewis & Boyer, 2007;Leonardi, 2011; Leach et al., 2012). However, little is known about the
7
effect of national government supportive policies on innovation activity. Even though some researchers, such as Cassiolato et al. (2014), have highlighted the necessity of government support for the enhancement of innovation, an empirical research related to this topic is still lacking.
Therefore, this thesis attempts to shed light on this topic. Specifically, the main question addressed here is whether the implementation of supportive governmental policies towards innovation in developing countries can bring about its enforcement or not. Therefore, this study, after having identified what is government support, seeks to find the potential relationship between government support and innovation in developing countries. Particularly, the formulated research question is:
“Do governmental policies affect innovation activity in developing countries?”
Furthermore, the study examines the effect of two moderating variables on the former relationship. As a matter of fact, it will be hypothesized that the degree of economic freedom in a developing country and its type of government system (democratic or non-democratic) will affect positively (when the degree of economic freedom is high and the government system is democratic) or negatively (when the degree of economic freedom is low and the government system is non-democratic) the main relationship.
Several important implications accrue from this research. Firstly, it gives insights in a rather interesting topic which not only is not saturated but also little research has been done on it. Secondly, deepening in the relationship between governmental support and innovation activity, offers useful information to managers, as knowledge of institutional environment, part of which is the national government, is integral component of the decision-making process when internationalize. Finally,
8
this research is also useful for the government itself, as it can function like a feedback to test whether the applied measures have an impact or not.
The rest of this paper is structured as follows. First, the main concepts are explicitly described and previous literature contribution to the topic is cited. Then, the theoretical framework is presented and the hypotheses of the research are formulated. Subsequently, the method followed is described. A chapter presenting the results accrued from the regression analysis follows. Latter section discusses the findings of the research, managerial implications, limitations and future research direction. Finally, the study is completed with few concluding remarks.
2.
Literature Review
2.1 Innovation
Nowadays, innovation plays a very significant role in human life. It is considered as a springboard for progress. But what actually is innovation? Scholars have given different definitions of innovation and they have shaped conflicting views about it (Robinson, 1997). The supporters of innovation have developed different ideas about what is innovation. Some of them believe that it is a linear process which through the implementation of new technologies leads to a new product or process and eventually to development and better performance, while others believe that innovation is about continuous change i.e. given technologies, products, processes, ideas etc. are reviewed and renewed (Robinson, 1997). Nonetheless, all of them recognize the importance of innovation for economic growth as well as social development. On the contrary, opponents to innovation assert that innovation creates inequalities in industrial sectors and consequently in the society. They also point out
9
the high risks that often-times innovation entails (Robinson, 1997; Hekkert et al., 2007).
In addition, the term is usually used interchangeably with creativity. Actually, it is a fact that creativity can lead to innovation and innovation needs creative ideas, but this is not always the case (Gurteen, 1998). Creativity is a multifaceted term which can take many forms but usually refers to personality characteristics of people. It is composed of knowledge, creative thinking and motivation (Veeraraghavan, 2009). Thus, creative ideas can bring about innovation, although creativity on its own is not adequate for the achievement of innovation (Veeraraghavan, 2009).
There are a lot of attributes relevant to innovation, its definition and origins in the existing literature. According to Crespi (2004), innovation is a compound process which is affected by a great number of factors like economic freedom, property rights, income and property. Robinson (1997) defines innovation as a competitive advantage for an organization and argues that innovation is not only a product or process but an advantage gained against competitors not only in business but in every aspect of human life.
Overall, innovation is about making humans’ life easier (Kotsemir & Meissner, 2013). Certainly, its existence is necessary for the survival of individuals, enterprises and the whole of the society (Kotsemir & Meissner, 2013). Additionally, innovation brings about changes that can be radical or incremental and which affect the economic sector of a nation (Mokyr,1992; Schumpeter,1934). It can also explain some of the main income differences between countries (Hall and Jones, 1999; Romer, 1990). As far as the society is concerned, businesses attempt to meet the needs of most of population through the development of innovative products,
10
processes etc.; thus, benefit the society (Ahlstrom, 2010). Therefore, if a society impedes innovation activity, then important benefits for that society get lost in long-term (Ahlstrom, 2010).
In conclusion, innovation is not perceived anymore as a new improved product or a new method or process, but as “a term free of values and comprehensive covering the whole spectrum of activities from discovery to first time practical application of new knowledge of any kind which aims at the fulfillment of requirements and meeting the goals of recipients in a new fashion and way there risk and uncertainty is inherent at any stage” and which is necessary for businesses’ societies’ and nations’ welfare (Kotsemir &Meissner, 2013, p. 3).
2.2 Innovation Systems
The system approach has extensively been used in scientific articles to study economic and technological shift from traditional models to new-ones (Markard & Truffer, 2008). A definition of the term system is: “a set or arrangement of things so related or connected as to form a unity or organic whole” (Carlsson et al., 2002, p. 233). For systems engineers a system is defined as “set of interrelated components working toward a common objective” (Carlsson et al., 2002, p. 234). Three main system characteristics accrue from this definition: first, components, which are considered as the functioning parts of a system, second, relationships, which piece together the components and third, attributes, which represent the properties of the components and of their linkages (Carlsson et al., 2002).
The development of innovation systems approach has accrued from the combination of evolutionary and institutional theories. Innovation systems mainly signalize the collective and individual nature of innovation (Hekkert et al., 2007). The
11
notion of innovation systems was introduced in 1980s, i.e. by the time it became clear that innovation is not a linear process which consists of adjacent steps, but a systemic process which demands the interaction of many agents (Cassiolato et al., 2014). Furthermore, at the same time, Freeman (1982) and Lundvall (1988) talked about national innovation systems, a concept that was largely applied in both developed and developing countries (Cassiolato et al., 2014).
