Selling information to competitive firms

• Author: Jakub Kastl,Salvatore Piccolo,Marco Pagnozzi
• DOI: http://doi.org/10.1111/1756-2171.12226
• Date: 01 March 2018
• Published date: 01 March 2018

RAND Journal of Economics

Vol.49, No. 1, Spring 2018

pp. 254–282

Selling information to competitive firms

Jakub Kastl∗

Marco Pagnozzi∗∗

and

Salvatore Piccolo∗∗∗

Internal agencyconflicts distort firms’ choices and reduce social welfare. Tolimit these distortions,

principals dealing with privately informed agents often acquire information from specialized

intermediaries, such as auditing and certification companies, that are able to ascertain, and

crediblydisclose, agents’ private information. Westudy how the structures of both the information

provision and the final good markets affect information accuracy. A monopolistic information

provider may supply impreciseinformation to perfectly competitive firms, even if the precision of

this information can be increased at no cost. This is due to a price effect of information: although

more accurate information reduces agency costs and allows firms to increase production, it also

results in a lower price in the final good market, which reduces principals’ willingness to pay for

information.

1. Introduction

Firms’ internal conflicts affect industry performance, even in perfectly competitive markets

(Legros and Newman, 2013). When “neoclassical black-box” firms are replaced by vertical org-

anizations whose members have diverging objectives, information asymmetries between princi-

pals and agents generate distortions that reduce profits and social welfare. As the severity of these

distortions depends on the verifiable information possessed by principals, in order to implement

proper incentives in their organizations, principals often acquire information from specialized

providers, such as auditors or certification companies, that are able to discover, and credibly

∗Princeton University,CEPR, and NBER; jkastl@princeton.edu.

∗∗Universit `

a di Napoli Federico II and CSEF; pagnozzi@unina.it.

∗∗∗Universit `

a di Bergamo and CSEF; salvapiccolo@gmail.com.

For very helpful comments, we would like to thank Aldo Pignataro, Alessandro Lizzeri, Bruno Jullien, Elisabetta

Iossa, Giacomo Calzolari, Jos`

e Penalva, Marco Ottaviani, Martin Peitz, Pierpaolo Battigalli, Piero Tedeschi, Stephen

Morris, Vincenzo Denicol`

o, the Editor and the referees, as well as seminar participants at Bergamo, Bocconi, Cat´

olica

Lisbon, Mannheim, Naples, Oxford, Padova,Venezia, the 2015 IO Workshopin Bergamo, the 2015 EIEF-Unibo-IGIER

Workshopin Industrial Organization, the First Marco Fanno Alumni Workshop, the Barcelona GSE Summer Forum, the

CSEF-IGIER Symposium on Economics and Institutions, the Applied Economics Workshop in Petralia, and the 2015

European Meeting of the Econometric Society. All errors are ours.

254 C2018, The RAND Corporation.

KASTL, PAGNOZZI AND PICCOLO / 255

disclose, agents’ private information.1The quality of this information affects firms’ production

choices and, through the market mechanism, impacts equilibrium prices and, hence, industry

performance. In this article, we analyze the interplay between firms’ incentives to acquire infor-

mation, endogenous information quality, and the structure of the markets in which information is

supplied and used.

Information acquisition and information disclosure are two aspects of the information man-

agement problem that, in recent years, has become central to the mechanism design literature

(see the survey by Bergemann and V¨

alim¨

aki, 2006). The emergence of endogenous information

structures provides both theoretical insights for mechanism design and policy implications for

market design and regulation. Little is known, however, about the role of information providers

in competitive environments, and about the effect of market competition on the quality of the

information that they supply.

We show that, although more accurate information enhances efficiency by reducing agency

costs and improving production, it also affects firms’ profits and welfare indirectly, through its

impact on equilibrium market prices and quantities. Our key insight is that, when buyers of infor-

mation compete in a product market, their production choices generate a (general equilibrium)

price effect that influences a provider’s choice of the aggregate amount of information to supply.

How much, then, are competitive firms willing to pay for information? How much infor-

mation do providers supply? What is the difference between firms’ individual and collective

incentives to acquire information? Do firms in competitive markets acquire too much or too little

information?

To address these questions, we begin by analyzing an environment in which a monopolistic

information provider sells an informative experiment to a large number of perfectly competitive

firms,2each composed by a principal and an exclusive agent who is privately informed about his

cost of production.3Principals take the market clearing price as given, simultaneously choose

whether to acquire information, and offer incentive-compatible mechanisms to agents. The in-

formation provider designs the accuracy of the experiment (which is the same for all firms) that

produces an informative signal (which is specific to each firm) about the agent’s cost. This signal

allows the principal to better screen the agent, thus reducing agency costs and distortions.

Our main result is that, even if information is costless for the provider (and even if the

provider has the same information as the principal ex ante), the optimal experiment is not fully

informative and does not maximize social welfare when aggregate demand is inelastic—that is, in

industries that face low competition from other markets where substitute products are sold—and

agents are likely to have high costs—that is, in industries far from the technological frontier. In

this context, the provider supplies less precise information in order to reduce competition in the

product market. In contrast to the previous literature (e.g., Admati and Pfleiderer, 1986), in our

model, the incentive to reduce information arises because of agency costs only.

A key role in the analysis is played by a firm’s incremental value of acquiring information.

This represents the price that a principal is willing to pay for the experiment, and is equal to

the difference between the profit of a firm that acquires information, and its outside option—that

is, the profit of a firm that does not acquire information, when all other firms do. Increasing

the experiment’s accuracy has two effects on the incremental value of information. First, a more

informative experiment increases principals’ willingness to paybecause, holding the market price

constant, it reduces agents’ information rent and increases production and profits: an incentive

effect of information. Second, because a more informative experiment increases the aggregate

1Alternatively, providers may supply information technology that makes information processing less costly and

facilitates decentralized decision making (Argyres, 1999).

2Considering a monopolistic information provider offers a useful benchmark. A concentrated market structure in

the certification industry often arises due to economies of scale or specialization—see, for example, Lizzeri (1999) and

Bergemann, Bonatti, and Smolin (2015) who also consider a monopolistic market for information.

3The production cost may be interpreted as a measure of the manager’s efficiency or of the extent to which his

preferences are aligned to those of the firm’sowner.

C

The RAND Corporation 2018.