Past interactions and new foreign direct investment location decisions: firm-specific analysis in the global tire industry.
* Analyzing the nature of competitive interaction among multinational firms in the tire industry, we find that the histories of the interactions between particular rivals matter.
* The decision to enter a new foreign market in the era of global consolidation is related to the identities of rivals in the market, characteristics of the firm and the market, and the extent of past competitive interactions with the international pioneering firm.
* Results suggest that, in an oligopolistic environment, aspects of multimarket competition are important to foreign direct investment decisions.
Keywords: Oligopolistic reaction. Firm-specific interactions' Multimarket competition-Pioneer vs. late-mover. FDI
Assume that companies A and B are active players in a global oligopoly. Company A is considering adding a new country to its foreign direct investment (FDI) portfolio. Company B already has an established subsidiary in that country. If the two companies have competed with each other repeatedly in other markets, does this affect Company A's decision regarding the potential new investment?
A large global company that operates in an oligopolistic industry tends to compete with its primary rivals in many different locations. We analyze the relationship between a firm's past interactions with particular rivals and its decisions regarding new FDI locations, using the frameworks of multimarket contact, oligopolistic reaction and herding behavior, first-mover advantages, and FDI. Our research contributes to the literature by incorporating a consideration of interaction histories with specific rival firms, investigating the phenomenon of how firms compete aggressively with particular rivals while remaining relatively passive with respect to others.
Examining the FDI behaviors of leading firms in the world tire industry, we consider factors related to the location of foreign subsidiaries. Binomial logistic regression modeling provides evidence that strategic behaviors differ, depending on both the presence of individual rival firms and the history of interactions between firms. Our results suggest that rivals pay particular attention to the investment decisions of, and past interactions with, the international pioneer, which was the first firm to establish an extensive network of international subsidiaries. The international investment behaviors of large tire companies imply a consideration of the extent of competition in individual markets. These insights into the strategies of the major firms in an important global industry enrich the literature related to international competition.
This paper is organized as follows. In the next section, we review the literature on FDI and multimarket contacts, in the context of oligopolistic reaction, herding behavior, and first-mover advantages. We also provide a brief background on the history of consolidation in the global tire industry. The following section describes our modeling approach, along with the data used in the logistic regressions. The results are then discussed, and the paper concludes with a summary and discussion of our findings.
FDI and Multimarket Contact
Bases of the FDI Literature
Nearly half a century of research has provided some important insights into why, how, where, and when firms may be expected to undertake FDI. Much of the theoretical underpinning associated with FDI is based on the work of Hymer (1960/1976), who noted the relationship between competitive conditions and foreign investment, and argued that foreign market entry must be supported by firm-specific or ownership advantage that allows the investing company to compete successfully against local firms, overcoming domestic firms' inherent advantages stemming from their superior knowledge of conditions in the host country. This argument forms the basis of our understanding of why FDI occurs (Pitelis 2006).
Internalization is the foundation for perspectives on how FDI takes place. This line of thought builds on Coase's (1937) work related to transaction costs and the theory of firms, which asserted that firm size is expected to expand until the costs associated with internally-organized transactions outweigh the benefits. Considering international operations, Buckley and Casson (1976) extended both the transaction costs and internalization frameworks, arguing that multinational enterprises will generally prefer transactions to be effected internally, rather than through external markets.
Our understanding of where FDI occurs has its foundation in Dunning's (1979) eclectic theory of international production. In the "OLI" framework, Dunning (1979) built on international trade theory to incorporate a consideration of location (L), in addition to the existing factors of ownership advantage (O) and internalization (I). The crux of this argument is that FDI locations are selected based on host countries' location-specific endowments that have the potential to facilitate the multinational's efforts to create value.
The fundamental work on temporal aspects of FDI, or when it is expected to take place, draws on Vernon (1966) and Johanson and Vahlne (1977). Vernon's (1966) product life cycle addressed patterns of international trade and investment in the context of national comparative advantage, dealing with issues of the introduction and evolution of products and technology diffusion. In this framework, international operation is viewed as sequential, based on life cycles at the product level. Focusing on psychic distance, the stages model of Johanson and Vahlne (1977) considers international expansion to be an incremental process, with increased commitment to a market developing on the basis of sequential gains in market-specific knowledge and experience.
These foundations of the FDI literature, along with the subsequent research that builds on them, have provided insights into important aspects of FDI. Of course, many aspects of this complex phenomenon remain understudied; one such aspect is the issue of competitive interactions, or how multinational firms compete with each other.
In an oligopolistic rivalry, firms recognize their mutual interdependence. Absent a collusive carving up of the world market, investing abroad eventually yields multimarket contact among international oligopolists. The literature indicates that the ensuing head-to-head competition does not always result in lower profits. Bernheim and Whinston (1990) analyzed the relationship between multimarket contact and firms' abilities to sustain noncompetitive or collusive behavior, suggesting that oligopolists with multimarket contacts can create particularly severe punishments (e.g., lower profits) to single-market competitors, enhancing collusive outcomes. Klemperer's (1992) model indicated that, when consumers are brand-loyal or have high switching costs, direct matching of rivals' product offerings results in less competition and higher prices, compared with carving up the product portfolio.
Empirically, Scott (1982, 1989, 1991) identified a relationship between multimarket contact and performance, based on lines of business for U.S. companies. These findings are consistent with the "mutual forbearance" hypothesis (Edwards 1955), which postulates that firms meeting in multiple markets can anticipate each others' potential reactions and generate credible threats of competitive retaliation. When an industry experiences global consolidation, the emerging international oligopolists should find themselves facing an environment of multimarket contacts. One of the behaviors often found in this type of competitive environment is oligopolistic reaction.
"Herding" behavior emerges when companies move in groups, e.g., choosing to invest in the same country over a short time span. Knickerbocker (1973) noted early empirical evidence of the oligopolistic reaction phenomenon among multinational firms. These results suggested that the extent of foreign investment is related to the form of the mutual interdependence among rival firms, such that firms in a domestic oligopoly tend, rather quickly, to follow their rivals' overseas investments.
Following the methodology used by Ball and Tschoegl (1982), Yu and Ito (1988) analyzed the FDI decisions of U.S. companies, and found the number of rivals in a host country to be positively related to the probability of an oligopolistic firm's investment in that country. Ito and Rose (2002) expanded the geographic scope of the research to include multiple home countries, and noted that oligopolists from different countries tend to locate their FDI similarly to that of their global rivals, particularly when a first-mover is involved. Their results suggest that both the number and identities of global rivals are related to a firm's FDI location decisions.
Models of Herding Behavior
Within the FDI literature, research pertaining to the phenomenon of oligopolistic reaction has shed some light on the investment behaviors of multinational firms. However, most of these studies have been empirical in nature, so a theoretical framework to explain oligopolistic reaction and herding behavior has not been fully developed. Some game-theoretic models from other areas of the literature provide some insight into imitative behavior among rivals.
For example, theoretical models in the spatial competition literature predict three different types of firm behavior. Firstly, a firm may choose to locate itself in close proximity to its rivals, in order to be competitive and capture more customers. Secondly, a firm may opt to distance itself spatially from its rivals, with the intention of reducing head-to-head competition and increasing profits. Thirdly, a firm may make its location decisions independently, regardless of its rivals' locations. However, the third option seldom dominates, which generally leads to predictions of either herding behavior or avoidance (Borenstein/ Netz 1999). While many models find that a firm's optimal strategy is to differentiate itself from its rivals, others suggest that imitation can be a better approach (Kennedy 2002). Theoretical model predictions depend critically on assumptions about price competition, heterogeneity of demand, the number of rivals, entry order, fixed costs, risk preferences, and managerial reward systems.
With a few exceptions (e.g., Netz/Taylor 2002), empirical work in the spatial competition literature has found evidence of clustering, similar to the results seen in FDI research. However, neither literature has given sufficient consideration to the relationship between location decisions and the identities--and history of interactions--of rival firms. In reality, FDI decisions are based on more than just the absolute number of firms operating in a market. Barron and Valev (2000), for example, suggested that smaller banks tend to follow larger ones overseas. In order to add this important dimension of competitor identity, we incorporate the notion of first-mover advantage.
Pioneers and Late-Movers
Some firms are early investors into markets, while others are latecomers. The theory of first-mover advantage posits that the first firm to enter a market may be able to gain advantage by preempting its rivals in acquiring scarce assets or by positioning in space (Lieberman/Montgomery 1988). In terms of international investment, early entrants have natural advantages with regard to market share in the focal country. However, the potential exists for the first-mover firm to enjoy additional advantages. These include obtaining new capabilities in technology, resources, management skills, and knowledge, all of which may be unfamiliar to the first-mover's rivals. If the new capabilities enable the first-mover to upset the competitive equilibrium of the oligopolistic industry at home and abroad, its rivals are left at a relative disadvantage.
