2

                              

Differences Across Countries

The CAGE Distance Framework

“There are no foreign lands. It is the traveler only who is foreign.”

—Robert Louis Stevenson, The Silverado Squatters, 1883

CHAPTER 1 EMPHASIZED the semiglobalized state of the real world, in which borders continue to matter. This chapter digs deeper into the question of why. The more obvious part of the answer is that large differences arise at borders. The less obvious part concerns how to think about such differences. Instead of treating differences versus similarities in absolute terms, this chapter allows for degrees of difference. It does so by modeling differences in terms of the distances between countries along a variety of Cultural, Administrative/political, Geographic and Economic (CAGE) dimensions. As a result, the CAGE framework not only helps identify the key differences in particular settings; it also affords insights into differences in differences by providing a basis for distinguishing countries that are relatively close, along the key dimensions, from those that are relatively far.

This chapter begins with two vignettes involving Google and Wal-Mart that illustrate the effects of the CAGE dimensions of distance. It then summarizes systematic evidence that multiple dimensions of distance still matter a great deal. This systematic evidence is extended and elaborated into the CAGE framework for understanding the differences between countries, and illustrated with an analysis of China versus India as seen from the United States. The chapter goes on to discuss how the effects of different types of distance between countries are conditioned by industry characteristics, suggesting that the CAGE framework usually has to be applied at an industry rather than cross-industry level. The chapter concludes by reviewing several such applications. The CAGE framework also recurs in the discussion of strategies for globalization and specific strategy levers in part 2 of this book.

Double Trouble with Distance

The example of Google’s difficulties in Russia and China, discussed in chapter 1, touches on all the components of the CAGE distance framework:

• Cultural distance: Google’s biggest problem in Russia seems to have been associated with a relatively difficult language.

• Administrative distance: Google’s difficulties in dealing with Chinese censorship reflect the difference between Chinese administrative and policy frameworks and those in its home country, the United States.

• Geographic distance: Although Google’s products can be digitized, it had trouble adapting to Russia from afar and has had to set up offices there.

• Economic distance: The underdevelopment of payment infrastructure in Russia has been another handicap for Google relative to local rivals.

For a second example of a company that has been very successful overall but has run into a great deal of trouble with distance, consider the case of Wal-Mart, the world’s largest enterprise in terms of sales. Despite its recent labor and nonmarket travails, Wal-Mart is lean and mean in its home base of the United States, where its $240 billion in revenue in 2005 accounted for close to 10 percent of nonautomotive retail sales, according to U.S. Census Bureau data. Wal-Mart’s international sales, while much smaller at $60 billion, have grown much faster and far outstrip those of any other international retailer. But the profitability of its international sales has been substantially less than that of its U.S. sales. Why?

While there are many contributing factors, the one I’ll focus on in this chapter is that Wal-Mart failed to account for distance, broadly defined. Several years ago, CEO Lee Scott was asked about Wal-Mart’s international prospects. His response: “People said we would struggle when we left Arkansas and got to places like Alabama, 600 miles from Arkansas. We even hired a person to work on the cultural differences between Arkansas and Alabama. Then we were told that in New Jersey or New York, our style wouldn’t be successful.”¹

His implication was clear: Our business model has performed well at home, despite the skeptics, so it should also perform well overseas. The predictable consequence: Wal-Mart transferred its basic business model from the United States to overseas and did better in countries similar to the United States than in very different ones.

Consider Wal-Mart’s profitability by major international market in 2004. The estimates in figure 2-1 suggest that only four out of nine countries generated accounting profits that year: Mexico, Canada, the United Kingdom, and Puerto Rico.² Even more interestingly, the profitable countries tend to resemble the United States along cultural, administrative, geographic, and economic dimensions whereas the unprofitable countries do not.

• Two of the profitable countries, Canada and the United Kingdom, share a common language with the United States, whereas none of the unprofitable ones do; the three are also linked by colony-colonizer ties.

• Unlike the unprofitable countries, two of the profitable countries, Canada and Mexico, partner with the United States in a regional free trade agreement, the North American Free Trade Agreement or NAFTA, whereas none of the unprofitable ones do. And a third profitable “country” as classified by Wal-Mart, Puerto Rico, is officially an unincorporated territory of the United States.

• The capital city of each of the four profitable countries is geographically closer to Wal-Mart’s headquarters (international as well as corporate) in Bentonville, Arkansas, than the capital cities of the five unprofitable ones; in addition, Canada and Mexico share a common land border with the United States.

• Economic differences seem to matter as well: it seems a bit harder for Wal-Mart to do well in very poor countries—although the number of data points is very limited.

Having presented two examples of distance undermining performance, I should add that distance isn’t always bad. Wal-Mart, for example, saves more money by procuring low-cost merchandise from China—that is, by exploiting economic distance—than it makes from its entire international store network. This example and, more broadly, arbitrage strategies that exploit distance instead of treating it as a constraint to be adjusted to or gotten around will be discussed at length in chapter 6. All that the discussion so far has meant to suggest is that distance does need to be taken seriously.

Taking Distance Seriously

The suggestion that distance can matter a great deal is borne out by more systematic data. The relevant evidence is potentially vast, encompassing as it does much of the literature on locational effects. A large amount of this literature is focused, however, on interactions over very short distances or at what is effectively a common location (e.g., the literature on agglomeration economies). This strand of work certainly demonstrates the general importance of location-specificity, but is just beginning to venture beyond the dichotomy of same location-different location. For finer-grained characterizations of how the intensity of economic interactions is affected by spatial (and other dimensions of) distance, the literature on so-called “gravity models” in international economics is a better starting point.

What Do the Numbers Tell Us?

International economists have adapted Newton’s law of universal gravitation to describe international economic interactions.³ Thus, the simplest gravity model of international trade predicts that the trade between two countries will be directly related to their economic sizes (a unilateral attribute of each) and inversely related to the physical distance between them (a bilateral attribute). In other words, bigger economies are predicted, as one would expect, to generate more trade in absolute terms, and greater distances between them should inhibit that trade. More sophisticated gravity models add in nongeographic dimensions of distance, as well as unilateral attributes other than the size of each economy. What do the attempts to fit such models to data on international economic interactions tell us about the world in which we live?

