Figure 4-3: ADDING value by opening up
More broadly, the existence of complementarities means that piecemeal evaluations of gains miss out on valuable cross-effects. To the extent that complementarities are important, the estimates discussed so far in this chapter are likely to understate the true potential associated with opening up further.
A second broad consideration that favors opening up further also involves links across choices—in this case, over time—in the form of a psychologically powerful commitment to reducing trade barriers. The nations of the world have sustained this commitment through several decades of global trade negotiations, a fact that in itself paves the way for future negotiations. A major setback, such as a failure to reach even a scaled-back agreement in the Doha trade talks, could signal a shift of momentum, making future gains harder to achieve. Policy expectations are important because trade doesn't just reflect day-to-day decisions based on spot market prices and tariffs. Rather, today's expectations drive long term commitments such as where factories are built, what relationships are cultivated, and even what children learn in school (and not just foreign languages) that shape the conditions of future cross-border engagement.47
In fact, given high unemployment and other macroeconomic pressures, the alternative to conclusion of the Doha round or, more broadly, continuing to move to increase cross-border integration might not be stagnation but reversion toward World 1.0. Illustrating the potential ramifications is a recent study that found that the costs of intensified protectionism might be “almost five times greater than the gains realized from trade creation resulting from the DDA [Doha].”48 Because of large past reductions in tariffs, losses from rolling them back loom much larger than gains from further tariff reductions (although as noted earlier, many, many sources of gains other than tariff reductions exist).
Summing up, I reckon that the considerations covered in this chapter increase estimates of the potential for globalization-related gains from 0.5 percent of global GDP to more than ten times that. But the precise numbers one ends up with are highly subjective. More robustly, I hope I have established that the usual economic models miss many of the ways freer trade can create economic benefits, and they barely begin to account for the ways that freer flows of capital, people, and knowledge would increase world prosperity. It's also important to remember that the gains aren't all about economics: more cross-border exchange also offers cultural and political benefits. And complementarities and commitment are other factors to consider.
Particularly at a time when the general economic outlook remains weak, this creates a strong temptation to simply push the pedal to the metal. But not so fast. The discussion in this chapter has focused, as mentioned at the outset, on the potential gains from globalization, particularly those left out of CGE models. We also need to consider the omitted factors whose inclusion might tilt things the other way if we want to assess net gains rather than just upside potential.
One type of generalization emerges when we recall that CGE models assume perfect markets. They do not offer much of a basis, therefore, for addressing the fundamental question with which this book began: should the rediscovery of market failures—factors that can lead unregulated markets to inefficient outcomes—affect how we view the cross-border integration of markets?
The first few chapters of Part II of this book start to answer this question by analyzing how globalization affects standard types of market failures. Specifically, chapter 5 focuses on how globalization affects concentration problems; chapter 6 studies global externalities; chapter 7 delves into informational imperfections and their risk implications. Then chapter 8 looks at a nonstandard type of failure, global imbalances, in line with concerns that the financial crisis stirred about the self-stabilization properties of markets.
A second type of generalization involves market outcomes that, while unrelated to standard market failures, spark widespread concerns or, frankly, fears about globalization. As an empirical matter, these fears mostly seem to focus on deprivation of various sorts: economic, political, and cultural. Instead of simply shrugging off such fears, I consider them at some length. Chapter 9 focuses on globalization and economic inequality; chapter 10 on political fears about globalization that all involve political disenfranchisement or dependence in some sense; and chapter 11 on cultural fears, particularly fears of homogenization.
The tack taken in all the chapters in Part II is frankly empirical. While there are trade-theoretic models relevant to the issues discussed, they tend to suffer from the same basic problem as the CGE models discussed earlier: their emphasis on embedding market interactions within a general equilibrium framework limits their ability to treat market failures and fears. Thus, models that don't allow for concentrated market structures can't shed much light on whether openness might exacerbate producer concentration. Models of pure exchange with no money or capital offer no visibility into issues of capital contagion. Models in which trade always ends up being balanced do not lend themselves to exploration of the risks, if any, associated with imbalances. Models featuring full or at least fixed employment cannot credibly address concerns that openness will lead to domestic job losses. And so on, although work is under way on each of these problems.
Recognizing these limitations, I haven't attempted in this book to cite or test deductions from trade theory. Rather I've worked inductively through specific cases of market failures and fears, examining how they affect or are affected by a range of cross-border flows, including trade. In other words, I have used the case method developed at Harvard Business School, but with “cases” focused on specific issues that instantiate broader failures and fears—for instance, global warming as an example of the worst kind of externality or food price spikes as a particularly worrying example of the risk of global contagion. The point is not just to come up with specific policy recommendations grounded in reality for the issues considered, but to uncover broader patterns in the nexus between cross-border flows and market failures.