Oligopolistic reaction implies that, when a first-mover establishes a subsidiary in a country, rivals can counter the threat to their market positions by matching the investment. This should result in the first-mover's being followed into foreign markets by other firms. In this way, a "pioneer" company, which has established an early global presence by being an international first-mover, becomes a target in global competition. (1) While the strategies of pioneers are in stark contrast with those of "late-movers," entry timing is not always completely subject to managerial choice; for example, firms with weaker innovative capabilities may be forced into later entry (Lieberman/Montgomery 1998).
Global competition often involves a small number of firms competing against each other across different markets. Intensive competition may lead to industry consolidation, forcing the remaining firms into even more careful examinations of the extent of their global market coverage through international subsidiaries. In this process, late-movers may have additional information on which to base their location decisions, given the experience gained through their history of taking on the pioneer in various markets. In addition to the fact that the pioneer is more likely to be incumbent in a potential new market, and has thus signaled the market's value to competitors (e.g., Bhardwaj/Dietz/Beamish 2007), a rival's familiarity with the process of competing against the pioneer is expected to facilitate its entry decision. This leads to our first hypothesis:
Hypothesis 1: The presence of the pioneer in a market is positively related to the probability of a rival firm's entering that market.
In the context of an international oligopoly, the nature of past interactions between pairs of rivals--pioneers or not--may also be an important factor in market entry decisions. Consider the situation of a late-mover contemplating an initial investment into a country in which another firm has already established a subsidiary. If the incumbent rival has a history of being a strong competitor, dominating the markets in which both firms have invested, the late-mover may be tempted to locate its subsidiary elsewhere, to avoid head-to-head competition with this particular firm. On the other hand, if the incumbent rival has not made full use of its first-mover advantages in the firms' prior meetings, then the late-mover may also be more likely to invest in that country. (2) However, the late-mover may choose to invest, even in the face of previous success by the first-mover:
... what follows from this first step can be a reaction of rivals that has much less to do with these firms' efforts to exploit their own proprietary advantages than their efforts to keep the pioneer from gaining even more advantage, and this reaction can entail FDI (Graham 1998, p. 81).
If the late-mover and the rival have not met extensively in the past, the two firms may not have had much interaction in their strategically important markets, and the late-mover may not view the incumbent as a crucial rival. In that case, the fact that the incumbent is operating in markets in which the late-mover does not have a presence may not particularly bother the late-mover. On the other hand, if the two firms have met extensively in other markets, the late-mover is likely to acknowledge the other as an important rival; as an extreme, a firm that competes with another firm in 99 of 100 markets is expected to view its interaction with that competitor as more important than its rivalry with a firm that it meets in only 1 of 100 markets.
Of course, market presence need not be all-encompassing. Karnani and Wernerfelt (1985) suggested that having a mutual foothold is a feasible strategy; that is, a firm may opt for a small presence in a market dominated by a rival. For example, when a rival firm enjoys high sales volume in a market and the focal firm does not, it is relatively inexpensive for the focal firm to attack in the market and relatively expensive for the rival firm to defend against the attack, making the attacking strategy rather attractive to the focal firm. On this basis, when the incumbent and the late-mover have had extensive meetings elsewhere, the presence of the incumbent and the absence of the late-mover in a market is expected to induce investment by the latter firm.
The identity of the incumbent should matter; we expect rivals' reactions to moves by pioneers and by late-movers to be quite different. Late-movers are more likely to have relatively recent histories as local or regional firms, typically operating closer to their home markets (e.g., Rugman 2005). Lacking the extensive network of international subsidiaries enjoyed by the pioneer, they are not apt to be such natural targets for chasing by rivals in the world market. Considering the relationships between the pioneer and all of the other major players in a global oligopoly, we hypothesize the following:
Hypothesis 2: More extensive past multimarket contact with a pioneer is related to a higher probability of a firm's entry into a new market in which the pioneer has a subsidiary.
Consolidation and Global Competition in the Tire Industry
The nature of the world tire market has changed dramatically during the past few decades. Before the 1980s, the tire industry was essentially multidomestic, with competition occurring on a country-by-country basis. By the late 1980s, the world tire industry had been transformed, and was dominated by a handful of large multinational firms. By analyzing investment positions at different times during this rapid industry consolidation, we can trace the different competitive strategies used by firms at various stages of the industry's evolution.
While tires are fairly simple products, they are not commodity items. The tire industry's consolidation was facilitated, in part, by the new radial tire technology perfected by Michelin, and the fact that some firms could not keep pace with the technological advance. The new steel-belted radial tires offered more than twice the average running life, compared with the bias-ply tires that had been the previous standard. In addition, the new tires provided greater safety, more sensitive road handling, and superior rolling efficiency. By the late 1970s, the quality gap between Michelin and its rivals, with respect to the radial tire technology, had been narrowed, increasing competition. The shift to radial tires in the 1970s and 1980s reduced the demand for replacements, resulting in excess capacity that precipitated price wars and plant closures (Scherer 1992), and led to a major international consolidation.
International mergers and acquisitions (M&As) in the tire industry during the 1980s involved companies from different countries. In 1986, Sumitomo Rubber of Japan acquired control of the U.K.-based Dunlop Tire. In that same year, Uniroyal (U.S.) and Goodrich (U.S.) merged their tire manufacturing operations and created Uniroyal/Goodrich, which was then acquired by Michelin (France) in 1989. Meanwhile, Continental Reifen AG (Germany) acquired General Tire (U.S.) in 1987. In 1988, Bridgestone, the leading Japanese tire producer, acquired Firestone for US $2.6 billion, and Pirelli (Italy) acquired Armstrong Tire (U.S.), leaving Goodyear as the only major U.S. tire manufacturer competing internationally. Pirelli tried to acquire Continental in the early 1990s, but was unsuccessful. In 1999, Goodyear formed a strategic alliance with Sumitomo Rubber (Japan). After all of the merger activity, the five major players in the world tire market were from France (Michelin), Germany (Continental), Italy (Pirelli), Japan (Bridgestone), and the United States (Goodyear). The estimated combined world market share of these five firms was over 75 percent by the early 1990s. The international M&As within the world tire industry provide an environment in which we can investigate the nature of global competition, while tracing the identities of individual rival firms.
Within the highly concentrated industry, one might have expected a large degree of tacit cooperation. However, this was not the case in the tire industry throughout the 1980s. Instead, price competition was keen. Tire manufacturers were subjected to "an almost uninterrupted series of price buffetings" (Scherer/Ross 1990, p. 281) and profits were lower, on average, than those of other industries with comparable market structures.
From early in the 20th century, Goodyear had assumed the role of the global pioneer, having been quick to establish an extensive network of international operations in the tire industry:
The world's largest tire producer since 1916, Goodyear became the world's largest rubber producer by 1926. By 1928, the company operated in 145 countries ... (Pascal/Barbour/Griffin 1992, p. 244).
In contrast, as of 1930, Michelin was only the 17th-largest tire company in the world (Pascal/Barbour/Griffin 1992), and Bridgestone was not yet incorporated. By 1980, prior to the start of the industry's consolidation, Goodyear had foreign subsidiaries in 33 countries. Its nearest rival, in terms of the number of international locations, was Continental, with foreign subsidiaries in 13 countries.
Model and Data
We are interested in factors that are related to the tire firms' decisions to undertake FDI into a new country, considering three time periods, with start/end dates of 1982/1987, 1987/1992, and 1982/1992. Models are estimated for each of the three time periods, using the firms' investment positions at the end of each period (i.e., 1987 or 1992), compared to the start (i.e., 1982 or 1987). The factors incorporated to explain FDI location decisions include the identities of rivals, competitive interactions, and aspects of both the host country and the investing firm. The three years (1982, 1987, and 1992) used to define the time periods represent very different stages in the tire industry's evolution. Due to rapid consolidation starting in the 1980s, the global tire industry went from 10 major players in 1987 to five by 1992.
The dependent variables for the various models are binary, indicating whether or not the firm had established a foreign subsidiary by the end of the period, in a country in which it did not have a subsidiary at the start of the period (e.g., 1987 and 1982, respectively, in the models for the 1982/1987 time period). Because the dependent variables are dichotomous, we use binomial logistic regression models of the form:
P[[y.sub.i] = 1] = [[1 + [e.sup.(-[alpha]-X[beta])].sup.-1].
When a firm has a subsidiary at the end of the period, in a country in which it did not have a presence at the start of the period, the dependent variable, [y.sub.i], assumes the value of 1; otherwise [y.sub.i] = 0. This measure incorporates all new initial investments, including those established through M&A activity. The vector of explanatory variables for the ith observation is X; [alpha] is the intercept parameter, and [beta] is the vector of coefficient parameters. A positive and significant estimated coefficient implies that an increase in the value of the explanatory variable is associated with an increased probability of the firm's having established a subsidiary in a new country during the period under consideration.
