1. Introduction
The global value chains’ (GVCs) participation index has become the most widely used indicator of the segmentation of multinational enterprises’ production processes among countries. GVCs’ activities take place in numerous locations with a variety of value-added in the countries when commodities cross the border more than once, the source of a rapid increase in interconnections among the developing countries during the last two decades. Multinational companies are the major source of the dispersion of production processes, so the share of intermediate inputs that cross the border more than once has increased in countries’ gross trade. As a result, the “trade in value-added” (TiVA) concept was introduced to designate the foreign and domestic value-added components in a country’s gross exports. The idea of participation in GVCs was explained by [
1,
2,
3] further crystallized it to define countries’/sectors’ participation in simple and complex GVCs, along with its development and name change over time. Similarly [
4,
5] provided a comprehensive analysis of GVCs’ contribution to economic development, and [
3,
6] provided renewed concepts of participation in GVCs, a summary of activities under GVCs, and its scope and geographic implication across countries.
Several measures have been used to calculate the value of the countries’/firms’ participation in GVCs. Best-known among them is that of [
1], later described by [
7], for the MRIO database. In GVCs, it has two indexes: forward participation (domestic value-added (DVA)) in foreign exports that cross borders more than once and backward participation (or foreign value-added (FVA)) in final goods that cross borders more than once. The value of the GVCs participation index goes from 0 to 1; the higher the value of the index, the more the country’s participation in GVCs, that is, the trade in intermediate products that cross the border more than once. We can say that trade that crosses borders is prevalent, and its production is more fragmented.
The selected countries’ values of the GVC participation index differ, ranging from 0.0879 to 0.8961. Malaysia, Hungary, and the Czech Republic have the highest participation in GVCs, and Morocco, Pakistan, and Egypt have the lowest. Reasons for these differences include that some of the emerging market countries (EMCs) have very good pro-foreign trade interconnections and investment policies, and some countries have weak performance in creating value chain activities. This study is an attempt to gain insights into the common determinants of bilateral participation in GVCs across countries to help minimize these differences.
A number of recent studies have emphasized the factors that determine the level of countries’/sectors’ GVC participation [
8,
9,
10,
11]. However, none of these papers has discussed the inter-connections of EMCs in terms of their participation in GVCs. The motivation for this study is the lack of research on the macroeconomic factors that can determine the overall and bilateral participations in GVCs across EMCs. The determinants of participation in GVCs and the origins of value-added in exports arise from trade and investment flows across borders. Generally, the level of economic development, the amount of skilled labor, the financial structure, tariffs and trade barriers, domestic business and labor market regulations, whether people and capital can move across borders (i.e., the level of economic freedom), rule of law, and technological advancements, among others, have the potential to influence these countries’ overall and bilateral GVC trade.
The novelty of this paper lies in its attempt to find and evaluate the macroeconomic and standard gravity model determinants of GVCs among the EMCs and its selection of the countries under analysis, which is based on the criteria of the Morgan Stanley Capital International Emerging Market Index (the MSCI Index). GVCs have sharpened the interdependence of trade in intermediate products and foreign direct investment (FDI), and the MSCI index claims that its selection of countries as emerging markets is due to the long-captured imagination of investors that continued to transform the global investment and trade landscape to embrace these countries during last two decades. Early in the 2000s, most trade in intermediate goods took place between developed countries; until 2003, 69 percent of world exports excluded the EMCs [
12]. However, by 2006, the exports of value-added trade from emerging economies realized a 25 percent average annual growth. Today, the value-added trade from emerging economies has overtaken the exports from developed countries. Another major component we consider with emerging markets’ integration into GVC participation is FDI. In recent years, emerging economies took a large share of the global investment of state-owned enterprises and multinational enterprises (MNEs) because of resource-seeking and the development of new markets for the dispersion of tasks and activities, respectively. According to UNCTAD (2016), the total value of FDI stock in emerging markets rose from
$60 billion to some
$2800 billion during the 2003–2015 period. The trends of value-added trade and FDI toward EMCs put them in the way of industrialization during the last two decades.
