Is international business good for companies? The evolutionary or multi-stage theory of internationalization vs. the transaction cost perspective.
* This paper's objective is to articulate, more precisely than has occurred in the past, the principal theory rationales underlying the Multinationality/Performance (M/P) link, by examining each claim for the negative or positive benefits of internalization, from the lens of the theory of the firm and the multinational enterprise. A concurrent objective of this paper is to respond to critiques of M/P theory and discuss methodology and operationalization problems in empirical testing.
* While international expansion of a firm will not necessarily always improve performance (during the initial international expansion stage, or in cases where a firm may have over-internationalized), for the most part, over the considerable middle range of expansion, net positive benefits accrue from internationalization. Underlying theory rationales are detailed in the paper.
* The results of over one hundred empirical studies over the past 30 years appear, on superficial examination, to be contradictory, but can be reconciled by the recently proposed 3-stage or S-shaped general theory.
General Theory, Multinationality-Performance Relationship, Past Empirical Results
We are in an era when everyone, from Deans to Captains of Industry, seems to subscribe to the notion that internationalization is good for firms. Indeed, the very field of International Business Studies rests upon the meta-hypothesis that increasing internationalization will generally augment company profits, other things being equal. However, empirical studies of the past 30 years have only shown mixed results for the link between financial performance and the degree of a firm's internationalization (e.g., Ruigrok/Wagner 2004). These seemingly contradictory results appear to put into question the entire sub-field of Multinationality/Performance (M/P) studies and the theory underlying them. This paper is a review of the theoretical underpinnings of this literature, as well as a response to Jean-Francois Hennart's article in this issue which views the debate through a Transaction Cost/Internalization (TCI) lens. By taking a deliberately agnostic view of the "Should international expansion produce positive results for companies?" question, Hennart has done our field a service, in forcing us to more critically articulate the theory underpinnings of the positive slope hypothesis for the link between internationalization and financial performance.
I will argue that there are indeed sound theoretical grounds for asserting a positive link between the degree of internationalization (DOI) of a firm and performance. In commenting on the Hennart paper, I propose that the Transaction Cost perspective is a useful, but rather limited lens with which to view this subject. Sometimes Hennart switches to stringent neoclassical views of the firm which abstract too much from current reality and the actual strategies of firms. (1) Moreover, we cannot treat this question in comparative statics terms. The more dynamic, or evolutionary perspective of the recent "multi-stage" models for the Multinationality/ Performance (M/P) field present a more insightful view, and can also reconcile the seemingly contradictory empirical results.
Since multi-stage theories for the multinationality/performance link seem yet to be generally accepted, and since some of Hennart's criticisms can be addressed using this more dynamic perspective, I begin this paper with a quick summary of this theory. Next, this paper specifically responds to Hennart's criticisms and shows how one can go beyond them to offer more robust a priori arguments for the M/P link. Finally, this article discusses methodological issues and offers an explanation of how the seemingly contradictory results of 30 years of empirical studies (e.g., Ruigrok/Wagner 2004) can be at least partially explained through the multi-stage, or evolutionary, lens.
A Multi-Stage or Evolutionary Perspective on International Expansion
The recently promulgated 3-stage theory (Contractor/Kundu/Hsu 2003, Lu/Beamish 2004, Thomas/Eden 2004) provides a dynamic, or longitudinal explanation of the effect of three sequential stages that companies go through when expanding internationally. At all stages, there are incremental benefits, and incremental costs, of adding an additional nation or market to the firm's existing portfolio of countries.
For most of the range, on the Degree of Internationalization (DOI) scale, incremental benefits outweigh the incremental costs of adding another nation, or dimension of internationalization. This is called Stage 2. However in Stage 1 (initial or early internationalization) as well as in Stage 3 (excessive internationalization) incremental costs are greater than incremental benefits.
The Hennart paper is cognizant of some of the advantages of internationalization mentioned below, but does not sequence them chronologically in the firm's evolution abroad, as is done in the 3-stage theory, depicted in Figure 1.
[FIGURE 1 OMITTED]
Stage 1 (Early Internationalization): On initial and early international expansion, a firm encounters liabilities of foreignness (Zaheer/Mosakowski 1997) in terms of their unfamiliarity with the foreign market's institutions and possible discrimination against foreign companies. There are also significant costs of learning about a new nation and culture (Doz/Santos/Williamson 2001), as well as local adaptation costs. (2) As Caves (1971, p. 5) puts it, "The foreign enterprise must pay dearly for what the native has acquired at no cost to the firm (because it was part of the entrepreneur's general education) or can acquire more cheaply (because as it were the native knows where to look)".
Initially, the large incremental set-up costs of setting up new international operations, including additional overheads, can only be amortized over one or few foreign markets. Such costs are likely to be high per unit of product sold abroad, per nation, or per whatever index of internationalization is used.
As a result, in Stage 1, the company's incremental costs of international expansion are greater than the incremental benefits or revenues it reaps. (3) The high upfront costs of international expansion can initially be spread over only a small base of foreign operations. Accordingly, the effect of international expansion on such a company's financial performance is negative. The Performance over Multinationality function has a negative slope.
Stage 2 (Later Internationalization): Having passed through Stage 1, further international expansion begins to yield incremental benefits that exceed the firm's incremental costs. (The length or duration that a firm may spend in Stage 1 will vary by sector, home and foreign market characteristics. In that sense, the M/P relation is context-dependent. The context or situation of the firm, or sector, does make a difference. But modification dependent on the context by no means obviates this general theory). For every additional international operation or market added, there would continue to be learning, coordination, local adaptation and legitimacy acquisition costs, but the following benefits begin to outweigh these incremental costs in Stage 2.
a. Knowledge acquired from abroad. The international firm's presence in several nations increases its ability to pick up on foreign knowledge that domestic or less internationalized rivals cannot access (Kogut/Zander 1993, Ghoshal/Bartlett 1990). (This is different from accumulated internationalization experience which refers to organizational learning and improved ability to reproduce the firm in foreign locations. This is discussed under point g. below).
b. Accessing or "arbitraging" cheaper inputs. (4) This may include lower cost labor, or any other inputs. (According to the Transaction Cost perspective in Hennart's paper, accessing cheaper inputs abroad can also be achieved through the market, e.g., contract production of footware by Adidas or Nike. Or if this advantage is internalized by rival multinationals who access the same cheaper input abroad, that advantage would erode away any "super-normal" profits, according to Hennart. This may be true in a comparative statics sense. However, this is a neoclassical, economics textbook view, of intense competition between undifferentiated firms--all having the same strategy--which leads to low or zero profits in all competing firms. This view ignores the whole concept of strategic differentiation, first-mover advantages, and of dominant market shares that lead to persisting oligopolies. In the real world, at a certain point in time, it is axiomatically true that some firms in an industry will be further along than others on their internationalization path and reap greater internationalization advantages (in Stage 2) compared to others which may still be in Stage 1 of their expansion). In sum, a company present in more nations is, ceteris paribus, in a better position to find arbitrage opportunities than its competitors that are present in few, or only one.
