ENTRY AND INTRA‐INDUSTRY MOBILITY IN THE UK CAR MARKET
| Author | P. A. Geroski,A. Murfin |
| Published date | 01 November 1991 |
| Date | 01 November 1991 |
| DOI | http://doi.org/10.1111/j.1468-0084.1991.mp53004001.x |
OXFORD BULLETIN
of
ECONOMICS and STATISTICS
Volume 53 November 1991 No. 4
OXFORD BULLETIN OF ECONOMICS AND STATISTICS, 53,4(1991)
0 305-9049 $3.00
ENTRY AND INIRA-INDUSTRY MOBILITY IN
THE UK CAR MARKET
P. A. Geroski andA. Mu,fin*
1. INTRODUCTION
Although few would doubt that entry depends on expected post-entry profits,
the assertion (when considered as a hypothesis) is untestable. Expected post-
entry profits are a classic latent variable, and the challenge of modelling entry
flows arises from the need to model around this unobservable. Many solu-
tions to this problem have been proposed in the literature (e.g. see Geroski,
1990, Chapter III), but few studies actually focus on the strategic interactions
that occur between entrants and incumbents (notwithstanding the enormous
theoretical literature on the subject). It seems plausible to believe that post-
entry profits are likely to depend on the responses made by incumbents and
the types of competitive weapons that they deploy against newcomers, and in
this paper we propose to focus on the role played by advertising in facilitating
and/or inhibiting entry into the UK car market.
An interesting characteristic of the UK car industry is that price competi-
tion has always been very weak. Therefore, there would have been little
reason, prior to entry, for entrants to believe that post-entry competition
would involve the use of price by incumbents to defend their markets. Nor
*We are obliged to Mike Waterson, Mick Keen, John Kay, Alex Bowen, John Cubbin, two
referees and Steve Nickell, to participants at the 1SF Conference in Boston, May 1987, the
ESRC Econometrics Study Group Meeting, January 1988, and to seminar audiences at the
London Business School and the KUL, Leuven, for comments on an earlier version of this
paper. The usual disclaimer applies.
341
342 BULLETIN
would there have been much reason for entrants to expect that undercutting
incumbents' prices would cause either passive accommodation or outright
exit, not least because (in the 1960's and 1970's) the UK government was
clearly committed to backing the 'national champion', British Leyland, as a
major volume producer. In fact, advertising and, more subtly, quality choice
might have been thought to be more likely to feature in post-entry battles
because (it appears that) 'too little' advertising was being undertaken by
incumbents prior to entry (e.g. Cowling and Cubbin, 1971) and because the
quality of domestically produced UK cars had deteriorated (relative to that of
entrants' cars).' All of this suggests, then, that to model the expected post-
entry profits of entrants, one needs to model non-price competition in this
market. By focusing on advertising and by using a particular model of
consumer choice that builds on several specific features of car purchasing
decisions, we shall develop - and then estimate - a model with an entry
decision rule that depends on the expected post-entry advertising shares in a
firm, market segment and time-specific manner.
A second interesting feature of competition in the UK car market is that
entry occurred first into the medium and large car market segments, and
then, after a lag, into the small car segment of the market. Explaining this
pattern of behaviour is an important part of any story of entry into this
market, and our explanation involves among other things examining the
extent to which previous experience in other segments of the market affected
entry in any particular segment at any given time. This we shall do by using
the estimated model developed below to simulate a counterfactual in which
no entrant contemplating an assault on any particular segment at any
particular time has any experience elsewhere in the market, and then
comparing the simulated pattern of entry to the actual, observed pattern.
The plan of the paper is as follows. Section II develops a simple empirical
model of entry, while Section III describes the results of applying this model
to describe the activity of entrants in three segments of the UK car industry,
1957-83. The model is then simulated to determine what might have
happened had entry into the various segments occurred de novo. Our
conclusions are summarized in Section IV.
II. MEASURING THE RELATIVE COSTS OF DIFFERENT ENTRY PATHS
The problem which makes the modelling of entry into markets less than
straightforward is that expected profits are not observable. However, if one
can observe the pattern and timing of actual entry, if one can observe the
determinants of expectations, and if one adopts the assumption that entry is
driven by expected profits, then it is possible to make some inferences about
'In the event, very little in the way of a price response to entry occurred (price cost margins
actually rose slightly post-entry), and the volume of industry advertising rose by a factor of
about eight; see Geroskj and Murfin (1990), for further details about what actually happened
post-entry.
To continue reading
Request your trial