Photograph by John Madere
But the simplest solution is not always the most effective. So says Pankaj Ghemawat, the most prominent flat-world skeptic in business circles today. Ghemawat is the Anselmo Rubiralta Professor of Global Strategy at the IESE Business School at the University of Navarra in Barcelona, the Jaime and Josefina Chua Tiampo Professor of Business Administration at Harvard Business School, and the author of four books on corporate strategy. His most recent book is Redefining Global Strategy: Crossing Borders in a World Where Differences Still Matter (Harvard Business School Press, 2007). In this intricately reasoned inquiry, grounded in economic research as well as in interviews with CEOs and other strategists, he explores the successes and failures of international forays made by such companies as Coca-Cola, Wal-Mart, the PepsiCo spin-off Yum Brands, Cemex, Whirlpool, Jinro (a Korean purveyor of popular alcoholic spirits in Asia), L’Oreal, Asea Brown Boveri, Cognizant, and IBM. (The book is also reviewed in “Books in Brief,” by David K. Hurst, s+b, Spring 2008.) Ghemawat’s conclusion: The world is in a state of only “semiglobalization.” A multinational expanding beyond its home country’s borders will find success only when it recognizes and manages the critical differences among countries.
Even differences among close neighbors cannot be ignored. For example, Canada and the United States share a 3,000-mile peaceful border, a common language (in many places), and a comprehensive free trade agreement. But the amount of commerce within their own national borders dwarfs the trade between the two countries by a factor of at least five to one. Not that a 20 percent rate of economic interchange should be sneezed at; indeed, Ghemawat himself argues that a well-designed policy for international expansion can make all the difference to a company or a nation. But this policy must be constructed with attention to one’s own strategic capabilities, the time it takes to expand organically, the culture and limits of the home country, and the particular requirements of the customers and industry. In other words, in-depth knowledge and solid business judgment make as much difference with a global strategy as they do with any other kind.
Ghemawat, the son of a botanist, spent part of his early life in a small town in India and part in a small town in Indiana (near Purdue University, where his father earned his Ph.D.). After entering Harvard as an undergraduate in the 1970s, Ghemawat remained to get a Ph.D. in business economics, became a protégé of strategy theorist Michael Porter and negotiation pioneer Roger Fisher, and in 1991, at age 31, became the youngest academic ever to be appointed a full professor at Harvard Business School. He spoke with strategy+business in August 2007, during a break in the annual meeting of the Academy of Management in Philadelphia. In person, Pankaj Ghemawat is a bit of a boundary crosser himself, combining academic formality, pragmatic business sense, and a habit of wide-eyed inquiry in the face of other people’s assumptions. “I imagine him, like Galileo Galilei before the inquisition,” wrote Nikos Mourkogiannis in the foreword to Redefining Global Strategy, “unable to keep from saying, ‘But it does move around the sun!’”
S+B: You start your book by staking a claim that, contrary to Thomas Friedman’s assertions, the world is not flat. Why is that distinction important?
GHEMAWAT: Let’s start with the business implications. I think believing the world is flat leads to one-size-fits-all strategies. This may be one reason that so many firms are disappointed with the performance of their overseas operations. If you’re looking for guidance on questions like where to compete, the “world is flat” thesis has only one message: “The market is almost infinite.” There’s no sense of place, no notion of distance to help you set priorities. So businesses go for bigger and blander strategies.
But I think there are broader implications as well. In both the academic and the political sphere, problematic consequences result from believing that no differences remain between countries. You’re less likely to believe that there is anything interesting to discover in the rest of the world. Right now, we’re in a fairly scandalous situation where only 5 percent of the articles published in the top 20 academic management journals have any cross-border content. Fewer people are conducting research in other parts of the world, looking to build distinctive awareness of other places, as opposed to reinforcing plain-vanilla, “one-approach-fits-every-country” strategies.
Furthermore, I think that the notion that borders don’t matter very much encourages an excessive faith in what Thomas Friedman calls the “golden straitjacket”: the idea that there’s nothing government can do to protect the weaker sections of society, because the capital markets will automatically punish every attempt. This further erodes support for economic integration by leading to alarmist visions about what’s going to happen tomorrow.
S+B: What do you say to the politician or citizen who takes the golden straitjacket message seriously?
GHEMAWAT: We’ve seen global competition before. We’ve seen East Asian economies powered by low labor costs penetrating world markets — but it literally took them decades to do it. And that strikes me as a much more reasonable time frame for some of these processes to unfold. To frame it as an instant, nearly universal economic integration is not accurate, and it leads to political overreaction: “There is no future for my children, so let’s man the barricades against imports and free trade.”