c. Exploitation of firm-specific assets carried to each foreign market (Caves 1996, Buckley 1988). This confers--for a considerable if not indefinite period--monopolistic advantage to the foreign firm (Hymer 1976). (At one point, Hennart appears to accept this idea, but then adds that internalized superior capabilities, enjoyed by a multinational company, would quickly be dissipated to the point where the foreign company would soon be reduced to earning only "normal" profits. This, once again, ignores a vast strategy and management literature, where internalized knowledge is sticky, not easily imitable, and where intellectual property (IP) such as brands and patents confers firm-specific advantages that last for a long time, if not indefinitely. Even if patent rights expire, continued investments in R&D can keep firms ahead of competitors. Moreover, the bulk of corporate capability, for all sectors in general, lies not in formally registered IP, but in internal organizational routines that confer lasting advantage).
d. Accumulation of market power because of wide multinational presence. There is a considerable body of work (e.g. Kogut 1985) that supports the view that, in many sectors, being large and multinational enables a company to establish its preferred technical standards and protocols, or create global brand equity, and even, in some cases, cartelize the industry to earn "super-normal" profits. A considerable degree of internationalization (DOI) is prerequisite to enjoying this financial muscle.
e. International scale. Hennart devotes a considerable fraction of his paper to critiquing this issue. I will return to this important question separately later in this paper. In sum, I agree with Hennart that the concept, of why a multinational scale should confer an advantage, has not been articulated clearly in the M/P literature, despite antecedent literature (Buckley/Casson 1976, Caves 1996, Hymer 1976). However, one needs to distinguish between scale effects in production, or pieces of the value-added chain--treated in the Hennart paper--from the scale effects that accrue to the multinational company by its ability to amortize high technology and centralized overheads over several nations. In the first case, lower cost production from scale economies may sometimes be realized by combining demand from more than one country--and sometimes not. (Here I agree with Hennart). But the simple fact remains, that the large expenditures needed for R&D and central overhead costs comprise a lower burden per unit of output, the more output there is. That is to say, a firm present and selling in many countries is (i) better able to amortize huge R&D outlays that are today needed in high-tech sectors, and (ii) is better positioned to assume the risks of larger R&D budgets than its less internationalized rivals. (5) If larger R&D budgets enable performance superior to that in rivals, then international scale does matter.
f. Geographical diversification. There has been unnecessary confusion in the M/P literature in using the term "diversification" which has been used loosely to suggest higher profits from internationalization. (6) The term should refer only to risk reduction from operating in several nations at once, because of asynchronous business cycles (e.g., Siddharthan/Lall 1982), or from smoothing out foreign exchange volatility with multi-currency cash flows, or because having plants in several nations confers operational and strategic flexibility on the firm. (Hennart correctly points out that the greater ability of the multi-plant firm (with plants in different nations) to respond to political disruptions, or supply chain disruptions, needs to be viewed against the possibly higher global total costs of maintaining multiple-country plants. But his caution applies only when there is substantial over-installed capacity). Hence geographical diversification can indeed be an advantage, but one needs to differentiate between reduction in risk (or volatility), from increases in profit, as a result of greater internationalization. In empirical studies one needs to measure risk or volatility reduction separately from increase in profits. Alternatively, as in the finance literature, one needs to create an overall index of risk-adjusted profit--as an indicator of performance.
g. Internationalization experience. This is another term that has been used to suggest different meanings to different authors. Traditional internationalization literature, such as Johanson/Vahlne (1977) simply describes the internationally expanding firm as gaining more foreign knowledge, and market opportunities, as it adds more markets abroad. With more experience it ventures farther afield to less culturally familiar territories. I refer here to something else--the accumulated and better organizational ability, over time, to reproduce the firm in foreign locations. This replication ability, at lower cost and time, is gained through greater international experience. Faster foreign entry, at lower cost, improves profit compared to rivals with less internationalization experience.
Based on the seven foregoing benefits, the internationalizing firm, over a considerable range in middle Stage 2, enjoys the net benefits of international expansion, as shown in Figure 1. In Stage 2 incremental benefits outweigh incremental costs. Ceteris paribus, there is a positive slope for the Performance vs. DOI (Degree of Internationalization) link.
Stage 3 (Excessive Internationalization): Internationalization is good for company profits and risk reduction over a wide range. But international expansion beyond a certain point (depending on the particular sector) again leads to a reduction in profits.
A fact not sufficiently articulated in the M/P literature is that in many sectors, after the first 40 to 60 nations, the lower ranked nations comprise very peripheral, and less promising, markets. (7) Beyond the top tier markets, the extent of cultural and economic diversity increases sharply (Bartlett/Ghoshal 1989, Sunderam/Black 1992). Adding incremental markets escalates managerial costs and information overload (Hitt/Hoskisson/Kim 1997, Hoskisson/Turk 1990) and global coordination costs increase by more than a linear relationship to the degree of internationalization (DOI) or number of nations. Hitt, Hoskisson, and Kim (1997, p. 769) put it succinctly as, "... the coordination required (multiple transactions among many geographically diverse units), costs more than the benefits derived from sharing resources and exploiting opportunities".
In Stage 3, a few firms may inadvertently (or knowingly) over-internationalize, and once more, suffer a net negative effect on performance. (The slope of Performance on Degree of Internationalization is negative). Hennart asserts that firms may deviate from an "optimal," position in the M/P map because of randomness. This supports the foregoing Stage 3 argument, but fails to take account of the chronological or evolutionary perspective, that companies may find themselves "over-internationalized" because (i) they enjoyed an increase in performance through many years of increasing internationalization and then did not know when to stop, or (ii) firms that knowingly over-internationalize for other long term strategic reasons, (iii) or firms that did not respond quickly enough to changes in technology or competition. We do not need TCI theory to tell us that "... firms which are not at their optimum degree of international diversification will experience lower performance" (Hennart 2007, p. 446). (8)
To summarize: the 3-stage model posits two relatively short periods (Stage 1 and Stage 3) where incremental internationalization produces a net negative effect on profits, and a longer middle Stage 2 wherein the effect of international expansion is (in net terms) positive. Overall, the theory thus posits a sigmoid M/P function. (In empirical practice, the statistically fitted curves may turn out to be U-shaped if Stages 1 and 2 predominate in the sample firms; or inverted-U-shaped if Stages 2 and 3 are heavily represented in other company samples; or indeed S-shaped if all three stages are well represented as was found in Contractor/Kundu/Hsu 2003, and Thomas/Eden 2004). (This is discussed and depicted later in Figure 3).
