Knowledge spillover, sectoral innovation and firm total factor productivity The case of manufacturing industries in Vietnam.Knowledge spillover, sectoral innovation and firm total factor productivity The case of manufacturing industries in Vietnam.Knowledge spillover, sectoral innovation and firm total factor productivity The case of manufacturing industries in Vietnam.Knowledge spillover, sectoral innovation and firm total factor productivity The case of manufacturing industries in Vietnam.Knowledge spillover, sectoral innovation and firm total factor productivity The case of manufacturing industries in Vietnam.
INTRODUCTION
PROBLEM STATEMENT
1.1.1 The significance of the research’s topic
1.1.1.1 The importance of the research the role of knowledge spillovers on innovation at sector level
Innovation and knowledge spillover are crucial for the economic development of countries, particularly in developing nations Theoretical frameworks, such as endogenous growth and new growth theories, highlight the significance of innovation and technological change as key drivers of long-term growth.
Innovation often stems from investments in knowledge and the resulting knowledge spillovers Griliches (1992) highlighted that such investments tend to benefit third-party firms that can commercialize these ideas without bearing the full costs of their development Furthermore, scholars like Romer (1986), Lucas (1988, 1993), and Grossman and Helpman (1991) emphasized that knowledge spillovers play a crucial role in driving endogenous growth.
Innovation plays a crucial role in the industrial success of four Asian NICs, yet Vietnam's Global Innovation Index score of 35.4, ranking 59th out of 128 countries, highlights its need for improvement Compared to ASEAN peers like Singapore (7th), Malaysia (33rd), and Thailand (48th), Vietnam must focus on enhancing its innovation capacity to drive economic growth Policymakers should prioritize understanding knowledge spillover mechanisms to foster innovation and promote sustainable development.
The capacity for innovation at the sector level, along with channels for knowledge spillovers, plays a crucial role in shaping effective policies for growth driven by innovation Analyzing sectors provides valuable insights into the relationship between knowledge accumulation and diffusion Mehrizi and Ve (2008) highlighted that sector-level analysis connects firm-level factors to macroeconomic conditions, while Malerba (2002) emphasized its importance in examining innovative and production activities Padoan (1999) also noted that a sectoral perspective allows for a deeper exploration of knowledge accumulation and diffusion processes.
The innovation capacity of a firm or sector is derived not only from its own knowledge and technology but also from knowledge spillovers from other firms and sectors, as highlighted by Aghion and Jaravel.
Innovations within a firm or sector often rely on knowledge generated by other firms or sectors, leading to a spillover effect that can diminish the incentive for companies to invest in Research and Development (R&D) for innovation However, recent studies by economists Cohen and Levithal (1989) and Onodera (2009) suggest that R&D activities can actually enhance a firm's absorptive capacity This, in turn, contributes to upgrading not only the firm itself but also its entire sector through internal spillovers, as well as benefiting other sectors through external spillovers.
Research on the influence of knowledge spillover channels on sector innovation capacity is limited While many studies have focused on how sector characteristics affect innovation (Castellacci, 2008; Yurtseven & Tandoğan, 2012; Hage et al., 2013; Piqueres et al., 2015), only a few recent investigations have explored the impact of R&D spillover (Autant-Bernard and Lesage, 2011; Ang and Madsen, 2013; Malerba et al., 2013) Additionally, there is a scarcity of studies that consider both sectoral and regional factors in relation to spillover effects (Buerger and Cantner, 2011; Aiello and Ricotta).
Vietnam, there are few studies on innovation and most of these studies focused on firm level (Jordan, 2015; Doan Thi Hong Van and Bui Le Nhat Uyen, 2017;
This study examines the impact of knowledge spillover on sectoral innovation, focusing on three key channels: research and development (R&D), foreign direct investment (FDI), and trade activities Utilizing spatial regression models, the research aims to uncover how these factors contribute to innovation across different sectors.
1.1.1.2 The importance of the research on heterogeneity of firms’ TFP in considering both firms’ characteristics and spillover effects from sectors and regions
Sustainable economic development cannot rely solely on capital and labor; it requires Total Factor Productivity (TFP), which represents the output not attributed to these inputs The Solow model (1956) identifies this residual as a key driver of technical change essential for sustainable growth Countries like Korea and Singapore have experienced significant TFP growth alongside their development For instance, Korea's TFP growth was 8.3% from 1970 to 1980, but surged to 31.5% from 1980 to 1990, highlighting the importance of enhancing TFP as a central focus of development policy.