S+B: You’re saying that companies and communities have ample time and strategic options for adapting to globalization.
GHEMAWAT: The press interest in my book has focused on my saying, “The world is not flat.” But that takes up literally only the first half of the first chapter. The rest of the book tries to answer the question, “OK, so what do you do about the world that we live in? How do you cope with this more textured and complex reality?”
If you talk to most business executives, you hear about one technique, which is balancing centralization against decentralization. “We’ll make some decisions centrally, and we’ll delegate some to the provinces or the country managers.” But there is a much richer menu of strategies for dealing with differences.
S+B: In Redefining Global Strategy, you write that natural barriers will keep most industries from achieving anything close to perfect integration. Do you mean primarily economic and structural barriers, or are you talking about culture as well?
GHEMAWAT: I’m referring to a range of barriers that I identified over years of studying corporate global efforts. I called them “distances,” because the distance between places does matter in setting business strategy. There are four basic types of distance: cultural, administrative, geographic, and economic, which form a framework I’ve named CAGE.
Many strategists recognize the importance of geographic distance; if you’ve ever been involved in transporting products, you know that geography is still significant. And certainly we are all attuned to economic distance, differences among nations in labor costs, and so on. But strategists often don’t expect administrative or cultural distances to have the force they do.
One powerful example of this bias is StarTV, which was acquired by Rupert Murdoch’s News Corporation in the early 1990s. The advent of satellite television had destroyed geographic distance as a boundary factor. A broadcaster no longer needed a TV tower on the ground; instead, a satellite source, about 23,000 miles above the earth’s surface, could cover a hemisphere. StarTV went ahead with pan-Asian, English-language programming, aiming at Asia’s wealthiest 5 percent as its audience. But it ran into problems on all the other kinds of distance.
Culturally, people prefer to watch TV programs in their own language; this was already established in research (and in common sense), and that barrier didn’t suddenly evaporate. Nor did the barrier of administrative distance. When Murdoch gave a speech in 1993 about satellite TV being “an unambiguous threat to totalitarian regimes everywhere,” the Chinese government reacted by banning the ownership of satellite dishes — which effectively choked StarTV’s business in that key country.
How could Murdoch make such a miscalculation? Well, one of his insiders told me, “Our operating experience was in the U.S., the U.K., and Australia. In none of these climates is it considered a big deal to rail against authoritarian regimes. We weren’t thinking. We just figured that this was a harmless little bit of filler.” The company has been trying ever since to climb out of the hole it dug with inattention to administrative differences.
The final barrier in the CAGE framework has been economic distance, the lack of developed infrastructure in these countries. In the U.S., there are already methods for measuring audiences, which is a necessity of advertising. Elsewhere, a satellite TV company must wait for Nielsen to install people meters.
All this explains why the net present cost of StarTV to News Corp. has been a couple of billion dollars. Yes, geographic distance doesn’t matter anymore. But the other kinds of distance continue to be critical. StarTV has had to provide programming in diverse languages. They’ve had to pay attention to local political sensitivities, particularly in China. And they’ve had to apply different business models, depending on the level of development of infrastructure in various Asian countries.
S+B: Are the same kinds of differences important for other industries and sectors?
GHEMAWAT: Every company faces its own issues. Let’s compare StarTV with Cemex, the Mexican cement company. Cement is the ultimate commodity. Culturally, it is not a high-touch industry; administratively, it’s certainly not as politically sensitive as media. But geographic distance is critical. It doesn’t involve just the physical number of miles between two countries, but also the ease and openness of transportation. Cemex earns a 20 percent premium, on average, over prices charged by other cement companies, and one reason is the deliberate way Cemex expands into countries where it has shipping access — and often where it controls the marine terminals through which cement is shipped.
The contrast between satellite TV and cement suggests that to make any global strategy actionable, you have to go down to the industry level and think about what kinds of distances matter the most and about your strategies for addressing those distances.
Note how different this is from a more typical approach of keeping the analysis purely at a country level. For example, “Are our prospects better for India or China right now?” Country-level, cross-industry analysis is generally not an adequate way to formulate actual strategy.
S+B: But why wouldn’t a company considering expansion into emerging markets want to compare the prospects of India and China?
GHEMAWAT: First, the comparison depends on where you are looking from. China will generally look closer to a company headquartered in Taiwan or South Korea, and India will look closer to a company from the Middle East.
Second, the relevant measures of closeness depend on the industry — as in the example of satellite TV versus cement.
