Global Value Chains, National Innovation Systems and Economic Development

  • Jan Fagerberg
  • Bengt-Åke Lundvall
  • Martin Srholec
Special Issue Article


The purpose of this paper is to examine the hypothesis that increased participation in global value chains (GVCs), such as assembly of imported parts for exports, leads to higher economic growth. The focus is particularly on the extent to which this holds for low-income countries, and the role that capability-building, i.e. development of the national innovation system, plays in the possibility of benefitting from GVCs. The analysis is based on evidence from 125 countries over the period 1997–2013. To analyse the issue, a comprehensive framework that allows for inclusion of a range of relevant factors, including GVC participation (measured by the foreign value added embodied in a country’s exports), is applied. The results suggest that countries that increase GVC participation do not grow faster than other countries, when other relevant factors are controlled for. Small countries, and countries with low capabilities, appear to be particularly disadvantaged.


global value chains capability national innovation system economic development economic growth 

Le but de cet article est d’examiner l’hypothèse selon laquelle une participation accrue aux chaînes de valeur mondiales (CVM), telle que l’assemblage de parties importées dans le but de les exporter, conduit à une croissance économique plus élevée. L’accent est mis surtout sur la pertinence de cet hypothèse pour les pays en développement. L’analyse est fondée sur des données probantes provenant de 125 pays, dont de nombreux pays à faible revenu, sur la période 1997–2013. Pour analyser la question, un cadre global qui permet d’inclure une série de facteurs pertinents, y compris la valeur ajoutée étrangère dans les exportations d’un pays (importations CVM), est appliqué. Les résultats présentés dans l’article suggèrent que les pays qui augmentent les importations de la chaîne de valeur mondiale ne connaissent pas une croissance plus rapide que d’autres pays, lorsque d’autres facteurs sont contrôlés. Les petits pays, et les pays à faible capacité, semblent particulièrement désavantagés.

JEL Classification

F43 O10 O30 O40 O57 


What explains the extent to which countries manage to exploit the worldwide pool of technological knowledge to their advantage? This has been a hotly contested issue in economic and development research for a long time. The so-called Washington Consensus, advocated by the World Bank and other international organisations, predicted that this would be easy as long as the country shied away from tampering with markets and practised openness to trade and foreign investment. However, empirical research has found the evidence on this proposition to be rather mixed (Görg and Greenaway, 2004; Fagerberg et al, 2010; Keller, 2010).

Several contributors to the debate point out that one explanation might be that successful exploitation of foreign knowledge crucially depends on the development of national “technological capability” (Kim, 1997; Lall, 1992) or “absorptive capacity” (Cohen and Levinthal, 1990; Criscuolo and Narula, 2008) within the framework of a “national innovation system” (Lundvall, 1992; Nelson, 1993). The emergence of the innovation-system approach has, from the early 1990s onward, led to a host of new research emphasising the role of national capability-building in economic development (Fagerberg and Srholec, 2008; Lundvall et al, 2009). However, since national innovation systems are increasingly dependent on foreign sources of knowledge, it is also important to assess if, how and in what forms openness to various channels for transfer of foreign knowledge matters for economic development.

One important strand of research on economic development has focussed on how and to what extent foreign direct investment (FDI) by multinational firms contributes to economic growth in countries at different levels of development (Narula and Dunning, 2010). An important insight from this literature is that spillovers to host-country firms require that those firms have sufficient absorptive capacity and linkages with the foreign affiliates (Castellani and Zanfei, 2006; Narula, 2014; Rojec and Knell, 2017). Econometric studies at country level also indicate that a positive impact of FDI is conditional on national capabilities and human resources (Borensztein et al, 1998; Xu, 2000; Filippetti et al, 2016).

A new stream of literature that emerged during the 1990s pointed out that the combination of the information and communications technology (ICT) revolution and innovations in transport technology had led to the development of new ways to produce and distribute goods and services globally (Sturgeon, 2002) in the form of global value chains (GVCs) coordinated and led by multinational companies, so-called lead firms (Gereffi and Korzeniewicz, 1994; Gereffi et al, 2005). It was argued that this might provide enterprises in developing countries with opportunities to upgrade technologically and in terms of functions through participating in such networks (Ernst and Kim, 2002; Gereffi and Fernandez-Stark, 2011; Gereffi, 2014). Following this line of argument, international organisations, such as the World Bank, have recommended developing countries to increase their participation in GVCs in order to spur economic growth.1

Much of the empirical research on GVCs has taken the form of case studies at the level of enterprises, geographical clusters or specific segments of vertically organised business activities.2 These studies have brought to light many examples of local firms in countries at different levels of development that have been able to upgrade products and processes in an interaction with lead firms in high-income countries. However, moving from case studies to analyses of entire countries or the global economy as a whole is a challenging step that was for a long time hampered by lack of data on participation in GVCs at national and global level. More recently, international agencies such as the Organisation for Economic Co-operation and Development (OECD) and United Nations Conference on Trade and Development (UNCTAD), as well as networks of researchers, have created data sets that in a better way than before account for trade in intermediate products (Timmer et al, 2015; Eora, 2016). These data may be used to illustrate the proliferation of GVCs (Koopman et al, 2010, 2014; Timmer et al, 2014 and Foster‐McGregor et al, 2015).

The purpose of this paper is to examine the hypothesis that increased participation in GVCs, such as assembly of imported parts for exports, leads to higher economic growth. The focus is particularly on the extent to which this holds for low-income countries, an issue that has received relatively little attention so far, often due to data limitations (Kummritz and Quast, 2016). Therefore, the present paper places emphasis on including as many low-income countries as possible, to some extent at the expense of richer data and availability of time series. To explore this issue, we apply a framework that allows for inclusion of a range of relevant factors, including not only GVC participation,3 but also other factors that may influence transfer and exploitation of knowledge and development of the national innovation system.

In “The Roles of Knowledge, Capabilities and GVCs in Economic Development” section, we discuss how different forms of capability-building and foreign sources of knowledge, including participation in GVCs, interact with the process of economic development. Based on the conclusions reached there, the subsequent “A Preview of the Data” section delves more deeply into the measurement of the various factors, including capabilities, participation in GVCs and other channels for foreign knowledge transfer, and explores the relationship with economic development. The “Does It Matter?” section of the paper considers, using regression analysis, the extent to which increased participation in GVCs, measured by foreign value added in a country’s exports, is associated with higher economic growth. The final section sums up the lessons from the study.

In general, the results presented in this paper suggest that countries that increase their participation in GVCs, e.g. through assembly of imported parts for exports, do not grow faster than other countries, when other relevant factors are controlled for. Countries with low capabilities, i.e. weakly developed innovation systems, appear to be particularly disadvantaged. The same holds for small economies. However, it should be noted that, due to the nature of the available data, the results should not be interpreted as proving causal relationships.

