Introduction
The phenomenon of multimarket competition (MMC) has received considerable attention in the last 15 years, including within the context of strategic management research. Jayachandran, Gimeno, and Varadarajan (Reference Jayachandran, Gimeno and Varadarajan1999) define the concept as competitive situations arising from inter-firm contact across multiple regions or markets. Introducing the concept of MMC to the management literature, Karnani and Wernerfelt (Reference Karnani and Wernerfelt1985) describe how this competitive structure leads to reduced inter-firm rivalry through a process called mutual forbearance. Mutual forbearance represents a form of tacit collusion (Gross & Holahan, Reference Gross and Holahan2003) that occurs as firms’ managers adopt a policy of restraint toward their multimarket rivals, expecting similar restraint in return from these rivals. This mutual restraint leads to better financial performance for multimarket firms by reducing potentially excessive costs resulting from direct, aggressive competition (Gimeno & Woo, Reference Gimeno and Woo1996; Guedri & McGuire, Reference Guedri and McGuire2011). The potential for significant performance improvement gives firms a substantial incentive to practice mutual forbearance once the levels of multimarket contact between them permit.
These studies have generally assumed that mutual forbearance will be adhered to by firms (and their managers) once the level of inter-firm contact across markets increases to a certain point (Jayachandran, Gimeno, & Varadarajan, Reference Jayachandran, Gimeno and Varadarajan1999; Stephan & Boeker, Reference Stephan and Boeker2001). Some studies have shown, however, that this is not always the case. One study found that when firms are engaged in exploration of new markets, rather than exploitation of familiar ones, they engage in competitive activity toward multimarket rivals even when their positioning vis-à-vis these rivals suggests that mutual forbearance should prevail (Anand, Mesquita, & Vassolo, Reference Anand, Mesquita and Vassolo2009). A second showed that the presence of a new CEO increases the likelihood that a firm will not behave toward its multimarket rivals in ways consistent with mutual forbearance (Stephan, Murmann, Boeker, & Goodstein, Reference Stephan, Murmann, Boeker and Goodstein2003). To address this lack of adherence to multimarket theory prescriptions, some work has indicated that in addition to contact with multimarket rivals, firms may need other conditions to exist for mutual forbearance to be sustainable. For example, Prince and Simon (Reference Prince and Simon2009) found that multimarket contacts across less-competitive markets, based on concentration levels, were more likely to produce mutual forbearance-type effects than contacts in more competitive (i.e., less concentrated) markets.
Taken together, this work represents important initial steps toward a more complete description of the conditions under which multimarket ties are likely to have their hypothesized effects. However, to date multimarket researchers have not undertaken a thorough examination of the range of factors that might affect whether and to what degree multimarket contact produces the expected mutual forbearance behavior between firms. Instead, this work has adopted an incremental approach that looks at potential factors individually. This one-at-a-time technique potentially obscures the underlying dynamics associated with multimarket environments and the behavior of multimarket firms and their managers. Given the significant financial penalties a firm might suffer through violations of mutual forbearance, more fully understanding the circumstances that are likely to reduce the incidence of such behavior seems warranted. This paper addresses this omission with a thorough examination of such conditions to inform both future theory-building efforts and empirical examinations.
Our approach builds on a game-theoretic view of MMC, to identify two key elements that underline the process: first, that firms (and their managers) understand the nature of the outcomes that particular competitive actions and combinations of actions are likely to yield; and second, that firms have an incentive to cooperate with each other rather than compete to produce the best overall outcome. This view points to the importance of understanding how (and to what extent) managers become aware of the possible outcomes of the actions and responses of their firms and their multimarket rivals as well as the requirements that are necessary for a firm to behave cooperatively with its rivals.
We then build a model that identifies characteristics drawn from environmental, organizational and individual levels that assist managers to better understand the nature of the outcomes that interaction with multimarket rivals may produce and those that have the capacity to encourage cooperation between firms. These characteristics provide the perceptual lens through which managers view the firm's environment (Seeck & Kantola, Reference Seeck and Kantola2009), which can help them more accurately gauge the potential outcomes of multimarket interactions and understand the necessity of cooperative behaviors (Heliot & Riley, Reference Heliot and Riley2010). The factors we identify along with the firm's multimarket positioning set the context in which competitive decisions are made (Seeck & Kantola, Reference Seeck and Kantola2009) and contribute to managers opting to abide by the norms of mutual forbearance to avoid potential negative financial outcomes. These three contextual settings provide separate motivations and enabling mechanisms for sustaining mutual forbearance and are consistent with recent work that has pointed to context as a critical factor affecting the nature of multimarket effects (Anand, Mesquita, & Vassolo, Reference Anand, Mesquita and Vassolo2009; Guedri & McGuire, Reference Guedri and McGuire2011).
Our approach expands on the existing literature devoted to understanding the nature of competitive interaction between firms. Much of this work has focused on the specific type and characteristics of a firm's action(s) for their impact on rivals’ motivation and/or ability to respond to an attack (e.g., Macmillan, McCaffery, & Van Wijk, Reference MacMillan, McCaffery and Van Wijk1985; Chen & Miller, Reference Chen and Miller1994). We go further by including the motivational effects of competitive positioning via multimarket contact on competitive interaction and develop the logic for how important contextual factors associated with individual managers and firms affect the likelihood of competitive behavior by firms within a multimarket environment – in this case the likelihood that mutual forbearance will occur when multimarket contact levels are sufficient to sustain it. Consequently, our model contributes to the strategy field by broadening the scope of competitive interaction research to include firm positioning while enhancing MMC theory through a broader investigation of the sources and conditions necessary for mutual forbearance to emerge. The recognition that the relationship between mutual forbearance and organizational performance may depend on other additional factors highlights the fact that firm strategy cannot be undertaken in a piecemeal fashion (Hambrick & Fredrickson, Reference Hambrick and Fredrickson2001). Similarly, since an organization's positioning vis-à-vis its multimarket rivals is not a guarantee that it will act in ways consistent with that position (e.g., Stephan et al., Reference Stephan, Murmann, Boeker and Goodstein2003), understanding the range of factors that are potentially involved is a crucial step in MMC theory development. Identifying the circumstances whereby competition receives primary attention and those where it does not will help to more fully understand the nature of interactions between multimarket firms. Our model also generates additional avenues for future research, including the examination of how the interaction between these types of influences affect competitive behaviors and, ultimately, firm performance.
