Strategic Groups Structure , Positioning of the Firm and Performance : A Review of Literature

A central theme of the literature on strategic groups is that group membership affects performance. The basic premise of strategic groups and their main interest is in fact that performance can be attributed to strategic groups and not only to the idiosyncratic character of the individual firm. The aim of this paper is therefore to characterize and synthesize studies that examined the relationship between strategic groups, considered from the economic or cognitive perspectives, and performance. We notice that these studies have often yielded conflicting results. We will focus specifically on the relationship between the positioning of the firm, the strategic group and performance, a link that has hitherto been little studied in the literature. In other words, is there for the firm an optimal positioning in relation to strategic groups in its industry? The few empirical studies focusing on this relationship, emphasizing a cognitive approach, were able to show that firms moderately differentiated with respect to their belonging strategic group are more successful.

A central theme of the literature on strategic groups is that group membership has an impact on performance (Cool & Schendel, 1987;Hodgkinson, 1997a).The basic premise of strategic groups and their main interest are in fact that performance can be attributed to strategic groups and not only to the idiosyncratic character of the individual firm.
There is no formal definition of the concept of strategic group unanimously accepted (Thomas & Venkatraman, 1988: 538).Hunt (1972), who was the first to formulate this concept, observed significant differences in the strategies followed by firms in the U.S. home appliance industry, even though many of them were pursuing similar strategies.Business combinations can therefore, according to this author, facilitate the identification of different types of generic strategies followed.Hunt therefore defined a strategic group as a group of companies within a sector, which are very similar in terms of cost structure, degree of product diversification, formal organization, control systems and perceptions and preferences of individuals.
Subsequently, many other definitions of strategic groups have been proposed.According to Porter (1979: 215): "An industry [...] can be considered as composed of groups of firms, each group consisting of firms pursuing similar strategies regarding key decision variables".The definition of Porter (1980: 129) is certainly the most commonly used: "A strategic group is a group of firms in an industry following a similar or identical strategy regarding relevant dimensions.An industry may have a single strategic group if all firms follow essentially the same strategy.At another extreme, each firm could be a different strategic group.However, there are usually a small number of strategic groups, summarizing the essential strategic differences among firms in an industry."Thus, in general, the concept of strategic group is defined in terms of a group of companies pursuing similar strategies with similar resources (Yami & Benavent, 2000: 3).
The aim of this paper is therefore to compile a literature review on the relationship between performance and strategic groups, showing that different works sometimes led to conflicting results.We will also try to identify new perspectives in this area of research, especially by focusing on the theory of strategic balance formulated by Deephouse (1999).This review of the literature should therefore allow us to propose an empirical test and could deepen the theory proposed by Deephouse.
Two research streams studied largely independently the strategic group concept, the first being derived specifically from the industrial economy and the second from the cognitive approach to strategy.We will see in the first section that these two approaches could be reconciled in order to achieve more conclusive results regarding the link between strategic group and performance.In a second part, we will focus on strategic groups' performance, by describing the factors that explain performance differences between groups and within groups.In the third part, we will analyze the research about the different positions of the firm relative to strategic groups.Finally, in a fourth part we will discuss the theory of strategic balance, which applied to strategic groups helps explain the link between the positioning of the firm relative to the strategic group and performance.

Strategic Groups and Performance: The Interest of Reconciling the Economic and Cognitive Perspectives
Research on strategic groups has not yet succeeded in proving the existence of a genuine link between strategic group membership and performance.While the "objective" approach has produced conflicting results, the cognitive approach insufficiently explored this question.Nowadays, as both economic and cognitive factors are likely to affect performance, a third perspective that integrates these two approaches seems likely to clarify this issue.

"Objective" Strategic Groups and Performance
Numerous studies have focused on strategic groups since the 1970s, some authors having proposed a synthesis of these studies (e.g.Ketchen et al., 2004).A wave of criticism on the "objective" approach to strategic groups is the lack of results demonstrating real performance variations between groups (Mehra & Floyd, 1998: 512;Yami & Benavent, 2000: 2).It was indeed assumed in theory that performance was relatively homogeneous within groups, the greater heterogeneity of performance being found between the various strategic groups (Thomas & Pollock, 1999: 131).Going against this proposal, it was found that firms belonging to the same strategic group could hold very significant performance differences (Fiegenbaum & Thomas, 1990Thomas & Venkatraman, 1988).The performance of the firm would not thereby be explained by its strategic group membership (Barney & Hoskisson, 1990: 187).In addition, performance is generally considered in the studies on strategic groups in the narrow terms of profitability, thus opposing a broader view including financial and operational measures (Peteraf & Shanley, 1997).

