Matching response to competitors' moves under asymmetric market strength

Journal of Business Research 82 (2018) 202–212

Contents lists available at ScienceDirect

Journal of Business Research journal homepage: www.elsevier.com/locate/jbusres

Matching response to competitors' moves under asymmetric market strength a

MARK

b,⁎

Kai-Yu Hsieh , Eunjung (E.J.) Hyun a b

Waseda Business School, Waseda University, 1-6-1 Nishiwaseda, Shinjuku-ku, Tokyo 169-8050, Japan College of Business Administration, Hongik University, 94 Wowsan-Ro, Mapo-gu, Seoul, South Korea

A R T I C L E I N F O

A B S T R A C T

Keywords: Asymmetric market strength Competitive strategy Entry behavior Matching response

This paper investigates a firm's inclination to match a rival's strategic move under asymmetric market strength. Drawing from the awareness-motivation-capability framework, we theorize that a firm is more likely to match modestly weaker competitors to sustain its current lead and match modestly stronger competitors to eschew lagging further behind; conversely, a firm is less likely to match far weaker competitors due to its lack of attention and match far stronger competitors due to its inability to compete. Event history analysis of a set of IT companies' entry moves into various locations in China exhibits support for our hypotheses. Our findings suggest that a firm's matching response most often occurs under a moderate level of asymmetric market strength.

1. Introduction A central topic in the strategy literature concerns competitor analysis: that is, who competes with whom in an industry. Early works on this subject drew mainly from industrial organization economics to study competition at the industry level. This approach implicitly assumes that all firms in the same industry are de facto competitors. Later studies distinguished between different strategic groups in an industry and suggested that firms belonging to the same strategic group are apt to identify one another as competitors (Duan & Jin, 2014; Panagiotou, 2007; Short, Ketchen, Palmer, & Hult, 2007). Although these studies provided an essential foundation for competitor analysis, it has been noted that they cannot fully account for intra-industry heterogeneity in interfirm rivalry. Hence, strategy research in competitive dynamics (Chen & Miller, 2012) proposed to conduct competitor analysis from the perspective of an individual firm with reference to a specific rival. This firm-centric, rival-specific approach has contributed to refined analysis of interfirm competitive relationships (Chen, 1996). One notable observation is that the competitive tension that a rival exerts on a firm is often different from the tension that the focal firm exerts on that rival. Scholars have long acknowledged such an asymmetry in a competitive relationship (Carpenter, Cooper, Hanssens, & Midgley, 1988), and subsequent development in analytical techniques helped researchers to better capture asymmetric competitive relationships. For instance, Chen (1996) and Peteraf and Bergen (2003) compared two firms in terms of their differential positions in overlapping product markets and resource types. DeSarbo, Grewal, and Wind (2006) examined customers' (revealed) preference and posited that customers

⁎

might view one company's offerings as substitutes for another company's offerings but not the other way around. Albeit with differences in the analytical approach, these studies all uncovered that asymmetric competitive relationships are manifested in a variety of business contexts, including airline, automobile, and mobile phone industries. Besides these efforts at identifying asymmetric competitive relationships across diverse settings, however, how such relationships affect interactive firm behavior remains understudied. To address this gap, we consider the potential influence of asymmetric competitive relationships on a firm's awareness of a rival's actions, motivation to react, and capability to carry out a response effectively. These three behavioral drivers are summarized as the awareness-motivation-capability (AMC) framework (Chen & Miller, 2012). Beyond existing research that highlights the existence of asymmetry in competitive relationships, we distinguish between varying levels of asymmetry, identify the primary behavioral driver underneath a given level of asymmetry, and demonstrate the corresponding behavioral consequences. In our examination of the behavioral implications of asymmetric competitive relationships, we focus on a specific form of interactive market behavior: namely, a firm's matching response to a competitor's entry into a new (geographic) location. Various literature traditions have investigated a firm's inclination to match, or to imitate, a rival's strategic move (Lieberman & Asaba, 2006). Some studies found that a firm is more inclined to model rivals with superior market position (e.g., Haunschild & Miner, 1997), other studies suggested that a firm is more likely to imitate rivals possessing a comparable competitive position (e.g., D'Aveni, 1994; Peteraf, 1993), and still other studies

Corresponding author. E-mail addresses: [email protected] (K.-Y. Hsieh), [email protected] (E.E. Hyun).

http://dx.doi.org/10.1016/j.jbusres.2017.09.038 Received 17 May 2016; Received in revised form 18 September 2017; Accepted 20 September 2017 0148-2963/ © 2017 Elsevier Inc. All rights reserved.

Journal of Business Research 82 (2018) 202–212

K.-Y. Hsieh, E.E. Hyun

showed that a firm is apt to follow relatively weaker rivals (e.g., Hsieh, Tsai, & Chen, 2015; Terlaak & King, 2007). We will argue and show that these seemingly contradicting findings in the literature can be consolidated through distinguishing between the primary AMC factors associated with different levels of asymmetry in a competitive relationship. In analyzing asymmetric competitive relationships, we focus on competing firms' asymmetric market strength in their shared product markets. Accordingly, we theorize that a firm is more likely to match modestly weaker competitors to sustain its current lead and match modestly stronger competitors to eschew lagging further behind; conversely, a firm is less likely to match far weaker competitors due to its lack of attention and match far stronger competitors due to its inability to compete. We test our idea in the context of a set of IT companies' entry moves into various locations in China, and find empirical support for our conjecture. Taken together, the theory and findings presented in this paper advance the competitor analysis literature by demonstrating that asymmetry in a competitive relationship matters not only because it exists—as prior studies have shown—but also because it affects interactive firm behavior.

even become infeasible if managers attempt to base their assessments on certain indicators that are difficult to process (Hsieh & Hyun, 2016). In a multimarket environment, one critical factor that attracts mangers' attention concerns other firms' presence in those markets that are particularly important to the focal firm (Chen, 1996; Peteraf & Bergen, 2003). Therefore, we conceptualize (and measure) asymmetry in a competitive relationship with a focus on firms' product market profiles. Following Chen (1996), we use market commonality as an indicator of market strength. Firm A's market commonality with firm B indicates B's presence in the markets that are important to A. Accordingly, we define the asymmetry in a competitive relationship as the situation in which firm A's market commonality with firm B exhibits non-trivial difference with B's market commonality with A. That is, asymmetry arises when a competitor's presence in the shared markets is notably stronger or weaker than a focal firm's own presence in those markets. By contrast, symmetry in a competitive relationship can be said to arise when the presence of a pair of firms (i.e., a focal firm and a given rival) in their overlapping markets is (nearly) identical.

