Nguồn tri thức nhân tố ảnh hưởng tới đổi mới sáng tạo ở cấp độ doanh nghiệp tại việt nam một nghiên cứu thực nghiệm tt tiếng anh

CHAPTER 1: INTRODUCTION 1.1 Background There is no doubt about the advantages that innovation can bring to a nation OECD (2007) reported strong evidence for a positive relationship between innovation and national competitiveness Innovation plays an essential role in economic growth and is an indispensable factor to contribute to creating more jobs More broadly, both past theory and practice suggest that countries will follow different development trajectories, depending on their ability to identify and capture technological progress Applying new technology translates technological and scientific advances into more productive economic activity As such, it is not surprising that many OECD member countries have adopted national strategies to boost innovation and enhance their economic performance through increased productivity and growth The vital role of innovation is not only affirmed in OECD countries but also developing countries (OECD, 2012) There is a plethora of reasons why innovation is important for developing and emerging countries (OECD, 2012) For example, innovation in agriculture plays an important role in reducing poverty and promoting economic growth Other advantages that innovation can bring to those countries are creating more jobs, improving welfare, access to business opportunities, reaching the world technological frontier in many industries, and especially having the way to avoid “middle income traps” At firm level, the ability to innovate leads to the wealth generation capacity Innovations reduce production cost and improve quality of firms’ goods and services Numerous empirical studies suggest that innovation enhances firm performance because the product of innovation increases firm competitiveness and the process of innovation transforms a firm’s internal capabilities making it more adaptive to change (Neely & Hii, 1998) De Jong and Brouwer (1999) confirmed the same idea in their empirical research with small and medium enterprises (SMEs) Those SMEs work in an environment of changing consumer preferences, increasing competition and changing technological requirements, in order to achieve business goals such as profit and growth, having a continuous flow of successful innovations is crucial for them They build their competitiveness position by offering high quality products and services that match their customers’ demand, so they need a permanent flow of innovations In lieu of the lack of research regarding innovation in developing countries, there is a specific dearth of firm-level innovation studies in those countries (N N Anh, Mai, Nhat, & Chuc, 2011; N N Anh, Ngoc, Chuc, & Nhat, 2008; Linh, Nghia, Dong, & Mai, 2019) Because firms are the dominant economic actors that drive innovation, and particularly the commercialization of innovations, it is necessary to have a study on innovation at the firm level Moreover, with innovation, firms can develop and create more value for the nation such as creating jobs, improving welfare, access to business opportunities, reaching the world technological frontier which could bring social benefits for a country Previous studies have shown that some firms are more capable of developing innovations than others This research attempts to understand the reason behind that and aims to fill in the gap in research in firm level innovation generally and contributes to literature regarding the significance of firm’s innovation determinants by deeply investigating the relationship between innovation and one of the key determinants – knowledge sources for innovation at firm level By doing so, the author expects to provide empirical evidence for policy makers in Vietnam to foster innovation at macro level 1.2 Problem Statement Innovation has received much attention over the last decades as it is considered as a source for economic development Most studies conducted in this area focus on developed countries even though for (firms on) developing countries innovation also is of crucial importance in order to grow and become internationally competitive However, the difference in stages of development can be a barrier when applying the outcomes of studies in developed countries and in developing countries Moreover, most of these studies, also those done in Vietnam, are about understanding innovation process at the macro level They predominantly focus on national innovation systems and government innovation policies As a result, despite its importance, there is very little knowledge about what drives firm level innovation in countries like Vietnam Therefore, this research includes firm level variables in order to analyze the relative importance of different resources that affect firm level innovation and access