Cognitive Strategic Groups and Performance
Very few studies have attempted to link, at the theoretical or empirical levels, cognitive strategic groups with performance.Peteraf and Shanley (1997), however, by proposing an identity theory of cognitive strategic groups, suggested that the homogeneity of perceptions within the same strategic group could have negative and positive effects on intermediate performance variables such as reactivity or strategic cooperation.They argue that such contradictory effects may explain the absence of significant performance differences between cognitive strategic groups.Reger and Huff (1993) for their part, studying the relation between cognitive strategic groups and performance, have found consistency in performance within groups and heterogeneity in performance between groups in the case of U.S. regional banks.Performance was measured by the fact that a bank has remained or not independent five years after the collection of perceptual data for the identification of strategic groups.Reger and Huff (1993: 114) point out, however, that the small size of their sample did not allow them to truly demonstrate the existence of a causal link between strategic group membership and performance.

Strategic Groups and Performance: a Double Lighting
Many authors, recognizing the potential complementarity of the "objective" and cognitive approaches to strategic groups, have therefore recommended to use them simultaneously to link strategic groups with other strategic variables.Bogner and colleagues (1994: 301) estimate that the two approaches to strategic groups lead to the same question for strategy makers: what are the implications of these groups on their members?It would be therefore preferable to avoid this theoretical «dead end" to develop a third model incorporating both paradigms a priori "incompatible" to produce a more comprehensive and relevant model (Bogner & Thomas, 1993).Scott (1981: 173) in a similar perspective suggests that "perception measures are necessary if one wants to predict the choices or behaviors of members of the organization, but they are not sufficient if one wants to predict the result of those choices».This author believes that these two visions of the environment make valid contributions, but on different points, so a complete model should include these two views.
As pointed out by Porac and colleagues (1989), using the theory of "enactment" (Weick, 1979), industry structure determines managerial perceptions and at the same time emerges from them.So there are interactions and feedbacks between economic and cognitive factors.Hanson and Wernerfelt (1989) simultaneously measured the impact of economic and behavioral factors on performance.In their research, a model based on the industrial economy, including the size of the firm and its market share and a model derived from the behavioral theory were integrated, and a significant contribution of each of these two sources on firm performance as a whole was observed.
As numerous studies that focused on "objective" strategic groups have not provided satisfactory results regarding their link to performance, while studies positioned on the cognitive approach have insufficiently explored this question, it therefore seems legitimate to investigate the link between strategic groups and performance considering simultaneously the arguments from both approaches.We nevertheless consider the case where top managers' representations of strategic groups are similar.In such a case, these cognitive strategic groups are close to the "objective" ones.

Strategic Groups Performance
As we noted earlier, many studies, mostly based on "objective" approach to competition, have focused on the link between strategic groups and performance including seeking the concepts of barriers to mobility and competition.However, the results were often contradictory: performance varies or does not vary both within strategic groups and between strategic groups.

Explaining Factors of Performance Differences between Strategic Groups
As the economic or "objective" approach to strategic groups further studied their relation to performance than the cognitive one, we mainly present the arguments and empirical studies of the authors positioned on the first approach.According to them, two interrelated key concepts explain the homogeneity of performance within a strategic group and the heterogeneity of performance among different strategic groups: barriers to mobility and competition.

The Mobility Barriers Concept
The concept of barriers to mobility is an extension of the concept of "barriers to entry" of an industry formulated by Bain (1956).Caves and Porter (1977), and later Porter (1980), have indeed applied this concept to strategic groups, suggesting that just as it exists in an industry barriers to entry, barriers to mobility between groups within an industry also exist.Leask (2004) defines strategic groups as stable structures in an industry separated by barriers to mobility and pursuing different strategies impacting performance.These barriers protect each strategic group from the competition of players located outside the group and thus lead to differences in performance (Short et al. 2007;Tywoniak et al., 2007).This concept implies that a company in a given sector makes decisions that cannot be imitated by other firms outside its strategic group unless they agree on an "increase of their costs, a loss of time and an uncertainty about the outcome of these decisions " (McGee, 1985).The height of barriers to mobility therefore depends on the cost of entering into a strategic group, that is to say non-recoverable investments that any newcomer must make to adopt the strategy applied in the group (Bidault, 1988).Such barriers may result from the actions of individual firms or the collusion between strategic group members (Dranove & Peteraf, 1998).
Investments in both tangible (e.g.plant and equipment) and intangible (e.g.name and reputation) assets and uncertainty about the ability of firms to copy successful competitors (Lippman & Rumelt, 1982) contribute to the construction of these barriers to mobility.The distribution system has been identified as constituting a major barrier in pharmacy and insurance industries (Cool & Schendel, 1987).The establishment of standard operating procedures, bureaucracy and organizational culture are also likely to inhibit the ability of a firm to enter a strategic group.