A great part of literature is dedicated to innovation systems. Freeman (1989) argues about innovation systems “The network of institutions in the public and private sectors whose activities and interactions initiate, import, modify, and diffuse new technologies”. Cassiolato (2014) defines innovation system as a group of firms and other actors who create new products, new processes and new forms of organization. For Hekkert et al. (2007, p. 415) it is defined as “all institutions and economic structures that affect both rate and direction of technological change in society”. Nevertheless, all the aforementioned scholars have pointed out that innovation systems’ fundamental characteristic is the interaction among actors. In a broader approach innovation systems include government policies, education, learning organizations etc., beyond science and technology policy (S&T) (Cassiolato et al., 2014). This idea was initially, underlined by Freeman (1982) who supported that governments should take action in promoting technology and innovation. This was one of the first papers in which governance was associated with innovation.
2.3 Developing countries
Traditionally, innovation has been developed by firms operating in developed countries. However, during the last decades, a number of developing countries like India, Russia, Brazil etc. started innovating and developing new technologies and capabilities, achieving to equally compete very big multinational firms from
12
developed economies (Li & Kozhikode, 2009). Several factors that enabled firms from developing countries to innovate are the need for quick response to MNEs products and services, the opportunity to get benefit of the spillovers from MNEs and the attitude of governments towards innovation (Li & Kozhikode, 2009).
The previous years, the peculiar economic, political and social environment of developing countries did not allow for the direct adoption of the accumulated innovation originating in developed countries (Zanello et al., 2015). Factors such as inappropriate business environment, government, low levels of education, different culture etc. deterred these countries from developing sufficient innovation systems and capabilities (Zanello et al., 2005).
However, in recent years, the phenomenon of globalization and the continuous technological improvement seem to have offered to developing countries an opportunity to keep up with developed countries in terms of innovation (Aubert, 2005). Indeed, many countries from the developing world have invested in innovation and have managed to compete with very large MNEs. In fact, several firms originating in China, India, Russia etc., have transformed their structure and in that way, they achieved the development of their national economy in a short time period (Luo & Tung, 2007). Certainly, some of those firms have also succeeded in expanding their activity in other countries, both developing and developed, engaging in outward foreign direct investment (Luo & Tung, 2007).
Needless to say, for truly achieving the enforcement and establishment of innovation activities, developing countries still need to acquire more knowledge and engage more in entrepreneurship (Aubert, 2005).
13
2.4 The institution of Government
Institutions have been characterized as one of the main drivers that are taken into consideration when a firm is forming its international strategy (Peng & Khoury, 2009). Their main contribution to international strategy is that they reduce uncertainty (North, 1991). Although many scholars have attempted to define the concept of institutions, North gave the most acknowledged definition. According to North (1990, p. 97), institutions are “the humanly devised constraints that structure human interaction”. Hodgson (2006, p. 2) quotes a similar definition: “we define institutions as systems of established and prevalent social rules that structure social interactions”. Both scholars point out the role of institutions as regulators of human interaction. This regulatory role of institutions allows them to enable or constrain certain behaviors (Hodgson, 2006).
In addition, North (1989) categorized institutions into two categories: formal and informal institutions. Formal institutions have been described as rules, as they determine formal contracts such as laws, constitution, legislation etc. (North, 1989; Zenger, Lazzarini & Poppo, 2000). Informal institutions refer to social norms and culture that can also impose constrains on human behavior, such as social values, religions, codes of conduct, behavior standards etc. (North, 1989; Zenger, Lazzarini & Poppo, 2000).
Among the determinants of institutional environment is national government. Government’s role is of great importance as it forms among others the business environment of a country; thus, indirectly affects business decisions about investments in the focal country (MacCarthy & Atthirawong, 2003).
14
In the 18th century, as Keynes (1927) notices, the role of the government was quietly distorting. Thus, many economists and philosophers of the following century viewed the state as having a less important role in the market place (Tanzi & Schuknecht, 1997). Specifically, they thought that the role of the state should be limited to some main responsibilities like law making, keeping order, defense system and protection of property rights. However, the following years, this distorting role weakened and the state started having a more active role (Tanzi & Schuknecht, 1997). In fact, several actions that prove that change are first, the development of the fiscal sector, which was receiving by then more public spending, and second the notable efficiency in public administrations performance (Saunders & Klau, 1985). Furthermore, ideas like redistribution of income, economy stabilization and economy intervention for dealing with externalities led to the stimulation of government spending (Saunders & Klau, 1985). Therefore, the role of the government was reinforced with new responsibilities in new sectors like education, health care, combating unemployment, innovation etc. (Tanzi & Schuknecht, 1997).
2.5 Supportive governmental policies on innovation
Part of a government’s duties is the implementation of supportive policies towards varied brunches of the national economy. One of these brunches, which is also very significant for the economic growth of a focal country, is innovation. Innovation is an evolutionary, non-linear and interactive process which requires intensive communication and collaboration between different actors; one major actor is the national government (Todtling & Trippl, 2005). According to Porter and Stern (2001, p. 5) “National innovative capacity depends in part on the technological sophistication and the size of the scientific and technical labor force in a given economy and it also reflects the array of investments and policy choices of the
15
government and private sector that affect the incentives for and the productivity of a country’s research and development activities.”
However, a great debate among economists has risen. This debate refers to whether government intervention on innovation activity is legitimate or not (Edler & Georghiou, 2007). Rationales for innovation policy rest on market and system failures (Edler & Georghiou, 2007). Innovation systems have been characterized as a frame of reference underlying government intervention (Etzkowitz & Leydesdorff, 2000). According to this opinion, the national government of a country is able to intervene in the economy through innovation, creating new markets or generally changing the rules of the game (Etzkowitz & Leydesdorff, 2000). However, it is firmly argued that government’s role has changed. This implies that government does not attempt to intervene innovation anymore, but stimulate innovation processes, promoting dialogue and building up social capital (Todtling & Trippl, 2005).
