The effect of innovation on export performance and the role of home-country business environment in transition economies

The effect of innovation on export performance and

the role of home-country business environment in

transition economies

Master Thesis

By

Alessio Ercole

S2946408

a.ercole@student.rug.nl

Rijksuniversiteit Groningen - Faculty of Economics and Business

MSc. International Business and Management

January 23

_{, 2017}

2

ABSTRACT

Innovation has been extensively demonstrated to be a determinant of export performance in open market contexts. The purpose of this research is to test the persistence of this relationship in transition economies, as the radical institutional changes particularly challenge firms’ activities and strategies. Moreover, we make use of the idiosyncratic context to investigate the potential moderating influence of the business environment on the relationship. Three aspects are considered: corruption, access to finance and intellectual property right protection. The analysis is conducted using a Tobit estimation model on data retrieved from the BEEPS database, which provide information on more than 15,000 firms from 30 former Communist countries. The findings confirm the positive effect of innovation on export performance within this context, and the positive moderating role of access to finance. While no support is found for corruption, intellectual property right protection, in contrast with our expectations, negatively moderate the relationship.

3

ACKNOWLEDGEMENTS

4

TABLE OF CONTENTS

2.4 Innovation and export performance ... 15

2.5 Business environment ... 18

2.5.1 Corruption ... 19

2.5.2 Access to finance ... 20

2.5.3. Intellectual property rights ... 22

2.6 Conceptual model ... 24

3. METHODOLOGY ... 25

3.1 Data source ... 25

3.2 Sample ... 26

3.3 Measures and variables ... 27

3.3.1 Independent variable ... 27 3.3.2 Dependent variable ... 28 3.3.3 Moderating variables ... 28 3.3.4 Control variables ... 29 3.4 Empirical approach ... 30 3.4.1 Model description ... 30 3.4.2 Preliminary analysis ... 31 4. RESULTS ... 32 4.1 Descriptive statistics ... 32 4.2 Baseline results ... 33 4.3 Robustness checks ... 35 5. DISCUSSION ... 37 6. CONCLUSIONS... 41 6.1 Theoretical implications ... 41

6.2 Managerial and policy-maker implications ... 42

6.3 Limitations and suggestions for further research ... 44

REFERENCES ... 46

5

LIST OF TABLES

LIST OF FIGURES

Table 1 Sample description 26

Table 2 Descriptive statistics 32

Table 3 Results of tobit estimation model 34

Table A1 LR test for heteroscedasticity 57

Table A2 VIF test for multicollinearity 57

Table A3 Correlation matrix 58

Table A4 Firm size robustness test 59

Table A5 EU membership robustness test 60

Table A6 Non-EU membership robustness test 61

Table A7

Robustness test with alternative measures of moderating

variables 62

Table A8 BEEPS 2012-2014 relevant questions 63

6

1. INTRODUCTION

Innovation has been regarded as “the life blood of corporate survival and growth” (Zahra & Covin, 1994, p.183). Due to constantly evolving technology and international competition, the value of existing products and services is continuously at risk. Therefore, enterprises have increased their innovative efforts as a corporate strategy to survive and have success in the domestic and global market. In addition, innovativeness represents a strategic orientation that helps firms to gain a competitive advantage over competitors, in terms of both price and quality (Gunday, Ulusoy, Kilic, & Alpkan, 2011). Moreover, because customer needs are continuously changing, it is generally accepted that all firms should innovate regardless of their size or sector in order to compete and survive in the market (Atalay, Anafarta, & Sarvan, 2013).

In business studies, not many issues have had as much agreement as the innovativeness being a determinant of firm performance (Rubera & Kirca, 2012). Similarly, the positive effect of innovation on export performance, conceptualized as the “success of a firm’s international

transactions in terms of its overall operations” (White, Griffith, & Ryans, 1998, p.193), has

been thoroughly confirmed. This is grounded on two main rationales. Firstly, export markets involve higher costs (i.e. entry or transportation costs) than domestic ones. Based on the resource-based view and the heterogeneity of firms, only the more productive firms self-select themselves into exporting. In particular, innovativeness is regarded as one of the main sources of higher productivity (Bernard & Jensen, 1999; Cassiman & Golovko, 2011). The second explanation is drawn from the “product life cycle” theory. In the early stage of new product, firms aim to maximize their revenues. However, since in the early innovation stage the domestic market is restricted, firms expand abroad in order to exploit their market power (Cassiman & Martinez-Ros, 2007; Hirsch & Bijaoui, 1985).

7 institutions affect the degree to which firms can take advantage of their innovative activity. For instance, the effectiveness of the regulatory system, and especially the enforcement of property rights, are likely to moderate the returns that firms can capture from innovation (Teece, 1986). Nevertheless, little attention has been given to whether the “rules of the game” might increase or decrease the effect on innovation on exporting.

This research would aim to fill several gaps in the literature. Firstly, we want to add value to the existing theory about the innovation-exports causality by investigating whether this relationship persists in the context of Central and Eastern European transition economies. These countries provide an interesting setting where to test established theories, since the transition process confers them a series of unique characteristics. The radical change from central planning to market competition has left a profound mark on the behaviour of firms, which are deeply conditioned by the ongoing and rapid changes of institutions (Meyer & Peng, 2005). However, effects of this dynamic environment have heterogeneously affected firms. While some companies have successfully adapted to the new market structure, others have failed to benefit from the transition to a market-oriented economy. This might be due to the planned economy heritage that still influences corporate-level activities and strategies (Boermans & Roelfsema, 2015). According to Meyer and Peng (2005), little light has been shed on which resources can contribute to establish a competitive advantage in a transition environment. The necessity to investigate is further highlighted by the fact that context-specific resources, such as business networks (Peng & Heath, 1996) and strategic flexibility (Uhlenbruck, Meyer, & Hitt, 2003), have been demonstrated to help prevail over competitors in these countries. Thus, it would be interesting to investigate whether, within this area, innovation represents a determinant of export performance, or whether the unstable institutional environment compromises the appropriability of returns to innovation.

8 environment: corruption, access to finance and protection of intellectual property rights. In fact, as Batra, Kaufmann, and Stone (2003) and Ayyagari, Demirgüç-Kunt, and Maksimovic (2008) have verified through firm-level data, these aspects represent the main obstacles to firms’ growth. We argue that these variables play a critical role in the innovation – exports relationship. Additionally, they adequately reflect the constraints that the idiosyncratic context of transition economies provides. For example, Teece (1986) shows that property rights enforcement hinders the possibility for enterprises to take advantage of their innovative activities. Moreover, Li and Ferreira (2011), examining the institutional quality of transition economies, take into examination the regulatory, political (corruption) and financial system, given the underdevelopment of these institutions in these countries. Therefore, our research question is: “Does innovation increase export performance in transition economies? Is this

relationship moderated by the local business environment?”