The Roles of Knowledge, Capabilities and GVCs in Economic Development

Today it is generally acknowledged that a very important source of differences in levels of economic development concerns differences in the command of knowledge (for an overview see Fagerberg and Srholec, 2009). Moreover, it is increasingly recognised that much economically useful knowledge is difficult and costly to identify, access, acquire and exploit and that, for most if not all nations, foreign knowledge-bases are much larger than domestic ones. Hence the ability to tap into these foreign knowledge-bases becomes of utmost importance for the economic development of a nation.

Several different channels for such knowledge transfer may be identified. Much knowledge, scientific knowledge for example, is in principle free, but that does not mean that it is easy to access and exploit. Above all, it requires a high-quality national education system, and a public and private R&D system that makes it possible to link up with advanced global research networks (Wagner and Leydesdorff, 2005). Some advanced knowledge is proprietary, and enterprises and governments can obtain access by paying for it, for instance, by licensing. Still, successfully exploiting the knowledge continues to be demanding and requires domestic engineering and design capabilities to succeed. FDI is another potential channel of knowledge transfer that may generate positive spillovers to domestic firms. Several studies of such spillovers demonstrate, however, that the benefits are conditional on developing sufficient indigenous capabilities on the receiving end (Bell and Marin, 2004; Castellani and Zanfei, 2006; Criscuolo and Narula, 2008; Narula, 2014). Participation in international trade, for example importing capital goods (Gomulka, 1971), may also contribute to knowledge transfer. Finally, knowledge may be embodied in people, i.e. skilled workers and experts moving across national borders (Saxenian, 2006). Sending students to obtain training abroad may be seen as one way to strengthen the domestic knowledge-base. Common to these different channels of knowledge flows is that effective use and diffusion of the knowledge absorbed will depend upon the strength of the national innovation system, e.g. its technological infrastructure, the skills of its labour force and firm-level capabilities (Fagerberg et al, 2010).

Participation in GVCs is a particular form of openness to trade in which knowledge transfer takes place in a more organised and interactive manner than in other forms of trade under the supervision of so-called lead firms governing the activities of the chain. Studies by GVC scholars have analysed how specific major multinational firms have organised production chains and how they have influenced formally independent firms operating as their preferred suppliers (Gereffi and Fernandez-Stark, 2011). Gereffi et al (2005) proposed five different modes of governance in global value chains: (i) hierarchy, (ii) captive, (iii) relational, (iv) modular and (v) market. According to the authors, the further down one goes on this list, the less dominating the lead firm. The dominance of the lead firm may be rooted in market control for a final product—such as when Walmart procures blue jeans from formally independent suppliers in Mexico (Gereffi, 1999). Alternatively, dominance may be rooted in technological capabilities—such as when Apple procures electronic components from formally independent producers in China (Linden et al, 2009).

Often, case studies of participation in GVCs have revealed long-term relationships and illustrated that the dominant firm under certain circumstances and to a certain degree will contribute to upgrading the supplier firms (Gereffi, 1999).4 Walmart needs good-quality products adapted to market needs, and Apple needs high-quality components that are designed so that they fit into final products, including new product generations. However, the literature has also demonstrated that there are limits to the willingness of dominant firms to share knowledge and build capabilities among suppliers. A crucial issue is branding and market access. Walmart does not want Mexican suppliers to become independent producers of a competing brand, and Apple will only share technological knowledge that is not at the core of the business. Actually, we would expect the dominant firm to take all kinds of precautions to avoid the supplier becoming a competitor (Humphrey and Schmitz, 2000).

Moreover, it is important to take into account that not all transactions in organised markets take place in GVCs dominated by multinationals and distributed worldwide. In fact, much of the trade in intermediate goods takes place between enterprises located in high-income countries and within supranational regions (Europe, Asia and Africa) rather than between continents, and sometimes regional trade agreements explain this kind of trade (Sturgeon, 2001); For example, the process of European integration was accompanied by a dramatic increase in this kind of trade. Such trade may of course involve long-term relationships between unequal partners, but it may also involve interaction between equal partners, and with suppliers in a quite strong position. Thus, the impact of GVC participation on the economy may differ a lot across different contexts.

While firms in high- and middle-income countries with a strong industrial base and knowledge infrastructure may be in a position to benefit from participation in GVCs,5 it is not obvious that this holds to the same extent for firms located in low-income countries with a weak national innovation system. According to literature on GVCs, the potential for upgrading will differ depending on the governance mode in the value chain, e.g. how dominating the lead firm is. Arguably, enterprises from countries with weak innovation systems may be expected to be predominantly operating in modes dominated by foreign multinational oligopolies (Gereffi and Fernandez-Stark, 2011; Gereffi and Lee, 2012). Therefore, it cannot be excluded that a major part of the economic value created goes to other parts of the value chain with more leverage (Linden et al, 2009; Ali-Yrkkö et al, 2011). Nor is it obvious that the local economy in which the enterprise is located benefits.

For example, a potential downside for the national economy might be that an enterprise joining a GVC, although advanced by local standards, decouples from interacting with domestic firms and thus undermines the potential for building dynamic national or regional clusters (Schmitz, 1995, 1999; Ponte and Ewert, 2009). Activities that become offshored to developing countries tend to be low value added and thus low commitment, which means that they are footloose and not embedded in the local economy (Castellani and Zanfei, 2006; Narula and Dunning, 2010; Giroud and Mirza, 2015). Furthermore, if the enterprise remains locked into narrow functions, the implications for the national economy may not be as favourable as policy-makers would have wished, at least not in the longer run. Development effects ultimately depend on the value-added intensity of activities undertaken locally. Several studies indicate that strong local capabilities are required for deriving substantial benefits from joining a GVC (Giuliani et al, 2005; Fu et al, 2011).

It is clear from this discussion that knowledge, including access to foreign sources of knowledge, is essential for economic development. However, it is also evident that there are several different channels for acquiring knowledge, that countries exploit these to a different extent and that the ability to do so depends on domestic capability-building. Therefore, to get a better grasp on the role of GVC participation in economic development, a broad framework including not only GVCs and other channels for knowledge transfer, but also domestic capability-building and other relevant factors, will be required.