Recent years, official reports of OECD, UNESCO and the EEC publications indicate that the number of government policies concerning innovation enhancement has surprisingly increased (Pavitt, 1976). This increase proves the importance of innovation as a means of national economy enforcement. Furthermore, the increasing involvement of the government in the innovation activity has other implications too, such as the reduction of the need for protection through patenting (Basberg, 1987).
A national government can implement a great number of measures in order to design a policy that can support and stimulate innovation activities. Although the most known policy is R&D enhancement through grants and subsidies (Todtling & Trippl, 2005), other policy areas include fostering of learning process, protection of
16
intellectual property, tax incentives, facilitation of trade and investments (Porter & Stern, 2001).
Large part of academic papers is dedicated to the identification of public policies towards innovation and many authors have attempted to classify innovation policies into broad categories. One of the oldest articles aiming to report governmental tools that support innovation, is that of Rothwell published in 1981. According to Rothwell (1981), all innovation policy tools can be classified into three broad categories: a) environmental (taxation, legislation, regulations), b) supply (scientific and technical infrastructure, and information -and financial) and c) demand (procurement and public services) (see Exhibition 1, Appendix).
On the contrary, the paper of Edler and Georghiou (2007), distinguishes between two categories of governmental policy tools, namely supply and demand. The former category consists of the sub-categories finance and services policies, which are composed of many lower levels of tools, while the demand-side measures are composed of four main groups of tools, systemic policies, regulation, public procurement and stimulation of private demand (Edler & Georghiou, 2007). (see Exhibition 2, Appendix)
Moreover, Borras and Edquist (2013) use a three-fold typology to categorize policy instruments. In their article, they classify into three broad categories, namely regulatory instruments, economic and financial instruments and soft instruments. Governments use regulatory instruments in order to designate the interactions among parts of the society and the economy (Borras & Edquist, 2013). Their implementation is obligatory and they aim at defining market conditions. Types of such policies include regulations for property rights, research in universities, public research
17
organizations, industrial sector research (Borras & Edquist, 2013). The second category, economic and financial instruments, provides incentives, which either encourage or discourage innovation, as well as support for specific social and economic activities enhancing innovation. It includes tools such as research funding, tax incentives for R&D, support for technology transfer and for venture. Finally, soft instruments are implemented voluntarily and are considered as non- coercive (Borras & Edquist, 2013). This implies that their indications are not obligatory but provide recommendations. Furthermore, they are based on mutual exchange of information. Campaigns, codes of conduct, voluntary agreements, private and public partnerships are perceived as main tools of this category (Borras & Edquist, 2013).
Concluding, although different classifications have been developed about policy instruments that are implemented for innovation enhancement, the majority of scholars argue that governments usually, implement mixed types of policies, after having examined the individual features and the synergetic effects of each tool (Edler & Georghiou, 2007; Borras & Edquist, 2013).
2.6 Government and innovation in developing countries
The improvement of innovative capacity of a country depends upon the investment and commitment to a variety of innovation drivers (Furman & Hayes, 2004). Developing countries share a lot of similarities but also a lot of differences. For example, geographic region of origin and national innovation systems differ significantly among developing countries (Furman & Hayes, 2004). This indicates that there is not a single suitable institutional configuration that can certainly lead to innovation enforcement (Furman & Hayes, 2004). However, most of developing countries use to implement policies such as investment in human capital, so as to enforce innovation (Hu & Mathews, 2005). Policy commitments and investments are
18
complementary for innovation enhancement; in the absence of the one component the other is insufficient (Hu & Mathews, 2005).
Nevertheless, in developing countries institutional regimes, including the national government, are not strong. Most of them face difficulties in adopting policies that are capable to enhance innovation activity, such as intensive learning and R&D subsidies, probably because they lack autonomy and competence (Gu,1999). In other words, government officials that are not skillful to cooperate with business actors (e.g. entrepreneurs) collectively or individually, have failed to encourage innovation (Intarakumnerd, Chairatanab &Tangchitpiboona, 2002). Furthermore, a factor that disables governments of developing countries to imply encouraging innovation measures is the problem of bureaucracy (Intarakumnerd, Chairatanab &Tangchitpiboona, 2002). Bureaucratic procedures, imposed by formal institutions, delay innovation processes and contribute to a non-supportive policy towards innovation.
A common practice that policy makers follow is the uncritical adoption of measures implemented in developed countries (Radas & Bozic, 2009). However, all measures implemented by developed counties are not appropriate for all developing countries. Conversely, the government of a focal developing country should design its innovative policy and select tools for enhancing innovation according to the strengths and the weaknesses of the country (Aubert, 2005).
Overall, not all developing countries are identical and not all governmental policies towards innovation have the same outcome (Intarakumnerd, Chairatanab &Tangchitpiboona, 2002). Countries like China, South Korea etc. have achieved to enhance innovation. China’s government, for example, changed some legislations in
19
order to allow public research institutions to be transformed into private technology service enterprises, which contributed to the increase of patents (Hu & Mathews, 2008).
3.
Theoretical Framework
Although there is a significant number of articles that have attempted to identify governmental policies on innovation in developed countries, only little research has been performed on governmental policies on innovation in developing countries. Authors like Rothwell (1981) and Pavitt (1976) have analyzed what a government can do to enforce innovation, but their research interest was focused only on developed countries.
However, developing countries are an integral part of global economy and play a major role in the formulation of today’s business environment. In fact, a lot of them have developed a worth mentioning innovation activity, which is capable to compete innovations created in developed countries.
Therefore, it would be interesting to study whether the national governments of those countries have contributed to the development of innovation activities or not. The focal study has also important implications for business leaders. Knowing how a government of a developing country triggers innovation can help decision makers take better decisions regarding entry and establishment modes.