the capabilities of a firm to utilize these resources Moreover, it is necessary to understand determinants of innovation in Vietnam to provide inputs for evidence-based policy makers Innovation is important as firms with innovation normally develop very fast and hence create lots of job opportunity which we could consider as one of the social benefits for a country Previous studies have shown that some firms are more capable of developing innovations than others This research attempts to understand the reason behind that and aims to fill in the gap in research in firm level innovation generally and contributes to literature regarding the significance of firm’s innovation determinants by deeply investigating the relationship between innovation and one of the key determinants – knowledge sources for innovation at firm level By doing so, the author expects to provide empirical evidence for policy makers in Vietnam to foster innovation at macro level Therefore, the author came up with the topic: “Knowledge Sources as Determinants of Firm Level Innovation in Vietnam: An Empirical Study” 1.3 Study Objective The main objectives of this dissertation are: • To review the literature on innovation and the determinants of innovation especially knowledge sources for innovation at firm level • To analyze the current situation of innovation of manufacturing firms in Vietnam • To identify the knowledge sources that manufacturing firms in Vietnam use for innovating and test the relationship between those knowledge sources and innovation outcome • To propose several solutions for manufacturing firms as well as policy makers in order to foster innovation in Vietnam based on empirical evidence 1.4 Study process 1.5 Scope of Study - Subjects of the research: knowledge sources affecting product innovation at manufacturing firms in Vietnam, including: (1) internal sources, (2) external sources (3) regional sources - Content: Role of each sources affecting product innovation at manufacturing firms in Vietnam - Venue: Manufacturing firms located in four Vietnamese regions, which are the Red River Delta, North Central area and Central coastal, South East and the Mekong River Delta - Timing: Secondary data collected by the World Bank in 2015-2017 1.6 Contribution of Study This research analyses the impact of different knowledge sources of product innovation in Vietnam using firm-level data Hence, it can help to answer the question that whether all kinds of knowledge sources have significant relationship with innovation or only specific one Firms in Vietnam could base on that and decide who and where they should take knowledge from if they intend to innovation Secondly, this study could also be a good evidence for policy makers to understand the determinants of innovation especially the impact of knowledge sources for innovation Innovation related knowledges are public foods and firms could benefit from the diffusion these knowledges Moreover, while most firms in Vietnam are small and medium size, they could lack of resources for conducting innovation To understand what could hinder innovation at firm level means we could provide grounds for new industrial policies and create social benefits for firms which pursuing innovation activities 1.7 Limitation of Study 1.8 Organization of the Dissertation CHAPTER 2: LITERATURE REVIEW This chapter provides an overview of numerous of studies on innovation in Vietnam and the world The author would like to review the research in order to find the gap in literature related to innovation field 2.1 Research on Innovation in the world 2.1.1 Ground theory research The groundwork of modern term “innovation” starts from 1930s, Joseph Schumpeter – a famous economist mentioned that innovation should be distinguished invention, innovation means “development” and “new combinations” (Schumpeter, 1934b) He announced the idea of so-called “new combinations” which refer to “the introduction of a new product or a new quality of a product, a new method of production, a new market, a new source of supply of raw materials or half-manufactured goods” (p.66) Later there are many followers in this field such as Drucker (1985); Eveleens (2010) 2.1.2 Determinant of innovation research There are evidence in literature proved that innovation might be affected by both internal and external factors namely Avermaete et al (2004); Hussen and Çokgezen (2019); Lim (2017) Research on innovation could also be categorized in term of sources for innovation, especially innovation idea Some researchers believed that innovation could only generated by the firms themselves or they follow resources-based view (RBV) However, there are also other line of the literature believed in knowledge-based view (KBV) when they study innovation Table 2.1: List of main innovation research categorized by theory Follow