Intra and Inter Strategic Group Rivalry
Behind the concept of mobility barriers implicitly lies the idea that the level of rivalry within and between different strategic groups differs (Cool & Dierickx, 1993;Peteraf, 1993;Porter, 1979).The concept of mobility barriers associated with the one of collusion implies that the rivalry within strategic groups would be low.The theory of oligopoly (e.g.Stigler, 1964) suggests that there exists in industries coordination mechanisms based on group consensus reducing the rivalry within the group.The intra-group rivalry is low when firms recognize their mutual dependence and cooperate or tacitly collude (Caves & Porter, 1977;Peteraf, 1993;Porter, 1979).In addition, many authors have defined strategic groups based on strategic alliances or inter-firm networks (e.g.Ketchen et al. 1993;Thomas & Carroll, 1994).Duysters and Hagedoorn (1995), in particular, believe that firms with similar resources are likely to have the same suppliers and customers, which facilitates communication and coordination within the strategic group.Lawless and Anderson (1996) concluded that competition was less intense among the most productive firms in a similar ecological niche.Lying on a cognitive approach, Peteraf and Shanley (1997) suggested that, when firms see themselves as belonging to the same strategic group, they benefit from the positive effects of coordination, efficiency and reputation.Yami and Benavent (2000), for their part, believe that strategic groups could be defined not by a principle of resemblance but by a principle of complementarity.Jayachandran and colleagues (1999: 50), however, believe that collusion between firms may be challenged by a major innovation: "companies may be inclined to see the mutual dependence as a preferable alternative to a frenzied rivalry, but they would not reject the possibility to dominate the market with a major innovation."As pointed out by McNamara and colleagues (2003: 167), research using "hard" data or perceptual one showed that rivalry was stronger within strategic groups than between strategic groups.For some authors, competition is actually greater among firms that are similar to each other on key strategic and structural dimensions (Dess & Davis, 1984;McGee & Thomas, 1986;Thomas and Pollock, 1999).
According to Porter (1979), the degree of rivalry within a group depends on its structure.Thus, large groups whose members differ significantly in terms of size and preference for risk have probably a greater internal rivalry than smaller groups.Members of a group may target the same consumers and quickly copy the competitive move of a rival (McGee et al. 1995: 258).According to the resource-based view, a firm that has chosen to differentiate its resources will be less exposed to the effects of competition (Barney, 1991, Hatten & Hatten, 1987, Nelson, 1991;Rumelt, 1984;Wernerfelt, 1984).Gimeno and Woo (1996) found that the similarity in the strategies of airlines was related to greater competition.Cool and Dierickx (1993) have also found a significant intra-group rivalry in the pharmaceutical industry by using historical data over a period of 20 years.
From a cognitive perspective, Fiegenbaum and Thomas (1995) suggest that firms are likely to focus on their competitive position within their group and be more attentive and responsive to the actions of members of their own group than to those of members of other groups.Other authors have shown that managers perceive a greater rivalry within their belonging strategic group (e.g.Hodgkinson & Johnson, 1994;Porac et al., 1995).
The results of Denglos (2002) on the computer industry confirm the explanatory power of strategic groups vis-à-vis competitive behavior.The author stresses that "the number of actions and reactions, the number of actions initiated, the frequency of price decreases and the rapidity of response to competition depends on the strategic group in question" (2002: 155).In all cases, the results do not confirm an attenuation of rivalry within groups, even if the nature of this rivalry differs according to groups.
Dranove and colleagues (1998) noted that competition between firms in different groups may be less vigorous than competition between firms within the same group.Hagedoorn (1993), for his part, believes that market and technology complementarity is a major cause of inter-firm partnerships.Complementarity thus suggests that firms in the same strategic group are not necessarily the most appropriate partners for partnerships, this suggesting a lower rivalry between strategic groups.
However, according to Porter (1980), a strong rivalry between groups occurs when the differences between groups involve heterogeneous resources and varied patterns of competitive behavior, as it makes it difficult to predict and coordinate actions with firms located in other groups.An increase in strategic distance should therefore, all things being equal, lead to more competition between groups than within strategic groups (Porter, 1979).In addition, as membership in certain groups may appear as more desirable because of their high level of performance or their lack of internal competitive rivalry, we can therefore expect greater competition between these groups and smaller groups performance (McGee et al. 1995: 258).Empirically, Peteraf (1993), in the airline industry, has shown that the rivalry between strategic groups was higher than the rivalry within strategic groups.In the case of the pharmaceutical industry over 20 years, Cool and Dierickx (1993) showed that intra-or inter group rivalry was alternately dominant.Denglos (2002) therefore concludes that the supposed superiority of the rivalry between groups with respect to the intra-group rivalry is subject to contradictory results.No theoretical perspective can properly adjudicate this issue very clouded by empirical results (Denglos, 2002: 137).He argues that this lack of convergent results could be primarily attributed to the paradox of the strategic groups theory: competition between groups is assumed higher than the intra-group rivalry, while the groups are formed from mobility barriers reducing competition between groups.He therefore suggested to mitigate the bilateral effect hypothesis of mobility barriers so that a group can be a strong rival for another without the opposite being true.This author also explains the different nature of the findings on inter-and intra-group rivalry by the diversity of the competition measures that have been used.
As theories and results are contradictory regarding the effect of barriers to mobility on rivalry within and between strategic groups, it is therefore somewhat not surprising, as we shall see that the results on the relation between strategic groups and performance have been conflicting.