This study seeks the factors that determine the level of EMCs’ participation in GVCs in light of this investment and trade landscape as a fast track to industrialization. We use instrumental variables (IV) and fixed effects as econometric tools to check the robustness of our findings on 23 emerging markets’ annual data from 1995 to 2018. To the best of our knowledge, the selection of our macroeconomic variables for a GVC-augmented model while controlling for endogeneity is unique in the literature. Our model includes GDP per capita, human capital development, the real effective exchange rate, the economic freedom index, the share of high technology exports, profit tax, and research and development expenditures as explanatory variables for country-level participation in GVCs, with trade openness and FDI as our external instruments. To avoid omitted variables’ bias and endogeneity, we use trade openness and FDI as our instrument variables as they both met the conditions of exogeneity and relevance with GVCs. We know that GVCs are the function of both trade openness and FDI with other factors, as well, and the explanation of these variables is provided in
Appendix C. For bilateral participation in GVCs, we use the structure of a standard gravity model with policy-related variables for GVCs trade among EMCs. Summary statistics, the correlation structure, and descriptions and data sources for variables are given in
Appendix A,
Appendix B and
Appendix C, respectively.
This paper contributes to the literature by focusing not only on the presentation and discussion of emerging economies’ GVCs but also on the empirical analysis of the policy and non-policy macroeconomic factors that may determine EMCs’ overall and bilateral participation in GVCs. As recent empirical studies on participation in GVCs have not discussed the possibility of reverse causation, this paper contributes by using trade openness and FDI as external instruments to control for the problem of endogeneity. For the bilateral participation in GVCs, we regress the standard gravity model for GVCs trade to identify the demography and institutional variables that determine the integration of EMCs into bilateral GVCs trade. In a same way [
13] determined some crucial factors for GVC participation at the macro-level for the ECOWAS countries, and [
11,
14] did the same for EU member states using micro- and macro-level data, respectively, to identify drivers of participation in GVCs. In contrast, this paper includes the demography and national economy variables of emerging markets to determine their overall and bilateral levels of participation in GVCs. The study’s findings can help the selected countries see where they stand in value-added production activities and find better opportunities to attract or fragment MNEs’ production processes, thereby increasing their level of participation in GVCs.
The rest of this paper is structured as follows:
Section 2 provides a discussion of the literature review, while
Section 3 discusses variables description and aggregate correlational relationships.
Section 4 presents model construction and econometric methodology;
Section 5 presents results and discussion;
Section 6 concludes the study.
2. Literature Review
A plethora of studies has focused on the participation of firms, industries, and countries in GVCs. This paper focuses on trade in value-added (TiVA) that crosses borders more than once and does not consider gross trade but a contribution of the countries’ value-added to domestic and foreign exports. GVCs facilitate the dispersion of production processes across the globe, but some countries are known as GVCs’ factory hubs for their regions: China, Germany, and the United States are the factory hubs of Asia, Europe, and North America, respectively. Here China has emerged as a hub of simple GVCs, and Germany and the United States are hubs of complex GVCs (CGVCs).
Ref. [
1] suggested the method by which the participation of firms/countries in GVCs is measured. This participation refers to the sum of DVA in intermediate exports and foreign value-added (FVA) in the production of final goods that cross the border more than once. The distribution of the value created by GVCs depends on their ability to produce and supply sophisticated and technologically sophisticated products and services to the value chains across the borders. Studies that analyze the selected countries’ participation in GVCs are lacking, although the factors that contribute to the participation of developed and developing countries in GVCs have been analyzed by, among others, (namely, [
10,
15,
16,
17,
18,
19,
20]).