Let’s begin by focusing on international trade. Fitted gravity models manage to explain one-half or even two-thirds of the variation in trade volumes by country-pair, which is remarkably good as economic models go. Looking across many such studies, we see that a 1 percent increase in the size of an economy is typically estimated to lead to a 0.7–0.8 percent increase in its total volume of trade. The effect of geographic distance goes in the opposite direction, and is somewhat larger: a 1 percent increase in the distance between (the capitals of) two countries is generally predicted to decrease trade between them by about 1 percent. In other words, the trade volume between countries one thousand miles apart is expected to be five times as large as it would be, other things being equal, if they were five thousand miles apart.⁴

The estimated sizes of the effects of other distance-related variables are even more impressive. Figure 2-2 summarizes the results of some statistical analysis (by Rajiv Mallick and me) of bilateral trade flows from this perspective.⁵ Basically, it implies that two countries characterized by all five of the commonalities listed in the figure should be expected to trade 29 times as much with each other (1.42×1.47×2.88×2.14×2.25) as an otherwise similar country-pair without any of these commonalities.

Such estimates are obviously meant to be indicative rather than exact, but the effects that they highlight do line up with actual cases. Canada, for example, is barely one of the world’s ten largest economies, yet its bilateral trading relationship with the United States is by far the biggest in the world. Geographic proximity is part of the answer, but so are the commonalities with the United States along four of the five dimensions listed in figure 2-2—more than any other country in the world can claim.⁶

But Canadian-U.S. trade data also remind us that economic integration is far from complete. In fact, the real mystery to economists about Canadian-U.S. trade is not why there’s so much, but why there isn’t much more. To see why, consider some more trade data. As of 1988, before NAFTA, merchandise trade levels between Canadian provinces—that is, within the country—were estimated to be 20 times as large as their trade with similarly sized and similarly distant U.S. states. In other words, there was a built-in “home bias.” NAFTA helped reduce this ratio of intranational to international trade—the home bias—from 20 to 1 to a ratio of 10 to 1 by the mid-1990s, and may have further shrunk it since, although it still exceeds 5 to 1. And these ratios are just for merchandise; for services, the ratio is still several times larger.⁷

So, international borders still loom very large, even if one looks at two countries that are very close to each other along most key dimensions. Once again, we’re running into the reality of semiglobalization.

The evidence concerning forms of international economic interaction other than trade generally confirms the importance of distance—both geographic and nongeographic. Thus, significantly negative (overall) distance effects have been detected in foreign direct investment (FDI), equity trading, patent citations, and e-commerce transactions—although the strength of these effects does vary across forms of interaction.⁸ Furthermore, a meta-analysis of nineteen separate statistical studies suggests that distance effects generally haven’t—unlike in the Canadian-U.S. case discussed above—decreased significantly over the course of the twentieth century!⁹

Frameworks for Country Analysis

The evidence just presented suggests that distance effects can be huge. So let’s look at existing tools for “country analysis”—for example, the kinds of due diligence that a company would conduct before deciding to set up shop in a new country—and see how well they account for the effects of distance. The answer, basically, is that they don’t!

Since this is not the place to undertake a detailed review of frameworks for country analysis, one example will have to suffice.¹⁰ Consider the competitiveness indices published by the World Economic Forum. While this is a useful source of cross-country data, most of the categories covered—for example, finance, technology, labor, management, and institutions—focus on the unilateral attributes of countries. The category of openness, which covers tariffs, hidden import barriers, and the like, is multilateral: it measures the administrative distance between a country and the rest of the world. But that still misses out on differences in differences: for example, the idea, which Wal-Mart would have done well to pick up on, that Germany and South Korea—which it has had to exit since figure 2-1 was prepared—are much farther from the United States than are Canada or Mexico. Picking up on such effects requires bilateral measures of distance.

The competitiveness indices are not unrepresentative of other widely used frameworks for country analysis, which also tend to assume that countries can be assessed one by one—that is, unilaterally—against common yardsticks. The trouble with this yardstick approach, though, is that it treats countries as discrete structural objects when they really should be treated as nodes embedded in a network at varying distances from each other. Adding bilateral measures of distance that capture such differences to more familiar unilateral or multilateral attributes is the key contribution of the CAGE framework for country analysis.

Note that the bilateral measures are based on differences between the home country and the foreign country or countries being analyzed, that is, they are anchored in the focal company’s home base. Chapter 1 provided part of the rationale for assuming a home base with its description and debunking of the fallacy of statelessness. Empirically, the clear identification of a company’s home base is generally not difficult and may actually have become easier in recent decades, so the few cases where this is an issue should not be allowed to hold up the rest of the discussion.¹¹ And from a prescriptive perspective, a home base or some other established base of activities is essential to operationalizing the idea that where a company has come from should influence where it goes.

The CAGE Framework at the Country Level

The CAGE framework, as noted, is an acronym for four broad components of distance: cultural, administrative, geographic, and economic. These four components often intertwine: for example, it is hard to imagine countries being close to each other administratively—say, part of a free trade area—unless they also happen to be close culturally, geographically, or economically. Still, it is useful to distinguish between the four components, because they have different bases and, partly as a result, present very different challenges and opportunities. In addition, these four headings also provide a useful way of grouping unilateral influences on cross-border interactions that are specific to particular countries as well as bilateral ones that are specific to particular country-pairs (and multilateral ones too). These various types of influences, summarized in table 2-1, are all discussed below, although the focus falls primarily on bilateral influences, reflecting their novelty as well as the impact identified by gravity models.

The idea of going beyond physical distance in thinking about cross-border strategy is not new. Thus, the idea that would-be internationalizers should go first to countries that presented the least psychic distance—that is, the least “distance between the home market and a foreign market resulting from the perception and understanding of cultural and business differences”—was first proposed thirty years ago.¹² But the CAGE framework takes a much broader view of distance, and has a much more solid empirical base.

