Review of Industrial Organization 23: 283–299, 2003.
© 2004 Kluwer Academic Publishers. Printed in the Netherlands.
Firm Size and Market Power in Carbonated Soft
, PATRICK PAUL WALSH
and CIARA WHELAN
Department of Economics, University of Venice, Italy;
Department of Economics, Trinity College,
Dublin, Ireland. Email: firstname.lastname@example.org;
Department of Economics, University College Dublin,
Abstract. Sutton (1998) offers us a simple way to model ﬁrm size distributions across differentiated
products industries. We analyse the implications of this approach for company markups using a
structural model for a speciﬁc industry. We incorporate the complexities of multi-product (brand)
companies operating with different (strategic) conﬁgurations of product characteristics and stores
to estimate brand markups, using Irish AC Nielsen retail data for Carbonated Soft Drinks. As a
second step we estimate that market power does not increase in companies with higher market share,
controlling for other factors. This challenges a traditional mind-set.
Key words: Company size and market power, differentiated products.
JEL Classiﬁcation: L11, L25, L66 and L81
A belief in a mapping of ﬁrm market share into market power has a long history and
is still at the centre of most merger investigations. One could argue that a positive
relationship is a good rule of thumb in homogenous goods industries.
for multi-product production with goods differentiated by product characteristics
and store coverage, we show that using such a rule of thumb is ill-advised in retail
Carbonated Soft Drinks.
Section II applies Sutton (1998) to show that differences in market shares in
retail Carbonated Soft Drinks result mainly from ﬁrms having a different number
of product and location segments covered by brands, and not from market share
heterogeneity within segments. Market shares within product and location seg-
ments suggest that small companies, even though they cover less of the market,
may have localized power. The aim of this paper is to investigate whether such a
Game theoretic models suggest that market share and power in homogenous goods indus-
tries can be positively related, though not necessarily so. The relationship is augmented by market
conduct. Techniques are available to empirically identify market conduct in homogenous goods
industries (Bresnahan, 1989 and 1982; Genesove and Mullin, 1998).