VoxEU Column Europe's nations and regions International trade

Global sourcing of heterogeneous firms: Theory and evidence

The share of international trade within firm boundaries varies greatly across countries. This column presents new evidence on how the productivity of a firm affects the choice between vertical integration and outsourcing, as well as between foreign and domestic sourcing. The productivity effects found in Spanish firm-level data suggest that contractual imperfections distort the sourcing of inputs in the global economy, and that firm boundaries emerge in response to mitigate this distortion.

One of the oldest and most intricate questions in economics is what determines the boundaries of firms. What motivates firms to seek control over certain parts of the value chain beyond the degree of influence afforded by market transactions? Why do some firms seek more control than others? Why might firms aim for different degrees of control when operating in different markets? Over the past 15 years, interest in these questions was spurred by the empirical observation that the share of international trade taking place within the firm boundaries of control varies greatly across countries, industries and firms (Antràs 2014).

New evidence on firm boundaries in global sourcing

In recent research we provide new evidence on the role of firm productivity in drawing firm boundaries in global sourcing (Kohler and Smolka 2015). Previous studies1 have shown that productivity matters for the allocation of ownership rights between the headquarters of a firm and its intermediate input supplier (vertical integration vs. outsourcing), as well as for the location of intermediate input production (offshore vs. domestic). Our study adds to this literature by investigating more explicitly how and why productivity matters.

We build on the idea, advanced by Antràs (2003) and Antràs & Helpman (2004), that the sourcing of inputs is distorted by contractual imperfections, and that adjusting the firm boundary is a direct response to mitigate this distortion – by assigning ownership rights to the party undertaking the more important investment, a firm can minimise the efficiency loss caused by lack of enforceable contracts (Grossman and Hart 1986, Hart and Moore 1990). The productivity of the firm plays an important role in this decision, because it magnifies the benefits of choosing the more efficient ownership allocation.

Consistent with these ideas, our empirical analysis suggests that firms react very differently to gains in productivity depending on which party (the headquarter or the input supplier) contributes the more important investment (measured by the sourcing intensity of production). We find a pattern of effects whereby productivity stimulates vertical integration in industries of low sourcing intensity, but favours outsourcing in industries of high sourcing intensity. Moreover, we find that productivity boosts offshoring throughout all industries, with the strength of the effect varying significantly across industries. Overall, our results suggest an important role for contractual imperfections in shaping input trade around the globe.

Heterogeneity along the distribution of sourcing intensities

We examine data for manufacturing firms in Spain over the period of 2006-2012. Figure 1 displays the prevalence of different sourcing strategies over time. The figure shows pronounced differences in the fractions of firms choosing a particular sourcing strategy, but a roughly stable pattern over time. Domestic outsourcing is by far the most common way of input sourcing (used by roughly 90% of all firms), followed by foreign outsourcing (30%), domestic integration (10%), and foreign integration (5%).

Figure 1. Sourcing strategies of Spanish firms 2006-2012

Our study requires a measure for the sourcing intensity of production. Antràs (2003) argues that the costs of physical capital are easier to share than the costs of labour inputs, and that headquarter firms provide (or pre-finance) machinery and specialised tools and equipment, or assist their suppliers in the acquisition of capital equipment and raw materials (as reported in Dunning 1993, pp455-456). We therefore use the (reversed) scale of industry-specific capital intensities to measure the sourcing intensity of production (normalised to the unit interval [0,1]).

Figure 2 shows that firms using vertically integrated production relationships are concentrated in industries of low sourcing intensity (e.g., those producing beverages, certain metal products, and chemical products). Outsourcing relationships, in contrast, are considerably more important – relative to vertical integration – in industries of high sourcing intensity (e.g., those producing leather and footwear or textiles).

Figure 2. Share of Spanish firms choosing vertical integration by industry

Note: Shares are averaged across the years 2006-2012 and across domestic and foreign sourcing.