[FIGURE 3 OMITTED]
One inescapable fact remains: In virtually all empirical M/P studies, whether we see a U, inverted-U, or S-shape, there is embedded in the results a positively sloped leg over some part of the Degree of Internationalization range, thus empirically supporting the notion that international expansion is "good" over some or much of the range. (See Figure 3 discussed later).
The Transaction Cost Lens: A Response to Hennart's Critiques
We have three views of the multinational corporation that can shape M/P theory, namely, the international firm (1) as an exploiter of internalized capabilities, (2) as a learning organization absorbing knowledge from abroad, and (3) as a coordinator and arbitrageur across national borders.
It may be useful to begin by asking how Transaction Cost/Internalization (TCI) theory views the international firm. The answer to this question is not clear. In his earlier papers Hennart argued that TCI theorists generally asserted that MNEs set up the bulk of their foreign subsidiaries to exploit intangibles held at corporate headquarters, and that these subsidiaries still remain principally involved in this exploitation task. This is followed by several pages arguing that the multinational company is not really a learning organization, and that therefore a principal leg of M/P theory--that profits are improved by acquisition of foreign knowledge (Stage 2 above)--is not really valid. However, later in the current version (Hennart 2007, p. 442) we read "... TCI theory highlights that having activities abroad arises from multiple motives. Foreign affiliates are set up to acquire parts or raw materials, to exploit tacit knowledge or reputation, or to access technology or brand names. Given this variety of motives for foreign expansion..".
In this second view, TCI theory does seem to allow for the internationally expanding firm to "access technology" via its foreign affiliates, and augment profits with its ability to coordinate across national borders. Elsewhere, Hennart cites Bartlett and Ghoshal (1989), to acknowledge the "transnational corporation", and Kogut and Zander's (1993) view of the firm as pooling knowledge from geographically scattered sources.
Which perspective predominates? The multinational firm as (1) Exploiter? (2) Learner and Transferor of Knowledge across borders? or (3) Coordinator and Arbitrageur? All three, in fact, support the notion of incremental benefits from international expansion, as we saw earlier in the foregoing discussion on Stage 2 of the 3-stage model. Hennart, however argues, earlier in his paper, that the multinational is principally still an Exploiter of internalized capabilities, and not a Learner or Coordinator. He then goes on to downplay the notion of exploitation advantages by saying that the internalized capabilities of a firm confer on it no more than a fleeting monopolistic power since competition would erode such firm-specific knowledge--or that in other cases, markets can substitute for the multinational company.
In short, Hennart first includes, under the ever-expanding rubric of TCI theory, many of the hallowed arguments for the benefits of internationalization, but he then simultaneously attacks them. If such an iconoclastic position is designed to force the field to be more exact in its exegesis of theory, then Hennart is doing the field of International Business a service. But it does leave his paper open to the charge that Hennart's thinking has evolved to the point where TCI theory has been stretched to include arguments that actually support M/P theory.
The Multinational Enterprise as Knowledge Exploiter versus Acquirer or Learner
Many pages of Hennart's paper are devoted to (a somewhat persuasive) argument that in multinational enterprises (MNEs)--and especially for firms based in the large home nations of Japan or the US--knowledge flows have been, in the past, far heavier from HQ to foreign affiliates, than vice versa, or between affiliates. He cites patenting and other data on R&D (e.g., Patel/Pavitt 1991) to show that the bulk of R&D expenditures (and hence presumably knowledge generation) occurred in or near the headquarters country. That being said,
a. This past view of the MNE is beginning to change. There is evidence that R&D activities are today beginning to be more dispersed (UNCTAD 2005, 2004).
b. The "reverse" flow of knowledge, from foreign markets to MNE headquarters, is growing in importance and accrues to the benefit of the international firm. Ceteris paribus, presence in more nations increases the likelihood of, and benefits from, foreign ideas. FDI in the US by Taiwanese companies is sometimes motivated more by the desire to learn from the local knowledge cluster in Silicon Valley, than to add value in the United States. An Indian garment firm may have a subsidiary in Milan mainly to be plugged into European design and distribution networks. The garments are sown in India, and then exported to Europe. But without the Italian subsidiary, fashion and delivery concepts cannot be effectively learned at the Indian HQ. Incidentally, this is one of thousands of instances when it is difficult to say whether such new multinational firms have more of an Export, or an FDI strategy. The answer is that both are indispensable to success.
c. Mergers and acquisitions explicitly predicated on acquiring foreign knowledge are increasing (Hoskisson/Kim/White/Tihanyi 2004).
d. Firms based in smaller nations would particularly benefit by learning from abroad. (Not all firms are based in large nations such as the US or Japan).
e. The firm need not depend on its own affiliates to learn about foreign ideas and knowledge. It can rely instead (i) on alliances and network relationships (Contractor/Lorange 2002, Contractor/Ra 2002), or (ii) on local distributors, agents and local contacts in case the foreign market is served by exporting. This point also relates to a methodological critique of M/P studies mentioned by Hennart and others, namely the operationalization of the Degree of Internationalization (DOI) variable. In M/P studies DOI is sometimes based on foreign affiliate sales or number, and sometime on the "Ratio of Foreign to Total Sales" which can include both affiliate sales as well as exports. (I will argue that perhaps the distinction does not matter as much as Hennart makes out).
f. There is another kind of learning, the ability to reproduce an organization and staff in a foreign location. We can describe it as an internationalization capability that is learned through experience. (9) This is indeed a superior management skill or distinctive competency that is learnt in only some firms. If Hennart (1994) allows for such learning and accumulated capability, it is difficult to understand why Hennart in his paper, in this issue, does not allow such learning to be considered a lasting monopolistic advantage which lowers the costs of incremental international expansion and raises net performance in Stage 2.
