Understanding Productivity and Employment in a Fragmenting Economy
The Global Value Chain Approach
Marcel Timmer
Introduction
Production processes in today's world no longer take place in one location: instead goods and services are produced in intricate regional production networks feeding into each other. This fragmentation of production across sectors and countries has pervasive implications for local labour markets, driving new patterns of productivity and employment growth. Take the production of the iPod, which is exemplary: designed in the United States (US), assembled in China based on several hundreds of components and parts that are sourced from around the world. In a seminal study Linden, Dedrick, and Kraemer (2011) found that 'in 2006, the iPod supported nearly twice as many jobs offshore as in the US. Yet the total wages paid in the US amounted to more than twice as much as those paid overseas. Driving this result is the fact that Apple keeps most of its research and development (R&D) and corporate support functions in the US, providing thousands of high-paid professional and engineering jobs that can be attributed to the success of the iPod.'[1] Anecdotal evidence like this suggests that advanced countries are increasingly specializing in skill- and capital-intensive activities within global value chains (GVCs), more popularly described as a process of turning into 'headquarter economies'. As a result firms and countries no longer trade goods, but tasks.
How to measure and analyse productivity, employment, and wages in such a fragmenting global economy? Foxconn in China is producing iPods using intangible designs and technology from Apple. But these services are typically not recorded in production and trade statistics, such that any study of the productivity of Chinese and US manufacturing is seriously hampered. Likewise, without the explicit modelling of substitution possibilities between Chinese and US workers, shifts in local labour demand are difficult to analyse. A new conceptual framework is needed which goes beyond the traditional analysis of separate firms, industries, or countries. In this chapter we introduce the GVC approach, which combines recent new insights in the literature on international trade, the so-called 'trade in tasks', see Grossman and Rossi-Hansberg (2008), and in labour economics, the 'task-approach to employment and earnings', see Acemoglu and Autor (2011).[2] In the GVC approach we model production as a set of discrete activities in distinct locations, which altogether form a supply chain starting at the conception of the product and ending at its delivery. We trace the value added by labour and capital in each activity by means of input-output (IO) analysis rooted in the seminal work by Leontief.[3] This provides new opportunities to analyse substitution possibilities between various types of labour, both domestic and foreign, as well as between capital and labour. It also offers for the first time the opportunity to measure the possible factor biases in technological change.
Apart from being conceptually appealing, this approach also bypasses some of the empirical problems that confront current productivity analyses. When fragmentation is high, accurate measurement of prices of intermediates becomes paramount to measure productivity. However, there is increasing doubt about the reliability of price indices for imported intermediates.[4] Even more serious is the problem of measuring flows and prices of intangible services such as the use of software, patents, brand names, or logistics. Intangibles are becoming increasingly important in production and are making up a major share of investment by firms. But so far the measurement of intangible output, and in particular use, appears to be challenging.[5] For example, the use of Apple's intangible designs and technology by Foxconn is typically not recorded in production and trade statistics. How to measure productivity of firms and sectors without information on the quantity and price of the most valuable inputs? The GVC approach offers a first step towards a new framework that takes this important but elusive characteristic of modern production systems as a point of departure.
The rest of the chapter is organized as follows. We first outline in Section 3.2 evidence for the pervasiveness of the international fragmentation process. The GVC approach is presented in Section 3.3 and illustrated with an analysis of German car manufacturing. Functional specialization in GVCs is discussed in Section 3.4. Section 3.5 offers some concluding remarks on whether the emergence of GVCs can explain, or raises doubt about, the recent slowdown in measured productivity in the EU.
- [1] Dedrick et al. (2010) show similar results for some other high-end electronic productssuch as notebooks. See also Ali-Yrkko and Rouvinen (2015) for a wide set of Finnish goods.
- [2] The GVC approach is also used to denote a longer research tradition in economicsociology, and was introduced by Gary Gereffi and co-workers. In this line of work emphasisis mainly put on analysis of the governance in production chains, and in particular theprospects for upgrading for less advanced countries and firms. Gereffi (1999) provides a goodintroduction to this work for economists.
- [3] See Miller and Blair (2009) for an overview.
- [4] See contributions in Houseman and Mandel (2015) for studies of the possible mismeas-urement of the import price of semi-conductors and its implications for measured productivity in US manufacturing.
- [5] See Corrado etal. (2012) for pioneering attempts.
