A summary of Pankaj Ghemawat's argument is that proponents of globalization use scant statistics to draw exaggerated conclusions, that such conclusions cannot reasonably be drawn from the data, and that "bridges" and "barriers at borders" are still very much in existence and of stronger consequence than suggested by the rise of Internet telephony. Ghemawat cites examples of how localization—the relevance of "geography, language, and distance"—is a determining factor in what can be accomplished and a deterrent to intensified globalization. The summation of Ghemawat's discussion is that he sees the exaggeration of globalization as threatening nations' sovereignty and antagonizing nations' "tendency to support protectionism." Ghemawat concludes by asserting, after presenting convincing evidence, that enthusiasts of globalization are "painting" a world that "doesn't exist" and that is "unproductive" and "dangerous."

One of Ghemawat's foundational premises is that, according to his calculations, in a "thoroughly globalized environment" foreign direct investment wouldn't be lingering between 10 and 20 percent but would be nearer to 90 percent. One point Ghemawat contests, as put forward by proponents of a globally integrated "flat earth," is that the pace of improvement and the declining costs for voice-over-Internet communication—which was "approaching zero" when "Why the World Isn't Flat" was published—can be asserted as proof that the importance of distance has been obliterated. Ghemawat uses the example of Google opening an office in Moscow to emphasize the continuing relevance and importance of local speech, place and distance.

About Pankaj Ghemawat: Pankaj Ghemawat is the Global Professor of Management and Strategy and Director of the Center for the Globalization of Education and Management at the Stern School of Business at New York University, among other notable accomplishments.

In this article, Pankaj Ghemawat attempts to counter assertions made by Thomas Friedman in his book The World is Flat to the effect that globalization has helped to create a “flat,” highly connected world. His arguments are based on the following points:

• A large percentage (90% or more) of fixed investments all over the world are domestic. In a more connected or flat world, this figure would be lower, pointing to the internationalization of business activities. Accompanying data for the period 2003-2005 shows that aggregate levels of internationalization across various industries for a range of activities does not exceed 10%, with the exception of activities such as patents, portfolio investments, and trade—and these do not exceed 25% or thereabouts, thus, the “10 Percent Presumption.” It could then be said that the world is actually semi-globalized, with ongoing assumptions relating to a highly globalized world being mere exaggerations.
• In spite of great achievements in connectivity across borders, brought about by inventions such as the internet, there still exist geographical boundaries that greatly limit people’s movements and actions. In other words, "we’re more wired, but not more 'global.'"
• The clash between “national sovereignty” and international economic integration greatly limits cross-border connectivity. For example, there is overwhelming evidence that individual countries tend to favor protectionist policies over free trade.

Basically, the author makes a couple of main points.  First, he says that there really is not nearly so much globalization as people think.  For example, he points out, about 90% of all investment is done within countries, not by people from one country investing in another country.  He also points out that even big companies like Google are often much stronger in one nation and much weaker in others.

Second, he says that what globalization there is is fragile and could go away.  He points out that other supposed trends (like the idea that liberal democracies were taking over and all countries would move in that direction) have not panned out.  He says there is no guarantee that globalization will increase.