Third, although country-level differences are meaningful, other levels of analysis may also be relevant. Both India and China are diverse in many ways. People speak many languages within them. There are tax barriers at provincial borders, huge differences in per capita income, and great geographic distances. Some multinationals, like Procter & Gamble, have realized that you can have a macro strategy for a whole country like China, but the provincial level is a more fruitful scale — especially when you consider how much economic activity is localized within each province.
Also note that the economic differences between the Chinese coastal regions and the Chinese hinterland have a parallel in the differences between south India and much of north India, which is being left behind.
S+B: Is that what you mean by “semiglobalization”?
GHEMAWAT: It’s one example. Levels of cross-border integration are currently increasing. They’re even setting new records in some cases. But they still account for only a fraction of economic activity. Instead of being global, most companies that succeed are highly regional: They expand only to the markets where they can navigate the distances.
I should also add that semiglobalization doesn’t just mean that we are striking a balance between extreme localization and extreme standardization. Even that dichotomy — local versus central — implies a single-country view of the world, in which there are two extremes: applying separate strategies country by country, or basically treating the world as one big country. It’s much better to develop a strategy for your own priorities that reflects the granularity of the real world.
S+B: What would that strategy consist of?
GHEMAWAT: It would be a combination, tailored to your own industry, needs, and capabilities, of three basic ways to add value in a world where differences still matter: adaptation, aggregation, and arbitrage.
Adaptation involves adjusting your business model and product offerings to different requirements and tastes around the world. In 1994, Whirlpool’s CEO, David Whitwam, based a global expansion strategy on the belief that consumers everywhere wanted the same kinds of refrigerators and washing machines. But Whirlpool’s financial performance didn’t rise accordingly; in fact, it fell. It turns out that many factors vary from country to country, and the manufacturer must adapt to them: electrical standards, tariffs, climate differences, the growth rate of new households, typical living space, availability of running water, income levels, and cultural tastes, among others.
And product variation is just one kind of adaptation. If you’re serious about expanding internationally, and you’re very profitable at home — like, say, the Turkish home appliance manufacturer Arçelik, which controls a key distribution channel at home with its 2,000 retail stores — you may have to change your metrics. Otherwise, no international opportunity will ever meet the screen that you apply to it.
Another kind of adaptation is product or business redesign to reduce the amount of complexity. For instance, Brunswick, a company that makes boats, figured out that boat engines were relatively similar and not costly to transport. So they focused on selling engines globally, rather than providing the whole end-to-end value chain.
Also of interest in this context is externalization — involving partners or customers in your globalization, so as to reduce the burden of adjustment. When entering an unfamiliar environment, you find a joint venture partner who knows that environment. Or you adopt some of the more creative approaches of YouTube, where customers provide the attractions that draw in other customers.
Innovation is yet another way of adapting, and one can distinguish several different kinds: transfer of insights from one country to another, localized innovation (like Unilever’s laundry soap bars for people who wash clothes by hand), recombination (melding elements of your business model with opportunities from the new country), and even transformation of the local country environment. When Starbucks entered Japan with stores that banned smoking, skeptics said chain-smoking Japanese businessmen would never go there. But instead they transformed the local market by drawing in a nonsmoking, largely female clientele.
The broader point is that there is a long list of strategies for adaptation, rather than just one or two blunt instruments such as product reformulation or decentralization.
S+B: What about aggregation?
GHEMAWAT: That is the recognition that, although there are big differences at the borders, the degree of difference varies a great deal, and therefore you can often group operations so as to minimize differences within a set of countries or some other group. You end up with a more reasonable situation for an individual manager, with geographic regional strategies being the most obvious manifestation.
Of course, there is nothing sacrosanct about using regions (or the minimization of geographic distance) as the basis for aggregation. Different aggregation strategies work for different companies at different points in their evolution. At IBM, I get the sense that [CEO] Sam Palmisano is effectively removing the regional heads from day-to-day line decision responsibility. Instead, he’s putting the product divisions together as the primary corporate structure.
By contrast, I recently talked to Toyota’s chairman, Fujio Cho. He emphasized that Toyota is focused on building up its competitive position at the regional level. One major driver is its sense of administrative distance, a sense that free trade across regions in auto parts and components is not going to happen. But Toyota does anticipate some free trade agreements within regions, which therefore become the fundamental building blocks for conquering new markets. And the other driver, of course, is economic distance: the large difference in gasoline prices between the U.S. — Toyota’s single most important market — and most of the rest of the world. This leads to distinct demands in terms of cars’ size and power.
A regional strategy is sometimes mistaken for a halfhearted approach to globalization: “They couldn’t quite cut it as a global company.” That’s why I like the Toyota example. Few people would regard Toyota as a weak global competitor.