The explanatory variables in our models are intended to provide insight into how firms factor information about, and past experience with, rivals into their FDI location decisions. Five explanatory variables reflect the identities of rivals; BRIDGESTONE, CONTINENTAL, GOODYEAR, MICHELIN, and PIRELLI are indicator variables that assume the value of I if the respective firm had a subsidiary in the host country two years before the start of the period (i.e., 1980 and 1985), and 0 otherwise. Because Goodyear is the FDI pioneer in the global tire industry, we expect the coefficient associated with GOODYEAR to be positive, in support of Hypothesis 1.
Another set of variables is used to reflect past interactions with specific rivals. These variables, [INT.sub.j], for j = Bridgestone (BS), Continental (CO), Goodyear (GY), Michelin (MI), and Pirelli (PR), are constructed as follows (3):
[INT.sub.ijm] = [D.sub.jmt] [summation over (m)] [D.sub.im](t-5)[D.sub.jm](t-5),
where t is the year, m is a given country, i is the focal firm, and j is a firm that competes with firm i in country m. The dummy variables [D.sub.im] and [D.sub.jm] assume the value of 1 if firm i (j) has a subsidiary in country m, and 0 otherwise. Thus, [INT.sub.BS], [INT.sub.CO], [INT.sub.GY], [INT.sub.MI] and [INT.sub.PR] represent the total past interactions between the focal company and Bridgestone, Continental, Goodyear, Michelin, and Pirelli, respectively.
For example, Bridgestone and Goodyear met in 10 countries in 1980. Consider a country that represented a potential new market for Bridgestone in 1985. If Goodyear was present in the country, [INT.sub.GY] = 10; if Goodyear did not have a presence in the country, then [INT.sub.GY] = 0. Thus, [INT.sub.j] measures the extent of prior interaction with a particular rival firm incumbent in a focal market. We anticipate that these firm-specific indicators of the history of multimarket interactions between pairs of firms will help to explain the probability of new investment. Because Goodyear is the global tire industry's FDI pioneer, we expect the coefficient associated with [INT.sub.GY] to be positive, consistent with Hypothesis 2.
The FDI literature suggests both host country- and firm-related attributes that must be controlled for in our models, in order to permit an examination of the marginal explanatory power of the variables that account for competitive interactions. Location-specific attractiveness (e.g., Dunning 1998) is an obvious example; some markets are inherently more attractive than others. In addition to the possibility that an external event may make particular markets more appealing to all competitors (Caves 1996), larger markets should, ceteris paribus, be more attractive than smaller ones, permitting more firms to operate profitably. This applies both internationally and domestically; in a U.S.-based study, Bresnahan and Reiss (1990) found that market size was a particularly strong predictor of the number of new automobile dealerships across different regions. The FDI literature includes numerous measures of a host country's location-specific attractiveness, including widely-employed ones such as political risk, tax rates, and gross domestic production (GDP).
In addition, proximity--both geographic and environmental--between the investing firm's home country and the potential host country is consistently viewed as important to FDI location decisions (e.g., Johansen/Vahlne 1977). Generally, investing in a more geographically distant market creates more complicated situations with respect to logistics, communication, language, and cross-cultural management, leading to higher costs associated with monitoring and controlling more distant subsidiaries.
We employ four control variables to capture location-specific factors. The host country's political risk is represented by its credit rating, COCREDIT. Countries with higher credit ratings are presumed to present safer investment environments, so we expect to observe a positive relationship between COCREDIT and the propensity to invest in that country. TAX is the corporate tax rate in the host country; higher tax rates are expected to be associated with a lower probability of investment. GDP is the host country's gross domestic production, representing market size; we anticipate a positive relationship between the host country's GDP and the probability of investment. DISTANCE is the air distance between the capital cities of the firm's home country and the host country. For Goodyear, the distance is measured to the host country capital from the closest U.S. city, among New York, Miami, San Francisco, and Los Angeles. We expect the estimated coefficient associated with the DISTANCE variable to be negative.
We also control for firm-level attributes, following Hymer's (1960/1976) observation that foreign firms must possess advantages over local firms, in order to compete effectively in the unfamiliar host country environment. Size is an advantage that investing firms may have over their local rivals (Horst 1972), and firm size is often used as a proxy for firm-specific advantages in the empirical FDI literature; we include ASSETS, which is the firm's total assets, in U.S. dollars (4), to represent firm size, and expect to observe a positive relationship between this variable and the propensity to invest. Another control variable included to incorporate firm-specific attributes is SUBS, the total number of foreign subsidiaries owned by the parent firm, which we also anticipate to be positively associated with the dependent variable. The third firm-specific variable identifies new subsidiaries resulting from the M&A activity that created the industry's consolidation. The past interactions variables ([INT.sub.ijm]) described earlier do not distinguish between actions initiated by the surviving firm and expansion that resulted from M&As in which one firm obtained the investment pattern of another. To differentiate between the firm's own choices and those it inherited, we use two approaches. Firstly, we incorporate a variable to identify expansions that are due strictly to acquisition. The dummy variable M&A assumes the value of 1 if the acquiring firm did not have a subsidiary in the country prior to the acquisition, but the acquired firm did; otherwise M&A = 0. Because one of the motivations for international mergers is the expansion of the acquiring firm's geographic portfolio, we expect the coefficient associated with this variable to be positive. Secondly, we re-estimate the models using the subset of the data for which M&A = 0, so that we consider only new market entry decisions that were not the result of M&A activity.
Table l summarizes the variables in the study, the data sources, and the expected signs of the associated coefficients. We use natural logarithm transformations for GDP, DISTANCE, ASSETS, and SUBS. In addition to providing more stability in the modeling, this approach incorporates a consideration of a decreasing marginal effect. Because of the relatively high correlations between some of the firm dummy variables and the firm-specific past interaction variables ([INT.sub.j]), models are estimated separately for each time period, with the two groups of variables. All of the explanatory variables, with the exception of the M&A indicator, are evaluated two years prior to the end of the period under consideration. (5)
Cross-sectional models are estimated separately for the three different timeframes: developments between 1982 and 1987, 1982 and 1992, and 1987 and 1992. This approach allows for the consideration of market entries in the pre- to mid-consolidation (1982/1987), mid- to post-consolidation (1987/1992), and pre- to post-consolidation (1982/1992) periods. Our sample includes data for the five largest global tire manufacturers--Bridgestone, Continental, Goodyear, Michelin, and Pirelli--for international markets in which at least one of the five firms had a subsidiary in at least one of the years of interest. Table2 shows the correlation matrices and descriptive statistics for the variables in our study, considering the full sample in each period. Despite some fairly high pairwise correlations, adequately low variance inflation factors indicate that multicollinearity does not have a serious impact on the estimation process.
Presence of Rivals
The outcomes of our modeling are shown in Tables 3, 4, 5 and 6. Tables 3 and 4 contain results for models that include the identities of incumbent rivals in each market, including and omitting M&A activity, respectively; these models are used to test Hypothesis 1. Tables 5 and 6 show the results for comparable models that incorporate a consideration of past interactions with incumbent rivals; these results are used to test Hypothesis 2. For each configuration, we estimate models using the full dataset. In addition, because host countries differ in terms of importance, based on location-specific factors, we stratify the data by market size and estimate each of the models separately for large and small countries; stratification is based on median GDP in 1980 or 1985, as applicable.
In Table3, the coefficients associated with GOODYEAR are positive and significant (p < 0.05) in models estimated using both the full sample and the larger-countries subsample for timeframes that include the post-consolidation period (1987/1992 and 1982/1992). These results provide strong support for Hypothesis 1, that the presence of the pioneer is associated with a higher probability of a firm's having a subsidiary in a new market, after controlling for measures of market attractiveness. The estimated coefficients associated with GOODYEAR are not significant for any of the smaller-country subsets, suggesting that the "Goodyear effect" in the post-consolidation period is driven by larger--and presumably more important--host markets. In addition, the presence of Bridgestone in a market is positively (p < 0.10) associated with the probability of investment in larger countries for the 1982/1987 period, while Michelin's presence is positively 09 < 0.05) associated with the probability of investment for both the full and larger-countries samples in the 1982/1992 period.
Table4 shows comparable results, having omitted observations with M&A involvement during each of the timeframes. Comparing Tables 3 and 4 indicates that the results are quite similar, with or without the M&A subsidiaries. The coefficients associated with GOODYEAR are positive (p < 0.05) for both the full and large-country samples that incorporate the post-consolidation period, providing additional strong support for Hypothesis 1. Interestingly, Hypothesis 1 receives no support in the context of smaller markets, across any of the timeframes. In addition, the coefficient associated with BRIDGESTONE is positive and significant (p < 0.10) for the full and large-country samples in 1982/1987, while the coefficient associated with the MICHELIN variable is significantly negative (p < 0.05) for the small-country subsample for the 1982/1992 period.