A Preview of the Data

This section is concerned with the empirical operationalisation of the factors discussed above, as well as their relationships with economic development given by gross domestic product (GDP) per capita. All of the variables are measured in two points in time: initial and final periods, which refer to data from the nearest available year to 1997 and 2013, and whenever appropriate used in logs to limit the influence of outliers.6 Although the selected indicators have broad coverage, in some cases there were missing values that had to be dealt with (further details on definitions and sources can be found in Appendix A1).7

As concerns capabilities, we take into account nine different indicators that together give a broad view on where a country stands with respect to the development of its national innovation system.8 The first four indicators reflect what Kim (1997) called “innovation capabilities”, i.e. the quality of a country’s science base (as measured by publications), R&D investments, patents and trademarks. The two next indicators on the list, namely ISO certification and internet users, are broader in character and may be seen as examples of what Kim (1997) labelled “production capabilities”. Finally, the set of indicators contains two measures referring to the educational level of the labour force and an index reflecting the quality of a country’s bureaucracy, both of which may be regarded as examples of what Abramovitz (1986) called “social capabilities”.

For the purpose of the analysis, the nine selected capability indicators are weighted together into a composite measure using factor analysis (for detailed results see Appendix A2). As shown by the factor loadings, the various capability indicators are closely correlated, giving strong empirical support to the use of a composite measure. Figure 1 plots the resulting capability measure against GDP per capita. The regression line between the two variables is also reported. As might be expected, GDP per capita is an increasing function of a country’s capability level. Poor countries generally have low capabilities. Furthermore, resource-rich countries tend in some cases to have far higher GDP per capita than their capability levels would indicate.
Figure 1

GDP per capita and capabilities, average 1997 and 2013.

At the centre of our interest is to explore how participation in GVCs relates to economic development. Following Koopman et al (2010, 2014), a country’s gross exports can be split up into a part capturing domestically produced value added and a part capturing imported value added that is incorporated into the country’s exports. The latter has become commonly used as an indicator for the extent of downstream GVC participation, such as assembly of foreign-produced parts for exports. In this study, this indicator is called GVC imports. The same indicator has also been dubbed “backward-linkage indicator” (Kummritz and Quast, 2016) or simply “foreign value added in exports” (Koopman et al, 2014; Timmer et al, 2014; Foster‐McGregor et al, 2015).9

Until recently, trade statistics were only available in terms of gross exports and imports, hence reflecting sales, not value added, and thus becoming increasingly biased due to the spread of GVCs. Nevertheless, international organisations, including UNCTAD and OECD, have put great effort into tracing how intermediate products move between countries using detailed data on international transactions recorded in input–output tables, resulting in the Eora multi-region input–output table (MRIO) database (Eora, 2016). More specifically, the GVC imports indicator used in this study is derived from the Trade In Value Added (TiVA) database (UNCTAD/Eora, 2016), which is based on the MRIO data set and provides evidence from 189 countries, including many developing nations.10 For more detailed explanation of the database and calculation of the GVC imports indicator see UNCTAD (2013, pp. 26–29).

A country’s exports does not consist of final goods only, but also of inputs for further processing and exports by other countries. Koopman et al (2010, 2014) refer to this indicator of upstream GVC participation as “indirect value-added exports”. It has been also interpreted as “forward-linkage indicator” (Kummritz and Quast, 2016) or “the extent of GVC participation for relatively upstream sectors” (Foster‐McGregor et al, 2015). However, this indicator is not included in the TiVA database (UNCTAD/Eora, 2016) that we are using. Moreover, for developing countries, this indicator primarily reflects their traditional roles as exporters of commodities, which has been extensively analysed elsewhere (see e.g. Morris et al, 2012; Fitter and Kaplinsky, 2001; Narula, 2018), and which, although interesting, is not the central focus of this paper.

Figure 2 plots the extent of GVC imports against GDP per capita. The figure reveals that this form of participation in GVCs is not as closely correlated with economic development as capabilities are (Figure 1). In fact, there is a lot of variation across countries at similar levels of development, and the degree of variation appears to increase as countries get richer.
Figure 2

GDP per capita and GVC imports, average 1997 and 2013.

Moreover, as pointed out above, there are other channels for knowledge transfer that countries may exploit and that need to be taken into account. The variables for which reliable data are available and hence that could be taken into account in the study are (i) capital goods imports, (ii) FDI (inward) and (iii) outbound mobility of tertiary students (to North America and Western Europe). We use imports of capital goods—rather than total imports—because the use of foreign capital goods is often cited as an important channel for knowledge transfer (see e.g. Gomulka, 1971). Another reason is that using total imports would lead to double-counting since GVC imports is a sizeable part of total imports. Unfortunately, there were no data for migration of highly skilled personnel that could be exploited in the analysis. Another potentially relevant indicator used in other studies is payments for import of proprietary knowledge (e.g. licences etc.), but this could not be taken into account separately in the present case since it is also already included in GVC imports.

Does It Matter?

Arguably, the level of economic development may be seen as the result of a process in which not just one but several channels for knowledge transfer interact with other national and international factors. Moreover, knowledge-based growth is not only about exploiting foreign knowledge, because domestic knowledge creation and the national innovation system matter too. Finally, economic development may also be influenced by factors that have little to do with knowledge, such as abundance of natural resources. To take all these factors into account, this section turns to multivariate regression analysis.

The purpose of the analysis that follows is to test whether increased GVC participation measured by the indicator of GVC imports is associated with higher economic growth, when the possible influence of other relevant factors is accounted for (so-called conditioning factors). However, it should be noted that the cross-sectional nature of the data does not allow for testing of causality or the impact of possible country-specific factors, and that the results should therefore be interpreted with caution.

To analyse this issue, we employ a so-called conditional growth regression (Cornwall, 1976; Barro, 1991):
$$y = a_{0} + a_{1}Y + a_{2} O + a_{2} o + a_{3} C + a_{4} c + a_{5} F,$$
where the dependent variable is growth of GDP per capita. Y refers to the initial level of GDP per capita. O/o represent the initial level/growth of various channels for transfer of foreign knowledge. As mentioned above, in addition to GVC imports, we also include capital goods imports, inward FDI and tertiary students abroad. C/c is the initial level/growth of relevant capabilities (as described in the previous section), and F represents other exogenous factors controlled for to reduce the possible omitted variable bias. The control variables taken into account here reflect differences in country size,11 industrial structure, disease ecology and nature. All variables are in logs, as already noted above, thus growth refers to log difference (a log approximation of the growth rate). The sample includes 125 countries between 1997 and 2013 (descriptive statistics on the variables that enter the regression analysis are provided in Appendix A3 and the list of countries is in Appendix A4).

The inclusion of the initial level of GDP per capita among the explanatory factors reflects the classical “catch-up” or “latecomer” hypothesis advanced by economic historians such as Gerschenkron (1962) and Abramovitz (1986), i.e. that low-income countries far from the technology frontier have greater scope to benefit from international knowledge spillovers than countries close to the frontier. Thus, the estimated impact of this variable should be expected to be negative, indicating slower growth for countries close to the frontier.