Our dataset is based on "EBRD-World Bank Business Environment and Enterprise Performance Survey (BEEPS), 2012". Although institutions have mostly been examined at the country-level due to lack of reliable data, this database allows to examine voids actually perceived by enterprises, increasing the relevancy of our study. In fact, some of previous studies have suffered from one particular drawback. Cross-country data is based on de jure regulations, therefore giving insight only into costs that firms would have to deal with if the situation was homogeneous among all enterprises. However, different firms face different constraints, making the use of subjective data fundamental (Aterido, Hallward-Driemeier, & Pagés, 2011). Data on intellectual property protection are retrieved from the Intellectual Property Right Index, which effectively combines both de facto and de jure measures. The heterogeneity in terms of institutional development at the country-level, registered within the sample, adds to the generalizability of our results to other contexts. In fact, our sample includes low-income economies of Central Asia, characterised by significant institutional voids, as well as rich Central European countries that, as members of the EU, have a fully developed market system (De Rosa, Gooroochurn, & Görg, 2010). Since a large portion of firms in our sample shows no exporting, Tobit estimation is utilised to test the hypotheses. This model has the main advantage of incorporating zero values in the estimation as the result of the decision about whether or not to export. Therefore, it is suitable for dual decision making process, that is the choice to whether to sell abroad and, if so, how much (Gashi, Iraj, & Geoff, 2014).

10

2. THEORETICAL BACKGROUND

In this section, the relevant literature for this study is presented. First, the concepts of innovation and export performance are defined, followed by an overview of the main features of transition economies. Next, the innovation-export performance relationship is discussed, both in terms of the main theories underlying it and the potential issues concerning the existence of this relationship in a transition economy context. Then, the potential influence of the business environment on the relationship is introduced. The aspects of the environment that are considered in this study, namely corruption, access to finance and intellectual property rights protection are hence explained and singularly combined with the main relationship in order to develop the hypotheses.

2.1 Innovation

Since innovation has been examined in several fields, from technology to business or economics, being therefore described from different standpoints, several definitions are provided (Baregheh, Rowley, & Sambrook, 2009). From a general perspective, innovation concerns the implementation of an idea that is new to the adopting organisation (Damanpour, 1996). Most of existing literature has mainly adopted the technological view on innovation, thus applying this concept only to the introduction of new products or processes. However, we argue that it is important to extend innovation even to non-technological modernisations (Schmidt & Rammer, 2007). These include both organisational, which regards the implementation of a new managerial technique, process or structure to purse organisational goals (Azar & Ciabuschi, in press), and marketing innovation, which involves “new product

design or packaging, product placement, product promotion or pricing” (OECD / Eurostat,

11

product (good or service), or process, a new marketing method, or a new organisational method in business practices, workplace organisation or external relations” (p.46).

Several studies have regarded internal research and development (R&D) activities as a reliable indicator of firm- or industry-level innovation (Raymond & St-Pierre, 2010). However, we argue that R&D expenditures capture only one dimension of the inputs required by innovation, but does not necessarily lead to the introduction of the new innovation. For instance, Parisi, Schiantarelli, and Sembenelli (2006) do not find any significant association between R&D spending and the introduction of new processes, as this might mainly be correlated with expenditures on fixed capital. Additionally, firms can choose market mechanisms as a mean to generate innovations. This is made possible by R&D outsourcing, technological consultancy or technological collaboration with other companies or public institutions (i.e. universities) (Santamaría, Nieto, & Barge-Gil, 2009). Finally, investments in innovation are not limited to R&D spending, but also involve expenditures for introducing a new technology into the market, new capital equipment or employee coaching (Hall, 2010).

12 complementary use of other assets (Teece, 1986) . For instance, without a high-quality manufacturing or marketing department, profiting from R&D would be significantly harder.

2.2 Export performance

An important decision related to the internationalization strategy of a firm is which entry mode to select for a foreign market (Agarwal & Ramaswami, 1992). The main modes of entry available to enterprises are exporting, contractual agreements, joint ventures, acquisitions, and greenfield investments. The choice between these five options has mostly been grounded on the transaction cost theory. Depending on the costs for finding, negotiating and monitoring a potential partner, firms tend to rely or on the market or on more hierarchical modes to deliver their products (Brouthers, 2002). Additionally, when evaluating each entry mode, managers appear to consider three main variables: risk, return and degree of control (Kumar & Subramanian, 1997). Exporting comprises the “physical transfer of goods from the firm to the

foreign market with or without an agent in exchange for the value of the goods in monetary terms” (Kumar & Subramanian, 1997, p.54). It is a low resource commitment and a low risk

entry mode. On the other hand, it is also associated with a low profit return and provides little control to the firm, which might hinder the possibility to improve its competitive power and to maximize the returns (Agarwal and Ramaswami, 1992). However, low control can also be beneficial for firms. These can take advantage of scale economies, while not having to deal with bureaucratic hinders associated with other modes of entry. Moreover, the enforcement of contracts can be problematic in the presence of a high external uncertainty or underdeveloped institutional environment. Finally, when expanding to low potential markets, exporting is found to reduce the risk of failure (Agarwal & Ramaswami, 1992; Anderson & Weitz, 1986) .

Katsikeas, Leonidou, and Morgan (2000), in a review of the existing literature, find three different groups of variables that directly influence export performance. Managerial

aspects embody all the characteristics of decision makers involved in the export process. Organisational characteristics include human and other critical resources, operating elements,

13 when analysing determinants of export performance. They demonstrate that export performance is strongly influenced by market-oriented capabilities of the firms, which consist on learning from customers, competitors and external environments and using this information to have success. However, foreign markets entail several obstacles, such as the institutional environment of the host country, which affects how the exporting firm will do business. Thus, the institutional distance between home and host county is found to moderate the effect that internal capabilities has on export performance. Additionally, the local business environment of the home country can be thought to moderate the relationship analysed, which is something we will discuss after offering an overview of transition economies.

2.3 Transition economies

The fall of Communism in Central and Eastern European Economies has been the starting point of a radical change, in terms of social, political and economic perspectives. This event has influenced 29 countries in Eastern Europe and the former Soviet Union. However, although showing similar features, the process of evolving from a centrally-planned to a market economy in Central and Eastern European economies begun under different conditions. While Eastern Europe countries were struck by waves of mostly peaceful revolutions in 1989, the collapse of Soviet Union in 1991 was the event that led the countries within the Soviet bloc to gain their independence (De Melo, Denizer, Gelb, & Tenev, 2001).

