Not That Flat: Pankaj Ghemawat Challenges Globalization’s Adherents

Globalization is a catchall term used to explain much of contemporary economic and social developments, from the prevalence of outsourcing to the popularity of fusion cuisine. But economist and professor Pankaj Ghemawat argues that a dive into different data reveals that much of the world is not as connected as it is portrayed to be.

He makes that argument in his book, World 3.0: Global Prosperity and How to Achieve It, and challenged the audience at the recent TED Global Conference in Edinburgh, Scotland, to measure the number of foreign Facebook friends they had. The percentages, he pointed out, usually turn out to be lower than what people expect.

Ghemawat is the Anselmo Rubiralta Professor of Global Strategy at IESE Business School in Barcelona, Spain. Prior to that, he was a professor at Harvard Business School for 25 years, having become the youngest person to become a full professor at Harvard.

The Arabic edition of his previous book Redefining Global Strategy has been released. He spoke to Arabic Knowledge@Wharton about "globaloney," the challenges facing a post-Arab Spring Middle East, and his concerns for the eurozone’s future.

An edited transcript of the conversation follows.

Arabic Knowledge@Wharton: One of the main points you make in World 3.0 is that the world is really semi-globalized, and not as globalized as people think it is. Can you explain?

Pankaj Ghemawat: We are far short of full integration. There are interesting things happening across borders. I wouldn’t be a very good professor of global strategy if I didn’t think there wasn’t anything interesting happening across borders. But the point is people have been talking about the world becoming one since Karl Marx and the Communist Manifesto. And I’ve asked people to look at things that could happen domestically or across borders, and look at the cross-border component as a percentage of the total. When you do that, if you really believe borders don’t really matter, you would expect numbers in the 95% range. Even if Americans formed their friendships on Facebook randomly, given the U.S. is 20 to 25% of Facebook users, some Americans will have other Americans for friends just through random processes. So that’s the complete integration benchmark? For Facebook, it might be that 95% is what we might expect. The actual number is 16%. And that’s actually quite high, compared to other cross-border flows.

If you look at international phone calls, voice-calling minutes is 2%. If you look at online news sources, the number might be 1 to 2% international sources. The same thing applies to non-informational flows; if you look at trade statistics, exports are about 30 % of GDP, and a lot of that is double or triple counting. An iPod component shipped from Japan to China and then shipped to the U.S., it ends up getting counted twice. So nobody really knows how bad the problem is. It’s a major project with the World Trade Organization [WTO]. I did ask Pascal Lamy [Director-General of the WTO] who is doing this project, what his best guess is. And he was saying it was more about 20% rather than 30% of the world’s GDP.

If you look at foreign direct investment and the most recent year we have data for (2010), it’s 10% of all the investment going on all over the world. If you look at people flows, they tend to be very localized as well. One of the things we talk about a lot in business school is the globalization of students. It’s really fun to ask people what percentage of all the university students are studying in the countries other than the ones they’re citizens of. The answer is 2%. You look very differently at higher education when you realize it’s 2% rather than 30%.

Those are the kind of data I tend to rely on to make the point that world is more semi-globalized. I’m using semi in the sense of partial or very partial. I once thought of using the term deci-globalization in the sense it’s more like 10%. If someone asked me to guess the cross-border component of the flow and I don’t have any data, the average of all these statistics I’ve been collecting is about 10%. That would be my presumption absent any information about how globalized things are.

Arabic Knowledge@Wharton: But it’s actually way lower than that?

Ghemawat: Yes, it’s way lower than that. It’s an interesting test for business school faculty. When students from business school graduate, what worldview do they leave Harvard or Wharton? My impression from the surveys I’ve done is that lots of students graduate with the basic belief that the world is completely globalized, differences don’t matter. Both from the business and social perspective, this seems really counterproductive.

Arabic Knowledge@Wharton: You talk about some of the advantages of making the world more globalized in World 3.0. Can you tell us about those?

Ghemawat: [The Qatari capital of] Doha would add 0.1% to the GDP. It’s a little hard for a politician to get really excited. The question is, if it was just that, this wouldn’t be worth bothering about. It would just be simpler to decouple and go our separate ways. It’s really important to understand why the gains are larger than the figures that get pushed out. One notion is the idea not to have complete liberalization. If we had complete liberalization of merchandized trade, the models that get churned out are around 0.5 %. Again, nothing to inspire humongous cheering in the streets, but it’s better than 0.1%.