Inter and Intra Strategic Group Performance
Whereas most research focused on performance differences between strategic groups, few studies have examined the performance differences between firms within the same group.Some studies, simultaneously considering these two questions, sought to determine whether differences in performance in an industry were more important between strategic groups or within strategic groups.

Performance Differences between Strategic Groups
The degree of rivalry is an intermediate variable of the effect of strategic groups on performance (Cool & Dierickx, 1993;Mehra & Floyd, 1998: 521).Barriers to mobility, as they create monopoly rents and promote cooperative practices within the group, allow member firms to achieve a sustainable superior performance (Caves & Porter, 1977;Mascarenhas & Aaker, 1989;Porter, 1980).It is therefore expected that the highest barriers protect the most profitable groups, their members having a higher capacity to prevent imitation by firms outside the group (McGee et al. 1995: 258).In addition, firms in different groups operate in heterogeneous competitive environments with varying munificence and profit potential (McNamara et al. 2003: 165).McGee and Thomas (1986) thus suggest that variations in industry performance can be explained by the homogeneity of firms' performance within groups and the heterogeneity of firm's performance between groups.Denglos (2002: 136) states that achieving higher profits within a group will be dependent upon the fact that this group includes a small number of firms, or six at most, and the existence of barriers to mobility.Mehra and Floyd (1998), meanwhile, believe that significant performance differences between strategic groups only occur in industries characterized by high product-market heterogeneity and inimitable resource configurations, including intangible ones (Wernerfelft, 1989), this situation forming barriers to mobility.
Numerous empirical studies found significant performance differences between strategic groups (e.g.Cool & Schendel, 1987, Dess & Davis, 1984;Dikmen et al., 2009;Hatten, 1974;Kale & Arditi, 2003;Mascarenhas & Aaker, 1989;Nair and Kotha, 2001;Oster, 1982;Schreyögg & Von Reitzenstein).Denglos (2002) found significant performance differences between strategic groups in the computer industry, as companies diversifying or specializing had a statistically different performance.He stressed that the structure of strategic groups has no significant influence on the results of firms: group size is equivalent when profits are significantly different.He therefore concludes that obtaining superior results to the ones obtained in a pure competitive situation is dependent on the existence of barriers to mobility but is independent of the number of firms (Denglos, 2002: 151).
Other authors, however, tended to demonstrate, both theoretically and empirically, that performance differences between strategic groups could not occur.Mehra and Floyd (1998: 513) argue in particular that it is perfectly conceivable that different competitive strategies, implemented in munificent environments, can achieve comparable results.Barney and Hoskisson (1990: 194-195) point out that when the number of firms in each strategic group is high enough to cause a pure and perfect competition, the observation of differences in performance between groups is compromised.
Empirically, many authors have thus failed to establish statistically significant financial performance differences between strategic groups (e.g.Claver et al. 2003;Claver et al. 2006;Lewis & Thomas, 1990;Porter 1979).Ketchen and colleagues (1997) conducted a meta-analysis of 33 studies comprising 40 independent samples.These studies were comparing performance differences between strategic groups.They concluded that there were significant differences in performance between groups, but group membership explained only 8% of these differences.Overall, there is indeed an effect of the strategic group on performance, but it may be small (McNamara et al. 2003: 167).