resources-based view (RBV) Research Barney (1991); Bates and Flynn (1995); Tarafdar and Gordon (2007); Terziovski (2010); Bakar and Ahmad (2010) Knowledge-based view (KBV) Cohen & Levinthal (1990); Grant (1996b); Nonaka and Takeuchi (1995); Nonaka and Peltokorpi (2006); Lu et al (2008); Quintane et al (2011); De Castro et al (2011); Isaksson et al (2016) Source: Collected by the author 2.1.3 Product innovation research Studies on product innovation seems to be the most popular in literature as it started very early and various of them available on academic sources (Bakar & Ahmad, 2010; Barasa, Knoben, Vermeulen, Kimuyu, & Kinyanjui, 2017; Chakrabarti, 1974; Cooper & Kleinschmidt, 1986) 2.2 Research on Innovation in Vietnam Studies on innovation in Vietnam starts from early 2000s, however, most of the studies is about innovation in agriculture field (Chairatana & Sinh, 2003; Martin, Castella, Anh, Eguienta, & Hieu, 2004; Spielman & Kelemework, 2009; Van Linh, 2001) The reason behinds it might be because Vietnam had just opened its economy and agriculture still played the main role Recent years, innovation is getting much attention in macro level and other fields as well Some examples are: N N Anh et al (2011); R.-J Lin et al (2013); N N Thang, Quang, and Son (2013); P T T Anh (2014); Tuan et al (2016); Voeten (2016); Vân et al (2018); D K Nguyen, Phong, and Hui (2019); Son, Cung, Thang, and Phong (2019) CHAPTER 3: THEORY FRAMEWORK 3.1 Definition of Innovation 3.2 Types of Innovation Radical innovation Incremental innovation Product innovation Process Innovation Marketing innovation Organizational innovation 3.3 Determinants of Innovation 3.3.1 Following Resources Based view theory 3.3.2 Following Knowledge based view theory 3.3.3 Conclusion Although the RBV struggles that a firm’s inner resources are necessary in nourishing competitive advantage Giving a brief review of the relevant academic literature we can see an expanding trend on resource application research that recommend value can solitary be obtained from resources by using them in a cleverer mode than the rivals (Barasa et al., 2017) Hence, knowing what information available from outside and learning from that seems to be a major influence on firm innovative outcome This is also in line with KBV that the role of knowledge becomes more and more essential Knowledge is believed to be a at most important resource of firms that could influence firm accomplishment (Argote et al., 2000, Oerlemans and Knoben, 2010, Agarwal and Shah, 2014) Current studies also concentrate on the effect of knowledge at nation level and highlight that firms could lessen poverty by leading to the knowledge based-economy (Cooke, 2001, Godin, 2006, Lehrer, 2018) In addition, in Vietnam, right from the 90s of the twentieth century, the Communist Party of Vietnam determined: "The revolution of modern science and technology is taking place strongly, attracting all different countries" Most recently, the Party has determined to associate industrialization and modernization with the development of the knowledge-based economy: "Promote industrialization and modernization with the development of the knowledge-based economy and environmental protection." (Document of the 11th National Congress of the Communist Party of Vietnam), in which the Party affirms, scientific and technological development is really a fundamental driver of the process of rapid and sustainable development Thus, it can be seen from the beginning, our Party attaches great importance to creating motivation for the formation and development of the knowledge economy The development of the knowledge-based economy in Vietnam is a major strategic change: transforming the economy from resource-based to primarily based on human knowledge and creative capacity With available resources, using new knowledge and new technologies to make more, better and more efficient Natural resources are restricted, the capacity to create of human being is limitless Therefore, the author would also want to test the relationship between knowledge sources and innovation in the context of Vietnam to see if it is in line with previous studies 3.4 Knowledge sources A number of authors have emphasized the vital role of knowledge in building and sustaining innovation (Schulze and Hoegl, 2008, Quintane et al., 2011, Leonard-Barton, 1995) In addition, innovation is defined as a knowledgebased commodity Hence, firms need to have knowledge to innovate and thus to profit from innovation (Lundvall, 1992, Lundvall, 1988) Therefore, this study also focuses on the impact of knowledge on innovation In doing so, it uses three different sources of knowledge: internal knowledge sources, collaborative knowledge sources and regional knowledge sources 3.4.1 Internal Knowledge Sources 3.4.2 Collaborative Knowledge Sources 3.4.3 Regional