Innovation, competition, and incomplete adoption of a superior technology

ABSTRACT

This article shows that competition exerts a feedback effect on market structure via the process of innovation. First, downstream competition increases the willingness to pay for a more efficient technology (the direct effect). Second, a sufficiently large innovation may offer the licensees a robust strategic advantage that forces non-adopters out of business. In turn, this raises the licensee's willingness to pay to survive in the market (the indirect effect). More specifically, if competition is intense, even a tiny innovation may result in drastic effects in the market. This work also demonstrates that royalties do not always imply the complete adoption of a superior technology because of competition's indirect effect on innovation. If granted the possibility of denying access to innovation to some downstream manufacturers, an inventor or rights holder may prefer to license a substantial patent to a subset of manufacturers at a discounted price, regardless of the contract scheme enforced. Finally, this article suggests that a ban on obsolete technology is not welfare-improving from a policy perspective.

KEYWORDS:

• Innovation
• licensing
• oligopoly
• competition

JEL CLASSIFICATION:

• L13
• L24
• O31

Acknowledgments

I am grateful to the Editor Cristiano Antonelli and the three anonymous referees for their useful comments and critics. Special thanks go to Fabio Manenti, Luciano Greco, José L. Moraga Gonzalez, Evgenia Motchenkova, Jos Jansen, Paola Valbonesi, Carlo Reggiani, and Stefano Comino for their precious help during the development of my research. An earlier version of the article has been awarded the 9th Giorgio Rota Best Paper Award by the ‘Centro di Ricerca e Documentazione Luigi Einaudi di Torino’; I am very grateful to the awarding committee. I am thankful to all the participants at the ASSET Conference 2019, the SIEP 2020 Annual Conference, and the SIE 2020 Annual Meeting.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

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

Notes

1 See Aghion, Akcigit, and Howitt (2015); Aghion et al. (2019) for a detailed survey of the literature on the evidence of inverted-U patterns.

2 Hermosilla and Wu (2018) provide an interesting analysis of how the market structure affects cooperation in the R&D stage. See also Capuano and Grassi (2019, 2020b) for a comparison between incentives to cooperate or compete in R&D.

3 The verticality of the market reflects the fact that often a product innovation (from the innovator's perspective) represents a process innovation as well (from the manufacturer's perspective).

4 If the innovator participates in the market for the final good (i.e. she is one of the downstream manufacturers) the increasing competition would generate a trade-off between the two forces, as mentioned above. Under some conditions, the inverted-U relationship illustrated by Aghion, Akcigit, and Howitt (2015) and Delbono and Lambertini (2020) would emerge.

5 Filippini (2010) shows that a technology transfer through licensing from a more productive to a less efficient firm can increase total profits in the industry. See also Capuano and Grassi (2020a) for an analysis of the relationship between patent protection and incentives to license a new technology.

6 See Thursby, Jensen, and Thursby (2001) for licensing of university technology.

7 See also Sen and Stamatopoulos (2016). In addition, Badia, Tauman, and Tumendemberel (2020) study how competition among innovators may delay innovation diffusion.

8 See also Zou and Chen (2020) for an analysis of the optimal contract type with exclusive and non-exclusive licensing schemes and Ferreira, Ferreira, and Bode (2021) for a comparison of licensing outcomes under either quantity or price competition with horizontally differentiated products.

9 In practice, patent holders have no duties to license their intellectual property (IP) to others. In the United States, the Antitrust Guidelines for the Licensing of Intellectual Property vaguely states that restraints in intellectual property licensing arrangements are evaluated under the rule of reason. In the European Union, Article 102 of the Treaties of the Functioning of the European Union establishes some exceptional circumstances that limit the freedom of a patent holder to refuse to license. See Petit (2017), Lianos and Dreyfuss (2013), and Espinosa (2017).

10 Absent the inefficient technology, licensing of the production technology by an outside innovator exacerbates market concentration's inefficiencies, as it would create a vertical monopoly.

11 In models with linear demand functions, licensing schemes that rely on two-part tariffs with non-negative linear per-unit royalties and fixed fees collapse to either pure per-unit royalty or pure fixed fee licensing. See Banerjee and Poddar (2019).

12 The results don't change if I assume r is independent of x. Indeed, any linear price r<x can be expressed as a proportional price $\text{β}x$, where $\text{β}=r/x$. The assumption of the price proportional to the cost-saving effect allows for a much clearer and simpler exposition.

13 An alternative licensing structure may be to assume that the manufacturers bid in an auction for the innovation licensed by the upstream innovator, as in Sen and Tauman (2007), where firms pay a royalty per unit of output and bid for the upfront fees (auction with royalties). However, I believe that the current pricing strategy represents a more conservative solution, as it mitigates the replacement effect when competition is more intense.

14 See the appendix.

15 See the appendix.

16 Notice that, in this case, the optimal royalty rate is r = x. Here, I do not consider the case in which m<k, as it is always a dominated strategy for the innovator. Moreover, if that is the case, the strategic disadvantage exerted by the m<k adopters on non-adopters would not be sufficiently strong to force the latter out of the market.

17 The ordering follows from Cournot competition properties. The first parenthetical refers to the size of the innovation and the second to the number of active firms in the market. Finally, the third term refers to the number of adopters. See the appendix.

18 Notice that the left-hand and right-hand sides of this condition are equal if the innovation is 1-drastic. In that case, a switch-off of the obsolete technology would not cause losses from a welfare perspective.

19 By replacing the left-hand side of the above condition with the level of production of an efficient m-oligopoly, that is $Q^{\mathrm{m}-\mathrm{O}\mathrm{l}\mathrm{i}\mathrm{g}}=\frac{m(a-c+x)}{m+1}$ and comparing it with the first row of the right-hand side, it is possible to see that the inequality holds only in the case of $x>\frac{a-c}{m}$, that is, if the innovation is k-drastic. However, if the innovation is k-drastic, no such policy is required.

Additional information

Funding

This research was supported by the OTKA Postdoctoral Execellence Program from the National Research, Development and Innovation Office (NKFIH) of Hungary under [grant ID 138543].