To derive valuable policy implications for development, it is essential to analyze the determinants of Total Factor Productivity (TFP) Technological progress is a key driver of sustained growth, as highlighted by Solow (1956) However, Acemoglu (2009) emphasizes that variations in TFP are not solely attributable to technology; even firms using identical technologies can exhibit different efficiencies in their application Capeda and Ramos (2015) further suggest that TFP can be influenced by factors such as economies of scale and resource combination methods, indicating that even firms with similar technologies can experience differing TFP levels due to sector characteristics and geographical location.
The variation in firms' total factor productivity (TFP) primarily stems from differences in their characteristics, including size, production technology, and human capital Additionally, sector-specific characteristics and external economies of scale significantly influence firm performance As noted by Krugman and Obstfeld (2009), regional concentration of industries can generate positive externalities, such as specialized suppliers, labor market pooling, and knowledge spillovers Furthermore, the geographical environment impacts firms, as regions vary in the quantity and quality of resources available, which can either benefit or hinder performance This localization of firms is evident in successful clusters like Silicon Valley and New York's financial sector, where proximity allows businesses to capitalize on external economies.
Examining the determinants of firms' Total Factor Productivity (TFP) through a multilevel cross-classified model is crucial, as it allows for the isolation of impacts from various factors, including firm, sectoral, and regional dimensions This multi-level analysis effectively addresses error correlation among firms operating at similar levels, which single-condition models often overlook By utilizing the entire sample size without differentiating levels, single-condition models typically underestimate variance; in contrast, the multilevel cross-classified model accurately assesses variance at each level, providing a more comprehensive understanding of TFP determinants.
However, most of studies on firms’ TFP focused on the determinants as firms’ characteristics (Sjoholm 1999; Blalock and Veloso, 2006; Waldkirch and Ofosu, 2010; Lopez, 2008; Baptist and Teal, 2014; Fernandes, 2008; Seker and Saliola,
2018) In Vietnam, studies on TFP are still very limited (CIEM, 2010) although
Total Factor Productivity (TFP) is increasingly recognized as a crucial factor in assessing development quality This study introduces a novel approach to analyzing TFP in Vietnam through the application of a multileveled cross-classified model Furthermore, the findings may inform policy recommendations that benefit not only individual firms but also entire sectors and regions.
1.1.1.3 The importance of research on the role of knowledge spillovers on sectoral innovation and firms’ TFP in the manufacturing industries in Vietnam
The manufacturing industry has significantly impacted the Vietnamese economy, contributing 18.35 percent to the GDP and emerging as the leading sector in 2016 Additionally, from 2010 to 2016, it ranked second in labor participation, following only the agriculture, forestry, and fishing sectors.
2016) In comparison to other countries in the region, the output growth of manufacturing sector in Vietnam was in the top five at 9.3% in the period from
2000 to 2014 Moreover, the manufacturing in Vietnam was also the leading sector with the biggest contribution share to output growth at 26,98%.
Despite its significant role in the economy, Vietnam's manufacturing sector experienced low labor productivity growth of just 2.7% from 2000 to 2014, ranking only above Brunei and Indonesia, and lagging far behind China's impressive 7.6% Additionally, the percentage of enterprises adopting modifications to existing technology was notably low at 0.87% between 2010 and 2014 While innovation activities, including securing patents and developing new technologies, showed a slightly higher engagement rate, only 9.02% of enterprises participated in such initiatives during the same period.
RESEARCH OBJECTIVES AND QUESTIONS
This study aims to explore how knowledge spillover influences innovation within various sectors The innovation capacity of a firm or sector is derived not only from its own knowledge and technology but also from the insights and advancements of other firms and sectors, highlighting the significance of knowledge spillovers.
2015) The first general objective is to investigate channels of knowledge spillovers on sectoral innovation in manufacturing industries in Vietnam, the study focuses on the following research questions:
1.1 Is sectoral innovation directly affected by R&D activities of that sector in manufacturing industries in Vietnam?
1.2 Is sectoral innovation indirectly affected by R&D activities of other sectors in manufacturing industries in Vietnam?
1.3 Is sectoral innovation directly affected by transactions with FDI enterprises in that sector in manufacturing industries in Vietnam?
1.4 Is sectoral innovation indirectly affected by transactions with FDI enterprises in other sectors in manufacturing industries in Vietnam?