Old Versus New Views of the Multinational Enterprise
The traditional view of the MNE as principally an exploiter of internalized advantages that it developed in its home nation, is being partially displaced by the new view of the firm as an international learner, coordinator, cross-border arbitrageur, multiple-network alliance partner, and integrator across borders (Palmisano 2006, Contractor/Lorange 2002). The new multinational may also be "born global" (Oviatt/McDougall 1997, p. 86). These are firms that quickly internationalize without the time or need to develop firm-specific internalized advantages in their home nation. Their advantage rests on learning derived from abroad, from their ability to coordinate and arbitrage across national borders, and from alliance network relationships.
At any rate, the advantages of international expansion described in the 3-Stage model above can accrue from either exploitation, multi-country learning, coordination, arbitrage, or access to cheaper inputs. These are not, moreover, mutually exclusive categories. The one common thread, or theme, is that geographical expansion enables the firm to utilize these advantages and reap improved performance (beyond the early Stage 1).
How Does International Scale Improve Performance?
The M/P literature has not articulated exactly how global scale results in incremental benefits from international expansion. There are production economies at the plant level. There are also firm-level economies resulting from global scale in terms of amortizing centralized expenditures over more markets and sales. The M/P literature, (where it does analyze the scale question, rather than make superficial references to "economies of global scale"), seems to be referring most often to firm-level economies (Hitt/Hoskisson/Kim 1997, Contractor/Kundu/Hsu 2003, Thomas/ Eden 2004). The Hennart paper makes a very useful contribution in questioning precisely how and why multinational scale should confer a benefit on a company.
Plant-Level Economies: There are two conditions under which international expansion may not result in further plant-level economies of scale, because MES (minimum efficient scale) is achieved already: (1) When the company is based in a very large domestic market so that MES is already reached, or (2) When, because of engineering reasons relating to a particular production method, MES for this product is smaller than the market size of many nations. (To my knowledge we do not have comprehensive data on how MES compares to national market size in each product area. Hence all-industry generalizations must necessarily remain theoretical abstractions). (10)
However, when MES exceeds the size of the home country market, it is difficult to avoid the conclusion that a firm based in a such a home market can lower its average costs by adding foreign demand (incrementally on to domestic demand) by exporting the additional amount. Massive multi-country scale is needed in sectors such as semi-conductors where a manufacturing "fab" plant requires an investment of over $ 2 Billion. Such production capacity may be concentrated in only a few nations, not only because of engineering scale reasons, but also because of political or other risk factors (Henisz/Macher 2004). It is true, as Hennart (2007, p. 433) states, that the exporting firm "... may have to shoulder the additional cost of overcoming trade barriers." But the reduction in average costs may more than offset the additional trade barrier costs. (For example, transportation costs are only a negligible fraction of value in shipping semiconductors). More pertinently, if the targeted foreign market size is itself significantly smaller than MES, the internationally expanding firm, by combining demand in two or more markets, will have lower costs than its domestic rivals. (11)
Firm-Level Economies of Internationalization: Many multinational enterprises are characterized by having high technology intensity, e.g. high R&D to Sales ratios, and very high ratios of Sales Value over Manufacturing Cost. A bottle of pills selling for $100, may typically have manufacturing costs of only $ 1 or $ 2. The difference of $ 98 or $ 99 is what one may describe as the "technology margin" which serves to amortize the massive R&D outlays of the pharmaceutical company in bringing a new drug to markets. Since (i) the location of R&D expenditures still remains relatively concentrated in a relatively few nations including the home nation of the multinational, (although this is beginning to change according to UNCTAD 2005) and (ii) in many high technology sectors competitive pressures have forced companies to escalate their R&D outlays well beyond levels sustainable by one or even a few nations, it follows that an international scale of marketing operations is imperative to recouping or amortizing the massive research and development outlays. Similar scale is also needed to justify large, sustained advertising expenditures to build up global brand equity.
Simply put, such companies must sell in multiple markets in order to justify and sustain the enormous centralized expenditures on R&D, brand equity, and other intangibles. For centralized expenditures to develop intangible assets there is no MES, as illustrated in Figure 2 below. It is a continuously decreasing function.
[FIGURE 2 OMITTED]
With no MES for amortization of centralized intangibles overhead, the incremental benefits of international expansion, in this respect, should continue to accrue at any scale level, albeit at a diminishing rate. In high-tech industries, with large outlays on R&D and intangibles, a multinational scale of amortization confers a lasting power to the firm over its less multinational rivals. This strategy does not conform to a neoclassical scenario where profits of undifferentiated firms regress towards an industry mean. (12) Instead, such firms compete by escalating their R&D expenditures against each other. Each firm explicitly understands that such expenditures have escalated to a point where a multinational (firm-level) scale is needed, and that since it has to match its rivals' R&D outlays, it will also have to match their multinational scale.
According to finance theory, a multinational firm can improve its "performance" along two separate dimensions (a) Increase in profitability or returns, (b) Reduction in risk or volatility. Here, I use the term "diversification" to refer only to risk reduction by virtue of presence in several geographical areas, (and not to augmentation of profits or returns).
How does geographic diversification reduce risk? The firm derives advantages by being geographically diversified (1) to the extent that the company's markets are in countries with asynchronous business cycles; (2) to the extent that a cocktail of currencies provides a lower overall volatility of cash flows over time; and (3) to the extent that multiple sources of production provide the firm greater flexibility in dealing with disruptions, and thus maintaining uninterrupted supply. By using the phrase "to the extent" I suggest that geographical diversification does not guarantee that all firms will always reduce their risk significantly, or benefit equally.
If, for example, the nations in which a firm operates have more or less synchronized business cycles, little risk reduction advantage will accrue to such a firm. But is this true for all firms? Hennart's 1966 data on US-based MNEs concluded that "... almost half of the US MNE assets abroad were held in Canada and Europe, countries with business cycles highly correlated with the US." (Hennart 2007, p. 448). But these data relate to a past period, and it is worth asking whether US multinational companies have increased their international reach since 1966. (They have done so). Second, Hennart's own findings suggest that more than half of US MNE assets were held in other countries whose business cycles were less highly correlated with the US. Third, using "Assets" as an index of geographical diversification can be misleading since firms sell in far more countries than they have invested assets. That is to say, the geographical reach of Sales (which are recorded via either via exports or FDI affiliates) is far greater than the reach or spread of Assets. Fourth, using more recent data, one wonders if the notion of "highly correlated" business cycles holds up. Using 1979-1999 data, Dueker and Wesche (1999) show that business cycles are closely correlated among France, Germany and Italy but much less so between them and the United Kingdom or the United States. (13)
Having multiple country production sources lowers risk and increases supply chain allocation flexibility. But if there is too much installed capacity in multiple nations, the increase in fixed costs may nullify the flexibility advantage.