Knowledge Sources 3.5 Hypotheses As explained above, different sources of knowledge can have a different effect on firm-level innovation Following this line of thought, the author hypothesizes that in a transition country like Vietnam, with a weak innovation system, knowledge even plays a more vital role compared to advanced economies Below, the author develop hypotheses that link the different knowledge sources to innovation 3.5.1 Internal knowledge sources Hypothesis 1a: The stronger a firm’s internal R&D, the higher the likelihood that that firm produces a product innovation Hypothesis 1b: The longer time the top manager of a firm working in this sector, the higher the likelihood that that firm produces a product innovation 3.5.2 Collaborative knowledge sources Hypothesis 2a: The stronger a firm’s collaborative knowledge gained from inside the supply chain, the higher the likelihood that that firm produces product innovation Hypothesis 2b: The stronger a firms’ collaborative knowledge gained from outside the supply chain, the higher the likelihood that that firm produces a product innovation 3.5.3 Regional knowledge sources Hypothesis 3a: The stronger the knowledge base of the region a firm is located in the higher the likelihood that that firm produces a product innovation Hypothesis 3b: The higher the population of the region a firm is located in the higher the likelihood that that firm produces a product innovation 3.6 Research model Figure 3.1: Research model with all the variables Internal Knowledge sources - Internal R& D - Managerial Experience Collaborative Knowledge Sources - Knowledge from inside the supply chain (from customers, competitors, suppliers) - Knowledge from outside the supply chain (Universities, Research institutes) Innovation outcome - Product innovation Regional Knowledge Sources - Knowledge benefit from the position of the firm even when firms not enter in any collaboration Source: The author composed and designed CHAPTER 4: METHODOLOGY AND DATA COLLECTION 4.1 Research context Although Vietnam has grown astonishingly in the past three decades, the transition to a flourishing and modern economy has only just begun Vietnam has per capita income still accounting for only a little part of the global average, the country is attempting to maintain its fast-growing path and follow the direction of other successful East Asian economies to join the ranks of high-middle income countries for the past half-decade Even though Vietnam has every potential to achieve this goal, victory is not by itself The population is aging rapidly, which moderate labor productivity Moreover, slow investment growth is weighing on Vietnam's medium-term growth potential How to cope with the resistance of domestic structures? According to the political report of the 10th Central Committee of the Party at the 11th National Congress of the Party, Vietnam needs to steer the external environment changing, where global trade structures are shifting, breakthrough technology, rapid innovation, and Climate change is shaping opportunities and creating new risks for the country Whether Vietnam can continue to maintain previous growth or not, this is extremely important Although growth is not the goal that the government are aiming for, growth is a necessary condition for the wider development It is the basis needed for creating jobs, alleviating poverty and mobilizing resources to invest in health, education and other social goals Vietnam have some advantages that we should not forget which are a geographic location, an open economy, a young and largely labor force in rural areas, and a high level of domestic savings Therefore, the country has conditions for high and sustainable growth Nonetheless, in order to take advantage of these fundamentals, the government needs to regularly focus on policy and institutional reforms targeted at creating increased productivity, effective investment in human capital and physical capital, sustainable and efficient use of natural resources Recently, the big trends are global transforming forces, shaping the future world by profound influence on businesses, society, economy, culture and human life Hence, Vietnam should be able to identify success, analyze and has action to face those major trends to build a long-term dynamic development strategy that contributes to successful economic development due to the fact that Vietnam is likely to be the country most affected by major global trends Vietnam is one of the most open countries in the world, with trade to GDP ratio reaching nearly 200% and FDI inflow accounting for about 7% of GDP in 2018 (GSO, 2019) Regarding geopolitical position and trade structure, Vietnam may be more vulnerable to climate change and US-China trade tensions Therefore, The Central Theoretical Council (2019) suggest four major global trends