Optimal R&D outsourcing

• Date: 01 June 2018
• DOI: http://doi.org/10.1111/ijet.12152
• Author: Patrick L. Leoni
• Published date: 01 June 2018

doi: 10.1111/ijet.12152

Optimal R&D outsourcing

Patrick L. Leoni∗

We study optimal contracts when R&D is outsourcedto another company. We find that, when

a realistic assumption holds, the optimal contract always leads in equilibrium to low chances of

having a successful technical innovation, given announced compensations. This assumption is

significantly different from those used in the literature, and it is necessary for the result to hold.

Our work shows that R&D outsourcing typically performs poorly.

Key wor ds R&D outsourcing, incentive mechanism, optimal contract, effort implementation

JEL classification D82, O31, O32

Accepted 26 September 2016

1 Introduction

Outsourcing research and development (R&D) is nowadays increasingly common. For instance,

Schubert (2012) argues that the pharmaceutical industry tends to entirely outsource its R&D.Reasons

are many, although the most obvious and potent reason is that one only rewards successful R&D

campaigns this way. Socially improving effects in terms of innovation availability are described in

Gorg and Hanley (2011), whereas some economic consequences are discussed in Glass and Saggi

(2001).

Westudy the optimal contract between a firm willing to outsource a R&D campaign (the princi-

pal), and another firm carrying out the operative task (the agent). This strategic setting is unique in

that rewards are granted only when the R&D is successful. When efforts are observable, we find that,

when a realistic assumption holds, the optimal contract always leads in equilibrium to the lowest

acceptable chance of a successful medical innovation, given announcedcompensations.

For the result to hold, we need to assume that the marginal odds of a successful R&D smoothly

converge to0, as investments keep on increasing. Exponential convergence to 0 works here, but much

slower speed will ensure that the result holds. This assumption is realistic in many R&D campaigns,

where massive financial investments will hardly overcomesignificant technological challenges, given

current knowledge.

The intuition for this result, with the above assumption, is that too large incentives to increase

the odds of success prove inefficient in termsof expected payoffs. With high necessary investments in

R&D and difficult technological challenges to overcome, it then becomes optimal to implement low

chances. The point is that implementing significantly higher odds becomes prohibitively expensive

from the principal’s viewpoint.

From a technical standpoint, the assumption that the marginal chances of a successful R&D

smoothly converge to 0, as investments keep on increasing, is significantly different from those in

∗Kedge Business School, Marseille, France.Email: patr ick.leoni@kedgebs.com

International Journal of Economic Theory 14 (2018) 201–205 © IAET 201

International Journal of Economic Theory