International expansion beyond a certain level may increase information and coordination costs too much, since (to use Hennart's (2007, p. 430) words) "... the costs of firm governance rise dramatically with cultural and geographical distance." But where Hennart seems to suggest this as a blanket indictment of M/P theory, his concern is really only a Stage 3 argument (in the 3-Stage general theory). That is to say, it is only in Stage 3 when the firm has over-expanded into smaller, peripheral, and culturally distant markets that incremental governance and information costs rise to outweigh the incremental benefits of further internationalization. Similarly, I would agree with Hennart that there is a "liability of foreignness" (Zaheer/Mosakowski 1997, p. 439) cost. But I aver that this is felt most keenly in Stage 1 of the 3-Stage theory when the early-internationalizing firm is relatively unfamiliar with foreign operations. (14)
Operationalization and Methodology
The Multinationality/Performance (M/P) field has been particularly vulnerable to the charge of decidedly mixed results in studies going back 30 years. This is partly because empiricism ran ahead of theory development, and partly because of operationalization and methodology issues discussed next. Many of Hennart's critiques of methodology are right on target.
Diverse Shapes for the M/P Link
The International Business and other fields has accumulated over 100 empirical M/P studies. For instance, Ruigrok and Wagner (2004) cite more than 180 references, and the 62 studies they analyze reveal no overall consensus as to the shape of the Multinationality/Performance link. Some studies' results seem to suggest a linear relationship between performance and degree of internationalization (DOI); other studies seem to show a U-shape curve; yet other results show an inverted-U-shaped relationship. The recently proposed 3-stage theory (Contractor/Kundu/Hsu 2003, Lu/Beamish 2004, Thomas/Eden 2004) can reconcile these seemingly diverse results by suggesting that linear, U-shaped and Inverted-U-shaped results are simply subsets of the general 3-stage sigmoid curve shown in Figure 3.
In all past studies prior to 2000 (with one exception, namely Riahi-Belkaoui 1998), no one had specified a third order term for multinationality or Degree of Internationalization (DOI). All prior studies had either specified a linear (or first order), and quadratic (second order) term for DOI. Hence it was impossible to find a sigmoid curve in most studies, even though the data in several of the past samples may possibly have supported the 3-stage model. Studies by Geringer, Beamish and Da Costa (1989), Hitt, Hoskisson and Kim (1997), and Gomes and Ramaswamy (1999) specified only a first and second order term and found an inverted-U-shaped relationship, i.e. a positive sign for the first order term and negative for the second order term. By contrast, with a similar first and second order specification, Qian (1998) and others have obtained a U-shaped curve, i.e., first order term negative and second order positive.
Simply put, with the exception of Riahi-Belkaoui (1998) who tested a 3-stage model without much theory development, no one had introduced the cubic term for DOI and tested for a S-shaped curve until Contractor and Newbert (2000), Newbert and Contractor (2001), Contractor, Kundu, and Hsu (2003), Lu and Beamish (2004), and Thomas and Eden (2004).
On the other hand, it is also possible that the data in studies prior to 1998, despite introducing a first, second and third order term for DOI, would not have resulted in a S-curve finding because the statistics for all three stages were not significant. This may also occur in empirical testing in the future, despite first, second and third order terms for DOI being included.
For example, as illustrated in Figure 3a, an inverted-U-shape may result because the first order term for DOI turns out to lack significance. The best statistical fit omits Stage 1 for the sample in question, because the sample belongs to a sector where there are relatively few Stage 1 firms. (Longitudinally speaking, the passage through Stage 1 may be quick and easy for firms in this sector because of conditions in this particular industry). Only Stages 2 and 3 are captured in the results, as illustrated in Figure 3a. The initial positive slope, suggests that the Stage 2 positive effects of international expansion are followed, in Stage 3 for "over-internationalized" companies, by a negative slope, as over-expansion again reduces performance.
On the other hand, as illustrated in Figure 3b, a U-shaped relationship may result because the best statistical fit, in this study, is for the first order term to have a negative sign, the second order term with a positive sign and the third (cubic) order term not significant. This is because only Stages 1 and 2 have statistical significance in this sample. That is to say, such a sample covers a sector with few firms that have over-internationalized. Stage 3 is only sparsely populated. (This could occur in relatively young sectors where few firms have reached maturity, or for emerging country based multinationals that have not yet undergone an extensive internationalization process).
Alternatively, as shown in Figure 3c, only a linear or first order term is statistically significant and represents the best fit for some samples.
Samples vary by sector and national origin. Each may produce different results. But the 3-stage model represents the general theory of the M/P literature. All three stages may not show up in empirical results in all studies. Even within the same study showing complete 3-stage results, the shape of the S-curve may vary by sub-sample, or sub-sector. (See for example results in Contractor/Kundu/Hsu 2003 and Thomas/Eden 2004). The general theory, however, covers all contingencies.
Operationalization of Multinationality or Degree of Internationalization (DOI)
Measures for DOI have ranged from a simple count of numbers of national markets the firm serves, to "Foreign Sales to Total Sales" (FSTS) or "Foreign Employees to Total Employees" or "Foreign Subsidiaries to Total Subsidiaries" or "Foreign Assets to Total Assets" (FATA)--to more complicated measures such as Herfindahllike indexes of geographical dispersion. There are also other composite indexes of DOI from the univariate measures mentioned above.
Hennart is particularly concerned about FSTS because both the numerator and denominator of the ratio can include sales by foreign FDI affiliates, as well as those achieved through exports. This is indeed a problem since the geographic reach (of overall sales) of a multinational usually exceeds the geographical coverage via its affiliates. The difference between the two can be large for some firms. The appropriate approach would be to treat FSTS, and say FATA, as different dimensions or factors (Thomas/Eden 2004) which may be treated separately or combined.
On the other hand one can argue that most of the a priori theory arguments for the M/P link apply regardless of the mix of FDI and Exporting used to reach foreign markets. In 21st Century companies both are simultaneous and combined strategies anyway. The old classic textbook contrast between exporting and FDI as distinct foreign market entry alternatives no longer applies to many companies. Take Tata Consultancy Services (TCS), an Indian information technology and business consultancy, with a European subsidiary based in Cologne, Germany and offices in several other European countries. Some fraction of the value added for European jobs is done in Europe itself, in functions such as customer interface and marketing. However, for most of the IT development and BPO (Business Process Outsourcing) the majority of value is added in India--from where the digitized output is beamed to European clients. A job done for a European client may be recorded in part as an export from India, as well as a sale from Tata Consultancy Services' (TCS) European affiliate.