Cultural Distance

Culture as used here refers to the attributes of a society that are sustained mainly by interactions among people, rather than by the state (as lawgiver or enforcer). Cultural differences between countries generally tend to reduce economic interactions between them. Languages’ effects in this regard are perhaps the most obvious: look at the first column of figure 2-2. Or for evidence of a different sort, consider classic cross-border faux pas such as the translation of Frank Perdue’s advertising tagline, “It takes a tough man to make a tender chicken” into the Spanish equivalent of “It takes an aroused man to make a chicken affectionate.” In fact, lists of marketing bloopers are usually dominated by such misadventures in foreign languages.¹³

Other aspects of cultural distance that can be measured systematically and have been shown to dampen economic exchange include differences in ethnicity and religion, a lack of trust, and variations in egalitarianism (defined as societal intolerance for abuses of market and political power).¹⁴ Yet other cultural attributes are highly idiosyncratic (e.g., preferences for certain colors) or subtle in the sense of being nearly invisible even to those whose behavior they guide.

Take, for instance, the traditional Chinese tolerance of copyright infringement. Many people ascribe this social norm to China’s recent communist past. But as William Alford argues in To Steal a Book Is an Elegant Offense, it probably reflects a Confucian principle that encourages replication of the results of past intellectual endeavors: “I transmit, rather than create; I believe in and love the Ancients.”¹⁵ Indeed, copyright infringement was a problem for Western publishers well before China’s current growth thrust. Back in the 1920s, for example, Merriam-Webster, about to introduce a bilingual dictionary in China, found that a local publisher had already begun to distribute its own, unauthorized version.

In addition to bilateral attributes that take the form of cultural differences, cross-border economic activity may also be affected by unilateral cultural attributes. Thus, it is intuitively clear that nations with cultures that are insular or even traditional will tend to be relatively closed to international trade and investment, that is, more isolated than others.

Prolonged contact between countries is likely to blunt at least some of the effects of cultural differences between them. Such contact increases mutual familiarity, “seeds” the institutions and organizations required to support cross-border economic activity, and eases cultural adjustments. Broadly speaking, it seems as if many differences in values, norms, dispositions, and unilateral, isolating attributes are likely to prove more malleable, even in the medium run, than differences in language, ethnicity, and religion.

Administrative Distance

Administrative attributes encompass laws, policies, and institutions that typically emerge from a political process and are mandated or enforced by governments. International relationships between countries, including treaties and international organizations, are included as well, on the grounds that these relationships are sustained by the countries that create or support them.

The administrative or political attributes highlighted by gravity models as affecting cross-border economic activity include colonial ties, membership in the same regional trading bloc, and the use of a common currency. The statistical analysis reported in figure 2-2 indicates that colonizer-colony links can multiply trade nearly threefold, even if they have lapsed a long time ago—for reasons that presumably range from cultural familiarity to similarity in legal systems. On FDI, given the limitations of the systematic data, it is more efficient cite specific instances. Thus, between 1997 and 2001, nearly one-half of a huge surge in FDI from Spain was directed at Latin America—about ten times Latin America’s share of world FDI—with Europe’s much larger and physically much closer regional economy getting pushed into second place. This pattern clearly reflects administrative (and cultural) commonalities rooted in colonizer-colony relationships that were formally terminated in the nineteenth century rather than the effects of size or geographic distance.

Preferential trading arrangements and a common currency can also increase trade substantially—even more than colony-colonizer links, if combined (again, see figure 2-2). The integration of the European Union over the last half-century is probably the best example of deliberate efforts to reduce administrative distance among trading partners. Conversely, bad relationships can increase administrative distance. Although India and Pakistan share a colonial past, a land border, and linguistic ties, their long-standing mutual hostility means that official trade between them is less than one-tenth of what gravity models would predict it to be. And in the wake of tensions such as Dubai Ports World’s forced relinquishment of the five U.S. port terminal facilities that it had acquired), various observers have noted a diversion of investments from the Middle East away from the United States.

As the last two examples suggest, administrative distance can be increased or decreased through unilateral measures. In fact, the policies of individual governments pose some of the most common barriers to cross-border trade. In some cases, the difficulties arise in a company’s home country. Companies from member countries of the Organisation for Economic Co-operation and Development [OECD], for example, have to deal with domestic prohibitions on bribery and have to conform to relatively stringent health, safety, and environmental standards—all of which can hinder their global operations. More often, though, it is the target country’s government that raises barriers to foreign investments—through trade quotas, restrictions on FDI, and preferences for domestic competitors in the form of subsidies and favoritism in regulation and procurement.

These are all examples in which a national government is powerful enough to affect outcomes. But a weak institutional infrastructure in a target country can also impede cross-border economic activity. For example, many companies shy away from doing business in countries known for corruption, unreliable legal systems, or social conflict. (Some research suggests that these negative local conditions, if left unchecked, can depress trade and investment far more than any explicit administrative constraint.) Conversely, when a country’s institutional infrastructure is strong, the level of cross-border integration is likely to be higher.

Geographic Distance

The geographic attributes of countries that can affect cross-border economic activity mostly grow out of natural phenomena, although some human interventions may also be involved. This is the part of the CAGE framework that most people first think of when they hear the world distance. And they tend to focus on physical distance, which is in line with the empirical finding—and the commonsense notion—that other things being equal, the farther away a country is, the harder it will be to conduct business there.

But geographic distance is more than simply a matter of physical distance between, say, the capitals of two countries. Other geographic attributes that must be considered include the presence or absence of a common land border, differences in time zones and climates, and, in unilateral terms, access to the ocean, topography, and within-country distances to borders. (I am reminded of former Canadian Prime Minister William MacKenzie King’s complaint, “We have too much geography.”) In addition, man-made “geographic” attributes, such as a country’s transportation and communication infrastructures, may also need to be taken into account—although they can also be treated as economic rather than geographic attributes.