that Vietnam needs to consider in the next two decades 10 which are population movement, disruptive technology, the rise of China and climate change We could see the vital role of innovation especially breakthrough technology which could be defined as emerging technologies that cause a change in costs or access to products or services, or quickly change the way we collect information, produce or interact operative (Lientz and Rea, 2016) Vietnam are currently in the midst of the 4th industrial revolution, characterized by rapid technological diffusion, multi-technology convergence, and emergence of global platforms Breakthrough technology is often based on technology and digital products, but they can go beyond the connection and potential of the internet This includes modern production methods such as robotics, artificial intelligence, and the internet of things They also include advances in nanotechnology and biotechnology - and new production lines such as batteries, drones, solar panels, self-driving cars, and exotic materials (Lientz and Rea, 2016) The vast opportunities will be associated with innovation and technological breakthroughs that contribute to expanding access to global markets, creating new goods and services, transforming production processes, changing business models, and make a significant impact on development The economic and social transformation that innovation and especially breakthrough technology could bring is that they could accelerate development of a country If a country cannot compete in the future global economy, they will be left behind Hence, to take advantage of the potential of new business models, ways of providing services, and shifting the source of competitiveness, Vietnam need to have a multi-sector and multi-pronged approach to increase opportunities especially foster innovation and technology for the country Assessing innovation policy in Vietnam Even though the Government has had several preferential policies, typically the National Technological Innovation Fund established under the Decision 1342/QD-TTg dated 5/5 8/2013 has the function of preferential loans, loan interest support, capital support for enterprises transferring technology research and innovation However, the sponsorship of such policies and funds for technology innovation is still limited Research results of Luong Minh Huan and Nguyen Thi Thuy Duong (2016) show that enterprises rarely access capital from the state budget to implement technology improvement activities, despite the chapter’s government funds and funds for this issue In addition, the government development fund for enterprises require that local authorities must have reciprocal capital while the locality has no reciprocal capital, so there is no fund This makes the policy seems not to be useful as it should be 4.2 Methodology approach Descriptive approach aims to examine and screen variables in the original research model While econometric approach can verify the logic and rules of observations in the study, helping verify the hypotheses mentioned above To measure the dependent variable, the author uses a dummy variable that takes the value of “1” if a firm has introduced any new or significantly improved innovative product and “0” if otherwise Hence, a binary logistic regression model is chosen for analysing the data 4.2.1 Logistic regression 4.2.2 Innovation research used logit regression 4.3 Data collection The data used in this study is from two main sources: (1) The World Bank Enterprise Survey (ES) conducted between November 2014 and April 2016 and (2) the Innovation Capabilities Survey (ICS) conducted from October 2016 to February 2017 The ES is an ongoing project covering over 155,000 firms in 148 countries, collecting data based on firms’ experiences and enterprises’ perception of the business environment and investment climate This firm-level survey comprises non-agricultural formal, private-sector firms The ICS in this study is a follow-up and complementary to the ES Respondents are randomly selected from the ES sample For Vietnam, 300 manufacturing firms have been included in the sample The ICS focuses on innovative activities and innovative capabilities of manufacturing firms The standardized questionnaires have been translated into local languages and back translated into English to check its accuracy The World Bank uses stratified random sampling as the sampling methodology, which means that all population units are grouped within 11 12 homogeneous groups and simple random samples are selected within each group This method helps to obtain unbiased estimates from different subdivisions of the population with some known level of precision as well as obtain unbiased estimates for the whole population In most cases, stratified sampling is more precise and may produce a smaller bound on the