Hence, this section includes first an investigation of the relationship between the variables government support and innovation activity, and then an examination of the moderating effects of economic freedom and the type of government system on this relationship follows.
20
3.1 Government Support and Innovation Activity
As previously stated, innovation is an integral part not only of businesses but also of human life, as progress and survival of individuals, enterprises and the society largely depend on it (Kotsemir & Meissner, 2013). Beyond private enterprises and universities, government initiatives are also considered as a main actor that enables the enhancement of innovation through communication and collaboration (Todtling & Trippl, 2005; Kang & Park, 2012).
Particularly, a national government of a country can intervene in the market through the design and the implementation of the appropriate policies aiming to support and foster the innovation output of that country (Kang & Park, 2012). The reasoning behind this governmental movement underlies in the theoretical economic rational. According to this rationale, innovation market is unable to “produce” the desirable innovative output, which can only be achieved with the contribution of the government (Arrow, 1962; Jaumotte et al., 2005). This failure of the market is attributed to knowledge spillovers, failures of financial market, shortages of skilled labor and informational imperfections (Jaumotte et al., 2005). Hence, governmental intervention functions as an effort to “fix” such failures.
Governments in developed countries tend to intervene in the market in order to support innovation through R&D financing. Expenditure on R&D are considered to generate high social rates of return (Feldman and Kelley 2006). Among the common supportive policies that a government can implement to empower R&D are grants, tax incentives and intellectual property rights protection (Porter & Stern, 2002; Todtling & Trippl, 2005). It has been reported that the governments of all OECD countries spend a large amount of public money on innovation, aiming to enhance private sector
21
innovation, fulfill socio-economic objectives like health care sector, and fund research in public research organizations (Jaumotte et al., 2005). They also use tax incentives to stimulate private sector R&D expenditures (Jaumotte et al., 2005). In fact, recent years, tax incentives have become more important than grants in many economies (Jaumotte et al., 2005).
Another extensively discussed topic in the existing literature is the relationship between government support and R&D stimulation in developed countries. Hall and Bagchi-Sen (2007) found support for their hypothesis that funding for research was positively associated with intensity of R&D in US biotechnology firms. In addition, similar scientific studies have found significant support for the positive impact of government support on innovation enhancement in developed countries (Lerner, 2000; Wallsten, 2000).
Moreover, in many transition economies governmental efforts to foster innovation seem to have achieved this objective. For example, in Russia, where the R&D sector has almost exclusively been supported by the state, a rise in R&D funding has been observed, which is accompanied by rise in growth rates (Gokhberg & Kuznetsova, 2011). However, R&D sector in Russia needs more funding in order to compete developed countries (Gokhberg & Kuznetsova, 2011).
Concerning the aforementioned remarks, it can be easily hypothesized that supportive practices on behalf of the government can stimulate innovation activity in developing countries too. A focal government should implement policies aiming to increase R&D activity and consequently, innovation, which would give the country significant competitive advantage. Nonetheless, it is necessary to clarify that a government should implement mixed types of policies, while having taken into
22
consideration the individual features and the synergetic effects of each tool as well as the appropriability for the country’s business and social environment (Borras & Edquist, 2013; Edler & Georghiou, 2007). Thus:
H1: Supportive governmental policies positively affect innovation activity in developing countries.
3.2 The moderating effect of economic freedom
According to economic theory, economic freedom refers to the freedom to choose and supply resources, compete in business, trade and secure property rights (North & Thomas, 1973). It is argued that economic freedom affects incentives, productive effort, and the effective use of resources (De Haan & Sturm, 2000). It is necessary to clarify that economic freedom should not be confused with political freedom, which refers to the equal participation in political processes and fair elections, or with civil freedom, which refers to freedom of speech, religion etc. (Berggren, 2003).
The core values that economic freedom shields, are personal choice, protection of private property, and freedom of exchange. An individual is supposed to have economic freedom when: he/she acquired his/her property without the use of force, fraud or theft, he/she is protected from physical invasions by others and when he/she is free to use exchange or give his/her property to another as long as his/her actions do not violate the identical rights of others (Gwartney & Lawson, 1996).
Some of the main components that compose economic freedom index are: size of government, structure of economy and use of markets, monetary policy and price stability, freedom to use alternative currencies, legal structure and property rights,
23
freedom to trade with foreigners, freedom of exchange in capital and financial markets (Gwartney &Lawson, 2001).
Generally, it is admitted that when institutions ensure economic freedom, a more valuable output is produced which eventually leads to growth (Berggren, 2003). This happens because of the implementation of low taxation which contributes to higher returns on production, independent legal system and protection of private property (Berggren, 2003). Moreover, Easton and Walker (1997, p. 330) as well as many other scholars, argued that “more economic freedom is associated with higher levels of income and, possibly, faster rates of income growth, ceteris paribus”.
As already known, government is a type of institution; thus, the implementation of policies which aim to enhance actions that stimulate growth, one of which is innovation activities, is positively affected by the degree of economic freedom of the country. In other words, a government of a country with a high rate of economic freedom is more likely to adopt supportive policies towards innovation. Furthermore, according to the Index of Economic Freedom 2015 (Miller et al., 2015, p. 5): “Governments that respect and promote economic freedom provide the best environment for experimentation, innovation, and progress, and it is through these that humankind grows in prosperity and well-being”. At this point, it should be mentioned that developed countries in which governments invest in innovation, are highly ranked in the Index of Economic Freedom.
Therefore, concerning the above arguments, the following hypothesis is formulated:
H2: Economic freedom positively moderates the relationship between government support and innovation activity in developing countries.
