بین المللی سازی در خرده فروشی: مدل سازی الگوی ورود به بازار خارجی

This article investigates retailers' choice of foreign markets using a spatial-interaction model that takes into account factors related to the attractiveness of a given market as well as distance to that market. While gravity and spatial-interaction models have a long tradition in retailing research, such models have not been applied to foreign market entry before. Data on UK retailers' cross-border operations provide good support for the model. Its explanatory power is particularly strong for the initial set of entries made by the retailers, which suggests that experience effects as well as firm-specific internationalization strategies eventually override the common effects of distance and market attractiveness on the choice of which foreign markets to enter.

Foreign market entry has been a popular topic in international business research during the last couple of decades. Entry modes as well as choice of foreign markets are analyzed in a large number of empirical studies (Datta et al., 2002 and Mitra and Golder, 2002). Manufacturing industries initially attracted the main interest in foreign market entry studies (see, e.g. Benito and Gripsrud, 1992 and Davidson, 1980), but the international expansion of service industries has recently gained increasing attention (Buckley et al., 1992, Contractor et al., 2003 and Gielens and Dekimpe, 2001). Retailing is a service industry where international expansion and cross-border activities have become much more pronounced in recent years (Akehurst and Alexander, 1996 and Petersen and Welch, 2000). Relatively little is in fact known about retailers' choice of foreign markets, but it is generally assumed that while manufacturing companies typically enter a foreign market to exploit location advantages in production costs and access to resources, an entry into a foreign market by a retailer is more likely to be motivated by demand factors. The present article deals with the choice of foreign markets when retail operations are to be internationalized. Which countries are being selected when a retailer goes abroad? What factors explain the pattern of foreign expansion pursued by retailers? Previous literature on foreign market entry has tended to look at such decisions from either economics (Dunning, 2000 and Hennart, 1982) or internationalization process models (Andersen, 1993, Johanson and Vahlne, 1977, Leonidou and Katsikeas, 1996 and Luostarinen, 1979) based on the behavioral theory of the firm (Cyert and March, 1963), and treating them as essentially competing approaches (Datta et al., 2002). Despite the emerging recognition that the theoretical perspectives complement each other (Benito and Welch, 1994), few attempts have been made to bring their strong points together; possibly due to a perceived lack of modeling tools that could adequately integrate the main attributes of the various approaches. Our aim is to develop a model of foreign market choice with a micro-foundation concerning human choice behavior. In this article we argue that a spatial-interaction model may be highly relevant for analyzing the impact of various factors on the choice of foreign markets to enter. That model can accommodate both factors drawn from economics approaches–e.g. the purchasing power and size of markets–and factors drawn from the internationalization process approach, such as the distance to a market. The proposed probability model summarizes the impact of various factors on the outcome of the location choices and makes it possible to estimate the effect of each factor. To the best of our knowledge, the model has not been applied to the choice of foreign markets before. The remainder of this article is organized as follows: In the next section, we briefly review literature dealing with retailers' choice of foreign markets. Following that, we present our model and describe the data and measures used in the study. Then, we present our results and conclude with a discussion of key findings, limitations of the study and possible directions for future research.

