Sie sind auf Seite 1von 18

Satish Jayachandran, Javier Gimeno, & P.

Rajan Varadarajan

The Theory of Multimarket


Competition: A Synthesis and
Implications for Marketing Strategy
Multimarket competition refers to competitive situations in which the same firms compete against each other in mul-
tiple markets. The theory of multimarket competition suggests that the phenomenon of mutual forbearance may re-
duce the market-level intensity of competition between two firms when the multimarket contact between them (the
number of markets in which they compete) increases. Mutual forbearance, a form of tacit collusion in which firms
avoid competitive attacks against those rivals they meet in multiple markets, is proposed to occur because multi-
market competition increases the familiarity between firms and their ability to deter each other. In this article, the
authors examine how multimarket contact increases familiarity and deterrence. Furthermore, they provide an ex-
tension of the theory of multimarket competition by developing a conceptual model that identifies competitive and
market factors that moderate the relationship between the degree of multimarket contact and the intensity of com-
petition. The authors also examine the implications of multimarket competition for marketing strategy in the context
of two marketing strategy issues: product line rivalry and entry strategy.

I
nterfirm rivalry, the extent to which firms compete against November 1989, Continental Airlines, the dominant airline
each other in a specific market through actions and reac- in Houston, retaliated against this “attack” by lowering
tions, influences their ability to gain and sustain competi- prices. However, instead of lowering prices for flights orig-
tive advantage (Dickson 1992; Porter 1980). Analysis of inating from Houston, Continental Airlines lowered prices
interfirm rivalry, an integral part of marketing strategy re- for flights out of Phoenix, Ariz., America West’s hub and
search, traditionally has been conducted by examining mar- dominant served market. Furthermore, Continental Airlines,
ket structure characteristics (e.g., concentration, entry in posting the lower fares on computerized reservation sys-
barriers), marketing strategy variables (e.g., product differ- tems, used a fare code intended to communicate its displea-
entiation), and characteristics of firms such as size and re- sure with and response to America West’s low-priced entry
sources (for example, see Buzzell and Gale 1987; Robinson into Houston. Subsequently, America West withdrew its low
1988). In an attempt to gain richer insights into this phe- introductory fare in the Houston market, following which
nomenon, marketing scholars in recent years have focused Continental Airlines withdrew its low fare promotion in the
on the behavioral aspects that drive interfirm rivalry and on Phoenix market (Nomani 1990).1
competitive decision making influenced by such rivalry The competitive interchange between these two airlines
(e.g., Clark and Montgomery 1996; Day and Nedungadi has an interesting characteristic: A competitive attack by a
1994; Heil and Robertson 1991; Mullins and Walker 1996). firm in the focal market of another firm sparked a reaction
In that tradition, we examine how multimarket or multipoint by the latter in the attacker’s focal market. Competitive in-
competition, “a situation where firms compete against each teractions of this type are common in multimarket compe-
other simultaneously in several markets” (Karnani and tition. Such competition is becoming prevalent in an
Wernerfelt 1985, p. 87), affects the competitive behavior of increasing number of industries, such as the global tire in-
firms. The following example is illustrative of multimarket dustry (Gimeno 1994); sets of related industries such as
competition. telephone and cable (Keller 1993; Parker and Roller 1997);
America West, a U.S.-based regional airline, started and industries characterized by geographically distinct
flights from Houston, Tex., with low introductory fares. In markets such as airlines, banks, and supermarket chains
(Gimeno 1994). The increase in multimarket competition
can be attributed to the increase in related product diversi-

Satish Jayachandran is Assistant Professor of Marketing, The Darla Moore


School of Business, University of South Carolina. Javier Gimeno is Assis-
1Continental Airlines and America West subsequently entered
tant Professor of Management, and P. Rajan Varadarajan is Professor of
into a code-sharing arrangement, an alliance in which they coordi-
Marketing, Lowry Mays College and Graduate School of Business, Texas
nate their flight schedules on certain routes and feed passengers in-
A&M University. The authors thank the three anonymous JM reviewers for
to each other’s connecting flights.
their comments on previous drafts of this article.

Journal of Marketing
Vol. 63 (July 1999), 49–66 Multimarket Competition / 49
fication and geographic market diversification by firms gic management (e.g., Chen 1996; Gimeno and Woo 1996;
and to the greater use of coordinated worldwide strategies Karnani and Wernerfelt 1985) and population ecology (e.g.,
by multinational enterprises (Gimeno and Woo 1999). Re- Barnett 1993; Baum and Korn 1996). The focus of the the-
search in multimarket competition suggests that firms com- ory is interfirm competition (Baum and Korn 1996). Multi-
peting in many markets may influence interfirm rivalry market competition research envisions a firm occupying a
(Baum and Korn 1996; Edwards 1955; Gimeno and Woo potentially unique market domain that is defined by activi-
1996). Because of the central nature of interfirm rivalry to ties in various geographic-product markets. There are many
marketing strategy issues, multimarket competition as a different ways of defining markets (for example, see Rao
phenomenon is of importance to marketing strategy re- and Steckel 1998, pp. 120–24). For the purpose of this arti-
search and practice. However, there is a dearth of research cle, a geographic market is defined as the lowest geograph-
in marketing literature addressing multimarket competi- ic unit that comes under the jurisdiction of a manager who
tion. Furthermore, as will become apparent from the litera- has decision-making authority over any of the competitive
ture reviewed, studies examining the relationship between strategy variables that might be employed in the context of
multimarket competition and interfirm rivalry have not competitive actions and reactions. For example, an individ-
considered the effects of various moderator variables on ual sales territory could be the lowest geographic unit if the
this relationship in a comprehensive manner. Finally, the sales manager in charge of the territory has the authority to
process by which multimarket competition affects inter- initiate actions along competitive strategy dimensions, such
firm rivalry has received only cursory attention in extant as price or advertising, to compete in that market. A product
literature. In an attempt to fill these gaps, we focus on the market is a set of goods or services that serves similar func-
following: tions, is created by the use of similar technology, and is used
by similar consumers (Abell 1980). The intersection of geo-
•The process by which multimarket competition affects inter-
firm rivalry. graphic and product markets constitutes a geographic-
•The factors that moderate the impact of multimarket competi-
product market. The market domain of a firm is the set of
tion on interfirm rivalry. geographic-product markets in which it operates. For exam-
•The implications of multimarket competition for marketing ple, if a firm operates in two geographic markets with two
strategy research and practice. products, its market domain will comprise four geographic-
product markets.
We attempt to shed insights on these issues by integrat- If the market domains of competing firms overlap in
ing the economic and behavioral underpinnings of multi- multiple geographic-product markets, the firms are engaged
market competition. A better understanding of the effect of in multimarket competition. The extent of market domain
multimarket competition on interfirm rivalry will enable overlap is characterized by the degree of mutimarket con-
marketing strategy researchers to analyze more objectively tact. Multimarket contact between two firms is the aggrega-
and marketing managers to respond better to the competitive tion of all contacts between them in geographic-product
dynamics in industries characterized by such competitive markets (Gimeno and Woo 1996). A contact occurs when
conditions. two firms compete in the same geographic-product market.
The rest of the article is organized as follows: First, we As the degree of multimarket contact between two firms in-
discuss the nature of multimarket competition and its influ- creases, they are likely to become more interdependent.
ence on interfirm rivalry by developing a process model. That is, the outcome of actions initiated by a firm becomes
Second, we propose a conceptual model that explicates the more contingent on the actions and reactions of its rivals.
moderating effects of certain competitive and market factors Furthermore, increasing multimarket contact provides firms
on the relationship between multimarket competition and in- with more opportunities to compete against each other and
terfirm rivalry. Third, we amplify the theoretical discussion retaliate in different markets. The degree of such rivalry is
in the context of two marketing strategy issues—product captured by the intensity of competition between firms, de-
line rivalry and entry strategy—to illustrate how multimar- fined as the aggressiveness and speed of the actions and re-
ket competition affects marketing strategy. Fourth, we out- actions they initiate to compete in the market (Chen 1996).
line future research directions. Fifth, we briefly discuss Intensity of competition is reflected in frequent and aggres-
some methodological issues regarding research in this area. sive changes in marketing mix variables (Chen 1996), such
Sixth, we conclude the article with a summary of the impli- as prices and advertising, that influence profit margins (Gi-
cations of multimarket competition. meno and Woo 1996, 1999).
Although multimarket competition increases the oppor-
Multimarket Competition and tunity that rivals have to compete with each other, the
greater market overlap may not translate into higher inten-
Interfirm Rivalry: The Mutual sity of competition. On the contrary, the theory of multi-
Forbearance Hypothesis market competition suggests that the intensity of
competition between firms with overlapping market do-
Multimarket Competition and Mutual Forbearance mains may be dampened by a phenomenon known as “mu-
The roots of the theory of multimarket competition can be tual forbearance” (Baum and Korn 1996, 1999; Clark and
traced to industrial organization economics (e.g., Bernheim Montgomery 1997; Edwards 1955; Gimeno 1994; Gimeno
and Whinston 1990; Edwards 1955) and sociology (Simmel and Woo 1996; Karnani and Wernerfelt 1985), which we
1950). It is a quickly evolving stream of literature in strate- discuss next.