Variations in Performance within Strategic Groups
As pointed out by McNamara and colleagues (2003: 168), although the possibility of performance differences among groups was demonstrated by Cool and Schendel (1988) and Lawless and colleagues (1989), the theoretical foundations of these performance differences were not examined in depth.Many authors positioning themselves on the perspective of the industrial economy have indeed suggested that performance within strategic groups is homogeneous (McGee & Thomas, 1986).
However, the resource-based view (Hatten & Hatten, 1987;Rumelt, 1984) and the contestable markets theory (Baumol et al. 1982;Cool & Dierickx, 1993, Hatten & Hatten, 1987) suggest that some firms in a group can develop strategic positions based on a unique product-market configuration or unique resources.Some resources may indeed be specific to the firm, while others are shared by all members of the strategic group (Mehra & Floyd, 1998: 517).Rumelt (1984), for his part, suggested that there exist isolation mechanisms limiting the equalization of profits within strategic groups.Some empirical studies following the resource-based view have noted differences in performance between firms within the same group (e.g.Cool & Schendel, 1987, 1988, Lawless et al., 1989).Nair and Kotha (2001), for example, found that the variables located at both the firm and group levels have significant effects on performance.Claver and colleagues (2006) found differences in performance between hotels in Alicante following the same strategy.McNamara and colleagues (2003), for their part, have found little evidence of consistency of performance within cognitive strategic groups.

The Superiority of Performance Gaps within and between Groups
Some authors have studied simultaneously the performance gaps within and between strategic groups.According to Denglos (2002: 138), the existence of performance differences within strategic groups in particular can explain why some authors have noted the absence of significant performance differences between strategic groups, the differences in resources within groups may be indeed more important than those between groups.Denglos(2002: 153), on the basis of his results, points out that "without questioning the existence of mechanisms limiting ex post the equality of profits within groups, which enlightens the role of asymmetric resources, performance variance between groups is greater than the one within groups".Mehra and Floyd (1998: 525), meanwhile, speculate that in industries characterized by low heterogeneity of products and markets and imitable resource configurations, differences in economic performance of firms will be higher within the groups that between them.In all cases, as pointed Denglos (2002: 138), "the relationship between group membership and performance is still an open question.No consensus emerged, both at the theoretical or empirical levels."However, it seems that literature, focusing primarily on performance differences between strategic groups since assuming homogeneity of performance within groups, is looking more and more to any performance differences within groups.
Literature, thus tending to take more into account any performance differences within strategic groups, is analyzing therefore the positioning of firms in relation to strategic groups.In other words, firms within the same strategic group may have differences in resources and performance.

Strategic Groups and Positioning of the Firm
Before deciding what type of positioning of the firm in respect to strategic group is associated with optimal performance, it is necessary to consider the different types of articulation that may exist between the positioning of the firm and strategic groups.

On the Existence of an Inter and Intra Strategic Groups Heterogeneity
Many researchers from different theoretical perspectives have suggested that strategic groups were not homogeneous in terms of both the number of companies composing them and the degree of similarity of their members.Thus, for example, Fiegenbaum and Thomas (1993) have identified six strategic groups in the insurance industry over the period 1983-84.The largest group consisted of 19 members while the other two groups included only 6 and 3 members.They also identified three solitary firms occupying unique strategic positions.Similarly, Mascarenhas (1989) identified a multiple groups structure in the refining industry: a group contained 80% of firms in the industry whereas a firm had a unique strategic position.Cool and Schendel (1987) and Reger and Huff (1993) also identified strategic groups consisting of a single firm.
Regarding the internal structure of strategic groups, some authors from both the economic and cognitive perspectives suggested that there were core and peripheral firms in a strategic group (Caves & Porter, 1977;Claver-Cortes, 2006;Desarbo & Grewal, 2007;Dharwadkar & Grewal, 2002;Ketchen et al. 1993;Peteraf & Shanley, 1997).Porac and Thomas (1990), for example, believe that industries are fuzzy sets of firms that differ in their degree of correspondence to the "ideal" of their belonging categories.Managers in the study by Reger and Huff (1993) on U.S. banks have identified a significant number of firms following the basic strategy of their strategic group, while a smaller number were following this strategy to a lesser degree, although being considered members of the group.According to Reger and Huff (1993: 114), a cognitive strategic group will tend to be composed of core members and a less distinct periphery.Firms located on the periphery of the strategic group (called secondary firms), although similar on many key strategic dimensions to the ones located in the heart of strategic group (called central firms), are also taking unique strategic decisions.Strategic group membership is therefore a matter of degree (Reger & Huff, 1993: 114).These authors also state that strategic groups may cross each other, some companies belonging to two groups or more simultaneously.