From a practical perspective, our framework can complement and expand managers’ strategic knowledge base and provide additional insight into key influences on firm performance and how managers determine what actions to direct their firms to take to achieve better performance. Our model provides some helpful guidelines to assist managers in, for example, personnel selection, organizational structuring and strategy formulation to improve the chances that competitive environments will be correctly perceived regarding their potential for mutual forbearance between firms and the role that intra- and inter-firm cooperation plays in realizing this reduced level of rivalry. More accurate perceptions of a firm's environment should lead to a decrease in the likelihood of unintentional violations of mutual forbearance by multimarket rivals.
MMC and game theory
The underlying rationale explaining why multimarket contact leads to mutual forbearance is illustrated in Figure 1, which shows a simplified case of two firms competing across two shared markets. In the case where firms compete in only one common market, an attack can only be responded to in the market where it occurred (the left side of Figure 1). The right side of Figure 1 depicts firms competing in multiple markets and shows an initial attack by firm A in market 2. Firm B now has several response options – it can respond in market 2 (labeled response R-1), which is identical to the single market case, or it can respond in market 1 (R-2), which is larger and more important to firm A and thus response within this market could be more costly to firm A than its initial attack on market 2. A third option is to respond across all shared markets (in this case, markets 1 and 2; R-3), which raises the costs of the initial attack substantially. The magnitude and complexity of the outcomes, of course, grows as the number of shared markets increases. These outcomes point to the conclusion that within an MMC environment, initiating competitive attack usually results in poor outcomes.
Multimarket competition versus single market competition
Game theory has been used in a number of multimarket studies where authors have relied on its underlying dynamics to help explain the interaction between multimarket rivals depicted in Figure 1 (e.g., Gimeno, Reference Gimeno1999; Jayachandran, Gimeno, & Varadarajan, Reference Jayachandran, Gimeno and Varadarajan1999). Applications of game theory to multimarket situations reveal two key similarities that identify the drivers of mutual forbearance by multimarket firms. Explaining the game theoretic approach, Schelling (Reference Schelling1960) described how a firm's (or player's) choice of an appropriate action is dependent on what it expects its rival(s) [or opponent(s)] to do in response. This dependence forms the essence of a strategic game with two key elements – the moves of each player are dependent on the choices all players make and the potential outcomes produced by the choices that are made.
Multimarket studies link mutual forbearance to these two elements of game theory in the following way. They assume that multimarket rivals are capable of understanding the nature of the outcomes that particular actions or combinations of actions (action-response pairs, for example) will yield and also that multimarket firms, under certain combinations of actions and outcomes, will have an incentive to cooperate rather than compete to produce the best overall outcome for all firms. If multimarket firms know that tacit cooperation (mutual forbearance) produces the best overall outcome and that cheating or violating such agreements will produce ruinous results, mutual forbearance should result as it yields the best overall outcome for each firm. That is, if the possible outcomes are properly perceived by competing multimarket firms, they should adopt a mutual forbearance approach to each other and be less likely to violate this arrangement once it is established. The underlying assumptions of both game theory and the multimarket studies are thus quite similar. Both assume that players (multimarket firms) know what the ruinous results might be to some degree of certainty and that players are willing and able to cooperate with each other to avoid them.
A simple example illustrates this similarity and reveals a critical assumption that may not exist when real-world multimarket firms are considered. In game theory studies, players are told what the various payoffs (outcomes) are from all of the possible combinations of moves they can jointly make. Figure 2 presents a typical version of what these outcomes might be for a game involving two players who have two possible actions to choose between (attack/respond to an attack or refrain from action)Footnote 1. The four cells of each scenario in Figure 2 represent four possible combinations of actions that the two players can take. The pair of numbers in each cell indicates the outcome or payoff that the players will receive if that combination of actions occurs. The first value is the outcome to player 1 and the second is the outcome to player 2 for each pair of moves.
Outcome possibilities for a simple two-player game
Figure 2 depicts the mutual forbearance situation where the best move for both players is to refrain from competing (cell I). If either chose to attack and the other responded, both would end up suffering large losses (cell IV). This is the potential outcome that should move players toward the cooperative option of cell I. Cells II and III depict the situation where the initial attacker anticipates no response to an attack (and none occurs) and thus a large payoff (with some potential loss for the other) results. These can also be seen as representing a firm that attempts to ‘cheat’ on mutual forbearance agreements. Empirical findings have shown that mutual forbearance can yield a range of desirable outcomes including being able to charge higher prices (Gimeno & Woo, Reference Gimeno and Woo1996; Jans & Rosenbaum, Reference Jans and Rosenbaum1996). A multimarket firm may conclude that cheating will increase its own outcomes along these lines at the expense of its rivals (cells II and III in Figure 1). As Chen (Reference Chen1996) notes, however, multimarket rivals are unlikely to let an initial attack go without response and the response options available to rivals make the initial attack quite costly (yielding cell IV of Figure 1). Consequently, cells II and III represent a best-case scenario for outcomes from cheating on mutual forbearance but also represent a real threat to maintaining mutual forbearance. However, if rivals respond, the outcomes shift to cell IV. Since multimarket firms are likely to respond to such provocation, cells II and III are less likely to appear in multimarket circumstances. The tension between cheating and cooperating forms the essence of the strategic choice in both the game and between multimarket rivals. All of a firm's businesses across all of its shared markets must be in agreement and acting consistently with a tacit agreement to cooperate to sustain mutual forbearance. This requires that managers of these businesses have access to and correctly interpret competitive intelligence.