In the models used to test Hypothesis 1 (Tables 3 and 4), country-specific factors provide some explanatory power. Geographical distance (In DISTANCE) is negatively associated (at least p < 0.10) with the probability of new market entry in four of the models, while the coefficients associated with the political risk variable (COCREDIT) are positive and significant (at least p < 0.10) in 11 of the 17 models. These latter results suggest that issues associated with political stability may have dominated other location-specific attributes, including market size and tax rate, in the market entry decisions of global tire companies during this time period. Our findings differ from those reported in the bulk of the FDI location literature, which has generally identified the size of the market as an important variable. It may be that, by the later stages of the industry's international consolidation, the surviving firms had already established themselves in the most crucial foreign markets, leading to different decision-making processes with regard to ensuing market entries.
The results for the three firm-specific variables are somewhat surprising. In contrast to much of the FDI literature (e.g., Caves 1996), company size (In ASSETS) adds consistent explanatory power only for the large-country samples, after controlling for the other explanatory variables. This is probably a function of the specialized nature of our sample, which consists of only the largest tire manufacturers. In this elite group, size may not be as distinguishing a factor, after incorporating other competition- and location-related effects. The coefficients associated with the number of subsidiaries (In SUBS) are positive and significant (at least p < 0.10) for the full and smaller-country samples in the models in the 1987/1992 and 1982/1992 periods. Rather unexpectedly, the M&A variable contributes little explanatory power; in Table 3, the marginal probability of market presence is significantly higher when the subsidiary has been acquired (p < 0.10) only for the small-country subsample in the 1982/1987 period, and it is significantly lower (p < 0.05) for larger countries in 1982/1992. In addition, the generally similar estimated models between the full sample (Table3) and the non-M&A subsample (Table4) suggest that acquiring firms were not routinely expanding their foreign subsidiary portfolios based heavily on the holdings of the acquired firms. We conjecture that the international tire firms had already established subsidiaries in key markets prior to the industry's consolidation. The M&A activity may have been motivated more by expansion of operational scope in those existing markets, rather than by the rapid expansion of geographical portfolios. Entry into new markets by the five large international players was driven by new investments, rather than acquisitions.
Past Multimarket Contact with Rivals
Replacing dummy variables representing the presence of rivals in market with variables representing company-specific prior multimarket contact yields remarkably similar decisions. In Table 5, the Goodyear-related variable ([INT.sub.GY]) provides strong explanatory power, with significant and positive (at least p < 0.10) estimated coefficients for both the 1987/1992 (full sample and larger countries subsample) and 1982/1992 periods (all three samples). The finding that more extensive prior multimarket contact with Goodyear is associated with a higher probability of entry into markets by 1992 provides support for Hypothesis 2. In contrast, none of the coefficients associated with [INT.sub.CO], [INT.sub.MI], and [INT.sub.PR] differs significantly from zero, and [INT.sub.BS] provides significant (p < 0.10) explanatory power only for larger countries for the 1982/1987 period, providing a clear distinction between the pioneer and later-movers.
The results shown in Table 6, for models estimated without M&A-associated subsidiaries, are quite consistent. More extensive prior multimarket contact with Goodyear (represented by the variable [INT.sub.GY]) is positively associated (at least p < 0.05) with the probability of investment in both the full and the larger-country samples for the 1987/1992 and 1982/1992 periods, providing support for Hypothesis 2. Among the other companies, only Bridgestone and Michelin show any significant explanatory power (each in one model, with p < 0.10).
Among the country-specific variables, political risk (COCREDIT) remains a useful contributor to explaining market entry decisions, with positive and significant (at least p < 0.10) coefficients in 10 of the 17 models in Tables5 and 6. Distance also adds some limited explanatory power, with negative and significant (at least p < 0.10) estimated coefficients in three of the models for the data including M&A-generated entries.
The three firm-specific variables provide similar decisions to those from Tables3 and 4. While company size (In ASSETS) adds some explanatory power, with a positive association with the probability of investment, the result is not consistent across the full complement of models; this is also the situation for the number of subsidiaries (In SUBS). Again, the M&A variable provides little marginal explanatory power, with coefficients that are positive (p < 0.10) for smaller countries in 1982/1987 and negative (p < 0.05) for larger countries in the 1982/1992 period.
Summary and Conclusions
Much of the literature on oligopolies has been developed in the context of domestic competition. The situation is more complicated when the oligopolists are based in different countries. Internationally operating oligopolists from a variety of home countries face more difficulty in developing an implicit consensus of who controls which markets. Particularly in the earlier stages of globalization (e.g., prior to an industry's consolidation), firms are not completely familiar with each other, increasing the potential for apprehension about the activities of poorly-understood foreign competitors (Bernheim/Winston 1990). Oligopolistic consensus, with the goal of maximizing joint profits, is more readily facilitated after the firms gain mutual familiarity through repeated investments in the same markets, and particularly after industry consolidation.
Past research related to multimarket competition has increased our understanding of the nature of competitive behavior among rival firms. Generally, oligopolistic firms have been found to establish subsidiaries in countries in which more of their rivals have operations. This study adds to the literature by showing that both the identities of rivals incumbent in markets and the histories of interactions between rival firms are important in subsidiary location decisions, after controlling for other factors suggested in the FDI literature. Our results suggest that the extent of prior multimarket contact with a particular rival may affect the likelihood that a firm will undertake a particular new international investment, and that the nature of the effect differs according to the identity of the competitor.
With respect to new investment decisions, we find that the impact of prior contact with Goodyear, the pioneer with respect to global presence, differs from that of prior contact with other rivals. More extensive prior multimarket contact with the pioneer is associated with a higher probability of entry into a markets in which Goodyear is resident, by the end of the industry's consolidation. This effect is not observed for the other firms. Thus, consideration of multimarket contacts with specific rivals reveals company-specific strategies.
Our results are also consistent with Graham's (1998) observation that late-movers may enter markets inhabited by the pioneer, even in the absence of superior products. After the shift to radial tires created excess industry capacity, tire firms expanded their international subsidiary networks. Had the late-movers not undertaken investments to match those of the pioneer, Goodyear might have gained even more advantages in international markets.
This paper sheds some light on strategy in the context of global industries. The definition of a global industry as one in which a firm's competitive position in one country is substantially affected by its position in other countries is widely employed. However, the more detailed, firm-level strategies within global industries have not been well documented in the oligopolistic rivalry and FDI literatures. The frameworks of multimarket contact, oligopolistic reaction and herding behavior, first-mover advantages, and FDI have strong applicability for understanding factors important to foreign subsidiary location decisions.
Acknowledgements: We acknowledge research support from the CIBER at the University of Hawai'i at Manoa, and are extremely grateful for the suggestions of the editors and two anonymous reviewers.
DOI 10.1007/s 11575-009-0010-y
Received: 21.04.2007 / Revised: 12.02.2008 / Accepted: 16.04.2008 [c] Gabler-Verlag 2009
Ball, C. A./Tschoegl, A. E., The Decision to Establish a Foreign Bank Branch or Subsidiary: An Application of Binary Classification Procedures, Journal of Financial and Quantitative Analysis, 17, 3, 1982, pp. 411-424.
Barron, J. M./Valev, N. T., International Lending by U.S. Banks, Journal of Money, Credit, and Banking, 32, 3, 2000, pp. 357-381.
Baum, J. A. C./Korn, H. J., Dynamics of Dyadic Competitive Interaction, Strategic Management Journal, 20, 3, 1999, pp. 251-278.
Bernheim, B. D./Whinston M. D., Multimarket Contact and Collusive Behavior, Rand Journal of Economics, 21, 1, 1990, pp. 1-26.
Bhardwaj, A./Dietz, J./Beamish, P. W., Host Country Cultural Influences on Foreign Direct Investment, Management International Review, 47, 1, 2007, pp. 29-50.
Borenstein, S./Netz, J., Why Do All the Flights Leave at 8 am?: Competition and Departure-Time Differentiation in Airline Industries, International Journal of Industrial Organization, 17, 5, 1999, pp. 611-640.
Bresnahan, T. F./Reiss, P. C., Entry in Monopoly Markets, Review of Economic Studies, 57, 192, 1990, pp. 531-553.
Buckley, P. J./Casson, M., The Future of the Multinational Enterprise, London: Macmillan 1976.
Caves, R. E., Multinational Enterprise and Economic Analysis, 2nd ed., Cambridge, United Kingdom: Cambridge University Press 1996.
Coase, R. J., The Nature of the Firm, Economica, 4, 16, 1937, pp. 386-405.
Cool, K./Chahid-Mourai, B., The Worm Tire Industry in 1993, Continental Gummi Werke, Fountainebleau, France: INSEAD-CEDEP 394-118-1 1994.
Dunning, J. H., Explaining Changing Patterns of International Production: In Defense of the Eclectic Theory, Oxford Bulletin of Economics and Statistics, 41, 4, 1979, pp. 269-96.