14 corruption, which instead aims to shape the implementation of these rules (Hellman, Jones, & Kaufmann, 2000). These two factors are likely to have hampered both firms’ and aggregate performance. Additionally, the EU membership has led certain countries to implement and enforce laws that have stimulated the local economy (Roland, 2002).

This weak institutional environment has affected both firms’ innovation activities and export performance. Although some transition countries have made important progresses in terms of economic development, innovation capabilities have not been successfully developed. This results in a notable gap between R&D expenditure of Central and Eastern European countries that have joined the EU and the rest of EU countries (Leskovar-Spacapan & Bastic, 2007). This distance can be explained by political reforms that have not successfully changed the managers’ attitude towards innovation. Therefore, the low motivation to enhance quality, the lack of competences, education or training and the responsibility avoidances are negatively hampering R&D expenditures in former centrally-planned economies (Leskovar-Spacapan & Bastic, 2007). The correlation between regulatory environment and under development of innovative activities is confirmed by Krammer (2009). According to his study, in transition economies, a more efficient bureaucratic system and a stronger intellectual property rights would be likely to lead to a substantial growth in R&D activities of firms.

Increasing exports has been a necessity for firms in transition economies (Shinkle & Kriauciunas, 2010). Firstly, after the collapse of the Communist system, enterprises could start selling their products abroad directly, as state-owned export agencies ceased to exist. In addition, the dramatic fall in domestic demand made the expansion to a foreign market a priority for firms in these countries. However, we should take into consideration the strong variability among countries. Kaminski, Wang and Winters (1996) claim that different conditions at the beginning of the reform period, geographical location and diverse market access to OECD countries have not allowed each country to equally benefit from the switch from a centrally planned to a market economy, especially in terms of exporting.

15 adapt to structural institutional changes (Meyer & Peng, 2005). Consequently, theories that are established in developed countries have demonstrated that are not fully applicable to this area. For instance, although privatization has often been linked to better performances, Aussenegg and Jelic (2007) find that privatised firms in transition economies did not achieve greater efficiency, profitability, capital investments or output. The reason for this is that an institutional void is more likely to lead to stagnation and decapitalisation rather than improvements in performance (Nellis, 1999).

2.4 Innovation and export performance

The critical role of innovation as determinant of export performance has received much attention in the literature. A significant amount of research has confirmed the validity of this causality at the firm-, industry- and country-level in both developed (Lages, Silva, & Styles, 2009; Rodríguez & Rodríguez, 2005; Roper & Love, 2002; Wakelin, 1998) and developing countries (Guan & Ma, 2003; Montobbio & Rampa, 2005). Most of these studies have drawn their hypotheses from the resource-based view of the firm, which gives emphasis to the heterogeneousness of enterprises. Due to their different endowments of resources, firms are able to generate a competitive advantage in the long term (Barney, 2001). Innovation is one of the resources that more contribute to a competitive advantage. In fact, it is characterised by a high degree of specificity, which limits its transferability to other firms, and complexity, as it is difficult to identify the factors associated with its generation (Rodríguez & Rodríguez, 2005). Grounding on this theory, innovative firms can have higher export performance for two reasons. Firstly, they can develop more efficient productive processes. Melitz (2003) builds a model of industry dynamics of firms with diverse endowment of resources, and so different productivity levels. In this model, only the most productive firms can overcome sunk costs of entry to export markets. Similarly, Bernard and Jensen (1999) claim that ‘‘good firms become

exporters’’ (p.2). Since operating in foreign markets increases firms’ costs, in terms of

16 2005). Cavusgil and Zou (1994), for example, find that the ability of enterprises to introduce new products increases its relevance when operating in foreign markets, since product management has to take the variety of demands in international markets into account. To examine which factor has a more significant influence on export behaviour for Spanish manufacturing firms, Caldera (2010) discovers that product upgrading has a greater impact than cost reduction. However, the product life-cycle approach to trade also underpins the causality. This theory maintains that, since the domestic market is limited in the early stage of the product cycle, innovative firms have incentives to expand into export markets to exploit their market power derived from innovation (Hirsch & Bijaoui, 1985). Consequently, their export performance is higher due to the exploitation of the market power achieved through the new technology.

Azar and Ciabuschi (In press) also prove the positive effect of organisational innovation on export performance. They contend that technological innovation does not improve export performance by itself. In contrast, it necessitates the complementary adaptation of managerial system to fully benefit from the new product or process. Introducing an organisational innovation would also enhance the organisational climate, teamwork or information sharing. Thus, by enhancing organisational climate, which is associated with labour productivity, or reducing costs of supplies, firms can take advantage of better export performance (Gunday et al., 2011).

17 distribution contracts and product line. However, this support vanished after the privatisation. Firms have so been left dealing with strategic decisions, institutional voids and external competitors on their own, despite the lack of technical and managerial capabilities. In particular, former state-owned firms found themselves in a context of market-based economy, yet with a set of resources suitable for satisfying the needs of a socialist economy. For example, the innovation effectiveness is supported and stimulated by flexible work arrangements. Therefore, the implementation of a new technology, in transition countries, might be hindered by the traditional organisational system grounded on the Fordist mass production. This might prevent firms to fully benefit from their innovative activities (Uhlenbruck et al., 2003).

18

Hyp1: In transition economies, innovative firms have better export performance

2.5 Business environment

Although the relation between innovation and export performance has received much attention in the literature, previous studies have not exhaustively examined the potential moderating effect of home country characteristics. An exception to this is the research of Yi et al. (2013), who discover that foreign ownership, government relationship, affiliation to a business group and regional marketisation shape the effect of innovativeness on export intensity. Additionally, López-Bazo and Motellón Corral (2013) demonstrate that the increase in the propensity of exporting due to innovation is larger in regions where the extensive margin of exports is high. Moreover, Prajogo (2016) find the environmental dynamism and competitiveness to increase the relationship between product and process innovation and firm performance. Consequently, there is a significant gap in the literature in relation to whether the business environment influence the effects of innovation on export performance. However, Meyer and Peng (2005) explain that, especially within the idiosyncratic context of transition economies, it is necessary to integrate the institution-based with the resource-based view. In fact, particular conditions influence which resources sustain a competitive advantage. Analysing the moderating effect of institutions in transition economies adds value to our research. In fact, this context has been characterised by an ‘institutional divergence’, or ‘great divide’ (Berglof & Bolton, 2002). While some countries have adopted modern standards and regulations, others are stuck in the ‘development trap’, where inefficient political institutions, such as high levels of corruption or bureaucracy inefficiencies, hurdle economic development (Sonin, 2013).