1.5 Is sectoral innovation directly affected by trade activities in that sector in manufacturing industries in Vietnam?
1.6 Is sectoral innovation indirectly affected by trade activities in other sectors in manufacturing industries in Vietnam?
This study aims to explore how firm-level, regional, and sectoral characteristics influence total factor productivity (TFP) in firms Key research questions will guide the investigation into these impacts.
2.1 How much heterogeneity in firms’ total factor productivity is explained by firm-level, sector-level and province-level determinants?
2.2 Does firms’ size have impact on firms’ TFP in manufacturing industries in Vietnam?
2.3 Does the capital intensity in firms have impact on their TFP?
2.4 Is there difference in TFP of exported firms and non-exported firms? 2.5 Is firms’ TFP affected by their sectoral innovation in manufacturing industries in Vietnam?
2.6 Does the human resource in a province have impact on the TFP of firms in that province?
RESEARCH METHODOLOGY and RESEARCH SCOPE
This study investigates the impact of three channels of knowledge spillovers on sector innovation capacity by utilizing Spatial Regression analysis Additionally, it employs a cross-classified model to assess the variations in firm productivity influenced by sector-specific, regional, and firm-level determinants The research leverages data from the Vietnam Enterprises Survey (VES) and the Vietnam Technology and Competitiveness Survey (TCS), along with Input-Output (I/O) data.
Vietnam in 2012 Besides, the study also uses the annually surveyed data on province of General Statistics Office (GSO).
The analysis focuses on the impact of R&D, FDI, and trade on sectoral innovation, using sector as the primary unit of investigation This unit is derived from data on Vietnamese manufacturing firms spanning from 2010 to 2014 The relationships between sectors are established through intermediary transactions within Vietnam's Input-Output framework.
2012 By spatial regression model, the study finds the direct as well as indirect impact of R&D, FDI and trade on sectoral innovation.
This study analyzes the impact of firm-level, regional, and sectoral characteristics on total factor productivity (TFP) in Vietnamese manufacturing firms from 2011 to 2014 Utilizing TCS and VES data, the research evaluates firm-specific attributes and incorporates sectoral characteristics derived from this data Additionally, annual province-level data on the Province Competitive Index (PCI) is employed to assess human resources within the provinces.
RESEARCH CONTRIBUTION
This study contributes to both theoretical understanding and policy implications by developing a framework and testing the hypothesis of knowledge spillover at the sector level Utilizing a Spatial Regression Model, it investigates knowledge spillovers among sectors and delves into the contextual factors influencing firms' total factor productivity (TFP) By applying a Cross-classified Model, the research examines the spillover effects of innovation activities at the sector level and human resources at the provincial level on firms’ TFP This approach is particularly valuable in the context of Vietnam, as it identifies key spillover factors that enhance sector innovation capacity, providing crucial insights for policymakers Additionally, the Cross-classified Model highlights the significance of firm characteristics and sectoral and provincial attributes in relation to firms’ TFP.
This study builds upon the frameworks of knowledge spillovers established by Cohen and Levinthal (1989) and Griliches (1992) to analyze both intra-sector and inter-sector dynamics By employing Spatial Regression Models, the research investigates the direct and indirect effects of R&D, FDI, and trade on innovation across sectors This approach provides crucial insights for identifying effective channels to boost innovation activities.
This study utilized the Cross-Classified Multilevel Model to investigate the heterogeneity in firms' Total Factor Productivity (TFP) in Vietnam, assessing the impact of firm-level, sectoral, and provincial factors It highlighted the significance of sectoral innovation spillover effects, based on Griliches' (1992) framework, as well as the influence of provincial human resources, as discussed by Moretti (2004b) The findings suggest that understanding these spillover effects can inform effective sectoral and provincial policies aimed at boosting productivity.
STRUCTURE OF THIS STUDY
This study is structured into five chapters The first chapter serves as the Introduction, while the second chapter presents a Literature Review that includes both the Theoretical Framework and Empirical Studies related to two general objectives The third chapter focuses on Methodology, detailing the Spatial Regression Model and the Cross-Classified Model.
This article includes a detailed Model Specification, Variable Measurement, and Data analysis The subsequent chapters focus on the Results and Discussion, with one chapter examining Sectoral Innovation and Spillover Effects, while the other explores the heterogeneity in Total Factor Productivity (TFP) among Vietnamese manufacturing firms The concluding chapter summarizes the findings and outlines the Policy Implications.