It has been found that economic development, infrastructure, and required skills for participation in GVCs are the preconditions for developing and emerging economies to integrate into GVCs and climb the value chain ladder. The study provided the basis for supportive public policies to increase their industrial competitiveness and industrial development to enable their participation in GVCs. The study also explores the structural change in selected countries that have emerged from participation in GVCs. Against the background of the rise of GVCs in Asia, the Ref. [
21] documented the key factors of GVCs’ development in allowing the selected economies to reap the economic benefits from their participation. Key findings of the study are that the share of the GVCs’ value-added is associated with upstream positioning in the production process and the economic complexity of the countries. Other factors, such as enhancing the quality of the infrastructure, reducing trade barriers, developing human capital, increasing research and development, and improving the quality of institutions, can also foster and expand the share of Asian economies’ DVA. Ref. [
15] found a wide range of factors that enhanced the productivity and sophistication of developing countries’ exports. Most of the countries in our selected panel are included in their sample. Their results indicate that structural factors, such as the size of a country’s market, its geography, and its level of economic development, are key factors in determining participation in GVCs. They also found that improvements in logistics and customs and reforms in trade and investment policies, infrastructure, and institutions can play a role in enhancing a country’s engagement in value chain activities.
In addition, [
16] emphasized the Philippines’ involvement in value chain activities with vertical fragmented production that is measured by its participation index. The findings of the study indicate that the growing recognition of the value of GVCs has increased the drive to develop competitiveness so the country can increase its participation in GVCs’ activities and enjoy more gains by a way of higher value-added, more employment, greater productivity, and improved spillover effects. Whether in goods or services, value chains play an integral role in overall value chains, and participation and upgrading rely on competitiveness in the value added for both goods and services in the fragmentation of the production process. Ref. [
17] found that, during the 1995–2011 period, South Korea radically internationalized its value chain activities. The study showed that the country continued widening the gap between gross exports and value-added exports with a change in employment structure while exacerbating wage inequality in domestic industries. Chung also found that the replacement of labor within the domestic manufacturing sector with skilled labor was another source of Korea’s active participation in GVCs. Ref. [
18] provided empirical evidence of some technological features of Turkey’s participation in GVCs using the World Input-Output Database (WIOD) for the 1995–2011 period. They classified the manufacturing sector with respect to technology use and found that, in the 2000s, the country’s share of the mid- and high-technology sectors of manufacturing that participated in GVCs increased more rapidly than low-technology sectors’ participation did. They concluded that Turkey’s participation in GVCs might be better if done through technology-intensive sectors when the technology used in GVCs’ activities is imported from developed countries.
Regarding the drivers of Southern African Customs Union countries’ (including South Africa) participation in GVCs, ref. [
19] summarized the general capabilities of these countries to participate in GVCs. Using a factor-content methodology, this study found that efficient logistics, proximity to markets, and strength of institutions are among the most important factors that can increase the capabilities of selected countries to participate in GVCs. However, the research also showed that each sector has unique requirements, such as fixed structural capabilities, that can limit the sectoral capabilities of a given country and that a country may be able to increase its competitiveness by reducing its policy-related gaps. Ref. [
20] stated that size, age, and foreign ownership are the most important determinants of participation in GVCs for firms in the information technology sector (IT) sector. They also demonstrated that investment in IT and computer systems promote both forward and backward participation by firms and that firms’ research and development (R&D) expenditure has no impact on their participation in GVCs.
Ref. [
21] analyzed the participation of new countries in East Europe and East Asia in GVCs as an important source of a rapid increase in the international production fragmentation process. Digital technologies can help small- and medium-sized enterprises participate in value chain activities [
22]. The arrival of new technologies played an important role in the dispersion of production by reducing the environmental cost and its hazardous impact on sustainable value chain activities [
23].
These studies make clear that the determinants of participation in GVCs by developed and developing countries differ. Some authors have argued that the quality of the infrastructure or some of its components are less important than other factors because of persistent heterogeneity in countries’ infrastructure as it relates to connectivity with international markets. Other authors have pointed to the role of technology use, the level of development, R&D expenditure, and the amount of skilled and educated workforce as the determinants of participation. The present paper seeks to include the indicators that reflect the above-mentioned objectives for EMCs in an isolated way.