Explaining Factors of the Heterogeneity of Strategic Groups Structure
As pointed out by McNamara and colleagues (2003: 165), researchers on strategic groups have used at least three theoretical perspectives to explain internal differences in a group: the cognitive perspective (e.g.Reger & Huff, 1993), the resource-based view (e.g.Wernerfelt, 1984) and the isolation mechanisms perspective (e.g.Rumelt, 1984).As the resource-based and isolation mechanisms views have already been mentioned when presenting the studies on the relation between strategic groups and performance, we focus specifically on the cognitive perspective and the arguments developed by Reger and Huff (1993): -The properties of cognitive categorization: a firm sharing many common points with its group will probably be frequently seen as a good example of this category (Porac & Thomas, 1990) and thus be considered a core firm of its strategic group.
-The degree of cognitive identification with the strategic group: Peteraf and Shanley (1997), using the concept of organizational identity (Albert & Whetten, 1985) argued that members of a strategic group might differ in their degree of identification with their strategic group.
-The industry realignment: in the case of an industry being in a transitional phase calling into question the key strategic dimensions, some observers may emphasize the positions of firms prior to this transition, while others may anticipate their future positions (Reger & Huff, 1993).
-Obscure strategies: as firms may have an interest in concealing their intentions to gain a competitive advantage derived from the effect of strategic surprise (Porter, 1980), the strategies of some companies, intentionally or not, are difficult to interpret and may be characterized in different ways by observers (Reger & Huff, 1993).
-Firms changing strategy: even during the most stable periods of an industry evolution, the strategies of some firms may be in a situation of flux, the observers then focusing on the longstanding, current or anticipated position of the firm.
-Firms with inconsistent strategies: some firms may not follow clear-cut and stable strategies over time, such firms being described as "stuck in the middle" (Porter, 1980) or "reactive" (Miles & Snow, 1978).
-Firms with new strategies: such firms have a consistent strategy but do not fit into the cognitive map that generally describes all the other competitors in the industry.

The Different Types of Positioning of Firms Regarding Their Strategic Group
Following mostly the arguments of Reger and Huff (1993: 117), it is therefore possible to distinguish different cases depending on the position of firms in relation to strategic groups: -"Solitary" firms: they are the only members of a strategic group, this group being defined according to the dimensions normally used to define strategic groups within the industry.
-Firms central to their strategic group: they are closely related to their strategic group and define its core strategy.
-Firms in the periphery of their strategic group: also called secondary firms, they have a strategy that is not exactly similar to the one of central firms.
-"Defector firms: they are undergoing a transition from a strategic position to another one along dimensions that are common to other firms in the industry.
-"Inconsistent" firms: they have no clear strategy over a long period of time and frequently change their strategy according to circumstances.
-"Idiosyncratic" firms: they have a strategy that can be described as new because they cannot be characterized based on the dimensions used to explain the strategies of most other firms in the industry.
-Firms belonging to several strategic groups: they share certain characteristics with a strategic group and others with another group.
Having shown that strategic groups are characterized by internal heterogeneity in terms of performance, the number of firms composing them and the positioning of their member firms, it is therefore necessary to examine the performance of the firm based on its position in its strategic group.

Firm Positioning Regarding Strategic Groups and Performance
Authors are mainly interested in three of the cases previously presented to characterize firm performance based on its position in relation to strategic groups: core firms, secondary firms and "solitary" firms.The issue of performance has therefore been addressed in terms of similarity or difference in relation to other firms in the industry, some authors, using conflicting theoretical perspectives, stressing that performance comes from being similar (case of core firms that very similar to other firms in their strategic group) or from being different (case of "solitary" firms that are different from other firms in the industry as being the only ones to pursue a certain strategy).A perspective integrating difference and similarity, called theory of strategic balance and mostly developed by Deephouse (1999), stresses that secondary firms (or moderately similar or moderately different) would be the most successful ones in an industry.