24
3.3 The moderating effect of the type of government system
The national government of a country is nominated according to the regime of that country. The term regime refers to the system of governance which can be broadly categorized into democratic or non-democratic, although a variety of different regimes that are subject to the aforementioned types, exist (Rose & Mishler, 2002). In democracies, the national government is elected through national elections where the population of the focal country votes the representatives that they prefer, while in the case of autocracies the government is imposed and human freedoms are limited.
The corner stones of democracy are political rights and civil liberties thus, democracy is associated with great openness, freedom of choice, political stability and higher quality of governance (DiRienzo & Das, 2015). These proxies of democracy have been described by many scholars and supporters of the regime as a national competitive advantage in the global market that can lead to effectiveness and growth (Clague et al., 1996; Collier, 2000). Furthermore, Rivera-Batiz (2002) correlated in his study democracy and good governance, proving that the combination of two can stimulate innovation activity in a country. In the same article, Rivera-Batiz (2002) argued that democratic institutions are more capable to enforce practices of good governance because rulers need support from the country’s population to get reelected.
On the other hand, rulers in authoritarian regimes exercise arbitrary the authority and their incumbency is long-term (Rivera-Batiz, 2002). Therefore, they usually do not have the incentive to promote good governance and lead their countries to growth through innovation, but instead they seek to misappropriate the nation’s wealth (Barro, 1996). Nevertheless, this is not always the case; there were
non-25
democratic countries where the governments promoted good governance and were interested in economic development like Chile and Peru (Barro, 1996).
Overall, it is inferred that the type of government system of a country can affect the amount of government support towards innovation. Specifically, in democratic countries government support to innovation activity is expected to be larger than that in non-democratic-ones. The main reason is that in democratic countries, population’s welfare and country’s growth are of primary importance and as previously discussed growth can be achieved through innovation. On the contrary, in non-democratic countries, less attention has been given to economic growth and consequently to innovation than the ruler’s welfare. Hence, it is argued that:
H3: A democratic type of government system positively moderates the relationship between government support and innovation activity in developing countries. On the contrary, a non-democratic type of government system negatively moderates the relationship between government support and innovation activity in developing countries. 3.4 Conceptual Framework Supportive Governmental Policies Moderator: Type of government system Innovation Activity in Developing Moderator: Economic Freedom + +/-
26
4.
Methodology
This section refers to the method used for testing the formulated hypotheses. Specifically, the sample and the data collection sources and process for each of the variables are discussed as well as the method used for testing the model.
4.1 Description of Sample
The total population for this study is developing countries. The term developing countries refers to countries in which the levels of income, education and technology are lower in comparison to developed countries (Aubert, 2005). Hence, according to this definition an initial list of 150 countries, that were perceived to have the characteristics of a developing country, was created. Then, the classification of those 150 countries as developing was tested using the official lists of the report “World Economic Situation and Prospects” which is issued by the United Nations. Specifically, all WESP reports from 2010 up to 2014, which is the time frame of this study, were examined for the classification of developing countries. This examination indicated that only 107 of the initial number were classified as developing during those years. Many of the countries that were included in the initial list are in transition, while others were listed by misperception, although they are actually classified as developed countries.
In addition, the Human Development Index (HDI) of those countries was examined, for further verification. The HDI is an index introduced by the United Nations Development Program (UNDP) in 1990, which aims to offer a simple but multidimensional approach to assess in a comparative manner the human development of various countries (Sagar & Najam, 1998). The first Human Development Report proposed that “development is much more than just the
27
expansion of income and wealth” and defined it as “the process of enlarging people’s choices” (UNDP, 1990, p. 1). According to this report, the HDI composed of three main elements, namely a long and healthy life, knowledge acquiring and access to resources that enable a decent standard of living (Sagar & Najam, 1998). This index has been proposed as an alternative to measurements of development such as the GDP and many scholars have used it in their studies for classifying countries into developed and less developed (Dumith et al., 2011; Anand & Ravallion, 1993). The index is a score between 0 and 1; countries that score less than 0.500 have low human development, between 0.500 and 0.799 have middle human development and above 0.800 have high human development (Dumith et al., 2011).
After applying this index in all 107 countries and taking into consideration the availability of data for all the variables, as well as WESP reports, the sample narrowed down in 30 countries. In the sample, there are countries located in Latin America, Asia and Africa and which present low, middle and high human development (see Table 1, Appendix). The decision to include in the research countries that scored highly in the HDI was based on the thinking that these countries are more likely to verify the hypotheses compared to those who score low, and this high score in the HDI can contribute to the explanation of such a result.
As previously mentioned, the study examines the variables in a five-year period and specifically from 2010-2014. Although the selected period is quite short, the data used are measurements of the very recent years. Besides, the availability of data for developing countries limits the extend of this study, mainly due to poor records in those countries.
28
4.2 Data Collection
4.2.1 Dependent Variable
Scholars have used a wide variety of measurements to assess the innovation activity. A few typical examples are the number of patents (Lanjouw & Schankerman, 2004; Acs & Audretsch, 2005) and scientific articles (Camisón-Zornoza et al., 2004). However, in this study, the magnitude of innovation activity is captured by the Global Innovation Index scores for the years 2010-2014 in the selected developing countries. The index has been published since 2010 by Cornell University, INSEAD, and the World Intellectual Property Organization (WIPO, agency of the United Nations). Because of its recent invention, it has not been widely used in studies. Nevertheless, an advantage of this index is that it presents innovation in terms of multi-dimensional facets, as it ranks innovation capabilities and results of world economies (Rejeb et al., 2008). On the contrary, a measurement of innovation through patents could lead to misleading results because of differences in legislation about patent protection (Crosby, 2000). Finally, it is considered that the number of scientific articles is not suitable for this study, as it does not capture the whole picture of innovation.