If you look at the basic structure of most of the academic papers in that vein, trade economists were stuck where business strategists were back in the 1960s and 1970s, focused on volume. I have a much more expanded conception of sources regarding value creation. I focus on value and I think about value a little more broadly than what these models implicitly tend to do, which is a fixed amount of resources and production and reshuffle it across countries to minimize costs. That picks up on some of the potential volume gains in globalization. It picks up on the notion of absolute cost differences.

Economies of scale are almost entirely absent from these models. While I certainly don’t believe economies of scale are as prevalent as a lot of business executives seem to think, nor do I think it is entirely absent from this world as trade economists tend to think. That is one source of value creation that tends to get overlooked in those models. Second, those models have nothing to say about differentiation as opposed to cost and yet we know that, as someone else said, 90% of life is about differentiation rather than costs. So the notion of better products, increased willingness to pay — that’s absent. The notion that globalization can Improve industry structure.

At a WTO conference, I remember a Mexican competition commissioner coming up to me and saying for a country like Mexico with their domestic monopolies and oligopolies, the biggest benefit of globalization is to reduce the likes of [billionaire investor] Carlos Slim. Similarly, an economist model might miss out on risk reduction. There’s no risk in general equilibrium models. They miss out on knowledge creation. It’s actually not that difficult.

Fortunately, we have a whole set of studies that look at this piecemeal. We have one study that looks at economy of scale, and another one that looks at industry structure.

So what I do in this particular chapter in World 3.0 is put together a lot of existing literature rather than just saying the gains are large, suggesting that when you pay attention to the customarily excluded components, you can bring the gains from trade liberalization itself up to 2% to 3% of the GDP without even really relying on very far-fetched estimates. Then there is service liberalization which people think can add and get us up to 4% or 5 % of GDP. Then there are flows other than products and services. There are flows of capital, flows of people, flows of information.

People flows are really the big numbers. Some people at Stanford have done an analysis suggesting complete liberalization of immigration could double the world’s GDP. Of course, that seems a bit optimistic given where we are. So I rely on estimates that more moderate increases might add another couple of percentage points to global GDP, around 3% to 4%. Around immigration, the gains are just huge as all the studies suggest.

Then there are the non-economic estimates. There’s the notion that globalization tends to go hand-in-hand with openness on the political front. There’s the idea that if you recognized the world is only 10% to 20% globalized, the implications of mixing are not homogenization. Not like two waves colliding when one wave takes over another. It’s more like two oceans lapping at the margins. And at the margins, you get some interesting new phenomenon happening. There are complementarities.

It’s well known that immigrants actually foster trade between the country they moved from and the country they moved to. You can imagine most of the linkages across these different categories are positive. It’s actually not that difficult to imagine the order of magnitude turning to 10% of GDP, and that’s a huge number. Policy makers are desperately scrabbling around for growth. And that is ignoring the fact that some people think this is controversial. Globalization may grow 10% larger each year than the previous year but there’s also some suggestion the economy may even grow faster. Plug that into the model and the thing just blows up. It’s the greatest thing since sliced bread. We really need to do it right away.

Arabic Knowledge@Wharton: In World 3.0, you state, "It’s about multiple parallel efforts and diversity, not single-track uniformity." In some ways, it looks like the perspective that wherever you are, you’re the center of the universe, whether it’s the U.S. or Europe. Why is it so hard for countries or corporations to grasp the idea of stylizing an approach specifically to a country?

Ghemawat: Some of it is just intellectual laziness. Some of it is the belief in the hype of the exaggerated levels of integration. In effect, a completely integrated market means you believe things are the same everywhere.

So there are two reasons to combat what I describe as "globaloney." The first reason has to do with what we just walked through — the notion that there are huge gains to additional integration. If our point of departure is we’re already there, then we wouldn’t be attentive to them at all. We wouldn’t spot these opportunities I was just talking about. The other reason to avoid globaloney is that at the business level, it does tend to push you to the "one size fits all" strategy. And then you can see natural managerial biases.

Companies usually go overseas when they run out of room to grow in their own market. So imagine someone like Wal-Mart who believes "we’re the best company in the toughest retail market in the world." I have a great quote from [former Wal-Mart CEO] Lee Scott who suggests, "Look, if we can move from Arkansas to Alabama, how different can Argentina be?" They used to really think that. This is the oldest mistake highlighted by international businessmen. With all the hype around the "flat world," it seems that people need periodic reminding that you need to think hard about what the differences are.