The influence of physical distance deserves additional elaboration. The most obvious impact of longer physical distances is to raise the costs of physical transportation. These, of course, are more important for trade than for FDI, which is why we see a tilt toward FDI as distance increases. But gravity models indicate that FDI also tends to drop off as distance increases—reflecting the fact that physical distance raises communication costs as well as transportation costs. Remember the example of Google’s having to set up offices in Russia to improve its local knowledge and responsiveness.

The broader lesson? Keep the “geography of information,” as well as the geography of physical transportation, in mind when you’re thinking through geographic influences on cross-border economic activity.

Economic Distance

Economic distance refers to differences that affect cross-border economic activity through economic mechanisms distinct from the cultural, administrative, or geographic ones already considered. In this regard, the gravity model flags not only economic size (which increases the absolute amount of trade but decreases trade as a percentage of GDP), but also per-capita incomes. Rich countries engage in more cross-border economic activity (relative to their economic size) than do their poorer cousins. And, as implied by the positive relation between per-capita GDP and trade and investment flows, most of this activity occurs with other rich countries.

Of course, high per-capita income goes hand in hand with higher labor costs. These can be looked at both directly and on a more disaggregated basis—in other words, in terms of different skill levels or types of training. Other factors of production whose cost or quality might be examined in this way include land, natural resources, capital, and more advanced man-made resources such as infrastructure and information.

Finally, it is worth noting that the rich-rich and rich-poor interactions tend to be associated, albeit imperfectly, with the performance of different economic functions. In particular, rich-poor interactions often involve arbitrage, in which a firm matches supply and demand across not within national markets but across them, by slicing up value chains internationally. While cultural, administrative, and geographic differences can also serve as bases of arbitrage, as discussed in chapter 6, economic arbitrage is particularly salient. It is, therefore, the best single reminder that while distance tends to have a dampening effect overall on cross-border economic activity, it may actually encourage such activity in specific situations.

A Country-Level Example: India Versus China from a U.S. Perspective

Let’s use the CAGE framework to look at a specific topic that I’m frequently asked to talk about: how India and China compare, from the perspective of U.S. companies.¹⁶ The comparison is summarized in table 2-2 and elaborated on in the following paragraphs.

Cultural factors. India’s main source of cultural proximity to the United States is arguably its greater use of English. The pool of Indians who know English is estimated to range from less than 100 million to more than 300 million—I’d go with the lower end of the range—but is generally agreed to be larger than China’s. China is generally thought to have an advantage in terms of the size and commercial orientation of its diaspora—although the Indian diaspora in the United States, in particular, tends to be better educated, more recently arrived, and more likely to be involved in the technology sector.

Unilateral cultural characteristics of the two countries yield less clear-cut conclusions. China is more homogeneous linguistically and ethnically, but whether that smoothes progress or makes for too much insularity is a matter of debate. And while India’s class- and caste-ridden social structure is deplorable on broader grounds, Westernized Indian elites may have reinforced Indo-U.S. economic ties.

Administrative factors. Colonization by Britain has created a number of commonalities between India and the United States. The most important of these is that the legal systems in both countries are based on English common law, with its emphasis on precedents and adaptation. China’s legal system, by contrast, relies on civil law—the German version—with its emphasis on principles that are absolute and therefore don’t need to be contextualized. In addition, the political relationships between the United States and India are currently very close. While that situation is subject to change, what does seem fairly certain is that political tensions between the United States and China will persist for decades, if not longer.

The outlook for unilateral administrative and political indicators depends on the time frame adopted. In the short run, multinationals currently seem to see themselves as facing fewer administrative and political obstacles to doing business in China than in India, partly as a result of special economic zones and enclaves such as Hong Kong, and until recently enjoyed preferential tax rates in China. But in the long run, China faces greater challenges than India in establishing the rule of law, protecting private property, restructuring insolvent state-owned enterprises and banks, and dealing with political change.

Geographic factors. Chennai (in India) is 60 percent farther away from Long Beach, California—the busiest U.S. container port—than is Shanghai. But the greater shipping distance is only part of India’s logistical problems: its ports are inefficient and slow, raising the estimated lead time in shipping to the United States to six to twelve weeks, versus two to three weeks from China, and exemplifying the relatively poor state of its infrastructure.

Another key geographic point is that China is the dynamo within a vibrant East Asian subregional economy, with regional partners that account for more than half of inbound FDI and three-quarters of imports. China’s trading relationships with the United States are embedded in, and in some respects enhanced by, this broader network. India, by contrast, is located in a far less economically dynamic subregion, and trade with its South Asian neighbors amounts to less than 5 percent of its total trade.

Economic factors. Unilateral factors warrant particular attention under this heading. China’s economy is reported to be more than twice as large as India’s—although China’s official statistics may overstate its actual economic growth rate by 2–4 percent!¹⁷ Moreover, the Chinese markets for many income-elastic products are more than five times as large as India’s, reflecting the effects of higher per-capita GDP. China’s labor productivity is also higher, in line with higher labor incomes, and its workers are generally better educated—although it trails India in a few higher-end categories (e.g., experienced managers and English-speaking graduates) and faces a somewhat worse demographic outlook, given its one-child policy. China has achieved better outcomes to date by reallocating labor from agriculture to manufacturing, and by mobilizing more domestic capital: its official savings rates—again, probably somewhat exaggerated—are 40–45 percent of GDP, versus 20–25 percent for India.¹⁸

The downside of China’s capital abundance is that it has depressed returns and led to overinvestment, especially in construction and infrastructure, by companies not noted for their self-restraint. Indian companies have been consistently more profitable. In addition, the best Indian domestic companies have typically gotten less support from their home government than have the companies that the Chinese government is trying to build up into global companies. Part of the Indian companies’ response has been a more disciplined, less investment-intensive approach.

Foreign companies account for about 20 percent of Chinese industrial production, and less of India’s. Foreign invested companies have had a disproportionately large impact on Chinese exports, in which the foreigners have grown their shares to more than 50 percent overall and to 80 percent in the higher value-added categories. Since foreign companies account for less than 10 percent of total Indian exports, and since India’s exports have recently run to one-tenth of China’s, foreign companies’ nominal exports out of China are about fifty times as large as those out of India. These figures also say something about the relative levels of development of supply chains in the two countries.