50 / Journal of Marketing, July 1999


Mutual forbearance. Mutual forbearance is tacit collu- strategies much more than “objective reality” (Porac and
sion as a consequence of firms competing in many markets Rosa 1996).
and the resulting increase in their interdependence. Tacit The social construction of rivalry implies that firms
collusion, as opposed to direct collusion, which is illegal, is might not, as a matter of routine, realize their interdepen-
a situation in which two firms understand each other’s mo- dence with other firms and often might engage in competi-
tives and strategies and implicitly coordinate to avoid com- tive actions with little consideration as to how these actions
peting intensely. Extant theory suggests that two different influence their rivals or how their rivals may respond. For
processes may be responsible for mutual forbearance as a firms to realize that their market actions and fortunes are in-
result of higher degrees of multimarket contact: familiarity terdependent with those of other firms (which then become
(Baum and Korn 1999) and deterrence (Bernheim and rivals), it is necessary for them to become aware of these
Whinston 1990; Edwards 1955; Porter 1980). Familiarity is firms and then recognize that their market fortunes overlap.
the extent to which tacit coordination is enhanced by a This awareness is a function of the market assessment ac-
firm’s awareness of the capabilities and actions of a rival. tivities undertaken by firms and would involve collecting in-
Deterrence is the extent to which a firm is able to prevent its formation, assimilating this information, identifying rivals,
rivals from initiating aggressive actions that may be harmful and comparing strengths and weaknesses with those rivals.
to its interests in the market. It is a consequence of the abil- When firms collect information about the market, multimar-
ity of the firm to cause its rivals serious financial loss by re- ket rivals may receive greater attention than those rivals en-
taliating aggressively to their actions. In the next section, we countered in fewer markets. Consequently, firms may
examine the impact of multimarket competition on familiar- become more familiar with the strategies, capabilities, and
ity and deterrence to elucidate how multimarket competition actions of multimarket rivals. Multimarket competitors thus
engenders mutual forbearance. may become more salient rivals than those firms that are en-
In this article, the scope of actions and reactions that are countered in fewer markets. For example, with higher de-
likely to be constrained by multimarket competition is lim- grees of multimarket contact, the firms involved would have
ited to the type of actions that are undertaken repeatedly by more common prior experiences in competitive engage-
firms, such as promotional activities (consumer and trade ments. Such experiences get captured in the collective mem-
sales promotions, sales force incentives, advertising cam- ory of the firm as “war stories” (see Weick 1995), further
paigns, and so forth). Competitive actions of firms in the enhancing the firm’s beliefs regarding the competitive char-
realm of more fundamental research and development- acteristics of that rival. This familiarization process there-
driven innovations might not be constrained by multimar- fore determines the history of competition between firms in
ket competition, because it is unlikely that firms would a market. Competitive history encompasses incumbents’
hold back on introducing dominant technological break- reputation for retaliation, their tactics and strategies, and the
throughs due to mutual forbearance. Such Schumpeterian nature of their relationships with suppliers and clients (Gru-
breakthroughs would change the very nature of competitive ca and Sudarshan 1995). When firms encounter each other
advantage in the marketplace in favor of the innovating in multiple markets, they are likely to share a much richer
firm by allowing it to differentiate its offerings substantial- history of competition.
ly from those of its rivals. Firms might be inclined to view In effect, firms that are encountered in many markets are
mutual forbearance as preferable to the alternative of all- more likely to figure as important rivals than firms that are
out rivalry, but surely not as a preferable alternative to encountered in fewer markets. Therefore, multimarket con-
dominating a market by introducing potentially valuable tact facilitates mutual learning by providing these firms with
radical innovations.2 the opportunity to become familiar with each other and rec-
ognize their interdependence—how the outcomes of a firm’s
Multimarket Contact and Familiarity actions are influenced and, consequently, how its actions are
Familiarity between rivals is likely to influence the extent to constrained by the possibility of competitive reactions.
which firms engage each other with actions and reactions However, as we noted previously, mere familiarity may not
(Chen and Miller 1994). In the complex environment that be sufficient to engender mutual forbearance. For mutual
characterizes most markets (with many firms), it becomes forbearance to occur, firms also need the ability to deter
virtually impossible for firms to garner all information about each other. Otherwise, powerful firms may have no incen-
rivals and become familiar with them. In effect, the infor- tive to refrain from aggressive rivalries with relatively
mation available to firms about the competitive environment weaker firms (Teece, Pisano, and Shuen 1997). The effect of
rarely meets the information requirements for competing ef- multimarket contact on deterrence is discussed next.
fectively, because firms are complex entities with a large
number of action variables available to them. In such com- Multimarket Contact and Deterrence
plex decision arenas, firms construct imperfect versions of Deterrence is a result of the ability of firms to cause serious
the external competitive reality, and these socially con- financial damage to their rivals and occurs when firms hold
structed versions of competition may guide decisions and credible threats of retaliation against each other. In such
conditions, firms might not compete aggressively because
the expected gains from aggressive moves may be lower
2The discussion relating to the scope of actions and reactions than the future losses due to competitive retaliation. Fur-
that might be constrained by multimarket competition has benefit- thermore, in contrast to familiarity, which is gained from
ed from the comments and suggestions of an anonymous reviewer. prior interactions with a rival, deterrence requires future in-

Multimarket Competition / 51
teractions between the firms so that a response can be made what can be gained by reneging on an implicit agreement
to a current attack. In effect, it is the possibility of repeated and (2) that other firms share the same belief (Schelling
interactions in the future that allows for retaliation and cre- 1960). By increasing the ability of firms to deter each other,
ates a link between the future payoffs for a firm and its pre- multimarket competition may make the payoff from mutual
sent actions. This link is called “the shadow of the future” forbearance more attractive than that from aggressive rival-
because “the future casts a shadow back upon the present, ry. By increasing the familiarity between firms, multimarket
affecting current behavior patterns” (Parkhe 1993, p. 799). competition may make it easier for firms to realize that they
In effect, the knowledge that competitors can retaliate in the share the same beliefs regarding the beneficial nature of mu-
future against current attacks may prevent firms from un- tual forbearance and to (implicitly) coordinate their expec-
dertaking aggressive actions (Heide and Miner 1992). For tations. As Schelling (1960) notes, what deters firms is not
deterrence to take place effectively, firms should believe the aggressive retaliation that could arise when a rival initi-
that their rivals have not only the ability to retaliate aggres- ates an action in a multimarket context, but rather the ex-
sively and hurt them, but also the opportunity to do so. Next, pectation that such retaliation could occur. If the firms
we discuss why multimarket competition is proposed to in- involved in multimarket competition share this expectation
crease the ability of firms to deter each other by examining (through increased familiarity), they may eschew aggressive
how it provides firms with greater ability and opportunity to competitive behavior.
hurt their rivals through retaliation. In summary, the theory of multimarket competition
Ability to hurt. In multimarket competition, the ability to draws a distinction between the breadth and intensity of ri-
retaliate in multiple markets implies that the future losses valry between two firms. As a result of mutual forbearance,
from multimarket warfare are likely to be considerably more broad or multimarket rivals (firms that compete in many
than would be the case in a single-market context (Edwards markets) may not be the most intense rivals in specific mar-
1955). The action that a firm takes in a particular kets (Gimeno and Woo 1996). Mutual forbearance is de-
geographic-product market (e.g., through a price promotion) fined at the interfirm level in specific geographic-product
may result in gains in that market, but also may cause much markets. That is, as a result of mutual forbearance, multi-
larger losses in other geographic-product markets. The pos- market contact is proposed to affect the intensity of compe-
sibility of higher losses arises because rivals may retaliate in tition between firms within the various geographic-product
many geographic-product markets to an action that the focal markets in which they compete. The foregoing discussion
firm initiates in one geographic-product market. Therefore, describing the process through which multimarket competi-
relative to the single-market context, severe rivalry in mul- tion influences interfirm rivalry is represented by the model
timarket competition may inflict much heavier losses on shown in Figure 1. The model suggests that multimarket
firms. competition may endow firms with greater ability to deter
each other from taking aggressive actions through a “thick-
Opportunity to hurt. Although the ability to hurt rivals is er” shadow of the future. Furthermore, it may increase the
a necessary condition to cause deterrence, a firm also must level of familiarity that firms have of each other’s capabili-
perceive its rivals as having the opportunity to hurt it by re- ties and expectations through a richer history of interaction.
taliating. In the single-market context, retaliation is only Increased deterrence and familiarity may lead to mutual for-
possible in future interactions in that market. If the attacked bearance, which lowers the intensity of competition be-
market is one in which the firm has an important position, tween these firms. This discussion is summarized in the
future retaliation in that market may even be a costly move mutual forbearance hypothesis, as follows:
for the attacked firm. In the case of multimarket competi-
tion, however, the larger number of markets of overlap pro- Mutual forbearance hypothesis: There will be an inverse rela-
vide more areas to retaliate against competitive attacks. tionship between the degree of multimarket con-
tact between firms and the intensity of competi-
Therefore, relative to the single-market competitive context, tion between them in specific geographic-
multimarket competition increases the opportunities for re- product markets.
taliation by extending the interdependence of firms from the
time dimension (future) to the time and space dimensions
(multiple markets). This extension enables firms to retaliate Multimarket Contact and Intensity of Competition:
in markets in which response is less costly or more conve- A Review of the Empirical Evidence
nient for them. A summary of the empirical studies that focus on the impact
of multimarket contact on the intensity of competition be-
Multimarket Contact and Intensity of Competition: tween firms is presented in Table 1. As is detailed in Table
The Mutual Forbearance Hypothesis 1, the recent empirical evidence, by and large, lends support
The foregoing discussions on (1) multimarket contact and to the mutual forbearance hypothesis. However, the support
familiarity and (2) multimarket contact and deterrence pro- for a negative relationship between the degree of multimar-
vide insights into how multimarket contact is capable of ket contact and intensity of competition has not always been
leading to reduced rivalry through mutual forbearance. As consistent. Some previous studies find multimarket contact
game theory suggests, the basic condition for mutual for- unrelated to the intensity of competition between firms (e.g.,
bearance (tacit collusion) to be sustained would be for each Rhoades and Heggestad 1985). It is conceivable that the
firm to perceive (1) that the payoff of engaging in tacit col- lack of support for the mutual forbearance hypothesis in pre-
lusion or mutual forbearance is likely to be greater than vious research may be due to the use of cross-sectional

52 / Journal of Marketing, July 1999


FIGURE 1
Multimarket Contact and Intensity of Competition: A Process Model

Multimarket contact
between focal firm and
rivals

Thicker shadow of the future

Ability to hurt Opportunity to hurt


Larger number of
interactions with rivals
Larger revenue exposure Rivals’ opportunity to
to rivals’ actions retaliate in multiple markets

