Barriers to Entry and Strategic Competition

P. Gilbert Geroski, A. Jacquemin

• Publisher: Taylor & Francis
• Year: 2013
• eBook ISBN: 9781136456961
• Edition: 1
• Topic: Business
• Subtopic: Business General
• Index: Business

This volume discusses crucial issues in the overlap between industrial organization and strategic management.

Barriers to Entry and Strategic Competition

PAUL GEROSKI

London Business School

RICHARD J. GILBERT

University of California, Berkeley, USA

ALEXIS JACQUEMIN

Université Catholique de Louvain, Belgium

I. INTRODUCTION

Entry is a source of competitive discipline on the industrial performance of firms. The threat of entry by new competitors puts a constraint on the latitude of existing firms to conduct their operations in ways that adversely effect consumers, while actual entry changes the structure of markets in ways that often brings about the same ends. Entry, actual or potential, can upset traditional patterns of market conduct, de-throne dominant firms, introduce new technology and fresh approaches to product design and marketing, and lead to more competitive prices.

As agreeable as these consequences of entry may seem to be in principle, there is considerable controversy over the extent to which they are achieved in practice. Observations of actual entry attempts reveal that direct entry typically has only a small effect on industry structure. Masson and Shaanan’s [1982] sample of 37 US industries over the period 1950–1966 yielded an average market share penetration by entrants of 4.5% over 6.8 years, a gain of less than 1 % per year. Yip [1982], in a sample of 59 entrants in narrowly defined markets, found a median gain of 6% (mean gain = 10%), with entry via acquisition achieving a penetration roughly three times that achieved by direct entrants. Biggadike [1976] investigated 40 entry attempts by 20 large US firms, and observed that less than 40% of these entrants achieved a penetration of at least 10% within two years. Hause and Du Reitz [1984] examined entry in Sweden over a 15 year period and observed that new entrants only managed 1.7% market penetration on average over that period.

Although these figures suggest that entrants manage to make only a fairly modest penetration into most markets, the number of actual entrants that appear in markets is often extremely large. For example, Dunne and Roberts [1986] examined about 400 four-digit US industries in 1967, 1972, 1977 and 1982, and observed 285, 347, 418 and 425 entrants on average per industry respectively. The predominant method of entry was new firms constructing new facilities (e.g. 69% in 1967 and 1982, 74–76% in 1972 and 1977), with diversifying firms numbering between 23–25% (1972, 1977) and 30% (1967 and 1982). In the UK for 1983, Geroski [1988c], observed an average of 197 entrants per three digit industry in 1983–4 (the maximum was 2556), roughly 12% of the stock of existing firms. However, the average gross penetration per entrant per industry was less than 0.1% and the average penetration for all entrants per industry was 5% (the maximum was 26.6%). Further, net import penetration averaged as much as four times the net domestic entry penetration rate.

Even more interesting, it appears from the data that the relationship between entry and exit is surprisingly close. Geroski (1988c) found that, on average, 130 firms exited per industry (the maximum was 1804), leaving a net increase in the number of firms of 67 (the maximum was 1030 and the minimum was − 34), and a net market share penetration of only 2%. There is also evidence to suggest that many of the entrants in year t entered in t − 1, t − 2 and t − 3, and, indeed, that a distressingly high percentage of entrants fail within a year of formation. Net and gross entry penetration and the net and gross number of entrants were each, however, highly positively correlated; the gross number of entrants and gross penetration were mildly negatively correlated. Fairly similar orders of magnitude were observed by Geroski [1988a] for the UK in the 1970s. Baldwin and Gorecki [1983] examined cumulative entry over the period 1970–1979 in 141 Canadian four-digit industries. All entrants post-1970 collectively acquired a market share by 1979 of about 26% on average; entry by birth accounted for 14% and entry by acquisition 12%. However high this market penetration appears to be, it must be noted that it was accompanied by a large turnover of firms; about 33% of the total 1979 population of firms were not present in 1970, and about 43% of the population of 1970 had, by 1979, disappeared (mainly by scrapping and not divestiture).

Thus, entry appears to be easy but post-entry market penetration and, indeed, survival is not. A question that this raises is whether the somewhat unimpressive market penetration by entrants reflects the existence of generally high barriers to entry in most markets or the differential efficiency advantages of established firms. Barriers to entry are often thought to reflect permanent disadvantages that entrants face, but may also be thought of as a kind of adjustment cost that entrants must overcome. Such adjustment costs can, of course, be affected by the strategic behaviour of incumbents, and this type of consideration broadens the range of potential factors which might account for what we observe. That is, it is of interest to ask whether the low post-entry market penetration that we have observed is caused by high levels of barriers or the strategic behaviour of incumbents, and why so many agents attempt entry when failure rates are high and post entry growth prospects are poor.

Although one may be interested in entry per se as an interesting phenomena to explain, the broader concern in looking at entry is, of course, with assessing market performance. If the extent of rivalry amongst active firms in a concentrated industry is too weak to generate a competitive outcome, it is important that new competitors exert a noticeable pressure on prices and productivity in order to have a positive welfare effect. In fact, many scholars believe that the mere anticipation of entry will induce incumbents to lower their prices toward more competitive levels, and thus that entry need not necessarily occur to have an effect on market performance. A benchmark in this respect has been established by Baumol et al [1982] who have applied the label ‘perfectly contestable’ to markets in which incumbent firms and potential entrants share the same technology and potential competitors can enter and exit without capital loss during the time taken by incumbent firms to change prices.