The business environment is made of “the myriad forces which are beyond the control

of management in the short run”. In particular, a critical role in shaping this environment is

19 and Martinez (2000) focus on government repudiation of contracts, risk of expropriation, corruption, rule of law, and bureaucratic quality. Also, Khanna and Palepu (1997) study product, capital, and labour markets, legal and regulatory system.

Following a literature review based on the nature of this study, we will rely on the research of Ayyagari et al. (2008). This study finds three particular aspects of the business environment to principally constrain firms’ growth: corruption, access to finance and property rights protection. Due to our focus on innovation, we concentrate on intellectual property rights. This choice is also motivated by our focus on transition economies: corruption, during the planned regime, used to be contemplated as an acceptable mean when dealing with rigid bureaucratic system (Iwasaki & Suzuki, 2012). Access to finance was also critical, due to the absence of any financial market in the former Communist bloc (Coricelli & Masten, 2004). Finally, the transition stage has led to one of the biggest privatisation of property rights in the history. Although previous studies have found few examples of countries with poor institutions but positive business environments (Gillanders & Whelan, 2014), we maintain that our chosen variables are good proxies of the institutional quality of legal (intellectual property rights), political (corruption) and financial (access to credit) system (Li & Ferreira, 2011).

2.5.1 Corruption

Corruption can be defined as “an act in which the power of public office is used for

personal gain in a manner that contravenes the rules of the game” (Jain, 2001, p.73). The most

common form of corruption, and the form we focus on in this study, is bribery. This is represented by payments made by companies to government officials in order to receive a favour, such as the granting of a licence (Cleveland, Favo, Frecka, & Owens, 2009). We argue that corruption has a primary role in the effect that innovation has on exports, as innovative firms are found to be more subject to rent-seeking behaviours. The need for having access to business permits, tax documents or other government-supplied resources increases the number of their contacts with officials. Thus, their perceived corruption is expected to be significantly higher than non-innovative firms (Murphy, Shleifer, & Vishny, 1993).

20 bribery helps agents circumvent these distortions. In contrast, the first-best option would be to remove the distortions themselves. On the other hand, a high number of studies have pointed out that corruption sands the wheels of growth, by significantly obstructing firms’ activities. The main argument is that the occurrence of bribery may alter resource allocation by increasing the returns to rent seeking compared to those of productive activities (Baumol, 1996). Dal Bó and Rossi (2007) discover a strong relationship between bribes and inefficiency at the firm level, since in more corrupt countries more labour is needed in order to produce a given level of output.

Given the damaging effect of corruption on firms’ operations, we claim that corruption might negatively affect the positive influence of innovation on export performance for two reasons. Firstly, the increase in productivity allowed by process innovations might be absorbed by the payment of bribes. Innovative firms would lose their competitive advantage in respect to their competitors. Thus, the level of productivity after the payment of bribe might not be high enough to overcome sunk costs and entry barriers imposed by foreign markets. Ayyagari, Demirgüç-Kunt, and Maksimovic (2010) prove that innovative firms, despite paying higher amount of bribes than non-innovative firms, do not receive better services from officials. As a result, in a corrupted environment, innovative firms might see their position with respect to non-innovative ones compromised, so that they do not have enough abilities and productivity to manage exports. Secondly, an extremely corrupted environment may persuade enterprises to reduce interactions with government, which is necessary in order to obtain licenses for operating abroad. Rather, they would prefer to expand more slowly or operate in the informal sector (De Rosa et al., 2010). Hence, we can expect that innovative firms in high-corrupted contexts may prefer to exploit their competitive advantage in the domestic market, rather than being more exposed to bribes by exporting. Following these considerations, we formulate the following hypothesis:

Hyp2: In transition economies, higher corruption decreases the positive effect of innovation on export performance

2.5.2 Access to finance

Access to finance “generally refers to the availability of quality financial services at reasonable

costs” (Barr, Kumar, & Litan, 2007, p.11) The dependence of the level of access to finance of

21 literature (Berman & Héricourt, 2010). Financial institutions are responsible for transferring funds from savers to investors. In particular, they have to overcome the obstacles of agency costs, adverse selection and asymmetric information associated with financing contracts and supervising the pooling of resources from savers to borrowers. However, dimensions of access to finance comprise both supply and demand of credit. While financial institutes only have a direct effect on the supply, the request of credit by firms might be affected by absence of financial knowledge and social trust or a burdensome paperwork requested (Aggarwal & Goodell, 2014).

Access to credit has a strong effect on economic growth. In fact, it influences the birth of new enterprises and therefore the competition that, in turn, increases the quality of products, the employment rate and the standard of living in general (Aggarwal & Goodell, 2014). Additionally, Ayyagari et al. (2008) reveal that financial constraint is the biggest obstacle to firm growth, regardless of the countries and firms considered. Previous literature has deeply shown that access to external finance influence the innovative activity of firms (Ayyagari, Demirgüç-Kunt, & Maksimovic, 2012; Kortum & Lerner, 2000). However, not much attention has been paid to whether external financing constraints affect the possibility to take advantage of innovative activity. It is important to state that innovative firms are subject to financial constraints to a much larger extent. In fact, financial institutions are less able to evaluate their activities, because either these are based on intangible assets or firms keep them secret to prevent other firms to copy ideas. Moreover, returns to innovation might be uncertain, as new products not always achieve the expected success. Finally, new technologies are mainly context-specific, therefore cannot be used outside the firm itself (Lee, Sameen, & Cowling, 2015; Magri, 2009).

22 some are still suffering from institutional voids, others have been able to develop significantly their credit market (Coricelli & Masten, 2004). Similarly, Konings, Rizov, and Vandenbussche (2003) find that there is a significant variance in access to credit in this area, as some firms can benefit from softer budget constraints, due to tax allowances or “preferential” bank credits.