Better understanding of Greater attention to rivals in


interdependence and market scanning and Lower expected payoff
overlapping market fortunes competitor information from rivalry
with rivals acquisition

Richer competitive history

Increased familiarity Increased deterrence

Mutual forbearance

Lower intensity of
competition

methods, in contrast with the use of longitudinal methods in three competitive factors (spheres of influence, resource
more recent research. similarity, and organizational structure of competing firms)
Although the mutual forbearance relationship has been and one market factor (seller concentration) in the relation-
examined in prior research, the factors that moderate the re- ship between multimarket contact and intensity of competi-
lationship between multimarket contact and intensity of tion. Spheres of influence refers to the extent to which
competition have not received adequate attention in extant different multimarket competitors have dominant market
literature. Although different studies have focused on some positions in different markets (Edwards 1955). Resource
of these moderators (see Table 1), there has been no attempt similarity is the extent to which the competing firms are
to examine integratively the impact of these moderators on comparable in terms of resource endowments, or resources
the relationship between multimarket contact and intensity critical for success in the market (Chen 1996). Organiza-
of competition. Further theoretical development of the liter- tional structure refers to the extent to which a firm is able to
ature stream on multimarket competition can be aided by control and coordinate its actions jointly in multiple
such an attempt. This issue is addressed next. geographic-product markets. Seller concentration is a mea-
sure of the number of firms in a market and the pattern of
dispersion of total industry sales among competitors in that
Multimarket Contact and Intensity of market. A conceptual model delineating the moderating in-
Competition: A Contingency Model fluences of these variables is presented in Figure 2 and dis-
Moderator variables affect the relationship between multi- cussed next.
market contact and intensity of competition between firms
through the familiarity between firms and/or their ability to Spheres of Influence
deter each other. Such moderating variables are typically Spheres of influence occur when firms engaged in multi-
characteristics of the competitive or market environment. market competition dominate different markets among those
Our discussion here focuses on the moderating effects of in which they overlap (Edwards 1955; McGrath, Chen, and

Multimarket Competition / 53
TABLE 1
Empirical Research in Multimarket Competition: A Summary
Authors Data and Sample Variables and Measures Major Findings

Heggestad and Rhoades 187 local banking areas in Dep.: Rivalry measured as Multimarket contact reduces
(1978) 1972 market share instability rivalry in banking areas.
Indep.: Count measure of
multimarket contact
Controls: Concentration,
growth, limited branding
dummy, unit banking
dummy

Scott (1982) 437 firms (among the 1000 Dep.: Profit/sale per Line of Interaction between
largest) included in the Business multimarket contact and
FTC Line of Business data Indep.: Probability of concentration has a
in 1974 observing less contact and significant effect. Profits
its dummy above the are approximately 3%
median value higher when both seller
Controls: Concentration, concentration and
minimum efficient scale, multimarket contact are
advertising intensity, high.
geographic market size,
assets/sales, market share

Alexander (1985) 67 market areas (banking) in Dep.: Performance (service Quadratic interaction effects
six states charge ratio in deposits, of multimarket contact with
interest rate on loans) concentration were more
Indep.: Multimarket contact significant.

Feinberg (1985) 391 manufacturing Dep.: Income/sales Company study supports


companies in the FTC Indep.: Multimarket contact mutual forbearance
Line of Business Data for Controls: hypothesis and finds a
1976 Company study: Market quadratic interaction with
share, average sales per concentration.
market, concentration, Industry study supports
assets/sales, advertising mutual forbearance
ratio hypothesis.
Industry study:
Diversification,
assets/sales, advertising
ratio, market growth,
imports, minimum efficient
scale, consumer focus,
cost disadvantage ratio

Rhoades and Heggestad A: Same as in Heggestad Dep:. Industry performance The results were contrary to
(1985) and Rhoades (1978) variables mutual forbearance
hypothesis.
B: 154 banking areas Dep.: Rivalry operationalized No relationship to
including 1074 banks as change in rank multimarket contact.
C: 210 banking areas with Dep.: Performance No relationship to
1443 banks (For study B and C, other multimarket contact.
variables were the same
as in Heggestad and
Rhoades 1978)

Mester (1987) 171 savings and loan firms Dep.: Rivalry measured as Contact affects behavior in
in California instability of market 77.8% of cases. The
shares, and performance interaction between
as firm profits and prices contacts and
Indep.: Multimarket contact concentration was more
as count and probabilistic significant.
measures Effects contradict mutual

54 / Journal of Marketing, July 1999


Authors Data and Sample Variables and Measures Major Findings
forbearance hypothesis. Sandler (1988) 123 airline markets during Dep.: Log of relative market
1974–1976 and share instability
1978–1980 Indep.: Concentration,
dummy for airports with slot
constraints, multimarket
contacts among rivals, new
entry dummy, labor strike
dummy, deregulation
dummy
Multimarket contact is Singal (1993) 14 mergers among airline Dep.: Yield per mile
related positively to rivalry. companies between 1984 (performance), change in
and 1987 yield per mile
Indep.: Multimarket contact

Multimarket contact has a Barnett (1993) Life history of every firm that Dep.: Exit rate from a market
significant positive effect operated in the consumer Indep.: Number of single-
on performance, thereby premises equipment and point and multipoint
supporting mutual service sector of the competitors
forbearance hypothesis. telephone industry in any
state from 1981 to 1986
If market is critical for a firm, Hughes and Oughton (1993) 418 manufacturing firms in Dep.: Industry price–cost
multimarket contact 1979 data from U.K. margin, industry rate of
reduces rivalry. Census of Production return
Indep.: Multimarket contact
Controls: Concentration,
minimum efficient scale, %
imports, % exports,
research and development
intensity, diversification,
growth rate, capacity
utilization, capital stock,
capital–output ratio,
presence of foreign
multinational enterprises
Multimarket contact has Evans and Kessides (1994) 1000 largest routes in U.S. Dep.: Log of average price
significant positive effect airline industry between Indep.: Direct flight, round-
on performance. 1984 and 1987 trip ticket, distance, route
market share,
concentration (route and
airport), airport market
share, multimarket contact
Multimarket contact has a Baum and Korn (1996) Data describing the route Dep.: Rate of market entry
strong positive effect on changes of California- and exit
prices. based commuter air Indep.: Market domain
carriers from January overlap, multimarket
1979 through December contact, concentration,
1984 spheres of influence

Multimarket contact and its Gimeno and Woo (1996) U.S. Department of Dep.: Yield (average price
interaction with spheres of Transportation data on 48 charged divided by the
influence are related airlines across 3171 distance of the market)
significantly to lower entry markets for fourth quarters Indep.: Multimarket contact,
and exit. However, the from 1984 through 1988 strategic similarity
interaction of multimarket Controls: Service attributes,
contact and concentration exogenous market
was insignificant. characteristics, cost
position, market structure
Multimarket contact strongly Jans and Rosenbaum (1996) 25 regional cement markets Dep.: Price
decreases rivalry, whereas in the United States Indep.: Multimarket contact,
strategic similarity (1974–1989) concentration
moderately increases Controls: Price of inputs,
rivalry. capacity, size, age,
process technology

Multimarket Competition / 55
Authors Data and Sample Variables and Measures Major Findings
Multimarket contact has a Boeker et al. (1997) 286 California-based Dep.: Market exit (as an
greater positive effect on hospitals from California indirect indicator of
price as home market Health Facilities intensity of competition)
concentration increases. Commission’s annual Indep.: Multimarket contact
disclosure survey (market overlap for
(1980–1986) specific service), mode of
service, chief executive
change, performance
Control: Density, contracting,
historic exit rate, statewide
service density, medical
doctors per capita,
ownership, hospital size

Multimarket contact reduces Parker and Roller (1997) Mobile (cellular) telephone Dep.: Price–cost margin.
the probability of exit from industry in the United Indep.: Regulation (none,
a market (because of States low, high), dummy
lower intensity of variables for specific firms
competition). and competitors in specific
markets, cross ownership,
multimarket contact, first
entrant’s lead over the
second entrant, age of the
cellular system

Multimarket contact explains Fernandez and Marin (1998) 2221 hotel establishments in Dep.: Price
noncompetitive prices (as Spain Indep.: Multimarket contact,
a result of lower intensity concentration
of competition). Controls: Age, hotel
category, quality distance,
local wages, local demand

Multimarket contact has a Baum and Korn (1999) Data describing the route Dep.: Rate of market entry
positive effect on collusion changes of California- and exit
at low levels of market based commuter Indep.: Multimarket contact,
concentration; negative aircarriers from January multimarket contact with a
effect at high levels of 1979 through December rival relative to the
market concentration. 1984 multimarket contact with
other competitors, size of
competitor
Control: Focal airline
characteristics, competitor
airline characteristics,
aggregate environmental
characteristics

Multimarket contact has an were considered the Gimeno (1999) U.S. Department of
inverted-U relationship aggregate measure of Transportation data on 48
with rate of entry and exit. interfirm rivalry. airlines across 2897
Relative multimarket markets
contact and the interaction
of multimarket contact with
firm size have significant
effects on entry and exit.
The rate of entry and exit

Dep.: Yield (average price firm dominance Reciprocal multimarket Gimeno and Woo (1999)
per mile), market share contacts decrease rivalry
Indep.: Reciprocal and and increase market share
nonreciprocal multimarket sustainability more than
contact nonreciprocal multimarket
Controls: Service attributes, contacts.
market attributes, cost
position, market structure,

U.S. Department of Transportation data on 28 airlines across 3000 markets from 1984 to 1988