In real-world competition between firms, however, managers are not told what these outcomes might be – they must deduce these for themselves along with what their rivals’ moves are likely to be. This requires an assessment of their rivals’ capabilities and motivation to act (Chen, Reference Chen1996) and requires observation of a rivals’ behavior over time in a variety of situations (Stephan & Boeker, Reference Stephan and Boeker2001). Thus, managers of multimarket firms need to be able to accurately understand the nature of the joint outcomes produced by their actions and those of their multimarket rivals. These circumstances are not guaranteed simply because a multimarket framework exists between firms, as emerging results have shown (Stephan et al., Reference Stephan, Murmann, Boeker and Goodstein2003; Anand, Mesquita, & Vassolo, Reference Anand, Mesquita and Vassolo2009; Guedri & McGuire, Reference Guedri and McGuire2011). Consequently, understanding the range of factors that enable managers and their firms to be able to correctly determine the negative outcomes that are likely to result from attacking multimarket rivals, and those which facilitate cooperation across the markets shared with multimarket rivals, is of considerable importance to permit firms to realize the performance enhancements that mutual forbearance yields.
Our model
Our model, shown in Figure 3, contains one set of factors that potentially affects managers’ perception of the outcomes that result from aggressive attacks upon multimarket rivals – that is, factors that increase the salience of negative outcomes and/or the costs associated with violating mutual forbearance. To the extent that these factors make the negative consequences of violating such agreements more apparent to a firm's CEO and managers, they are likely to lead the firm to act in ways that are less likely to violate the tacit arrangements underlying mutual forbearance. These factors are likely to have a cumulative effect. That is, their effect on facilitating mutual forbearance should increase over time. This notion is consistent with the fact that multimarket rivals do not interact (or have the potential to interact) with each other only once, but engage each other many times over the course of their relationship. As a result, managers develop expectations for their rivals’ behavior – the so-called ‘rules of the game’ – over time and not through a one-shot encounter (Saloner, Reference Saloner1991).
Theoretical model
The second element drawn from game theory is the ability of multimarket firms to tacitly agree to refrain from attacking each other or attempting to cheat on such agreements. Figure 3 identifies a second set of factors that focuses on enhancing a firm's willingness to tacitly cooperate with its multimarket rivals.
Since our primary focus is on whether and to what degree firms recognize and act on their multimarket positioning to agree to mutual forbearance with rivals and the additional conditions that help facilitate this, the existing positioning of the firm is a critical factor. Stephan and Boeker (Reference Stephan and Boeker2001) describe how not all levels of multimarket contact are sufficient to induce mutual forbearance. They argue that only after some level of multimarket contact is achieved between firms (an inflection point, i.e., where rewards from mutual forbearance equal or are greater than the losses incurred from forbearance violation) will mutual forbearance be strategically prudent. Since our model presumes that mutual forbearance is the appropriate decision for the firm, we correspondingly assume that this point of inflection has been reached and therefore the opportunity for mutual forbearance exists. We begin by discussing organizational-level factors because these permit us to lay out some of the key dynamics to sustaining mutual forbearance that are then referenced in following sections. We then move on to industry/market characteristics and conclude with individual issues.
Factors promoting cooperative firm behavior
Organizational: Firm's prior experience with explicit agreements to cooperate
In order to maintain a regime of mutual forbearance among rivals, multimarket firm mangers must feel they are able to produce and maintain the levels of tacit coordination necessary to reduce aggressive competition among multimarket rivals and count on their rivals to do the same. Mutual forbearance agreements lack written and/or verbal contracts and do not provide legal recourse should a firm renege on its tacit agreements. The tacit collusion at the heart of mutual forbearance is effected through the behaviors, signals and specific cues firms use to indicate their intentions to other firms (Barney & Hesterley, Reference Barney and Hesterley1996). The performance incentives to remain committed to such tacit agreements, noted earlier, provide additional motivation to abide by these agreements. However, there is always a risk that some firms will cheat and attempt to increase their own performance at the expense of their multimarket rivals. Because of this possibility, the incentives to remain committed to the relationship are ultimately psychologically embedded and implicit (Turner, Reference Turner1987). The behaviors and cues that support tacit collusion must be communicated within each multimarket firm to all businesses in the firm that compete in shared markets and they must be interpreted consistently by the managers of these businesses in order for the firm to coordinate action (or inaction) across these business in their markets to sustain mutual forbearance. A firm's (and its manager's) prior experience with a range of inter-organizational arrangements may help build an ability to successfully manage this information and sustain these implicit relationships.
Ring and Van de Ven (Reference Ring and Van de Ven1994) describe generally how implicit relationships emerge between firms. They note that most inter-organizational relationships begin as explicit, formal agreements. Over time, these evolve into informal, psychologically based relationships via an institutionalization process that builds on the increased familiarity, socialization and social capital facilitation that develops as these relationships mature (Adler & Kwon, Reference Adler and Kwon2002). Therefore, the extent to which a firm has greater experience with formal agreements between itself and other firms can translate, over time, into greater facility with and commitment to informal, psychological contracting (i.e., tacit arrangements), whereby perceptions and trust replace laws and direct monitoring as a deterrent to opportunism. Wu (Reference Wu2008) notes how repeated interaction between firms can lead to higher levels of trust. Similar findings have emerged from management scholars examining a variety of organizational relationships including alliances and joint ventures (Gulati, Reference Gulati1998) and vendor–customer relationships along a firm's supply chain (Heide & Miner, Reference Heide and Miner1992).
Following this logic, we propose that firms with prior successful experiences in explicit cooperative agreements will be more likely to mutually forbear and remain committed to this forbearance when embedded in a network of multimarket contacts. The extent to which a firm has experienced this evolution from relationships based on explicit agreements to those grounded in implicit understandings is likely to make it more adept at the internalization process (Nonaka, Reference Nonaka1994) that is required for psychological constructs such as trust to engender inter-organizational cooperation. In addition, it is likely to make a firm more able to recognize the nuances of such a relationship without misinterpretation that helps to build trust between the partners. Although the lack of formal governance and policing may provide an avenue for parties to mutual forbearance arrangements to cheat, the extent to which the firms recognize the psychological ties between them and what they represent, acts to counter this influence:
Proposition 1a : Multimarket firms with more experience with direct explicit cooperation (e.g., alliances) will be more likely to exhibit mutual forbearance when their multimarket positions support it.