LITERATURE REVIEW
DEFINITION AND CONCEPTS
Knowledge spillovers refer to the benefits gained from others' investments in knowledge without incurring the full costs, stemming from the public good nature of knowledge, which is non-rival and non-excludable Arrow (1962) highlights that knowledge's non-rival characteristic means it remains available for use without depletion, while its non-excludable nature allows others to benefit from it without restriction Kaiser (1960) noted that these spillovers often arise from inadequate protection of knowledge within innovative firms, leading to what is termed ‘knowledge spillover.’ Building on this, Griliches (1992) suggested that investments in knowledge are likely to spill over for commercialization by third-party firms that do not bear the full costs of accessing and implementing these ideas.
Knowledge spillovers can occur both voluntarily and involuntarily among agents, as noted by Romer (1990) Knowledge is typically categorized into codified and tacit forms Codified knowledge, which can be partially protected through explicit means like patents, contrasts with tacit knowledge that is more challenging to safeguard due to its embodiment in the skills of employees Consequently, tacit knowledge serves as a significant source of knowledge spillover.
Firms can access external knowledge through various methods, including market transactions, formal collaborations, and informal interactions They may purchase knowledge assets directly from other organizations or engage in partnerships to enhance their knowledge base Additionally, the non-rival and non-exclusive nature of knowledge allows firms to acquire insights without formal transactions Key external sources of knowledge include customers, suppliers, and competitors, enabling firms to identify market opportunities and emerging trends by understanding customer needs and expectations, as well as leveraging the knowledge embedded in goods provided by suppliers and competitors.
Most of studies on knowledge spillovers made distinguish between
Knowledge flows can be categorized as 'horizontal' or 'vertical' spillover Horizontal spillover occurs when both the knowledge sending and receiving firms operate within the same industry, while vertical spillover takes place when these firms belong to different industries To effectively recognize, access, or adopt external knowledge, firms must possess a certain level of absorptive capacity, which refers to their ability to recognize, absorb, and utilize new external information (Cohen and Levinthal, 1989).
Knowledge spillover is crucial in both economic literature and public policy Griliches (1992) highlighted that the "New" growth economics suggests that economic growth cannot continue at a steady pace without significant externalities, spillovers, or other factors contributing to social increasing returns.
A recent study from 2018 highlights that the interplay between spillover effects and other externalities significantly influences whether an economy underinvests or overinvests in knowledge In developed nations, strong spillover effects have led to substantial subsidies for knowledge investment For developing countries, leveraging knowledge spillovers could be a key strategy to boost economic growth.
Innovation is defined by its degree of novelty, area of change, and measurement, encompassing five key cases: the introduction of new products that consumers have not previously encountered, the implementation of untested production methods, the opening of new markets previously unexplored by specific domestic industries, the acquisition of new raw materials or semi-finished products, and the establishment of new organizational structures within an industry The OECD (2005) elaborates on this by defining innovation as the implementation of a new or significantly improved product (goods or services), process, marketing method, or organizational method in business practices, workplace organization, or external relations.
Innovation is distinguished from invention by its degree of novelty, where invention is limited to being a "global first" and holds little value unless applied In contrast, innovation encompasses concepts that are "new to the firm" or "new to the market" (Onodera, 2009) This distinction leads to two main types of innovation: radical and incremental Radical innovations often stem from inventions or intellectual creations, while incremental innovations typically emerge from specific activities executed without a formal process Consequently, the terms improvement and incremental innovation can be used interchangeably, with continuous improvements representing ongoing incremental innovations.
Innovation, while inherently linked to new developments, is primarily defined through the Schumpeterian approach (1943) Martin (2016) notes that in the 1960s, 'innovation' was largely understood and measured as technology-driven advancements in manufacturing, heavily influenced by R&D and patenting due to the manufacturing sector's dominance in developed economies Early definitions focused significantly on product and process innovation (Pavitt).
The trend of economic development has led to a growing focus on organizational development and marketing, particularly in relation to innovation, as defined by the OECD (2005) Innovation is characterized by the introduction of new or significantly improved products, processes, marketing methods, or organizational practices It can take various forms, being new to the firm rather than the world, and plays a crucial role in enhancing productivity and employment.
Innovation measurement varies based on research objectives, methodologies, and interpretations of novelty Martin (2016) highlights that in the technology-driven innovation landscape, which often involves patenting, pioneers in innovation studies have developed measurement tools utilizing indicators such as R&D funding, researcher numbers, and patent counts Patent counts serve as a key metric in assessing regional innovation (Ponds et al., 2010; Capello & Lenzi, 2016; Wang et al., 2016) as well as firm-level innovation (Guan et al., 2015; Lin, 2015; Blazsek & Escribano, 2016; Li et al., 2017; Qiu et al., 2017).