Right now, I get lots of companies coming to talk to me about strategy for emerging markets, especially but not just India. If you just figure out there are a lot of Indians, that’s not a proprietary insight. If you’re trying to achieve any kind of uniqueness you need to go beyond that, from the perspective of your sector, rather than staying at the aggregate levels, the macroeconomic level. You really have to think about what differences matter the most in my sector. In cement, it’s more likely to be geography. In software programming, it’s more likely to be language to a greater extent than geography. In pharmaceuticals, it’s going to be regulations.

The framework I have in my book is called the CAGE model, which is widely used in most business schools in some fashion in their globalization courses. It is just a way to give students who are naturally inclined to focus on the economic and maybe the geographical differences between countries. China is a low-level labor cost country. If you go to China with just that understanding, you’ll be grossly disappointed. CAGE is a little way to give people a broader sense of what the differences are.

Unfortunately, what’s happened in business school is people are subdivided into camps. The scholars of cross-cultural management only believe that culture matters. The institutionalists only believe that only institutions or administrative factors matter. The geographers believe only geography matters. I’m taking more of a practitioner perspective. It’s not particularly useful to look at the world through one set of rather limited perspectives. It’s one of the things that really bother me about what’s going on in academia in developing new knowledge because it seems there are some people who never get beyond looking at culture.

CAGE is a way to think of a laundry list, to think of what differences matter in your business, and to think of strategies for dealing with that. In addition to consciousness-raising function, for trade, we have some estimates down at the industry level; which industries are particularly prone to which kind of differences and to what extent. That’s a rough description.

Information is very sensitive to language, whether you’re talking about followers of Twitter, which is a new data set I’ve constructed. Phone calls are ten times more intense between countries that share a common language.

Arabic Knowledge@Wharton: Recently, Wal-Mart expanded into Africa with the purchase of Massmart. Has Wal-Mart learned its lesson on how to achieve semi-globalization by partnering (or purchasing) with a similar corporation in a local economy?

Ghemawat: It’s complicated because the political situation in South Africa is complicated. The deal was very, very contentious. It nearly got blocked. They did manage to revive it. I wrote my first case about Wal-Mart when they were US$6 billion large. I’ve tracked the company for about 25 years now. If you do a simple growth rate since Sam Walton first disclosed the numbers in the 1970s, the problem is US$6 billion is not going to matter in the Wal-Mart scale of things. That company is still flailing around looking for the next big growth vehicle. They thought developed markets would work but that hasn’t worked out too well with Germany and various other experiences. ASDA (in Great Britain) is okay, but they’ve paid so much for it. So they do have an interest in emerging markets. Most of the Wal-Mart watchers I know don’t see this as shifting things very materially. Personally I think if they can find some way to get around the regulatory barriers, [they can accomplish a] longstanding desire to extend into domestic financial services, which might be a bigger deal than international stores. They’ve tried and the banks won’t let them in. They’ve started to offer nonbanking financial services like money orders. If you think of the typical Wal-Mart market base, lower income and lots of immigrants from Mexico who will send money back. Western Union grossly overprices their stuff.

I also use Wal-Mart, not just as an example of this bias of thinking that "the world is one," but also because they’re a relatively adaptive company. They’ve figured out things so they’re a pretty good illustration in my business-oriented work in three strategies dealing with differences — aggregation, adaptation, and arbitrage.

Arabic Knowledge@Wharton: With the recent unrest in Arab countries, many citizens protested about the lack of economic opportunities and the frustration while their government leaders enjoyed immense wealth. What are some of the recommendations you have for new governments forming in Tunisia? Libya? Egypt?

Ghemawat: This is both a historic opportunity and a huge challenge. If you look at demographics in that part of the world, there is an enormous youthful population looking for jobs. There’s a lot of discussion what’s called the demographic dividend but that’s only if they find jobs for the young, particularly for the men. Otherwise the demographic dividend turns into a disaster. I think we’re at potentially a fateful point in the evolution of this. Because for so many people, if they don’t get jobs, very, very bad things will happen.

It’s important to remember there is a lot of stereotyping that goes on in the West. I think it’s important not to get swayed by this. The India I grew up in was an India where "Made in India" was supposed to be synonymous to the idea that stuff didn’t work. And the Hindu growth rate was naturally dictated by Hinduism to be just 2% to 3%. It’s far better to give people a chance than to say, "Certain cultures just don’t have it."

Thirdly, the level of government heavy-handedness in Arab countries; fundamental institutional change is necessary as opposed to traditional large government, a private sector that is heavily subsidized and doesn’t expand much. I think the policy changes are required. I’m convinced the latent potential is there. On the other hand, the regulatory framework I see in places like Egypt… Until those get changed, it’s hard to imagine local enterprise and growth, as opposed to just people who are historically close to the regime and just grab a disproportionate share.

