Trade, technology and skills: Evidence from Turkish microdata

Abstract

In this paper we report evidence on the relationship between trade openness, technology adoption and the relative demand for skilled labour in the Turkish manufacturing sector, using firm-level data over the period 1980–2001. In a dynamic panel data setting, using a unique database comprising data from 17,462 firms, we estimate an augmented cost share equation whereby the wage bill share of skilled workers in a given firm is related to international exposure and technology adoption.

It emerges that R&D expenditures are positive and significantly related to skill upgrading. This result supports the skill-biased technological change argument in the case of a middle-income country such as Turkey.

Moreover, the firm-level analysis reveals a positive impact of technological transfer from abroad, foreign ownership and exporting status on the demand for skills, highlighting the role of increasing international openness in fostering skill upgrading within firms.

Our microdata also allow us to investigate the direct impact of import flows in shaping the relative demand for skills. The results show that those firms belonging to the sectors experiencing rapid increase in the share of inputs imported from industrialised countries also experience a higher increase in the labour cost share of skilled workers. This finding provides further support for the hypothesis that imports from industrialised countries imply a transfer of new technologies, in turn leading to a higher demand for skilled labour (the so-called skill-enhancing trade hypothesis).

Introduction

This paper examines the relationship between trade openness, technology adoption and the relative demand for skilled labour¹ in Turkish manufacturing firms. Turkey started a marked process of liberalisation in the early ’80s, and volumes of exports and imports have continued to grow since then, linking the Turkish economy increasingly to the world market. This increasing trade openness has particularly affected the manufacturing sector, where most of the growth in imports and exports has occurred. An important aspect of this process could be its impact on labour demand, and, more specifically, its impact on the relative demand for skilled labour. Indeed, over the same period, the relative demand for skilled labour increased substantially, leading to higher wage-gaps between skilled and unskilled workers. Whether these two simultaneous phenomena are linked has not yet been established.

The theoretical economic literature provides different predictions regarding the impact of trade liberalisation on labour demand in a developing country. On the one hand, according to the central tenet of traditional trade theory, as expressed in the Heckscher–Ohlin theorem and in its Stolper–Samuelson corollary (HOSS hereafter), we should expect a relative decrease in the demand for skilled labour. Openness should benefit a country's relatively abundant factor, because trade specialisation should favour sectors intensive in the abundant factor. Therefore, in the case of Turkey, which is abundant in unskilled labour compared to its main trading partner, the EU, trade openness should have increased the demand for unskilled workers and raised their relative wages. On the other hand, if the HOSS assumption of homogeneous production functions² among countries is relaxed, then international openness may facilitate technology diffusion from more industrialised countries. Imports, exports and foreign direct investment (FDI) may act as a channel of technological upgrading, and shift the production function towards more skill-intensive technologies. In other words, trade and FDI may induce and foster skill-biased technological change (SBTC).

This paper contributes to the debate by presenting new empirical evidence at the firm level. We estimate the impact of trade openness on labour demand by using a unique detailed panel of Turkish manufacturing firms (Annual Manufacturing Industry Statistics by the Turkish Statistical Institute, TurkStat). The dataset, by covering all manufacturing firms employing 10 or more people and comprising data from 17,462 different firms, represents about 90% of Turkish (formal sector) manufacturing output and is therefore particularly well-suited to quantifying the possible impact of firms' international exposure on the labour demand for skilled workers.

Turkey is a particularly interesting context for studying the relationship between trade openness, technology adoption and the relative demand for skilled labour. Indeed, it is a middle-income country with sizeable trade flows with more advanced countries, especially the European Union.³ This makes it a net technology importer, and hence the ideal country for investigating the impact of potential imported SBTC. Indeed, trade openness is supposed to have a greater inequality-enhancing impact in middle-income countries, where “social capabilities” (Abramovitz, 1986) and “absorptive capacity” (Lall, 2004) are higher and where technology adoption from more advanced countries effectively acts as a channel for technological upgrading, thus leading to an increase in the relative demand for and wages of skilled workers (see Meschi and Vivarelli, 2009, Almeida, 2009). From a policy perspective, Turkey presents an illustrative example for investigating the social impact of international openness: during the ’80s, Turkey carried out a significant trade liberalisation programme, shifting from a protectionist model characterised by heavy state intervention to a more outward-looking one.

The remainder of this paper is organised as follows: the next section reviews the theoretical and empirical literature on the interaction between trade openness, technological change and the relative demand for skilled labour, focusing mainly on developing countries. Section 3 introduces and describes the data. In Section 4 we discuss historical trends in the Turkish economy and present some descriptive statistics. In Section 5 we present our econometric analysis and discuss the empirical results. Finally, the last section puts forward some concluding remarks.