International trade has long been considered to be a channel of knowledge transfer. Firms that export are expected to absorb new knowledge from foreign markets and buyers, thereby enjoying better productivity performance than non-exporting firms. Yet, empirical evidence on ‘learning by exporting’ is mixed at best. Although there is pervasive evidence on the superior productivity performance by exporters, such advantage has been mostly attributed to the self-selection of originally productive firms into export (Bernard et al. 2007, Wagner 2007, 2012). Furthermore, supportive evidence suggests that learning by exporting is far from guaranteed, but instead is conditional on several factors. For instance, it tends to occur more when firms are exporting to advanced economies (de Loecker 2007), or when they are exporting multiple products to multiple destinations (Masso and Vahter 2015). In our latest working paper (Benkoviskis et al. 2017), we explore whether learning by exporting can be conditional on the type of activities that exporters engage in global value chains (GVCs).
GVCs encompass a complex network of production and non-production activities. Past case studies have observed that those activities vary greatly in the size of value added they generate (Gerrefi 1999, Dedrik et al. 2010). In particular, highly unique, knowledge-intensive activities like new product development, design, manufacturing of key components, as well as marketing and branding, are disproportionally better remunerated than standardised and labour-intensive activities such as the assembly of final goods or supply of generic materials. Firms engaging in these high value-added activities capture the lion’s share of the total value added generated by a GVC, while other firms face fierce global competition that drives down their profit margins. The uneven distribution of value added among GVC participants is often illustrated as a U-shaped smile curve, because high value-added activities are often concentrated in the upstream and far downstream of a value chain.
Evidence of learning by exporting is more likely to be found for exports that correspond to high-value-added activities within GVCs. Exports of technologically sophisticated or high quality intermediate goods or knowledge-intensive services are likely to be such case. Conversely, the learning effect can be ambiguous for exports of final goods or non-technical services like transport services, because the potential productivity gains from learning may be offset by the strong competition that compresses profit margins.
We test this hypothesis using microdata on Latvian and Estonian firms. For both countries, firm-level data on corporate activities are matched with custom data and data on service trade to capture a firm’s productivity performance and its export contents. We distinguish exports of intermediate goods and those of final products using the OECD Bilateral Trade Data by Industry and End-Use (BTDIxE) classification.1 Service exports are separated into exports of transport services and those of non-transport services which include knowledge-intensive activities such as R&D, information and communication technologies and other professional services. We also highlight re-exports, which are considered to account for 28%, on average, of Latvia’s goods exports between 2005 and 2013 (Beņkovskis et al. 2016). Following Beņkovskis et al. (2016), re-exports are defined as imports and exports of the same goods within the 8-digit Combined Nomenclature within a period of 12 months.
As in many countries, exporters in Latvia and Estonia earn larger value added per employee (labour productivity) than non-exporters. However, the premia of exporters vis-a-vis non-exporters differ across types of exports (Figure 1). After controlling for factors that may affect a firm’s performance, exporters of intermediate goods have larger premia than exporters of final goods. Moreover, exporters of non-transport services enjoy particularly larger premia than other exporters. Large premia on exports of non-transport services and, to a lesser extent, on those of intermediate goods is consistent with the observation that knowledge-intensive activities in the upstream (or far downstream) of GVCs are better remunerated. Interestingly, re-exporters also enjoy significantly larger premia than other goods exporters. This suggests that re-exporting is a knowledge-intensive activity that generates informational rent by intermediating trade between parties with large information asymmetries (Feenstra and Hanson 2004).
Figure 1 Exporters’premia by type of exports, 2006-2014
Note: The figure describes the exporters ‘advantage over non-exporters in labour productivity level. For example, Latvian exporters of intermediate goods on average have 75% higher productivity than non-exporters. The premia was estimated by pooled OLS over 2006-2014 (1995-2014 for Estonia). The OLS regressed log labour productivity against a dummy variable indicating export status while controlling for firm age, foreign ownership, location in capital region as well as 2-digit NACE sector and year fixed effects.