Dunning, J. H., Location and the Multinational Enterprise: A Neglected Factor?, Journal of International Business Studies, 29, 1, 1998, pp. 45-66.
Edwards, C. D., Conglomerate Bigness as a Source of Power, in Universities-National Bureau Committee for Economic Research (ed.), Business Concentration and Price Policy, Princeton: Princeton University Press 1955.
Graham, E. M., Market Structure and the Multinational Enterprise: A Game-theoretic Approach, Journal of International Business Studies, 29, 1, 1998, pp. 67-83.
Horst, T., Firm and Industry Determinants of the Decision to Invest Abroad: An Empirical Study, Review of Economics and Statistics, 54, 3, 1972, pp. 258-266.
Hymer, S. H., The International Operations of National Firms: A Study of Direct Foreign Investment, Cambridge, Massachusetts: MIT Press 1960 (1976).
Institutional Investors, September 1980, 1985, 1990.
International Air Transport Association and International Aeradio, Air Distances Manual, Geneva: International Air Transport Association 1994.
International Monetary Fund, International Financial Statistics Yearbook, Washington, D.C.: International Monetary Fund 1993.
Ito, K./Rose, E. L., Foreign Direct Investment Location Strategies in the Tire Industry, Journal of International Business Studies, 33, 3, 2002, pp. 593-602.
Johanson, J./Vahlne, J.-E., The Internationalization Process of the Firm: A Model of Knowledge Development and Increasing Foreign Market Commitments, Journal of International Business Studies, 8, 1, 1977, pp. 23-32.
Karnani, A./Wernerfelt, B., Multiple Point Competition, Strategic Management Journal, 6, 1, 1985, pp. 87-96.
Kennedy, R. E., Strategy Fads and Competitive Convergence: An Empirical Test for Herd Behavior in Prime-Time Television Programming, Journal of Industrial Economics, 50, 1, 2002, pp. 57-84.
Klemperer, P., Equilibrium Product Lines: Competing Head-to-Head May Be Less Competitive, American Economic Review, 82, 4, 1992, pp. 740-755.
Knickerbocker, F. T., Oligopolistic Reaction and Multinational Enterprise, Boston: Division of Research Graduate School of Business Administration, Harvard University 1973.
Lieberman, M. B./Montgomery, D. B., First-Mover Advantages, Strategic Management Journal, 9, Special Issue, 1988, pp. 41-58.
Lieberman, M. B./Montgomery, D. B., First-Mover (Dis)Advantages, Strategic Management Journal, 19, 12, 1998, pp. 1111-1125.
Moody's Investors Service, Moody's International Manual, New York: The Service (various years).
Netz, J./Taylor, B. A., Maximum or Minimum Differentiation? An Empirical Investigation into the Location of Firms, Review of Economics and Statistics, 84, 1,2002, pp. 162-175.
Pascal, D. B./Barbour, P. A./Griffin, J. (eds.), International Directory of Company Histories, Volume V, Detroit and London: St James Press 1992.
Pitelis, C., Stephen Herbert Hymer and/on the (Theory of the) MNE and International Business, International Business Review, 15, 2, 2006, pp. 103-110. Price Waterhouse, Corporate Taxes: A Worldwide Summary, New York: Price Waterhouse Center for Transnational Taxation (various years).
Rugman, A. M., The Regional Multinationals. MNEs and "Global" Strategic Management, Cambridge, United Kingdom: Cambridge University Press 2005. Scott, J. T., Multimarket Contact and Economic Performance, Review of Economics and Statistics, 64, 3, 1982, pp. 368-375.
Scott, J. T., Purposive Diversification as a Motive for Merger, International Journal of Industrial Organization, 7, 1, 1989, pp. 35-47.
Scott, J. T., Multimarket Contact among Diversified Oligopolists, International Journal of Industrial Organization, 9, 2, 1991, pp. 225-238.
Scherer, F. M., International High-Technology Competition, Cambridge: Harvard University Press 1992.
Scherer, F. M./Ross, D., Industrial Market Structure and Economic Performance, 3rd ed., Boston: Houghton Mifflin 1990.
United Nations Statistical Office, Statistical Yearbook, New York: United Nations (various years).
Vernon, R., International Investment and International Trade in the Product Cycle, Quarterly Journal of Economics, 80, 2, 1966, pp. 190-207.
Yu, C. M./Ito, K., Oligopolistic Reaction and Foreign Direct Investment: The Case of the U.S. Tire and Textile Industry, Journal of International Business Studies, 19, 3, 1988, pp. 449-460.
(1) While each country has its own first-mover, we use the term "pioneer" to represent the first company to establish an overall global presence in an industry. In the case of the tire industry, this was Goodyear.
(2) Baum and Korn (1999) analyzed the dyadic interactions of California commuter airlines, and found an inverted U-shaped relationship between firms' entry into and exit from each other's markets and the level of multimarket contact.
(3) Note that the i and m subscripts are dropped for ease of exposition.
(4) We used the reported dollar values if firms disclosed them. When dollar values were not reported, we computed them based on the average exchange rates reported by the International Monetary Fund (1993), for the years concerned.
(5) A conceptually logical addition to the control variables could involve information on automobile plants. However, FDI undertaken by tire companies does not appear to have a direct relationship with the locations in which home country automobile companies establish plants. This is probably due to the nature of the sourcing process in the original equipment market during the time we are considering. It was essentially an auction process, in which tire manufacturers were asked to submit bids for particular vehicle models. Since the car manufacturers would then pressure all firms to match the lowest bid, this system led to extremely low prices. The supply contracts typically lasted for about three years, with annual price re-negotiation (Cool/Chahid-Mourai 1994, p. 7). The short-term nature of OEM contracts meant that there was little economic justification for tire companies to follow specific auto manufacturers to new production locations.
Prof. E. L. Rose ([mail])
