Selection Biases in Complementary R&D Projects

• Date: 01 December 2014
• Published date: 01 December 2014
• Author: Heiko Gerlach,Jay Pil Choi
• DOI: http://doi.org/10.1111/jems.12071

Selection Biases in Complementary R&D Projects

JAY PIL CHOI

School of Economics

YonseiUniversity

50 Yonsei-Ro,Seoul 120-749, Korea

jay.choi@unsw.edu.au

HEIKO GERLACH

School of Economics

University of Queensland

St. Lucia, QLD 4072, Australia

h.gerlach@uq.edu.au

This paper analyzes selection biases in the project choice of complementary technologies that are

used in combination to produce a final product. In the presence of complementary technologies,

patents allow innovating firms to hold up rivals who succeed in developing other system com-

ponents. This hold-up makes innovation rewards independent of project difficulties and firms

excessively cluster their R&D efforts on a relatively easier technology in order to preemptively

claim stakes on component property rights. This selection bias is persistent and robust to several

model extensions. Implications for the optimal design of intellectual property rights are discussed.

1. Introduction

The patent system was designed with a paradigm invention in mind where a new

product or machine was covered by a single patent. Although, historically, this paradigm

was probably a fairly good description, the nature of innovations has changed drastically

over the last decades. A typical product in today’s high-tech industries incorporates

not a single new invention but a combination of many different components, each

of which is essential and may be the subject of one or more patents. For example, in

biomedical research, gene fragments (or,more generally,DNA sequences) are considered

patentable innovations. However, commercial products such as therapeutic proteins or

genetic diagnostic tests, require the use of multiple gene fragments. Or, for instance,

products such as mobile devices, microprocessors, memory devices, or software code

may encompass hundreds, in the case of 3G mobile phones even thousands, of different

technologies covered by patents.

This paper analyzes the impact of technological complementarity on firms’ in-

centives to allocate their R&D efforts across system components. In the presence of

complementarity, patents allow innovating firms to hold up rivals who succeed in

developing other system components. The hold-up potential makes the value of a patent

independent of how hard it is to achieve the underlying innovation. This introduces

a selection bias in the project choice of complementary innovations. Our analysis has

Weare grateful to the co-editor and two anonymous referees for their valuable comments and suggestions. We

also thank Nancy Gallini, Eric Maskin, and Jean Tirole, and participants in various conferences and seminars

for useful comments and discussions. Sanghyun Kim provided excellent research assistance. All remaining

errors are ours.

C2014 Wiley Periodicals, Inc.

Journal of Economics & Management Strategy, Volume23, Number 4, Winter 2014, 899–924

900 Journal of Economics & Management Strategy

important implications for the optimal design of intellectual property rights. Recognition

of the complementary nature of innovations requires a new way to reward innovators

in order to eliminate wasteful R&D duplications and align private and social incentives.

To analyze the nature of selection biases in complementary R&D projects and

the role of R&D competition as the main driver of these selection biases, consider the

following basic setup. There are two complementary innovations, Aand B,bothof

which are needed to produce a final product and have no stand-alone value. There are

two periods. A firm can engage in only one project in each period. In such a case, if there

is only one firm that can engage in the R&D projects, the sequence of project choice is

irrelevant because the firm needs to make both innovations.1However, project choice

becomes important when there is competition and the first firm that innovates receives

a patent on the component innovation. We allow asymmetry in the research projects

in terms of the difficulty of success and show that there is a tendency for preemptive

duplication in the easier project compared to the socially optimal allocation of research

project choices. The intuition for this discrepancy between private and social incentives

is that firms care about the division of market value of the final product as well as

the overall probability of completing both projects by the second period. The privately

optimal strategy to maximize their share of market value is to stake an early claim on

the patent that is easily achievable because it allows firms to hold up against anyone

who succeeds in the other project.

We demonstrate that the selection bias in complementary project choices is per-

sistent and robust. We first develop the argument in a simple two-firm example and

then extend to allow for more firms, free entry to the R&D race, an infinite horizon and

asymmetric firms. In all variations, we find that the bias toward preemptive duplication

in the easier project persists. We also explore policy implications of the analysis and

discuss mechanisms to mitigate this selection bias, such as prior user rights, licensing,

and patent scope.

Our paper departs from most of the R&D literature in two respects. First, the main

focus of the literature has been on the aggregate amount of R&D and the comparison

of equilibrium R&D spending to the socially optimal level.2However, as pointed out

by Mansfield (1981), “the composition of R&D expenditures may be as important as their

total size (italics added, p. 614).” The R&D management literature also emphasizes the

importance of R&D project selection decisions. For instance, Ofek (2008) states that

“one of the most important dilemmas confronting firms” in the product development

process is “where should these efforts be directed?”. Second, most of the R&D literature

analyzes R&D competition for a single isolated innovation. Even though this may be a

good description of many innovations in the past, it is increasingly at odds in today’s

high-tech industries. In our paper, we consider complementary innovations and focus

on the allocation of R&D resources across projects to reflect this reality.

Early contributions that analyzed the issue of R&D resource allocations across

projects include Dasgupta and Maskin (1987), Bhattacharya and Mookherjee (1986),

and Klette and de Meza (1986). They consider an isolated innovation case and show

that the market is biased toward the selection of more risky projects compared to the

social optimum. The main intuition comes from the “winner-takes-all” nature of patent

1. This holds true as long as there is no learning effect, or other economies of scale and scope, from

successfully completing one project before another.

2. The classical contributions to this question include Loury (1979), Lee and Wilde (1980), Dasgupta and

Stiglitz (1980), and Tandon(1983), among others. See Reinganum (1988) for an excellent survey of the literature

on R&D.