We argue that good access to finance increases the effect that innovation has on export performance. In fact, innovators have better opportunities of exporting into foreign markets, due to their higher productivity and quality products. However, they might have to deal with the finance gap, that is defined as the lack of capitals to fund a profitable investment chance (Mina, Lahr, & Hughes, 2013). Therefore, a good access to finance can help innovative firms to exploit their advantages in a foreign market. Financial loans are indeed necessary for overcoming costs associated with starting operating abroad, such as international shipping, that have to be faced before making revenues (Contessi & De Nicola, 2012). Moreover, Melitz (2003) provides a model where firms with higher productivity self-select into exporting. However, Manova (2013) integrates this model with the access to finance variable. He demonstrates that in a context of low access to finance, despite every firm above a certain productivity level can start selling abroad, only most successful firms can export at their best without being affected by credit constraints. In contrast, firms with productivity below a certain level will decide to limit their exports, in order to reduce their need of external capital. Following these considerations, we can hypothesise:

Hyp3: In transition economies, higher access to finance increases the positive effect of innovation on export performance

2.5.3. Intellectual property rights

Intellectual property rights can be defined as “government-protected rights granted to

an inventor or creator to exclude others from using the technology or product in question”

23 public revelation of what is protected, they increase the amount of openly available knowledge. Finally, they incentivise the development of technology markets, as patents or trade secrets give the possibility to trade protected knowledge through formal contracts. Some studies argue that a strong protection hinders growth, due to the restricted diffusion of knowledge, along with permitting innovators to obtain a monopoly power, thus deterring potential entry into a certain industry. However, Gould and Gruben (1996), in a cross-country study, demonstrate that stronger intellectual property protection promotes the economic growth and does not hamper the entrance in a market. Additionally, although new technology results to be cheaper, a weak enforcement of property rights makes the acquisition of innovation much more difficult. Finally, more innovative firms from developed countries limit their investments in countries with a weak property rights protection, so that these can less benefits from the advantages of globalisation.

According to the resource-based view, certain resources allow firms to gain a competitive advantage over the competitors, due to their value, rareness, non-substitutability and non-imitability (Barney, 1991). Consequently, the subsistence of a market power depends also on the degree to which a firm can protect its innovations from imitation (González-Álvarez & Nieto-Antolín, 2007). In order for innovative firms to make profits from their innovative activity, patents are fundamental. Innovators can keep the control over the new technology, as well as decide to transfer the control to others, either by selling or leasing it (Lunn, 1985). Previous studies have found that patents and other types of property rights allow innovators to appropriate the biggest portion of the technology value, while a weak enforcement limit the possibility to capture rents (Anton & Yao, 1994).

Within the former Communist regimes, despite some countries still suffering from consistent institutional weaknesses, several transition countries have toughened the protection of intellectual property rights after the ratification of the TRIPs Agreement (Della Malva & Santarelli, 2016). In addition, the implementation of property rights reforms has been complementary to open market policies. Opening up the economy to FDI aims to increase the flow of international technology available to domestic firms. Within a context of strong protection of property rights, multinationals from developed countries are more prone to diffuse their technology. Thus, we might consider market liberalisation and tighter regime of intellectual property rights as two complementary reforms (Maskus, 2000).

24 imitate technology implemented by a certain firm. The competitive advantage that innovators can take advantage of might therefore be lost, since imitators can still achieve the same levels of productivity or quality. As Teece (2006) theorised, the limited intellectual property protection prevents pioneers from obtaining the returns of their technologies, since the first mover advantage is likely to be lost.

Therefore, innovative firms will benefit more from better export performance with a strict regime of appropriability. Thus, we can theorise:

Hyp4: In transition economies, higher intellectual property right protection increases the positive effect of innovation on export performance

2.6 Conceptual model

25

3. METHODOLOGY

This section offers an insight into the method used to test the hypotheses presented in the theoretical chapter. A description of data source and the final sample is given. This is followed by an explanation of how the variables of this research are measured. Finally, the estimation model used in this research, namely the Tobit model, is illustrated.

3.1 Data source

Our data source is mainly based on the 2012-2014 EBRD-World Bank Business Environment and Enterprise Performance Survey (BEEPS)”, which has featured several relevant studies in the literature (S. Lee & Weng, 2013; McCann & Bahl, 2016). Since the aim of this survey was to analyse the idiosyncratic business environment in transition economies, its usage is appropriate for answering our research question.

The fifth wave of BEEPS survey is chosen because this is the first one to utilise the definition of innovation suggested by the Oslo Manual, thus comprising product, process, marketing and organisational innovation. Moreover, as direct information about innovation are included, there is no need to rely on indirect proxies (i.e. R&D intensity) for our independent variable. The usage of subjective data also adds value to our research. In fact, these become particularly effective when evaluating the influence of the business environment on firms’ processes and performance, as economic agents ground actions and decisions on their perceptions of external constraints (Kaufmann, Kraay, & Mastruzzi, 2011). In particular, objective data tend to significantly vary from the de facto situation, especially when related to external condition. For instance, some aspects of governance, such as corruption or financial constraints, often do not leave any trail on paper, thus being difficult to be objectively measured. On the other hand, information collected through questionnaires might be subjected to perception bias. At the country level, this might originate from different cultural norms and diverse degrees of political freedom, which influence the choice of specific ratings. However, Fries, Lysenko, and Polanec (2003) have checked for potential perception bias in the BEEPS 2002, but have not found any impartiality. Since the BEEPS 2009 follows a similar approach, we expect that the results of our analysis will not be compromised.

26 the Property Rights Alliance, and in particular with the Intellectual Property Rights (IPR) section, as previous studies have already done (Balsmeier & Delanote, 2015; Della Malva & Santarelli, 2016). This index draws on several different sources. These are opinions of experts on actual enforcement on these rights, objective data regarding participation in international treaties, restriction or length of the protection and the Office of the US Trade Representative “Special 301” list, which analyses to what extent piracy rates are present in business software. Therefore, consistent with BEEPS structure, de facto measures are taken into account (Aggarwal & Goodell, 2014).

3.2 Sample

Country Freq. Percent Cum.

Albania 360 2.58 2.58 Armenia 360 2.58 5.16 Azerbaijan 390 2.8 7.96 Bosnia Herzegovina 360 2.58 10.54 Bulgaria 293 2.1 12.65 Croatia 360 2.58 15.23 Czech Republic 254 1.82 17.05 Estonia 273 1.96 19.01 FYR Macedonia 360 2.58 21.59 Georgia 360 2.58 24.17 Hungary 310 2.22 26.4 Kazakhstan 600 4.3 30.7 Latvia 336 2.41 33.11 Lithuania 270 1.94 35.05 Moldova 360 2.58 37.63 Montenegro 150 1.08 38.7 Poland 542 3.89 42.59 Romania 540 3.87 46.46 Russia 4,220 30.27 76.73 Serbia 360 2.58 79.31 Slovak Republic 268 1.92 81.24 Slovenia 270 1.94 83.17 Turkey 1,344 9.64 92.81 Ukraine 1,002 7.19 100 Total 13,942 100

27 BEEPS data has been gathered through surveys conducted by the World Bank and the European Bank for Reconstruction and Development in 15,883 firms of 30 European and Central Asian transition countries between 2012 and 2013. The subset of enterprises was selected through a stratified random sampling, in order to provide a representative sample of the firms in the region, in terms of age, size or location. Moreover, in each country, the industry composition of the sample was based on its comparative contribution to GDP. In the BEEPS survey, interviewers evaluated to what degree respondents were providing trustworthy information: inferring from McCann and Bahl (2016), we will exclude observations if interviewees were considered as not truthful (n=107). However, since the IPR index does not cover Belarus, Georgia, Kirgiz Republic, Mongolia, Tajikistan and Uzbekistan, these countries will therefore be excluded from our analysis, reducing the final dataset to 13,942 observations. Table A1 illustrates the composition of our sample.