Organizational: Top management team (TMT) homogeneity
In order for firms to effectively sustain mutual forbearance across multiple shared markets, joint decision making by multimarket managers within a firm is required (Haveman & Nonnemaker, Reference Haveman and Nonnemaker2000). The ability to take coordinated action across a number of markets requires that the managers responsible for these business units, which may be structurally independent, clearly understand the need for joint action and be willing and able to perhaps sacrifice their own immediate goals for those of the overall firm. The same prerequisite holds with respect to refraining from competitive engagement with multimarket rivals. Thus, the degree to which decision makers are able to reach agreement on what course(s) of action the firm will take jointly is a crucial component of a firm's ability both to present a credible response threat and effectively engage in mutual forbearance. The makeup of the decision-making group and the experiences it brings to the task are of critical importance in this regard. Consequently, TMT-related characteristics that affect the likelihood of agreement are essential antecedents to this agreement. This outcome is dependent on the team's inclination to share information among the members of the team and its ability to reach consensus on what action(s) to take or avoid.
Hambrick and Mason (Reference Hambrick and Mason1984) proposed that homogeneity among TMTs increases strategic decision-making efficiency of firms by decreasing the incidence of destructive conflict and communication gaps. Glick, Miller, and Huber (Reference Glick, Miller and Huber1993) note that less diversity on TMTs can improve strategic planning efficiency because homogenous teams are able to reach resolution on particular strategic issues more readily. Recent TMT studies have explained this by using the concept of behavioral integration. In particular, TMTs with high levels of behavioral integration exchange information in greater quantities and of better quality and as a result are better able to collaborate in decision making and are more likely to develop organization-wide strategies. Simsek, Veiga, Lubatkin, and Dino (Reference Simsek, Veiga, Lubatkin and Dino2005) argued that higher levels of diversity among TMTs leads to lower levels of behavioral integration as rapport among team members drops and less information is exchanged between members. In addition, they described how greater diversity is likely to hinder the team's chances for mutual understanding, making jointly reached decisions harder to achieve (Simsek et al., Reference Simsek, Veiga, Lubatkin and Dino2005). In order to openly exchange information, team members must be able to trust each other that ideas will not be unduly criticized when raised (Olson, Parayitam, & Bao, Reference Olson, Parayitam and Bao2007). This is more likely to occur when team diversity levels are lower. Consequently, more diverse, less behaviorally integrated teams, are less likely to share the information that taking joint action across many business units requires.
Homogeneity is also likely to permit top managers to reach consensus on a course of action more quickly. When teams are more heterogeneous, they exhibit increased cognitive diversity (i.e., differences in the preferences and beliefs of individuals) as well (Miller, Burke, & Glick, Reference Miller, Burke and Glick1998). Increased cognitive diversity has been found to bring more varied ideas into decision-making processes, but it negatively affects group cohesion (Condon & Crano, Reference Condon and Crano1988) making it harder for TMTs to reach consensus on strategic issues (Finkelstein & Hambrick, Reference Finkelstein and Hambrick1996). In addition, diversity can result in the formation of sub-groups that can become polarized if the differences between the groups loom larger than those within them (Rico, Molleman, Sanchez-Manzanares, & Van der Vegt, Reference Rico, Molleman, Sánchez-Manzanares and Van der Vegt2007). The emergence of such polarized sub-groups makes reaching consensus more difficult.
Pfeffer (Reference Pfeffer1983) traced this ability to reach consensus to the idea that cohesion produces a common perspective by facilitating the creation of shared organizational schemas. Consequently, factors that contribute to group cohesion and the sharing of organizationally based schemas can impact the degree to which particular organizational actions occur. Since mutual forbearance relies on a firm-wide agreement not to aggressively engage the firms’ multimarket rivals in any shared market, the possession of an agreed-upon schema supporting this approach facilitates forbearance. More diverse teams are likely to possess different schemas and consequently be less likely to agree on the benefits or the possibility of mutual forbearance and thus the actions the firm should take (Simsek et al., Reference Simsek, Veiga, Lubatkin and Dino2005; Olson, Parayitam, & Bao, Reference Olson, Parayitam and Bao2007), all of which lessen the likelihood that joint action consistent with mutual forbearance will result:
Proposition 1b : Firms with more homogenous TMTs will be more likely to exhibit mutual forbearance when their multimarket positions support it.
Environmental: Differences across shared markets/industries
Studies of MMC have almost exclusively conducted empirical tests of MMC phenomena using firms that compete in single industries, with their multimarket contact being based on geographically defined markets within a single country. For example, Baum and Korn (Reference Baum and Korn1996, Reference Baum and Korn1999) used the California airline industry, Haveman and Nonnemaker (Reference Haveman and Nonnemaker2000) studied California banking firms, Barnett (Reference Barnett1993) utilized US PBX switching systems while Jans and Rosenbaum (Reference Jans and Rosenbaum1996) investigated the US cement industry. One exception was the use of hospitals in California by Stephan et al. (Reference Stephan, Murmann, Boeker and Goodstein2003), where they defined markets as different therapeutic classes of medical treatment. While these were not geographically based markets, they still are more similar to each other than they would be to many other types of unrelated markets or industries that could be shared by firms.
It remains an open question whether multimarket effects will emerge among firms that have contact with each other across significantly different industries or markets or across geographic regions that span radically different cultural norms – that is, when the characteristics of the industries and the competencies necessary for success in them are very different from one another or when cultural norms do not closely align. Yet, many firms find themselves in such positions. For example, Procter and Gamble competes with Colgate–Palmolive across several different markets including toothpaste (Crest vs. Colgate), dishwashing liquid (Palmolive vs. Dawn and Joy), deodorant (Secret vs. Speedstick) and cleansers (Comet vs. Ajax), among others.
From an international perspective, General Motors competes with Toyota across the US, European and Asian markets. These cultures differ dramatically along many dimensions. For example, Hofstede's (Reference Hofstede1979) description of the differences between South Korea, France and the USA reveals significant cultural inconsistencies. On the individualistic/collectivist dimension, South Korea is far more collectivist, the USA is more individualist and France somewhere in between (Itim International, 2011). These kinds of differences are likely to have implications for a firm's ability to easily coordinate actions across these markets when attempting to address multimarket concerns.