56 / Journal of Marketing, July 1999


FIGURE 2
Multimarket Contact and Intensity of Competition: A Contingency Model

Organizational
Seller
structure of
concentration
competing firms

P3 P4
Degree
of Intensity of
multimarket competition
contact
P1 P2

Spheres of Resource
influence similarity

MacMillan 1998). When firms are involved in multimarket mechanisms for the transfer of power; retaliatory power in
competition, their market position may not be similar in all one market can be deployed to sustain a dominant position
overlapping markets. Differences may arise from factors in another market (Bernheim and Whinston 1990). In the ab-
such as differing area and levels of technological knowl- sence of spheres of influence, Firm A may have no reason to
edge, specialization, and different transportation costs be- cooperate with Firm B in Market Y. In effect, Firm B has
cause of different plant locations (Bernheim and Whinston transferred the power to retaliate against Firm A in Market
1990; Gimeno 1999). This will result in spheres of influ- X to enforce Firm A’s cooperation in Market Y. However,
ence, that is, different firms being dominant in different for spheres of influence to take effect, firms need credible
markets. If the firms that have multimarket contact also have threats of retaliation in each other’s dominant markets.
different spheres of influence, this may accentuate the in- Footholds in rivals’ spheres of influence manifest them-
verse relationship between multimarket contact and intensi- selves as credible threats (Karnani and Wernerfelt 1985).
ty of competition further (Baum and Korn 1996). Bernheim and Whinston (1990) suggest that it is spheres
Consider, for example, a two-firm, two-market situation, of influence that lead to the reduction in the intensity of
in which Firms A and B compete against each other in mar- competition in a multimarket scenario and that mere aggre-
kets X and Y, so that Firm A is dominant (the market leader) gation of contacts might not lead to mutual forbearance.
in Market X and Firm B is dominant in Market Y. If A at- However, it seems likely that aggregation of contacts, as de-
tacks B in Market Y, B is likely to retaliate by attacking A in scribed previously, will lead to mutual forbearance because
the market in which A has more to lose, namely, Market X. of increased deterrence and familiarity. Spheres of influence
At the same time, Firms A and B will want to prevent severe will enhance deterrence because firms will be interested in
rivalry in the markets they dominate, because if such an protecting specific key markets from retaliatory moves by
eventuality arises, the dominant firm is the one that has rivals and therefore may refrain from taking actions that are
more revenue or profits at risk. This situation may lead to likely to provoke reactions in such markets. In effect, at a
reciprocal tacit collusion, in which A will allow B to domi- given level of multimarket contact among firms, the pres-
nate Market Y as long as it is allowed to dominate Market ence of different spheres of influence may make the firms
X. In a way, the asymmetry of competitive market positions even less prone to taking aggressive actions (Baum and Ko-
in different markets leads to symmetry in overall retaliatory rn 1996). In such cases, firms would want to prevent multi-
capability for the firms. This, in turn, may enhance mutual market competitive reactions that are targeted at their
forbearance and lower the intensity of competition between spheres of influence. Spheres of influence affect the rela-
these firms. Spheres of influence limit rivalry by serving as tionship between multimarket contact and intensity of com-

Multimarket Competition / 57
petition by influencing the ability of firms to deter each oth- Proposition 2: The negative relationship between the degree of
er. This argument has received empirical support in Gi- multimarket contact and the intensity of compe-
meno’s (1999) work. tition between firms will be stronger in condi-
tions of high resource similarity than in condi-
Proposition 1: The negative relationship between the degree of tions of low resource similarity.
multimarket contact and the intensity of compe-
tition between firms will be stronger in condi-
Organizational Structure of Competing Firms3
tions of spheres of influence than otherwise.
The preceding discussion makes it apparent that appropriate
Resource Similarity coordination and control among the different organizational
units that manage the activities in the different geographic-
Resource similarity is the extent to which a firm’s tangible product markets are critical for the effectiveness of multi-
and intangible resource endowments (resources critical for market strategies (Gimeno and Woo 1999). For example,
success in the market) are comparable to those of its com- consider Firm A and Firm B, which compete with each oth-
petitors. Resource similarity may affect interfirm rivalry be- er in Markets X and Y, with different administrative units
cause competitive advantage is derived from the ability of a managing the operations of the firms in each market. For
firm to develop and sustain valuable resources that are dif- mutual forbearance to take place, the two firms should have
ferent than those of its competitors (Teece, Pisano, and administrative units that are willing and capable of coordi-
Shuen 1997). Consequently, firms with similar resources are nating their strategies across the two markets. That is, mul-
likely to have similar strategic strengths and weaknesses in timarket competition will lead to mutual forbearance and
the marketplace (Chen 1996) and compete using similar lower intensity of competition only if each firm achieves ef-
strategies (Gimeno and Woo 1996). fective coordination between the administrative units that
As we noted previously, rivalry is a result of social con- manage the operations in the two markets. In the absence of
structions of the competitive environment by firms. Firms such intrafirm coordination, competition converges to
with similar resources may be more likely to view each oth- market-by-market competition. Such competition may be
er as significant competitors because they will be more alert less than optimal for overall firm performance. For example,
to each other’s actions (Porac and Thomas 1990). However, an administrative unit within a firm, in its attempt to maxi-
this increased focus also means that these firms may be able mize its performance within its geographic-product market,
to understand each other’s strategies and capabilities better may initiate actions that lead to multimarket retaliation by a
and become more familiar with each other compared with better coordinated rival, which leads to losses in many mar-
firms that are less similar in terms of resource bundles. The kets, even if the administrative unit that initiates the original
ability of these firms to become familiar with each other is action benefits from the outcome. In effect, what is optimal
aided further by the likelihood that firms with similar re- behavior at the administrative unit level may be suboptimal
sources will pursue similar strategies. at the firm level. Therefore, organizational structures that
In addition, firms with similar resource bundles may be provide firms with the ability to coordinate actions across
in a better position to sustain a cooperative arrangement, markets may improve their ability to deter rivals in a multi-
such as mutual forbearance, because they hold credible market context. This discussion is summarized in Proposi-
threats of retaliation against each other (Chen 1996). That is, tion 3:
firms with similar resource bundles may be able to deter Proposition 3: The negative relationship between the degree of
each other better because they can match competitive ac- multimarket contact and the intensity of compe-
tions, thus rendering the actions invalid (Teece, Pisano, and tition between firms will be stronger when the
Shuen 1997). In signaling terms (Heil and Robertson 1991), firms have organizational structures character-
ized by a greater degree of coordination of
a signal from a competitor endowed with similar resources actions across the units that manage decision
is likely to be considered more credible than a signal from a making in the different markets than otherwise.
competitor endowed with dissimilar resources. Consequent-
ly, regardless of multimarket contact, if one firm has a re-
Seller Concentration
source advantage over its rival, it may not be motivated to
forbear from aggressive competition because of the percep- Seller concentration is a measure of the oligopolistic na-
tion that it can outmaneuver the competition, which cannot ture of the market and an important predictor of compet-
match its resources. Therefore, the ability of multimarket itive behavior. It is often computed as the percentage of
contact to reduce intensity of competition between firms be- market sales accounted for by the largest few firms in the
comes greater when the firms are similar in their resource market (e.g., four- and eight-firm concentration ratios) or
endowments. Empirically, Baum and Korn (1999) find that by the Herfindahl-Hirschman index, defined as the sum
size differential of firms, an indicator of resource dissimi- of squared market shares (Oster 1994). Multimarket con-
larity, interacted with multimarket contact in influencing the
rate of entry and exit in a market. In summary, resource sim-
ilarity enhances the familiarity between multimarket com- 3We gratefully acknowledge that organizational structure-related
petitors and increases the credibility of retaliation issues raised by two anonymous reviewers were instrumental in
expectations, thus increasing the forbearance effects of mul- our refining the conceptual model and including organizational
timarket contact. structure as a moderator variable.

58 / Journal of Marketing, July 1999


tact may fail to sustain a tacit collusive arrangement, such action effect for focal market concentration and multimarket
as mutual forbearance, in conditions of low seller con- contact on intensity of competition. That is, they find that
centration. It is likely to have little additional effect on multimarket contact is most effective as a mechanism for
mutual forbearance when seller concentration is very enforcing mutual forbearance in markets of moderate con-
high. In markets of moderate concentration, however, the centration. The curvilinear moderating impact of seller con-
impact of multimarket contact on intensity of competition centration on the relationship between multimarket contact
is likely to be most pronounced. A more detailed discus- and intensity of competition is due to multimarket contact
sion follows. and seller concentration being substitutes at high levels of
concentration, though they are complementary at moderate
Low seller concentration. In conditions of low seller
levels of concentration. These arguments are summarized in
concentration, firms do not possess the ability to deter each
the following proposition:
other, because no firm has enough of a stake in the market
(Areeda and Turner 1979). Furthermore, the difficulty that Proposition 4: The negative relationship between the degree
firms may encounter in becoming sufficiently familiar with of multimarket contact and the intensity of
each other when there are many competitors also suggests competition between firms will be stronger in
that multimarket contact may not be effective as a learning conditions of moderate seller concentration
mechanism to ensure tacit collusion in fragmented markets. than in conditions of either high or low seller
concentration.
Moreover, retaliatory moves in less concentrated markets
influence many firms, which may lead to escalation of ri- In summary, the relationship between multimarket con-
valry rather than deescalation. tact and intensity of competition is posited to be moderated
High seller concentration. In conditions of high seller by the impact of the four variables discussed in this section.
concentration, because a significant share of the total market Further advancement of the theory of multimarket competi-
revenues will be dispersed among fewer firms, firms are tion calls for exploration of the implications of the phenom-
more likely to refrain from intense rivalry because they enon for marketing strategy and research and empirical
would have sizable revenues and market shares at stake. For verification of the moderating effects of the variables speci-
example, the absolute monetary loss from an aggressive fied in Figure 2.
price war may be higher compared with what it would be in
conditions of low seller concentration. Therefore, the poten-
tial for higher losses that already exists when seller concen- Implications for Marketing Strategy
tration is high, along with increased opportunities for and Research
retaliation, provides the motivation for deterrence in multi- The implications of multimarket competition for marketing
market competition. strategy and research are discussed here in reference to two
Furthermore, when seller concentration is high, firms issues germane to marketing strategy: product line rivalry
have a higher likelihood of becoming more familiar with and market entry.
each other because, when faced with fewer competitors,
they are likely to be able to monitor their competitors better The Multimarket Nature of Product Line Rivalry
(Porac and Rosa 1996). Implicit coordination of activities Product line rivalry. A product line is defined as “a
between firms therefore may be easier when familiarity en- group of products within a product class that are closely re-
ables them to recognize their interdependence. In effect, lated because they function in a similar manner or are sold
firms may not even require multimarket contact to become to the same customer groups or are marketed through the
familiar with each other or attain a market position at which same types of outlets or fall within given price ranges”
it serves the firm better to refrain from intense competition. (Kotler 1994, p. 434). Product line rivalry is defined as the
That is, concentrated markets may provide firms with the competitive actions and reactions taken by firms when they
necessary and sufficient conditions for mutual forbearance compete with one another using product lines. Firms often
that multimarket contact provides—familiarity and deter- compete using product line strategies in multiple product
rence—and thereby serve as a substitute for multimarket markets. For example, the product line rivalry among the
contact as far as creating conditions for tacit collusion is consumer products multinationals Colgate-Palmolive,
concerned. As a result, multimarket contact may have little Henkel, Procter & Gamble, and Unilever spans across a va-
additional effect on forbearance in concentrated markets. riety of geographic markets (e.g., several countries in
Moderate seller concentration. In markets of moderate Africa, Asia, Europe, North America, and South America)
concentration, in which firms may lack the necessary power and product lines (e.g., laundry detergents, toiletries, house-
to deter rivals, as well as the necessary familiarity to coor- hold cleansers).
dinate activities tacitly, multimarket contact may be able to Product lines emerge as a result of intrafirm product
enhance both these conditions to foster mutual forbearance. differentiation. Product differentiation is the process by
Therefore, multimarket contact is more likely to ensure mu- which marketing managers make a product physically
tual forbearance in markets of moderate concentration and/or perceptually different from competing products. It
(Areeda and Turner 1979; Bernheim and Whinston 1990) is a generic strategy for achieving competitive advantage
compared with markets of high or low concentration. In sup- (Porter 1980). Often, firms introduce differentiated prod-
port of this argument, Alexander (1985), Feinberg (1985), ucts within a product market as a growth strategy (to
and Gimeno, Marin, and Woo (1998) find a quadratic inter- cater to varying consumer tastes or exploit economies of