4.2.2 Independent Variable
The independent variable of this study is government support which usually involves the implementation of policies aiming to enhance innovation in one country. For the measurement of this variable, the R&D expenditure performed by the government as a percentage of the country’s GDP is used. The data were provided by UNESCO Institute for Statistics. This measurement allows for a precise assessment of the amount of money that a focal government spends on R&D enhancement, which is as previously stated integral part of innovation (Feldman and Kelley 2006).
29
Furthermore, a number of studies (Hu & Mathews, 2008; Kang & Park, 2012) have also used R&D expenditure to measure innovation.
4.2.3 Moderators
The second hypothesis of this thesis proposes that economic freedom positively moderates the relationship between government support and innovation activity. For the measurement of the variable, the index of Economic Freedom issued by the Heritage Foundation is used. The choice of economic freedom as a moderator was based on the plurality of components that economic freedom encompasses such as size of government, structure of economy and use of markets, monetary policy and price stability, freedom to use alternative currencies, legal structure and property rights, freedom to trade with foreigners, freedom of exchange in capital and financial markets (Gwartney & Lawson, 2001). Moreover, it is an index widely used in former studies (Pieroni et al., 2013; Doucouliagos & Ulubaşoğlu, 2008).
The second moderator used in this study is the type of government system which refers to the regime of a country. There are a lot of types of government system, but in this research two broad types are accounted, namely, democratic and non-democratic governments; thus, the type of government system is dummy variable which takes price 1 for democratic governments and price 0 for non-democratic governments. Each of the countries used in this study was characterized as democratic or non-democratic based on the Economist Intelligence Unit (see Table 2, Appendix).
4.2.4 Control Variables
The GDP per capita, the total amount of R&D expenditure and the size of each country in terms of population were used as control variables in this study. These
30
control variables will allow for a better insight in the hypotheses testing and will control for any possible bias.
GDP per capita. The innovation environment of a country largely depends on the economic development of the focal country (DiRienzo & Das, 2015). Economic development affects business, financial etc. issues that influence innovation activity in a country (DiRienzo & Das, 2015). Measuring economic development by GDP per capita allows for the “capture of the ability of a country to translate its knowledge base into a realized state of economic development” (Stern, Porter, Furman, 2000, p. 20). Besides, a great number of studies have been controlled for biases with the use of GDP per capita (Stern, Porter, Furman, 2000; Anokhin & Schulze, 2009; De Clercq, Meuleman & Wright, 2012). The data for GDP per capita used in this study were obtained from The World Bank.
R&D expenditure (total). The expenditure for R&D can be performed by different actors. The main actors that fund R&D activity are business enterprises, government, higher education, and private non-profit enterprises (Smith, 2005). The total amount of R&D expenditure performed in a country is used as a control variable because “innovation feeds on knowledge that results from cumulative R&D experience on the one hand, and it contributes to this stock of knowledge on the other” (Coe & Helpman, 1995, p. 860). Other studies have also used total R&D expenditure as a control variable (Srholec, 2011). The United Nations Educational, Scientific, and Cultural Organization (UNESCO) Institute for Statistics provided the data for measuring R&D expenditure as % of GDP.
Size. Finally, another factor that could cause bias in the study results is the size of country in terms of population. Country size can explain some differences in the
31
economic development and thus, the intensity of innovation activity between unequally populated countries (Anokhin & Wincent, 2012). The World Bank provided the data for population size.
Table 3- Summary Variables
Variables Description Source
Innovation Activity Measured by Global Innovation Index (GII)
WIPO
Government Support Measured by R&D expenditure as percentage of GDP
UNESCO Institute for Statistics
Economic Freedom Measured by Economic Freedom Index (EFI)
Heritage Foundation
Type of government system
Dummy variable taking price 1 for democratic and price 0 for non-democratic countries
Economist Intelligence Unit
GDP per capita Control Variable World Bank
R&D expenditure total Control Variable UNESCO Institute for Statistics
Size of population Control Variable World Bank
4.3 Method
This study aims to identify and measure the relationship between government support and innovation activity in developing countries and examine how this relationship is affected by factors like economic freedom. The formulated hypotheses were tested using a quantitative method of analysis which provides for several advantages. Firstly, it generates more reliable, objective and countable results. Additionally, it simplifies a complex problem to a limited number of variables and
32
identifies the relationship among them. Finally, it has been used from many scholars of every scientific sector to test hypotheses (Newman & Benz, 1998).
Specifically, the three hypotheses were tested using a hierarchical multiple linear regression analysis, which is considered as the most valid tool to predict the effect of the independent variable on the dependent-one (Field, 2009). For the regression four models were created which tested the relationship between the dependent variable with a) control variables, b) moderators c) dependent variable and d) interaction effects among them.
5.
Results
In this chapter, the study results are reported. The results include the descriptive statistics of the overall dataset, the correlation analysis and the regression models which test the hypotheses.
5.1 Descriptive Statistics
The initial dataset consisted of 150 observations of 30 developing countries between the years 2010-2014. However, running a first analysis, it was obvious that due to a great number of missing values in R&D expenditure performed by government (N=98) as well as in the total number of R&D expenditure (N=108), the initial sample of the 150 observations was reduced to 97. This huge reduction of the initial sample (<50 cases) limited the strength of the whole study as it could possibly lead to biased results. Furthermore, the control variables presented large variances, resulting in a non-normal distribution. For a normal distribution, it is necessary that the statistic measures, skewness and kurtosis, take prices between -1 to +1 (Field, 2009). Nevertheless, in this first analysis, all three control variables abstained from the accepted values.
33
Consequently, the data underwent some transformations in order to fix the limitations and make the research results stronger and more valid. The first set of transformations concerns the data of R&D expenditure performed by government. Particularly, out of the 98 observations that were available for this variable, 19 were totally excluded (values> 0,22) as outliers i.e. observations with very high values that largely diverge from the mean value (0,14). Then, a new mean value was calculated from the remaining observations (new Mean= 0,09), which consequently replaced the missing values and thus, the sample was enriched with more observations (new valid N=131). This was also the case for the two missing values in GII, which were also replaced by the mean value (33,9). These transformations led to an increase of the sample from 97 observations to 131, a number that is able to provide more valid results for this study. In addition, the control variables were transformed into logarithms, which improved the skewness and kurtosis statistics.