3.3 Measures and variables

3.3.1 Independent variable

Innovation. To measure innovation, we refer to Gorodnichenko and Schnitzer (2013) and

28 the reliability of data when gathering information in a large set of countries (Bergmann, Müller, & Schrettle, 2014).

3.3.2 Dependent variable

Export performance: We will measure this variable by quantifying the percentage of the total

sales that is exported directly. As Katsikeas et al. (2000) argue, this is the most widely recognised measure of export performance in the literature. In fact, beside ruling out the size effect, it mirrors the success of the international transactions of the firm in respect of its overall domestic sales (White et al., 1998). Since we are interested in investigating whether innovation helps to overcome entry and operating costs of foreign markets, we have decided not to include indirect exporting. In fact, indirect intermediaries allow firms to reduce costs of internationalising, such as those correlated with negotiation in distant markets (Hessels & Terjesen, 2010). Hence, due to the nature of our study, we will quantify export performance on the basis of the firms’ answer to the question “In fiscal year [insert last complete fiscal year],

what percentage of this establishment’s sales were direct exports?”.

3.3.3 Moderating variables

Corruption is constructed on the basis of the answer to the question about the frequency of informal payments/gifts to get things done: never (=0), seldom (=1), sometimes (=2), frequently

(=3), very frequently (=4) and always (=5). The reliability of self-reported measures of bribes might be questionable, as corruption relates to illicit activities. However, as Ayyagari et al. (2010) maintain, the data collection of the BEEPS aimed to obtain as reliable answers as possible. For instance, government officials did not participate to the data gathering. Also, since identities of firms and managers were confidential information, interviewees did not run any risk of being legally prosecuted. Finally, several papers, such as the one of Hallward-Driemeier and Aterido (2009), have found a strong correlation between firms’ answers about obstacles in the business environment in BEEPS survey and objective data from other sources.

Access to finance. This variable is captured by the question: “To what degree is Access to Finance an obstacle to the current operations of this establishment?”. Higher values

29

Intellectual property rights protection is quantified by the Intellectual Property Rights

Index, which ranges from 0, associated with a weak IPR regime, to 10, that represents the ideally strongest protection context. The final ratio is determined by the average of three sub-measures, namely (1) an evaluation of the IPR regime made by experts, (2) a measure of patent protection associated with the Ginarte-Park Index of Patent Rights and (3) a quantification of the strength of the IPR enforcement on the basis of the volume of copyright piracy.

3.3.4 Control variables

In this study, we control for firm-, industry- and country-level variables. First, we control for

firm size. Larger firms tend to have more resources that can be invested in certain activities,

such as expansion to international markets. Thus, due to the fixed costs associated with exporting, such as dealing with additional uncertainty or gathering information, large enterprises might take advantage of economies of production and marketing (Rodríguez & Rodríguez, 2005; Wakelin, 1998). The lack of related data prevents us from using continuous variable. Thus, following the approach of Beck, Demirguc-Kunt, & Maksimovic (2005), we generate a discontinuous variable equal to 1 for small (less than 20 employees), 2 for medium (between 20 and 100 employees) and 3 for large firms (more than 100 employees). Also, we include the firm age, since older firms have accumulated experience and knowledge that are found to allow for higher export performance (Yi et al., 2013). This is quantified by the difference between the beginning of firms’ operations and the year of each interview. Moreover, we build a dichotomous indicator for foreign ownership. In fact, the higher experience in operating in foreign markets is found to stimulate indeed the exporting (Rodríguez & Rodríguez, 2005). An enterprise is considered as foreign if more than 10% of it is owned by a private foreign individual, company or organisation.

We also control for the potential industry variability by creating a sector dummy. Regarding the country-level, we do not use of country dummies as these would also capture the variability in the country-level intellectual property protection regime. Thus, four country-level dummies are included. Market size is captured by the natural log of GDP: by operating in a bigger market, firms develop certain capabilities that can be exploited by internationalizing (Deng, 2004). Additionally, the degree to which a market is open to export, measured by the

total exports as a percentage of GDP, is expected to have an influence on our dependent

30 implement market liberalisation policies. Hence, firms located in EU countries might be favoured in terms of less barriers to exports (Awokuse, 2007). Furthermore, Lall (1980) has demonstrated the level of human competence to be correlated with either internationalisation propensity and intensity of American firms. Thus, we also control for the percentage of labour

force with tertiary education.

3.4 Empirical approach

3.4.1 Model description

In a review of the previous studies concerning the determinants of export performance, Sterlacchini (1999) argues that the shortcoming of several researches lies in having used OLS estimates. In fact, when the export share variable has most of observations in its lower limit, then OLS regression is biased, since it might predict values of export performance outside its actual range. Moreover, this distribution compromises the assumption of linearity (Amemiya, 1984). Thus, as the same author suggests, a censored Tobit estimate has to be utilised.

31

3.4.2 Preliminary analysis

As Wooldridge (2002) contends, in order for Tobit model to provide consistent results, the assumptions of normality and homoscedasticity have to be met. First, we maintain that, according to the central limit theorem, if the sample consists of a significantly high number of observations, no attention has to be paid to distribution of data (Ghasemi & Zahediasl, 2012). Considering that our sample consists of more than 13,000 observations, the central limit theorem is found to apply and issues of normality can therefore be disregarded. Heteroscedasticity is found when the values that a dependant variable take are inhomogeneous across the range of values of its predictors. Inferring from Engle (1984), a Likelihood Ratio (LR) test is performed. Table A1 in the Appendix displays the results. Since Prob>chi2 is significant, the null hypothesis of homoscedasticity is rejected. Therefore, robust standard errors will be used.

In addition, multicollinearity involves the dependence of explanatory variables from one another. This would compromise both the specification and the estimation of the relationship analysed through the regression (Farrar & Glauber, 1967). Therefore, to check that multicollinearity does not reduce the consistency of our results, we examine the correlations among our variables by inspecting the Variance Inflation Factor. According to Graham, 2003, multicollinearity is detected when VIFs > 10. Thus, we can conclude that our explanatory variables do not suffer from this issue (see Table A2 in the Appendix).

32

4. RESULTS

The next section presents the empirical results of our analysis. The chapter starts with the descriptive statistics. After that, the results of our estimation model are provided. Finally, robustness checks are performed in order to test the consistency of our findings.

