Concentrating on Globalization

Pankaj Ghemawat is certainly right to draw attention to the myths surrounding globalization ("The World’s Biggest Myth," November/December 2007). His interpretation of our arguments in our book, The Rule of Three, however, must be clarified.

We argue that through competitive market forces, markets that are largely free of regulatory constraints and major entry barriers eventually become organized into two kinds of competitors: generalists and specialists. Generalists compete across a range of products and markets, and are volume-driven companies for whom financial performance improves with gains in market share. Specialists tend to be margin-driven companies, who actually suffer by increasing their share of the market. Contrary to traditional economic theory, evolved markets tend to be both oligopolistic and monopolistic. Most markets that we have observed end up with three big players who control anywhere from 50 to 80 percent of the market.

There are relatively few industries that today can be said to be truly globalized, in the sense of operating worldwide with few restrictions on where they can produce and where they can sell. Examples include tires and consumer electronics. And indeed, these industries do exhibit high levels of concentration. Many of the industries cited by Ghemawat, such as soft drinks and cement, are simultaneously becoming more concentrated at one end of the market while seeing the emergence of niche players at the other end. The exceptions to this process are industries that are strongly dominated by state-owned enterprises (mining), more subject to nationalistic impulses (national airlines), or consist of natural submonopolies (software). In our book, we clearly identify such examples as exceptions to the Rule of Three. Unrestricted economic globalization, then, does lead to more market share for the biggest players — this is not a "myth."

But it is important to emphasize again that a bigger market share for the biggest players doesn’t necessarily lead to fewer players in the market. Ghemawat fails to observe our point that the emergence of a "Big Three" (companies offering all major product types in the industry to all major market segments) on a global level does not result in a decrease in the total number of companies in the industry. On the contrary, the data show that the total number of firms operating in an industry tends to rise once the industry has arrived at a Big Three structure, which renews itself when the industry moves from being regional to national to global. The reason is that the Big Three tend to withdraw from areas of the market that are not volume-oriented, creating opportunities for niche players to prosper in the shadow of the giants. So although Ghemawat is right to highlight the emergence of greater competition, he does not adequately explain why.

— Jagdish N. Sheth
Professor of Marketing
Emory University
Atlanta, Ga.

— Rajendra Sisodia
Professor of Marketing
Bentley College
Waltham, Mass.

Pankaj Ghemawat replies:

Jagdish Sheth and Rajendra Sisodia’s defense of the Rule of Three is unsurprising. But it is also unconvincing.

Conceptually, Sheth and Sisodia still seem conflicted about whether to measure competition by the raw number of competitors in an industry, or by concentration measures that attempt to capture their size. A large body of work in industrial economics favors the use of concentration measures. Yet Sheth and Sisodia end up favoring the raw numbers. Their letter concludes by arguing that any increases in competition are the result of a rise in the number of niche players in an industry that outweigh the anti-competitive effects of a more concentrated industry. As I mentioned in my article, this focus on raw numbers supplied a bogus rationale for the disastrous DaimlerChrysler megamerger.

Sheth and Sisodia also seem confused about the role of entry barriers in determining concentration levels. They assert that the absence of major barriers tends to lead to highly concentrated industries. Common sense, however, as well as dozens — if not hundreds — of empirical studies by industrial economists, suggests the exact opposite.

Empirically, the assertion that three big players generally control 50 to 80 percent of most markets significantly overstates the concentration levels in the sample of industries I present in my article. Can Sheth and Sisodia be more systematic in presenting their data by, for example, offering a cross-industry compilation of their findings as opposed to providing stray examples? As far as I can tell, they do not do so in their book.

Also note a fundamental measurement problem. Even if concentration within an industry is high or rising, its effects might be blunted by indirect competition that broadens industry boundaries. A good example of that is banking. Many studies demonstrate increasing concentration in banking (albeit to levels well below Sheth and Sisodia’s 50 to 80 percent), but financial innovations and new kinds of financial intermediaries have actually undercut the market power of traditional banks.

The example of banking also reminds us that even if concentration, properly measured, is high, that outcome may still be driven by a desire for empire-building or simple egotism, for instance, instead of the economic logic of competition. At a time when behemoths such as Citigroup are being characterized as unmanageably large, it would seem quaint to celebrate corporate dinosaurs. Yet that is effectively what the Rule of Three does.