S+B: Didn’t it take Toyota about 50 years to reach its current international position?
GHEMAWAT: Toyota went through successive stages of a regional strategy. They began selling cars outside Japan in the 1950s, but with a center-and-satellite configuration. All the production was concentrated in Toyota City in Japan, and they shipped the cars overseas.
Then they created freestanding regional hubs — for instance, with manufacturing plants located around the world, and some product development capabilities in East Asia and North America.
Next, they shifted toward regional platforms. They still had engineering and other commonalities directed from Toyota City, but they embraced the notion that, ultimately, people in different regions want sufficiently different automobiles that the cars should be designed locally. This is now evolving into networks among the regions, to tap economies of scale. For instance, Southeast Asian plants basically supply the transmission for Toyota’s pickup trucks worldwide.
S+B: What is the limiting factor in an expansion strategy like that? Is it Toyota’s own capabilities?
GHEMAWAT: That is certainly one limiting factor. With the rapid buildup of manufacturing capabilities outside Japan and Southeast Asia, quality is a real issue. In parts of the world, you will see a lot of Toyotas in the repair shop, and they’re concerned about that. They are still trying to figure out how to implant the Toyota Way more deeply in their new regional manufacturing facilities. And they have an impressively deep way of thinking about it. They know there are problems with their Southeast Asia components; they know those problems come up when those components are plugged into downstream manufacturing processes around the world. “But this is exactly how we expose problems,” they say, “with our manufacturing operations.” As they ship stuff back and forth, it becomes clear which regions have conformance problems.
S+B: Your third strategy, arbitrage, is usually associated with getting more labor for less money.
GHEMAWAT: I think of arbitrage as involving a broader range of factors. You can build arbitrage strategies around all four types of distances: cultural, administrative, geographic, and economic.
For instance, the Benihana company has one restaurant in Japan and a couple hundred in other parts of the world, primarily the U.S. The Japanese restaurant is in Tokyo, and it primarily serves Westerners who visit there. It’s really a form of performance art, customized to the U.S. market, wrapped up in a somewhat deceptive mantle of Japanese cooking. To most Japanese, there’s nothing Japanese about Benihana.
But at least there’s a connection, however tenuous, with Japan. The Häagen-Dazs brand was invented in the Bronx; its name is a sheer play on a fictitious cultural association. “It sounds Scandinavian, from a cold country; it must be pure stuff.” The brand was so successful that competitors like Frusen Glädjé followed, also with no direct Scandinavian link. That’s pure cultural arbitrage; it’s commanding a premium price through association with another culture, even though there’s no substantive connection with that culture.
Administrative arbitrage encompasses all the measures that companies take regarding taxes and regulations, including environmental regulations. Tax planning shows up, in anonymous surveys of multinational corporations, as a major factor in their geographic decisions — which is understandable given the huge variations in tax rates around the world. It’s also understandable when, say, tanneries move to less-developed countries because the environmental or health and safety regulations aren’t as strict. I don’t want to celebrate tax dodging or sweatshops, but just to point out that there’s still an awful lot of arbitrage around variations in regulations, both within countries and internationally.
S+B: Is administrative arbitrage feasible as a long-term strategy?
GHEMAWAT: There are some real “dos” and “don’ts” for companies. A multinational from an advanced country has to be particularly careful when running operations in countries where the health, safety, and environmental standards are more lax, not necessarily because it’ll get into trouble in that jurisdiction, but because of the reaction it may face at home.
An executive at a major manufacturer was telling me about the big problem they have in China: To compete cost-effectively, they have to ignore emissions. This company has spent 20 years trying to burnish its environmental credentials. The blowback to its worldwide reputation could be significant. And yet it feels it can’t afford to ignore China, because the market is so large.
Such considerations are, incidentally, one of the reasons for discounting the argument that established brands from industrial countries will tend to win. An India-based foundry is in a better position to take advantage of variations in environmental regulations than a Europe-based company that might be subject to massive challenges in its home market.
S+B: But couldn’t a European company up and move its headquarters to India?
GHEMAWAT: Well, there have been a few examples of companies moving to other nations. And companies may often threaten to move as a way of bargaining for lower tax rates or less onerous regulations. But actually moving is hard. The notion of the utterly footloose corporation is at variance with any careful look at the historical record.
When Accenture moved its headquarters to Bermuda, it must have seemed like a good idea. But if you’ve just relocated to a tax haven, you can’t bid as easily on government contracts. In 2004, a [US]$10 billion Homeland Security contract was held up for precisely this reason. Halliburton is currently moving its headquarters from Texas to Dubai, which makes sense given the number of Middle Eastern clients it has. But it will be interesting to see what negative repercussions arise in the U.S., especially given Halliburton’s high-level political connections.