Guan and Pang, 2017) and sectoral innovation (Buerger and Cantner, 2011). However, Martin (2016) argued that such indicators may be ‘missing’ much innovative activities that is incremental or is not patented.
2.1.3 Knowledge production function and the determination of innovation in this study
This study utilizes the knowledge production function (KPF) proposed by Pakes and Griliches (1984) to analyze innovation and its determinants The authors provided a simplified diagram to illustrate the KPF model, highlighting its relevance in understanding the relationship between knowledge and innovation.
Figure 2.1 The framework of knowledge production function
The diagram illustrates that K˙ is generated by a knowledge production function (KPF) that converts previous research expenditures (R) and a disturbance term (U) into inventions This disturbance term accounts for the influence of various nonformal R&D inputs and the inherent randomness involved in the invention process.
Pakes and Griliches (1984) examined the transformation function from r to k˙, proposing a Cobb-Douglas form for the KPF that incorporates firm-specific constants and a time trend Their model is expressed as k˙ it = ai + bt + ∑ 8 c r i,t–c + u i,t, where ui,t denotes an independent and identically distributed disturbance that is uncorrelated with r, reflecting the randomness in the KPF The term ai accounts for firm-specific variations in the productivity of research efforts, influenced by factors such as appropriability environments, opportunities, and managerial ability differences.
Next, the relationship between p and k˙ was presented as follows: pi,t = dt + þ kt˙
Where þ is the elasticity of patents with respect to knowledge increments, and d is a measure of the trend in factors determining the propensity to patent.
The variance in patents, denoted as v ∗, represents the portion that cannot be explained by changes in knowledge increments; thus, this variance is considered "noise" within the patent measurement.
Combining two above equations, they reached the preferred functional form in their analysis as follows: pi,t = α + yt + ∑ wc r i,t–c + ∑ 5 Φ c r i,– c
O wc = þ8c , y = þ b + d, Φ = þƒ, 5i = vi + þ ui, and ci,t = vi,t + þ ui,t
THEORETICAL FRAMEWORK
2.2.1 Developing model on Knowledge Spillovers at sector-level
Due to the limited existing empirical and theoretical research on the connections between R&D, FDI, trade activities, and innovation at the sector level, we initiated the development of a functional equation to link these variables in our data analysis.
Cohen and Levinthal (1989) constructed a model of firm’s stock knowledge as follow: zi = Mi + yi (8 ∑j*i
A firm's stock of technological and scientific knowledge is influenced by its investment in research and development (R&D), denoted as Mi The firm's ability to assimilate and utilize knowledge from the public domain, represented by yi, reflects its absorptive capacity Additionally, the degree of intra-industry spillovers (8) and the level of extra-industry knowledge (T) play crucial roles in enhancing the firm's innovation potential Furthermore, investments in R&D by other firms also contribute to this dynamic knowledge ecosystem.
Mj for j≠i also contribute to zi This model implies the intra as well as inter sectoral knowledge spillover among sectors.
We defined Zs to be the total output of knowledge in the sector s: Zs n i=
Similarly, Ms is the total input of knowledge in the sector s: Mc =∑ N Mi. (2.21)
Zc = Mc + 8 ∑ yi ( Mc − Mi) + N T ∑N
In the industry, N represents the total number of firms, while n indicates the number of firms engaged in innovation activities Additionally, m denotes the number of firms involved in research and development (R&D) activities.
Then the average knowledge stock in the industry is Z̅ = Zc
The average R&D investment in the industry is M¯ = Mc
We assume that zi = Z̅ + ði and Mi = M¯+ &i
n Zc + ∑ n ði = Mc + Mc 8 ∑ yi - 8 ∑N
Replace (2.28) into (2.31) we have: n Z c + ∑ ði = (N N
n Zc + ∑n ði = N Mc + (N − 1) Mc 8 ∑N yi + (Mc − N Mc ) +
(2.38) Besides intra-industry spillover, Griliches (1992) suggested the more complicated spillover when a whole array of industries which “borrow” different i= 1 i=
Industries acquire varying amounts of knowledge from diverse sources based on their economic and technological proximity Griliches (1992) formulated a method to quantify the total knowledge that the ith industry borrows from all accessible sources.