2. Conceptual background

2.3. The awareness-motivation-capability perspective

2.1. Location decision

In developing our argument, we utilize the awareness-motivationcapability (AMC) framework, which was first advanced by competitive dynamics research (Chen & Miller, 2012) and has recently been adopted in a wider range of studies (e.g., Angeli & Jaiswal, 2015; Keil, Laamanen, & McGrath, 2013; Nair & Selover, 2012; Peng & Liang, 2016). Following the AMC approach, we reason that a firm will match a rival's move when it is aware of that move, feels motivated to react, and is capable of carrying out an effective matching response. In the context of a firm's entry into a new location, we define matching response as the correspondence between a rival's recent entry move and a focal firm's inclination to enter the same location. We highlight a rival's ‘recent’ move because a move occurring in the distant past, once being considered a part of the status quo, is less likely to elicit additional competitive tension and new response. Also, focusing on a rival's recent entry move helps us to distinguish competitive interplays in the form of action and response (or research focus) from agglomeration economies, which tend to be the result of the accumulated investments made by various firms over a longer time span.1

A sizable volume of literature has examined firms' decision to enter an overseas location. Concerning where firms locate their overseas investment and business activities, the literature has identified three types of determinants: local conditions in a target location, capabilities of a parent firm, and firm-location fit (Nielsen, Asmussen, & Weatheralld, 2017). Local conditions such as institution environment, industry infrastructure, products demand, and supply of production factors and strategic resources determine the general attractiveness of a target location (Beugelsdijk & Mudambi, 2013). Technological, marketing, and management capabilities of a parent firm lay the foundation of competitive advantages that the firm can potentially leverage overseas (Kirca et al., 2011). Finally, a higher similarity between a target location and a parent firm's home environment increases firm-location fit and facilitates the firm to leverage its current advantages across geographic boundaries (Rugman & Verbeke, 2004). Location decision has important implications for interfirm competition. Rivals entering an overseas location might benefit from unique local conditions (e.g., attractive production factors, growing demand, and supporting infrastructure) and thereby enhance their competitive position relative to non-entrants; to avoid lagging behind rivals in the pursuit of overseas opportunities, a firm is under pressure to respond to rivals' entry moves (Head, Mayer, & Ries, 2002; Lieberman & Asaba, 2006). In what follows, we will examine how asymmetry in a competitive relationship affects a firm's inclination to respond by matching a rival's entry move and entering the same location.

2.4. Different levels of asymmetry When managers judge their company's competitive relationship with an industry rival as being asymmetric, they can further assess the size of the asymmetry. Accordingly, we account for the level of asymmetry in a competitive relationship, which captures the level of difference between firm A's market commonality with firm B and B′s market commonality with A. More generally, the level of asymmetry between a pair of competing firms refers to the size of the two parties' differential strength in their shared markets. Drawing from the AMC framework, we will examine how varying levels of relative market strength affect a firm's likelihood of matching a rival's entry move.2

2.2. Asymmetric market strength Asymmetry arises in a competitive relationship when two firms exert differing competitive tension on each other. In analyzing competitive relationships, scholars have utilized a range of indicators including firms' market and resource profiles as well as customers' (revealed) preferences (Chen, 1996; DeSarbo et al., 2006; Peteraf & Bergen, 2003). Among these indicators, a firm's market profile is usually most visible to managers and thus is more likely to have a direct impact on how managers evaluate and act upon their company's asymmetric competitive relationship with a rival. As prior research has shown, an easily observable indicator of competitive relationships typically exhibits greater influence on rivalry behavior (Chen & Miller, 2012). The task of monitoring and evaluating numerous competitors is highly demanding for managers with limited attentive capacity; it may

1 Empirically, we include a comprehensive set of control variables (listed in Table 1) to account for potential agglomeration economies as well as other factors that can affect the general attractiveness of a location. 2 In predicting interactive firm behavior using market commonality-based indicators (Chen, 1996)—including the asymmetry concept we define here—an implicit assumption is that strategic decisions across multiple product markets are coordinated at the corporate level. This assumption reasonably characterizes those focused firms specializing in an industry consisting of multiple, highly related market segments (as in our empirical context). By contrast, this assumption is less applicable to large and highly diversified firms, who often operate multiple business units as semi-autonomous profit centers with own decision-making authority.

203

Journal of Business Research 82 (2018) 202–212

K.-Y. Hsieh, E.E. Hyun

3. Hypotheses

rival's relative market strength (i.e., inferior or superior). Two additional factors can contribute to the likelihood of matching a modestly stronger rival's entry move. First, a rival with superior market performance in the past is often viewed as better informed about what constitutes an effective course of action (Haunschild & Miner, 1997). After this rival has entered a certain location, managers may interpret the observed entry move as revealing the potential of that location, and thus feel compelled to follow suit (Bikhchandani, Hirshleifer, & Welch, 1998). When the relative advantage of a rival is not very large but modest, firm managers may reason that their company still has a reasonable chance of competing with this rival in the same location for local customers, production factors, and institutional support (Hsieh et al., 2015). Second, even if managers decide to follow a rival into a new location that turns out to be a bad one, such a faulty decision tends to carry less repercussion for the reputation of decision makers. As Scharfstein and Stein put it: “an unprofitable decision is not as bad for reputation when others make the same mistake” (Scharfstein & Stein, 1990: 466). On the contrary, choosing not to follow a rival into a location that turns out to be promising can be deemed as a personal mistake, for which decision makers might be held responsible (Hsieh & Vermeulen, 2014). Therefore, we expect that matching response is more likely to occur corresponding to a rival who is modestly stronger than a focal firm in their shared product markets. Formally,

3.1. Moderate levels of asymmetry In what follows, we will first consider the situation in which a rival's market strength is either modestly weaker or stronger relative to a focal firm. We expect that matching response is more likely under these moderate levels of asymmetry, due to either a proactive intent to sustain the company's lead or a defensive intent to eschew falling further behind. 3.1.1. Weaker rival Previous research suggests that firms often concentrate on tracking more prominent rivals (Haunschild & Miner, 1997). Yet firms do not always overlook a relatively weaker rival if the gap in market strength with it is not large but moderate (Hsieh et al., 2015; Terlaak & King, 2007). We posit that a rival with modestly weaker market strength can be identified as a potential challenger and be placed under surveillance. After a modestly weaker rival has entered a potentially attractive location, managers may worry that not responding to the move will give this rival an opportunity to surpass their company (Ferrier, Smith, & Grimm, 1999). In other words, insofar as managers are aware of a modestly weaker rival and perceive it as a challenger, they are likely to be motivated to secure their company's lead through monitoring and reacting to this rival's move. For the objective of maintaining a lead, matching response represents an effective measure even when the true potential of a new location is highly uncertain. Whether the target location turns out to be a promising destination or not, following a modestly weaker rival there can help to sustain the competitive status quo, which favors the focal firm. As Dixit and Nalebuff (1991: 10) put it: “if you have the lead, the surest way to stay ahead is to play monkey see, monkey do.” Riskaverse managers may decide to follow a weaker rival with some potential to surpass their company into a new location, regardless of their own assessments of that location's potential (Head et al., 2002; Hsieh & Vermeulen, 2014). Thus, we expect that matching response is more likely to occur corresponding to a rival who is modestly weaker than a focal firm in their shared product markets. Formally,

Hypothesis 1b. A rival's entry move is more likely to increase a firm's likelihood of entering the same location when this rival's product market strength is moderately stronger relative to the focal firm in their shared markets. 3.2. More extreme levels of asymmetry So far, we have considered the situation in which a rival's market strength is either modestly weaker or modestly stronger relative to a focal firm. Next, we will examine the more extreme cases in which the level of asymmetry is quite high or low, indicating that a given rival is either much stronger or weaker than a focal firm in their shared markets. We expect that matching response is less likely under these more extreme levels of asymmetry, due to either a lack of awareness of a rival's move or an expected inability to contest.

