This article is published in collaboration with Vox EU.

Over the last three decades, the perception that unskilled workers are being ‘left behind’ has become a matter of political concern in industrialised economies. The increase in the skill premium, i.e. the wage of skilled workers relative to that of unskilled workers, and the larger unemployment rates among unskilled workers have prompted researchers and policymakers to look into the causes of and the potential remedies to this state of things.

Figure 1. Relative unemployment rate for high-skilled (tertiary education or more) versus low-skilled (lower secondary education or less)

Source: ILO and own calculations.

Culprits for the skill premium

In the search for culprits, two rather apparent phenomena have been the usual suspects. Both skill-biased technical progress and globalisation raise the relative demand for skilled to unskilled workers in the rich world. According to the technical progress explanation, the ICT revolution has biased relative labour demand in favour of skilled workers since modern technologies are operated with more brain and less brawn than those of the industrial era. Computers and skilled workers, in other words, complement each other well, leaving unskilled workers less and less space in the labour market. The globalisation argument is usually framed in terms of the Heckscher-Ohlin (HO) model; the integration of unskilled-labour abundant countries (first the East Asian tigers and Mexico, then China and India) into the world economy leads to an expansion of the rich (skill-abundant) world’s industries that use skilled labour intensively and a contraction of its unskilled-labour-intensive industries, thereby triggering a fall in the demand for unskilled labour and an increase in the demand for skills.

The majority of economists have usually tilted towards the ICT explanation, partly due to the empirical shortcomings of the HO model. Rather than a rise in the relative demand for skills via a change in industrial composition and a decrease in the skill intensity of production processes due to the higher skill premium – as the HO model would predict for skill-abundant countries in the face of freer trade – we observe increases in the skill intensities of all industries across the board (Grilicheset al. 1994). Moreover, the increase in skill intensities has not only occurred in skill-abundant countries, but also in developing countries, where the HO model predicts precisely the opposite effect due to a change in industrial composition away from skill-intensive industries. Finally, aggregate country or industry studies found only weak support for the causal effect of trade on changes in the demand for skills.

Offshoring and skills upgrading: New evidence

This traditional view has only recently started to change as micro-level data, which allow tracking the impact of international trade on individual firms and workers much more precisely, have become widely available. A number of empirical papers (e.g. Autor et al. 2013 for the US, Mion and Zhu 2013 for Belgium) have provided convincing evidence that import competition from low-wage countries, in particular from China, indeed has a large impact on labour demand in rich countries, hurting mostly unskilled workers, while benefiting skilled ones. Similarly, Hummels et al. (2014) find negative effects of offshoring decisions of Danish firms on wages of unskilled workers, while we take a detailed look at the offshoring behaviour of manufacturing firms in France in the period 1996-2007 (Carluccio et al. 2015). We point out a number of stylised facts that suggest a strong link between offshoring and skill upgrading. Within a given French industry, firms are very different in the amount of skilled relative to unskilled workers they employ in production. Firms that import from labour-abundant countries are much more skill-intensive in their domestic production than non-importers, and the positive trend in offshoring to these countries since the mid-1990s runs parallel to the surge in the skill intensities of those importers.

Figure 2. Share of imports from labour abundant countries in total French manufacturing imports and log skill ratio

Note: Log skills ratio is defined as the ratio of skilled workers to unskilled workers in French manufacturing.

Figure 3. Skill ratio of importers from labour-abundant countries, all firms and non-importers against time

These results and the evidence discussed below suggest that HO forces are operating at the within-industry rather than the between-industry level, but economists still lack theoretical models that explain how these forces manifest themselves. We develop such a model to understand the effects of offshoring practices on the relative demand for skilled workers. Its intuitions build on Feenstra and Hanson’s (1997) offshoring model, but the combination with insights from the heterogeneous-firm trade literature delivers a novel set of empirical predictions, which can be tested with firm-level data.