Figure 3 shows that gains in productivity stimulate vertical integration in those industries where the sourcing intensity is low, but favour outsourcing where it is high. Moreover, there is evidence for a monotonic adjustment, with the productivity effect becoming gradually more favourable to outsourcing as we move along the distribution of sourcing intensities. Strikingly, we find almost the same pattern of productivity effects when we look at domestic sourcing (h = d) as we do when looking at offshoring (h = f). As for the location of sourcing, we find that gains in productivity increase the likelihood of offshoring throughout all industries, but the strength of the effect varies significantly along the distribution of sourcing intensities. Again, this holds true when sourcing takes place through vertical integration (j = v) as well as for outsourcing (j = o).2

Figure 3. Productivity effects by quintiles of sourcing intensity

Note: The figures show the marginal effects of productivity on the probability of vertical integration (upper panel) and offshoring (lower panel), respectively, obtained from a panel regression including a host of industry- and firm-level controls.


Our results have important implications for scenarios of technological change. For instance, suppose that firms in industrialised countries experience a uniform increase in their productivity. Our results imply that the share of multinational firms will consequently increase in those industries where headquarter inputs loom large, and decrease where they are of minor importance. Because multinational firms are effectively transmitting technological knowledge across borders (Guadalupe et al. 2012), this has far-reaching implications for technological diffusion and industrial structure in developing countries. In particular, what starts out as a neutral technological improvement in industrialised countries will, through differential ownership allocation, arrive as a non-neutral technological improvement in developing countries, favouring input suppliers in capital-intensive industries.

As for the location of input production, our results suggest that in such a scenario of technological progress firms in all industries will see offshoring become more attractive, but the effects will be felt differently across industries, thus accelerating job turnover in some industries more than in others. The interesting point here is that the cross-industry heterogeneity in offshoring is explained by the sourcing intensity of production, independently of the standard measures of offshorability (e.g., the prevalence of routine and impersonal tasks; Blinder 2009).


Antràs, Pol (2003), “Firms, Contracts, and Trade Structure”, Quarterly Journal of Economics 118(4):1375-1418.

Antràs, Pol (2014), Global Production: Firms, Contracts, and Trade Structure, Harvard University mimeo.

Antràs, Pol and Elhanan Helpman (2004), “Global Sourcing”, Journal of Political Economy 112(3):552-580.

Blinder, Alan (2009), Offshoring: Big Deal, or Business as Usual? In: B. M. Friedman (ed.) Offshoring of American Jobs. What Response from U.S. Economic Policy?, chap. 2, (pp. 19-60). Cambridge, MA: MIT Press.

Corcos, Gregory, Delphine Irac, Giordano Mion, and Thierry Verdier (2013), “The Determinants of Intrafirm Trade”, Review of Economics and Statistics 95(3):825-838.

Defever, Fabrice and Farid Toubal (2013), “Productivity, Relationship-Specific Inputs and the Sourcing Modes of Multinational Firms”, Journal of Economic Behavior & Organization 94:345-357.

Dunning, J H (1993), Multinational Enterprises and the Global Economy, Addison Wesley Longman, Inc.

Grossman, Sanford and Oliver Hart (1986), “The Costs and Benefits of Ownership: A Theory of Vertical and Lateral Integration”, Journal of Political Economy 94(4):691-719.

Guadalupe, Maria, Olga Kuzmina, and Catherine Thomas (2012), “Innovation and Foreign Ownership”, American Economic Review 102(7):3594-3627.

Hart, Oliver and John Moore (1990), “Property Rights and the Nature of the Firm”, Journal of Political Economy 98(6):1119-1158.

Kohler, Wilhelm and Marcel Smolka (2014), “Global Sourcing and Firm Selection”, Economics Letters 124(3):411-415.

Kohler, Wilhelm and Marcel Smolka (2015), “Global Sourcing of Heterogeneous Firms: Theory and Evidence”, CESifo Working Paper No. 5184.


1 Defever and Toubal (2013), Corcos et al (2013), Kohler and Smolka (2014).

2 The effect for the fifth quintile is missing for due to an insufficient number of observations.

6,824 Reads