The results are reported in Table 1. Ordinary least squares (OLS) robust to outliers is used in the estimates based on the procedure suggested by Li (1985). Beta coefficients are reported, i.e. the variables enter the analysis standardised with mean of zero and standard deviation of one, thus the estimated coefficients refer to the impact of change by one standard deviation. The first column in Table 1 reports estimates of the model without controls, while in the second column the control variables are added. However, since the estimates for some of the variables were not statistically significant, a backward search for the best model was conducted, using a 20 per cent significance level as criterion for exclusion/re-inclusion in the model, the results of which are reported in the third column.
Table 1

Explaining growth of GDP per capita: regression results, iteratively re-weighted least squares, 1997–2013





GDP per capita














Δ capabilities







GVC imports







Δ GVC imports







Capital goods imports







Δ capital goods imports






FDI inward






Δ FDI inward






Outbound mobility of tertiary students






Δ outbound mobility of tertiary students







Control variables

 Size (population)











 Natural resources rents




















R 2




Number of observations




The dependent variable is log difference of GDP per capita [purchasing power parity (PPP), constant 2011 international USD] divided by the number of years (a log approximation of the annual growth rate). Absolute value of robust t statistics in parentheses. *, **, ***denote significance at the 10, 5 and 1 per cent levels. Beta coefficients reported.

The results suggest that capabilities, whether measured by the initial level or subsequent growth, have a strong, positive relationship with growth of GDP per capita, and the estimated relation is quite robust with respect to changes in specification. However, only two of the eight variables for channels of knowledge transfer included in the test can be shown to be positively correlated with economic growth, namely capital goods imports (initial level) and sending tertiary students abroad (growth). The estimates indicate that countries that increase GVC imports tend to grow more slowly than other countries, when a number of other relevant factors are controlled for. Nevertheless, as already noted, these results do not provide information on the direction of causality.

It is possible, however, that this estimate, which is for all countries in the sample, masks quite different relationships for subgroups of countries with common characteristics, such as development level and size. The relatively small sample does not allow for extensive testing of this possibility. Nevertheless, to throw some light on this issue, we report in Table 2 a test for the parameter stability across groups of countries by adding to the model dummy variables for membership in various groups and allowing for their interactions with the increase of GVC imports. All other variables remain the same (third column of Table 1). The dimensions taken into account are income level (as defined by the World Bank), development level [as defined by the International Monetary Fund (IMF)], initial capability level (as derived from the factor analysis), geography (continents) and country size (population).12 The base category is low-income countries in the World Bank classification, developing countries in the IMF case, low-capability countries according to the factor score, African countries in the version with continents and small countries when it comes to size. Along all five dimensions, there are indications of parameter variability, although often not statistically significant at conventional levels. The strongest support for parameter variation is for large and medium-sized economies, countries that are classified as advanced by the IMF and countries with high capabilities. For these country groups, the estimated relationship between increased GVC imports and economic growth (i.e. the sum of the estimated coefficient for the base category and the interaction dummy) is close to zero.
Table 2

Testing for differences in the impact of increased GVC imports across country groups







Income level (World Bank)

Development level (IMF)

Capabilities level (factor analysis)

Geography (continents)

Size (population)

Δ GVC imports











Δ GVC imports × medium income




Δ GVC imports × high income




Δ GVC imports × transition





Δ GVC imports × advanced





Δ GVC imports × medium capabilities





Δ GVC imports × high capabilities





Δ GVC imports × Asia and Oceania





Δ GVC imports × America





Δ GVC imports × Europe





Δ GVC imports × medium size




Δ GVC imports × large size




The dependent variable is log difference of GDP per capita (PPP, constant 2011 international USD) divided by the number of years (a log approximation of the annual growth rate). All other variables remain the same as in the third column of Table 1, i.e. in the preferred model, except that the group dummies are added to the regression. Absolute value of robust t statistics in parentheses. *, **, ***denote significance at the 10, 5 and 1 per cent levels. Beta coefficient of Δ GVC imports reported.

The interpretation of these findings is plain. For advanced countries with well-developed capabilities, it does not matter much whether they participate in GVCs a little more or a little less. Hence, it appears that, when they participate in GVCs, they tend to have sufficient leverage to get their fair share of the economic benefits. This does not hold for countries without such capabilities, however, and not for very small countries either. A possible explanation, then, consistent with literature on GVCs, would be that firms in these countries have little leverage when it comes to decisions on how to share the economic benefits from the value added created in the GVCs.

Thus, the results suggest that countries with less well-developed capabilities actually lose from taking a more active part in GVCs, but that this does not hold for countries higher up on the capability ladder. An interesting question, then, is what the level of capabilities at which GVC participation starts to make more sense economically is.13 Figure 3 provides a take on this issue. The figure shows (from left to right) how the estimate of an interaction term between the increase of GVC imports and a country dummy (when countries are ranked by the initial capability level) changes as the number of countries covered by the dummy becomes gradually smaller (i.e. increasingly limited to high-capability countries); For example, 75 on the horizontal axis indicates that a dummy for the top 75 countries in terms of capabilities is used in the estimate, etc. The results show that the sample divides in two parts at a medium level of capabilities. For the more capable countries, starting from a level close to the poorest member countries of the European Union (Bulgaria and Romania, for example) or the more advanced countries in Latin America (such as Argentina, Chile or Uruguay), the relationship between increased GVC imports and economic growth (the sum of the baseline and the interaction term) edges up slightly above zero. However, for less capable countries, the relationship remains clearly negative.
Figure 3

Impact of increased GVC imports at different capability levels—an explorative analysis.

Note: The baseline is the estimated beta coefficient of Δ GVC imports. Interaction is the estimated coefficient of an interaction term between Δ GVC imports and a dummy variable for countries with capabilities exceeding a certain threshold level. Total is the sum of the baseline coefficient and the interaction term. All other variables remain the same as in the third column of Table 1. Smoothed values (the lines) are derived from kernel-weighted local polynomial smoothing (using Epanechnikov kernel function).

As pointed out above, cross-sectional data are not well suited for analyses of the direction of causality. For example, while—as pointed out earlier—the usual assumption is that participation in GVCs affects economic growth, an effect in the opposite direction may not be ruled out. While interesting, this is not an issue that can be explored with the available data. Such problems may also be relevant for some of the other indicators. We tested the robustness of the results with regards to removing the growth of capital goods imports and outbound mobility of tertiary students from the preferred model (Table 1, third column), however, the main conclusions, including with respect to parameter stability, did not change. 14 Furthermore, multi-collinearity may be a problem in small, cross-sectional data sets. A particular concern might be the correlation between GDP per capita and capabilities (see Figure 1 and Appendix A5 for details). Nevertheless, both variables come out with statistically highly significant coefficients, and results of the variance inflation factors (VIFs) test confirm that there is not a serious collinearity problem.15

Concluding Remarks

In the 1980s, international organisations such as the IMF and World Bank forged the so-called Washington Consensus which emphasised openness to trade and FDI and a hands-off approach with respect to markets as essential ingredients for development. The consensus soon started to crack, however, as research indicated that the empirical support for the underlying assumptions was far from robust (Rodrik, 1994; Chang, 2002; Fagerberg and Godinho, 2004).