Formally, Baumol et al [1982] define a ‘perfectly contestable market’ as a market in which a necessary condition for an equilibrium outcome is that no firm can enter taking prices as given and earn strictly positive profits using the same technology as existing firms. In a perfectly competitive market, entrants can and will enter to take advantage of even transient profit opportunities at current prices. This behaviour is most reasonable when the costs of entry are completely reversible so that there are no capital losses in the event of exit. If these conditions are satisfied, a perfectly contestable market mirrors a competitive environment in which entry and exit are frictionless and barriers to entry and exit are non-existent. The assumption of identical costs insures that whenever incumbents can make profits, so can entrants, and, therefore, if a perfectly competitive market equilibrium exists, firms cannot sustain prices in excess of average cost. If more than one firm operates a positive levels of output in a perfectly contestable market equilibrium, the possibility of incremental changes in the output of a rival firm ensures that, in equilibrium, price cannot deviate from the competitive ideal of marginal cost. Thus, the outcome of competition in a market with unconstrained entry is perfectly competitive whenever the market is not a natural monopoly. If it is a natural monopoly, potential competition ensures that the behaviour of the monopolist is ‘regulated’ in the sense that total revenues are not more than total costs.

The theory of contestable markets is a useful benchmark to use in analyzing the effect of entry on market performance, but both the theory and its empirical relevance are open to question. A key assumption that is implicit in the theory of perfectly competitive markets is that capital can move with little risk of loss into and out of an industry over a period of time that is short compared to the time required for existing firms to respond with competitive price changes. If this assumption is satisfied, a firm can base its entry decision on current prices, and ‘hit and run’ can be a rational entry strategy that will police the pricing decisions of incumbent firms. However, it is unlikely that entry into any industry is perfectly reversible; so any entry attempt will probably entail some risk of capital loss if the firm should subsequently exit the industry. Farrell [1986], Gilbert [1986], and Stiglitz [1987] have argued that even if the sunk costs of entry are vanishingly small, the possibility of prompt, aggressive pricing by established firms can make entry unattractive and, by deterring ‘hit and run’ entry, permit existing firms to price at noncompletitive levels.

Empirical studies of entry also cast doubt on the contestability hypothesis. Many scholars (e.g. Bailey and Baumol [1984]) initially argued that transportation services, and, more specifically, the airline industry, were typical cases of contestable markets where fixed costs, although important, were not sunk because aircraft could be easily diverted to other uses. However, as recognized by Baumol and Willig [1986], several econometric studies have shown that the threat of entry in airline markets does not suffice to keep profits at normal levels (see, e.g., Graham, Kaplan, and Sibley [1983]; Call and Keeler [1985], Morrison and Winston [1987]). The intrinsic mobility of aircraft suggests that entry and exit may be easier in airlines than in most industries, and the fact that pricing in airlines is not determined by the threat of entry provides a good reason to question whether contestability is a force in most industries (see also Schwartz [1986] and Gilbert [1989]). Whether airlines are not contestable because airline fare schedules are capable of being changed with exceptional speed, because ground support facilities are scarce and expenditures on capital other than aircraft are largely sunk costs, or because incumbents are able to deter entry strategically through the use of non-price competitive weapons is not clear. However, the question is important enough to demand both theoretical and empirical work in the different types of entry barrier or deterrence strategies that might be relevant in any particular case.

Contestability theory is an extreme characterization of market performance that relies on the assumption that entry and exit are costless. At the other end of the spectrum is the view that infinitesimal sunk costs are sufficient to protect monopoly behaviour when incumbents respond rapidly and aggressively to entry attempts. In between, there is sufficient ground for divergent views on the effects of potential competition and on the ability of established firms to engage in strategic behaviour that deters entry. The ‘structuralist school’, typified by the work of Joe Bain [1956], maintains that the efficacy of potential competition depends on determinants of the conditions of entry such as economies of scale, technological advantages, and access to marketing or natural resources. Moreover, members of this school also argue that the conditions of entry can often be manipulated by established firms in various ways to reduce the likelihood that entry will occur and to mitigate its effects. In contrast, the ‘Chicago school’, typified by the work of Stigler [1968] and Demsetz [1982], maintains that market concentration reflects the differential efficiencies of established firms, and that most of the important types of entry barriers arise from restrictions on market conduct imposed by the government. Absent these, entry conditions are generally considered to be fairly easy and market performance is generally thought to approximate competitive outcomes fairly closely. Trying to explore the determinants of entry and to ascertain which of these several views about the effect of entry on market performance is most consistent with the facts is not an easy task, and our goal here is to introduce the reader to the basic issues involved in that choice.

We shall proceed through six stages.

Section II presents several different definitions of barriers to entry that have been proposed in the literature. Section III describes Bain’s determinants of barriers to entry and introduces the role of behaviour in the determinat...

Table of contents

• Cover
• Halftitle
• Title
• Copyright
• Contents
• Introduction to the Series
• Acknowledgements
• I Introduction
• Index