Department of Marketing and Management
Helsinki School of Economics, Helsinki, Finland
Prof. K. Ito
Department of Management, Shidler College of Business
University of Hawai'i at Manoa, Honolulu, HI, USA
Table 1: Variables, Expected Signs, and Sources Variable Expected Variable Description Name Sign Data Source Dependent variable: y Moody's Investors Presence of the Service focal company's subsidiary in host country Presence of BRIDGESTONE ? Moody's Investors Bridgestone in host Service country Presence of CONTINENTAL ? Moody's Investors Continental in host Service country Presence of Goodyear GOODYEAR + Moody's Investors in host country Service (Hypothesis 1) Presence of Michelin MICHELIN ? Moody's Investors in host country Service Presence of Pirelli PIRELLI ? Moody's Investors in host country Service Prior interaction [INT.sub.BS] ? Moody's Investors with, and presence Service in host country of, Bridgestone Prior interaction [INT.sub.CO] ? Moody's Investors with, and presence Service in host country of, Continental Prior interaction [INT.sub.GY] + Moody's Investors with, and presence Service in host country of, Goodyear (Hypothesis 2) Prior interaction [INT.sub.MI] ? Moody's Investors with, and presence Service in host country of, Michelin Prior interaction [INT.sub.PR] ? Moody's Investors with, and presence Service in host country of, Pirelli Political risk COCREDIT + Institutional Investor Tax rate TAX - Price Waterhouse Size of host country In GDP + United Nations Statistical Office Air distance between In DISTANCE - International Air home country and Transport host country capital Association and International Aeradio, Ltd. (1994) Firm's total assets In ASSETS + Moody's Investors Service Number of foreign In SUBS + Moody's Investors subsidiaries created Service by firm Subsidiary owned by M&A + Moody's Investors the acquired firm Service Table 2: Correlation Matrix and Descriptive Statistics (Data for the Period Between 1982 and 1987, Given no Presence in 1982) (1) y BRI CON GOOD MICH PIR y BRIDGE- 0.25 STONE CONTINEN- 0.27 0.07 TAL GOODYEAR -0.03 0.17 -0.03 MICHELIN 0.11 -0.07 0.36 -0.14 PIRELLI 0.37 0.32 0.69 -0.05 0.15 [INT.sub.BS] 0.20 0.95 0.09 0.12 -0.07 0.35 [INT.sub.CO] 0.29 0.11 0.96 -0.07 0.27 0.68 [INT.sub.GY] -0.04 0.15 -0.02 0.99 -0.12 -0.04 [INT.sub.MI] 0.14 -0.04 0.26 -0.17 0.94 0.10 [INT.sub.PR] 0.32 0.34 0.53 -0.04 0.17 0.89 COCREDIT 0.27 0.26 0.40 -0.18 0.14 0.28 TAX 0.08 -0.13 0.02 0.25 0.11 -0.04 In GDP 0.16 0.39 0.23 0.03 0.25 0.30 In DISTANCE -0.24 0.02 -0.27 0.24 0.01 -0.27 In ASSETS 0.09 -0.06 0.16 -0.10 0.04 0.09 In SUBS 0.02 0.11 -0.05 -0.27 -0.05 0.00 M&A 0.12 0.14 -0.09 0.15 0.04 0.03 y BRIDGE- 0.00 STONE CONTINEN- 0.21 -0.03 TAL GOODYEAR 0.07 0.25 -0.03 MICHELIN 0.12 -0.09 0.57 -0.12 PIRELLI 0.08 -0.03 0.39 0.32 0.37 [INT.sub.BS] 0.00 0.96 0.03 0.20 -0.08 -0.05 [INT.sub.CO] 0.18 0.01 0.88 -0.06 0.36 0.26 [INT.sub.GY] 0.17 0.26 -0.03 0.92 -0.12 0.21 [INT.sub.MI] 0.02 -0.08 0.48 -0.14 0.93 0.31 [INT.sub.PR] -0.05 -0.01 0.38 0.25 0.38 0.92 COCREDIT 0.24 0.28 0.40 -0.25 0.09 -0.01 TAX -0.01 -0.07 -0.09 0.08 0.02 0.01 In GDP 0.20 0.29 0.30 -0.02 0.25 0.23 In DISTANCE -0.03 0.17 -0.16 0.28 -0.01 0.02 In ASSETS 0.17 -0.08 0.06 0.01 -0.02 0.01 In SUBS 0.29 -0.04 0.06 -0.18 -0.01 -0.19 M&A -0.14 0.16 -0.12 0.17 -0.10 0.06 y BRIDGE- 0.00 STONE 0.07 CONTINEN- 0.29 0.07 TAL GOODYEAR 0.03 0.17 -0.03 MICHELIN 0.19 -0.07 0.36 -0.14 PIRELLI 0.26 0.32 0.69 -0.05 0.15 [INT.sub.BS] 0.04 0.95 0.09 0.12 -0.07 0.35 [INT.sub.CO] 0.23 0.11 0.96 -0.07 0.27 0.68 [INT.sub.GY] 0.03 0.15 -0.02 0.99 -0.12 -0.04 [INT.sub.MI] 0.14 -0.04 0.26 -0.17 0.94 0.10 [INT.sub.PR] 0.20 0.34 0.53 -0.04 0.17 0.89 COCREDIT 0.33 0.28 0.44 -0.22 0.15 0.32 TAX -0.01 -0.10 -0.09 0.10 0.09 -0.13 In GDP 0.27 0.42 0.31 0.02 0.21 0.37 In DISTANCE -0.16 0.02 -0.27 0.24 0.01 -0.27 In ASSETS 0.19 -0.07 0.18 0.01 0.00 0.08 In SUBS 0.19 -0.04 0.10 -0.17 0.03 0.04 M&A -0.05 0.14 -0.09 0.15 0.04 0.03 [INT.sub.BS] [INT.sub.CO] [INT.sub.GY] y BRIDGE- STONE CONTINEN- TAL GOODYEAR MICHELIN PIRELLI [INT.sub.BS] [INT.sub.CO] 0.14 [INT.sub.GY] 0.10 -0.07 [INT.sub.MI] -0.04 0.22 -0.15 [INT.sub.PR] 0.39 0.51 -0.03 COCREDIT 0.23 0.40 -0.18 TAX -0.12 0.02 0.25 In GDP 0.36 0.20 0.03 In DISTANCE 0.04 -0.36 0.27 In ASSETS -0.07 0.16 -0.11 In SUBS 0.16 0.08 -0.33 M&A 0.09 -0.10 0.13 y BRIDGE- STONE CONTINEN- TAL GOODYEAR MICHELIN PIRELLI [INT.sub.BS] [INT.sub.CO] 0.09 [INT.sub.GY] 0.24 -0.05 [INT.sub.MI] -0.08 0.29 -0.14 [INT.sub.PR] -0.03 0.30 0.14 COCREDIT 0.27 0.39 -0.21 TAX -0.06 -0.09 0.07 In GDP 0.28 0.23 -0.01 In DISTANCE 0.14 -0.19 0.26 In ASSETS -0.09 0.10 -0.01 In SUBS 0.01 0.16 0.04 M&A 0.10 -0.11 0.05 y BRIDGE- STONE CONTINEN- TAL GOODYEAR MICHELIN PIRELLI [INT.sub.BS] [INT.sub.CO] 0.14 [INT.sub.GY] 0.10 -0.07 [INT.sub.MI] -0.04 0.22 -0.15 [INT.sub.PR] 0.39 0051 -0.03 COCREDIT 0.24 0.43 -0.22 TAX -0.10 -0.09 0.10 In GDP 0.40 0.28 0.02 In DISTANCE 0.04 -0.36 0.27 In ASSETS -0.08 0.16 -0.06 In SUBS 0.03 0.12 -0.24 M&A 0.09 -0.10 0.14 [INT.sub.MI] [INT.sub.PR] COCRE- TAX In DIT GDP y BRIDGE- STONE CONTINEN- TAL GOODYEAR MICHELIN PIRELLI [INT.sub.BS] [INT.sub.CO] [INT.sub.GY] [INT.sub.MI] [INT.sub.PR] 0.15 COCREDIT 0.12 0.21 TAX 0.12 -0.08 -0.03 In GDP 0.21 0.30 0.52 -0.04 In DISTANCE -0.05 -0.15 -0.25 -0.12 -0.04 In ASSETS 0.02 0.16 0.05 -0.04 0.02 In SUBS 0.06 0.03 0.05 -0.09 0.00 M&A 0.07 0.09 0.05 0.00 0.20 y BRIDGE- STONE CONTINEN- TAL GOODYEAR MICHELIN PIRELLI [INT.sub.BS] [INT.sub.CO] [INT.sub.GY] [INT.sub.MI] [INT.sub.PR] 0.38 COCREDIT 0.05 0.02 TAX 0.02 0.02 -0.15 In GDP 0.20 0.22 0.53 -0.16 In DISTANCE -0.08 -0.05 -0.25 -0.13 -0.09 In ASSETS -0.09 -0.07 0.00 0.02 -0.06 In SUBS -0.01 -0.21 0.10 -0.04 0.01 M&A -0.10 0.09 -0.03 -0.03 0.13 y BRIDGE- STONE CONTINEN- TAL GOODYEAR MICHELIN PIRELLI [INT.sub.BS] [INT.sub.CO] [INT.sub.GY] [INT.sub.MI] [INT.sub.PR] 0.15 COCREDIT 0.13 0.25 TAX 0.10 -0.12 -0.19 In GDP 0.17 0.36 0.62 -0.14 In DISTANCE -0.05 -0.15 -0.29 -0.07 -0.11 In ASSETS -0.08 0.04 0.05 -0.02 0.00 In SUBS 0.02 -0.05 0.06 0.00 -0.03 M&A 0.07 0.09 0.02 -0.04 0.17 In DIS- In In SUBS Mean SD TANCE ASSETS y 0.08 0.28 BRIDGE- 0.13 0.34 STONE CONTINEN- 0.12 0.32 TAL GOODYEAR 0.51 0.50 MICHELIN 0.09 0.29 PIRELLI 0.10 0.31 [INT.sub.BS] 0.69 1.89 [INT.sub.CO] 0.88 2.53 [INT.sub.GY] 4.65 4.59 [INT.sub.MI] 0.58 1.90 [INT.sub.PR] 0.57 1.92 COCREDIT 49.06 24.29 TAX 0.41 0.10 In GDP 10.50 1.27 In DISTANCE 8.32 0.77 In ASSETS 0.07 14.78 0.90 In SUBS -0.12 0.14 2.91 0.31 M&A 0.06 0.06 0.00 0.13 0.34 y 0.21 0.41 BRIDGE- 0.08 0.27 STONE CONTINEN- 0.10 0.30 TAL GOODYEAR 0.51 0.50 MICHELIN 0.76 0.27 PIRELLI 0.19 0.39 [INT.sub.BS] 0.44 1.53 [INT.sub.CO] 0.65 2.24 [INT.sub.GY] 5.61 5.93 [INT.sub.MI] 0.50 1.86 [INT.sub.PR] 1.67 3.78 COCREDIT 46.32 24.22 TAX 0.37 0.09 In GDP 10.95 1.35 In DISTANCE 8.35 0.84 In ASSETS 0.08 16.05 0.48 In SUBS 0.01 0.55 3.21 0.36 M&A 0.08 -0.09 -0.30 0.12 0.32 y 0.27 0.44 BRIDGE- 0.13 0.34 STONE CONTINEN- 0.12 0.32 TAL GOODYEAR 0.51 0.50 MICHELIN 0.09 0.29 PIRELLI 0.10 0.30 [INT.sub.BS] 0.69 1.89 [INT.sub.CO] 0.88 2.53 [INT.sub.GY] 4.65 4.59 [INT.sub.MI] 0.58 1.90 [INT.sub.PR] 0.57 1.92 COCREDIT 48.42 24.99 TAX 0.37 0.09 In GDP 10.97 1.39 In DISTANCE 8.32 0.77 In ASSETS 0.02 16.07 0.48 In SUBS 0.01 0.53 3.20 0.35 M&A 0.06 -0.12 -0.32 0.13 0.34 (1) To economize on space, these statistics are for the full sample; comparable statistics for the non-M&A subsample are available from the authors. Table 3: Entry Model Estimates with Identities (Standard Errors in Parentheses) 1982/1987 Variables Full Sample Large Small Countries Countries (1) Constant -6.37 -13.32 -4.84 (7.46) (13.39) (18.06) BRIDGE- 1.45 2.59 * -- STONE (0.94) (1.54) CONTINEN- -1.24 -1.46 -16.22 TAL (1.59) (2.09) (27783.58) GOODYEAR -0.23 0.06 -0.18 (Hypothesis 1) (0.81) (1.38) (1.67) MICHELIN 0.78 0.89 3.02 (0.99) (1.73) (3.08) PIRELLI 2.34 2.38 -- (1.50) (2.04) COCREDIT 0.04 * 0.07 * 0.00 (0.02) (0.04) (0.04) TAX 4.52 3.14 1.14 (4.01) (6.17) (6.56) In GDP -0.24 0.35 -1.57 (0.38) (0.65) (1.18) In DISTANCE -0.61 -1.37 ** 0.68 (0.39) (0.73) (1.56) In ASSETS 0.52 1.67 * -0.16 (0.40) (0.61) (0.84) In SUBS -0.36 -2.28 3.64 (1.17) (2.11) (3.12) M&A 0.96 -0.62 4.50 * (0.85) (1.48) (2.33) Nagelkerke's 0.32 0.590 0.38 [R.sup.2] n 179 92 87 Firms Bridgestone, Continental, Goodyear, Michelin, Pirelli 1987/1992 Variables Full Sample Large Small Countries Countries (1) Constant -25.97 ** -36.75 ** 11.48 (12.30) (17.82) (23.13) BRIDGE- -1.13 -0.55 -- STONE (0.97) (1.17) CONTINEN- -0.39 0.74 -7.80 TAL (0.93) (1.14) (30511.79) GOODYEAR 1.47 ** 2.22 ** 1.69 (Hypothesis 1) (0.59) (0.94) (1.12) MICHELIN 0.82 1.08 -17.85 (1.09) (1.72) (17935.00) PIRELLI 0.21 0.39 -16.09 (0.70) (1.06) (16320.08) COCREDIT 0.03 ** 0.02 0.07 ** (0.01) (0.02) (0.03) TAX 0.38 -1.60 0.15 (2.57) (3.32) (6.72) In GDP 0.25 -0.54 0.18 (0.22) (0.47) (0.82) In DISTANCE -0.05 -0.12 -1.12 * (0.30) (0.48) (0.72) In ASSETS 0.77 2.28 ** -1.51 (0.74) (1.07) (1.30) In SUBS 2.24 *** 1.44 4.63 ** (0.77) (1.06) (1.63) M&A -20.82 0.98 (1.15) (9687.67) (1.43) Nagelkerke's 0.32 0.46 0.53 [R.sup.2] n 171 90 81 Firms Bridgestone, Continental, Goodyear, Michelin, Pirelli 1982/1992 Variables Full Sample Large Small Countries Countries (1) Constant -20.94 ** -39.18 *** 25.98 (9.45) (15.12) (20.51) BRIDGE- -0.70 -0.52 -- STONE (0.67) (0.78) CONTINEN- -1.02 -1.86 -29.28 TAL (0.90) (1.39) (23367.58) GOODYEAR 1.11 ** 2.05 *** 1.48 (Hypothesis 1) (0.49) (0.80) (0.95) MICHELIN 1.46 ** 3.69 ** -2.98 (0.74) (1.56) (1.96) PIRELLI 1.52 2.35 * -- (0.96) (1.29) COCREDIT 0.03 *** 0.03 * 0.11 *** (0.01) (0.02) (0.04) TAX 0.30 -2.30 9.33 (2.30) (2.90) (6.35) In GDP 0.15 -0.46 -0.70 (0.21) (0.42) (0.65) In DISTANCE -0.30 -0.14 -2.17 *** (0.27) (0.41) (0.81) In ASSETS 0.89 2.66 *** -1.59 (0.57) (0.91) (1.24) In SUBS 1.21 * -0.02 4.37 *** (0.67) (0.92) (1.60) M&A -0.43 -2.44 ** 0.60 (0.70) (1.09) (1.34) Nagelkerke's 0.31 0.43 0.55 [R.sup.2] n 179 92 87 Firms Bridgestone, Continental, Goodyear, Michelin, Pirelli *** p<0.01; ** p<0.05; * p<0.10 Omitted variables are constant for the subset. Table 4: Entry Model Estimates with Identities without M&As (Standard Errors in Parentheses) Variables 1982/1987 Large Small Full Sample Countries (1) Countries (2) Constant -12.26 15.18 (9.21) (13.33) BRIDGE- 2.04 * 2.99 * STONE (1.07) (1.58) CONTINEN- -0.37 -- TAL (2.57) GOODYEAR -0.02 0.40 (Hypothesis (0.96) (1.40) 1) MICHELIN -18.28 -18.09 (10148.25) (10655.93) PIRELLI 1.36 0.57 (2.34) (1.12) COCREDIT 0.03 0.06 (0.03) (0.04) TAX 7.04 0.08 (4.80) (5.87) In GDP -0.11 0.30 (0.51) (0.72) In DISTANCE -0.46 -1.31 * (0.46) (0.71) In ASSETS 0.61 1.33 ** (0.44) (0.67) In SUBS 0.08 -1.94 (1.34) (2.12) Nagelkerke's 0.40 0.60 [R.sup.2] n 155 80 75 Firms Bridgestone, Continental, Goodyear, Michelin, Pirelli Variables 1987/1992 Large Small Full Sample Countries Countries (2) Constant -43.69 ** -36.75 ** -301.54 * (18.45) (17.82) (162.02) BRIDGE- -0.82 -0.55 -- STONE (1.06) (1.17) CONTINEN- -0.55 0.74 -8.06 TAL (0.94) (1.14) (31163.85) GOODYEAR 1.33 ** 2.22 ** 0.49 (Hypothesis (0.61) (0.94) (1.30) 1) MICHELIN 0.67 1.08 -17.58 (1.16) (1.72) (16448.86) PIRELLI 0.53 0.39 -14.17 (0.75) (1.06) (14348.88) COCREDIT 0.04 ** 0.02 0.12 ** (0.01) (0.02) (0.05) TAX 0.44 -1.60 1.93 (2.65) (3.32) (7.40) In GDP 0.30 -0.54 0.30 (0.23) (0.47) (1.02) In DISTANCE 0.07 -0.12 -0.11 (0.32) (0.48) (0.77) In ASSETS 1.69 2.28 ** 14.89 * (1.04) (1.07) (8.49) In SUBS 2.57 *** 1.44 14.45 ** (0.88) (1.06) (6.29) Nagelkerke's 0.36 0.40 0.68 [R.sup.2] n 151 77 74 Firms Bridgestone, Continental, Goodyear, Michelin, Pirelli Variables 1982/1992 Large Small Full Sample Countries (1) Countries Constant -44.04 *** -44.34 *** -113.30 (16.97) (16.97) (117.64) BRIDGE- -0.41 -0.20 -- STONE (0.75) (0.82) CONTINEN- -0.34 -- -24.28 TAL (1.00) (26457.20) GOODYEAR 1.04 ** 2.12 *** 0.13 (Hypothesis (0.51) (0.79) (1.11) 1) MICHELIN 0.56 1.65 -5.15 ** (0.87) (1.15) (2.57) PIRELLI 0.75 0.49 -- (1.08) (0.88) COCREDIT 0.03 ** 0.02 0.13 *** (0.01) (0.02) (0.05) TAX 1.06 -2.31 18.66 * (2.47) (3.08) (9.72) In