Is this an export or FDI strategy? Both are indispensable to business success. The comparative advantage of TCS is its location in India and the availability of trained IT professionals, at one third the salary levels in Europe. On the other hand, local European presence, through their European affiliate, is imperative because development of IT systems requires deep understanding of a customer company's administrative and accounting processes. Local presence in Europe by TCS affiliate personnel is also imperative in winning bids and getting orders. TCS Europe's engineers and sales personnel have to be embedded inside European client companies to deliver effective IT, BPO and consulting solutions, as well as to network with alliance partners, since most IT service providers cannot deliver all pieces of a package themselves.
A similar story can be told about Indian garment companies with FDI affiliates in Milan. The Indian production and export base provides the lower cost labor in the Indian garment factory. But without the European FDI affiliate and local Italian presence, fast changing fashion trends and customer needs would not be understood. Production and marketing are separated and occur in India and Italy respectively.
The FSTS ratio thus encompasses both exports as well as Sales by FDI affiliates. For integrated companies such as the ones described above, the distinction is not crucial in terms of their Degree of Internationalization (DOI). M/P theory arguments work congruently for both. (This raises an important issue: What exactly do we mean by "internationalization" or "degree of internationalization"? At the least, it is incumbent on authors to state the exact construction of their DOI variable in their papers).
Other Methodological Issues
As observed above, the nature of the firms in a sample, their national origin, and sectoral characteristics, can powerfully influence the actual shape of the statistically fitted curves obtained. I agree with Hennart that, in past studies, not enough control variables have been introduced, such as controls for sector, MES in relation to typical market size, or country of origin. (15) Hennart also identifies a major problem in M/P studies, of using cross-sectional data to measure what is ultimately an evolutionary, or time-based phenomenon. Rather than using cross-sectional samples, longitudinal studies would provide richer and more precise understanding of international expansion effects. After all, many of the theory arguments are based on a time or experience related process. I agree. (However, this criticism applies to many social science studies. Longitudinal research, spanning years or decades, is understandably more difficult and, by sometimes focusing on a more limited set of companies, runs the risk that the results have limited general applicability).
The economics of production or marketing in different sectors suggest different length of passage through the initial Stage 1. As a result, Stage 1 may appear much shorter in the plotted curve (or not be statistically significant at all) for some sectors, compared to others. (See for instance Contractor/Kundu/Hsu 2003 or Thomas/ Eden 2004). Hence in samples which include different sectors, a sector control variable may be appropriate. Alternatively, a partitioning of the sample to show different shaped curves may be warranted.
The "country of origin" issue has two aspects with somewhat different methodological considerations, (i) the size of the home country market and (ii) the number of significant-sized contiguous markets. Historically, most studies on the M/P link have used data on the internationalization of US-based companies. Only in the last decade or so, more have included European or Japanese firms. For US-based companies, the continental scale of the home market means that using a simple country count measure, or an un-weighted Herfindahl index, DOI for US-based firms would be lower, in relation to company size, at initial stages of international expansion (i.e., with a few foreign markets). Take the example of a company with 80 percent of its sales in the US, and 20 percent in Germany. (I am grateful to the Editor of this special issue for suggesting this example). If the corporate HQ is in the US, Foreign Sales to Total Sales (FSTS) is a ratio of 0.20. If the corporate HQ were to shift to Germany, the FSTS would be 0.80, even though the operations of the firm have not changed. It is only for very large US-based companies operating in say over 50 nations, and hence having the bulk of their sales outside the US, that their DOI would begin to approach that of European firms, ceteris paribus. (16)
In terms of contiguous markets, using the Johanson and Vahlne (1977) argument of incremental investment in culturally, physically, or psychically proximate markets, European firms are surrounded by far more nations whereas the US touches upon only two foreign countries. This means that a European company has a larger number of potential neighboring markets. But it also means that the mean size of each "home" European market is much smaller. Hence domestic market saturation and market-share competition are reached at an earlier point for European, as opposed to US firms. For companies based in emerging nations (e.g. India) that have only recently begun to internationalize, few, if any may have reached Stage 3 (over-internationalization). Hence their fitted curve would, most likely, be U-Shaped, since only Stages 1 and 2 would be well represented in an emerging nation company sample.
Introducing other control or independent variables to explain "Performance" should not make us lose sight of DOI (the main explanatory variable). The fact that control or other variables add explanatory power to the model does not suggest that the general 3-stage theory is invalid. Similarly if some studies find merely U-shapes (Stages 1 and 2 only), or inverted-U-shapes (Stages 2 and 3 only), such findings do not invalidate the general 3-stage theory.
Conclusion: International Business is Good For Companies
The field of International Business Studies rests on the fundamental assumption that international business is good for companies, and that international expansion will benefit financial performance. This paper marshals and articulates the principal theory rationales underlying the Multinationality/Performance sub-field. It concludes that, while internationalization does not necessarily always improve performance (during the initial international expansion stage, or in cases where a firm may have over-internationalized), for the most part--over a considerable range of expansion--international expansion does indeed result in net positive benefits to a firm. Based on modern theories of the firm, this paper articulated seven arguments why internationalization is "good" for company performance over much of the international expansion range: (1) Knowledge Derived from Abroad, (2) Accessing or Arbitraging Cheaper Inputs, (3) Exploitation of Firm-Specific Assets in Foreign Markets, (4) Accumulation of Global Market Power, (5) International Scale, (6) Lowering of Volatility from Geographical Diversification, and (7) Accumulated Internationalization Experience.
The results of over one hundred empirical studies over the past 30 years appear, on superficial examination, to be confusing and even contradictory. Some show a linear positive slope on a Performance versus Degree of Internationalization (DOI) graph. In recent years, as scholars have specified a second-order term for the Multinationality or DOI variable, other studies have shown a U-shaped, or inverted-UShaped result. These seemingly diverse results can be reconciled by the recently proposed 3-stage or S-shaped general theory of international expansion (Contractor/ Kundu/Hsu 2003, Thomas/Eden 2004, Lu/Beamish 2004) which shows that linear, U-shaped, as well as inverted-U-shaped curves are merely sub-sets of the overall S-curve, or 3-stage, general theory. (This is illustrated in Figures 3 a, b, and c in this article). Since different samples used in past studies are based on different industries, each produced different statistical significance for different stages of the internationalization process, resulting in different shaped curves.
But the unassailable fact remains, that in virtually every one of the over one hundred empirical studies of the past, there is a positively sloped segment (within the overall curve) for the Performance vs. Multinationality link. This demonstrates beyond cavil that international expansion produces a net positive effect for companies for some, or considerable, portion of the internationalization range.