5. Results and Discussion
The basic objective of this study is to analyze the impact of environmental patents and other macroeconomic variables on participation in global value chains in emerging market countries. The important thing from
Table 1 is to note that with other determinants, the environmental patents have the highest positive correlation with GVCs participation. Further, a negative correlation between the real effective exchange rate and tax on profit is also notable for further analysis.
Empirical findings from the regression of our baseline specification model are presented in
Table 2. The panel of 23 EMCs is unbalanced because of the missing values of some variables in the period from 1995 to 2018. Our regression process includes the lag value of the dependent or predetermined variable and two lags for the endogenous variables (internal instruments and external instruments) to deal with the problem of endogeneity.
Using various estimation methods, the model regression was augmented with possible determinants of participation in GVCs, as reported in
Table 2.
Table 2 shows the country effects and common-time effects. As for the standard determinants of participation in GVCs, the signs are according to expectations, with the tax on profit indicator, which carries a negative significant coefficient, the sole exception. On the other hand, the coefficients obtained through the use of time-effects are high, although their magnitude is close to within-group estimators. In both these estimators, the regressors’ potential endogeneity is ignored, so the potential endogeneity of the determinants is not corrected by these estimators. The GMM difference estimator has been employed in our model to address the endogeneity problem, as shown in
Table 2. The table shows that most of the variables have the same result and significant and positive coefficients. In addition, the null hypothesis of over-identification restrictions, the Sargan test for Difference GMM, shows the null hypothesis can be rejected, and the second-order correlation (AR-2) does not show any indications of misspecifications.
As shown through our preferred estimate (SYS-GMM, along with external instruments used as robustness checks) in
Table 2, the strongest and most significant influences on participation in GVCs are the lag of the dependent variable and the environmental patents. The coefficients for the other determinants are also according to our expectations. The question that arises concerns the contribution of these variables. For a positive and significant impact of GDP per capita on GVCs, there is evidence in the literature that economic growth (GDP per capita in our case) in emerging countries has led to shifting end markets in GVCs as more trade in value-added terms has occurred in these countries. After the 2000s, China, South Korea, India, and Mexico became major exporters of final and intermediate goods, and Brazil, Russia, and South Africa became major exporters of primary products. Human capital development has a weaker but still significant and positive impact, while the real effective exchange rate, financial development, economic freedom, and environmental patents have positive and significant impacts. The results for the real effective exchange are in accordance with our expectation that lower elasticity of REER would increase the manufacturing exports, which plays an important role in determining the countries’ competitiveness in the international trading system. The positive and significant coefficient of financial development is also in line with the argument that a sound financial structure helps to ensure that a country is compatible with the international business environment and increases its participation in value chain activities. The positive and significant coefficient of environmental patents is also in line with the argument that environment-friendly innovations are considered and regarded as a solution for sustainable value chain activities. The positive significance of the coefficient of economic freedom has some important implications for the analysis, as the impediments to a business-friendly environment and market openness affect a country’s ability to participate in GVCs. The trend of economic freedom in emerging economies is increasing over time, so the higher the value of the economic freedom index, the more integrity there is with which domestic legal and economic institutions can facilitate participation in GVCs. Rates for tax on profit have a negative and significant impact on participation in GVCs, so most EMCs are extending their tax networks and introducing tax barriers to cross-border trade and foreign investment to increase the competitiveness of their domestic firms and the development of GVCs. R&D expenditure has a positive significant impact, as it is the source of technological advancement and more sophisticated production to increase value-adding activities across the borders. As discussed in previous studies on participation in GVCs, these findings are in line with the economic theory.