From the 1990s onwards, a sizeable literature has emerged on the increasing role played by GVCs, coordinated by multinational companies, in the world economy, and the possibilities that participation in such chains may entail for firms in developing countries. As pointed out in the introduction to this paper, the very same international organisations that were behind the now defunct Washington Consensus now actively promote participation in such chains as a way forward for development. The question then arises of whether this is just old wine in a new bottle, or if it represents a decisive new turn in the process of global economic development with significant new opportunities for low-income countries to escape the poverty trap. This paper has attempted to throw new light on the issue, using a framework that also takes into account other factors that may be of importance for growth and development, and data for a broad sample including many low-income countries. Having extensive country coverage is essential to produce reliable evidence on the matter, but the cross-sectional nature of the available data means that the analysis presented in this paper is explorative.

It is certainly true, as shown by e.g. Kummritz and Quast (2016), that participation in GVCs has increased steadily over the last decades.16 However, as pointed out in Sect. 2, positive effects for all participating countries cannot be taken for granted, because it cannot be excluded that most of the benefits go to the multinationals that coordinate the chains, and that spillovers in the context of developing countries, being pecuniary or technological in nature, are small and possibly less than they would have been had the human and other resources been devoted to something else. The results of this paper suggest that, in general, countries gain little if at all in terms of economic growth from increasing their participation in GVCs as measured by foreign value added in exports. Moreover, the analysis suggests that countries with less well-developed capabilities—about half of the sample analysed here—that increase this form of participation in GVCs perform worse economically than other countries with similar characteristics. The same applies for very small countries. Unfortunately, there are not many other studies with which the results presented here can be compared. However, it is noteworthy that Kummritz (2015), using different methods and a smaller sample of countries, also finds that low-income countries do not benefit economically from participating in GVCs.17

Nevertheless, the results presented here confirm, in line with other research (Fagerberg and Srholec, 2008, 2017), that there is a strong link between developing technological and social capability and economic development. Furthermore, the results suggest that building the innovation system is not only important for economic development more generally, but also for the possibility of benefitting from GVC participation. So, placing emphasis on improving such factors, i.e. developing the national innovation system, appears to be a fruitful direction for policy. How to do that is a challenging issue that we cannot address in the necessary detail in the present paper. As emphasised in literature on national innovation systems, specific policy recommendations need to be based upon deep insight into the unique characteristics of the national innovation system (Lundvall et al, 2009). Moreover, the different dimensions included in the capability measure employed here might warrant separate attention, as they may be of different importance in countries at different levels of development (Cirera and Maloney, 2017).

Several issues raised in this paper merit more research. Better data and longer time series may allow for more elaborate tests, particularly with respect to causality, than those employed in this paper. The time period taken into account here is relatively short and includes a major global economic crisis. It cannot be excluded that this has influenced the results. There may also be long lags in cause–effect relationships that would be easier to detect if longer time series were available. Moreover, as also suggested in other recent research (Narula, 2018; Pietrobelli and Staritz, 2018), the questions concerning the role of—and interaction between—building the innovation system and different forms of openness in the process of economic development certainly deserve more attention. Finally, the GVC measure used here is derived from national accounting and as such does not discriminate between different types of governance of the value chains, which remains a challenge for future research (Altenburg, 2006; Kaplinsky and Morris, 2015).


  1. 1.

    See for example the following programmatic statement on the World Bank’s website: “Participation in global value chains (GVCs), the international fragmentation of production, can lead to increased job creation and economic growth. The World Bank Group is helping developing countries catch the GVCs wave and realize the benefits GVCs can deliver”. (, accessed on 11 April, 2018).

  2. 2.

    For a combination of macroeconomic and sectoral analysis in a specific region see Del Prete et al (2017).

  3. 3.

    GVC participation is measured in this paper as foreign (imported) value added embodied in a country’s exports, which is the only measure that is available for a broad sample of countries. A typical example of such GVC participation is assembly of imported parts for exports. See Sect. 3 of this paper for an extended discussion.

  4. 4.

    Humphrey and Schmitz (2002) make a distinction between four forms of industrial upgrading: new process, new product, new function and new sector. While lead firms may have an interest in stimulating the development of new processes and products among suppliers, they also might use their position in captive and relational forms of governance to block suppliers’ attempts to move into new functions (building strong internal R&D capability or establishing own brand).

  5. 5.

    For example, enterprises in countries such as Korea, Singapore and China have entered into international interactions as suppliers to multinationals in electronics and used the experience to move from being dependent suppliers to developing their own brands, ending up as important multinational enterprises (Lee, 2013).

  6. 6.

    If necessary, unity was added to avoid logs of zero. Unity was also added to variables with values very close to zero to avoid generating outliers with high negative values. The index of the quality of a country’s bureaucracy is not used in logs, as countries are ranked on a fixed five-point scale.

  7. 7.

    Missing data were imputed (in particular about 16 and 17 per cent of the sample for R&D expenditures and trademark applications, respectively) using the impute procedure in Stata 11.2 (for more information see Stata, 2005, pp. 217–221). The procedure, which is regression based, uses information from other variables in the data set to fill in missing values.

  8. 8.

    The capability indicators taken into account below strongly resemble those included in the “innovation system” measure proposed by Fagerberg and Srholec (2008) based on a similar methodology as the one applied here. However, their study also contained a rich set of indicators on governance and institutions, which factored out in separate dimensions.

  9. 9.

    For caveats regarding the use of input–output tables for measuring the participation in GVCs see Nomaler and Verspagen (2014).

  10. 10.

    Another database with multi-country input–output tables that could be used to compute GVC imports is the World Input–Output Database (WIOD) (Timmer et al, 2015), which, however, provides evidence on only 43, predominantly advanced countries.

  11. 11.

    Size, represented by population, is pertinent to control for, as firms in large countries naturally engage more with domestic customers, suppliers and investors than do firms in smaller economies.

  12. 12.