Managerial Implication: The trick then, for companies, as a "managerial implication" of this sub-field, is to quickly get through the net negative effects of Stage 1 (initial internationalization) into the net positive zone of Stage 2--but without expanding excessively into Stage 3, a zone where overly-internationalized companies once again encounter a net diminution of financial performance. The middle zone (Stage 2, in Figure 1) is the strategic and financial target. This is not merely a theoretical abstraction. A statistically fitted curve on a carefully selected sample of comparable firms (within a particular sub-sector), would reveal where each firm stands, in relation to others, on the Performance vs. Degree of Internationalization map, thus revealing whether an individual company is near, or far, from its optimum degree of internationalization.
The Firm in Theory and Alternate Views of the Multinational Enterprise: A second objective of this paper was to respond to the critiques of M/P theory in the Hennart paper in this issue. Hennart makes pertinent comments on methodological issues in this field, and shows that some of the theoretical bases and assumptions may not hold up, under certain conditions. For instance, he argues that the claimed benefits of strategic flexibility, resulting from the firm's presence in several nations, has to be set off against higher global total costs--if there is underutilized capacity in the assets invested worldwide. Similarly, he dissects the almost reflexively trotted out argument in many past studies that economies of an international scale are beneficial. My own conclusion on the "scale" issue is to partially agree with Hennart in that plant-level scale effects are not necessarily, or always, beneficial during internationalization. However, I aver that firm level scale effects definitely produce positive effects during international expansion. By taking a deliberately challenging and agnostic position, Hennart is forcing the field to be more precise and articulate in its defense of the purported benefits of internationalization and the multinational firm.
But which multinational firm? There is no single theory of the firm, and Hennart seems to swing between a more modern Transaction Costs/Internalization perspective and an older neoclassical view where the firm is mainly driven by exogenous prices and costs, and where creative management, proprietary distinctive knowledge, and long term strategy play a zero or marginal role. As Demsetz (1983, p. 377) famously wrote about neoclassical assumptions, "... it is a mistake to confuse the firm of economic theory with its real-world namesake."
Where I found several of Hennart's assumptions about multinational companies to be too stringent, or the Transaction Cost lens as providing too limited a view, I have indicated alternative perspectives from the strategy, management and IB literatures. All in all, the debate in this special issue should have a constructive effect in sharpening underlying theory in the M/P field.
But in engaging in such a debate, let us make no mistake about the basic notion--that internationalization is indeed generally "good" for companies.
Manuscript received June 2005, revised June 2006, final version received September 2006.
Bartlett, C./Ghoshal, S., Managing Across Borders: The Transnational Solution, Boston: Harvard Business School Press 1989.
Buckley, P.J., The Limits of Explanation: Testing the Internalization Theory, Journal of International Business Studies, 19, 2, 1988, pp. 181-194.
Buckley, P.J./Casson, M. C., The Future of the Multinational Enterprise, London: Homes & Meier 1976.
Caves, R., International Corporations: Industrial Economics of Foreign Investment, Economica, 38, 149, 1971, pp. 1-27.
Caves, R., Multinational Enterprise and Economic Analysis, Cambridge, MA: Cambridge University Press 1996.
Contractor, F.J., Valuation of Intangible Assets in Global Operations, Westport, CT: Quorum Books, Greenwood Press 2001.
Contractor, P.J./Kundu, S. K./Hsu, C., A Three-stage Theory of International Expansion: The Link between Multinationality and Performance in the Service Sector, Journal of International Business Studies, 34, 1, 2003, pp. 5-18.
Contractor, F. J./Lorange, P., The Growth of Alliances in a New Knowledge-Based Economy, International Business Review, 11, 4, 2002, pp. 485-503.
Contractor, F. J./Newbert, S., The Relationship Between Multinationalitv, Size, and Performance: A Research Note and Empirical Examination, Unpublished Manuscript, Rutgers University, December 2000.
Contractor, F. J./Ra, W., How Knowledge Attributes Influence Alliance Governance Choices: A Theory Development Note, Journal of International Management, 8, 1, 2002, pp. 11-27.
Demsetz, H., The Structure of Ownership and the Theory of the Firm, Journal of Law and Economics, 26, 1, 1983, pp. 375-390.
Demsetz, H., The Firm in Economic Theory: A Quiet Revolution, American Economic Review, 87, 2, 1997, pp. 426-429.
Doz, Y./Santos, J./Williamson, E J., From Global to Metanational: How Companies Win in the Knowledge Economy, Cambridge, MA: Harvard Business School Press Books 2001.
Dueker, M./Wesche, K., European Business Cycles: New Indices and Analysis of their Synchronicity, Working Paper No. 019-B, Federal Reserve Bank of St. Louis 1999.
Geringer, J./Beamish, P./daCosta, R., Diversification Strategy and Internationalization: Implications for MNE Performance, Strategic Management Journal, 10, 2, 1989, pp. 109-119.
Ghoshal, S./Bartlett, C., The Multinational Corporation as an Interorganizational Network, Academy of Management Review, 15, 4, 1990, pp. 603-625.
Gomes, L. K./Ramaswamy, K., An Empirical Examination of the Form of the Relationship between Multinationality and Performance, Journal of International Business Studies, 30, l, 1999, pp. 173-188.
Henisz, W./Macher, J., Firm- and Country-Level Trade-offs and Contingencies in the Evaluation of Foreign Investment: The Semiconductor Industry 1994-2002, Organization Science, 15, 5, 2004, pp. 537-554.
Hennart, J.-F., The Comparative-Institutional Theory of the Firm, Journal of Management Studies, 31, 2, 1994, pp. 193-207.
Hennart, J.-F., The Theoretical Rationale for a Multinationality/Performance Relationship, Management International Review, 47, Special Issue, 2007, pp. 423-452.
Hitt, M./Hoskisson, R. E./Kim, H., International Diversification: Effects on Innovation and Firm Performance in Product Diversified Firms, Academy of Management Journal, 40, 4, 1997, pp. 767-798.
Hoskisson, R./Kim, H./White, R./Tihanyi, L., A Framework for Understanding International Diversification by Business Groups from Emerging Economies, in: Hitt, M./Cheng, J. (eds.), Advances in International Management, 16, 2004, pp. 137-163.
Hoskisson, R./Turk, T., Corporate Restructuring: Governance and Control Limits of the Internal Market, Academy of Management Review, 15, 3, 1990, pp. 459-477.
