تعادل رشد در سرتاسر رویکرد تنوع بخشی جغرافیایی و تنوع بخشی محصول: رویکرد اقتضائی

We theorize that firms simultaneously seek to balance their growth across both the geographic and product diversification domains. To achieve this balance, businesses commonly adopt a strategy of expanding an under-diversified direction at the expense of an over-diversified one. Accordingly, we depict geographic diversification and product diversification as being an endogenous relationship, from which we hypothesize that firms that have under-diversified in a given direction and over-diversified in the other will expand the former at the expense of the latter. Meanwhile, firms that have under-diversified in both directions will expand both diversification paths, while firms that have over-diversified in both directions will contract in both diversification routes. We investigate these predicted relationships and show them empirically using a sample of leading Japanese multinationals in the 1990–2000 period.

Geographic diversification and product diversification have long been acknowledged as two dominant growth strategies of firms (Caves, 1996 and Mudambi and Mudambi, 2002). The ability to introduce multiple products to multiple countries can increase sales and reduce operating costs provided a firm does not over expand (Geringer, Tallman, & Olsen, 2000). Although both the separate and joint impact of geographic and product diversification on firm performance have been extensively researched (Delios and Beamish, 1999, Geringer et al., 1989, Hitt et al., 1997 and Tallman and Li, 1996), relatively little attention has been given to the inter-relationships between these two growth strategies (Peng & Delios, 2006). The few studies that have considered this relationship found contradictory relationships between geographic and product diversification, including: a positive linear relationship indicating complementarity between the two strategies, a negative linear relationship indicating a substitution effect between these two growth strategies, and more complex curvilinear relationships (e.g. Davies et al., 2001, Kumar, 2009, Meyer, 2006, Pearce, 1993, Wiersema and Bowen, 2008 and Wolf, 1977). Hence, the exact nature of the relationships between geographic diversification and product diversification remains vague and unclear, in both a conceptual and an empirical sense. We address these issues by developing a conceptual perspective on the relationship between geographic and product diversification that takes into account the current levels of both types of diversification when predicting future levels of these two types of diversification. We ground our arguments in the Resource Based View (RBV) and Transaction Costs Economics (TCE) to contend that firms seek to increase diversification and growth in underutilized directions while trying to avoid over-diversification in specific paths. In this sense, we regard a firm's growth as an attempt to balance the two diversification dimensions. Specifically, we hypothesize that firms with low geographic and product diversification levels will increase both types of diversification. However, firms with low product diversification and high geographic diversification will increase the former and reduce the latter, while firms with high product diversification and low geographic diversification will reduce the former and increase the latter. Finally, firms with high levels of both geographic diversification and product diversification will reduce both types of diversification. This view offers an important contingency to our understanding of the relationships between geographic and product diversification as it implies that such relationships will differ for firms at different levels of both diversification types. We test these predictions on panel data of 288 Japanese multinational corporations (MNCs) covering the years 1990–2000. During the last two decades of the 20th century, Japanese firms went through a period of rapid geographic expansion and also expanded product-wise, which allows us to capture the longitudinal profile of Japanese firms at varying stages of geographic diversification and with a good range in their level of product diversification. Two important features of our empirical estimation method are noteworthy. First, the methodology involves separating the firms in our sample into four different quadrants representing their respective levels of geographic and product diversification as a means of distinguishing between firms that are under-diversified and over diversified in terms of geographic and product diversification. Secondly, the methodology corrects for the fact that geographic and product diversification decisions are likely to be made simultaneously and endogenously by running Two Stage Least Squares (2SLS) within-firm fixed effects regression models. The rest of the paper is organized as follows. In Section 2 we present our conceptual framework. In Section 3 we describe our data and estimation methods and forwarding Section 4 we discuss our results. In Section 5 some preliminary performance implications of our study are drawn and finally in Section 6 we discuss our results and future research directions, and draw relevant conclusions.