                              

Summarizing very broadly, China seems more attractive than India to general-purpose U.S. investors on many geographic and economic grounds, but less attractive on a number of cultural and administrative grounds.

I’ll add four elaborations to that bald summary. First, the choice of perspective is key. From West Europe, the comparison looks different: China is farther geographically, but on the other hand, India’s English language capabilities are of narrower relevance. And East Europe and North Africa may be more attractive offshoring alternatives than either China or India.

Second, both China and India are very large countries with a great amount of internal variation. For example, both countries’ coastal regions are significantly more vibrant than their hinterlands, suggesting that the CAGE framework can be applied intranationally as well as internationally. Thus, glass manufacturer Saint-Gobain has overtaken longer-established foreign competitors in India by focusing on the coastal south rather than north.

Third, many comparisons of China and India focus on the last column of table 2-2, particularly the points about larger markets and higher labor productivity in China. But the table reminds us of the need to take a broader perspective, the most unexpected conclusion from which is India’s comparative cultural and administrative closeness to the United States. Not coincidentally, these are the two CAGE dimensions that most often get overlooked.

The fourth point is a logical extension of the third. Presumably, India should look more attractive than China as an investment destination in industries that are more sensitive to cultural or administrative distance. The software services industry provides a good example. Culturally, this is a business in which speaking English is particularly important and in which the Indian diaspora in the United States—variously reported to account for more than a third of the workforce of technology companies in Silicon Valley and to run 10 percent of new technology ventures there—has been directly helpful. In addition, geographic distance from the United States matters less and less, especially since the shift toward offshore development, and India has benefited economically from its much larger graduate talent pool. The result: India accounts for more than two-thirds of software services offshored from the United States, compared to roughly one-tenth for China.¹⁹

The software example leads directly to industry-level CAGE analysis, discussed in the next section.

The CAGE Framework at the Industry Level

An investment fund making portfolio investments might be satisfied with an answer to the question of how attractive China is in general terms relative to India. But most executives comparing the two countries are likely to want to do so from the perspective of a particular industry. In such situations, the impact of the differences between countries is conditioned by industry characteristics, which must be taken into account for most applications to company strategy. Table 2-3 summarizes the kinds of industries that are particularly sensitive to each component of distance and cites examples; the rest of this section elaborates.

Cultural Sensitivity

What kinds of products or services are most sensitive to cultural differences? Given our earlier discussion of language as one key determinant of cultural distance, linguistic sensitivity is one obvious indicator: differences in languages matter more in software or TV programming than they do, for instance, in cement. Similarly, one can think of products with a specifically ethnic appeal that are particularly sensitive to ethnic differences or products for which religious differences loom large. Thus, in cross-country statistical regression analyses, food products turn out to rank among the ones most sensitive to cultural distance partly for such ethnic and religious reasons, and partly because they trigger other associations related to the identity of a consumer as a member of a particular community. For example, Americans think of rice as a commodity—like noodles or potatoes—but this foodstuff matters much, much more to the Japanese.

Other cultural differences at the industry level are partially derived from, and therefore blur into, economic differences. The Japanese, for example, prefer their cars to be small, reflecting social norms as well as considerations of economy and convenience in countries where space is limited and therefore treasured.

Finally, while the preceding country-level discussion of culture noted that such differences tend to reduce cross-border economic activity, this general tendency can be counteracted, to some extent, by industry-level considerations. The major countervailing force is strong vertical differentiation by country of origin that makes customers in different countries rank products from a particular country or particular countries as “best.” The leading French champagne houses, for example, have demonstrated that you can use local cachet to build up a global business. The purveyors of American pop culture, from Disney to denim, have made the same point, in a way that reminds us that strong country-of-origin effects need not always be associated with very high quality.

These two examples of vertical differentiation—champagne and Mickey Mouse—fit the adage that the two most global segments in consumer products are the luxury segment and the youth segment. There are two broader takeaways as well about analyzing cross-country variations in preferences:

• Distinguish vertical differentiation from horizontal differentiation, defined as a situation in which consumers in different countries rank the same products very differently (i.e., tastes are different rather than similar).

• Conduct the analysis at a micro level, for example, at the level of champagne rather than beverages, or in terms of bakery products (which are relatively sensitive to distance) versus protein products such as pork and poultry (which are not) instead of lumping both categories into “food.”

Administrative Sensitivity

Administrative distance most often grows out of the desire to protect or regulate domestic industries: local governments see some reasons to intervene to shield industries from outside competition and erect barriers of one kind or another (e.g., tariffs, regulatory complications, local-content laws). In general, these kinds of barriers are most likely to be built if a domestic industry meets one or more of the following criteria:

• It produces staples. Governments are highly likely to interfere in local markets for goods that are perceived to be essential to their citizens’ everyday lives. Food staples, fuel, and electricity, for example, fall into this category.

• It produces an “entitlement” good or service. Similarly, some industries, such as the health-care sector, produce goods or services to which people believe they are entitled as a basic human right. Governments often intervene to set quality standards and to control pricing in such industries as well.

• It is a large employer. Industries that represent large voting blocs often receive state support in the form of subsidies and import protection. Farmers and textile and garment workers are cases in point.

• It is a large supplier to the government. If governments are major buyers (e.g., for mass transit equipment), that obviously widens the scope for governmental intervention as well.

• It is seen as a national champion. Some industries or companies serve as symbols of a country’s modernity and competitiveness. The shoot-out between Boeing and Airbus in the large-passenger-jet market, for example, has generated disproportionate passion on both sides of the Atlantic. This industry is about more than the jobs and dollars (or euros) directly involved.

• It is, or is construed as, vital to national security. Governments will intervene to protect those industries that they deem to be closely linked to national security. Thus, recent examples from the United States include the Dubai Ports World case cited above and the resistance to China National Offshore Oil Corporation’s attempts to buy Unocal.

• It controls natural resources. Other cases from the oil and gas sector—for example, Bolivia’s recent renationalization of its natural gas reserves—illustrate that a country’s natural resources are often considered part of the national heritage and that foreign companies seeking to exploit them can be viewed as plunderers.