When multimarket contact between these types of firms exists, the organizational challenges are especially strong regarding the ability to manage coordinated decision making and ensure the priority of organizational-wide goals (Gimeno & Woo, Reference Gimeno and Woo1999). When multimarket firms meet across very different types of industries, the business units within each firm are apt to be part of a corporate strategy of unrelated diversification (e.g., Rumelt, Reference Rumelt1982) and consequently the interaction between them is quite limited. More importantly, the activities and objectives across these units can and will differ markedly and their experience with coordinating actions across businesses is slim. These conditions are likely to make coordination for competitive purposes quite difficult as these business units are unlikely to share similar perspectives or schemas on their competitive positions (Pfeffer, Reference Pfeffer1983). An analogous situation may emerge when business units span diverse cultures. Conversely, when firms have contact with rivals across more similar markets and/or industries or similar cultures, their ability to engage in tacit cooperation with their rivals and ensure internal coordination across all business units is greater as the impediments to shared understanding and agreement are less, as existing results have indicated (e.g., Gimeno & Woo, Reference Gimeno and Woo1996):
Proposition 2 : Multimarket contact between firms across or within industries that are more similar will be more likely to exhibit mutual forbearance when their multimarket positions support it.
Individual: Personal disposition to cooperate
Hambrick and Mason (Reference Hambrick and Mason1984) laid out the underlying rationale for why top managers’ predispositions are likely to have a significant effect on the actions taken by the organizations they lead. With respect to mutual forbearance, the degree to which the CEO of a firm possesses a personal disposition to cooperate is likely to affect the kinds of actions the firm takes when positioned in a multimarket environment. Numerous studies have described how an individual's predisposition to cooperate can result in a cooperative culture within an organization and for cooperative interaction among organizational members to dominate over competition. Chatman and Barsade (Reference Chatman and Barsade1995) noted that an individual's propensity to cooperate can exert both an independent and an interactive effect with the organization's culture to produce cooperative behaviors within the organization. Ling, Zhao, and Baron (Reference Ling, Zhao and Baron2007) find that a greater tendency of CEOs to behave in cooperative ways results in more involvement in decision-making situations across the organization, higher levels of commitment to decisions and higher levels of motivation to implement the decisions being made by those involved.
CEOs with a preference for cooperation over competition within their organizations are more likely to exploit the avenues within an existing organizational culture and structure that support a cooperative and firm-wide posture toward the firm's rivals, rather than a market-by-market approach (Golden & Ma, Reference Golden and Ma2003). Such CEOs are more likely to ensure that information essential to cooperation within the organization is acquired and shared among internal organizational members (Ling, Zhao, & Baron, Reference Ling, Zhao and Baron2007) to facilitate a coordinated response to multimarket rivals across all shared markets. A CEO who favors cooperation is likely to rely on devices such as built-in lateral relationships, open information channels and team-based reward systems (Rozemeijer, van Weele, & Weggeman, Reference Rozemeijer, van Weele and Weggeman2003) that address corporate-level objectives to reinforce firm-wide (rather than business-level) views of multimarket rivals and to facilitate firm-level responses to engage with multimarket rivals. A CEO's own tendency to take advantage of organizational structures that enhance joint problem solving and action would assure that all managers are on the same page with respect to how aggressive the firm will be toward its multimarket rivals across the markets they share (Nowak, Lewenstein, & Szamrej, Reference Nowak, Lewenstein and Szamrej1993):
Proposition 3a : Firms with a CEO with a personal disposition to cooperate will be more likely to exhibit mutual forbearance behaviors when their multimarket positions support it.
Individual: CEO experience in team-oriented organizations
In addition to a CEO's own predisposition to cooperate with others, the environment that she or he finds himself or herself in can also influence behavior (Bandura, Reference Bandura1977), depending on the accessibility and saliency of specific experiences (Dutton, Dukerich, & Harquail, Reference Dutton, Dukerich and Harquail1994). That is, the context in which managers find themselves is a conduit through which expectations, or norms regarding (among other things) which environmental elements deserve the most attention, are transmitted (Heliot & Riley, Reference Heliot and Riley2010). After exposure to organizational expectations and norms, managers eventually adopt them as their own (or, in some cases, choose to leave the organization – i.e., they self-select out).
Organizations, as carriers of prescribed norms, guide and facilitate the attitudes and beliefs of organizational members through the social processes that occur within them (Ashkanasy, Wilderom, & Peterson, Reference Ashkanasy, Wilderom and Peterson2000). In line with this, we suggest that experience in organizational environments centered on teamwork shape CEO perceptions of the efficacy of cooperation and team-based arrangements (Chatman & Barsade, Reference Chatman and Barsade1995) and increase the likelihood that they will encourage individual cooperation and promote team-oriented activity by providing guidelines for appropriate behavior (O'Reilly & Chatman, Reference O'Reilly and Chatman1996).
Experience in cooperative environments is expected to positively shape attitudes about cooperation, which in turn can make managers more willing to exchange information, pursue joint goals and assist others when necessary in addition to setting these standards for others. This willingness directly affects the ability of a firm to coordinate its approach to its rivals across all affected business units. Having experienced cooperation on an internal basis, it is reasonable to expect this experience of success to affect the likelihood of and expectations of the success of such behavior with respect to external entities. As familiarity with cooperative organizational norms and institutions increases, individuals are more likely to identify with these norms and adopt corresponding attitudes and behaviors. In addition, since those involved in cooperative endeavors are likely to experience some successful outcomes, there is also a greater likelihood that cooperation between the firm and its multimarket rivals will be seen as more feasible which would lead the firm to be more willing to sustain mutual forbearance. Consequently, cooperation is less likely to be seen as an inferior approach to direct competitive confrontation. In summary, CEOs who have worked and succeeded in team environments will be influenced by the cooperative norms inherent in these environments and be more willing to pursue cooperative behaviors toward their multimarket rivals:
Proposition 3b : Firms with CEOs with experience in team environments will be more likely to exhibit mutual forbearance behaviors when their multimarket positions support it.