Where Kj measures the levels available sources of knowledge while wij is the
“weighting” function that can be interpreted bas the effective fraction of knowledge in sector j by industry i One of earlier suggestions was based on
The "vertical" borrowing approach, as outlined by Griliches (1992), utilizes input-output tables to assess the interconnectedness of industries based on their purchasing relationships This study focuses on exploring rent knowledge spillovers both within a sector and across different sectors, distinguishing these from pure knowledge spillovers by emphasizing the spillover embedded in transaction goods Consequently, employing a transaction matrix derived from the Input and Output table is deemed the most suitable method for this analysis.
Griliches (1992) quantified the total aggregate knowledge utilized by the ith industry from various sources in equation (2.39), denoted as T in equation (2.38) This T represents the combined technological and scientific knowledge derived from all other sectors (k ≠ s), leading to the formulation of n N.
The equation (2.40) serves as a foundation for analyzing the direct and indirect effects of knowledge spillovers stemming from R&D, foreign direct investment (FDI), and trade activities within our model, as detailed in the methodology chapter.
In the equation (2.41), the left-hand side express the percentage of firms in the industry having innovation activities This percentage may depend on the R&D
The relationship between industry investment and the percentage of firms engaged in R&D activities highlights the significant impact of research and development on innovation within the industry, as detailed in the methodology chapter.
2.2.2 Channels of knowledge spillovers and the research hypothesis of the first objective
The percentage of firms engaged in innovation activities within an industry is influenced by the spillover of technological and scientific knowledge from other industries This knowledge often stems from R&D investments made in those industries, which in turn affects the percentage of firms involved in R&D activities Griliches (1979) emphasized that the knowledge within any sector is not solely derived from its own R&D investments but is also shaped by insights gained from other sectors Consequently, the productivity of an industry relies on the R&D investments of various other industries This study aims to test the hypothesis that R&D in a specific sector positively impacts its innovation, while also examining the indirect influence of R&D from other sectors on that innovation.
H12: The sectoral innovation in the sector i may be positively affected by the R&D from other related sectors.
In addition to R&D investments, the technological and scientific advancements within an industry are significantly influenced by the vertical linkages formed through foreign direct investment (FDI) transactions and the embodied knowledge gained from trade activities Previous studies on technological progress highlight the importance of these factors, emphasizing that innovation is not solely reliant on research and development but also on the interconnectedness of global trade and investment.
According to Engelbrecht (1997), international trade and foreign direct investment (FDI) serve as the primary channels for transferring external knowledge and foreign technologies between countries Engaging in activities like importing, exporting, and FDI not only facilitates this transfer but also fosters communication channels that enhance cross-border learning in areas such as production methods, product design, organizational strategies, consumer preferences, and market conditions.
Knowledge spillover effects are significantly enhanced by foreign direct investment (FDI), as highlighted in studies by Jaffe (1986), Jaffe et al (1993), and Keller (2002) Geographical proximity plays a crucial role in these spillovers, suggesting that FDI may generate more externalities than international trade Hofmann and Wan (2013) identified various channels through which knowledge spillovers from FDI occur, including horizontal spillovers within the same industry, vertical downstream and upstream spillovers between different industries, and spillovers facilitated by the labor market.
Hofmann and Wan (2013) identified four channels through which foreign direct investment (FDI) creates horizontal externalities that impact domestic firms in the same industry: competition, imitation and adoption, labor turnover, and second-round effects via input suppliers Multinational enterprises (MNEs) can exert competitive pressure on local firms, potentially leading to reduced market share and prompting domestic firms to innovate Additionally, MNEs possess advantages in technology, management, research and development, and human capital, which can result in knowledge spillovers as domestic firms imitate products, production methods, and market strategies—a phenomenon known as the demonstration effect or learning by watching (Wang and Blomstrom, 1992) Labor turnover effects also facilitate knowledge transfer, as domestic firms can gain insights by hiring former MNE employees, while information exchange between employees of domestic firms and foreign affiliates occurs through social contacts Lastly, domestic firms may benefit from improved input suppliers, driven by the demands of MNEs, leading to enhanced overall efficiency and productivity.
Foreign Direct Investment (FDI) creates vertical externalities that impact domestic firms across various industries, categorized into backward and forward vertical FDI Backward vertical externalities occur when multinational enterprises (MNEs) act as suppliers to domestic customers, while forward vertical externalities arise when MNEs serve as customers to domestic suppliers MNEs can facilitate direct knowledge transfer and demand higher quality products from suppliers, often providing technological support and managerial training This increased demand for intermediate goods can lead to economies of scale for domestic suppliers, resulting in positive externalities through enhanced knowledge transfer and a broader range of inputs.