4.1 Descriptive statistics

In the table below, main statistics of our variables are presented.

The data suggests that a significant number of total firms have introduced at least one between product, process, organisational or marketing innovation (5,557). Moreover, a relatively low average percentage of direct export over total sales is registered (6.53), thus showing that local markets seem to be the main area of operation for most firms in transition economies. With regards to moderating variables, corruption seems not to strongly hit firms, as roughly half of total interviewees have answered that is never common to have to pay some

Variable Obs Frequency Mean Std. Dev. Min Max

Innovation 13,942 0=8,387 1=5,557 0.3985798 .4896235 0 1 Export 13,782 6.530982 19.96037 0 100 Accessfin 13,655 0=817 1=1,660 2=2,443 3=2,338 4=6,347 2.863274 1.284484 0 4 Corruption 12,730 0=6,590 1=2,394 2=2,336 3=955 4=244 5=211 0.93967 1.198097 0 5 IPRindex 13,942 4.651198 1.119898 2.3 6.9 firm_age 13,788 14.46729 11.68866 0 174 firm_size 13,942 1=7,190 2=5,128 3=1,624 1.600775 0.6876362 1 3 foreign_own 13,942 0=13,071 1=871 0.062473 0.2420215 0 1 logGDP 13,942 26.11401 1.953919 22.13125 28.40582 logExpo 13,942 24.19504 5.051317 3.64154 27.11047 TertEdu 13,944 35.00854 15.61164 13.3 55.5 EUmember 13,942 0=10,226 1=3,716 0.266533 0.4421618 0 1

33 “additional gifts” to get things done. Similarly, very few managers consider access to finance as a major (817) or very severe (1,660) hurdle to their firms’ growth. The statistics of intellectual property right index illustrate how institutional development vary across single countries, since the minimum value is significantly lower than highest one (2.3 in comparison to 6.9, on a scale from 1 to 10). In relation to control variables, firms have an average age of 14 years. Furthermore, most of the dataset regards small and medium enterprises, since only 1,624 of the respondents are managers of large firms, and domestically-owned firms, with only one-quarter of the total being foreign-owned. Additionally, three-one-quarter of enterprises are settled in countries not members of EU. In addition, the table A3 in the Appendix displays the correlation between our main variables.

4.2 Baseline results

Table 3 shows the results of the Tobit estimation. Since only pseudo-R2 _{is given, we cannot} evaluate the extent to which the model predicts the variability. AsVeall & Zimmermann (1996) contend, while pseudo R2 can be used to make comparisons across different estimation models, no conclusions can be drawn when a singular type of estimation is utilised. Therefore, pseudo-R2 is not included in the findings.

Model 1 examines the direct effect of innovation on exporting. The most critical result regards the statistically significant effect (p<0.01) of innovation on export performance, thus confirming the validity of an established relationship even in a particular context as transition economies. As expected, firm size (firm_size), firm age (firm_age) and foreign-ownership (foreign_own) are strong predictors (p<0.01) of export performance, due to their assets or higher experience with internationalising; A similar significance is registered, at the country-level, by membership to European Union (EUmember). In contrast, GDP (logGDP) is negative strongly significantly correlated with export performance. A significant result, but also with a sign different from our expectations, is registered for the level of human skills in the country (TertEdu). Given these results, we find support for hypothesis 1.

34 interacted with innovation, corruption has a negative relationship with exporting. However, since the coefficient is not statistically significant, we have to reject hypothesis 2.

In model 3, the results of the moderating effect of access to finance (accessfin) on innovation-export performance are displayed. Access to finance is not a direct significant predictor variable of export performance: this diverges from previous studies that demonstrated

(1) (2) (3) (4)

VARIABLES Export Export Export Export

Innovation 14.34*** 12.61*** 8.380** 27.86*** (1.593) (2.046) (3.832) (6.656) foreign_own 36.34*** 35.70*** 36.59*** 35.58*** (2.482) (2.582) (2.494) (2.478) firm_size 18.48*** 17.73*** 18.22*** 19.33*** (1.121) (1.160) (1.124) (1.119) firm_age 0.212*** 0.225*** 0.218*** 0.156*** (0.0550) (0.0569) (0.0554) (0.0554) logGDP -1.914*** -1.664** -1.960*** -9.311*** (0.684) (0.704) (0.685) (0.899) EUmember 22.57*** 22.81*** 22.37*** -12.07*** (1.867) (1.937) (1.871) (3.651) logExpo 0.144 0.114 0.141 0.380* (0.217) (0.219) (0.217) (0.227) TertEdu -1.174*** -1.177*** -1.169*** -0.985*** (0.0739) (0.0759) (0.0741) (0.0722) Corruption -1.834 (1.153) Innovation*corruption 1.480 (1.483) Accessfinance -0.902 (0.916) Innovation*accessfinance 2.145* (1.219) IPRindex 18.65*** (1.594) Innovation*IPRindex -2.602** (1.324) Constant -20.40 -22.51 -15.98 81.43*** (14.40) (14.96) (14.63) (16.52) Sigma 57.69*** 56.89*** 57.31*** 57.18*** (0.969) (0.997) (0.966) (0.965)

Industry dummies YES YES YES YES

Observations 13,638 12,470 13,367 13,638

Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

35 how firms without financial constraints can benefit from higher export performance. In addition, the coefficient of the interacting effect of innovation with access to finance is marginally significant (p<0.1). Hence, we find support for hypothesis 3.

In model 4, we have considered the moderating effect of intellectual property right protection (IPRindex). In this case, the moderator is found to be a strong significant predictor (p<0.01) of export performance. Thus, a strong protection local regime is found to promote the international expansion of firms. Furthermore, the interaction between the IPR index and innovation is negative significant at the 90% level. This entails that an interaction of innovation and consistent protection of property rights hinders exports rather than favouring them. Consequently, hypothesis 4 has to be rejected.

4.3 Robustness checks

Robustness checks are executed in order to test how regression coefficient behaves when the original model undergoes some modifications, especially in terms of removal or addition of regressors or investigation of sub-samples. These sensitivity analyses help detect misspecification in the original estimation model (Lu & White, 2014).

Firstly, firm size can be argued to have an influence on both internationalisation and ability to cope with constraining business environment. In fact, firms with limited assets, might face more obstacles when it comes to expanding abroad or can find institutional constraints as more binding in comparison to big enterprises (Audretsch & Elston, 2002). Therefore, we decide to test our regressions only on small and medium businesses, namely those with less than 100 employees, removing firm size as a control variable. Results are shown in table A4 in the Appendix. As before, the effect of innovation on exporting is strongly significant (p<0.01). In model 2, corruption significantly hinders exports (p<0.1), while its interaction with innovation is insignificant. Moreover, while the interacting effect of access to finance is statistically significant (p<0.05), the one of intellectual property right protection is strongly significant (p<0.05), but with a negative coefficient, thus contradicting our original hypothesis. Since this further test reflects the findings of the original sample, we can contend that our results are robust to firm size.