    Countries are assigned to income level groups according to their classification in 1997 by the World Bank (2015) and to development level groups following the classification used by the IMF (1997). Based on the capabilities index in the initial period, as derived from the factor analysis (Table 1), countries are assigned to three groups as follows: (i) low capabilities with the index more than one standard deviation below the mean (15 countries), (ii) medium capabilities with the index within the range of one standard deviation below and above the mean (85 countries) and (iii) high capabilities with the index more than one standard deviation above the mean (25 countries). The results are qualitatively similar if the medium-capabilities group is split into medium–low (58 counties) and medium–high (27 countries) depending on whether the index is below or above the mean, respectively. Countries are assigned to the continent where most of the population resides. Size groups are defined in terms of the initial population reported by the World Bank (2016) as follows: (i) small size with less than 5 million (40 countries), (ii) medium size with 5–20 million (45 countries) and (iii) large size with more than 20 million (40 countries).

  13. 13.

    We wish to thank one of the editors of this journal for proposing this question.

  14. 14.

    Results of these additional tests are available on request.

  15. 15.

    In the preferred specification (Table 2, third column), mean VIF = 2.91 for the whole model and maximum VIF = 7.18 for the most correlated single variable. For more details on the VIF test, see the estat vif command in Stata (2017, pp. 2276–2280).

  16. 16.

    Whether this also should be expected to hold in the future is another matter, which we cannot pursue here. See IRC Trade Task Force (2016) for an interesting take on this issue.

  17. 17.

    Tajoli and Felice (2018) present estimates of a knowledge production function with patenting as dependent variable and R&D spillovers transmitted through GVC participation among the independent variables. Patenting is well known to be a problematic indicator in a developing country context (Fagerberg et al, 2010). Nevertheless, they conclude that GVC participation plays a positive role for knowledge transmission, and that this holds also for developing countries. However, their sample includes very few developing countries, and not a single low-income country (as classified by the World Bank) or country from Africa.




Financial support from the VINNOVA Core Funding of Centers for Innovation Systems Research project 2010-01370 on “Transformation and Growth in Innovation Systems: Innovation Policy for Global Competitiveness of SMEs and R&I Milieus” and the Czech Science Foundation (GAČR) project 17-09628S on “Innovation activities in global production networks: Evidence from Czech business enterprises” is gratefully acknowledged. Earlier versions of the paper were presented at the 2016 OBEL Symposium on Innovation Systems, Globalization and Development, May 10–12, 2016, Aalborg, Denmark, the 14th Globelics International Conference, October 12–14, 2016, Bandung, Indonesia and the workshop on “Innovation Systems in the era of Global Value Chains”, April 24–25, 2017, Copenhagen, Denmark. We thank participants at these events, especially John Humphrey, Ned Lorenz and Bart Verspagen, and reviewers and editors of the journal for useful comments and suggestions. All usual caveats apply.