Hypothesis 1a. A rival's entry move is more likely to increase a firm's likelihood of entering the same location when this rival's product market strength is moderately weaker relative to the focal firm in their shared markets.

3.2.1. Weaker rival Earlier we have posited that managers pay attention to a moderately weaker rival. However, not all rivals with a weaker market strength will get equal attention from firm managers. For a weaker rival to be recognized by managers as a significant player that warrants attention, it needs to have a comparatively non-trivial strength in the shared markets. Due to firms' limited attentive capacity (Hsieh & Hyun, 2016), they often overlook far less prominent competitors. Accordingly, matching response is less likely to occur corresponding to a rival with a much weaker strength in relation to a focal firm in their shared product markets. Firm managers may not even notice this rival's move if they have not been tracking this rival carefully. Even if managers do notice the move, they may see it as an inconsequential event posing insignificant competitive threat. (Chen et al., 2007), Therefore, we expect that:

Stronger rival. Rivals with stronger market presence are more salient and thus they are easier to recognize (Haunschild & Miner, 1997). Hence, firm managers might place a rival with modestly stronger market presence under close surveillance. Research suggests that a rival's superior market strength tends to raise managers' anxiety about the potential consequences of this rival's strategic moves (Chen, Su, & Tsai, 2007). Accordingly, the primary concern for managers is to avoid lagging further behind a rival possessing more strength in the shared product markets. Given this, when a modestly stronger rival has entered a new location, managers may fear that not responding to its move would allow this rival to become even more dominating, thereby placing their company in jeopardy. This thought process is likely to put a focal company into a defensive mode. As such, the firm may follow this modestly stronger rival into the same location to avoid lagging further behind. This mechanism underlying the decision to match a modestly stronger rival can be distinguished from the one at play in the decision to match a modestly weaker rival. In the previous section, we have highlighted managers' proactive intent to sustain their company's lead vis-à-vis a modestly weaker rival. By contrast, here we emphasize managers' defensive intent to eschew falling further behind a modestly stronger rival. Whether a firm's matching response is triggered by the proactive or defensive mechanism is determined by the direction of a

Hypothesis 2a. A rival's entry move is less likely to increase a firm's likelihood of entering the same location when this rival's product market strength is much weaker relative to the focal firm in their shared markets. 3.2.2. Stronger rival In addition to the lack of awareness, another important factor that can reduce matching response concerns a firm's inability to benefit from following a certain rival into the same location. Prior studies on matching response have implicitly assumed that members of an 204

Journal of Business Research 82 (2018) 202–212

K.-Y. Hsieh, E.E. Hyun

Together, our multiple hypotheses present a comprehensive model showing how varying levels of asymmetry in a competitive relationship will affect matching response likelihood. The model indicates that matching response is most likely to occur when the level of asymmetry is moderate. In the two moderate zones, a firm is likely to be aware of a rival's entry move, feel motivated to respond by entering the same new location, and be capable of establishing a comparable operation there. When the level of asymmetry approaches either extreme, however, matching propensity will be reduced, due to lack of awareness or inability to successfully contest. Finally, when the level of asymmetry is so small such that the relationship in question can be viewed as symmetric, the focal firm will become more reluctant to directly engage an equal rival by entering the same territory.

industry are equally competent in operating in a newly entered location (Head et al., 2002). Under this assumption, a firm that follows a rival into the same location can benefit as much if the target location turns out to be promising, and will suffer no more if the location turns out to be a bad destination. In both scenarios, the firm's competitive stand relative to the rival remains roughly unchanged. However, a rival with a much stronger market presence is likely to possess superior resources and skills, which it can leverage to assist its endeavor in establishing a solid operation in a new overseas location (Caves, 2007). When a rival is far more competent in operating in a new location, following it into the same location can no longer help a firm to maintain the competitive status quo. A firm may still pursue entry into the same location, but that firm can hardly establish a comparable operation there without the support of preexistent advantage (Kirca et al., 2011). As such, a rival with a far stronger market strength will be able to reap greater benefits if the target location turns out to be promising, whereas the focal firm will take more losses if the location turns out to be a bad destination (Hsieh et al., 2015). In both scenarios, the focal firm will be outperformed by a far stronger rival. In other words, if a firm attempts to match a rival's entry move but fails to set up a comparable operation in the new location, its relative competitive stand will be further weakened, not sustained. The expectation that a firm in this situation would be unsuccessful in competing against a far stronger opponent in the newly entered location can restrict its likelihood of matching this rival's entry move (Chen & Hambrick, 1995). Thus, we expect that:

4. Methods 4.1. Empirical context, data, and sample To test our hypotheses, we compiled a dataset on personal computer (PC) companies from Taiwan and tracked their entries into various locations in mainland China. Our sample was made up of all listed Taiwanese companies who had ever produced PCs or ancillary hardware using in-house manufacturing facilities. We identified these companies from two local bourses: the Taiwan Stock Exchange (where larger and more established companies were listed) and the GreTai Securities Market (for smaller, more entrepreneurial companies). Using Standard Industry Classification (SIC), we defined our focal firms as all those operating in the following five product markets: computers, monitors and terminals, computer peripherals, audio and video equipment, and communication equipment. This procedure yielded 205 focal firms, whose entries into China were modeled. Our observation period starts from 2000 and ends in 2005. During this time, many Taiwanese PC firms actively invested in mainland China, which represented the single most important destination for overseas investments. For our sample firms, China not only provided attractive sites for low-cost production but also increasingly became an important end market. Economic development accelerated in numerous locations, and local demand surged. Although relatively few firms had a major presence in China before its admission to the World Trade Organization (WTO) in the December of 2001, by the end of 2005 our sample companies' investment in China accounted for nearly 70% of their total investment abroad. By entering different locations in China, our sample firms improved their cost structure and better positioned themselves for seizing the growing demand from both emerging Chinese customers and western clients expanding their business interest in China. In contrast to the growth in China, in the U.S., for example, the average retail price of a PC decreased from $1800 to under $1000 during 1997 to 1998. In 2000 to 2001, the global demand for PCs declined for the very first time since IBM introduced its PC infrastructure in 1981. Under industry consolidation, opportunities for cost reduction and growth in China were highly crucial. Falling behind rivals in seizing these opportunities could jeopardize a firm's development and even survival. Constructing our main variables requires that we keep track of all entry moves made by every product market competitor facing a focal firm. Many of the 205 focal PC firms diversified into other related markets of information technology (IT) products. Accordingly, they competed not only among themselves, but also with other IT companies there. The related IT product markets where our sample firms had presence include telephones and cellular phones, storage media, cameras, optical instruments, measuring and control equipment, electromedical equipment, integrated circuits, discrete devices, semiconductors packaging and testing, electronic passive devices, bare printed circuit boards, printed circuit boards assembly, electronic parts and components, liquid crystal panel, and optoelectronic materials and components. Hence, we identified additional 344 companies (i.e.,