36 Market, strengthen or weaken the sensibility of the innovation-exports relationship to the influence of aspects of business environment. The affiliation to the EU has also allowed some firms to conform their regulatory system with that already established in European developed market economies. Consequently, institutional constraints might be less binding in European countries (Manolova, Eunni, & Gyoshev, 2008). Table A5 in the Appending displays the results of the estimation model with the European countries subsample, while table A6 of the one of countries outside the EU. While our results for the non-EU members are equal to those previously originated, within EU countries none of the three interactions between innovation and the three moderators is significant. Therefore, we might presume that our results are sensitive to the EU membership.

37

5. DISCUSSION

The aim of this research is to investigate whether the positive relationship between innovation and export performance, persists even in an idiosyncratic context such as the one of transition economies. Moreover, the specificity of this environment has provided the opportunity to investigate whether this relationship is moderated by three aspects of the business environment, namely corruption, access to finance and intellectual property rights protection. In the remainder part of this chapter, we analyse more in depth the results of our analysis.

Hypothesis 1 is confirmed, thus supporting the assumption that, in transition economies, innovative firms can benefit from better export performance. This result extends the already consistent literature about the significance of the relationship (Rodríguez & Rodríguez, 2005; Wakelin, 1998; Yi et al., 2013). Uhlenbruck et al. (2003) argued that the collapse of the Soviet Union, and the following privatisation of most of the firms, left local managers dealing with strategic decisions, despite lacking abilities critical to operate in an open market. In the light of what has been found, we can argue that local firms seem to have managed to overcome the obstacles that the transition stage involves, at least those hampering the development of a competitive advantage exploitable in international markets. As expected, we have found foreign ownership, firm size and firm age to be positively associated with firms’ export performance. In contrast, two peculiar results need further attention. Both level of human skills, measured by the percentage of employees with tertiary education, and market size are found negatively correlated with firms’ export performance. We maintain that the nature of exports plays a critical role in the first case. In fact, firms in this context, when exporting, are likely to compete on price rather than quality. Thus, since employees with tertiary education require higher salaries, and therefore increasing costs, firm in countries with a lower educational level tend to export more. Additionally, the significant negative effect of market size (logGDP) on exporting can be explained by the fact that, larger domestic markets allow enterprises to exploit their competitive advantage in the home country, reducing the need for internationalising.

38 our hypothesis might be due to measurement bias. In fact, we have focused on a particular aspect of corruption, namely informal payments required when dealing with officers’. Nevertheless, corruption is a much broader concept that cannot be quantified in one single definition. As Pellegata and Memoli (2016) explain, corruption has to be related to the overall confidence of citizens in institutions, as only this way it would be possible to effectively measure the extent to which the rent-seeking behaviour of politicians or officials influence the firms’ behaviour. Additionally, subjective measures, although capturing those aspects that do not leave trail on paper, are not exempt from flaws. Firstly, less educated people or those living in rural parts are demonstrated to be not as sensitive to corruption as other people; secondly, interviewees are likely to generalise or overestimate based on their own experiences (Pellegata & Memoli, 2016). Further studies should therefore measure corruption in a broader sense, by also considering state capture or trustworthiness of legal system, or triangulate experience-based data with objective ones.

39 Finally, the results of our last hypothesis have turned out to be unanticipated. A stronger intellectual property right protection regime seems to negatively affect the innovation-exporting relationship, which is the outcome opposite to that expected. In other words, the more intellectual rights are protected, the less innovative firms can take advantage of better export performance. Though, intellectual property rights should provide firms with the certainty that they can appropriate their innovative efforts. However, as Meyer and Peng (2005) maintain, transition economies provide an idiosyncratic context where the ongoing institutional change affect firms in a way different from that expected in an open market context. In addition, strengthening patent laws have already been found to have an ambiguous influence on trade flows in a significant number of studies (Maskus & Penubarti, 1995). One potential explanation might stem from the fact that governments in developing countries have strengthened IPR in order to attract FDI from developed countries (Tanaka & Iwaisako, 2014). Since only the most productive firms self-select themselves into foreign markets, in countries with a tight protection regime, the larger presence of MNEs might have eroded the competitive position of local firms that, despite their innovative efforts, still lag behind multinationals in terms of productivity. Another reason could be associated with the determinants of exporting in countries affected by significant institutional voids. Witt and Lewin (2007) demonstrate that, the misalignment between firms’ needs and home country institutional development, increases the likelihood that firms internationalise, thus considering FDI as an escape response. Therefore, in our case, once IPRs have been tightened, innovative firms might not have felt the need to escape from a negative environment, but rather they have preferred to exploit their competitive advantage in the home-country. However, further research should shed more light on this finding. Our estimation model showed also the positive significant effect of intellectual property rights on export performance. This finding adds empirical support to the theoretical model developed by Yang and Maskus (2009), that contend that stronger IPR protection regime in developing countries spurs know-how transfer from developed countries. The minor marginal cost raises the amount transferred, thus making firms from emerging countries more competitive when operating abroad.

41

6. CONCLUSIONS

The environmental uncertainty and the radical institutional changes, which have been affecting transition economies during the last 20 years, have already challenged and rejected several theories that were taken for granted in market economies (Meyer & Peng, 2005). Nevertheless, in this study, we have demonstrated the significance of the innovation-export performance relationship even in such a particular context. Thus, we can conclude that innovation represents a critical factor of firms’ survival and success regardless of the type of economy in which a firm operates. Moreover, we have shed more light on the potential contingency of this relationship in transition economies. In this case, however, the results have reflected the unpredictability of behaviours of constructs in an ongoing transition process. While innovative firms with higher access to finance are more likely to take advantage of better export performance, no support has been found for the moderating role of corruption. Furthermore, an unanticipated finding is related to the negative effect of interaction of strong intellectual property right protection with innovativeness on export performance. However, all these results are found to be strongly contingent on the level of marketization of the country. In fact, when testing our hypotheses only on EU members, we have discovered the effect of innovation on export performance to not be significantly moderated by the business environment. It seems clear that transition countries that have joined the EU have had the opportunity of both opening up their economies and promoting and internal institutional development. In the following paragraphs, drawing on our results, we define theoretical and practical implications of our results, as well as describe the limitations of our study and provide ideas for future research.

6.1 Theoretical implications