• It involves high sunk costs. Industries that require large, irreversible, geographically specific investments—including many of the heavy industries discussed above—are highly vulnerable to holdup by governments once those investments have been made.

An example of an industry that scores high on most of these dimensions—and feels the pain from ignoring them—is the electricity business, defined here to include generation, transmission, and distribution. One of the key “high-tech” fields of the late nineteenth century, this industry witnessed significant foreign investment early on, despite capital intensity of an order previously experienced only with steam railways. But because of the administrative pressures to which foreign ownership in this industry was particularly subject, a tide of “domestication” swept around the world, beginning with the Russian Revolution and running through the late 1970s and early 1980s.

This deglobalization was followed by a revival in interest in foreign direct investment as the electricity sector began to be deregulated around the world. The result has been a global investment bubble, particularly in electricity generation.²⁰ This bubble was fed by more than $400 billion in FDI between 1992 and 2002, and has resulted in more than $100 billion in value destruction, much of it—particularly in emerging markets—due to renegotiation and expropriation by local governments. But it would be even more surprising if this were the last example of a widely shared but wildly inaccurate sense of administrative security.

Geographic Sensitivity

What kinds of industries are most sensitive to geographic distance? The answers, as far as trade flows are concerned, are mostly intuitive: products that have low value-to-weight/bulk ratios (e.g., cement), products prone to hazards or high perishability in transport (e.g., fast foods), or products that present significant local-presence requirements.

Corresponding influences on cross-border investment are harder to specify cleanly, since such investment can serve as either a substitute for or a complement to trade. Thus, researchers have argued both that high local performance or supervision requirements tend to decrease FDI (by constraining trade) and that they tend to increase FDI (by causing investment to substitute for trade). Remember, however, that physical distance has been demonstrated to have a negative influence overall on FDI, as well as on trade. This increases the likelihood of trade and FDI moving hand in hand.

One corroborative example of geographic distance’s having a powerful impact on FDI is provided by the case of Cemex, the Mexican cement company that will be discussed at much greater length in chapter 3. Cemex originally focused on expanding internationally through acquisitions in emerging markets and, after exhausting opportunities in Latin America, went as far afield as Indonesia (which is about as far as you can get from Mexico and still remain on planet Earth). But its more recent acquisitions—as well as an informed source—suggest that it has in fact refocused on the Western Hemisphere in an attempt to build a geographic fortress around itself.

Economic Sensitivity

To take a micro, industry-level perspective on economic distance, it is useful to decompose value for a representative firm in an industry into costs on the supply side, and willingness to pay on the demand side. This microeconomic perspective will be elaborated on further in chapter 3. What will be discussed here are the supply-side and demand-side determinants of sensitivity to economic distance.

On the supply side, economic distance is likely to have the greatest impact on products whose cost structures are dominated by factors with absolute costs that vary a lot internationally. While products with high labor-intensity stand out in this respect, the reality of semiglobalization reminds us that even the costs of factors such as capital are subject to some degree of location-specificity and associated variation.

On the demand side, large differences in willingness to pay—usually associated with per-capita incomes—create incentives to look beyond national borders. But income differences are likely to hurt rather than help international economic activity when they imply preferences for very different kinds of products. Industries that demand lots of variety, agility, or responsiveness are also likely to experience relatively low levels of cross-border international exchange because of the extra complexity costs.

Then there are other, less specific but still useful measures of economic sensitivity. For example, the extent to which economic distance leads to differences in customers, channels, or business systems—or, most broadly, industry structure—across countries is also relevant in assessing the impact of distance at the industry level. Thus, one study has suggested that domestic margins—the costs of domestic transportation, wholesaling, and retailing—play a bigger role in erecting barriers to imports into the United States than do international transportation costs and tariffs combined.²¹

To recap this section, the CAGE framework is usually most usefully applied at the industry level. The task, in other words, is not just to identify the differences between countries but to understand which ones matter the most in the industry of interest to you. This helps bring the analysis down from the macro level to the micro level.

Some Applications

The CAGE framework, once it is taken down to the industry level, lends itself to a very broad array of applications. Let’s focus on five of the most important ones.

Making Differences Visible

One application of the CAGE framework is to make key differences visible. While this application may seem too obvious to be worth belaboring, a case study of Star TV helps show why it merits additional emphasis.²²

Star was launched in 1991 as a satellite TV service for the top 5 percent of Asia’s population. At that time, the use of satellites as gigantic transmission antennae was dissolving the constraints of geographic distance to which terrestrial broadcasters had traditionally been subject. Within its pan-Asian footprint, Star focused on a relatively cosmopolitan elite, which was expected to be able to afford the service, attract advertisers, and be willing to view recycled English-language programming. (This would spare Star the costs of creating new local-language programming.) Rupert Murdoch’s News Corporation, which was betting on satellite instead of cable TV, was sufficiently taken with this business model and the idea of leveraging its English-language programming across Asia—particularly the 20th Century Fox movie and TV program library—to buy out Star’s founder, Hong Kong billionaire Li Ka-Shing, for a total of $825 million by mid-1995.

By 2006, Star was finally making operating profits. Nevertheless, it seems to have been a poor investment for News Corp. The reasons are related to distance. Satellite TV did reduce geographic distance, but it did not address other aspects of distance that Star discounted initially—to its later regret:

• Cultural distance: Star initially assumed that Asian viewers would be satisfied with English-language programming, simply because many in the target demographic spoke English as a second language. The company paid no attention to evidence already available from continental Europe that given a choice, audiences strongly prefer local-language content, even if they do speak foreign languages.

• Administrative distance: News Corporation seemed administratively tone-deaf—especially in a business in which foreign ownership is always politically loaded, given TV’s power to influence people. Shortly after acquiring Star, Rupert Murdoch pronounced satellite TV “an unambiguous threat to totalitarian regimes everywhere,” because it permitted people to circumvent official news sources!²³ The Chinese government reacted by banning the domestic reception of foreign satellite TV services. Much of Murdoch’s China strategy has since involved digging out of this hole.