This article applies a spatial-interaction approach to modeling retailers' choice of foreign markets. To the best of our knowledge, this is a novel approach to modeling that particular, and important, dimension of retailer internationalization. Appealing features of the model are its flexibility–it can easily accommodate a range of variables drawn from diverse theoretical perspectives–and its rigorous, yet simple structure; variables are grouped into two main sets of factors. The model is derived assuming a rational mode of strategic decision making in retail companies, and it is generically equivalent to the spatial-interaction model developed in retailing to analyze consumers' choice of shopping centers. We use data on British retailers' foreign operations in 1996 to test the model. On the whole, we get good empirical support for the model, and find that the model is convincingly robust across different specifications. In all, we find evidence that the selection of which foreign markets to enter is influenced both by factors that make a particular market attractive and by the distance to that market. The results indicate furthermore that experience effects are at work. In agreement with the internationalization process model, the data show that distance factors are particularly important early-on in companies' internationalization: retailers tend to seek out close markets initially, but may then move on to geographically and culturally more distant countries. We also find that characteristics that improve the attractiveness of a country, like a sizeable customer concentration, are important at both the early and the intermediate phases of internationalization. At the more advanced stages of internationalization, the patterns were less apparent. This suggests that there is considerable variation between highly internationalized companies with regard to their market choice policies. Hence, in order to uncover behavioral patterns, if any, one may have to explore in more detail the investment history of specific companies. Some limitations of the study should be observed. First, the findings presented here concern the behavior of a sample of UK retailers, and future studies should examine to what extent the results can be generalized to other empirical settings. Second, although the data set does provide a longitudinal sequencing of the entries, it is limited to foreign operations that existed at a particular point in time. This means that past ventures that were dissolved prior to 1996 are not included. It is, obviously, probable that at least some of the retailers have made a larger number of foreign entries than that actually recorded in our data set. Ventures that proved to be unsuccessful may have been closed down or sold out, while successful operations may have been taken over by other retailers at a point in time preceding our registration of the data. Without additional information, it is generally difficult to assess how “left-censoring” issues affect results, and efforts should be made to collect more complete data. Third, the cross-border activities covered in the database include different methods of entry: wholly owned subsidiaries, joint ventures, franchising, and concessions/licensing. This means that the data encompass entry modes associated with different degrees of risk and commitment. It is reasonable to assume that when faced with a distant, and perhaps less known, market an investor will tend to offset the higher risk involved by choosing a low risk entry mode (Petersen and Welch, 2000). Investors may similarly be more concerned about risk for large investments than for small ones and be hesitant to commit resources when entries are regarded as risky. It would clearly be preferable to have more information about the actual entries into the various markets, but it may be argued that the data used in the present study actually underestimate the effect of, in particular, cultural distance. The reason is that foreign entries at higher levels of risk and commitment are likely to be more sensitive to an uncertain environment. Finally, this study uses simple empirical specifications of the model, which nevertheless work rather well. We believe that simplicity is commendable per se, but future research could try out various alternatives regarding what variables to include in the model as well as other measures. The economics and behavioral approaches represent rich research traditions in international business from which additional variables can be identified. The present study investigates both cultural and physical dimensions of distance. Even though geographical distance appears to be more important than cultural distance in determining foreign market choices, our findings suggest that cultural distance has an independent effect. While the estimates are significant for the total number of moves made by the companies, the model performs really well only for the first few moves. It seems reasonable that cultural distance would have the highest impact in the first couple of foreign ventures (Johanson and Vahlne, 1977, Luostarinen, 1979 and Vida, 2000). A possible explanation is that entries into particular countries may form the basis for consecutive moves. In that case, measuring cultural and geographical distances from the home country may be misleading after the first initial steps. Companies that have chosen different directions initially may actually face quite different choice situations at a later stage, including different distances to particular markets and a different consideration set. The model as such may be relevant, but the implicit assumption that the retailer each time starts in the home country and has the same consideration set may not be warranted. One way of investigating such issues is to conduct in-depth case studies of specific companies. Such studies would also make it possible to investigate to what extent cognitive biases are present in decision processes and the cognitive maps of the decision-makers need to be explored more in depth (Hodgkinson et al., 1999). Future studies could also analyze increments in distance between subsequent entries and thereby unveil whether and when foreign expansion levels off. A less expected finding is that the market attractiveness factors also lose much of their explanatory power when we look at higher investment numbers. Another reason why the performance of the model is weaker for higher investment numbers is that after the initial moves have been made, a number of more subtle firm-level and context specific factors will eventually override the impact of the macro factors included in the model. When the most attractive markets have been entered, the remaining possibilities may differ only marginally on the characteristics that we have included in the model. In such a case, a much more specific model is needed to capture the decision process in different retail firms and the attraction of the various markets (Mitra and Golder, 2002). Gielens and Dekimpe (2001) have for instance analyzed how five strategic entry decisions (scale of entry, mode of entry, order of entry, adaptation of the retail format to local market conditions, and familiarity of the store format to the parent company) have an impact on the performance of foreign entries in grocery retailing. It turns out that the order of entry into a foreign market is particularly important: The fewer international competitors a retailer face when entering a foreign market, the more likely it is that the performance will be high. This means that when a retailer evaluates the attraction of a foreign market the number of international competitors already established in the market should be taken into account. Analyzing specific location choices requires information regarding a number of characteristics, and recent developments concerning GIS (Geographic Information System) has made such information much more available and easier to use. Such data could form the basis for future studies. Taken together, our study demonstrates the applicability of a spatial-interaction approach to modeling the foreign market entry patterns of international retailers. We hope that other studies will follow that remedy some of its limitations. Useful extensions of the research would be to collect more detailed and complete information about the characteristics of foreign entries, especially the modes of entry used by companies and the timing of the entries.