Multimarket Competition / 59
scale and scope) or as an entry deterrent strategy (to pre- foothold retaliation may limit firms to deterring each other
vent rivals from entering the market), thereby creating with limited presence in each other’s markets.
product lines (Oster 1994). When competing firms seg- The pet food industry: An illustration of the multimarket
ment the market in similar ways, product line rivalry be- nature of product line rivalry.4 The case of the U.S. pet food
comes a form of multimarket competition. This is industry (Stuart 1991) provides additional insights into the
explained next. multimarket nature of product line rivalry. The U.S. pet food
Product line rivalry and multimarket competition. industry consists of dog food in five segments (dry, semi-
Product line strategies associated with competing firms moist, canned, snack, and soft-dry) and cat food in three
segmenting the market in similar ways is likely to bring segments (dry, moist, and canned). The industrywide market
firms into greater contact with each other, leading to in- share leader in 1985 was Ralston Purina (26.9%), followed
creased multimarket competition. That is, the more simi- by Carnation Foods, a subsidiary of Nestlé (12.2%). The
lar the product lines of two firms, the greater the other leading players in the market were Kal Kan Foods, a
multimarket contact possibilities are, provided the firms subsidiary of Mars (8.4%); H.J. Heinz (7.8%); Quaker Oats
operate in the same geographic markets. Product line sim- (7.1%); Alpo Pet Foods, a subsidiary of Grand Metropolitan
ilarity implies that firms have lines of similar breadth PLC (7.1%); and Gaines Pet Foods (6.6%).
(number of products) targeted at similar market segments. The segmentwise market share figures (Maxwell 1986)
Given the notion of similar product lines bringing firms reveal that different firms dominated different segments in
into greater contact with each other, product line rivalry 1985. For example, Ralston Purina was the market share
may be influenced by the characteristics of multimarket leader in the dry dog food and dry and moist cat food seg-
competition. When the product lines of two firms evolve ments and was the second-biggest player in snack and soft-
to become more similar and their product markets in- dry dog food segments. Carnation was the market share
creasingly overlap, the greater multimarket contact may leader in the canned cat food segment and the second-largest
brand in the dry cat food segment. Furthermore, Carnation
reduce the intensity of competition between firms in a spe-
had a strong presence in the canned dog food segment.
cific market. Furthermore, if different firms dominate dif-
Heinz led the snack dog food segment and was a strong sec-
ferent product markets (segments), these markets
ond in the canned cat food segment. Gaines dominated the
(segments) may serve the function of spheres of influence,
semimoist dog food segment, and Quaker Oats led the soft-
thereby providing added motivation for the firms to re-
dry dog food segment. Alpo dominated the canned dog food
duce the intensity of competition.
segment. All the leading firms, except Kal Kan, dominated
However, before mutual forbearance takes effect, firms
at least one of the different segments in the market. This
may extend their product lines and enter different markets. suggests that the pet food industry was an example of an in-
The motivation for this strategy normally is attributed to the dustry with product line competition in which different
desire of firms for growth or entry deterrence, but it also firms dominated different segments of the market.
could follow from firms’ attempts to establish footholds in The theory of multimarket competition suggests that the
markets dominated by their rivals to signal their intent and firms may observe mutual forbearance in these circum-
ability to compete aggressively in those markets, if warrant- stances. However, if the mutual forbearance behavior is up-
ed (Karnani and Wernerfelt 1985). Therefore, firms may use set by an action taken by a competitor, the reactions may be
product lines to establish footholds in the segments domi- swift and aggressive. In this particular case, actions taken by
nated by their rivals and deter these rivals from becoming Quaker Oats sent the industry into a spiral of competitive
overly aggressive in their own spheres of influence. activity. By acquiring Gaines in 1986, Quaker entered the
Although establishing these footholds may lead to retal- dry dog food segment dominated by Ralston Purina. Ralston
iation and, in turn, to costs associated with combating such Purina retaliated by acquiring Benco in 1987 and establish-
rivalry, firms still may establish footholds because they ing a foothold in the semimoist dog food segment dominat-
serve as credible commitments of retaliation to future at- ed by Quaker Oats. Following this, all major competitors in
tacks in their spheres of influence. Deterrence is often pos- the industry attempted to broaden their product scope (from
sible only if a firm establishes a credible commitment to dog foods or cat foods to dog foods and cat foods), conceiv-
respond should it be subjected to competitive attacks in its ably to establish footholds in their competitors’ spheres of
markets (Schelling 1960). A firm’s actual presence in a mar- influence. For example, Heinz, a dominant firm in the cat
ket is likely to be perceived by rivals as a more credible food segments, entered the dry and canned dog food seg-
commitment of the firm’s potential to retaliate than the mere
possibility that it may enter the market. Furthermore, actual
presence may be required to highlight interdependence,
through heightened familiarity, to prevent future attacks by
a firm’s rivals. Therefore, even if a firm is aware that a com- 4This example is meant to illustrate and clarify the nature of

petitor is likely to respond in kind to its attempts to establish multimarket competition. We do not intend it to be considered ro-
bust empirical evidence of the phenomenon, because this example
a foothold in its competitor’s market, it still may attempt to
is susceptible to alternative explanations. At the very least, we are
do so because the cost of building these footholds may be not in a position to rule out such alternative explanations. Never-
considered lower than the cost of aggressive rivalry. Even if theless, this example illustrates the nature of firm behavior in mar-
the original entry may have been with aggressive intent, a kets characterized by multimarket competition.