Hypothesis 2b. A rival's entry move is less likely to increase a firm's likelihood of entering the same location when this rival's product market strength is much stronger relative to the focal firm in their shared markets. 3.3. Symmetric competitive relationship Finally, we consider the implications of a (nearly) symmetric competitive relationship: the situation where a given rival and a focal firm possess highly similar strength in their shared product markets. The strategy literature has long suggested that strategic similarity between competing firms—in terms of both internal operations and market postures—can facilitate tacit coordination and reduce the intensity of rivalry (D'Aveni, 1994; Peteraf, 1993). This occurs because similar firms can more easily identify each other and recognize their interdependence, and such a mutual understanding increases firms' motivation and ability to tacitly coordinate. In the setting of new location entry, tacit coordination can take the form that a firm avoids entering a territory where a rival with an equal market strength has displayed a major interest via its recent entry move. This form of tacit coordination implies a lowered likelihood of matching response. Instead of matching an equal rival's entry move, a firm can instead focus on a different destination. To the extent that different locations are characterized by different local conditions, focusing on a different destination can also help a firm to differentiate itself from an equal rival and avoid the undesirable scenario of competitive convergence (Porter, 1997), where competing firms become increasingly indistinguishable and face growing pressure of fighting head to head against each other for the same customers via identical value propositions. This line of thinking leads to the following conjecture: Hypothesis 3. A rival's entry move is less likely to increase a firm's likelihood of entering the same location when this rival's product market strength is (nearly) symmetric to that of the focal firm in their shared markets. 3.4. Summary: A theoretical model Fig. 1 presents a theoretical model that summarizes our hypotheses. 205

Journal of Business Research 82 (2018) 202–212

K.-Y. Hsieh, E.E. Hyun

Fig. 1. A theoretical model of asymmetric market strength and matching response.

22 locations in China. Our data were organized as firm-location level observations. As our sample firms were in the manufacturing sector, many entry moves involved establishing a production site. We used the Cox's semi-parametric model to estimate a focal firm's entry rate:

product market rivals of the focal, sample firms) that had presence in these related IT product markets, and collected information on these additional firms to construct our independent variables. For all the firms identified through the above snowball approach, we traced their moves into different locations in mainland China. We obtained data from a regulatory agency in Taiwan: Investment Commission under the Ministry of Economic Affairs. Also, Taiwan's listed companies reported their activities in China to an information platform managed by the Taiwan Stock Exchange. Relying on these supplementary information sources, we created a dataset with a comprehensive coverage of investment activities in China. We delineated a location as a province (e.g., Jiangsu; Guangdong) or municipality (e.g., Shanghai; Beijing) in China. We defined an entry move as the establishment of a company's first subsidiary in a location. A few prior studies utilized the same data sources to investigate different issues. For instance, Hsieh and Hyun (2016) examined the spatial and temporal distribution of investment in China, Hsieh et al. (2015) unveiled the escalation of commitment to loss-incurring investment projects, and Hsieh and Vermeulen (2014) mapped the structure of encounter between multiple rivals facing a focal firm. Despite (partial) overlap in data sources, the current paper has a different analytical locus and sampling scheme. Whereas Hsieh and Hyun (2016) analyzed the distribution of investments at the whole China level, the current study examines matching response at the level of a province/ municipality. Whereas Hsieh et al. (2015) traced continuing commitment to loss-incurring investments over a prolonged period (up to 2011), the current study focuses on a narrower window (before and after China's admission to WTO) because firms' competitive engagement in China was most intensive during this time window. By 2005, investments in China from our sample firms kept increasing, but such a situation began to change following the 2007–2008 financial crisis. In recent years, a growing number of firms are divesting and relocating their production-related activities in coastline China due to rising cost of operations.

h(u) = h0(u)⋅exp(X ′β ) In the above equation, h(u) denotes entry rate given a spell of duration u. Therefore, h(u) is the product of the baseline rate h0(u) and an exponential linear function of covariates X. In this semi-parametric model, the baseline rate was left unspecified except for being non-negative, and the effects of covariates were estimated through maximum partial likelihood. 4.3. Explanatory variables In this study, we defined the asymmetry in a competitive relationship as the situation in which a rival j's strength in the product markets that it shares with a focal firm i is weaker or stronger than focal firm i's strength in their shared markets. Accordingly, different levels of asymmetry were identified through juxtaposing industry members' distinct market profiles. Let Sim and Sjm denote i's and j's revenues in product market m, respectively, and let Sm denote the size of market m. The market strength difference between i and j was computed as:

market strength differenceijm, t =

Sjm, t − Sim, t Sm, t

As i could encounter j across multiple product markets, we next aggregated market strength difference across all markets. In doing so, we weighted each market m by m's contribution to focal firm i's total revenue. Let Si denote the total revenue of firm i, and let Sim denote i's revenue from market m. The level of asymmetry experienced by focal firm i in its competitive relationship with rival j was computed as:

level of asymmetryij, t = 4.2. Analytical approach and dependent variable

∑m ⎛⎜

Sim, t

⎝ Si, t

×

Sjm, t − Sim, t ⎞ Sm, t

⎟

⎠

A negative asymmetry score indicates that rival j was weaker relative to focal firm i in markets key to i, whereas a positive asymmetry score indicates that j was comparatively stronger. The value of this measure approaches zero when the two parties possessed highly similar market strength in their shared product markets. Based on the above asymmetry score, we divided all rivals facing a focal firm into nine categories, including substantially weaker rivals (score < − 0.07), nearly equal or equal rivals (score > −0.01 & < 0.01), and significantly stronger rivals (score > 0.07) by an interval of 0.02. Next, we constructed nine corresponding variables, as listed in Table 2, to indicate whether any rival in a certain category has made an entry move in a location recently (within the last year/four quarters). For example, if at least one equal rival of a firm had made a recent entry move into Shanghai, the corresponding firm-location level binary indicator “move by an equal rival” would take the value of one;

Our hypotheses predicted a firm's likelihood of entering a new geographic location. We operationalized our dependent variable as entry rate: the instantaneous probability that a firm will enter a given location in which it did not have a subsidiary before. Each entry move made by a focal firm with respect to a target location was specified as an ‘event.’ The temporal horizon was modeled as the time elapsed since at least one product market rival of a firm had entered a target location, in that the focal firm became ‘at risk’ of matching a rival's entry move since then. To avoid the issue of left-censoring, all companies' operations in China were traced all the way back to 1991 to determine when a ‘spell’ should begin if it was before 2000. This ‘spell,’ measured by quarters, ended in an event if a firm entered a location during the period of our observation; otherwise, a spell was right-censored. Of the 205 focal firms in our sample, 171 made a total of 249 entry moves into 206

Journal of Business Research 82 (2018) 202–212

K.-Y. Hsieh, E.E. Hyun

Table 1 Exploratory factor analysis of location conditions. Item

International-ization

Wealth of population

Supply of skilled labor

Transportation infrastructure

1. Total foreign capital 2. Exports 3. Imports 4. Gross domestic product per capita 5. Disposable income per capita 6. Household expenditure per capita 7. No. of professional personnel 8. Population with a college degree 9. No. of recent college graduates 10. Highway density 11. Railway density Cumulative variance explained