Hymer, S., The International Operations of National Firms: A Study of Direct Foreign Investment, Cambridge, MA: MIT Press 1976.
Johanson, J./Vahlne, J. E., The Internationalization Process of the Firm: A Model of Knowledge Development and Increasing Foreign Commitments, Journal of International Business Studies, 8, 1, 1977, pp. 23-32.
Kogut, B., Designing Global Strategies: Comparative and Competitive Value Added Chains, Sloan Management Review, 27, 1, 1985, pp. 27-38.
Kogut, B./Zander, U., Knowledge of the Firm and the Evolutionary Theory of the Multinational Corporation, Journal of International Business Studies 24, 4, 1993, pp. 625-645.
Lu, J./Beamish, P., International Diversification and Firm Performance: The S-Curve Hypothesis, Academy of Management Journal, 47, 4, 2004, pp. 598-609.
Newbert, S./Contractor, F. J., An Empirical Analysis of the Relationship Between Multinationality and Performance: A New Internationalization Cycle Theory, unpublished manuscript, Rutgers University, October 2001.
Oviatt, B./McDougall, P., Challenges for Internationalization Process Theory: The Case of International New Ventures, Management International Review, 37, 2, 1997, pp. 85-99.
Palmisano, S., The Globally Integrated Enterprise, Foreign Affairs, 85, 3, 2006, pp. 127-136.
Patel, P./Pavitt, K., Large Firms in the Production of the World's Technology: An Important Case of 'Non-globalisation', Journal of International Business Studies, 22, 1, 1991, pp. 1-23.
Qian, G., Determinants of Profit Performance for the Largest US Firms 1981-92, Multinational Business Review, 6, 2, 1998, pp. 44-51.
Riahi-Belkaoui, A., The Effects of the Degree of Internationalization on Firm Performance, International Business Review, 7, 3, 1998, pp. 315-321.
Ruigrok, W./Wagner, H., Internationalization and Firm Performance: Meta-analytic Review and Future Research Directions, Working Paper, University of St. Gallen 2004.
Siddharthan, N./Lall, S., Recent Growth of the Largest US Multinationals, Oxford Bulletin of Economics and Statistics, 44, 1, 1982, pp. 1-13.
Sunderam, A./Black, S. The Environment and Internal Organization of Multinational Enterprises, Academy of Management Review, 17, 4, 1992, pp. 729-743.
Thomas, D./Eden, L., What is the Shape of the Multinationality-Performance Relationship?, Multinational Business Review, 12, 1, 2004, pp. 89-110.
United Nations Conference on Trade and Development, World Investment Report 2004. New York: United Nations 2004.
United Nations Conference on Trade and Development, World Investment Report 2005. New York: United Nations 2005.
Zaheer, S./Mosakowski, E., The Dynamics of the Liability of Foreignness: A Global Study of Survival in Financial Services, Strategic Management Journal, 18, 6, 1997, pp. 439-464.
(1) By "neoclassical" I mean a world where prices, costs and knowledge are available identically to all, and where prices and costs, rather than companies exercising active management and strategy, play a central role in theory. (See Demsetz 1997)
(2) These include not just end-product adaptation, but internal organizational and cultural adaptation
(3) In Stage 1 the effect on performance is negative. However, longitudinally speaking, a firm that continues to expand into additional markets abroad will, by Stage 2, have added enough new territories so that the incremental benefits of international expansion will then outweigh the costs of further expansion. Cross-sectionally speaking, a sample of firms that includes companies at various stages of internationalization, may well find companies in both Stages 1 and 2. If both stages are statistically significant, the fitted curve will be U-shaped. The limitations of using cross-sectional data to test an evolutionary theory are discussed in a later section.
(4) Hennart does not wish to give much credence to this argument as a significant advantage of internationalization. I will return to this and other of his criticisms in more detail later.
(5) This is not a "one nation" scale perspective treated in detail by Hennart. Rather, I am referring here to the amortization aspect of international scale in a global total sense, i.e., over all markets of the company put together.
(6) Many scholars prefer to reserve the term "diversification" only for risk reduction from presence in multiple nations, and not use the term as a general substitute for internationalization, or benefits from international expansion.
(7) An examination of 10-K or Annual Reports of most multinationals reveals that, even in the largest of companies, fewer than 50 nations are served via controlled affiliates. Sales via exporting cover a much larger number of countries. However the limits on internalized or FDI presence suggests the hypothesis that total internal coordination and administrative costs may escalate too much beyond the top 50 or so nations.
(8) The use of "international diversification" instead of "degree of internationalization" is in my view something to be avoided, unless one is specifically referring to a reduction in overall risk, or volatility, from geographical diversification - as opposed to a change in overall profits.
(9) This benefit would be most strongly seen in improved performance in Stage 2 of the 3-stage model, since in Stage 1(early internationalization), the company is still digesting its learning from its initial forays abroad.
(10) This also relates to the methodological concern in empirical M/P studies which analyze multiindustry samples.
(11) That is true even if all firms have the same cost function, or identical technology (as drawn in Hennart's Figure 1). But in fact, a keystone of MNE theory is that multinational firms are likely to have better technology and lower cost production processes than other, or purely local firms (Hymer 1976, Bartlett/Ghoshal 1989)
(12) The fact that high-technology companies may exhibit high returns in comparison with their accounting (or tangible) assets (e.g., Contractor 2001) does not affect the firm-level scale argument above, although it does have a bearing on methodology in M/P studies where ROA (Return on Assets) is used as an index of performance. This is of methodological concern because the accounts of high technology companies, by ignoring intangible assets, may overstate ROA.
(13) The correlation between the national indexes for the UK and US against the European index averaged near zero during the 1990s. See Figure 8 in Dueker/Wesche (1999).
(14) This is not to deny that, with further internationalization, the expanding firm will continue to experience "liability of foreignness" costs in entering subsequent markets. But these costs will reduce incrementally with greater firm learning and international experience until perhaps in Stage 3 when very culturally and institutionally distant countries are encountered.
(15) Introducing control variables, or context-specific additional variables, does not obviate the general 3-stage theory. Additional variables only qualify or modify the shapes of the M/P curves.
(16) Mixing firms from different nations into one analysis, or not controlling for country of origin, is therefore a methodological problem.
Farok J. Contractor, Professor of Management and Global Business, Rutgers University, Newark, New Jersey, USA.
|Printer friendly Cite/link Email Feedback|
|Author:||Contractor, Farok J.|
|Publication:||Management International Review|
|Date:||Jul 1, 2007|
|Previous Article:||The theoretical rationale for a multinationality-performance relationship.|