60 / Journal of Marketing, July 1999


ments, and Alpo, a major player in the dog foods market, en- which are the reasons commonly attributed to the use of
tered the dry cat food segment and enhanced its market po- product line strategies by firms (Oster 1994). As we de-
sition in the canned cat food segment from .4% to 9.1%. By scribed previously, a firm may use a product line strategy to
1990, all firms had established a presence in three or more establish footholds in competitors’ key markets to protect its
of the dog food segments and two or more of the cat food interest in the segments that it dominates. This dimension of
segments (see Maxwell 1991). As a consequence, in 1990, the strategic intent of a firm is likely to become apparent on-
the industry was characterized by much higher multimarket ly when the multimarket nature of product line rivalry is
contact than in 1985. Subsequent to the episodes of mutual considered.
invasion, that pattern and intensity of competition changed Furthermore, for the firms involved in product line ri-
somewhat. In July 1990, Ralston Purina, the industry leader, valry, even simple actions such as a price promotion in a
announced its first price increase in two years, which was particular geographic-product market could have disastrous
followed by most rivals (Stuart 1991). The firms also consequences if they send firms on a competitive spiral that
stopped their stream of product introductions in their rivals’ has effects in many markets. Therefore, in such situations,
main markets and began launching superpremium exten- firms may need to consider the consequences that their ac-
sions of their existing brands (e.g., Purina O.N.E. by Ralston tions conceivably could have on multiple markets that nor-
Purina, Expert by Kal Kan, and “advanced formula” Cycle mally might not enter into their strategic calculus when
by Quaker Oats). The segment positions of the dominant contemplating a plan of action. In addition, as we noted in
firms changed little between 1990 and 1992.5,6 All these fac- the section on the moderating impact of organizational
tors imply a likely decrease in the intensity of competition structure on mutual forbearance, there is a need for efficient
among the firms in the period between 1990 and 1992. This information sharing and coordination among the managers
example serves to illustrate the dynamics of multimarket responsible for various products in a product line in order to
competition when firms compete using product lines. implement efficient multimarket strategies. The need for
such coordination may have been an underlying motivation
Observations. The foregoing discussion demonstrates
for the increasing use of category management by firms.
how firms might use product lines to establish footholds in
Category management, in which a single decision maker is
the segments dominated by their rivals. When these firms
responsible for a product line (Zenor 1994), facilitates bet-
established footholds in the critical markets of their rivals
ter coordination of strategies for the various brands within a
and signaled their capability to retaliate against the segment
product line and thereby enables firms to formulate and im-
leader, each firm seemed to focus on enhancing its position
plement multimarket strategic actions for the whole product
further in those market segments in which it already held a
line. The category manager may be willing to forbear from
dominant position. In effect, multimarket competition fos-
competitive actions in some products to deter rivalry in
tered by the similarity in the product lines of firms may con-
some other products within the product line.
strain their strategic flexibility. Furthermore, product line
strategy may have a strategic import that goes beyond the
Market Entry Strategies
desire of firms to block competition, serve varied consumer
tastes, and exploit potential economies of scale and scope, Entry strategy and competitive history. Market entry
from the perspective of the incumbent, as well as the per-
spective of the new entrant, has been a topic that has re-
ceived a significant amount of attention in marketing
(Clark and Montgomery 1997). Entering a new market with
5The end of mutual invasion strategies can be identified by
a new or existing product is a commonly used growth strat-
tracking the changes in the firms’ revenue concentration in a few egy, but it is also prone to extremely high failure rates. New
segments. A Herfindahl-Hirschman index is constructed for each
product failure rate is often as high as 80%, which suggests
firm and reflects the concentration of revenues in a few segments.
A higher index reflects that firms are focusing on fewer segments the need for a well-conceived plan for entering the market
for a greater portion of their revenues (greater revenue concentra- that considers the various facets that may influence the suc-
tion in a few markets), and a lower index represents greater diver- cess of the entrant (Green, Barclay, and Ryans 1995). Gru-
sity across segments (revenues for firms are spread out more ca and Sudarshan (1995), in developing a framework for
evenly across various segments). The revenue weighted index for entry deterrence strategy, suggest that the competitive en-
firms was highest in 1983 and fell quickly between 1985 and 1990. vironment is a key factor that, through its dimensions of
After 1990, the index increased again, which implies that firms
were relying more on their spheres of influence for increasing their
cost conditions, demand conditions, legal climate, and
revenues and initiating fewer entries into other segments. competitive history, shapes an incumbent’s response to an
6The market share instability (an indicator of intensity of com- entrant’s strategy. Competitive history, which encompasses
petition) among major firms was evaluated using market share in- incumbents’ reputation for retaliation, their tactics and
stability indexes for each segment (Ogur 1976), with market shares strategies, and the nature of their relationships with suppli-
defined relative to all major firms in the industry (to eliminate the ers and clients (Gruca and Sudarshan 1995), can act as an
effect of marginal firms). The revenue weighted average across effective barrier to competitive entry. Competitive history,
segments indicates that market share instability peaked in 1986 and
was also high in 1989–90. After 1990, market share instability de- in this sense, is the knowledge that potential entrants have
creased, with the 1992 index being the lowest in the 1983–1992 pe- about incumbents’ ability and motivation to respond, and it
riod. This also implies that intensity of competition in this industry may influence their decision to enter specific markets. As
decreased after 1990. Clark and Montgomery (1997) argue and experimentally

Multimarket Competition / 61
demonstrate, a potential entrant’s perception of the aggres- Observations. Firms, when they formulate market entry
siveness and intelligence of incumbents determines its own strategies, may benefit from considering the potential im-
propensity to enter a market. Therefore, competitive histo- pact that their decision to enter new markets can have on
ry is likely to be a critical input as firms devise entry and other markets in which they currently compete. This calcu-
entry deterrence strategies to gain and sustain competitive lation may change the estimates of future revenues that
advantage might result from entry. Entry strategy formulation based on
Multimarket competition and competitive history. When a broader construal of the competitive history construct also
incumbents and potential entrants are single-market firms, may sensitize a firm to other markets it can enter more ef-
potential entrants must evaluate competitive history indi- fectively. In some of these markets, though the projected
rectly, because by definition, they are not present in the mar- gross revenue may be lower, the competitive retaliation also
ket that they are trying to enter. When firms compete in may be lower and the actual net revenue therefore higher. In
multiple markets, however, the potential entrant can use the other words, consideration of the impact of multimarket
lessons from the rich history of interactions with the incum- competition may significantly alter the outcome of the
bent in multiple markets for strategic decisions on market cost–benefit analysis that firms undertake before a market
entry. Multimarket contact fosters greater familiarity among entry, thus influencing the choice among alternative entry
firms about the ability and motivation of their rivals to act strategies.
and react in a given competitive environment. Therefore,
multimarket competition contributes an additional dimen- Future Research Directions
sion of competitive history among firms in a market. The in- The research stream in multimarket competition is still
creased number of interactions, as a result of multimarket evolving, thus providing several avenues for further re-
contact, will both highlight and shape the rivalry among search. Although the mutual forbearance effect of multimar-
firms. Therefore, the behavior of firms in a multimarket con- ket competition (the inverse relationship between the degree
text, rather than in a single market, constitutes the entire of multimarket contact and intensity of competition) has re-
competitive history among them. These considerations sug- ceived empirical support, the contingency effects of perti-
gest the need for broadening the scope of the competitive nent competitive and market factors have not been
history construct to encompass the history of competition in examined in an integrative manner. Against this backdrop,
not only the focal market, but also other markets in which an empirical examination of the conceptual model devel-
the fortunes of the firms are interdependent. oped here could lend additional insights into the nature of
The entry strategy of U2: An illustration of multimarket the relationship between multimarket contact and intensity
contact and competitive history.7 In 1994, United Airlines of competition.
announced plans for launching a short-haul subsidiary, The process by which multimarket competition influ-
named U2, to cater to several California commuter airtravel ences the intensity of competition between firms also calls
markets dominated by Southwest Airlines. As of December for empirical verification. That is, the role that multimarket
1993, United and Southwest were competing head-to-head contact plays in enhancing deterrence and familiarity among
in 95 other city pair markets in the industry, from which firms should be verified to lend substance to the theory that
United and Southwest obtained 9% and 31% of their air- underlies the relationship between multimarket contact and
travel revenues, respectively.8 Southwest signaled its intent intensity of competition. This would call for approaches that
to retaliate in multiple markets to this proposed action by capture the mental maps of managers in firms that engage in
publicizing its plans to enter some longer-haul routes served substantial multimarket competition. The extent to which
by United Airlines, conditional on United’s behavior. Over multimarket contact plays a role in the determination of a
time, Southwest Airlines has developed a reputation of be- firm’s familiarity with other firms and the degree to which it
ing a competitor that does not capitulate, a reputation that deters these firms from acting against each other can be
was built on the basis of its history of rivalry with other air- gauged through detailed interviews with managers, observa-
lines in many markets. Although U2 initially tried to com- tion of actual behavior, or self-report measures. Further-
pete directly with Southwest, it swiftly made its exit from more, such studies will help shed light on how managers
many routes because Southwest, true to reputation, did not mentally map their market and determine their rivals and
yield. As of now, U2 has only minimal market overlap with thereby contribute substantially to strategic decision-making
Southwest Airlines. It is conceivable that the rapid reposi- knowledge. The influence of multimarket competition on
tioning of U2 away from Southwest Airlines’ routes in Cal- the cognitive construction of rivalry is, therefore, another
ifornia was also based on considerations, such as the latter’s promising stream of research.
competitive history in mutually contested markets and the Another critical issue regarding multimarket competi-
possibility that it might retaliate in multiple markets, as tion is the extent to which managers use multimarket con-
threatened.9 tact as a deliberate strategy to elicit cooperation from
their firm’s competitors. Much research in this area as-
7See footnote
sumes or implies that there is a deliberate strategic inten-
4.
8Calculations are based on the Department of Transportation’s
tion in a firm’s use of multimarket contact to ensure
Origin and Destination database. mutual forbearance through increased deterrence of com-
9For additional insights, see Carey (1998), McCartney (1996), petitors and increased familiarity with competitors. The
and O’Brian (1994). question then is whether multimarket contact is an out-

