The two ways of FDI R&D spillovers: evidence from the French manufacturing industry

ABSTRACT

Using French firm-level panel data, this study investigates R&D spillovers from inward foreign direct investment (FDI) with respect to both horizontal and vertical linkages (backward and forward). Using a Crepon, Duguet and Mairesse (CDM) model, we estimate an R&D-augmented Cobb–Douglas production function to assess the impact of R&D spillovers on firm performance. The results emphasize that international spillovers (from foreign affiliates to local firms) have a greater effect on firm performance than reverse spillovers (from local firms to foreign affiliates) and are more likely to be backward than forward. Moreover, the effect of backward spillovers depends on a firm’s absorptive capacity and is amplified in the case of outsourcing relationships.

KEYWORDS:

• Backward linkages
• foreign direct investment
• forward linkages
• outsourcing
• R&D spillovers
• reverse spillovers

JEL CLASSIFICATION:

• F23
• O32

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 For a survey, see Spencer (2005).

2 Griffith, Redding, and Van Reenen (2004) confirm the importance of absorptive capacity using a panel of industries across 12 OECD countries (1974–1990), a finding also obtained by Girma (2005) based on a sample of the U.K. manufacturing firms (1989–1999).

3 As the database used in the present research does not provide reliable information on patents or other proxies for innovation output, the second step of the CDM model has not been considered. This choice could be regarded as a limitation of our model. However, patents do not necessarily capture all the innovation efforts of firms (Cohen, Nelson, and Walsh 2000).

4 Note that $D\mathrm{\_}R\text{\&}{D}_{it}=0$ encompasses two alternative cases: R&D stock is either zero or is not available because R&D investment is not reported.

5 This approach allows us to combine two opposite effects. First, there is an escape-competition effect in the sense that competition may stimulate firms to improve their productive efficiency. Second, according to the Schumpeter effect, in a market with fierce competition, firms have few incentives to develop innovations because competition decreases the monopoly rents of prospective innovators.

6 We follow the classification defined by Eurostat: small firms with fewer than 49 employees, medium firms with 50–249 employees, intermediate firms with 250–4,999 employees and large firms with more than 5,000 employees.

7 This method also prevents a potential collinearity problem between the set of regressors in the equation of interest and the inverse Mill’s ratio.

8 We adopt the Eurostat classification for high-, medium- and low-tech sectors.

9 To control for absorptive capacity, Kokko, Tansini, and Zejan (1996) use the extent of the technology gap of domestic firms vis-à-vis MNEs, whereas Girma and Wakelin (2001) consider the size and skill intensity of domestic firms.

10 This corresponds to the classification developed by the French National Institute of Statistics and Economic Studies (INSEE), which is compatible with the two-digit NACE rev. 2.

11 According to the International Monetary Fund, a firm is defined as a foreign-owned firm if a non-resident investor controls at least 10% of the ordinary shares or voting power.

12 For the sake of brevity, these results are not reported.

Additional information

Funding

This work is supported by a public grant overseen by the Agence Nationale de la Recherche (ANR) as part of the ‘Investissements d’avenir’ programme (reference: ANR-10-EQPX-17 – Centre d’accès sécurisé aux données – CASD).