Of course, the large premia associated with some types of exports can be the result of higher entry costs which only allow most productive firms to engage in such exports. We therefore explore the causal effect of those exports on labour productivity using the propensity score-matching method, which allows us to control for the self-selection of better performing firms into each type of exports.
We find that that exporting in general boosts the labour productivity of Latvian firms by 21% and that of Estonian firms by 14% during the three years period that follows the export entry. However, the size of productivity gains differs significantly across exports (Figure 2). Exports of intermediate goods and non-transport services result in sizable and statistically significant gains in productivity of Latvian firms. However, exports of final goods and transport service result in small or insignificant productivity gains. For Estonian firms, all types of exports are associated with positive and significant gains in productivity, but the gains from exports of non-transport services are considerably larger than those from other types of exports. Those findings are consistent with our hypothesis that learning by exporting occurs most when exports correspond to well remunerated activities within GVCs. Interestingly, re-exports are associated with large productivity gains for firms in both countries, perhaps because its knowledge-intensive nature makes learning by exporting particularly important.
Figure 2 Average productivity gains over the three years period following export entry
Note: The chart reports the estimated productivity gains from export entry averaged over the three years period following the export entry (the year of export entry and the two following years). The blank bar indicates statistically insignificant estimates.
Participation in GVCs provides emerging economies with opportunities for fast-track development and technological upgrading (OECD 2013). However, countries need to diversify their exports into knowledge-intensive products and services that generate high value added within GVCs. Due to the very uneven remuneration among GVC participants, countries that specialise in standardised, generic products or services may not enjoy sufficient improvements in productivity, even if such exports channel knowledge transfer. Participation to high value added activities requires strong capabilities. For instance, the share of exporting firms in non-transport services is significantly higher in Estonia which appears ahead of Latvia in terms of R&D expenditure and the share of working age adults with tertiary education degree. Policies that stimulate innovation and strengthen skills would allow countries to move up the value chain and draw larger value added from GVC.
Authors' note: The views expressed here are those of the authors and do not necessarily reflect those of the institutions they represent.
Beņkovskis, K, S Bērziņa and L Zorgenfreija (2016), “Evaluation of Latvia’s re-exports using firm-level trade data”, Baltic Journal of Economics 16(1): 1-20.
Benkovskis, K, J Masso, O Tkacevs, P Vahter and N Yashiro (2017), "Export and productivity in global value chains: Comparative evidence from Latvia and Estonia", OECD Economics Department Working Paper No. 1448.
Bernard, A, B J Jensen, S Redding and P Schott (2007) “ Firms in International Trade”, Journal of Economic Perspectives 21(3): 105-130.
Dedrick, J, K L Kraemer and G Linden (2010) “Who Profits from Innovation in Global Value Chains?: A Study of the iPod and Notebook PCs”, Industrial and Corporate Change 19(1): 81-116.
De Loecker, J (2007), “Do exports generate higher productivity? Evidence from Slovenia”, Journal of International Economics 73: 69–98.
Feenstra, R C and G H Hanson (2004) “Intermediaries in Entrepôt Trade: Hong Kong Re-Exports of Chinese Goods”, Journal of Economics and Management Strategy 13(1): 3–35.
Gereffi, G (1999), “International Trade and Industrial Upgrading in the Apparel Commodity Chain”, Journal of International Economics 48: 37–70
Masso, J and P Vahter (2015). “Exporting and productivity: The effects of multi-market and multi-product export entry”, Scottish Journal of Political Economy.
OECD (2013), Interconnected Economies: Benefiting from Global Value Chains, Paris.
Wagner, J (2007). “Exports and productivity: A survey of evidence from firm-level data”, The World Economy 30(1): 60–82.
Wagner, J (2012), “International trade and firm performance: A survey of empirical studies since 2006.” Review of World Economics/Weltwirtschaftliches Archiv 148(2): 235–267.
 The Bilateral Trade Data by Industry and End-Use (BTDIxE) is used to construct that OECD-WTO Trade in Value added (TiVA) Database, which has been widely used for GVC analysis.