62 / Journal of Marketing, July 1999


come of rational strategic planning by managers or an ac- Data Sources
cidental outcome of marketplace dynamics. Or, is it an in- Much of the empirical research in multimarket competi-
cremental adaptive strategic learning process in which tion has been conducted using secondary data sources.
multimarket contact is both a result of strategic intent and For example, Alexander (1985) and Mester (1987) use
an outcome of random marketplace dynamics? The an- data from the banking industry, and Baum and Korn
swers to these questions may lie in an empirical examina- (1996) and Gimeno and Woo (1996) use data from the
tion of the process that underlies the emergence of mutual airline industry. In these cases, overlap in geographic
forbearance as a result of multimarket contact (Korn and markets is considered the determinant of multimarket
Baum 1999). contact. Other industry-specific data used include data
Although the focus of this article has been multimarket from the telephone industry (e.g., Barnett 1993; Parker
contact, it must be highlighted that multimarket contact is and Roller 1997). Multi-industry data sources used for
only one of the many forces that influence the market be- studying the impact of multimarket contact include Feder-
havior of firms. The role that multimarket contact might al Trade Commission (FTC) Line of Business data (e.g.,
play in shaping the competitive behavior of firms along Feinberg 1985; Scott 1982) and U.K. Census of Produc-
with other influences provides many opportunities for re- tion data (e.g., Hughes and Oughton 1993). The impact of
search. The moderating role that Clark and Montgomery multimarket competition on product line rivalry could be
(1997) observe for multimarket contact in the relationship studied by isolating industries such as the pet food indus-
between competitive reputation and entry deterrence is an try, in which such competition is prevalent, and then lon-
example of one such outcome. In continuing with this gitudinally examining how multimarket competition
stream of research, researchers can examine how multimar- shapes market structure, as well as the nature of rivalry.
ket competition influences entry strategies, entry deterring
strategies, and other competitive market behavior of firms Research Methods
jointly with other factors, such as other market and cost
characteristics. Cross-sectional and longitudinal approaches have been used
Much of the empirical research in multimarket competi- to conduct research in this area. Because the phenomenon of
tion conceptualizes multiple markets along the geographic multimarket competition delineates an impact that is shaped
market dimension. This leaves the empirical aspects of the over time, longitudinal approaches such as panel data meth-
multimarket impact of overlapping product markets, charac- ods (e.g., Evans and Kessides 1994; Gimeno and Woo 1996)
terized by product line rivalry, considerably underre- and event history analysis (e.g., Barnett 1993; Baum and
searched. With regard to the multimarket impact of product Korn 1996) seem more appropriate. Furthermore, as demon-
line rivalry, possible research directions include (1) the im- strated by Clark and Montgomery (1997) and Phillips and
pact of product line similarity on the intensity of competi- Mason (1992), experiments can be used effectively to study
tion between firms, (2) the constraining effects of the impact of multimarket competition in controlled situa-
multimarket competition on the product line strategies of tions. This approach could be particularly valid for examin-
firms, and (3) the extent to which product line strategies are ing the process that underlies the impact of multimarket
used to establish footholds in important markets. contact on intensity of competition. That is, experiments
In the case of multinational enterprises (MNEs), it is could be useful for studying the impact of multimarket con-
possible to envision a firm reacting to being attacked in one tact on deterrence and familiarity. Furthermore, the impact
national market through actions in a different national mar- of multimarket competition on how firms define rivals and
ket. This could be more prevalent for firms that use the rivalry (the extent to which multimarket contact contributes
transnational form of coordination (Bartlett and Ghoshal to the social construction of rivalry within firms) could be
1989). In this form of organization, an MNE coordinates its analyzed using qualitative methods, such as depth inter-
activities on a global basis. Therefore, it is likely that an at- views of the managers involved in making the decisions.
tack in a specific country market would cause an MNE or-
Measurement
ganized transnationally to evaluate the merits of retaliating
in a different country market (Watson 1982). Likewise, The constructs central to the proposed conceptual model are
when contemplating attacking a multimarket rival in a spe- multimarket contact, intensity of competition, spheres of in-
cific country market, an MNE might take into account the fluence, resource similarity, organizational structure, and
country market(s) in which the rival is most likely to retali- seller concentration. In Table 2, we outline alternative oper-
ate. The dearth of empirical research examining multimarket ationalizations and studies that have employed these opera-
competition in the global marketing context provides oppor- tionalizations. A critical issue regarding operationalization
tunities for additional research in this area. of multimarket contact and spheres of influence is the defi-
nition of the market. Typically, research in multimarket con-
tact is conducted in industries in which clear geographic
Methodological Issues delineation of markets is possible (e.g., airline, banking).
Extant research on multimarket competition provides in- However, marketing strategy researchers would be equally
sights into some methodological issues to which marketing interested in multimarket competition in the context of prod-
scholars contemplating empirical research in multimarket uct line rivalry, for which market definition would need to
competition must be sensitive. This section captures these encompass the two dimensions of product markets and geo-
insights.

Multimarket Competition / 63
graphic markets. Therefore, market definition in such cases Conclusion
may benefit from the approach we have used in defining a
In this article, we examine the impact of multimarket com-
geographic-product market.
petition on the competitive behavior of firms. Multimarket
It also should be noted that there are different ways of competition may affect the strategies of firms by influenc-
evaluating intensity of competition. One method is direct ing their ability and motivation to deter competitors, as
observation of competitive attacks and responses by rivals well as by fostering better understanding of competitors. In
(for example, see Chen 1996). Another method focuses on this competitive scenario, rivalry between firms in the mar-
the impact of these actions and reactions on price or perfor- ket is expected to be shaped, in part, by their level of mul-
mance. For example, Gimeno and Woo (1996) use the price timarket contact. The mutual forbearance agreement that
charged by the firm (after controlling for costs) as a measure can emerge effectively will constrain the strategies that
of the intensity of competition. Studying actual actions and firms can use in these circumstances. Even a price promo-
reactions would provide a finer-grained approach to analyz- tion for a brand in a particular market may elicit multimar-
ing intensity of competition. It should be noted, however, ket competitive retaliations if the promotion is considered
that such data are difficult to obtain, and looking at output predatory by the firm’s competitors. An extension of a
measures such as price is often the only practical route to firm’s product line by launch of a variant or an attempt to
operationalizing intensity of competition. Alternative opera- expand the firm’s scope of operations geographically may
tionalizations, however, are possible in the context of survey cause competitors to retaliate against the firm in other mar-
data (for a measure of intensity of competition, see Jawors- kets in which they compete.
ki and Kohli 1993). Resource similarity often is opera- The notion of mutual forbearance and lower rivalry aris-
tionalized using proxies such as firm size and strategic es from the possibility of firms comprehending the multi-
group membership. Researchers may need to move to less faceted nature of such competition and, as a result,
coarse measures of this construct by studying actual re- recognizing tacit collusion as a more viable strategy. There-
sources available to firms (Chen 1996). A more detailed dis- fore, multimarket competition may act as a constraint, re-
cussion of measurement issues is beyond the scope of this ducing the degrees of freedom firms have in implementing
article. strategies. However, such multimarket retaliatory tactics

TABLE 2
Research on Multimarket Competition: Operationalization of Constructs

Construct Measure Representative Research


Multimarket contact Count measure (sum of the number of Baum and Korn (1996); Evans and Kessides
markets in which the firms compete (1994); Feinberg (1985); Gimeno and
outside the focal market) Woo (1996); Mester (1987)

Probabilistic measure Mester (1987); Scott (1982)


Intensity of competition Performance based:
Price-cost margin Feinberg (1985); Gimeno and Woo (1996);
Price levels Heggestad and Rhoades (1978); Hughes
Income/sales and Oughton (1993); Mester (1987)
Market share instability

Action based:
Market entry and exit Barnett (1993); Baum and Korn (1996,
1999)

Spheres of influence Market dominance measures such as Baum and Korn (1996); Gimeno (1999)
market share
Market dependence (percentage of overall Gimeno (1999)
firm revenues from the market)
Resource centrality (the extent to which the Gimeno (1999)
market unit draws on critical firm
resources).

Resource similarity Competitor’s size Baum and Korn (1999)


Strategic group membership Barnett (1993); Gimeno and Woo (1996)

Seller concentration Herfindahl-Hirschman index Gimeno (1994)

Structure of the firm Centralization of decision making for the Jaworski and Kohli (1993)
relevant markets

64 / Journal of Marketing, July 1999


could be objectionable from an antitrust perspective. Al- only the potential for mutual forbearance and, therefore,
though U.S. antitrust explicitly has ignored multimarket re- lower intensity of competition. Thus, mutual forbearance is
actions (focusing only on horizontal dominance of markets), not a deterministic outcome of multimarket contact. For ex-
it is likely that strong multimarket reactions against small ample, General Motors and Ford compete aggressively in
competitors could be interpreted as predatory behavior. most markets, despite substantial multimarket contact. Even
Dodgson, Katsoulacos, and Pryke (1990) prepared a report when conditions are ripe for mutual forbearance between
to the Competition Directorate of the European Commission two firms in a specific market, a third firm can prevent them
(the equivalent of the FTC or Department of Justice) in
from competing less intensely. It is possible that the third
which they discussed multimarket responses in the aviation
market as predatory behavior. Therefore, multimarket reac- firm does not share the incentive that propels the other two
tions could be construed as predatory behavior.10 toward mutual forbearance and may unleash intense compe-
Finally, we do not mean to imply that multimarket con- tition in the market that may prevent or unravel mutual for-
tact always reduces the intensity of competition through mu- bearance among its rivals.11 Regardless, for the reasons
tual forbearance. Akin to the observation made by Gatignon, described in this article, marketing strategy researchers and
Weitz, and Bansal (1990) in reference to seller concentra- marketing managers stand to benefit from considering the
tion—namely, that it only provides the potential for collu- impact of multimarket competition on the competitive dy-
sion—we note that multimarket competition also provides namics between firms.

10We thank a reviewer for drawing this issue to our attention. 11We thank a reviewer for drawing this issue to our attention.