Markusen and Venables (1997) developed an analytical framework to evaluate the impact of industrial linkages, suggesting that foreign direct investment (FDI) can alter supply and demand across related industries While FDI introduces increased competition that may negatively affect local businesses, it can also provide advantages to firms in other sectors For example, the influx of FDI can result in lower prices, benefiting the overall market.
EMPIRICAL STUDIES
2.3.1 Empirical Studies on determinants of sectoral innovation
There are three main unit levels in investigating determinants of innovation.
Research at the sectoral level is less prevalent compared to national and firm levels, as noted by Buerger & Cantner (2011), Chamberlin et al (2010), Piqueres et al (2015), and Malerba et al (2013) Various theories, including R&D mainstream theory, industry lifecycles, evolution theory, and sectoral innovation systems, help synthesize the determinants of sectoral innovation Malerba et al (2013) explored national and international intersectoral R&D spillovers and their impact on innovative activities across six large, industrialized countries from 1980 onwards, building on existing literature regarding R&D spillovers.
In 2015, Piqueres et al validated the importance of various factors influencing sectoral innovation capacity, drawing on industry lifecycles, evolution theory, and sectoral innovation systems They developed a model highlighting key determinants of sector innovation capacity, which include prior innovation experience, innovation infrastructure, and the structural characteristics of the sector.
Most studies on sectoral innovation rely on firm-level data to generate sectoral variables due to the scarcity of sectoral data Typically, sectoral innovation is assessed through metrics such as patent counts (Capello and Lenzi, 2016; Piqueres et al., 2015; Malerba et al., 2013; Li et al., 2017; Lin, 2015), innovation counts (Castellacci, 2008), or the sales value of products (Hashi & Stojcic, 2013) A common approach to estimate sectoral innovation from firm-level data is to calculate the number of firms with at least one type of innovation (Aiello and Ricotta, 2015) or the proportion of firms engaged in innovative activities, as defined by the Oslo Manual (Piqueres et al., 2015).
To assess the reliability of sectoral innovation proxies, Piqueres et al (2015) utilized factor analysis to develop a sectoral innovation variable Additionally, Capello and Lenzi (2016) examined sectoral innovation by applying weights based on the significance of innovation levels.
Spillover variables, crucial for understanding sectoral innovation, are complex and encompass various types, including R&D spillover, innovation spillover, and FDI spillover A straightforward method for constructing these variables involves summing values without weighting For example, Tian (2016) assessed intra-sector FDI spillover by considering foreign presence within a sector, while external FDI spillover was analyzed based on foreign presence outside the domestic firm's sector Researchers often apply weights based on sector linkages, similarities, or geographical proximity Goya et al (2016) differentiated between intra-industry and inter-industry spillover innovation by evaluating total R&D stock within a sector and the R&D stock from other sectors, weighted by intermediate purchases, typically derived from Input and Output analysis.
Various estimation methods are employed to investigate the determinants of sectoral innovation, ranging from OLS regression to multi-level analysis Multiple regression techniques, often utilizing two or three-stage least squares (2SLS or 3SLS), are commonly used (Chen et al., 2015; Kováč & Žigić, 2016; Lee, Kim, & Lee, 2017; Luo, Guo, & Jia, 2017) Studies focusing on dependent variables like patent counts frequently adopt Tobit or negative binomial regression methods (Lọpple et al., 2016; Buerger & Cantner, 2011; Castellacci, 2008) Additionally, research that considers the geographical characteristics of sectors often employs spatial regression techniques (Autant-Bernard & Lesage).
Several studies have examined the combined impact of sectoral and regional factors on firm innovation, often utilizing multi-level regression analysis (Guan & Pang, 2017; Aiello and Ricotta, 2015).
Research on sectoral innovation highlights several key findings Some studies, such as those by Piqueres et al (2015) and Chamberlin et al (2010), indicate that certain determinants are not linked to spillover effects Conversely, research by Autant-Bernard & Lesage (2011) and Kaygalak & Reid (2016) demonstrates the positive influence of R&D on sectoral innovation, with Autant-Bernard & Lesage confirming that both private and public R&D contribute positively Additionally, Kaygalak & Reid found that innovation processes in Turkey are geographically concentrated rather than organizationally proximate Malerba et al (2013) emphasized the significant role of intersectoral R&D spillover effects on innovation activities, noting that domestic R&D exerts a stronger impact than international R&D.
