The emergence and increased importance of global value chains has meant that the production process for most goods and services has become increasingly fragmented and geographically dispersed. For the developed countries within the OECD, this has involved that substantial parts of manufacturing and assembly within the value chains have been relocated to low-wage countries. More value-creating parts, particularly at the beginning of the value chains, for example R&D and design, but also at the end of the value chains, however, still largely remain in the developed countries.
This is a pattern that emerges when we study the Swedish multinational companies and how the composition of the workforce at the parent companies in Sweden is affected when they increase activity in their foreign subsidiaries. The proportion of skilled labour then increases at the parent company and the more routine tasks become fewer.
A growing relative demand for skilled labour due to the development described above and in combination with skilled-biased technological change seems, however, to have been counteracted by the substantial expansion of higher education in Sweden in recent years. Apart from a small increase in the relative wage for labour with post-secondary education at the end of the 1990s, this has remained largely constant over the past fifteen years.
The increasingly important role that service-related parts have come to play within the framework of the global value chains in the more developed countries becomes evident when we divide the Swedish economy into three parts: tradable services, non-tradable services and manufacturing. Here we find that within that part of the Swedish economy that is exposed to international trade, tradable services has over the past twenty years grown at the expense of manufacturing.
From a regional perspective, it is worth noting that tradable services industries are concentrated to regions with dense populations and a high proportion of skilled labour. The location pattern of manufacturing is related to neither the size of the region nor the human capital intensity.
It is also notable that the share of skilled labour has increased faster in the tradable sectors than in the non-tradable service sector. One possible explanation is that in these highly internationalised parts of the economy there has been a particularly strong trend towards less skilled jobs disappearing at the same time as more skilled jobs are created.
A person’s job also seems to involve a wage premium for those who work in industries and occupations that are tradable; these people’s wages are almost 12% higher compared to those who have similar education and other observable characteristics but who work in other industries and occupations. The premium might partly constitute some form of compensation because the probability of losing one’s job is higher in the tradable industries and the loss of income more noticeable for those who lose their jobs in these industries. On the other hand, the probability of finding a new job is greater for people who were employed in tradable services when they lost their jobs than for those who had worked in non-tradable services and in manufacturing. The adjustment costs due to structural changes in the tradable part of the economy seem generally to be higher for those who are displaced in manufacturing. The probability of losing one’s job is less, it is true, but the loss of income is greater and the probability of being re-employed is lower for people being displaced in manufacturing than in tradable services.
Finally, it is clear that the long-debated Swedish R&D paradox might very well be able to be explained within the framework of an argument based on global value chains. Swedish multinationals tend quite simply to locate their R&D activities in Sweden and their manufacturing facilities elsewhere. That what might at first glance appear to be a paradox can very well be explained by means of a simple economic model underlines the importance of a well-founded analysis being made of a perceived problem before beginning to consider conceivable policy measures. Another lesson that can be drawn from the discussions concerning the R&D paradox and the increasing importance of global value chains is that it has become much more difficult to identify international competitiveness using traditional measures of competitiveness based solely on production values.
Global value chains and international competitiveness