error of estimation than using a simple sampling method The strata for the surveys are firm size, business sector and geographic region within a country The data for this study is merged from the most recent version of the ES and the ICS conducted in Vietnam Unsurprisingly, the data contain missing observations, hence our analyses will use fewer observations than the full sample The data is analyzed by using logit and tobit models Logit models will be applied when the dependent variable is binary (process innovation present yes/no) whereas tobit models will be utilized for censored dependent variables (percentage of sales from new products) 4.4 Variables Table 4.1: Variable measurement Questio Reference Variable Measurement Source n No (Avermaet e et al., 2004); Innovation (Baumann and Kritikos, 2016) Firm introduced any new H3a, Product product or service: "1" Yes ICS H3b, innovation "0" No H3c Internal (Díaz et al., Knowledge 2016) sources Manager Top manager’s number of ES B7 experience working experience year in 13 Internal R&D this sector Dummy variable: "1" Yes "0" No ICS B01 (Caloghiro u et al., 2004) Collaborative knowledge sources Innovation developed with competitors, customers, and supplier: "1" if b1b is Yes, Inside the supply if b1b and b1c or b1j is Yes chain and “3” if all b1b, b1c and b1j are Yes, "0" none of the three is “Yes” Information or ideas from Competitors competitors: "1" Yes "0" No Information or ideas from Suppliers suppliers: "1" Yes "0" No Information or ideas from Customers customers’ feedback: "1" Yes "0" No Information or ideas from Outside the universities and research supply chain institutes: "1" Yes "0" No Regional Knowledge sources % of firms conducting internal R&D within a region Regional R&D using mean of the internal (log) R&D over the regions in Vietnam City with population of less Firm location than 50.000 equal “1” City with population from 14 ICS ICS B1b ICS B1c ICS B1j ICS B1e (Barasa et al., 2017) ICS ES B1 A3 50.000 to less than 250.000 equal “2” City with population from 250.000 to less than million equal “3” City with population over million equal "4" Control variables Age Size Number of years since ES B5 establishment Number of permanent, full ES L1 time employees Source: The author composed and designed CHAPTER 5: RESEARCH RESULTS 5.1 General description of the ES and ICS sample 5.1.1 Distribution of firms by sector and region 5.1.2 Descriptive statistics 5.1.3 Innovation 5.1.3.1 Product innovation 5.1.3.2 Innovation activities 5.1.3.3 Sources of information for innovation 5.1.3.4 Barriers to innovation 5.2 Descriptive statistics of the sample merged from ES and ICS 5.3 Knowledge Sources and Product Innovation A binary logistic regression model is used for the hypotheses Model is a baseline model, in which the author included only control variables to evaluate the independent variables explanatory value The author added internal knowledge sources in Model Model tests the effect of collaborative knowledge sources Model includes the regional knowledge sources Model assesses the effects of all independent variables simultaneously Given that the AIC/BIC indicate that model 15 has the best model fit, the results is mainly interpreted based on this model Table 5.1 and 5.2 report all the results of the models In order to check for multicollinearity, it is required to calculate VIFs The mean of VIF is 1.26, which is well below 10 as are all individual VIFs As such, multicollinearity is not an issue in this study’s data Our results show that the control variables (firm age and size) have no significant association with firms’ likelihood to innovate With regard to the direct effect of internal knowledge sources on innovation, this study finds that internal R&D is positively and statistically significant correlated with firm innovation Hypothesis 1a is supported: a firm’s likelihood to innovate increases when there is an increase in internal R&D On the other hand, for hypothesis 1b the result is positive, but not significant Therefore, the study could not say if managerial experience of firms in Vietnam has any relationship with innovation With regard to the relationship between collaborative knowledge sources and product innovation the results confirm that a firm’s collaborative knowledge gained from inside the supply chain (customers, suppliers, competitors) is positively related to with product innovation of that firm On the other hand, it shows no significant relationship between collaboration with universities or research institutes and innovation Khanna and Palepu (2005) mention that in developed economies, firms can rely on a variety of institutions to minimize market failure, while firms in emerging markets are, for example, confronted with institutional voids, i.e weak linkages between firms and universities and/or research institutes