Performance Related to the Fact of Being Different
Many theoretical perspectives highlight the benefits associated with having a unique strategic positioning.Thus, the theory of contestable markets (Baumol et al., 1982) applied to strategic groups (e.g.Cool & Dierickx, 1993, Hatten & Hatten, 1987) highlights the value of being strategically different, highly similar firms facing more rivalry.
According to the resource-based view, a firm that takes advantage from differentiating its resources will be less exposed to the effects of competition (Barney, 1991, Hannan et al. 1990;Hatten & Hatten, 1987;Henderson, 1981;Porter, 1991;Rumelt, 1984;Wernerfelt, 1984) and can then expect a higher profit level.The market is indeed supposed to have a finite level of resources at a certain point of time, these resources being then divided among the firms competing for the same strategic positions (Baum & Mezi, 1992;Baum & Singh, 1994).Solitary firms would seek resource configurations associated with unique and inimitable niche-market products to achieve a sustainable competitive advantage and benefit from local monopolies (Barney, 1991;Peteraf, 1993;Wernerfelt, 1984).Peteraf and Schanley (1997) note that firms located at the heart of their strategic group tend to be more resistant to change and to have a myopic vision of the industry because of their strong identification with the group.These core companies, as being more visible, should be more frequently identified as competitors by members of their strategic group and then face more rivalry (Porac and colleagues, 1995).These theoretical perspectives suggest that "solitary" firms, that is to say not belonging to a strategic group made up of several members, should be more successful than central or secondary firms to their strategic groups, secondary firms being at the same time more successful than central firms (McNamara et al. 2003: 169).

Performance Related to the Fact of Being the Same
Several arguments and theoretical perspectives, particularly institutional-theory through the concept of legitimacy, suggest that the most similar firms are the most efficient.
According to Reger and Huff (1993), the deviation from the heart of the strategic group may reflect the inability of a firm to implement the strategy of the strategic group.Peteraf and Shanley (1997), for their part, have proposed that firms who identify strongly with their strategic group will be more effective to exchange information, transfer resources and work collectively for their mutual benefit.Oligopoly theory applied to strategic groups suggests that firms that identify strongly with the group recognize their interdependence and act together more effectively to create competitive barriers (Caves & Porter, 1977;Reger & Huff, 1993).McNamara and colleagues (2003: 169), meanwhile, noted that solitary firms are vulnerable to competitive actions taken against them by strategic groups composed of several members.
The main benefit for the firm from being similar to others that have been stressed by institutional theory or organizational isomorphism is related to legitimacy (Chen & Hambrick, 1995, DiMaggio & Powell, 1983;Hybels, 1995;Meyer & Rowan, 1977).This greater legitimacy allows firms to acquire resources on preferential terms from suppliers and customers for at least three reasons (Pfeffer & Salancik, 1978;Suchman, 1995).First, potential exchange partners are more willing to interact with firms whose strategies are easily understood or seen as rational (McNamara et al., 2003).Second, exchange partners are likely to offer more favorable terms and higher quality products to legitimate firms, the legitimacy of a business partner being reinforced by the fact of contracting with a legitimate firm (Deephouse, 1999;Peffer & Salancik, 1978;Wood, 1991).Third, exchange partners may require additional risk premiums from less legitimate firms because of their greater likelihood of failure (Baum & Oliver, 1991, DiMaggio & Powell, 1983;Singh et al., 1986).An organization that does not pay attention to its legitimacy could therefore disappear faster than its competitors that have gained outside support necessary for their survival (Singh et al., 1986).Empirical research has thus indirectly supported the proposition that conformity strategy leads to higher performance thanks to a strong legitimacy (Abrahamson & Hegeman, 1994, Chen & Hambrick, 1995;Deephouse, 1999, Miller & Chen, 1995).Some authors have applied institutional theory and especially the concept of legitimacy to strategic groups.Peteraf and Shanley (1997) estimate that strategic group membership promotes the legitimacy of individual firms.Firms described by Reger and Huff (1993) as "unclassifiable" and "idiosyncratic" do not benefit from gains in legitimacy due to belonging to a group: even if they begin a move towards a particular strategic group, they will not capture the benefits of legitimacy as they will not become sufficiently similar to group members to be recognized as part of this group (Deephouse, 1999).
Taken together, these perspectives involving strategic similarity and performance imply that firms central to strategic groups should be more successful than secondary and solitary firms and secondary firms should be more successful than solitary firms (McNamara et al., 2003: 168).