Of course, I'm not an expert on all the issues covered in chapters 5–11; in fact, the issue of global concentration that is addressed in chapter 5 is the only one on which I am recognized as an authority of any kind. Why read the other chapters? They offer concise, and I hope even-handed treatments of some of the key issues of our times—anchored in and illuminated by the worldviews discussed in chapter 1. They therefore constitute a train of argumentation that readers who disagree or want to go farther can follow at least part of the way. And as noted above, they are an essential complement to the argument in this chapter that trade experts tend to think of the potential benefits from globalization too narrowly. Since trade models typically ignore market failures, it is plausible that they underestimate the potential costs of globalization as well as the benefits, leaving a bias of indeterminate direction—unless we actually try to size up the issues discussed in Part II.
To provide a brief preview of the conclusions: the bottom line from Part II is that the market failures and fears commonly cited as reasons to curb cross-border integration are overblown—and that where real dangers exist, so do preventive and corrective measures that typically do not involve closing off borders.
Part Two
Seven Possible Problems
Chapter Five
Global Concentration
Source: Terry Mosher/Aislin Inc.
HOW SHOULD THE rediscovery of market failures—factors that can lead unregulated markets to inefficient or undesirable outcomes—affect how we think about markets' cross-border integration? This chapter focuses on one market failure that has come in for particular discussion in the context of globalization: small numbers of competitors, or high concentration. Both pro- and antiglobalizers associate globalization with what the late William Safire called “the norm of enormity”: they believe that most industries are coming to be dominated by a small number of competitors as markets become more integrated.
This kind of concentration argument is not new
. More than a hundred years ago, Karl Marx wrote, “One capitalist always kills many … [leading to] a constantly diminishing number of the magnates of capital, who usurp and monopolize all advantages.”1 Numerous antiglobalizers have rallied to this point of view. So, more oddly, have some proglobalization business types who should know better. During the 1970s, Boston Consulting Group founder Bruce Henderson promulgated his Rule of Three: “A stable competitive market never has more than three significant competitors.”2 In the early 1980s, Jack Welch seemed to reduce the magic number further with his widely remarked insistence that General Electric be either first or second in its various fields of business. And in the late 1990s, Mercer Management Consulting went to the winner-take-all extreme by popularizing the “plight of the silver medalist”: You're either number one, or you're nowhere.3 Obviously, the next level of hype is the Rule of Zero, in which no one, not even a monopolist, makes money (the eventuality Marx seems to have had in mind).
Many professional managers still buy into such notions. In an online survey I conducted before the global financial crisis, 58 percent of the several hundred managers who responded agreed that “globalization tends to make industries become more concentrated.” In addition, 64 percent believed that “the truly global company should aim to compete everywhere.” These results are yet another manifestation of the tendency toward globaloney and belief in World 2.0 that was discussed in chapter 2. If you feel that cross-country differences don't matter much, you probably find it much easier to think of multinational companies (MNCs) as a few colossi carrying the world on their broad shoulders—or in the case of antiglobalizers, as a few great vampire squid with their tentacles increasingly tightly wrapped around the face of humanity. Either way, a World 2.0 perspective goes hand in hand with a focus on global concentration levels and, as the surveys indicate, an expectation that those levels are increasing over time.
Much of the rhetoric about globalization leading to global monopolies or oligopolies is, as we shall see, overheated. Still, it is useful to start out by understanding why the possibility of concentration is worrisome. In perfectly competitive markets with many equally efficient firms, sellers can't raise prices above costs. As a result, products and services are available to consumers at the lowest possible prices that make production worthwhile for sellers. This is Adam Smith's famous invisible hand. By contrast, (some) economists worry that the few players in concentrated markets may possess market power, that is, be able to raise prices and profits at customers' expense and hurt social welfare. High, stable levels of concentration can also dampen entrepreneurial zeal: why try to start a new firm if the big guys rule? And then there are also concerns about how business concentration might affect political processes—concerns that are addressed in chapter 10.
This chapter focuses on how globalization affects the incidence of this kind of market failure rather than on its consequences. It shows that even if we focus on the global concentration measures preferred by World 2.0 enthusiasts (we'll critique these later in the chapter), there is no clear evidence that globalization leads to increasing industry concentration. In fact, in most situations, globalization appears to promote more competition, not more concentration. Furthermore, a focus on global concentration can miss out on the real market power problems, identification of which requires attention to distance sensitivity in defining the scope of markets. And when a concentration problem is diagnosed, domestic regulation is generally the recommended public policy response, rather than restrictions on cross-border competition. But we also need to be wary of too much government intervention: many restraints on competition are in fact due to governmental action rather than to its absence. It is useful to start out with a detailed case study that illustrates some of these points before developing them more generally.
Auto Delusions
The auto industry is a heavyweight sector that for many people epitomizes big global companies. Five of the world's top twenty MNCs are automakers, more than any other sector except big oil.4 And industry executives themselves have looked to global concentration as the wave of the future since at least the 1980s, when Fiat's chairman, Giovanni Agnelli, envisioned the industry as affording room for no more than six major players worldwide.