These findings refer to national economies’ participation in GVCs, not to specific sectors of the countries under analysis, and are found to be robust in various types of regressions. The signs and significance of the coefficients of our selected variables are in keeping with economic theory. Environmental patents, the level of economic development, and the lag of GVC participation are the variables that determine participation in GVCs across countries. EMCs are known as the homes of MNEs, and the influence of most of the variables is robustly significant, so these findings are in line with the argument that MNEs’ activities in these countries have already taken place and that their businesses involve the purchase and production of intermediate products that cross the border more than once at a large scale. Therefore, MNEs increase the participation of most of our 23 countries in GVCs. The results also confirm that sustainable innovations, i.e., total patents registered for technology and innovation related to the environment, are an important driver of the comparative advantage of the location of production, which is in accordance with production theory. The sign of the coefficient of the real effective exchange rate is also positive, so an increase (or stability) in the real effective exchange rate increases GVC participation significantly, perhaps because MNEs prefer to invest in countries with stable exchange rates (which is in accordance with FDI theory). The positive impact of economic freedom can be explained by the increasing importance of the domestic institutional arrangements for value chain activities; the economic freedom index is increasing over time, or these countries have already built strong institutional systems to facilitate businesses that are involved in value chain activities. The negative coefficient of tax on profit indicates that countries with a high tax on profit have relatively less participation in GVCs, as a high rate of taxes on MNEs’ profits discourages the businesses and FDI.
Determinants of Bilateral Participation in GVCs
In the previous analysis, we used an aggregated “world level” measure of participation in GVCs and its determinants for EMCs. However, that analysis does not address how these countries integrate into bilateral sector/industry-level participation in GVCs with each other. For this purpose, we decompose our country-level GVCs participation index further by sources and destinations to reveal bilateral GVCs relationships for industries with major production-sharing partners across EMCs. Following ref. [
29], we use FVA exports, also known as backward linkages, to capture the demand side of the intermediate inputs among EMCs for bilateral GVC integration. The best method to analyze bilateral trade is a structural gravity equation. Ref. [
30] argues that the standard structural gravity equation does not explain bilateral trade flow across countries as it is unable to consider multilateral resistance terms. Another argument given by [
31] against the use of the traditional gravity model in value-added trade is that purchases are now driven by both consumer demand, where income is the determinant of demand elasticity, and intermediate demand, where total production is the determinant of demand elasticity. However, the solution to these problems is to augment the standard gravity equation with importer and exporter fixed effects. Following the above arguments, our model for bilateral GVCs relationship is as follows:
where
,
and
are the parameters to be estimated,
is the income of the country i and country j,
is a preferential trade agreement between respective countries,
is the total number of environmental patents related to value-added trade,
is the ease of doing business in countries involved in bilateral GVCs trade, and
and
are the dummy characteristics of importer and exporter countries. It is also predicted that
, which is the unit elasticity of income/GDP.
The stochastic version of our gravity equation is:
Here it is natural for the reader to raise a question about how this model deals with zero-value observations as we have countries with zero bilateral GVCs participation. The answer to this question is the presence of individual effects that we included in Equation (3), and we argue here that this may reduce the problem of zero-value observations of countries’ GVCs participation and selection bias, as well. However, whether that happens or not is an empirical issue, and we deal with it in our previous section when we calculated an individual country’s value-added trade using “backward linkages” and “forward linkages”. The analysis of this part also covers 23 countries over the 1995–2018 period, and the dataset consists of 12,144 observations of bilateral GVCs trade flows (23 × 22 × 24 country pairs with the 24-year time span, respectively). The list of variables and countries is reported in
Appendix B and
Appendix C.
We used a theory-grounded gravity model to evaluate the bilateral participation in GVCs among EMCs. The gravity model has been used extensively in the trade literature because of its empirical and theoretical usefulness, and we augmented it in accordance with valid arguments from different studies.