Factors increasing the salience of negative outcomes
Organizational: Past experience with forbearance violation
Reinforcement theory (Skinner, Reference Skinner1969) and theories of social power (Raven, Reference Raven1992) argue that when contingent rewards and punishments are attached to specific behaviors, they are capable of effectively controlling them. In particular, punishments that result from deliberate action increase individuals’ motivation to avoid such negative outcomes in the future by refraining from behaving in ways that caused the outcomes. Reinforcement theory has shown that the greater the number of contingent punishments issued for particular proscribed behavior, the less likely is the behavior to be repeated in the future (Skinner, Reference Skinner1969). Trevino and Ball (Reference Trevino and Ball1992) note that it need not be necessary for an actor to experience the punishment and negative outcome directly to produce a reduction in the behavior that gave rise to these consequences. Consequently, both direct and indirect experience (Bandura, Reference Bandura1977) of the negative outcomes of mutual forbearance violations can reduce the incidence of such violations in the future, although we expect that direct experience will have a stronger effect (Fazio & Zanna, Reference Fazio and Zanna1981).
From a multimarket perspective, this means that multimarket firms that have had first-hand experience with negative performance impacts from engaging multimarket rivals when mutual forbearance should have been practiced are more likely to avoid such encounters in the future:
Proposition 4a : Firms that have experienced greater numbers of violations of mutual forbearance will be more likely to exhibit mutual forbearance behaviors when their multimarket positions support it.
In addition to the effect of experiencing greater numbers of incidents of forbearance violation, the magnitude of a violation's consequences may also exert an independent effect. When violations of tacit agreements to refrain from aggressive conduct toward each other produce more significant performance declines for an aggressor, firms are more likely to recognize this connection and refrain from repeating such behavior in the future:
Proposition 4b : Firms that have experienced greater performance decreases due to violations of mutual forbearance and will be more likely to exhibit mutual forbearance behaviors when their multimarket positions support it.
Finally, more recent violations of mutual forbearance are likely to exert a disproportionate influence on subsequent firm behavior. The saliency of the negative impact of forbearance violation is likely to be greatest when these violations have occurred in the recent past (Lord & Hanges, Reference Lord and Hanges1987). We therefore posit that the likelihood of forbearance violation will decrease as the recency of information about the negative outcomes of violation increases:
Proposition 4c : Firms that have more recently experienced violations of mutual forbearance will be more likely to exhibit mutual forbearance behaviors when their multimarket positions support it.
Environmental: Industry/market uncertainty
The degree of predictability or uncertainty in a firm's environment has long been a focus of investigation with respect to many organizational outcomes. Highly uncertain environments are those that are less stable (i.e., they are changing rapidly, a condition known as dynamism) and more complex (have more factors that are critical to firm success; Duncan, Reference Duncan1972; Dess & Beard, Reference Dess and Beard1984). These highly uncertain environments make reaching strategic consensus among top managers about the causes of firm outcomes and the actions that a firm should take to respond to its environment, much more difficult (Homburg, Krohmer, & Workman, Reference Homburg, Krohmer and Workman1999).
In order to correctly comprehend what the consequences of firm actions are regarding multimarket rivals, multimarket firms must be able to link expected outcomes to specific actions taken by themselves and their rivals. This becomes harder to do when environments are more uncertain. First, top managers are less likely to reach consensus on what the relationship is between firm actions and outcomes (Homburg, Krohmer, & Workman, Reference Homburg, Krohmer and Workman1999). Consequently, they will be less likely to agree on what kinds of outcomes resulted from actions regarding multimarket rivals and the magnitude of these effects. Managers have been shown to construct biased explanations for firm performance when the environment provides uncertain cues as to the actual causal relationships (Wagner & Gooding, Reference Wagner and Gooding1997). That is, when explaining firm outcomes that they are responsible for, managers tend to attribute good results to their firm and the actions it has taken and poor ones to environmental factors, despite evidence to the contrary. In unpredictable environments, managers are more likely to settle on inaccurate interpretations of events.
Because of this, when multimarket settings also have high levels of environmental uncertainty, it is likely to be harder for managers of multimarket firms to accurately identify the outcomes that are linked with specific organizational actions. This inability to correctly identify outcomes may make it more difficult for them to establish and abide by agreements to refrain from high levels of competition with their multimarket rivals. If they do not see the link between firm actions and outcomes as producing the best overall results with cooperative behavior between firms, managers are unlikely to refrain from additional competitive attacks on their rivals.
Some circumstantial evidence exists to show that managers of multimarket firms in more uncertain environments are less likely to cooperate with their multimarket rivals when their competitive position would suggest this is the appropriate action to take. Anand, Mesquita, and Vassolo (Reference Anand, Mesquita and Vassolo2009) found that mutual forbearance was less likely when firms competed in markets characterized as explorative – that is, with which the firm had less experience and where environmental changes were more frequent. Similarly, Roy and Prescott (Reference Roy and Prescott2002) found that mutual forbearance was more likely when multimarket firms were competing in environments where the nature of the innovations was more readily understandable by managers and thus less uncertain compared with environments that were characterized by novel, difficult-to-imitate innovations:
Proposition 5 : Multimarket contact between firms in less uncertain (more stable, less complex) industries will be more likely to exhibit mutual forbearance when their multimarket positions support it.
Individual: CEO Job Tenure
While CEO tenure has not been specifically studied in relationship to MMC, Stephan et al. (Reference Stephan, Murmann, Boeker and Goodstein2003) introduce the concept of CEO succession as an influence on a firm's mutual forbearance behavior. They argued that new CEOs are more likely to engage their firms in actions that break away from the firm's status quo, a position consistent with previous research which observed that a change in CEO is often a catalyst for strategic change (Tushman, Virany, & Romanelli Reference Tushman, Virany and Romanelli1985). In a multimarket context, Stephan et al. (Reference Stephan, Murmann, Boeker and Goodstein2003) find that this is likely to lead such newer CEOs to violate mutual forbearance regimes that have been honored by the firm in the past. Longer job tenure inherently increases the likelihood that CEOs not only participated in the firm's decision to mutually forebear vis-à-vis its multimarket rivals, but also personally created at least some of the multimarket ties that facilitate this tactic.
