Horizontal mergers between multisided platforms: Insights from Cournot competition

• Author: Joao Correia‐da‐Silva,Joana Pinho,Yassine Lefouili,Bruno Jullien
• Published date: 01 January 2019
• DOI: http://doi.org/10.1111/jems.12309
• Date: 01 January 2019

Received: 20 August 2018

|

Accepted: 20 August 2018

DOI: 10.1111/jems.12309

Horizontal mergers between multisided platforms:

Insights from Cournot competition

Joao Correia‐da‐Silva

1

|

Bruno Jullien

2

|

Yassine Lefouili

3

|

Joana Pinho

4

1

CEF.UP, Faculdade de Economia,

Universidade do Porto, Porto, Portugal

2

Toulouse School of Economics (TSE),

CNRS, University of Toulouse Capitole,

Toulouse, France

3

Toulouse School of Economics (TSE),

University of Toulouse Capitole,

Toulouse, France

4

Catolica Porto Business School, CEF.UP,

Universidade do Porto, Porto, Portugal

Correspondence

Bruno Jullien, Toulouse School of

Economics (TSE), CNRS, University of

Toulouse Capitole.

Email: bruno.jullien@tse-fr.eu

Funding information

European Regional Development Fund;

FCT, Grant/Award Numbers: POCI‐01‐

0145‐FEDER‐006890, PEst‐OE/EGE/

UI4105/2014; H2020 European Research

Council, Grant/Award Number: 670494;

European Commission, Marie

Sklodowska Curie Fellowship,

Grant/Award Number: H2020‐MSCA‐IF‐

2014‐657283

Abstract

This paper discusses the literature on horizontal mergers between multisided

platforms and argues that the Cournot model can provide useful insights into

the welfare effects of such mergers. To illustrate those insights, we develop a

simple model in which two‐sided platforms offer a homogeneous service and

compete à la Cournot, and derive the effects of “average‐marginal‐cost‐

preserving”mergers on consumers on both sides of the market. We conclude

with a discussion of several research avenues that could be explored to

understand better the impact of horizontal mergers between multisided

platforms.

KEYWORDS

Cournot competition, mergers, multisided platforms

JEL CLASSIFICATION

D43, L41

1

|

INTRODUCTION

It is widely recognized that traditional merger analysis cannot be directly applied to multisided platform industries.

1

There

are at least two reasons for that. First, a horizontal merger between multisided platforms not only affects the level of prices

but also their structure. In particular, such a merger may lead to a lower price on one or more sides of the market even in the

absence of cost‐related efficiency gains.

2

Second, a merger between two or more platforms may generate network‐related

efficiencies that are absent in traditional markets. For instance, if the merger leads to the creation of a single platform or if

the merging platforms are maintained but made interoperable, agents on each side of the market may end up interacting

with more agents on the other side, which would affect positively their surplus if the relevant network effects are positive.

Several merger cases have involved multisided platforms.

3

Interestingly, the way competition authorities have

addressed the competitive effects of mergers in these markets has been varying over time and across jurisdictions. While

the multisided nature of the relevant market was overlooked in some cases, it was crucial to the decision in other cases.

For instance, in 2004, when analyzing the merger between the two weekly local newspapers Archant and

Independent News and Media,

4

the UK Competition Commission focused on the impact of the merger on advertisers

and ignored its effect on readers. In other media merger cases involving TV channels, such as BSkyb’s acquisition of

J Econ Manage Strat. 2019;28:109–124. wileyonlinelibrary.com/journal/jems © 2019 Wiley Periodicals, Inc.

|

109

ORIGINAL ARTICLE

24% of KirchpayTV

5

in 2000 and News Corporation’s acquisition of 25% of Premiere

6

in 2008, the effect on the

advertising side and the existence of network externalities seem to have been neglected to a large extent.

