Vertical integration and disruptive cross‐market R&D

• Author: Tianle Zhang,Ping Lin,Wen Zhou
• Date: 01 January 2020
• DOI: http://doi.org/10.1111/jems.12328
• Published date: 01 January 2020

J Econ Manage Strat. 2020;29:51–73. wileyonlinelibrary.com/journal/jems © 2019 Wiley Periodicals, Inc.

|

51

Received: 28 May 2017

|

Revised: 2 July 2019

|

Accepted: 4 July 2019

DOI: 10.1111/jems.12328

ORIGINAL ARTICLE

Vertical integration and disruptive cross‐market R&D

Ping Lin

1

|

Tianle Zhang

1

|

Wen Zhou

2

1

Department of Economics, Lingnan

University, Tuen Mun, New Territories,

Hong Kong

2

Faculty of Business and Economics, The

University of Hong Kong, Pok Fu Lam,

Hong Kong

Correspondence

Department of Economics, Lingnan

University, Tuen Mun, New Territories,

Hong Kong.

Email: plin@ln.edu.hk

Abstract

We study how vertical market structure affects the incentives of suppliers and

customers to develop a new input that will enable the innovator to replace the

incumbent supplier. In a vertical setting with an incumbent monopoly

upstream supplier and two downstream firms, we show that vertical integration

reduces the R&D incentives of the integrated parties, but increases that of the

nonintegrated downstream rival. Strategic vertical integration may occur

whereby the upstream incumbent integrates with a downstream firm to

discourage or even preempt downstream disruptive R&D. Depending on the

R&D costs, vertical integration may lower the social rate of innovation.

KEYWORDS

innovation, replacement effect, structural change, vertical integration

JEL CLASSIFICATION

L13; L42; O31

1

|

INTRODUCTION

There are many real‐life situations in which downstream producers in vertically related industries enter backward into

the upstream market as a result of internal R&D or through the acquisition of independent innovating firms. For

example, Apple Inc. once considered acquiring Imagination, a major supplier of the graphics processors used in

iPhones, but eventually decided to develop the processors in‐house to reduce its reliance on Imagination’s technology.

1

On the software side, Apple recently launched the mobile payment system Apple Pay, which is viewed by many

analysts as posing a direct competitive threat to the incumbent PayPal, the dominant leader in online payment services.

2

Similarly, in 2012, Dell created its own software division, Dell Software Group, after conducting a series of acquisitions

in the software and service sectors.

3

Motivated by these observations, we aim to address the following questions. How does market structure affect firms’

incentives for developing innovations that may disrupt a vertically related industry? What are the effects of such cross‐

market R&D on incumbent suppliers, downstream producers, and overall level of innovation? What competitive

strategies might the affected firms use to fend off such threats? To answer these questions, we develop a model of

innovation that consists of an incumbent monopoly supplier and two downstream producers. Initially, the downstream

firms produce differentiated products using an input supplied by the upstream monopolist. The input is sold via a two‐

part tariff contract.

4

Successful R&D results in a new input with higher quality, or equivalently at a lower cost, than the

existing input, and enables an innovating downstream firm to enter backward into the upstream market. Central to our

model are the structural changes associated with the post‐R&D entry into the upstream sector and the incumbent

supplier’s exit if the innovator is a downstream firm.

We first consider the case in which the market structure is exogenously given as one of two types: vertical separation,

where all three firms are independent entities, and vertical integration, where the upstream firm merges with one of the

two downstream firms. Our analysis provides the following insights.

First, we show that vertical integration reduces the R&D incentives of the integrated firms but raises that of the

nonintegrated downstream rival firm. The driving force for this result is the replacement effect of innovation

identified by Arrow (1962) and corroborated by extensive studies on horizontal innovations, which show that a

firm’s R&D incentive is inversely related to its preinnovation profit (see, e.g., Gilbert, 2006a, 2006b; Reinganum,

1989; Tirole, Tirole, 1988). If a downstream firm integrates with the incumbent upstream supplier, the downstream

firm’s innovation incentive will be dampened, because the new input will replace its upstream unit’scurrent

business. As the integrated firms reduce their R&D investments, the nonintegrated downstream firm steps up its

R&D effort in response.