REFERENCES
Abell, D.F. (1980), Defining the Business: The Starting Point of Day, George S. and Prakash Nedungadi (1994), “Managerial Rep-
Strategic Planning. Englewood Cliffs, NJ: Prentice Hall. resentations of Competitive Advantage,” Journal of Marketing,
Alexander, Donald L. (1985), “An Empirical Test of the Mutual 58 (April), 31–44.
Forbearance Hypothesis: The Case of Bank Holding Compa- Dickson, Peter Reid (1992), “Toward a General Theory of Com-
nies,” Southern Journal of Economics, 52, 122–40. petitive Rationality,” Journal of Marketing, 56 (January),
Areeda, Phillip and Donald Turner (1979), “Conglomerate Merg- 69–83.
ers, Extended Interdependence and Effects on Interindustry Dodgson, J.S., Y. Katsoulacos, and R.W.S. Pryke (1990), Predato-
Competition as Grounds for Condemnation,” University of ry Behavior in Aviation. A Report to the Competition Direc-
Pennsylvania Law Review, 127, 1082–103. torate of the European Commission, Commission of the
Barnett, William P. (1993), “Strategic Deterrence Among Multi- European Communities.
point Competitors,” Industrial and Corporate Change, 2, Edwards, Corbin D. (1955), “Conglomerate Bigness as a Source of
249–78. Power,” in Business Concentration and Price Policy, National
Bartlett, Christopher A. and Sumantra Ghoshal (1989), Managing Bureau of Economic Research Conference Report. Princeton,
Across Borders: The Transnational Solution. Boston, MA: Har- NJ: Princeton University Press, 331–52.
vard Business School Press. Evans, William N. and Ioannis Kessides (1994), “Living by the
Baum, Joel A.C. and Helaine J. Korn (1996), “Competitive Dy- ‘Golden Rule’: Multimarket Contact in the U.S. Airline Indus-
namics of Interfirm Rivalry,” Academy of Management Jour- try,” The Quarterly Journal of Economics, 109, 341–66.
nal, 39 (2), 259–91. Feinberg, Robert M. (1985), “Sales-at-Risk: A Test of the Mutual
——— and ——— (1999), “Dynamics of Dyadic Competitive In- Forbearance Theory of Conglomerate Behavior,” Journal of
teraction,” Strategic Management Journal, 20 (3), 251–78.
Business, 58, 225–41.
Bernheim, Douglas and Michael D. Whinston (1990), “Multimar-
Fernandez, Nevea and Pedro L. Marin (1998), “Market Power and
ket Contact and Collusive Behavior,” RAND Journal of Eco-
Multimarket Contact: Some Evidence from the Spanish Hotel
nomics, 21, 1–26.
Industry,” Journal of Industrial Economics, 46, 301–16.
Boeker, Warren, Jerry Goodstein, John Stephan, and Johann Peter
Murmann (1997), “Competition in a Multimarket Environment: Gatignon, Hubert, Barton Weitz, and Pradeep Bansal (1990),
The Case of Market Exit,” Organization Science, 8 (2), 126–42. “Brand Introduction Strategies and Competitive Environ-
Buzzell, Robert D. and Bradley T. Gale (1987), The PIMS Princi- ments,” Journal of Marketing Research, 27 (November),
ples. New York: The Free Press. 390–401.
Carey, Susan (1998), “Aided by Its Shuttle, United Air Is Taking Gimeno, Javier (1994), “Multipoint Competition, Market Rivalry,
Los Angeles by Storm,” The Wall Street Journal, (January and Firm Performance: A Mutual Forbearance Hypothesis in
16), A1. the U.S. Airline Industry, 1984–1988,” doctoral dissertation,
Chen, Ming-Jer (1996), “Competitor Analysis and Interfirm Rival- Krannert Graduate School of Management, Purdue University.
ry: Toward a Theoretical Integration,” Academy of Manage- ——— (1999), “Reciprocal Threats in Multimarket Rivalry: Stak-
ment Review, 21 (1), 100–34. ing Out ‘Spheres of Influence’ in the U.S. Airline Industry,”
——— and Danny Miller (1994), “Competitive Attack, Retalia- Strategic Management Journal, 20 (2), 101–28.
tion, and Performance: An Expectancy–Valence Framework,” ———, Pedro L. Marin, and Carolyn Y. Woo (1998), “How Do
Strategic Management Journal, 15, 85–102. Market Structures Influence Multimarket Competition? An Em-
Clark, Bruce H. and David B. Montgomery (1996), “Perceiving pirical Test of Bernheim and Whinston’s Hypothesis,” working
Competitive Reactions: The Value of Accuracy (and Para- paper, Department of Management, Texas A&M University.
noia),” Marketing Letters, 7 (2), 115–29. ——— and Carolyn Y. Woo (1996), “Hypercompetition in a Mul-
——— and ——— (1997) “Competitive Reputations, Multimar- timarket Environment: The Role of Strategic Similarity and
ket Competition and Entry Deterrence,” Research Paper No. Multimarket Contact on Competitive De-Escalation,” Organi-
1414, Graduate School of Business, Stanford University. zation Science, 7, 322–41.

Multimarket Competition / 65
——— and ——— (1999), “Multimarket Contact, Economies of Ogur, Jonathan D. (1976), “Competition and Market Share Insta-
Scope, and Firm Performance,” Academy of Management Jour- bility,” Staff Report to the Federal Trade Commission. Wash-
nal, 42 (3), 239–59. ington, DC: U.S. Government Printing Office.
Green, Donna H., Donald W. Barclay, and Adrian Ryans (1995), Oster, Sharon (1994), Modern Competitive Advantage. New York:
“Entry Strategy and Long-Term Performance: Conceptualiza- Oxford University Press.
tion and Empirical Examination,” Journal of Marketing, 59 (4), Parker, Philip M. and Lars-Hendrik Roller (1997), “Collusive Con-
1–16. duct in Duopolies: Multimarket Contact and Cross-Ownership
Gruca, Thomas S. and D. Sudarshan (1995), “A Framework for En- in the Mobile Telephone Industry,” RAND Journal of Econom-
try Deterrence Strategy: The Competitive Environment, Choic- ics, 28 (2), 304–22.
es, and Consequences,” Journal of Marketing, 59 (3), 44–55. Parkhe, Arvind (1993), “Strategic Alliance Structuring: A Game
Heggestad, Arnold A. and Stephen A. Rhoades (1978), “Multimar- Theoretic and Transaction Cost Examination of Interfirm Co-
ket Interdependence and Local Market Competition in Bank- operation,” Academy of Management Journal, 36 (August),
ing,” The Review of Economics and Statistics, 60, 523–32. 794–829.
Heide, Jan B. and Anne S. Miner (1992), “The Shadow of the Fu- Phillips, Owen R. and Charles F. Mason (1992), “Mutual Forbear-
ture: Effects of Anticipated Interactions and Frequency of Con- ance in Experimental Conglomerate Markets,” RAND Journal
tact on Buyer–Seller Cooperation,” Academy of Management of Economics, 23, 395–414.
Journal, 35, 265–91. Porac, Joseph F. and Jose Antonio Rosa (1996), “Rivalry, In-
Heil, Oliver and Thomas S. Robertson (1991), “Toward a Theory dustry Models, and the Cognitive Embeddedness of the
of Competitive Market Signaling: A Research Agenda,” Strate- Comparable Firm,” Advances in Strategic Management, 13,
gic Management Journal, 12, 403–18. 363–88.
Hughes, Kirsty and Christine Oughton (1993), “Diversification, ——— and Howard Thomas (1990), “Taxonomic Mental Models
Multimarket Contact and Profitability,” Economica, 60, of Competitor Definition,” Academy of Management Review,
203–24. 15 (2), 224–40.
Jans, Ivette and David I. Rosenbaum (1996), “Multimarket Contact Porter, Michael E. (1980), Competitive Strategy: Techniques for
and Pricing: Evidence from the U.S. Cement Industry,” Inter- Analyzing Industries and Competitors. New York: The Free
national Journal of Industrial Organization, 15, 391–412. Press.
Jaworski, Bernard and Ajay K. Kohli (1993), “Market Orientation: Rao, Vithala R. and Joel H. Steckel (1998), Analysis for Strategic
Antecedents and Consequences,” Journal of Marketing, 57 (Ju- Marketing. New York: Addison Wesley.
ly), 53–70. Rhoades, Stephen A. and Arnold A. Heggestad (1985), “Multimar-
Karnani, Aneel and Birger Wernerfelt (1985), “Multiple Point ket Interdependence and Performance in Banking: Two Tests,”
Competition,” Strategic Management Journal, 6, 87–96. The Antitrust Bulletin, 30, 975–95.
Keller, John J. (1993), “For Years Bell Atlantic Has Been Obsessed Robinson, W. (1988), “Marketing Mix Reactions to Entry,” Mar-
with Being More than a Phone Company,” The Wall Street keting Science, 7 (4), 368–85.
Journal, (October 14), A14. Sandler, Ralph D. (1988), “Market Share Instability in Commercial
Korn, Helaine J. and Joel A.C. Baum (1999), “Chance, Imitative, Airline Markets and the Impact of Deregulation,” Journal of In-
and Strategic Antecedents to Multimarket Contact,” Academy dustrial Economics, 36, 327–35.
of Management Journal, 42 (2), 171–93. Schelling, T.C. (1960), The Strategy of Conflict. Cambridge, MA:
Kotler, Philip (1994), Marketing Management. Englewood Cliffs, Harvard University Press.
NJ: Prentice Hall. Scott, John T. (1982), “Multimarket Contact and Economic Per-
Maxwell, John C. (1986), “Pet Food Category Off Its Feed, but formance,” The Review of Economics and Statistics, 64,
Snacks Gain,” Advertising Age, (August 18), 70. 368–75.
——— (1991), “Specialty Shops Spice Up Sales for Pet Foods,” Simmel, Georg (1950), “Superordination and Subordination and
Advertising Age, (September 16), 43. Degrees of Domination and Freedom,” in The Sociology of
McCartney, Scott (1996), “Competitors Quake as Southwest Air Is Georg Simmel, Kurt H. Wolff, ed. New York: The Free Press,
Set to Invade Northeast,” The Wall Street Journal, (October 268–303.
23), A1–10. Singal, Vijay (1993), “Interdependence Among Firms that Com-
McGrath, Rita G., Ming-Jer Chen, and Ian C. MacMillan (1998), pete in Many Markets: The Airline Industry and its Mergers,”
“Multimarket Maneuvering in Uncertain Spheres of Influence: working paper, Department of Finance, Virginia Polytechnic
Resource Diversion Strategies,” Academy of Management Re- Institute and State University.
view, 23 (4), 724–40. Stuart, Toby (1991), “Cat Fight in the Pet Food Industry,” Case 9-
Mester, Loretta J. (1987), “Multiple Market Contact Between Sav- 391-189, 195, 196, 197. Cambridge, MA: Harvard Business
ings and Loan,” Journal of Money, Credit, and Banking, 19, School.
538–49. Teece, David J., Gary Pisano, and Amy Shuen (1997), “Dynamic
Mullins, John W. and Orville C. Walker (1996), “Competency, Pri- Capabilities and Strategic Management,” Strategic Manage-
or Performance, Opportunity Framing, and Competitive Re- ment Journal, 18 (7), 509–34.
sponse: Exploring Some Behavioral Decision Theory Watson, Craig M. (1982), “Counter-Competition Abroad to Pro-
Perspectives,” Marketing Letters, 7 (2), 147–62. tect Home Markets,” Harvard Business Review, 60 (1),
Nomani, Asra Q. (1990), “Airlines May Be Using a Price-Data 40–42.
Network to Lessen Competition,” The Wall Street Journal, Weick, Karl E. (1995), Sensemaking in Organizations. Thousand
(June 28), A1. Oaks, CA: Sage Publications.
O’Brian, B. (1994), “Southwest Says It Will Add Longer Flights,” Zenor, Michael J. (1994), “The Profit Benefits of Category Man-
The Wall Street Journal, (July 5), A3. agement,” Journal of Marketing Research, 31 (May), 202–13.

66 / Journal of Marketing, July 1999