Many scholars have noted that managers, particularly where they are responsible for particular organizational outcomes, are predisposed to persist with past policies and actions (e.g., Miller, Reference Miller1991). The likelihood of such persistence increases when organizational and personal beliefs are highly correlated (Miller, Burke, & Glick, Reference Miller, Burke and Glick1998) leading CEOs who are personally associated with their organizations and its outcomes to become personally resistant to change (Salancik, Reference Salancik1977). These tendencies are most likely to occur when past actions have produced positive outcomes (Miller, Reference Miller1991). When firms competing in multimarket structures benefit from the adoption of mutual forbearance (Gimeno & Woo, Reference Gimeno and Woo1996; Baum & Korn, Reference Baum and Korn1999), a strong precedent is established and managers are more likely to continue with this cooperative arrangement. The longer a particular CEO occupies the top position and benefits from this arrangement, the less likely is he or she to deviate from it.
A second characteristic that accompanies longer tenure is likely to work in concert with strategic persistence to make firms with longer-tenured CEOs less likely to violate mutual forbearance agreements. Gupta and Govindarajan (Reference Gupta and Govindarajan1986) propose that as top managers enjoy longer tenures, their familiarity with organizational practices and people increases. Essentially, top managers become central players in organizational networks and longer tenure not only increases centrality but promotes greater awareness of it. From a central position, CEOs act both as information acquirers and communication conduits. They are therefore both the informed and the informant. With this centrality comes a greater ability to pass along information that communicates the kinds of outcomes mutual forbearance can produce. Since longer-tenured CEOs are likely to be more aware of the nature of their firm's competitive environment and the competitive ramifications of particular courses of action, when they pass along this information to others in the firm, the likelihood of misinterpreting potential outcomes is likely to decrease:
Proposition 6a : Firms with CEOs with longer job tenures will be more likely to exhibit mutual forbearance behaviors when their multimarket positions support it.
Individual: CEO Industry Tenure
Managers can also acquire information about particular firms, their rivals and competitive positioning from general experience in their industry. It is often common for CEOs to have gained information and experience about their organization's industry before job attainment (whether from inside the firm or externally). Since multimarket ties essentially embed an organization in a network of relations with other firms, the managers of these other firms are also in a position to understand the dimensions of this network (Gnyawali & Madhavan, Reference Gnyawali and Madhavan2001). To the extent that particular multimarket structures are confined to individual industries, managers can acquire, through general experience in the industry, familiarity with these external networks and a firm's embeddedness within these inter-firm frameworks. Consequently, CEOs with longer tenures in their respective industries are likely to possess knowledge about their firm's network position and the kinds of outcomes generated by various combinations of firm actions, even if they have only recently stepped into the CEO position with a specific firm. This knowledge will affect competitive orientations (Baum & Dutton, Reference Baum and Dutton1996) and should cause firms to compete less aggressively with multimarket rivals when contact levels rise to the point where mutual forbearance is possible to institute and sustain (Gnyawali & Madhavan, Reference Gnyawali and Madhavan2001).
In addition to exposure to more general industry characteristics, longer industry experience enables managers to build informal networks within the organizations they belong to, creating social capital (Adler & Kwon, Reference Adler and Kwon2002). This social capital can act as a link between firms, increasing familiarity among competitors and more importantly, exchanging the information that will permit managers to understand the potential outcomes of competitive engagement and forbearance more accurately:
Proposition 6b : Firms with CEOs with longer tenures in the industry will be more likely to exhibit mutual forbearance behaviors when their multimarket positions support it.
DISCUSSION
Early research on MMC focused on conceptualizing the phenomena and discovering the inter-firm linkages that supported MMC and mutual forbearance (Karnani & Wernerfelt, Reference Karnani and Wernerfelt1985). Recent contributions have begun to move beyond this narrow approach and introduce various environmental, organizational and individual factors that could affect the MMC–mutual forbearance connection (Prince & Simon, Reference Prince and Simon2009; Guedri & McGuire, Reference Guedri and McGuire2011). This research trend has demonstrated the importance of considering multiple factors that can impact the emergence, efficacy and durability of tacit collusion in a multimarket environment.
The current paper builds on this work by identifying and discussing a number of factors that accentuate the likelihood that CEOs and top managers, and by extension their firms, will remain committed to mutual forbearance. Drawing on game theory concepts, two categories were identified: (1) factors that increase the salience of the negative outcomes and costs of violating tacit agreements of mutual forbearance and (2) factors that promote cooperative behavior by firms. We proposed that both aversion to financial losses due to violation of tacit cooperation as well as willingness to cooperate both at the intra-firm and inter-firm level will increase the likelihood of firm commitment to mutual forbearance when the firm's multimarket position permits such tacit agreements.
By framing the issue from the perspective of managerial decision making in a competitive environment, our model allows these factors to be treated as influences on personal and firm action. This treatment positions the CEO as a vital element in strategy formulation (Hambrick & Mason, Reference Hambrick and Mason1984) and therefore introduces concepts such as individual predispositions and experiences as relevant to the study of MMC. It also recognizes that because mutual forbearance requires firm-wide cooperation and coordination, other top managers play influential roles as well. Further, we discuss how characteristics of the industries and/or markets that firms share as well as the relative stability of these environments may influence a firm's propensity to forbear. Consequently, it offers the opportunity to expand our knowledge of when and how the competition matters to individual firms.
Academics and practitioners alike benefit from refinements to MMC theory. Incorporating individual, industry and organizational factors within an overarching framework treats firms as multi-dimensional and recognizes the role each tier plays in explaining firm-level behavior. Consideration of these dimensions acknowledges the interdependency between firm strategy and environmental realities while positioning the CEO and top managers as critical integrating actors of internal and external elements.
Organizations seeking to sustain performance through mutual forbearance may make use of the discussion presented here as a general guideline to creating organizational institutions (e.g., cultures, reward systems, formal intra- and inter-firm arrangements and coordinating mechanisms) that promote cooperative behavior. Personal and group characteristics such as cooperative dispositions, industry experience and shared team tenure can help firms embedded in multimarket networks to select and maintain the personnel that are most likely to help the firm create and sustain mutual forbearance. In addition, firms might be well advised to devote additional attention to the strategic choices made to enter into particular markets or industries, as such moves may increase the difficulty of clarifying the outcomes of competitive interaction that could make sustaining mutual forbearance more difficult.