In this context, it is also worth remembering that the benefits of arbitrage tend to vary by company as well as industry. In the information technology sector, IBM cannot beat Tata Consultancy Services or Infosys on labor costs in India — in part because of the premium IBM pays as a foreign company. Nor can IBM tell its 200,000 Western employees, “Just bide with us a moment while we shift all the jobs over to India.” And so the IBM strategy is to develop services, software, and hardware that will differentiate it.
Sam Palmisano provides a very interesting perspective on arbitrage with his explanation of why IBM expanded from 6,000 people to about 60,000 in India in three years. To be sure, some of it had to do with labor costs, but much of it was driven by talent shortages. Some skills are becoming hard to recruit in the U.S.; if you want 100 qualified voice engineers, you may have better luck finding them elsewhere. The ultimate cost of not arbitraging labor in such a situation is the opportunity cost associated with turning business away. Moreover, some Indian offshore companies, simply by virtue of having to learn to work at a distance, have operations that outpace their Western competitors in documentation and process quality. The Western companies have to learn to meet those standards.
S+B: Michael Porter calls you “one of those rare individuals who combine world-class scholarship with a deep knowledge of business practice.” In that light, what can you say about the state of business school research today?
GHEMAWAT: Actually, I think few business schools take the practice of business very seriously. In 2006, MIT Sloan School Dean Richard Schmalensee published an interesting article in Business Week [“Where’s the ‘B’ in B-Schools?” November 27] in which he said that business schools are focused on research aimed at other researchers, uninformed by practice. [See “Knowledge Review: Lessons for Business Schools,” by Andrea Gabor, s+b, Spring 2008.] I think Harvard, IESE, and a few other business schools do emphasize real-world problems, but they’re more the exception than the rule — particularly among the top B-schools in the U.S.
For instance, in the academic research that does get done on international business, there’s a fascination with firms as knowledge transfer networks. Researchers talk about information flows in network-theory terms, and so on. But when you ask the CEOs of large international companies what they would like to learn more about, knowledge flows don’t rank that high. For Sam Palmisano, the crucial challenge was shifting IBM’s geographic center of gravity. For A.G. Lafley at P&G, it was coordinating outsourcing with a customer focus — and stitching together global business units and market development organizations. At Cisco Systems, appointment of a chief globalization officer has been a key initiative. And on every one of these topics, I have trouble locating academic literature.
S+B: Much of the research for your book involved evaluating the success of corporate strategies. How do you judge whether a company has proven successful? And how long does it take to know that something’s working?
GHEMAWAT: It depends on the industry. In some industries, like financial services, given their low level of disclosure, it’s awfully hard to figure out whether one operation is propping up the rest or the company has actually created a coherent global strategy. Given the corporate-level problems that many “global” financial behemoths have experienced as a result of problems in the U.S. market for subprime mortgages, one often suspects the former to be the case.
But the more general point is the prevailing low level of analysis of global strategies that exists in the business literature. Even a little bit of diligence in looking at perfor-mance data can reveal major misinterpretations of cases of apparent success — and failure.
One case in point is Cemex. Many studies of Cemex have emphasized that it is a model company in postmerger integration, the use of information technology, and the transfer of knowledge from one location to another. And these accolades do have some basis.
But having said that, no one — apart from one article in the Wall Street Journal — bothered to go back and look at price per ton versus costs per ton for Cemex and its competition. It turns out that, although Cemex’s costs per ton are similar to those of its largest global competitor, Holcim, the Cemex price per ton is much higher. This suggests a different interpretation of Cemex’s success: that its deliberate approach to picking structurally attractive markets, or restructuring them — an explanation based on market power rather than efficiency — accounts for most of its superior performance. Such data are often underplayed in analyzing businesses, even though without them, it’s hard to get real about performance.
In a way, that’s a symptom of the same issue that affects writing about globalization in general. Jean de La Fontaine’s aphorism “Everyone believes very easily whatever they fear or desire” captures much of the utopian/dystopian quality of publications about the flat world, the death of distance, the end of history, and so forth. But a reality-based perspective on global strategy leads to different prescriptions. To which I should add, of course, that realism is not a recommendation to stay at home. Columbus managed to believe that the world was round but still took a pretty interesting trip — and discovered some unexpected things on the way!
Reprint No. 08110
Art Kleiner (kleiner_art@ strategy-business.com) is editor-in-chief of strategy+ business and author of Who Really Matters: The Core Group Theory of Power, Privilege, and Success (Doubleday, 2003).