It is particularly surprising that Murdoch missed the last point, given his personal history—he had to become a U.S. citizen to buy the TV stations that anchored the Fox network—as well as his generally good political instincts. But his international experience and that of News Corp were confined to English-speaking democracies. As it turned out, this was poor preparation for dealing with China.

My broader point is that making key differences more visible—as the CAGE framework does—is important in part because in a very diverse world, many foreign contexts will be alien to many of the managers who must decide on cross-border issues. In such situations, personal experience is not enough. It might not occur to a U.S. speechwriter that there could be a problem with antiauthoritarian rhetoric. Such blind spots can be minimized by being careful to attend to all dimensions of the CAGE framework.

Understanding the Liability of Foreignness

A second application of the CAGE framework is as an antidote to the visions of triumphant multinational companies (MNCs) that often accompany the visions of the globalization apocalypse discussed in chapter 1. The application involves using the framework to pinpoint all the differences between countries that might handicap MNCs relative to local competitors—the so-called liability of foreignness—or more generally affect their relative positions.²⁴ This can be a useful exercise for MNCs, their local competitors or both.

To help stretch minds and overcome a bias toward believing in the inevitable triumph of MNCs, table 2-4 provides a fairly comprehensive list of all the disadvantages that MNCs might suffer relative to local competitors. Consider, as a specific example, the case of beauty products, in which a handful of MNCs, led by L’Oréal of France and Procter & Gamble of the United States, have orchestrated a significant increase in global concentration over the last few decades and now lead in most major markets around the world. One of the biggest exceptions is South Korea, where “local beauty” AmorePacific accounts for more than 30 percent of the market for cosmetics—versus 8 percent for its leading local competitor and 5 percent for L’Oréal, the leading multinational competitor—and has posted operating margins that are among the highest in its industry worldwide. Why has Korea proved such a tough market for MNCs?

The CAGE framework suggests a number of answers to that question. First of all, beauty care products must represent the absolute peak in terms of ego-expressive products subject to cultural biases. Korea, in particular, is obsessed with skin care and makeup, product areas that permit horizontal differentiation around distinct Asian skins and conceptions of beauty—especially white skin in East Asian markets. These influences have combined to limit the cultural appeal of MNCs’ global product lines. In addition, MNCs face extra administrative hurdles that include tariffs, discriminatory product regulations, and initiatives such as the Korean Cosmetics Industry Association’s “Made in Korea products are good for Koreans” campaign. And economically, MNCs have lacked access to door-to-door sales, a very important distribution channel in Korea, which has confined MNCs to the small, high-priced department store channel and denied them scale economies. These are all critically important considerations for a beauty care MNC thinking of entering or expanding in the Korean market—or rethinking its presence there.

Probably the most obvious expedient for trying to overcome the liability of foreignness is to acquire a local competitor. But whether buying a local competitor makes you one depends on the circumstances. Many think that Star TV might have fared better in China if Murdoch had maintained his initial partnership with Li Ka-Shing—and access to Li’s deep relationships with the Chinese government—instead of buying him out completely.

Assessing Natural Owners and Comparing Foreign Competitors

Even if MNCs can be confident that they are going to win out over local players in a particular market, the CAGE framework can be used at a finer level of resolution to shed light on the relative position of MNCs from different countries. Consider, for example, the interesting question of what will happen in Cuba after Fidel Castro’s passage from the scene. Assuming that the country opens up further, will European or U.S. companies win out there?

Cuba’s political relationships with Europe are currently much better, and the nation also has linguistic and colonial ties to one European country, Spain. But the United States is much closer to Cuba along most other dimensions. Its geographic proximity is quite obvious: on a clear night, the violet glow of Miami’s lights is visible from Havana’s harbor. And then there is proximity along at least some cultural dimensions: Cuba is part of the baseball zone rather than the soccer zone, for example. Spain’s language advantage is at least partially offset by the use of Spanish as a second language in the United States, particularly around Miami, which has become a regional hub for Latin business, despite the fact that it is outside the region. Miami is also the hub for the very large Cuban diaspora in the United States, potentially widening the channels of contact between the two countries, even if the channels are not put to good use currently.

In addition, although Cuba was never a U.S. colony (despite repeated U.S. attempts to purchase the island), U.S. big business, including organized crime, dominated the Cuban economy in the decades before Castro’s revolution. Most likely, the large claims that these business interests and the Cuban diaspora in the United States still have pending against Castro’s government will result in significant transfers of assets in any post-Castro normalization. For this reason, I would bet on U.S. companies winning out over European ones except in industries where first-mover advantages can be secured by European companies before U.S. ones move in.

In line with the theme of the previous section, such analysis can also be pushed down to the industry level. Some success has recently been reported with broader efforts to predict which countries’ firms will win out at the industry level in which markets. For example, U.S. firms are somewhat more successful as a group in achieving market leadership in India than in China—and are much more successful in Mexico, where they outperform even Spanish firms in terms of their rate and scope of success.²⁵ But having noted such “natural ownership” advantages, we must understand that they can easily be trumped by other factors—for example, particularly good or bad international strategies.

Comparing Markets

The CAGE framework can also be used to compare markets from the perspective of a particular company. Since I’ve already laid out the basic ideas on this subject, here I’ll present an application that brings together a company and a pair of countries already considered earlier: AmorePacific looking at China versus India.

From the perspective of a Korean company, China has several attractions over India. Perhaps most obviously, New Delhi is nearly three thousand miles away from Seoul, versus less than six hundred miles for Beijing. Reinforcement is provided by an array of historical links between Korea and China: ethnic commonalities that reflect, in part, significant cross-migration; the influence of Confucianism and Buddhism; the ancient kingdom of Kogoryo, which stretched from northeast China to North Korea; and Korean use of Chinese script for a millennium. More recently, Korean movies, TV programs, and musicians have enjoyed such popularity in China that the media in both countries refer to this infatuation as the “Korean Wave.”