Although we tended to focus on the key role of the CEO in developing our propositions, we did indicate that the TMT also played a crucial role in a firm's decision regarding whether or not to mutually forbear. Proposition 2b directly addressed the role of TMT homogeneity regarding the ability of top management to reach consensus on a course of mutual forbearance and carry it out. Clearly, several other of our propositions that focused on the CEO are applicable to the TMT as well. For example, TMT tenure and experience in the industry is likely to lead the team members to be much more aware of the multimarket structures in the industry and the firm's own position. We elected to concentrate on the CEO as the key organizational actor.
By describing the effects of these factors on firms with multimarket contacts, this paper has put the multimarket construct into a much broader organizational context, thus making it possible for organizations to think more comprehensively about ways to leverage their multimarket position and what additional issues may arise while doing so.
Our model draws on game theory but incorporates the fact that with real-world competitors, a number of standard characteristics of game theory models and the way they are tested need modification. Although game theory studies frequently yield interesting and unexpected results (Camerer, Reference Camerer1991), the theory typically assumes that players know the outcomes that various moves and combinations of moves produce and that players are willing and able to cooperate if that situation produces the best outcomes. Our approach recognizes that managers in multimarket environments learn the key elements of competing with their rivals over time – that is, the rules of the game including what the likely outcomes are and what actions rivals are likely to take, are accurately perceived and understood only through repeated interactions with a firm's rivals (Saloner, Reference Saloner1991). This learning process is a key element in whether multimarket firms can sustain mutual forbearance among them. Thus, there is a necessary temporal dimension to effective multimarket interaction. A key challenge for managers of multimarket firms is to correctly assess where in the learning process their firm and its rivals current stand as to potential outcomes and rival's tendencies for competitive interaction. The possibilities for additional research along these lines are rich. For example, studies can seek to understand how this process of learning the rules develops, both within individual multimarket firms and across multiple rivals.
Our model posits influence from three primary sources – individual, organizational and environmental (Hambrick & Finkelstein, Reference Hambrick and Finkelstein1987). Our discussion did not attempt to assign primacy to any one level over the others. That is, we did not argue that, for example, individual level factors have a stronger influence regarding whether managers direct their firms to abide by mutual forbearance than those associated with organizational or environmental sources. It is worth remembering that theories of the spatial geometry of social influence (e.g., Nowak, Lewenstein, & Szamrej, Reference Nowak, Lewenstein and Szamrej1993) have discussed how the relative impact of such influences is a function of spatial distance. From this perspective, because of its immediacy, self-influence is likely to be most salient, followed by factors from the organization and then the environment. Incorporating this hierarchical arrangement went beyond the scope of the current paper, but represents an important and potentially revealing opportunity for future theorizing and research. Therefore, while the factors we outline here may all lead to greater likelihood of mutual forbearance by a firm, it is likely that some factors are more influential than others. It may be the case that in multimarket firms, the environment plays a more influential role because environmental turbulence may be particularly salient (Roy & Prescott, Reference Roy and Prescott2002). In addition, the possibility of investigating whether particular types of influencers (e.g., organizational or individual) complement or oppose each other would permit researchers to address inter-factor relationships and the complexity of decision-making and strategic outcomes associated with multimarket environments.
Our model also treats each factor directly, relating it in a linear fashion to its effect on the likelihood that a firm will act in ways consistent with mutual forbearance. We did not model interactions between these factors, nor consider non-linearities, which are increasingly used to describe multimarket phenomena (Stephan et al., Reference Stephan, Murmann, Boeker and Goodstein2003; Guedri & McGuire, Reference Guedri and McGuire2011). For example, while we presented CEO job tenure and industry tenure as independent, direct effects on the firm's tendency to engage in mutual forbearance, it is possible that they also interact whereby firms with CEOs with extensive industry experience and tenure in the CEO position exhibit the greatest tendency to act in ways consistent with a mutual forbearance regime in place. Alternatively, changes in one factor may lead to changes in others, with a resultant change in firm behaviors. For example, reward systems can motivate employees to share information, thereby coordinating strategic efforts toward firm-wide mutual forbearance, which in turn facilitates a cooperative, team-based environment that may potentially increase CEO proclivity to participate in cooperative arrangements.
Finally, the effects of factors we describe on managers’ (and their firms’) tendencies to abide by mutual forbearance agreements are likely to be especially relevant for multimarket structures that span different industries (as Proposition 1 notes) or different national cultures. In the latter case, tendencies toward cooperative behaviors are likely to differ between cultures (Hofstede, Reference Hofstede1979). This means that firms from such cultures need to be aware that the tendency to cooperate is not uniformly present in firms or business units from other countries or cultures. Expectations for cooperative behavior from firms based in more individualistic cultures may take longer to yield the tacit agreement mutual forbearance represents. In addition, firms that have multimarket contacts spanning such different cultures may require stronger and more publicly communicated norms for cooperation, supported by more prominent organizational mechanisms that encourage cooperation. If these are lacking, the different orientations toward cooperation among subsidiaries in different countries may make both internal firm coordination and coordination with multimarket rivals more difficult.
CONCLUSION
These shortcomings, though constricting, do provide substantial avenues for future development and research. Incorporating interaction effects between factors increases the complexity of the model but may more realistically represent the means by which mangers and firms make decisions within multimarket environments. Relaxing the constraint of environment stability can also allow reevaluation of these factors, thereby possibly modifying their saliency, efficacy and potentially their individual effects.
As presented, our model attempts to flesh out some of the boundary conditions associated with multimarket effects due to contact between the same firms across multiple markets. Because of the incidence of multimarket contact between actual firms, being able to provide some guidance to both academics and practitioners regarding when and how multimarket contact influences managers’ competitive tendencies is vitally important. It can help firms better understand the possible causes of rivals’ actions, which may minimize unintended competitive escalation. Although our model is directed toward multimarket environments, it may also shed light on competitive interactions more generally
Acknowledgements
The authors would like to thank the numerous reviewers for their constructive advice and knowledge.