0.90 0.87 0.83 0.49 0.45 0.44 0.40 0.28 0.12 0.44 0.12 0.30

0.25 0.33 0.36 0.63 0.81 0.81 −0.43 0.20 0.31 0.49 0.18 0.54

0.25 0.29 0.27 0.19 0.21 0.18 0.75 0.91 0.90 0.06 −0.01 0.77

0.16 0.15 0.29 0.55 0.27 0.27 −0.07 0.08 0.02 0.66 0.96 0.94

Note: Orthogonal varimax rotation. Factor loadings displayed in italics indicate the measures used to construct the variables.

firm established its first subsidiary in China. These covariates were updated at the end of each quarter in our Cox models. Third, due to substantial location differences in China, we controlled for local conditions in each location. Number of all incumbents captures the sheer number of a firm's rivals who had entered a given location by far (regardless of when they entered), which might create agglomeration economies and hence affect the focal firm's propensity to invest in a location. In addition, we created location-level variables that capture differences between 22 locations in China. As shown in Table 1, we conducted exploratory factor analysis using eleven macroeconomic indicators for each location using data from the China Statistical Yearbooks. These indicators loaded on four distinct factors: internationalization, wealth of population, supply of skilled labor, and transportation infrastructure. These location-level covariates were updated at the end of each year in our Cox models. Finally, we accounted for unobserved temporal heterogeneity by including year dummies. In addition, since our data were organized at the firm-location level, each focal firm was associated with multiple observations corresponding to its rate of entering different locations, and observations of the same focal firm were likely to be correlated to one another. Hence, we computed cluster-robust variance estimates (Williams, 2000) to adjust for this potential non-independence between observations.

otherwise, the variable would take a zero value.3 By examining the effect of these nine variables on a firm's entry rate, we could tell whether the firm is inclined to match a weaker, equal, or stronger rival. These explanatory variables entered our Cox models as time-varying covariates, with values being updated at the end of each quarter. Our measure extends Chen's (1996) notion of market commonality, which can be expressed as:

market commonalityij, t =

∑m ⎜⎛

Sim, t

⎝ Si, t

×

Sjm, t ⎞ ⎟

Sm, t ⎠

Market commonality reflects rival j's absolute market strength (Sjm) without considering focal firm i's market strength (Sim). Nevertheless, firms with differing market strengths are likely to experience different levels of competitive tension from the same rival. For example, a rival with a low market strength might appear marginal to a firm with a high market strength, but may appear significant to another firm with a similarly low market strength. Our asymmetry score accounts for this complexity by capturing rival j's relative market strength with respect to each focal firm i (Sjm − Sim). 4.4. Control variables We included numerous control variables as covariates. First, we controlled for each focal firm's own market strength and operating performance, which could influence its overseas operations. Own market strength was measured as the weighted average of a firm's shares (by revenues) in all the product markets in which it operated. Specifically:

own market strengthi, t =

∑m ⎛⎜

Sim, t

⎝ Si, t

×

Sim, t ⎞

⎟

Sm, t ⎠

4.5. Endogeneity considerations Our research design accounted for three potential sources of endogeneity: sample selection bias, simultaneity, and endogenous explanatory variables. To reduce selection bias, we incorporated all firms in the selected product markets—not only firms who had taken an action—in our sample. To reduce simultaneity issue, we adopted the event history analysis framework and updated time-varying covariates at the end of the previous time interval before an action event can potentially occur. Endogenous explanatory variables become a major concern when important factors determining both the dependent and independent variables are omitted in the regression models. In our context, local attractiveness in China is one crucial factor that can affect both a focal firm’ action (dependent variable) and its competitors' actions (independent variables). Hence, we included comprehensive indicators of location conditions (as reported in Table 1) to address this issue.

.

Operating performance, on the other hand, was measured as return on assets. Because data on firms' sales and returns were available annually, these two covariates were updated yearly in our Cox models. Second, we controlled for each focal firm's prior experience in China, which could facilitate subsequent entry into other locations in China, using two variables. The variable of first entry in China is a binary variable indicating whether the focal firm has (not) entered China before. Experience in China captures how much time had elapsed since a 3 Of the observations taking the value of one, about 15% were associated with two or more rivals making an entry move into a given location within the last year. As a robustness check, we constructed alternative variables that aggregated the number of entry moves made by multiple rivals, and obtained fully consistent results. We prefer binary indicators because our theoretical argument pertained to a firm's response to a certain rival (dyadic interaction) instead of the aggregate assessment of multiple rivals' impact.

5. Results 5.1. Hypotheses testing Table 2 displays the summary statistics of all the variables used in 207

Mean

0

0

0

0.04

0.05

0.18

Move by a stronger rival with 0.03 0.18 6. Asymmetry score: 0.01–0.03 7. Asymmetry 0.02 0.13 score: 0.03–0.05 8. Asymmetry 0.01 0.12 score: 0.05–0.07 9. Asymmetry 0.04 0.21 score: > 0.07 10. Own market 0.03 0.05 strength 0.06 0.11 11. Firm performance (ROA) 12. First entry in 0.30 0.46 China 13. Experience in 1.96 2.14 China (years) 14. No. of all 5.08 9.99 incumbents 15. Internationaliz0.23 1.02 ation 16. Wealth of 0.37 1.00 population 17. Skilled labor 0.22 1.10 18. Transportation 0.19 1.23

1

1

1

1

0.32

0.44

1

8

117

2.80

3.85

3.38 3.16

0

0

0

0.01

−0.94

0

0

1

−1.37

−1.85

−1.50 −1.44

1

1

1

1

1

Max

0

0

0

Min

0.05

SD

Move by an equal rival with 0.07 0.26 5. Asymmetry score: − 0.01–0.01

Move by a weaker rival with 1. Asymmetry 0.00 score: < − 0.07 0.00 2. Asymmetry score: − 0.05–−0.07 0.00 3. Asymmetry score: − 0.03–−0.05 0.03 4. Asymmetry score: − 0.01–−0.03

Variable

Table 2 Summary statistics.