Hence, as seen in Table 4 (see Appendix), the first observation concerns GII for which the maximum value is 56.0 while the minimum is 17.7 and the mean value is 33.09. For the independent variable R&D expenditure by the national government, the minimum value is 0.01 and the maximum value is 0.22, while the mean value reached 0.09. The first moderator, economic freedom, ranged from 46.9 to 79.0 and the mean value was 62.2. The second moderator i.e. the type of government system (whether it is democratic or not) is a binary variable which takes prices 1 for democratic countries and 0 for non-democratic countries. The LogGDP per capita was accounted for each observation with minimum price 3.17 and maximum 7.03. The mean value for this variable was 4.91. The variable LogR&D total takes prices from -1.80 to 0.62 with a mean value -0.40. Lastly, the LogSize ranges between 6.12 and 9.11 with a mean value of 7.30. Moreover, for all the values the statistic measures
34
skewness and kurtosis are normal except from the kurtosis of GII(SMEAN) which is slightly higher (a=1,484) and of Government System (a= -2,012), which is a binary variable.
5.2 Correlation Analysis
After the calculation of descriptive statistics, the correlation analysis was proceeded. The correlation analysis aims to identify whether the relationship between two variables is significant or not. The results of the Pearson’s correlation coefficients retrieved from the SPSS analysis are presented in Table 5. As seen in the table, there is no significant correlation between the dependent and independent variable namely a= .014 (r < .01). However, the GII correlates significantly with both moderators (economic freedom and type of government system), a= .576 and a= .379 respectively, LogSize (a=- .173) and the logarithm of the total number of R&D expenditure (a= .502). In all the aforementioned correlations, the correlation is significant at the 0.01 level (2-tailed) except from LogSize for which it is significant at the 0.05 level. The R&D performed by government correlates significantly also with economic freedom (a= - .232, p< .01), LogR&D total (a= .354, p< .01) and LogGDP per capita (a= - .204, p< .01). Economic freedom presents significant correlation with the type of government system (a= .319), LogGDP per capita (a= .339, p< .01) and LogSize (a=- .297, p< .01). The second moderator correlates significantly with LogGDP per capita (a= .319). Finally, it is observed a significant correlation between LogR&D total and LogSize (a= .355, p< .01).
Table 5- Correlation Coefficients
Mean SD 1 2 3 4 5 6
GII (SMEAN)
33,095 7,0449
35 Econ. Fr. 62,283 7,0341 ,576** -,232** Gov. Syst. ,53 ,501 ,379** -,161 ,319** LogGDPcap 4,9122 1,03198 ,166 -,204* ,339** ,319** LogR&Dtot -,4087 ,47617 ,386** ,354** -,091 -,069 ,104 LogSize 7,3099 ,62131 -,173* ,132 -,297** -,158 -,061 ,355**
**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).
5.3 Regression Analysis
The final step in this research is the regression analysis which will determine the results of the study. However, before running the regression analysis, it is necessary to test a number of assumptions that must be met in order to end up in valid results. These assumptions are: Sample size, Normality, Linearity, Independence and Homoscedasticity. After examining the former assumptions, the formulated hypotheses are tested through a hierarchical multiple regression.
5.3.1 Assumptions
Sample size: an adequate sample size is a prerequisite for a research in order to end up in valid results. There are a lot of contradicting opinions about a proper sample size. In this study, Green’s rules (1991) were used to determine whether the sample size is suitable or not. According to Green (1991), if a researcher needs to test the overall fit of the regression model, then the type that counts the minimum acceptable sample size is: 50+8k, where k represents the number of predictors. On the contrary, in the case that the researcher intends to test the individual predictors within the model, then the type that counts the minimum acceptable sample size is: 104+k,
36
where k represents the number of predictors. In the focal research, in both cases the sample size is adequate as k=6 and 50+8*6=98 < 131 and 104+6=110 < 131.
Normality: a distribution is characterized as normal when it takes the form of a symmetric bell-shaped curve. The standard normal distribution is one with a mean of 0 and a standard deviation of 1 (Garson, 2012). In order to check if the variables in the focal study are normally distributed, it is necessary to measure Skewness and Kurtosis statistics. These measurements are displayed in Table 4 of the Appendix. As seen in this table, the value of skewness for the variables SMEAN (GII) (β= .891), R&D performed by government (β= .563), LogGDP per capita (β= .396) and LogSize (β= .330) is positive which means that the “tail” of the distribution points to the right. On the other hand, the value of skewness for the variables Economic freedom (β= - .298), government system (β= - .139) and LogR&D total (β= - .630) is negative which means that that the “tail” of the distribution points to the left.
Regarding the value of Kurtosis, the variables SMEAN (GII) (β=1.484), R&D performed by government (β= .577), economic freedom (β= .002) and LogR&D total (β= .251) has a positive value which indicates that the distribution has heavier tails and a sharper peak than normal distribution. On the contrary, the variables government system (β=-2.012), LogGDP per capita (β=- .406) and LogSize (β= - .006) are negative which indicates that the distribution has lighter tails and a flatter peak than the normal distribution. Although the values of SMEAN (GII) and the moderator type of government system are higher than the acceptable price, it is considered that this divergence does not affect the study.
Linearity: testing linearity is necessary for determining whether there is a linear relationship between independent and dependent variables or not. A linear
37
relationship between dependent and independent variables is a prerequisite that must be met so as that the regression analysis provides valid results. In this research, linearity was tested with an ANOVA test of linearity. The results of this test are presented in Table 6. As seen in the table the value sig. Deviation from Linearity is lower than 0.05 (0.000 < 0.05) which indicates that the relationship between the two variables, SMEAN (GII) and R&D by government, is linear.