The main contribution of our study rests in its identification of the asymmetric set of relationships that exist between geographic diversification and product diversification. These arise, arguably, from differences in the current levels of firms’ geographic and product diversification and the endogenous relationship between simultaneous decisions regarding future expansion or contractions along these two diversification paths. In this respect, the paper offers a possible way to reconcile past conflicting findings on the relationship between geographic and product diversification by explicitly considering the contingency of distinguishing between the effects of firms’ under- and over-diversification in each diversification trajectory. We have argued and shown empirically that the relationships between geographic and product diversification are not uniform as implicitly assumed in most extant research. Instead, these relationships are likely to differ according to the current levels of diversification in both directions. Our core arguments are: (1) under-diversification in one path and over-diversification in the other will lead to an expansion of the former at the expense of the latter, hence leading to a negative correlation between geographic and product diversification. Furthermore, (2) firms that are either under-diversified or over-diversified in both directions will respectively increase diversification or decrease it along both paths, hence implying a positive correlation between geographic and product diversification. We have demonstrated that the relationship between geographic and product diversification is not identical for all firms and have specified the conditions under which a positive linear relationship and a negative linear relationship may co-exist at different levels of geographic and product diversification. Likewise, our results also mark the possible existence of a curvilinear relationship between product and geographic diversification, as such relationships can result from the joint analysis of firms belonging to different quadrants (as portrayed in Fig. 1). In other words, curvilinear relationships are likely to be observed for samples of firms drawn from quadrants [2] and [4] (where a U-shape relationship is likely to be observed) or in samples of firms drawn from quadrants [1] and [3] (where an inverted U-shape relationship is likely to be observed). Our conceptual framework and findings further imply that future research needs to consider how and whether for some firms simultaneous expansion in both directions will enhance performance, while for other firms simultaneous contraction will enhance performance. For another set of firms, expansion in one direction paralleled by a contraction of the other will enhance performance. Thus, future work should focus on incorporating the contingencies in these relationships between growth through geographic and product diversification into studies of the performance implications of these growth decisions. Nevertheless, it should be noted that our results need to be tempered against the setting we used to test our hypotheses. Japanese MNCs have had a strong export-orientation to serving foreign markets, with foreign direct investments (FDI) forming an important component of the production aspects of that strategy. Geographic diversification is often motivated by local production considerations, not necessarily by local market entry considerations. To the extent that firms from other nations have a greater local market-seeking motivation for their foreign investments, the relationships we have identified here might differ. Likewise, Japanese MNCs are traditionally considered to be relatively late to pursue FDI in their lifecycles and do so less extensively and in a more limited value chain of activities than MNCs from the US, UK, and other European countries. Furthermore, Japanese firms are often said to focus on sales growth more than other MNCs (Bartlett and Ghoshal, 1989 and Geringer et al., 2000). Both these considerations mean that our findings cannot necessarily be generalized to non-Japanese MNCs. In addition, our sample is comprised of MNCs (and excludes domestic firms), however it is not clear whether the analyzed firms actually compare themselves to other MNCs or to their industry peers (including non MNCs). Future work examining geographic and product diversification patterns of domestic and non-domestic firms from different countries are hence required to examine the external validity of our results. It should be noted that, while our analyses combine resource based view logic with that of transaction costs, the data available to us did not enable us to directly test the impact of the nature of firm specific resources and transactions on the chosen path of growth. Such a test requires identifying how the use of resources or their release coupled with the complexity of transactions faced by the firm affect its future diversification moves. Data allowing the examination of the accumulation of such resources and transactions are likely to be difficult to develop, but it would be important to directly analyze how resources and transactions affect the growth paths of firms. The latter point is especially relevant given the fact that a change in the market environment, such as a change in technology or a change in the diversification strategy of the firm's competitors, is likely to affect a firm's resources as well as its transaction costs. In fact, even if a firm is, at a given point of time, at an “optimal” level of diversification geographically and product-wise, such changes may drive it to change its levels of diversification. Finally, changes in external factors were not taken into account in the current study because of data availability limitations. Future research, expanding the contingency approach presented in this paper, may well need to take into account such factors as they are likely to affect geographic and product diversification decisions. Overall, this study highlights the complexities of simultaneously pursuing geographic and product diversification paths. We theorize that firms seek to simultaneously optimize their levels of geographic and product diversification. Hence firms that have under-diversified in either path will seek to expand their diversification in this path while firms that have over-diversified in either path will seek to contract their diversification in that path. Accordingly, we expect and observe positive correlations between the geographic and product diversification moves of firms that are either under-diversified or over-diversified in both paths and negative correlations between the diversification moves of firms that are under-diversified in one path but over-diversified in the other. Bearing in mind the limitations of this study, as stated above, we therefore conclude that decisions about resource allocations and the emphasis placed on strategies promoting geographic or product expansion need to be balanced against a firm's current levels of diversification and the strategic intents of expansion into both diversification realms.