Table 3 presents the results for the augmented gravity determinants and policy and the environment-related variables for EMCs’ insertion into GVCs, measured by backward integration. The Poisson Pseudo Maximum Likelihood (PPML) regression of Equation (5) reveals that for the positive GVCs sample, physical proximity (size of GDP) with other country-pair characteristics, such as distance and a common border and colonial history, are important determinants of bilateral participation in GVCs, measured by FVA among EMCs. These results suggest that a positive change in the intra-community GVCs trade between industrial sectors has a positive impact on EMCs’ participation in GVCs. The results of including policy-related variables in our standard gravity specification show that both preferential trade agreements and the ease of doing business increase countries’ bilateral participation in GVCs. The coefficient of the environmental technology in the partner country is also positively significant for bilateral participation in GVCs. GVC theory suggests that the industrial sector is positively related to FVA/backward linkages, so by using the backward linkages of EMCs’ industry-level participation in GVCs, we add to the theory that geographical proximity is relatively more important for the manufacturing industry than policy and environmental variables.
However, backward linkages capture the demand side of the value chains, so these results suggest for EMCs that growing industrialization increases demand for intermediate inputs for exports. As we argued earlier, the global investment and trade landscape has been transformed toward EMCs, so participation in bilateral GVCs is now more accessible to these countries from the demographic, policy, and institutional perspectives. On the basis of these findings, we can say that investment and foreign trade boost various manufacturing sectors, increase competitiveness, and improve the business environment for the expansion of value chain activities through backward linkages (demand side of the intermediate input) among these countries.
6. Conclusions and Policy Implications
Foreign direct investment and international trade remain a topic of interest because they constitute a strong weapon for countries’ economic development. Although trade using GVCs has rapidly developed since the 2000s, the empirical work on its links, determinants, and implications is not well perceived. This paper identifies the key determinants that affect the participation in GVCs of emerging market countries. The study also examines the bilateral participation of EMCs in GVCs by discussing the role of standard gravity determinants, policy, and environmental variables, using data from 23 countries from 1995 to 2018. To prevent potential endogeneity, we employed numerous instrumental variables (IV) techniques for the determinants of participation in GVCs. Although consensus has already been reached regarding the determinants of participation in GVCs across countries, the literature is divided between the choice of countries and empirical strategies.
Participation in GVCs of the emerging market countries is a fast track to industrialization. The findings of this study broaden the discussion of GVCs and their determinants. As FDI inflows and the share of foreign trade in EMCs have increased rapidly over the last two decades, we estimated the impact of various macroeconomic variables that affect multinational enterprises and other foreign investors’ decisions about the location/segmentation of their production processes and investments that determine value chain activities. The study’s findings led to useful insights into the extent and potential of EMCs’ policies, use of sustainable and environmentally friendly technologies, and geographical proximities of bilateral integration and participation in GVCs. We used an instrumental variables approach to avoid reverse causation, and the common findings are in accordance with our expectations, as the lags of GVCs participation, environmental patents, and level of development (GDP per capita) have the highest positive and significant impact on participation in GVCs; except for tax on profit, all variables have a positive influence on EMCs’ participation in GVCs. The findings are in line with economic theory and studies that have dealt with value-added trade across countries. Further, our findings show that geographical proximity and environmental patents have a positive and significant impact on bilateral participation in GVCs through backward linkages. These findings suggest that intra-regional and intra-community GVCs trade, backed by strong policy and institutional measures, is positively related to growing backward participation and that this backward integration can be a powerful instrument with which emerging market countries can insert themselves into the global trading system.
The results of the present research complement previous findings on the determinants of participation in GVCs at the aggregate/national level and provide insights and lessons for individual economies in creating or joining GVCs. In sum, emerging market countries must create an enabling environment for foreign investors and multinational enterprises to insert themselves into the GVCs, especially in the production of more sophisticated and environmentally friendly export bundles and in increasing the diversification of value-added exports. Guiding principles on foreign trade and investment policies that are conducive to bilateral and country-level value chain activities could also be developed and adopted in EMCs.