The ladder of internationalisation modes

Gábor Békés, Balázs Muraközy 28 March 2018



Globalisation and market integration not only provides an opportunity for more firms to serve foreign markets, it has also opened the door to choose from a number of ways (or modes) to serve those markets. Foreign direct investment (FDI) and outsourcing, for example, have become viable options for a growing number of firms. This phenomenon raises a number of questions. For example, does exporting via intermediaries provide a low-cost alternative for many small and medium-sized enterprises (SMEs) to enter foreign markets, thereby making subsidising direct export less relevant? Similarly, can direct export be made more effective by operating sales affiliates in the foreign market?

A simple but informative way to study internationalisation mode choice is to compare the difference, or premium, in the average performance of firms choosing the different modes. It has been documented that some modes can be ‘ranked’ in terms of productivity premia. The most productive firms choose modes which may require a large up-front investment but require a lower marginal cost of supplying the market. Helpman et al. (2004) document such a sorting between FDI, exporting and domestic production, while similar patterns have been suggested for the choice between indirect and direct export (Ahn et al. 2011, Felbermayr and Jung 2011, Lu et al. 2014) and between export, outsourcing, and FDI (Greenaway and Kneller 2007, Tomiura 2007).

However, mainly because of the scarcity of information, most studies have focused on a few internationalisation modes only. In a recent study we investigate how firms sort into a rich set of internationalisation modes – a setup which we call the ‘ladder of internationalisation’ (Békés and Muraközy 2018). The study extends earlier work in three directions. First, we can investigate a fuller set of modes and their combinations than most previous authors. Second, by observing a rich set of firm attributes, we can show that innovation plays a similarly important role in mode choice to total factor productivity (TFP), which has also been suggested by an earlier Vox column using the same dataset (Altomonte et al. 2014). Third, we can focus on activities when the primary aim of foreign production is to serve foreign customers (horizontal motive) rather than to produce intermediate inputs for further processing (vertical motive).

How do firms internationalise?

We use an established survey database of European manufacturing firms with at least ten employees gathered in the European Firms in a Global Economy (EFIGE) project. This cross-sectional dataset covers five European countries (France, Germany, Hungary, Italy, and Spain) and includes 8,600 firms surveyed in 2009. The dataset is unique as it includes direct information on internationalisation activities, innovation and financial performance at the firm level. The questionnaire distinguishes several modes of serving the foreign market (summarised in Table 1).

Table 1 Definition and prevalence of internationalisation modes

Table 1 shows the share of firms conducting each of these modes. First, 61.5% of these (at least ten employees, manufacturing) firms are internationalised. The dominant way of internationalisation is direct exporting – the overwhelming majority of internationalised firms conduct it. The role of indirect exporting is relatively small. This suggests that indirect exporting is generally not an effective alternative to direct exporting (but this may also partly result from the restrictive definition of indirect export in the EFIGE survey). While the number of firms with FDI is not very large, such firms tend to have many employees, therefore their employment share is far beyond their share in terms of numbers. Interestingly, service FDI seems to be more frequent than manufacturing FDI, suggesting that sales or other service affiliates may provide an effective way of penetrating foreign markets without investing into host country manufacturing.

Another interesting feature of these data, shown in the last column, is that 20% of firms serving foreign markets are engaged in more than one mode. Some modes are particularly frequently complemented this way. For example, more than half of the firms conducting indirect exporting also perform other modes – mostly direct export. Similarly, 83% of firms engaged in outsourced manufacturing will conduct at least one other mode. This phenomenon also raises methodological issues when estimating premia. How to assign premia to different modes when a firm conducts multiple modes? Are these additive? Or should one assign the premia to the most ‘demanding’ mode (Békés and Muraközy 2016)? In line with the logic of the self-selection models, we choose the latter option, which we dub as ‘top-coding’.

The ladder of TFP

Figure 1 shows the productivity premia, in terms of TFP, of firms conducting each mode. There is a clear sorting pattern. Direct exporters are about 40% more productive than non-internationalised firms, while firms conducting FDI are more than 100% higher TFP than non-internationalised firms. Our regression results, which control for conducting multiple modes, show that indirect exporters are not significantly more productive than non-exporters, while outsourcing firms are very similar to direct exporters.