We assume that producers of final goods use intermediate inputs that differ in their skill intensities, and that countries have different relative factor endowments, as in the traditional HO theory. Firms are heterogeneous in terms of productivity, and offshoring of intermediates requires the payment of per-input fixed offshoring costs. Firms in a skill-abundant country must therefore weigh the lower cost resulting from offshoring a labour-intensive input to a labour-abundant country against the fixed costs implied by such a decision.

  • Low-productivity firms end up producing all inputs domestically.

The cost reduction from offshoring does not compensate the corresponding investment due to the small revenues generated by these firms.

  • Sufficiently productive firms instead offshore the most labour-intensive inputs to very labour-abundant countries.
  • Firms with even higher productivity levels find it profitable to also import relatively more skill-intensive inputs from not-so-labour-abundant countries.

Such imports substitute for domestic unskilled employment, making domestic production more skill intensive. Thus, at the firm level, productivity and skill intensity of domestic production are positively correlated, generating endogenous within-industry variation in skill intensities.

Our econometric analysis provides empirical support for all these correlations, which obviously point to a direct link between offshoring and changes in the demand for skilled workers relative to unskilled workers. Finally, we exploit exogenous supply shocks in France’s trading partners and reductions in EU external tariffs in order to provide causal evidence that the increase in imports from labour-abundant countries has led to a substantial increase in the French manufacturing industry’s skill intensity over the sample period (around 10 percentage points). We find that virtually all of the observed within-firm changes in skill intensity can be explained by increased offshoring to labour-abundant countries.


The empirical evidence suggesting that globalisation is indeed affecting the income distribution in industrialised countries is much stronger than initially thought, at least as far as the manufacturing industry is concerned. In fact, the productivity gains from having access to cheaper inputs through offshoring are not being distributed equally between different economic actors in our rich societies. Consequently, political resistance to free trade by certain groups has a clear rationale as long as the effects discussed above are not addressed through more effective redistribution policies.

Authors’ note: The views expressed here are those of the authors only and do not necessarily reflect the views of the Banque de France or the Eurosystem. 


Autor, D, D Dorn and G Hanson (2013), “The China Syndrome: Local Labour Market Effects of Import Competition from China”, American Economic Review 103(6): 2121–2168.

Berman, E, J Bound, and Z Griliches (1994), “Changes in the demand for skilled labour within U.S. manufacturing: Evidence from the annual survey of manufacturers”, Quarterly Journal of Economics109, 367-397.

Carluccio, J, A Cunat, H Fadinger and C Fons-Rosen (2015), “Offshoring and Skill-Upgrading in French Manufacturing: A Heckscher-Ohlin-Melitz View”, CEPR Discussion Paper 10864.

Feenstra, R and G Hanson (1997), “Foreign direct investment and relative wages: Evidence from Mexico’s maquiladoras”, Journal of International Economics, 42 (3-4), 371-393.

Hummels, D, R Jorgensen, J R Munch and C Xiang (2014), “The Wage Effects of Offshoring: Evidence from Danish Matched Worker-Firm Data”, American Economic Review, 104(6), 1597-1629.

Mion, G and L Zhu (2013), “Import competition from and offshoring to China: A curse or blessing for firms?” Journal of International Economics, 89(1), 202-215.

Publication does not imply endorsement of views by the World Economic Forum.

To keep up with the Agenda subscribe to our weekly newsletter.

Authors: Juan Carluccio is a Research Economist, Banque de France; Senior Lecturer in International Trade, University of Surrey. Alejandro Cuñat is professor of economics at the University of Vienna and a research fellow at CES-ifo and CeFiG. Harald Fadinger is an Associate Professor of International Economics at the University of Mannheim. Christian Fons-Rosen is an Assistant Professor of Economics at Universitat Pompeu Fabra.

Image: Workers on the assembly line replace the back covers of 32-inch television sets at Element Electronics in Winnsboro, South Carolina. When Walmart pledged in 2013 to buy an extra $250 billion in U.S.-made goods over the next decade, it appeared to be just what was needed to help move America’s putative manufacturing renaissance from rhetoric to reality. REUTERS/Chris Keane.