GDP 0.28 -0.25 0.18 (0.22) (0.43) (0.88) In DISTANCE -0.08 -0.28 -0.65 (0.29) (0.43) (0.89) In ASSETS 2.01 ** 2.91 *** 4.66 (0.94) (1.00) (6.22) In SUBS 1.67 ** 0.11 8.17 ** (0.79) (0.94) (4.05) Nagelkerke's 0.35 0.42 0.65 [R.sup.2] n 155 81 74 Firms Bridgestone, Continental, Goodyear, Michelin, Pirelli *** p<0.01; ** p<0.05; * p<0.10 (1) Omitted variables are constant for the subset. (2) Estimation did not converge for the subset. Table 5: Entry Model Estimates with Interactions (Standard Errors in Parentheses) Variable 1982/1987 Full Sample Large Small Countries Countries (1) Constant -4.91 -7.49 -6.42 (7.41) (12.11) (19.01) [INT.sub.BS] 0.21 0.45 * -- (0.16) (0.26) [INT.sub.CO] 0.01 0.02 -3.03 (0.12) (0.15) (4325.15) [INT.sub.GY] -0.04 0.06 -0.04 (Hypothesis 2) (0.08) (0.16) (0.19) [INT.sub.MI] 0.10 0.17 0.42 (0.14) (0.25) (0.39) [INT.sub.PR] 0.15 0.05 - (0.14) (0.20) COCREDIT 0.03 * 0.06 0.00 (0.02) (0.04) (0.04) TAX 3.83 1.06 1.23 (3.92) (6.07) (6.50) In GDP -0.01 0.44 -1.54 (0.34) (0.64) (1.19) In DISTANCE -0.66 * -1.71 ** 0.94 (0.39) (0.74) (1.73) In ASSETS 0.40 1.12 * -0.16 (0.41) (0.68) (0.84) In SUBS -0.66 -2.91 3.39 (1.19) (2.16) (3.15) M&A 1.13 -0.10 4.43 * (0.84) (1.41) (2.36) Nagelkerke's 0.35 0.56 0.39 [R.sup.2] n 179 92 87 Firms Bridgestone, Continental, Goodyear, Michelin, Pirelli Variable 1987/1992 Full Sample Large Small Countries Countries (1) Constant -27.77 ** -35.76 ** 5.69 (12.02) (17.11) (22.24) [INT.sub.BS] -0.28 -0.28 -- (0.18) (0.22) [INT.sub.CO] 0.02 0.15 -0.91 (0.11) (0.13) (5035.78) [INT.sub.GY] 0.12 *** 0.22 *** 0.08 (Hypothesis 2) (0.04) (0.08) (0.07) [INT.sub.MI] 0.05 0.25 -3.40 (0.17) (0.28) (2764.85) [INT.sub.PR] -0.07 -0.12 -2.44 (0.09) (0.13) (1999.00) COCREDIT 0.03 ** 0.02 0.06 ** (0.01) (0.02) (0.03) TAX 0.64 -2.14 0.80 (2.55) (3.29) (6.23) In GDP 0.34 -0.43 0.37 (0.21) (0.45) (0.79) In DISTANCE -0.04 0.00 -1.00 (0.30) (0.47) (0.66) In ASSETS 1.02 2.45 ** -1.18 (0.72) (1.05) (1.26) In SUBS 1.30 -0.28 3.88 ** (0.76) (1.14) (1.53) M&A -1.25 -20.28 0.92 (1.13) (9937.13) (1.44) Nagelkerke's 0.33 0.47 0.51 [R.sup.2] n 171 90 81 Firms Bridgestone, Continental, Goodyear, Michelin, Pirelli Variable 1982/1992 Full Sample Large Small Countries Countries (1) Constant -22.11 ** -39.76 *** 26.61 (9.54) (13.99) (20.60) [INT.sub.BS] -0.15 -0.09 -- (0.12) (0.13) [INT.sub.CO] -0.11 -0.06 -4.50 (0.10) (0.11) (4281.58) [INT.sub.GY] 0.13 ** 0.22 ** 0.19 * (Hypothesis 2) (0.05) (0.08) (0.11) [INT.sub.MI] 0.16 0.30 -0.35 (0.11) (0.18) (0.24) [INT.sub.PR] 0.20 0.16 -- (0.13) (0.14) COCREDIT 0.03 *** 0.03 * 0.12 *** (0.01) (0.02) (0.04) TAX 0.23 -2.32 9.09 (2.30) (2.84) (6.42) In GDP 0.20 -0.21 -0.76 (0.20) (0.39) (0.66) In DISTANCE -0.41 -0.26 -2.22 *** (0.27) (0.40) (0.82) In ASSETS 0.92 2.53 *** -1.65 (0.58) (0.87) (1.26) In SUBS 1.58 ** 0.29 4.72 *** (0.70) (0.95) (1.70) M&A -0.39 -1.94 ** 0.84 (0.70) (0.95) (1.35) Nagelkerke's 0.32 0.39 0.55 [R.sup.2] n 179 92 87 Firms Bridgestone, Continental, Goodyear, Michelin, Pirelli *** p<0.01; ** p<0.05; * p<0.10 (1) Omitted variables are constant for the subset. Table 6: Entry Model Estimates with Interactions without M&As (Standard Errors in Parentheses) Variables 1982/1987 Full Sample Large Small Countries Countries (1) Constant -15.98 -46.12 (10.01) (36.23) [INT.sub.BS] 0.21 0.49 * (0.18) (0.29) [INT.sub.CO] 0.18 0.56 (0.18) (0.46) [INT.sub.GY] -0.03 0.08 (Hypothesis 2) (0.09) (0.18) [INT.sub.MI] -4.45 -5.09 (1876.44) (1785.20) [INT.sub.PR] -0.09 -0.80 (0.24) (0.68) COCREDIT 0.02 0.08 (0.02) (0.06) TAX 5.54 -4.02 (4.53) (6.75) In GDP 0.28 0.86 (0.41) (1.00) In DISTANCE -0.32 -0.60 (0.49) (0.93) In ASSETS 0.66 2.89 (0.50) (1.89) In SUBS -0.15 -3.24 (1.30) (2.64) Nagelkerke's 0.36 0.62 [R.sup.2] n 155 80 75 Firms Bridgestone, Continental, Goodyear, Michelin, Pirelli Variables 1987/1992 Full Sample Large Small Countries Countries (2) Constant -43.50 ** -35.76 ** -326.21 ** (17.35) (17.11) (165.97) [INT.sub.BS] -0.24 -0.28 -- (0.19) (0.22) [INT.sub.CO] -0.01 0.15 -1.46 (0.11) (0.13) (4774.85) [INT.sub.GY] 0.11 *** 0.22 *** -0.02 (Hypothesis 2) (0.04) (0.08) (0.09) [INT.sub.MI] 0.03 0.25 -3.21 (0.19) (0.28) (2453.93) [INT.sub.PR] -0.04 -0.12 -2.09 (0.10) (0.13) (1698.03) COCREDIT 0.04 *** 0.02 0.11 ** (0.01) (0.02) (0.05) TAX 0.68 -2.14 2.00 (2.63) (3.29) (7.15) In GDP 0.38 * -0.43 0.56 (0.22) (0.45) (1.05 In DISTANCE 0.10 0.00 0.06 (0.32) (0.47) (0.74) In ASSETS 1.80 * 2.44 ** 16.05 * (0.98) (1.05) (8.63) In SUBS 1.64 * -0.28 15.14 ** (0.87) (1.14) (6.56) Nagelkerke's 0.36 0.40 0.68 [R.sup.2] n 151 77 74 Firms Bridgestone, Continental, Goodyear, Michelin, Pirelli Variables 1982/1992 Full Sample Large Small Countries Countries (2) Constant -44.69 *** -51.08 *** -80.63 (16.80) (17.41) (109.36) [INT.sub.BS] -0.13 -0.08 -- (0.13) (0.14) [INT.sub.CO] -0.03 0.15 -4.41 (0.11) (0.16) (4253.73) [INT.sub.GY] 0.12 ** 0.22 ** 0.06 (Hypothesis 2) (0.06) (0.09) (0.13) [INT.sub.MI] -0.02 -0.07 -0.81 * (0.14) (0.26) (0.43) [INT.sub.PR] 0.09 -0.11 - (0.15) (0.19) COCREDIT 0.03 ** 0.02 0.13 *** (0.01) (0.02) (0.04) TAX 1.17 -2.12 17.76 * (2.46) (3.05) (9.36) In GDP 0.35 -0.04 0.08 (0.22) (0.44) (0.86) In DISTANCE -0.13 -0.06 -0.79 (0.30) (0.45) (0.87) In ASSETS 1.98 ** 3.06 *** 2.93 (0.94) (1.02) (5.88) In SUBS 1.96 ** 0.09 7.49 ** (0.81) (1.01) (3.78) Nagelkerke's 0.35 0.41 0.65 [R.sup.2] n 155 81 74 Firms Bridgestone, Continental, Goodyear, Michelin, Pirelli *** p<0.01; ** p<0.05; * p<0.10 (1) Estimation did not converge. (2) Omitted variables are constant for the subset.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||RESEARCH ARTICLE|
|Author:||Rose, Elizabeth L.; Ito, Kiyohiko|
|Publication:||Management International Review|
|Article Type:||Industry overview|
|Date:||Sep 1, 2009|
|Previous Article:||Meta-analyses of international joint venture performance determinants: evidence for theory, methodological artifacts and the unique context of China.|
|Next Article:||Fifteen good years: an analysis of publications in management international review.|