Table 6- Linearity Test ANOVA Table Sum of Squares df Mean Square F Sig. SMEAN(GII) * RDgov Between Groups (Combined) 2276,076 20 113,804 2,998 ,000 Linearity 1,307 1 1,307 ,034 ,853 Deviation from Linearity 2274,769 19 119,725 3,154 ,000 Within Groups 4175,972 110 37,963 Total 6452,048 130
Independence: according to this assumption the residuals must be independent with each other. The independence of the variables used in this study was tested with a Durbin-Watson test. This test indicated that the dependent and independent variables are non-independent as the value of the Durbin-Watson test is 1.190, which is lower than the accepted range of values (1.5-2.0). In this case, the assumption of independence practically is not met, but as the data used for the research are cross-sectional, it is “assumed” that independence is met (Beck, 2006).
Table 7-Independence Test
Model Summaryb Model R R Square Adjusted R Square Std. Error of the Estimate Durbin-Watson 1 ,014a ,000 -,008 7,0715 1,190
38 a. Predictors: (Constant), RDgov
b. Dependent Variable: SMEAN(GII)
Homoscedasticity: for this assumption to be met, the relationship between the dependent and the independent variable must be the same for the entire range of the dependent variable (Garson, 2012). Homoscedasticity has been tested with the use of a scatterplot graph. As seen in the graph (see Appendix), the dependent variable and the residual show a random pattern across the range of the former. This means that the regression model is equally accurate across the range of the dependent (Garson, 2012).
5.3.2 Hierarchical Multiple Linear Regression
Afterwards, a hierarchical multiple regression was conducted to investigate the relationship between the independent variable, i.e. R&D performed by national government of a country, and the dependent variable, i.e. innovation activity, as well as whether the two moderators, namely economic freedom and the type of government system (democratic or non-democratic) affect this relationship, after controlling for GDP per capita, the total number of R&D performed in a country and the size of population.
Considering the model of the current research, it is obvious that four models are needed for the regression. In each model, a new set of predictors is added, starting from the less important predictors and ending up to the main question in investigation of this study and the interaction effects.
In the first model, the three control variables were entered: LogGDP per capita, LogR&D total and LogSize of population. Table 8 shows that the R square in model 1 is 0.268 which means that 26.8% of variance of GII is been accounted by
39
control variables. Furthermore, this model is statistically significant as F (3, 127) =15.460, p < 0.001. Finally, the Beta values of the control variables are LogGDP per capita = 0.094, LogR&D total = 0.498 and LogSize = -0.345. Out of all three, only LogGDP per capita is not significant (β= 0.226, p > 0.005).
Then, in the second model the two moderators were added. This addition leaded to a 31.5% increase of the R square (0.583) which indicates an increased predictive capacity. This model is also statistically significant as F (5, 125) =34.920, p < 0.001. The Beta values show that with the addition of the two moderators the controls’ values changed. Specifically, the Beta values in model 2 are: LogGDP per capita β = -0.134, LogR&D total β = 0.525, LogSize β = -0.183, Economic Freedom β = 0.532 and Type of Government System β = 0.169. Lastly, in this model LogGDP per capita (β = 0.036, p > 0,005) and Type of Government System (β = 0.014, p > 0.005) are not significant.
In model 3, the independent variable, R&D performed by government is also entered in the analysis. It is spotted a negligible increase of 0.1% in the R square value (0.583) which does not enforce predictive capacity. Model 3 is statistically significant as F value is (6, 124) =28.951, p < 0.001. The Beta values in this model are: LogGDP per capita β = -0.140, LogR&D total β = 0.536, LogSize β = -0.185, Economic Freedom β = 0.528 and Government System β = 0.169 and R&D by government β = -0.030. The variables that are not significant in this model are LogGDP per capita (β = 0.033, p > 0,005), Government System (β = 0.014, p > 0.005) and R&Dgov (β = 0.647, p > 0.005). Thus, Hypothesis 1 is not confirmed.
Finally, in the fourth model of the regression analysis the interactions among the independent variable and the two moderators were added. In this final model, the
40
R square had also only an increase of 0.1 % (0.585). The F value is (8, 122) = 21.481, p < 0.001 which is statistically significant. The Beta values in this model are: LogGDP per capita β = -0.144, LogR&D total β = 0.541, LogSize β = -0.189, Economic Freedom β = 0.528 and Government System β = 0.247, R&D by government β = 0.046, R&Dgov * EconFr β = 0.028 and R&Dgov * GovSyst β = -0.091. In this model only LogR&D total (β = 0.000) and Economic Freedom (β = 0.000) are significant. Therefore, Hypotheses 2 and 3 are also rejected.
Table 8- Hierarchical Multiple Regression Analysis Summary
Model 1 Model 2 Model 3 Model 4
LogGDP per capita 0,094 -0,134 -0,140 -0,144
Log R&Dtot 0,498 ** 0,525 ** 0,536 ** 0,541 ** LogSize -0,345 ** -0,183 * -0,185* -0.,189 * Econ. Freedom 0,532 ** 0,528 ** 0,528 ** Type of government system 0,169 0,169 0,247 Main Effect R&D gov. -0,030 0,046 Moderating Effects R&Dgov. x Econ. Freedom -0,028 R&Dgov .x Type of government system -0,091 R square 0,268 0,583 0,583 0,585 Adjusted R square 0,350 0,566 0,563 0,558 R square change 0,268 0,315 0,001 0,001 F 15,460 34,920 28,951 21,481
6.
Discussion
This chapter constitutes a discussion of the aforementioned results. In particular, the study results are further explained and a justification of the reasons why
