Allocation of the Public R&D Budget: The Impact of International Competitive Advantages and R&D Alliances

Abstract

I consider a two-country model, in which two asymmetric firms invest in R&D to increase their competitiveness and compete over the supply of a homogeneous product, and the government grants R&D subsidies to increase welfare. In this setting I show that optimal R&D policy is affected by industrywide international competitive advantages. Similar to conventional wisdom on strategic trade policy, competitive advantages have a positive impact on the optimal amount of R&D subsidy in the case of R&D competition. With international R&D cooperation, this conclusion is reversed: subsidising the more competitive firm may have, if any, very little impact on domestic welfare.

Keywords:

• International R&D cooperation
• R&D policy
• international competitive advantage
• R&D spillover

Acknowledgments

I am grateful to Catia Montagna, Simone D'Alessandro, Domenico Buccella and two anonymous referees for helpful comments. This paper was produced as part of my PhD thesis at the University of Siena, which I acknowledge for financial support. Opinions, conclusions and mistakes are strictly personal. The usual disclaimer applies.

Disclosure Statement

No potential conflict of interest was reported by the author.

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Notes

1 http://ec.europa.eu/research/participants/data/ref/h2020/wp/2018-2020/main/h2020-wp1820-intro_en.pdf

2 Works on international R&D cooperation belong to the wider strand of literature concerning competition policy. The list of notable contributions includes Katz (1986) and Kamien et al. (1992), describing different types of research cooperative agreements; Spence (1984), Jaffe (1989) and Feldman (1994) discuss technological spillovers and their effect on R&D investments. d'Aspremont and Jacquemin (1988) formalise the effect of R&D coordination on technological spillover internalisation; Suzumura (1992) and Ziss (1994) extend their work by considering many firms, general demand function and price competition. Leahy and Neary (1997, 1999) describe optimal output and research subsidy policy in a Cournot competition framework. Zhou et al. (2002) focus on strategic trade policy with endogenous choice of quality of goods in the case of two heterogeneous countries. The authors prove that both a tax or a subsidy to the investment on quality may be motivated depending on the competition model (Bertrand vs Cournot) and degree of coordination of the policy (one-country policy vs coordinated policy). Brander and Spencer (1985), Lahiri Ono (1988), Eaton and Grossman (1986), De Meza (1986), Neary (1994) and Collie (1993) discuss export subsidies. Haruna and Goel (2017), Lee et al. (2017), Kesavayuth and Zikos (2013), Lee and Tomaru (2017), Gil-Molto et al. (2011) and Lee and Muminov (2020) analyse the dynamics of R&D in the case of mixed oligopolies, in which both private and public firms compete in the same market.

3 The parameter setting b>1 is required to ensure positive values for the optimal amount of investment and output decided by the firms.

4 In terms of competitiveness, country- and firm-heterogeneity are equivalent in this set-up because I assume one firm in each country.

5 The quadratic function, largely adopted in the literature (Leahy & Neary, 1997), describes falling marginal returns, which means that the amount of investment required to obtain a given level of cost reduction is progressively larger. An economic justification is that a large fraction of innovations derive from small firms even if the capital they can invest is relatively low (Dasgupta, 1986).

6 As observed by Spence (1984) and formalised by d'Aspremont and Jacquemin (1988), the presence of technological spillovers affects the private incentive of firms towards R&D investment.

7 The third-country assumption (Brander & Spencer, 1985) is largely adopted in international trade literature, because it has the upside to have greater tractability and allows to focus on the impact of trade policy on international competition. However, it ignores the effect of trade policy on internal markets, the discussion of which requires different approaches.

8 In this set-up, the R&D subsidy is considered per unit of innovation rather than per dollar of investment. Such simplification is conceptually harmless because K and x are in the one-to-one relationship described by Equation (1(1) $x=\sqrt{K}$(1) ) and in fact is largely adopted in the literature; see Leahy and Neary (1997, 1999).

9 A similar guidance is provided by Katz (1986).

10 Various types of cartel may arise according to the sharing rules of innovation. The members may agree, for example, to fully share the innovation produced by the cooperation. In this case, the spillover rate within the cartel becomes equal to unity, since the knowledge produced by one member is shared with the other. Alternatively, the participants may decide to coordinate their research activity, for instance to agree on the direction of the R&D effort or to conduct basic research, without any innovation sharing. This type of agreement does not have an effect on the spillover rate β. A discussion around the different sorts of cooperative agreement can be found in Katz (1986). In the present paper, I focus on the latter case because it appears to be more general, since it embeds the fact that innovation-sharing is not an essential characteristic to achieve spillover internalisation: the critical feature is represented by the coordination over the amount of R&D investment.

11 There are some reasons to believe that the government is not always willing to trade off one dollar of extra profits with one dollar of subsidy payments. For instance, the resources for the subsidy may be obtained by raising a distortionary tax on the population, with negative effects on efficiency; or the firm may be partially owned by foreign investors, with a shift in profits from the national account towards foreign countries. For a thorough discussion, see Lahiri et al. (2000) and Leahy and Montagna (2001). In this analysis, the introduction of a subsidy social cost would have the unambiguous effect of reducing the optimal amount of R&D subsidy granted by the government.

12 To increase robustness, the simulations are checked on a wide range of the parameters. See the appendix.

13 This result is already known in the literature, see d'Aspremont and Jacquemin, 1988.

Additional information

Notes on contributors

David Silei

David Silei Worked at the European University Institute in Florence, Italy, as a Research Fellow. Holds a Ph.D. in Economics from the University of Siena, Italy. Research interests mainly focus on the Economics of Innovation and Public Economics.