The Theory of Strategic Balance
Deephouse (1999), developing the theory of strategic balance and then applying it to the case of cognitive strategic groups (McNamara et al., 2003), integrates into a single theory the conflicting perspectives outlined above, namely that the firm performance is associated with similarity or difference.In established markets whose competitive and institutional forces are high, the proposals for differentiation and similarity would both be significant according to this author.The two streams are then synthesized by Deephouse in a proposal on strategic balance, which states that moderately differentiated firms perform better than firms that are similar to other ones or that are very different.Performance would then result from the balance between similarity and difference, companies for their strategic decisions arbitrating in a perspective of intermediate differentiation between the benefits of increased legitimacy from being more similar to their rivals and the benefits of reduced competition from being different.Because members of the organizational field do not perceive or are indifferent to some degree of differentiation, firms may be different from their competitors to some degree while maintaining their legitimacy.Deephouse (1999) has therefore emphasized in the research perspectives he identified that the validity of the theory of strategic balance can measured regarding strategic groups.A firm belonging to a strategic group should according to him try to distinguish itself from other firms in this group to reduce the intensity of competition it faces.However a firm, to maintain its legitimacy, should not adopt a strategy that deviates from the range of strategies acceptable to its group.McNamara and colleagues (2003), following the research perspectives identified by Deephouse (1999), have therefore studied the performance of core, secondary and solitary firms in the case of U.S. banks located in the same urban area.These authors found results consistent with the theory of strategic balance.Indeed, secondary banks had superior financial performance compared with solitary or central banks, the central banks belonging to the same group as the secondary firm or belonging to a separate group.These authors also showed that the strategic intentions of the firms have maintained or increased over time the strategic differences between core and secondary firms.Thus, it appears that secondary firms expect to maintain their unique position whereas core firms have no intention of imitating them.This may then result in stable differences in performance levels between them.
Many authors have suggested that the strategic group has an effect on performance, mostly through the effect of mobility barriers and competition.However, if this issue was mainly addressed by the authors positioned on the "objective" approach to strategic groups, without leading to conflicting results, the results obtained by researchers clearly positioned on the cognitive approach (Reger & Huff, 1993), or seeking to integrate the two approaches (McNamara et al., 2003) seem to lead to concluding results.The results of the first studies on the relationship between the positioning of firms, strategic groups and performance, in particular, seem promising.Consistent with the theory of strategic balance, they tend to show that the most successful firms are those located on the outskirts of strategic groups as they simultaneously receive the benefits of legitimacy and differentiation.Nevertheless, the performance of some firms identified by Reger and Huff (1993), for example, as firms belonging simultaneously to several strategic groups, has not yet been studied empirically.As highlighted by Deephouse (1999), these firms could benefit from the legitimacy of several groups while facing less competition from their members.

Conclusions and Future Research Directions
To conclude, the performance of firms based on their position in relation to strategic groups should be tested in different contexts.In the case of the banking sector, for example, the institutional environment is strong (Scott, 1995) which may explain the need for a minimum of legitimacy to be successful (McNamara et al. 2003: 178).However, the methodology for identifying strategic groups, whether based on the economic paradigm, the cognitive paradigm or a synthesis of both, is likely to have a very strong influence on the research results.It may therefore seem questionable to compare the results of studies that used different methodologies to identify strategic groups.It would therefore be desirable in future work to use, in one study, several different methods for constructing groups, which is rarely done.
In addition, performance is generally treated in the studies on strategic groups in the narrow terms of profitability, thus opposing a broader view including financial and operational measures (McGee et al. 1995;Peteraf & Shanley, 1997).It therefore seems important that multiple measures of performance are incorporated in the analysis of strategic groups.
In all cases, we believe that research on the relation between strategic groups and performance, although it led to contradictory results, should not be abandoned but rather focus on models that incorporate cognitive and economic perspectives.Empirical studies taking into account the model of Bogner and Thomas (1993), in particular, have not yet been conducted.It would also be necessary to identify more precisely in what conditions an industry tends to favor a certain type of relation between strategic groups and performance.
Other studies examining the performance of firms based on their position regarding strategic groups, for instance by considering the different cases identified by Reger and Huff (1993), should also be conducted in different contexts of industries.We propose, in continuation of the study of McNamara and colleagues (2003), to conduct an empirical study measuring the performance of firms with distinct positions in relation to strategic groups.Such a study would allow us at first to generalize the results found by MacNamara and colleagues (2003), namely that the most successful firms are the ones located at the periphery of their strategic group.Second, this study could allow us to characterize the performance of firms that are not taken into account by these authors, namely "defector", inconsistent and idiosyncratic firms, and the ones belonging to several strategic groups.This will require therefore to incorporate a dynamic perspective to the theory of strategic balance of Deephouse, performance being also considered regarding the evolution of the position of firms regarding their strategic groups.Theoretical developments should also enable us to make assumptions about the type of positioning of the firm in respect to strategic group and performance.We also believe that taking into account specific strategic groups (number of firms within the group, size, degree of rivalry, etc.) should allow us to improve the theory of Deephouse, allowing us to distinguish differences in performance between firms that share a similar positioning in the strategic group.Such a study would also require to complete and to improve the tools used by McNamara and colleagues (2003) to measure the distance from the heart of a strategic group.It is quite possible indeed that the results obtained by McNamara and colleagues (2003) depend heavily on the way the distance between firms was measured in this study.