Second, because of these incentives, there are cases in which it pays the incumbent supplier to strategically

acquire a downstream firm to preempt its disruptive innovation. We call this type of vertical integration strategic

vertical integration. This may occur, for example, if the R&D project is inaccessible or very costly to the other

downstream firm.

It is worth noting at this point that in contrast to models with linear pricing on input, vertical integration in our

setting is never profitable when the R&D incentives are not considered. In particular, with two‐part tariff contracts

under vertical separation, the input supplier is able to extract the entire monopoly profit of the industry, whereas

vertical integration creates some cost distortion in the downstream market and hence lowers the total profits of the

integrated entity.

5

In other words, without R&D, vertical integration will never emerge in our model. However, in

the presence of input innovation by the downstream firms, which may disrupt the existing upstream input business, the

incumbent supplier may choose to merge with an innovating downstream firm to defend and maintain its upstream

market position. We also show that there are circumstances under which it pays the incumbent upstream firm not to

merge with its target firm to avoid promoting greater R&D investment from the other downstream firm.

6

Third, we analyze the effects of vertical integration on the overall level of innovation and show that vertical

integration reduces the social rate of innovation (the probability that at least one firm succeeds in R&D) if R&D is

sufficiently costly. As argued above, vertical integration reduces the R&D incentives of the integrated firms but

increases the R&D incentive of the nonintegrated firm. When an R&D project is sufficiently costly, the increase in R&D

incentive by the nonintegrated firm is insufficient to offset the decrease in the R&D incentive for the integrated firms,

resulting in a lower social rate of innovation.

Our analysis reveals that vertical mergers, toward which antitrust agencies are usually more lenient than horizontal

mergers, can preempt R&D and damage innovation. Antitrust authorities should therefore be aware of the possible

preemptive motive of vertical integration and assess its possible negative effects on innovation and welfare.

7

Unlike in a standard horizontal setting (e.g., Gilbert & Newbery, 1982) in which a firm’s innovation leaves the

market structure intact, market disruption through innovation may occur in a vertical setting because firms are

naturally asymmetric (i.e., some operate in the downstream and some in the upstream) and an innovator has an

incentive to enter the vertically related industry. The finding that market power in a vertical setting (in the form of

vertical integration) reduces the integrated firms’R&D incentives resonates with Arrow’s argument, and part of the

driving force is indeed the replacement effect. Nevertheless, the other part of the driving force concerns the reduced

ability of the incumbent upstream firm to fully benefit from its innovation because vertical integration forces its

downstream unit to compete too aggressively. This is a new force and is specific to vertical settings. Another feature of

our model is that we allow multiple firms to conduct R&D simultaneously. This is not only more realistic, but also

provides new insights. In particular, vertical integration discourages R&D in some firms but encourages it in others, and

the latter effect can dominate so that market power in a vertical setting may increase the overall amount of innovation.

This is consistent with Schumpeter’s view but again works through a new channel.

Most of the studies on innovation in a vertical setting assume a single innovator that never changes the market

structure. Y. Chen and Sappington (2010) find that vertical integration encourages upstream process innovation if the

downstream competition is Cournot, as any upstream cost reduction is passed fully onto the downstream unit in

vertical integration, which generates a competitive advantage in Cournot competition. Loertscher and Riordan (2019)

study the R&D incentives of upstream suppliers that compete to sell to a monopoly buyer. They find that vertical

integration encourages the integrated supplier’s R&D but discourages the nonintegrated suppliers’R&D because the

buyer favors its in‐house supplier in procurement. Allain, Chambolle, and Rey (2016) demonstrate that vertical

integration creates a hold‐up problem for an independent downstream firm and therefore discourages its investment.

By contrast, vertical separation preserves upstream competition, thus preventing the upstream firms from appropriating

downstream firms’investment benefits.

8

In all of these papers, in contrast to our setting, innovation only affects the

innovator’s competitive profit without changing the market structure.

52

|

LIN ET AL.