  1. Abramovitz, M. (1986) Catching up, forging ahead, and falling behind. Journal of Economic History 46: 386–406.Google Scholar
  2. Ali-Yrkkö, J., Rouvinen, P., Seppälä, T. and Ylä-Anttila, P. (2011) Who captures value in global supply chains? Case Nokia N95 smartphone. Journal of Industry, Competition and Trade 11(3): 263–278.CrossRefGoogle Scholar
  3. Altenburg, T. (2006) Governance patterns in value chains and their development impact. European Journal of Development Research 18(4): 498–521.CrossRefGoogle Scholar
  4. Barro, R.J. (1991) Economic growth in a cross section of countries. Quarterly Journal of Economics 106: 407–443.CrossRefGoogle Scholar
  5. Bell, M. and Marin, A. (2004) Where do FDI-related technology spillovers come from in emerging economies? An exploration in Argentina in the 1990s. European Journal of Development Research 16: 653–686.CrossRefGoogle Scholar
  6. Borensztein, E., De Gregorio, J. and Lee, J.-W. (1998) How does foreign direct investment affect economic growth? Journal of International Economics 45: 115–135.CrossRefGoogle Scholar
  7. Castellacci, F. and Natera, J.M. (2011) A new panel dataset for cross-country analyses of national systems, growth and development (CANA). Innovation and Development 1: 205–226.CrossRefGoogle Scholar
  8. Castellani, D. and Zanfei, A. (2006) Multinational Firms, Innovation and Productivity. Cheltenham: Edward Elgar.CrossRefGoogle Scholar
  9. Chang, H.-J. (2002) Kicking Away the LadderDevelopment Strategy in Historical Perspective. London: Anthem.Google Scholar
  10. Cirera, X. and Maloney, W.F. (2017) The Innovation Paradox: Developing Country Capabilities and the Unrealized Promise of Technological Catch Up. Washington, DC: The World Bank.CrossRefGoogle Scholar
  11. Cohen, W.M. and Levinthal, D.A. (1990) Absorptive Capacity: A New Perspective on Learning and Innovation. Administrative Science Quarterly 35(1): 128–152.CrossRefGoogle Scholar
  12. Cornwall, J. (1976) Diffusion, Convergence and Kaldor’s Law. Economic Journal 85: 307–314.CrossRefGoogle Scholar
  13. Criscuolo, P. and Narula, R. (2008) A novel approach to national technological accumulation and absorptive capacity: aggregating Cohen and Levinthal. European Journal of Development Research 20(1): 56–73.CrossRefGoogle Scholar
  14. Del Prete, D., Giovannetti, G. and Marvasi, E. (2017) Global value chains: New evidence for North Africa, International Economics. Scholar
  15. Eora. (2016) The Eora multi-region input-output table (MRIO) database., accessed 1 June 2017.
  16. Ernst, D. and Kim, L. (2002) Global production networks, knowledge diffusion, and local capability formation. Research Policy 31(8): 1417–1429.CrossRefGoogle Scholar
  17. Fagerberg, J. and Godinho, M.M. (2004) Innovation and Catching-up. In: Fagerberg, J., Mowery, D. and Nelson, R. (eds.) The Oxford Handbook of Innovation. Oxford: Oxford University Press, pp. 514–542.Google Scholar
  18. Fagerberg, J. and Srholec, M. (2008) National innovation systems, capabilities and economic development. Research Policy 37: 1417–1435.CrossRefGoogle Scholar
  19. Fagerberg, J. and Srholec, M. (2009) Innovation systems, technology and development: Unpacking the relationship(s). In: Lundvall, B.-A., Joseph, K.J., Chaminade, C. and Vang, J. (eds.) Handbook of Innovation Systems and Developing Countries: Building Domestic Capabilities in a Global Context. Cheltenham: Edward Elgar, pp. 83–115.Google Scholar
  20. Fagerberg, J. and Srholec, M. (2017) Capabilities, economic development, sustainability. Cambridge Journal of Economics, 41: 905–926.CrossRefGoogle Scholar
  21. Fagerberg, J., Srholec, M. and Verspagen, B. (2010) Innovation and economic development. In: Hall, B. and Rosenberg, N. (eds.) Handbook of the Economics of Innovation, Vol. II. North Holland, The Netherlands: Elsevier, pp. 833–872.Google Scholar
  22. Filippetti, A., Frenz, M. and Ietto-Gillies, G. (2016) The impact of internationalization on innovation at countries’ level: The role of absorptive capacity. Cambridge Journal of Economics 41(2): 413–439.Google Scholar
  23. Fitter, R. and Kaplinsky, R. (2001) Who gains from product rents as the coffee market becomes more differentiated? A value-chain analysis. IDS Bulletin 32(3): 69–82. Scholar
  24. Foster‐McGregor, N., Kaulich, F. and Stehrer, R. (2015) Global value chains in Africa. UNU-MERIT Working Paper Series #2015-024.Google Scholar
  25. Fu, X., Pietrobelli, C. and Soete, L. (2011) The role of foreign technology and indigenous innovation in the emerging economies: Technological change and catching-up. World Development 39: 1204–1212.CrossRefGoogle Scholar
  26. Gallup, J.L., Sachs, J.D. and Mellinger, A. (1999) Geography and economic development (CID Working Paper No. 1). Cambridge: Harvard University.Google Scholar
  27. Gereffi, G. (2014) Global value chains in a post-Washington Consensus world. Review of International Political Economy 21: 9–37.CrossRefGoogle Scholar
  28. Gereffi, G. (1999) International trade and industrial upgrading in the apparel commodity chain. Journal of International Economics 48: 37–70.CrossRefGoogle Scholar
  29. Gereffi, G. and Fernandez-Stark. K. (2011) Global value chain analysis: A primer. Center on Globalization, Governance & Competitiveness (CGGC), Duke University.Google Scholar
  30. Gereffi, G. and Korzeniewicz, M. (1994) Commodity Chains and Global Capitalism. Santa Barbara, CA: Praeger.Google Scholar
  31. Gereffi, G. and Lee, J. (2012). Why the world suddenly cares about global supply chains. Journal of Supply Chain Management 48(3): 24–32.CrossRefGoogle Scholar
  32. Gereffi, G., Humphrey, G.J. and Sturgeon, T. (2005) The governance of global value chains. Review of International Political Economy 12: 78–104.CrossRefGoogle Scholar
  33. Gerschenkron, A. (1962) Economic Backwardness in Historical Perspective. Cambridge: Belknap.Google Scholar
  34. Giroud, A. and Mirza, H. (2015) Refining FDI motivations by integrating global value chains’ considerations. Multinational Business Review 23(1): 67–76.CrossRefGoogle Scholar
  35. Giuliani, E., Pietrobelli, C. and Rabellotti, R. (2005) Upgrading in global value chains: Lessons from Latin American clusters. World Development 33(4): 549–573.CrossRefGoogle Scholar
  36. Gomulka, S. (1971) Inventive Activity, Diffusion and the Stages of Economic Growth. Aarhus: Aarhus University Press.Google Scholar
  37. Görg, H. and D. Greenaway (2004) Much ado about nothing? Do domestic firms really benefit from foreign direct investment? The World Bank Research Observer 19: 171–197.CrossRefGoogle Scholar
  38. Humphrey, J, and Schmitz. H. (2002) How does insertion in global value chains affect upgrading in industrial clusters? Regional Studies 36: 1017–1027.CrossRefGoogle Scholar
  39. Humphrey, J. and Schmitz, H. (2000) Governance and upgrading: Linking industrial cluster and global value chain research. IDS Working Paper, No. 120, Institute of Development Studies, University of Sussex, Brighton.Google Scholar
  40. IMF. (1997) World Economic Outlook, October 1997. International Monetary Fund, Statistical Appendix,, accessed 1 June 2017.
  41. IRC Trade Task Force. (2016) Understanding the weakness in global trade. What is the new normal? European Central Bank, Occasional Paper Series No 178/September 2016,, accessed 25 September 2017.
  42. ISO. (2014) ISO Survey 2014. International Organization for Standardization,, accessed 1 June 2017.
  43. Kaplinsky, R. and Morris, M. (2015) Thinning and thickening: productive sector policies in the era of global value chains. European Journal of Development Research 28(4): 625–645.CrossRefGoogle Scholar
  44. Keller, W. (2010) International trade, foreign direct investment, and technology spillovers. In: Hall, B. and Rosenberg, N. (eds.) Handbook of the Economics of Innovation, Vol. II. North Holland, The Netherlands: Elsevier, pp. 793–829Google Scholar
  45. Kim, L. (1997) Imitation to Innovation: The Dynamics of Korea’s Technological Learning. Boston: Harvard Business School Press.Google Scholar