208

0.00 0.02

−0.01 −0.01

0.01 −0.02

0.02

0.04

0.02

−0.01 −0.01

0.00

0.04

0.04

0.00

0.01

−0.02 0.02

0.00

0.02

0.07

0.00

−0.01

0.19

0.00

0.02

0.07

0.10

2

−0.01

−0.01

0.03

0.05

0.11

1

− 0.01 − 0.02

0.00

0.06

− 0.03 − 0.08

− 0.01

0.22

0.29

− 0.03

− 0.01 0.07

0.05

0.03

0.03

0.12

0.05

0.11

0.15

0.51

4

0.01

0.00

0.06

0.01

0.00

0.01

0.05

0.05

0.09

3

0.05 −0.12

0.00

0.28

0.48

−0.05

0.04

0.04 − 0.08

− 0.01

0.26

0.35

− 0.06

0.05

0.00

− 0.03

−0.07 0.03

0.21

0.13

0.15

6

0.21

0.10

0.16

0.24

5

0.02 − 0.04

0.01

0.17

0.19

− 0.05

0.04

0.01

− 0.03

0.14

0.05

7

0.02 − 0.03

0.03

0.15

0.16

− 0.03

0.04

− 0.01

− 0.04

0.09

8

0.07 −0.09

0.04

0.28

0.33

−0.04

0.03

0.01

−0.06

9

0.00 − 0.02

0.01

−0.07 0.00

−0.08

0.00

0.00

− 0.08 0.00

−0.14

0.08

11

0.24

− 0.17

0.12

10

−0.24 0.03

−0.25

0.18

0.03

−0.60

12

0.27 −0.04

0.26

−0.18

−0.05

13

0.12 −0.01

0.17

0.46

14

−0.06 −0.05

0.08

15

−0.08 0.10

16

− 0.18

17

K.-Y. Hsieh, E.E. Hyun

Journal of Business Research 82 (2018) 202–212

Journal of Business Research 82 (2018) 202–212

K.-Y. Hsieh, E.E. Hyun

Table 3 Cox models of firm's entry rate. Variable

Model 1

Model 2

Model 3

Model 4

Model 5

Model 6

0.676⁎⁎ (0.245)

0.238 (0.676) 0.666⁎⁎ (0.248)

0.232 (0.781) 0.235 (0.684) 0.680⁎⁎ (0.249)

−0.205 (0.755) 0.229 (0.795) 0.246 (0.684) 0.690⁎⁎ (0.253)

− 0.024 (0.229)

− 0.050 (0.227)

−0.070 (0.229)

−0.084 (0.229)

0.185 (0.196)

0.162 (0.196) 0.686⁎⁎ (0.232)

0.149 (0.195) 0.696⁎⁎ (0.231) 0.593⁎ (0.276)

3.174⁎⁎⁎ (1.768) 0.490 (0.591) 0.940⁎⁎ (0.209) 0.152⁎⁎⁎ (0.091) 0.011⁎⁎ (0.004) 0.838⁎⁎ (0.129) 0.493⁎⁎ (0.120) 0.373⁎⁎ (0.130) 0.207⁎⁎ (0.064) Included − 1668.852 437.649 0.100

3.240⁎⁎⁎ (1.786) 0.510 (0.588) 0.955⁎⁎ (0.212) 0.156⁎⁎⁎ (0.093) 0.011⁎⁎ (0.004) 0.826⁎⁎ (0.130) 0.487⁎⁎ (0.120) 0.365⁎⁎ (0.130) 0.210⁎⁎ (0.064) Included − 1664.865 440.944 0.102

3.370⁎⁎⁎ (1.827) 0.507 (0.590) 0.937⁎⁎ (0.212) 0.153⁎⁎⁎ (0.093) 0.011⁎⁎ (0.004) 0.815⁎⁎ (0.130) 0.484⁎⁎ (0.120) 0.360⁎⁎ (0.130) 0.210⁎⁎ (0.065) Included −1662.571 450.672 0.103

0.141 (0.196) 0.681⁎⁎ (0.233) 0.589⁎ (0.275) 0.173 (0.205) 3.534⁎⁎⁎ (1.927) 0.490 (0.596) 0.938⁎⁎ (0.211) 0.152 (0.093) 0.011⁎⁎ (0.004) 0.810⁎⁎ (0.131) 0.486⁎⁎ (0.121) 0.358⁎⁎ (0.131) 0.212⁎⁎ (0.065) Included −1662.180 452.793 0.103

Move by a weaker rival with Asymmetry score: < − 0.07 Asymmetry score: − 0.05–−0.07 Asymmetry score: − 0.03–−0.05 Asymmetry score: − 0.01–−0.03 Move by an equal rival with Asymmetry score: − 0.01–0.01

0.279 (0.208)

Move by a stronger rival with Asymmetry score: 0.01–0.03 Asymmetry score: 0.03–0.05 Asymmetry score: 0.05–0.07 Asymmetry score: > 0.07 Own market strength Firm performance (ROA) First entry in China Experience in China (years) No. of all incumbents Internationalization Wealth of population Skilled labor Transportation Year dummies Log (pseudo)likelihood Wald Chi-sq Pseudo R-sq

3.509⁎ (1.665) 0.522 (0.583) 0.948⁎⁎ (0.211) 0.160⁎⁎⁎ (0.092) 0.016⁎⁎ (0.003) 0.857⁎⁎ (0.124) 0.492⁎⁎ (0.121) 0.364⁎⁎ (0.130) 0.195⁎⁎ (0.062) Included −1674.425 414.563 0.097

3.583⁎ (1.684) 0.497 (0.587) 0.944⁎⁎ (0.211) 0.158⁎⁎⁎ (0.092) 0.013⁎⁎ (0.004) 0.851⁎⁎ (0.125) 0.496⁎⁎ (0.120) 0.365⁎⁎ (0.130) 0.202⁎⁎ (0.063) Included − 1673.421 420.144 0.097

Notes: Cluster-robust standard errors (by firms) are in parentheses. ⁎ p < 0.05; all two-tailed test. ⁎⁎ p < 0.01; all two-tailed test. ⁎⁎⁎ p < 0.1; all two-tailed test.

Each explanatory variable indicates whether a rival belonging to a certain asymmetry category (as well as the symmetry category) has made a recent entry move. Hence, a positive and significant coefficient

this study. Table 3 reports the results of Cox models. Model 1 includes only control variables. The nine explanatory variables enter Models 2 to 6 successively.

Fig. 2. Estimated effect and 95% confidence interval (based on Model 6).

209

Journal of Business Research 82 (2018) 202–212

K.-Y. Hsieh, E.E. Hyun

associated with any of these nice explanatory variables suggests that a firm's rate of entering the same location is increased by a certain rival's recent move—or that the focal firm is inclined to matched that rival's move. Throughout Models 2–6, only three coefficients corresponding to a rival with a modest level of asymmetry (asymmetry score: − 0.03–− 0.01, 0.03–0.05, and 0.05–0.07) are positive and significant. The other six coefficients corresponding to a (nearly) symmetric rival (score: − 0.01–0.01) and to a far weaker and stronger rival (score < − 0.03 and > 0.07) are all insignificant. This overall pattern is consistent with our conjecture, which predicts that matching response is most likely to occur under a moderate level of asymmetry in market strength. 5.2. Additional analysis

Fig. 4. Estimated effect based on alternative continuous variables.

