Prices and heterogeneous search costs

• DOI: http://doi.org/10.1111/1756-2171.12170
• Author: Zsolt Sándor,José Luis Moraga‐González,Matthijs R. Wildenbeest
• Date: 01 March 2017
• Published date: 01 March 2017

RAND Journal of Economics

Vol.48, No. 1, Spring 2017

pp. 125–146

Prices and heterogeneous search costs

Jos´

e Luis Moraga-Gonz´

alez∗

Zsolt S´

andor∗∗

and

Matthijs R. Wildenbeest∗∗∗

We study price formation in a model of consumer search for differentiated products in which

consumers have heterogeneous search costs. We provide conditions under which a pure-strategy

symmetric Nash equilibrium exists and is unique. Search costs affect two margins—the intensive

search margin (or search intensity) and the extensive search margin (or the decision to search

rather than to not search at all). These two margins affect the elasticity of demand in opposite

directions and whether lower search costs result in higher or lower prices depends on the

properties of the search cost density.

1. Introduction

Throughout history, technological improvements have made it possible for consumers to

participate in markets that were previously beyond their reach. A recent example of such a

technological development is the Internet. The introduction and sustained growth of e-commerce

has significantly lowered the transaction costs consumers experience while shopping. This has

not only made it easier for consumers to search for and compare products but it has also made

∗Vrije Universiteit Amsterdam and University of Groningen; j.l.moragagonzalez@vu.nl.

∗∗Sapientia University Miercurea Ciuc; zsosan@gmail.com.

∗∗∗Indiana University; mwildenb@indiana.edu.

Earlier versions of this article were circulated under the title “Do higher search costs make markets less competitive?”We

wouldlike to thank the Editor, Mark Armstrong, and two anonymous referees for their helpful comments. Simon Anderson,

Guillermo Caruana, Alex de Corni`

ere, Andrea Galeotti, Knut-Eric Joslin, Silvana Krasteva, Theo Dijkstra, Dmitry

Lubensky, Andras Niedermayer, Marielle Non, Vaiva Petrikait˙

e, R´

egis Renault, Mariano Tappata, Makoto Watanabe,

and Jidong Zhou are also thanked for their useful remarks. Special thanks go to Paulo K. Monteiro for his comments,

help, and useful references. The article has also benefited from presentations at Antwerpen, Bologna, Carlos III, CEMFI,

Copenhagen, CPB, East Anglia, EUI, Mannheim, Oxford, PSE, Sauder Business School, the Fifth Workshop on Search

and Switching Costs (Bloomington, IN), the IIOC 2014 (Chicago), the 2014 MaCCI Workshop on Consumer Search in

Bad Homburg, the SaM 2015 Workshop at Bristol, the SED 2015 (Warsaw), the EARIE 2015 (Munich), and the 4th

WIPE (University Rovira i Virgili).Financial support from the Marie Curie Excellence Grant MEXT-CT-2006-042471,

grant ECO2011-29533 of the Spanish Ministry of Economy and Competitiveness, and grant PN-II-ID-PCE-2012-4-0066

of the Romanian Ministry of National Education, CNCS-UEFISCDI, is gratefully acknowledged. Moraga-Gonz´

alez is

also affiliated with the CEPR, the Tinbergen Institute, CESifo, and the Public-Private Sector Research Center (IESE,

Barcelona).

C2017, The RAND Corporation. 125

126 / THE RAND JOURNAL OF ECONOMICS

it easier for consumers to get access to new markets and hard-to-find products. This has also led

to the Internet’s Long Tail phenomenon: the Internet allows consumers to access a much larger

product selection, including niche products that were previously difficultto find, thereby creating

a Long Tail in the sales distribution (see, e.g., Brynjolfsson, Hu, and Simester, 2011).1

How are technologicaladvances that make it easier for consumers to search expected to affect

market competitiveness? The conventionalanswer is that a reduction in search costs increases the

elasticity of demand, and as such reduces prices. However, as illustrated by the Internet example

above, this is not the complete story. Lower search costs also allow consumers to search for

products that previously were not part of their consideration sets. As the new consumers that

enter the market are likely to have higher search costs (because otherwise, they would have been

actively searching before), a reduction in search costs leads to compositional changes in the active

consumer population that may decrease the elasticity of demand. More consumer participation as

a result of a technology-driven reduction in search costs may end up being the dominating factor,

which means firms will raise prices, even when search costs go down for all consumers.

Hortac¸s uand Syverson (2004) present empirical evidence that is consistent with this partic-

ular mechanism in their study of the US mutual fund industry during the late 1990s. Not that long

ago, investors had to go through significant effort to search for adequate investmentoppor tunities

while managing their financial investments. However, the rise of Internet banking and online

brokerages during the late 1990s decreased transaction costs and made it easier for individuals

to search for new investment opportunities. Hortac¸su and Syverson (2004) find that during this

period, investment fund fees went up and that search costs decreased at the lower percentiles of

the search cost distribution but increased at the upper percentiles. According to the authors, this

surprising finding (i.e., a second-order stochastic dominance decrease in search costs leading to a

price rise) is explained by the large increase in the number of households that participated in the

mutual fund market for the first time. These new investors arguably had higher search costs and

were less savvy. Hortac¸ su and Syverson argue that the entry of these new investors changed the

composition of the investor population and made demand more inelastic, explaining the increase

in mutual fund fees.

The main objective of our article is to present conditions under which a mechanism like the

one described above occurs. In particular, our goal is to derive general conditions under which a

first- or second-order stochastic dominance decrease in search costs triggers sufficient entry of

consumers with high search costs such that the elasticity of demand decreases and equilibrium

prices go up. In order to do so, we develop a model of consumer search for differentiatedproducts

in which consumers have heterogeneoussearch costs and the decision to participate in the market

is endogenous. Our model, which is presented in Section 2, builds on the seminal model of

consumer search for differentiated products introduced by Wolinsky (1986), and further studied

by Anderson and Renault (1999).2We extend this model by allowing for arbitrary search cost

densities. This extension is crucial for studying the second-order stochastic dominance (SOSD)

decreases in search costs observed in the mutual fund industry by Hortac¸su and Syverson (2004).3

1The expansion of the railroad network in the United States in the 19th century also led to significantly lower

transportation costs and greater market integration. As with the Internet, lower transportation costs were not only

beneficial to consumers because of reduced transaction costs but also because they allowed consumers to get access to

new markets (see Donaldson and Hornbeck, 2016, for a recent study on the historical impact of railroads on the US

economy that focuses on market access).

2Arguably,Wolinsky’s frameworkhas become the workhorse model of consumer search for differentiated products.

Recent work that builds on this article includes Bar-Isaac, Caruana, and Cu˜

nat (2012), who extend the model to the case

of quality-differentiated firms and study design differentiation; Armstrong, Vickers, and Zhou (2009), who study search

and pricing behavior in the presence of a prominent firm; Haan and Moraga-Gonz´

alez (2011), who study the emergence

and the price effects of prominence; Moraga-Gonz´

alez and Petrikait˙

e (2013), who examine the effect of search costs on

mergers; and Zhou (2014), who studies multiproduct search.

3Our results are not specific to the model we present: wehave obtained similar results when firms sell homogeneous

products and charge a price randomly drawnfrom a distribution, while consumers search nonsequentially for good deals.

The details can be found in the web Appendix to this article, which is placed at the Journal’swebsite.

C

The RAND Corporation 2017.