Review of Industrial Organization 16: 399–415, 2000.
© 2000 Kluwer Academic Publishers. Printed in the Netherlands.
Double Marginalization and Privatization in Liquor
DOUGLAS S. WEST
Department of Economics, University of Alberta, Edmonton, Alberta T6G 2H4, Canada
Abstract. The purpose of this paper is to examine the price and store count predictions of the
spatial and non-spatial models of vertical integration between an upstream monopolist and a down-
stream monopolistically competitive retail industry using data from Alberta’s recently privatized
liquor retailing industry. This industry, which had been a government owned monopoly, became
monopolistically competitive under privatization. The models predict that vertical disintegration will
lead to higher retail prices and an increase in the store count in markets that can support multiple
stores. Both predictions are supported by liquor store count and price data.
Key words: Double marginalization, liquor stores, privatization, vertical integration.
It is well known that when there are successive monopolies at two stages of produc-
tion, like manufacturing and retailing, a double marginalization problem can arise
(see Spengler, 1950; Tirole, 1988, pp. 174–176), leading to a higher price, lower
output, and lower proﬁts than under vertical integration. As compared to successive
monopolies, vertical integration eliminates one of the monopoly markups, leading
to a lower retail price and an increase in welfare.
A double marginalization problem also can arise when the downstream industry
is monopolistically competitive rather than a monopoly. The welfare consequences
of eliminating double marginalization via vertical integration will depend on the
change in price and the number of retailers, and can vary depending on whether
an address approach (e.g., Mathewson and Winter, 1983) or non-address approach
For helpful comments, the author would like to thank Greg Dow, Tom Ross, Nicolas Schmitt,
John Huston, Robin Lindsey, Mel McMillan, seminar participants at the Faculty of Commerce and
Business Administration at the University of British Columbia, the Roundtable sponsored by the
Centre for the Study of State and Market, the Canadian Law and Economics Association meetings,
the Privatization Roundtable held at the University of Toronto’s Faculty of Law, and participants at
the Liquor Retailing Conferences in Vancouver and Toronto organized by the Fraser Institute. I am
also grateful to William Shepherd and two anonymous referees for their comments. I would also like
to thank John Szumlas and the Alberta Liquor Store Association for providing me with data, and the
Centre for the Study of State and Market, Faculty of Law, University of Toronto, for ﬁnancial support
of this research.