Sectoral innovation research has predominantly concentrated on developed countries within the European Union (EU) Numerous studies have either examined specific EU nations, such as France (Autant-Bernard & Lesage, 2011), Germany (Bade et al., 2015; Buerger and Cantner, 2011), Spain (Piqueras et al., 2015), Turkey (Yurtseven and Tandoğan, 2012), or Canada (Chamberlin et al., 2010), or analyzed groups of countries like six developed nations (Barcenilla-Visús et al., 2014) and ten Asian countries.
Research on spillover effects, particularly in R&D and knowledge, is predominantly conducted in developed nations like the U.S and France, as noted in various studies (Blazsek & Escribano, 2016; Capello & Lenzi, 2016; Hashi & Stojcic, 2013; Lin, 2015; Autant-Bernard & Lesage, 2011) Additionally, innovation spillovers have been examined across ten Asian countries (Chen & Yang, 2011) and in Spain (Stare & Damijan, 2015) A significant portion of FDI spillover research has concentrated on China, highlighting its unique context (Li et al., 2017; Liu et al., 2010; Qiu et al., 2017; Tian, 2016; Wang & Wu, 2016; Zhang, 2017).
Research on sectoral innovation and R&D spillover in Vietnam is scarce, with most studies focusing on firm-level analysis (Jordan, 2015; Doan Thi Hong Van and Bui Le Nhat Uyen, 2017; Tran Thi Bich et al., 2017; Srholec, 2011; Nham Tuan et al., 2016; Nguyen Ngoc Anh et al., 2008) A questionnaire survey conducted in November highlights this gap in the literature.
Between 2015 and February 2016, a study by Tran Hoai Nam et al (2017) involving 500 firms across Hanoi, Danang, and Ho Chi Minh City identified six key determinants of innovation: awareness of innovation, innovation strategy and policy, organization for innovation, human resources for innovation, and absorptive capacity Additionally, research by Doan Thi Hong Van and Bui Le Nhat Uyen (2017) with a sample of 380 enterprises in various sectors, including electronics and biotechnology, indicated that innovation capacity is positively influenced by total quality management, internal human resources, absorptive capacity, government support, and collaboration networks Furthermore, Jordan (2015) highlighted that firms in the Red River Delta Region, particularly Hanoi, were more likely to achieve product innovation compared to those in other regions Nham Tuan et al (2016) found that process, marketing, and organizational innovation positively impacted the performance of supporting industries in Hanoi Lastly, Nguyen Ngoc Anh et al (2008) demonstrated that innovation, defined by new products and production processes, significantly affects exports in Small Medium Enterprises (SMEs) in Vietnam.
The enhancement of current products is a crucial factor influencing exports among Vietnamese SMEs Additionally, the research highlights the presence of endogeneity in exports, which could result in skewed estimates of innovation in earlier studies.
2.3.2 Empirical Studies on channels of knowledge spillover and applications of Spatial Regression Model
Numerous studies have explored the impact of regional knowledge spillover on innovation and growth, highlighting that advancements in communication technology have minimized the effects of spatial distance on knowledge transfer (Huggins and Thompson, 2015; Lọpple et al., 2016; Ponds et al., 2010; Shang et al., 2012) This section reviews empirical research on the channels of knowledge spillover and the applications of Spatial Regression Models A significant source of knowledge generation is research and development (R&D) activities, which have been shown to positively influence firm innovation and total factor productivity (Raymond and St-Pierre, 2010; Wieser, 2005) The focus on R&D spillover effects has gained traction since Griliches (1979) estimated the returns to R&D, leading to further studies on inter-firm spillover effects (Bloom and Reenen, 2007; Kovac and Zigic, 2016; Khazabi and Quyen, 2017; Savin and Egbetokun, 2016; Yang and Maskus, 2001) Additionally, R&D spillovers have been examined at both regional (Rodriguez-Pose and Villarreal, 2015) and national levels (Ang and Madsen, 2013; Tientao et al., 2016).
Malberba et al (2013) conducted an analysis of the impact of R&D spillovers on innovation activities across six industrialized countries, examining both national and international levels, as well as intra-sectoral and inter-sectoral dynamics Their research spans a significant timeframe, providing insights into how these spillovers influence innovation across various sectors.
1980-2000 Moralles et al (2016) found the positive R&D spillover effects among industries in Brazil.