7

In other cases, however, competition authorities have analyzed the effect on all the sides of the market when

assessing mergers between platforms. An example is the decision by the US Department of Justice and the Federal

Communications Commission to clear, in 2008, the merger between Sirius and XM, the only two US satellite digital

radio services. As discussed by Belleflamme and Peitz (2015), the merger would probably have been blocked if the two‐

sided nature of the market had not been recognized since it would have been a “2‐to‐1”merger. However, by taking into

account the advertisers’side, authorities widened the product market (by including other kinds of broadcast) and

cleared the merger. The two‐sidedness of the market was also critical in the decision of the European Commission to

clear, in 2007, the merger between Travelport and Worldspan.

8

The Commission claimed that this merger would not be

detrimental to agents on one side of the market (travel agents) due to the need of the platforms to build a sufficiently

large network to attract agents on the other side (travel service providers).

9

Note that in the latter two cases, competition authorities did not (at least explicitly) account for the potential

efficiencies stemming from an increase in indirect network effects after the merger. These efficiencies played, however,

a key role in The Netherlands Competition Authority’s assessment of the merger between the (only) two Dutch yellow

pages directories, European Directories and Truvo.

10

In particular, the merger was approved on the basis that

advertisers and users would benefit from using a larger platform. This makes this case a “milestone”in the assessment

of mergers between multisided platforms (Camesasca, Meulenbelt, Chellingsworth, Daems, & Vandenbussche, 2009).

Despite the existence of many merger cases involving competing multisided platforms, the literature on such

mergers remains scarce and competition authorities still lack clear guidance as to how they should be assessed. In this

paper, we discuss this literature and argue that the Cournot model can help address some of the challenges it faces.

To illustrate the insights that can be drawn from the Cournot model regarding mergers between multisided

platforms, we develop a simple model in which

K

(2

)

≥

twosided platforms offer a homogeneous service and compete à

la Cournot. We derive the effects of average‐marginal‐cost‐preserving (AMCP) mergers, that is, mergers that do not affect

the industry‐wide average marginal cost on any side of the market.

11

Considering first general demand functions, we

show that the comparison between the premerger “externality‐adjusted price”on each side of the market and the

average marginal cost on that side plays a key role in determining the effect of a merger on consumers.

12

When both

externality‐adjusted prices are above (below) the corresponding average marginal costs, a merger harms (benefits)

consumers on both sides. We then restrict attention to linear demand functions and establish that a merger harms

consumers on both sides if (total) network effects are small; benefits consumers on one side and harms consumers on

the other side, if network effects are intermediate; and benefits consumers on both sides if network effects are large.

The remainder of the paper is organized as follows. In Section 2, we summarize the main results of the existing literature

on horizontal mergers between multisided platforms. In Section 3, we discuss the Cournot model in multisided markets and

explain why it can provide novel insights into the impact of mergers between multisided platforms. In Section 4, we lay out

our model and derive predictions regarding the effect of a merger between twosided platforms on consumers (on both sides

of the market). In Section 5, we conclude with a discussion of avenues for future research.

2

|

A PROGRESS REPORT ON THE EXISTING LITERATURE

In a quite specific but provocative paper, Chandra and Collard‐Wexler (2009) show that in a twosided market, even

mergers to a monopoly may not lead to higher prices on either side of the market. Their model features two newspapers

and two advertisers of differentiated products and assumes that consumer preferences for newspapers and products are

correlated. Marginal readers are less valuable to advertisers than captive consumers and some of them may be

unprofitable to the newspaper because the advertising revenue they generate is insufficient to cover the subsidy they

receive when buying the newspaper. However, a newspaper cannot exclude these readers because it is unable to engage

in price discrimination. In a duopoly equilibrium, a small decrease in the price of a newspaper would attract the least

valuable consumers away from the rival newspaper and, therefore, may increase the rival’s profit. A merger makes

platforms internalize this externality, which explains why it may lead to a decrease in newspaper prices.

In a similar vein, Leonello (2010) analyzes a merger to monopoly between two platforms located at the extreme

points of a Hotelling line on both sides of the market. She finds that merging firms may have incentives to lower their

prices on at least one side of the market if indirect network externalities are strong enough. Baranes, Cortade, and

Cosnita‐Langlais (2016) extend her analysis by considering four platforms equidistantly located on a Salop circle.

110

|

CORREIA‐DA‐SILVA ET AL.