As such, hypothesis 2a is strongly supported, while there is no support for hypothesis 2b Table 5.1: Logistic Regression Results of each individual knowledge source models Age (log) Size (log) Manager Model DV= Product Innovation == B SE P>|z| -0.05 0.23 0.83 -0.11 0.17 0.49 0.02 0.02 0.42 Model DV= Product Innovation == B SE P>|z| -0.00 0.31 0.99 -0.11 0.16 0.50 16 Model DV= Product Innovation == B SE P>|z| 0.04 0.24 0.86 -0.02 0.14 0.91 experience Internal R&D Inside supply chain knowledge Outside supply chain knowledge Regional R&D Firm Location Constant 2.06*** -0.64 Observations Prob> Chi2 Pseudo R-Square AIC BIC 0.24 0.75 284.00 0.00 0.09 344.77 355.72 0.00 0.40 1.04*** 0.30 0.00 0.22 0.21 0.30 0.32 0.00 1.71*** 284.00 0.00 0.26 281.04 291.98 0.01 0.13 0.02 0.09 0.64 0.13 -1.21 1.09 0.27 284.00 0.00 0.00 377.82 388.77 For regional knowledge sources the study finds no significant relationship between regional R&D, product innovation Hence, the author could not accept hypothesis 3a However, the location of firms is significant positively correlated with product innovation Firms in a city with larger populations are likely to produce more product innovation than their counterparts in less crowded cities That might be explained by the fact that in big cities more facilities and infrastructure are available for firms to utilize Moreover, in densely populated cities, firms might be able to find more suitable personnel who bring new knowledge (Glaeser and Mare, 2001) As such, hypothesis 3b is accepted Table 5.2: Logistic Regression Results of the all independent variables simultaneously model Age (log) Size (log) Model DV= Product Innovation == B SE P>|z| -0.09 0.25 0.73 -0.13 0.19 0.50 17 Manager experience 0.03 0.02 0.20 Internal R&D 1.46*** 0.21 0.00 Inside supply chain 1.09*** 0.34 0.00 knowledge Outside supply chain 0.39 0.94 0.03 knowledge Regional R&D -0.03 0.03 0.26 Firm Location 0.27* 0.15 0.07 Constant -2.63*** 0.32 0.00 Observations 284.00 Prob> Chi2 0.00 Pseudo R-Square 0.31 AIC 261.56 BIC 272.51 As mentioned in our theoretical background several studies from the open innovation literature, suggest that specific sources of knowledge might be more relevant than others and that over-search of external knowledge might be detrimental to innovation (Laursen and Salter, 2006, Bayona-Saez et al., 2017, Hung and Chou, 2013) To probe the existence of such effects in our setting, the author performed two additional explorative analyses Specifically, the author checked the impact of the knowledge sources separately and tested for an inverted U-shape effect The results of these analyses are reported in Table 5.15 With regard to the former, this study found that only knowledge from customers is positively and significantly associated with product innovation This is in line with the study of Doloreux and Lord-Tarte (2013), which emphasizes that the tastes of customers are paramount and customer ideas are a highly important source of information for product development With regard to the latter, the results in Table 5.3, which are plotted in Figure 5.1, demonstrate an inverted Ushape relationship between knowledge from inside the supply chain and product innovation Moreover, the author tested whether the point estimate for the highest value of knowledge use differs significantly from the second highest score and find a significant different (p=0.003) This indicates that the downward sloping 18 part of the inverted U-shape is statistically significant and that this study truly finds negative relationship of high levels of usage of knowledge from the supply chain That means searching for knowledge from collaboration is important, but too much openness might have a negative impact on product innovation The author finds this correlation highly surprising given the low-tech environment that she studies and the limited number of different knowledge sources that this study distinguishes The author will get back to the implications of this finding in the discussion section below Table 5.3: Robustness tests Age (log) Size (log) Manager experience Internal R&D Inside supply chain 3.68*** 1.09 knowledge Inside supply chain 0.27 knowledge0.86*** squared Knowledge from competitors Knowledge from suppliers Knowledge from customers 0.00 0.00 Model DV= Product Innovation == B SE P>|z| -0.13 0.24 0.58 -0.15 0.18 0.41 Model DV= Product Innovation == B SE P>|z| -0.09 0.24 0.72 -0.19 0.17 0.24 0.02 0.02 0.02 0.31 1.45*** 0.10 0.00 3.88*** 1.25 0.30 0.90*** 0.03 0.40 1.30*** 0.03 0.00 0.00 0.00 0.57 0.44 0.20 0.34 0.00 2.34*** 284.00 0.00 Constant 0.45 0.57 0.42 -0.05 0.04 0.22 0.27** 0.14 0.05 0.57 0.00 3.08*** 284.00 0.00 Observations Prob> Chi2 Pseudo R0.34 0.39 Square AIC 253.19 233.71 BIC 264.14 244.65 All reported standard errors are robust clustered standard errors at the regional level * p