Further examination of Table 3 reveals that the standard errors associated with a rival in the three weakest categories (asymmetry score: −0.03–− 0.05, − 0.05–−0.07, and < − 0.07) are generally larger. This situation is presented graphically in Fig. 2, which depicts the means and the 95% confidence intervals of entry rate multipliers based on the results of Model 6. A wider confidence interval (or a larger standard error) indicates a lower precision of estimates. A possible reason behind the lower precision is that rivals with a weaker market presence were generally less active and took fewer moves (as can be seen in Table 2), and hence data points available for estimating the influence of these rivals' moves are more limited. To evaluate the robustness implications of this data limitation, we aggregated the information associated with the moves by rivals in the three weakest categories. Instead of estimating three separate parameters, we estimated one parameter to summarize the influence of a move taken by a rival in any of the three weakest categories. Results based on this alternative approach are presented in Fig. 3, where the confidence interval associated with a far weaker rival's move is narrowed. But still, in line with our main analysis, the corresponding average effect is statistically insignificant (i.e., multiplier is indifferent from one). Our asymmetry scores reflect the difference between a rival's market strength and a firm's own market strength. Nonetheless, the same level of difference might appear marginal to a firm with a high absolute market strength but appear substantial to a firm with a low absolute market strength. As a robustness test, we normalized the asymmetry scores by a firm's own market strength and constructed an alternative set of explanatory variables accordingly. Results based on the normalized asymmetry scores are fully consistent with those obtained using raw asymmetry scores. Still, normalization is less preferred due to various issues concerning the use and interpretation of ratios (Wiseman, 2009). We have divided rivals into nine categories associated with various cut points of asymmetry scores. By transforming continuous asymmetry scores into discrete categories, however, we discarded variation and

assumed homogeneity within each category. As a robustness test, we constructed two continuous variables—asymmetry relative to stronger rivals and asymmetry relative to weaker rivals—by calculating the average of the asymmetry scores of all relatively stronger and all relatively weaker rivals who recently took an entry move. We then included these two alternative variables as well as their square terms as predictors. As presented in Fig. 4, this additional analysis yields consistent results. 6. Discussion In this study, we have extended the awareness-motivation-capability framework to examine the behavioral implications of differing levels of asymmetry in market strength. We predict that a firm is most likely to match a rival's entry move when the rival's strength in shared markets is either modestly weaker or modestly stronger relative to the focal firm. We reason that such behavioral tendency arises because a firm is likely to be aware of a rival with a moderate difference in market strength, feel motivated to respond by matching this rival's move, and be capable of competing with this rival in the same territory. Event history analysis of a set of IT companies' entry moves into various locations in China lends support for our conjecture. 6.1. Research implications Our study has demonstrated the value of considering multiple mechanisms driving interfirm rivalry behavior. While previous work implicitly assumed the joint existence of awareness, motivation, and capability factors in generating competitive interaction, little progress was made in assessing their relative influence under different competitive situations (Chen & Miller, 2012). Our findings suggest that multiple mechanisms are at work underneath a firm's decisions to match its rival's strategic actions, and that the primary mechanism can shift from one to another depending on the level of asymmetry between a pair of Fig. 3. Estimated effect and 95% confidence interval (constrained model).

210

Journal of Business Research 82 (2018) 202–212

K.-Y. Hsieh, E.E. Hyun

benefit from economies of agglomeration. For example, firms previously entered may foster the development of local supplier networks, which can attract followers. Whereas agglomeration economies are the result of accumulated investments made by numerous firms in one location over a longer time span, our analysis only focuses on the correspondence between a rival's recent move and a firm's potential matching response over a shorter time span. Future research can examine how economies of agglomeration interact with competitionbased motivations in shaping interactive firm behavior. Fourth, our results imply that in the face of competitive symmetry, a given rival's prior entry will not significantly increase a firm's rate of entering the same location. It is important to note that such an observation does not imply that a firm does not respond to a symmetric rival at all. The firm may respond to this rival's move in other ways, such as entering a different location or enhancing the firm's existing operations. A careful analysis of alternative response options would be a fruitful next step for future research. Finally, we treat entry move as an umbrella concept encompassing a range of value chain activities that a firm might perform in a location. These activities—including production, logistics, marketing and sales, and research and development—might be associated with different competitive dynamics. For instance, our sample firms typically invested in both production and logistics activities when entering a new location in China. Depending on managers' assessment of the risk of escalating capacity competition, a firm might limit the scale of its up-front investment in production activities; on the other hand, investment scattering across various logistics activities is less likely to attract awareness and competitive reaction, yet making such investment helps a firm to better prepare itself for future scale-based competition. These complexities warrant future investigation.

competing firms. The present study also informs organizational research on matching response. Some prior studies found that a firm is more inclined to model rivals with superior market position (e.g., Haunschild & Miner, 1997), other studies suggested that a firm is more likely to imitate rivals possessing a comparable competitive position (e.g., D'Aveni, 1994; Peteraf, 1993), and still other studies showed that a firm is apt to follow relatively weaker rivals (e.g., Hsieh et al., 2015; Terlaak & King, 2007). These seemingly contradicting findings can be consolidated through paying attention to relative strength of firms in their shared markets and considering different levels of asymmetry in strength between them in those markets. The results of this study indicate that a firm is more inclined to follow other organizations who are neither far superior nor far inferior in their market strength; yet, among these organizations, those who are modestly ‘stronger’ or modestly ‘weaker’ are more likely to be followed. As suggested earlier, this pattern emerges perhaps because different behavioral mechanisms govern a firm's interactions with different industry members. Our conceptualization and measure of asymmetric market strength are grounded on a multimarket industry. Scholars in the fields of industrial organization and strategy have long examined competition in multimarket settings (Bernheim & Whinston, 1990). A central idea in the multimarket competition research is that repeated market encounter facilitates tacit coordination and reduces the intensity of rivalry, as managers worry that aggression in one market can trigger costly retaliation across multiple markets by a multimarket competitor (Head et al., 2002). More recent research further suggests that multimarket contact reduces aggressive rivalry particularly when competing firms exhibit differentiated interests in their shared markets, such that a firm respects a rival's territory in exchange for its rival's respect for the firm's own territory (Hsieh & Vermeulen, 2014). Our research adds to this literature tradition by unveiling an additional scenario in which a firm avoids engaging a multimarket rival in the rival's territory (location) of interest; that is, when the two parties possess highly similar strength in the shared markets important to the focal firm. This article also has practical implications. Managers can use our measure of asymmetric market strength to map their industry environment and distinguish among their rivals. For a company planning to make a strategic move—including entering a new market, launching a new product, or adopting a new technology—such a mapping exercise can help managers to identify likely followers and design preemptive or defensive actions accordingly.