  46. Kiszewski, A., Mellinger, A., Spielman, A. and Malaney, P. (2004) A global index of the stability of malaria transmission. American Journal of Tropical Medicine and Hygiene 70: 486–498.Google Scholar
  47. Koopman, R., Powers, W., Wang, Z. and Wei, S.-J. (2010) Give credit where credit is due: Tracing value added in global production chains. NBER Working Paper No. 16426. Cambridge, MA., Accessed 17 April 2018.
  48. Koopman, R., Wang, Z. and Wei, S.-J. (2014) Tracing value-added and double counting in gross exports. American Economic Review 104(2): 459–494.CrossRefGoogle Scholar
  49. Kummritz, V. (2015) Global value chains: Benefitting the domestic economy? CTEI Working Papers, CTEI-2014-05,, accessed 17 April 2018.
  50. Kummritz, V. and Quast, B. (2016) Global value chains in low and middle income countries. CTEI Working Papers, CTEI No 2016-10,, accessed 17 April 2018.
  51. Lall, S. (1992) Technological capabilities and industrialization. World Development 20(2): 165–186.CrossRefGoogle Scholar
  52. Lee, K. (2013) Schumpeterian Analysis of Economic Catch-up. Knowledge, Path-Creation, and the Middle-Income Trap. Cambridge: Cambridge University Press.Google Scholar
  53. Li, G. (1985) Robust regression. In: Hoaglin, D.C., Mosteller, F. and Tukey, J.W. (eds.) Exploring Data Tables, Trends, and Shapes. New York: Wiley, pp. 281–340.Google Scholar
  54. Linden, G., Kraemer, K.L. and Dedrick, J. (2009) Who captures value in a global innovation network? The case of apple’s iPod. Communications of the ACM 52(3): 140–144.CrossRefGoogle Scholar
  55. Lundvall, B.-Å. (1992) National Innovation Systems: Towards a Theory of Innovation and Interactive Learning. London: Pinter.Google Scholar
  56. Lundvall, B.-A., Joseph, K.J., Chaminade, C. and Vang, J. (2009) Handbook of Innovation Systems and Developing Countries. Building Domestic Capabilities in a Global Context. Cheltenham: Edward Elgar.CrossRefGoogle Scholar
  57. Morris, M., Kaplinsky, R. and Kaplan, D. (2012) One thing leads to another: Commodities, linkages and industrial development. Resources Policy 37(4): 408–416.CrossRefGoogle Scholar
  58. Narula, R. (2014) Foreign direct investment as a driver of industrial development: Why is there so little evidence? In: Van Tulder, R., Verbeke, A. and Strange, R. (eds.) Progress in International Business Research, Vol. 8. Bingley: Emerald Group, pp. 45–67.Google Scholar
  59. Narula, R. (2018) Multinational firms and the extractive sectors in the 21st century: Can they drive development? Journal of World Business 53(1): 85–91.CrossRefGoogle Scholar
  60. Narula, R. and Dunning, J. H. (2010) Multinational enterprises, development and globalisation: Some clarifications and a research agenda. Oxford Development Studies 38: 263–287.CrossRefGoogle Scholar
  61. Nelson, R. (1993) National Innovation Systems: A Comparative Analysis. New York: Oxford University Press.Google Scholar
  62. Nomaler, Ö. and Verspagen, B. (2014) Analysing global value chains using input-output economics: Proceed with care. UNU-MERIT Working Paper No. 2014-070, UNUMERIT, Maastricht.Google Scholar
  63. OECD. (2016) OECD.Stat. Paris: OECD.Google Scholar
  64. Pietrobelli, C. and Staritz. C. (2018) Upgrading, interactive learning, and innovation systems in value chain interventions. European Journal of Development Research. Scholar
  65. Ponte, S. and Ewert, J. (2009) Which way is ‘up’ in upgrading: Trajectories of change in the value chain for South African wine. World Development 37(10): 1637–1650.CrossRefGoogle Scholar
  66. PRS Group (2014) International Country Risk Guide: Table 3B: Political Risk Points by Component. Syracuse, NY: PRS Group.Google Scholar
  67. Rodrik, D. (1994) King Kong meets godzilla: The World Bank and the East Asian miracle. In: Fishlow, A., Gwin, C., Haggard, S. and Rodrik, D. (eds.) Miracle or Design? Lessons from the East Asian Experience, Vol. 11. Washington, DC: Overseas Development Council, pp. 13–38.Google Scholar
  68. Rojec, M. and Knell, M. (2017) Why is there a lack of evidence on knowledge spillovers from foreign direct investment? Journal of Economic Surveys. Scholar
  69. Saxenian, A. (2006) The New Argonauts: Regional Advantage in the Global Economy. Cambridge: Harvard University Press.Google Scholar
  70. Schmitz, H. (1995) Small shoemakers and Fordist giants: Tale of a supercluster. World Development 23: 9–28.CrossRefGoogle Scholar
  71. Schmitz, H. (1999) Global competition and local cooperation: success and failure in the Sinos Valley, Brazil. World Development 27(9): 1627–1650.CrossRefGoogle Scholar
  72. SCImago. (2016) SCImago Journal & Country Rank. Scopus database, Elsevier,, accessed 1 June 2017.
  73. Stata. (2005) Stata Data Management, Reference Manual, Release 9. College Station: Stata Press.Google Scholar
  74. Stata. (2017) Stata Base Reference Manual, Release 15. College Station: Stata Press.Google Scholar
  75. Sturgeon, T. (2001) How do we define value chains and production networks? IDS Bulletin 32: 9–18.CrossRefGoogle Scholar
  76. Sturgeon, T.J. (2002) Modular production networks: A new American model of industrial organization. Industrial and Corporate Change 11(3): 451–496.CrossRefGoogle Scholar
  77. Tajoli, L. and Felice, G. (2018) Global value chains participation and knowledge spillovers in developed and developing countries: An empirical investigation. European Journal of Development Research. Scholar
  78. Timmer, M.P., Dietzenbacher, E., Los, B., Stehrer, R. and de Vries, G.J. (2015) An illustrated user guide to the world input–output database: The case of global automotive production. Review of International Economics 23: 575–605.CrossRefGoogle Scholar
  79. Timmer, M., Erumban, A., Los, B., Stehrer, R. and de Vries, G. (2014) Slicing up global value chains. Journal of Economic Perspectives 28(2): 99–118.CrossRefGoogle Scholar
  80. UN. (2016) UN comtrade database. United Nations, New York,, accessed 1 June 2017.
  81. UNCTAD. (2016) UNCTAD statistics on-line. Geneva: UNCTAD,, accessed 1 June 2017.
  82. UNCTAD. (2013) Global value chains and development: Investment and value added trade in the global economy. Geneva: UNCTAD,, accessed 1 June 2017.
  83. UNCTAD/Eora. (2016) The trade in value added database (TiVA)., accessed 1 June 2017.
  84. UNDP. (2016) Human development report 2015. United Nations Development Programme. New York,, accessed 1 June 2017.
  85. UNESCO. (2016) On-line Data Centre. Geneva: UNESCO Institute for Statistics,, accessed 1 June 2017.
  86. USPTO. (2016) Extended year set - patents by country, state, and year utility patents (December 2015). U.S. Patent and Trademark Office,, accessed 1 June 2017.
  87. Wagner, C.S. and Leydesdorff, L. (2005) Network structure, self-organization, and the growth of international collaboration in science. Research Policy 34: 1608–1618.CrossRefGoogle Scholar
  88. WIPO. (2016) WIPO statistics database. World Intellectual Property Organization,, accessed 1 June 2017.
  89. World Bank. (2015) How does the World Bank classify countries? World Bank,, accessed 1 June 2017.
  90. World Bank. (2016) World development indicators 2015. World Bank, Last Updated on 17 February 2016Google Scholar
  91. Xu, B. (2000) Multinational enterprises, technology diffusion, and host country productivity growth. Journal of Development Economics 62(2): 477–493.CrossRefGoogle Scholar

Copyright information

© European Association of Development Research and Training Institutes (EADI) 2018

Authors and Affiliations

  • Jan Fagerberg
    • 1
    • 2
  • Bengt-Åke Lundvall
    • 2
  • Martin Srholec
    • 3
    • 4
  1. 1.Center for Technology, Innovation and Culture (TIK)University of OsloOsloNorway
  2. 2.IKE, Department of Business and ManagementAalborg UniversityAalborgDenmark
  3. 3.Center for Economic Research and Graduate Education-Economics Institute (CERGE-EI)a joint workplace of Charles University and the Economics Institute of the Czech Academy of Sciences111 21 PragueCzechia
  4. 4.Centre for Innovation, Research and Competence in the Learning Economy (CIRCLE)Lund UniversityLundSweden

Personalised recommendations