These country-level commonalities were reinforced at the industry level by the great influence of the Chinese herbal medicine system on the Korean one—Korea had historically been a transshipment point for herbal medicine from China to Japan—and at the company level, by Amore-Pacific’s focus on ginseng, green tea, and bamboo sap as proprietary ingredients, which also resonated with Chinese traditions. India wasn’t as close to South Korea along any of these dimensions and therefore seemed much more of a challenge.

Discounting by Distance

The examples discussed above are qualitative, but it’s also possible to take a more quantitative approach to assessing the effects of distance. Consider the most common tool that companies use in deciding where to compete: country portfolio analysis (CPA), which includes some measure of market size as one of its principal components. Unfortunately, this is a recipe for exactly the sort of “size-ism” that I described and decried in chapter 1. One remedy is to discount (specifically, divide) raw measures of market size or potential with measures of distance, broadly defined. While such discounting involves numerous approximations, making some adjustments for distance is a better idea, given how much it matters, than refraining from making any adjustments at all.

Consider the case of Yum! Brands, the parent of the Pizza Hut, Taco Bell, and KFC fast-food chains, which was spun off from Pepsi in 1997. At that time, its international operations were very dispersed, with restaurants in twenty-seven countries (although two-thirds of international revenues, and an even higher proportion of profits, came from just seven markets). Furthermore, its debt-service obligations and limited international profitability left it with less than one-tenth as much money as archrival McDonald’s to invest outside the United States. As a result, the head of international operations at Yum! Brands, Pete Bassi, decided to cut the number of its primary equity markets to just ten. But which ten?

Figure 2-3 maps twenty major international markets for Yum! in terms of per-capita income, per-capita fast-food consumption, and total fast-food market size (the areas of the bubbles in the figure). The logic of such country portfolio grids would tend to steer the company toward the larger bubbles to the center and right of the chart in picking its ten primary markets. But note that this would entirely miss out on the effects of distance!

To get a sense of how much difference accounting for distance might make, consider the case of Mexico, which is labeled in the figure, and ranked as the sixteenth of the twenty major markets in terms of total fast-food consumption.²⁶ When this ranking is combined with estimates of low per-capita income and consumption, it appears that Yum! should bail out of Mexico. But when the market size numbers for each country are adjusted for their geographic distance from Dallas—the company’s home base—Mexico jumps to sixth place in terms of market opportunity. And when one further adjusts the numbers to reflect a common land border (the absence of which is assumed to halve the business opportunity) and Mexico’s membership in NAFTA along with the United States (the absence of which is again assumed to halve the opportunity), Mexico climbs all the way into second place—behind only Canada. Of course, not all the obvious adjustments are positive—Mexico’s lack of a common language with the United States pushes it down a bit in the rankings, although it still remains in the top three, along with Canada and the United Kingdom. But the overall message is clear: reasonable attempts to account for distance imply that the market opportunity in Mexico is relatively large.²⁷ In contrast, country portfolio analysis unadjusted for distance might have suggested to Yum! that it pull its equity out of Mexico!

Bassi’s perspective? “Mexico is one of our top two or three priorities.”

In addition to my warning about the procedure being approximate, I should add two other caveats. First, the efficacy of distance discounting depends on the parameters of the situation. It works best when distances between the home base and the various foreign markets being considered vary greatly—a condition satisfied in the Yum! case.

Second and more important, market analysis is only part—and sometimes only a small part—of success. Big successes often require creative thinking about competitive positioning or other dimensions along which new-and-improved strategies might be devised, rather than just mechanical resizing of market potential.

The evolution of Yum! since the time Bassi was deciding how to restructure the non-U.S. operations provides a good example. China, which accounted for 263 units in 1998, has grown to 1,800 units in 2005 and generates more operating income than all the company’s international operations did in 1998. Returns on invested capital in China exceed 30 percent, versus a corporate cost of capital of 9 percent, and Yum! now describes building dominant China brands as its key corporate strategy. It also claims that KFC in China is on track to be “as big as McDonald’s [in the United States] some day.”²⁸ What lies behind this stunning performance?

Very briefly, Yum! repositioned KFC in China to offer extended menus, full table service, and better facilities, reckoning that while China had developed very rapidly, there was a dearth of affordable, casual dining options, particularly ones with assured quality. Yum! China still faces no serious challenger in this booming category.

Note the divergence between this outcome and the predictions of a pure distance-discounted analysis of the market as it was in 1998—which, if performed as described earlier, would have led to China’s barely making the top ten. To make the same point more broadly, attempts to adjust for distance are usually warranted. But they have to complement, rather than substitute for, thoughtful competitive positioning and other elements of strategy—which are explored at greater length in chapter 3 and beyond.

Conclusions

The box “Global Generalizations” summarizes the specific conclusions from this chapter. To provide a broader recap, while the previous chapter, on semiglobalization, argued the importance of the distinction between home and abroad, this chapter pushed farther by recognizing not only this distinction but also the finer-grained one that countries differ greatly in the extent to which they are different from each other. The key innovation in this chapter has been the presentation of a framework, the CAGE framework, that captures such “differences in differences” in terms of bilateral measures of distance along various dimensions. The addition of bilateral measures of distance to traditional models for country analysis permits countries to be represented as the nodes—embedded at varying distances from each other—in a global network.

Having explored the CAGE distance framework and its possible applications, it is worth concluding by noting that distance is not a sufficient basis for setting international strategy, which is why this book doesn’t stop here. The CAGE framework helps us map the global landscape. But to decide how to move across that landscape, we need a more granular understanding of the costs and benefits from crossing borders. For an example, reconsider Wal-Mart’s market entry decisions. While it is striking that the profitability of its store operations declines with distance from Bentonville (figure 2-1), it is more useful to unbundle that relationship and determine that Wal-Mart accounts for 5 percent-plus of retail sales in the non-U.S. markets where it is profitable, versus less than 2 percent in its unprofitable markets. Clearly, its approach to procurement and logistics requires relatively large local market shares to work. The question then becomes this: in light of distance, but also given the company’s strategy, mind-set, and so forth, does the required share seem attainable in a particular target market? Finer-grained analysis along these lines of value creation and its drivers is the topic of chapter 3.