Acknowledgements Financial support from Waseda University and Hongik University is gratefully acknowledged. References Angeli, F., & Jaiswal, A. K. (2015). Competitive dynamics between MNCs and domestic companies at the base of the pyramid: An institutional perspective. Long Range Planning, 48(3), 182–199. Bernheim, B. D., & Whinston, M. D. (1990). Multimarket contact and collusive behavior. RAND Journal of Economics, 21(1), 1–26. Beugelsdijk, S., & Mudambi, R. (2013). MNEs as border-crossing multi-location enterprises: The role of discontinuities in geographic space. Journal of International Business Studies, 44(5), 413–426. Bikhchandani, S., Hirshleifer, D., & Welch, I. (1998). Learning from the behavior of others: Conformity, fads, and informational cascades. Journal of Economic Perspectives, 12(3), 151–170. Carpenter, G. S., Cooper, L. G., Hanssens, D. M., & Midgley, D. F. (1988). Modeling asymmetric competition. Marketing Science, 7(4), 393–412. Caves, R. E. (2007). Multinational enterprise and economic analysis (3rd ed). Cambridge, UK: Cambridge University Press. Chen, M.-J. (1996). Competitor analysis and interfirm rivalry: Toward a theoretical integration. Academy of Management Review, 21(1), 100–134. Chen, M.-J., & Hambrick, D. C. (1995). Speed, stealth, and selective attack: How small firms differ from large firms in competitive behavior. Academy of Management Journal, 38(2), 453–482. Chen, M.-J., & Miller, D. (2012). Competitive dynamics: Themes, trends, and a prospective research platform. Academy of Management Annals, 6, 135–210. Chen, M.-J., Su, K.-H., & Tsai, W. (2007). Competitive tension: The awareness-motivationcapability perspective. Academy of Management Journal, 50(1), 101–118. D'Aveni, R. A. (1994). Hypercompetition: Managing the dynamics of strategic maneuvering. New York: Free Press. DeSarbo, W. S., Grewal, R., & Wind, J. (2006). Who competes with whom? A demandbased perspective for identifying and representing asymmetric competition. Strategic Management Journal, 27(2), 101–129. Dixit, A. K., & Nalebuff, B. J. (1991). Thinking strategically: The competitive edge in business, politics, and everyday life. New York, NY: W. W. Norton. Duan, X., & Jin, Z. M. (2014). Positioning decisions within strategic groups: The influences of strategic distance, diversification and media visibility. Management Decision, 52(10), 1858–1887. Ferrier, W. J., Smith, K. G., & Grimm, C. M. (1999). The role of competitive action in market share erosion and industry dethronement: A study of industry leaders and

6.2. Limitations and future directions The present research has several limitations that warrant future research. First, we do not directly observe how firm managers perceive a market rival. Without information on managerial perceptions, we rely on firms' objective market profiles to characterize their competitive relationships. Similarly, while we base our theoretical argument on managers' aversion to being out-competed and their aversion to escalating rivalry, we do not directly observe managerial motivations. As such, our empirical analysis represents a reduced-form test. Future research may seek to capture managerial perceptions and motivations directly through surveying firm managers and industry experts. Second, in this study, when a focal firm encounters a rival across multiple product markets, we only consider the relationship between two parties in totality by aggregating their market strength differences across all common markets. We adopt this simple conceptualization as an early step because prior research on asymmetric competitive relationships has been limited. In doing so, we ignore market heterogeneity. Future research might distinguish between the levels of (a) symmetry in each market and examine both within- and cross-market effects. Third, we focus primarily on competition-based motivations. Yet, it is possible that a firm may follow certain rivals into a new market to 211

Journal of Business Research 82 (2018) 202–212

K.-Y. Hsieh, E.E. Hyun

Peteraf, M. A., & Bergen, M. E. (2003). Scanning dynamic competitive landscapes: A market-based and resource-based framework. Strategic Management Journal, 24(10), 1027–1041. Porter, M. E. (1997). What is strategy? Harvard Business Review, 74(6), 61–78. Rugman, A. M., & Verbeke, A. (2004). A perspective on regional and global strategies of multinational enterprises. Journal of International Business Studies, 35(1), 3–18. Scharfstein, D. S., & Stein, J. C. (1990). Herd behavior and investment. American Economic Review, 80(3), 465–479. Short, J. C., Ketchen, D. J., Palmer, T. B., & Hult, G. T. M. (2007). Firm, strategic group, and industry influences on performance. Strategic Management Journal, 28(2), 147–167. Terlaak, A., & King, A. A. (2007). Follow the small? Information-revealing adoption bandwagons when observers expect larger firms to benefit more from adoption. Strategic Management Journal, 28(12), 1167–1185. Williams, R. L. (2000). A note on robust variance estimation for cluster-correlated data. Biometrics, 56, 645–646. Wiseman, R. M. (2009). On the use and misuse of ratios in strategic management research. In D. D. Bergh, & D. J. Ketchen (Eds.). Research methodology in strategy and management (pp. 75–110). Bingley, UK: Emerald.

challengers. Academy of Management Journal, 42(4), 372–388. Haunschild, P. R., & Miner, A. S. (1997). Modes of interorganizational imitation: The effects of outcome salience and uncertainty. Administrative Science Quarterly, 42(3), 472–500. Head, K., Mayer, T., & Ries, J. (2002). Revisiting oligopolistic reaction: Are decisions on foreign direct investment strategic complements? Journal of Economics and Management Strategy, 11(3), 453–472. Hsieh, K.-Y., & Hyun, E. (2016). Concentration of rivals' actions and competitive reaction. Management Decision, 54(9), 2188–2207. Hsieh, K.-Y., Tsai, W., & Chen, M.-J. (2015). If they can do it, why not us? Competitors as reference points for justifying escalation of commitment. Academy of Management Journal, 58(1), 38–58. Hsieh, K.-Y., & Vermeulen, F. (2014). The structure of competition: How competition between one's rivals influences imitative market entry. Organization Science, 25(1), 299–319. Keil, T., Laamanen, T., & McGrath, R. G. (2013). Is a counterattack the best defense? Competitive dynamics through acquisitions. Long Range Planning, 46(3), 195–215. Kirca, A. H., Hult, G. T. M., Roth, K., Cavusgil, S. T., Perryy, M. Z., Akdeniz, M. B., ... White, R. C. (2011). Firm-specific assets, multinationality, and financial performance: A meta-analytic review and theoretical integration. Academy of Management Journal, 54(1), 47–72. Lieberman, M. B., & Asaba, S. (2006). Why do firms imitate each other? Academy of Management Review, 31(2), 366–385. Nair, A., & Selover, D. D. (2012). A study of competitive dynamics. Journal of Business Research, 65(3), 355–361. Nielsen, B. B., Asmussen, C. G., & Weatheralld, C. D. (2017). The location choice of foreign direct investments: Empirical evidence and methodological challenges. Journal of World Business, 52(1), 62–82. Panagiotou, G. (2007). Reference theory: Strategic groups and competitive benchmarking. Management Decision, 45(10), 1595–1621. Peng, Y.-S., & Liang, I.-C. (2016). A dynamic framework for competitor identification: A neglecting role of dominant design. Journal of Business Research, 69(5), 1898–1903. Peteraf, M. A. (1993). Intra-industry structure and the response toward rivals. Managerial and Decision Economics, 14(6), 519–528.

Kai-Yu Hsieh ([email protected]) is an associate professor in global management at Waseda Business School, Waseda University. He obtained his PhD in strategic and international management from London Business School. His current research examines internationalization strategy and multinational organizations, with a focus on crossborder joint ventures and acquisitions. His work has appeared in leading journals including the Academy of Management Journal and Organization Science, among others. Eunjung (E.J.) Hyun ([email protected]) is an assistant professor in management at Hongik University in Seoul. She obtained her PhD in organizations and strategy from Chicago University's Booth School of Business and taught at Hitotsubashi University in Japan. Her research areas include cross-border expansion of services firms and the role of reputation and status in organizations and markets.

212