The top firms have yet to drive such changes through, however: as figure 5-1 shows, the share of the industry output that they control has, if anything, declined since 1970. Over longer time frames, the decline is even more pronounced. In 2010, the top six companies accounted for 50 percent of global auto production. Back in the 1970s, five companies did; in the 1950s, two; and in the 1920s, just one, Ford. Hardly a chronicle of concentration. And for those who worry about country rather than company concentration, the high point of hegemony was again back in the 1920s, when the United States accounted for 95 percent of world auto production!5 Overall, cross-border integration seems to have intensified competition in the auto industry rather than reduced it; that is, it has improved the I in the ADDING value scorecard. And in addition to reducing bloated margins in formerly cozy oligopolies (e.g., in North America), it seems to have wrought real improvements in cost efficiencies, quality, and design.
If you find these concentration data surprising, you're not alone. When I've presented them to auto executives, disbelief has been the most common reaction. While the CGE modelers discussed in chapter 4 were prone to ignore economies of scale, business executives typically overestimate them. The standard example from the auto industry is the assertion that you can't make less than several million small cars and still be profitable. Yet I have trouble squaring this with the performance of an Indian automaker I know, Maruti Suzuki, which made a million cars for the first time in 2009. Comparing Maruti with a sample of ten top automakers indicates that the former's profitability has exceeded its larger competitors' every year since 2004: its operating margin averaged 11 percent over 2004–2009, versus 6 percent for Honda and Toyota, the most profitable of the ten top competitors, and an average of 3 percent for the ten as a whole. At the same time, Maruti managed to expand unit production at a 13 percent annual clip, versus stagnant production for the top ten as a whole. Back in 2004, Maruti made and sold only half a million vehicles.6
Figure 5-1: Global concentration levels in the auto industry
Source: 1970–1999 based on Center for Global Business Studies; 2000–2009 based on Organisation Internationale des Constructeurs d'Automobiles (OICA), http://oica.net/category/production-statistics/.
How did Maruti outpace its larger rivals? Partly through the help with small-car technology that it receives from its Japanese parent, Suzuki. But Suzuki itself accounts for the production of only 2 million cars a year, which made it the world's tenth-largest automaker in the world in 2009. So while the Maruti Suzuki link might narrow the scale-based disadvantage that Maruti faces versus larger players, it shouldn't, by itself, lead to superior performance.
The most obvious part of the answer is that Maruti still holds nearly 50 percent of the Indian market for automobiles, although that number is coming under pressure as interest in that market intensifies. Local scale/share clearly counts for more than global scale when it comes to localized production, distribution, marketing, and postsales service. The more general point is that except in the extreme case of zero distance and border effects (World 2.0), local market position and structure still matter—even in a “global” industry like autos.
The less obvious part of the answer has to do with how Maruti has managed its early mover advantage in applying, in the early 1980s, up-to-date foreign automotive technology to the Indian market. A highly politicized joint venture between a Japanese automaker and the Indian government was not necessarily fated to succeed. Ever since I met him nearly twenty years ago, I have been struck by the ability of Maruti Suzuki's long-time managing director and chairman, R. C. Bhargava, to negotiate a minefield of competing interests while overseeing a gradual but inevitable migration of capabilities and decision making from the Japanese parent to the In
dian jewel in its crown. The game isn't just about scale or other structural factors: management matters, even and perhaps especially in the “public” sector.
Global Concentration Increases That Weren't
Looking beyond the auto industry, how globally concentrated have industries generally become over the last few decades? While there are no comprehensive data on this topic, I have for more than a decade now (mostly in collaboration with Georgetown University's Fariborz Ghadar) been tracking global concentration levels for selected industries and writing about the patterns in those data. This chapter will summarize some of this earlier work as well as check whether the conclusions from it continue to hold up.
In 2000, Ghadar and I published an article in Harvard Business Review presenting detailed concentration data on three industries: auto production, oil production, and aluminum smelting. We argued that despite feverish merger and acquisition activity in these industries, the hype about global consolidation seemed overblown. While all three had experienced some increases in concentration between the late 1980s and the late 1990s, that trend seemed modest compared to the large decreases in concentration experienced by these industries since the early 1970s—and the even larger ones since World War II.7
In a 2006 article, Ghadar and I followed up by adding eight other industries to the three we had discussed previously.8 These eight were selected because global concentration data were available in volume terms (as opposed to revenue terms) for ten to fifteen years over the 1980s and 1990s.9 Of them, four showed increases in the five-firm concentration ratio: carbonated soft drinks, cement, steel, and paper and board. And four showed decreases: cargo airlines, copper, iron ore, and passenger airlines. Overall, average global concentration in the eleven-industry sample increased slightly between the 1980s and the late 1990s, from 35 percent to 38 percent. But as in the case of autos, changes since the 1980s were greatly overshadowed by the decreases reported since 1950 by the seven industries out of the eleven for which data were available. We concluded that rising hype about global consolidation was not matched by actual changes in global concentration levels—and that over longer time frames, global concentration appeared to have declined steeply!10
World 3.0 Page 10