Figure 1 Average TFP of firms conducting different internationalisation modes (no. foreign sales=100), after ‘top-coding’

Notes: Wooldridge (2009) TFP. Firms are classified to the most ‘demanding’ (highest average TFP premium) mode they conduct.

Based on these patterns, one may claim that the ladder of internationalisation has three main rungs: non-internationalisation, direct export, and FDI (either service or manufacturing). This pattern is very much in line with the framework of Helpman et al. (2004).

A ladder of innovation?

Based on the questionnaire, we also define a measure of innovativeness. We use three innovative input and output variables when creating our innovativeness measure. We find that the innovation variable is highly significant in our mode-choice regressions even when controlling for TFP; if anything, its magnitude seems to be larger than that of TFP. One interpretation is that both product attributes (proxied by Innovativeness) and physical efficiency (proxied by TFP) are important determinants of internationalisation mode choice (Figure 2). (Alternatively, the two measures may proxy one ‘deep’ firm-level capability.)

Interestingly, innovativeness is more continuously correlated with internationalisation modes than TFP. Indirect exporters are more innovative than non-internationalised firms, while outsourcing is associated with more innovation than direct exporting. This highlights that the sorting process in indeed multi-dimensional, different steps requiring different combinations of capabilities. 

Figure 2 The semi-elasticity of the probability of choosing the different modes with respect to TFP and innovativeness


Internationalisation mode choices are, indeed, similar to a ‘ladder of internationalisation’. In terms of TFP, there are two main steps of this ladder besides non-internationalisation: one leading to direct exporting, and another to FDI. While TFP premia provide a good characterisation of the choice process, product innovation also seems to be an important part of successful internationalisation strategies.

We can also draw a few conclusions regarding the individual internationalisation modes:

  • Direct exporting and manufacturing FDI are still the dominant global sales platforms.
  • Service FDI seems to provide and important avenue to reach foreign customers.
  • In contrast, indirect exporting via home country intermediates offers a channel to international markets only for relatively few firms.

These results are relevant for at least two reasons.

  • First, identifying how firms choosing diverse internationalisation modes differ from each other helps managers or policymakers in understanding what capabilities are needed to make different internationalisation modes profitable. In particular, our results suggest that both productivity and product market innovation are needed for stepping up on the ladder.
  • Second, the ladder of internationalisation hints at an additional margin of adjustment to policy shocks besides entering and exiting foreign markets, i.e. the mode of foreign sale. This may provide an important source of flexibility for managers and can also affect the aggregate behaviour of the economy during shocks.


Ahn, J, A K Khandelwal and S J Wei (2011), “The role of intermediaries in facilitating trade”, Journal of International Economics 84(1): 73-85.

Altomonte, C, T Aquilante, G Békés and G Ottaviano (2014), “Internationalisation and innovation of firms: Give them one roof”,,  21 March.

Békés, G and B Muraközy (2016), “Measuring productivity premia with many modes of internationalization”, Economics Letters 139: 61–64.

Békés, G and B Muraközy (2018), “The ladder of internationalization modes: Evidence from European firms”, CEPR Discussion Paper No. 12639; forthcoming in Review of World Economics.

Felbermayr, G  and B Jung (2011), “Trade intermediation and the organization of exporters”, Review of International Economics 19(4): 634-648.

Greenaway, D and R Kneller (2007) “Firm heterogeneity, exporting and foreign direct investment”, The Economic Journal 117(517).

Helpman, E, M J Melitz amd S R Yeaple (2004), “Export versus FDI with heterogeneous firms”, American Economic Review 94(1): 300-316.

Lu, J, Y Lu and Z Tao (2014), “Pure exporter: Theory and evidence from China”, The World Economy 37(9): 1219-1236.

Tomiura, E (2007), “Foreign outsourcing, exporting, and FDI: A productivity comparison at the firm level”, Journal of International Economics 72(1): 113-127.

Wooldridge, J M (2009), “On estimating firm-level production functions using proxy variables to control for unobservables”, Economics Letters 104: 112-114.



Topics:  International trade

Tags:  exports, FDI, internationalisation mode

Assistant Professor at Central European University in Budapest, Research Fellow at Institute of Economics in Hungary and CEPR

Research Fellow, Institute of Economics, Hungarian Academy of Sciences


CEPR Policy Research