We have a DHL Global Connectedness Index. The most striking thing is if you look at international interactions, in general they’re governed by the law of distance and a lion’s share of interactions occurs that are close to each other — geographically, culturally, and politically. From the DHL Global Connectedness Index, you look at the Middle East and North Africa, they are ranked the lowest of any other region in the world. There is an obvious explanation for this because this is the region that it has a significant number of oil producers and the oil gets shipped overseas.

It does raise some broader questions about what the future regional integration is going to be. People are happy with the model if you’re Qatar. If most people think the model is to simply ship commodities to Western markets or Eastern markets, at this point like China, the basic point is the developed parts of the world typically tend to have more intercontinental or interregional connectedness than the Middle East does. It’s just striking how badly the region performs. If one was serious about building up different levels of integration, then this will have implications for what kind of trade flows might occur within the Middle East and North Africa. It’s probably going to be light manufacturing and moving away from this pure dependence on primary products to recognizing the state needs to move away from the export mix of these countries.

This is a point at about high regionalization. If you look at Germany’s export maps, the great global exporter, about 65% of their exports are within Europe. Typically, a country is about 50% to 60 % regionalized and you see this across trade, phone calls, financial flows, investment, and people. And sometimes it doesn’t get talked about.

Arabic Knowledge@Wharton: You must have a first-hand perspective of what’s going on in Spain and the eurozone crisis since you’re based in Barcelona at the IESE Business School?

Ghemawat: They seem unable to get growth and productivity. They focus on growth. What I was trying to tell them there are other things besides labor and market reforms. Economists think of things in terms of market failure. I’ve spent enough time in business schools to think of things like management failure. Why are a whole country’s companies subject to certain biases? But it does seem a little hard to say it’s just the government. It has nothing to do with the private sector.

Arabic Knowledge@Wharton: With the euro and the eurozone under threat from countries like Italy and Spain, the country you live in now, what does this say about semi-globalization?

Ghemawat: It’s a reminder that things aren’t perfectly globalized. They’re talking about eurozone problems with some spill over effect on the rest of the world. I think the very discourse we’re having reflects the sense of semi-globalization. For instance, if Spain has problems, it’ll affect France more than the U.S. This would not be the case in a perfectly globalized world. Its not like Spain is entirely unrelated to the U.S. but the estimates are that most of the direct effects are going to be felt locally. When angst rises, other nastier things also come to the fore.

Arabic Knowledge@Wharton: Do you think there is a possibility that the eurozone will disintegrate and will it be a mistake?

Ghemawat: There are a few problems here. One has to do with the countries in Southern Europe. They have a productivity problem because they grew wages and they didn’t grow productivity virtually at all since 1990. So Spain joined the European Union [EU] in 1986 and things seemed to go well for a while. Then the EU expanded into Eastern Europe and those became the low-cost countries. It’s hard for me to see Spain climb out of the productivity hole within the constraints of the eurozone, unless Germany turns massively inflationary or something like that. That seems very unlikely.

The other big problem is there is a tendency to think about what divides Europe very narrowly. Now all the focus is on administrative measures to make the EU stronger. But the point is, just look at the German map. The EU is not one integrated zone. Where does Germany have the highest shares? It’s basically German-speaking places. Austria, Switzerland and Czech Republic have a significant German-speaking population.

The former Central Bank president said Europe is like the U.S. without a fiscal union. The subtext being, if it was fiscally the same, all the problems would go away. So I took that apart. For instance, all along the categories or barriers, Europe is much more fragmented than the U.S. So culturally, it’s not like people in Massachusetts think people in Texas are just like them. That’s nothing compared to what the average German looks at the average Greek or the average Spaniard looks at the average Romanian. The cultural barrier goes unaddressed by the bureaucrats. Geographically, apart from high-density vehicle transportation, it’s still more expensive to move stuff within Europe because of petrol and the roads than in America.

This is slightly different than transportation costs but there is much less mobility. If you’re living in Catalonia, you’re not going to move 100 miles outside Catalonia if you have children in school. Geographically, it’s much harder to get roads, people, etc., moving around Europe than within U.S. Finally, in terms of per capita income, the dispersion factor in Europe is several times as large as what you would see across the U.S. That again, there are dispersions around productivity growth and a whole set of levels.

I think the barriers are addressable. My sense is many European policymakers forget that life is not just about bureaucratic decrees. There are these other categories of barriers that need to be addressed if European integration is going to uphold and progress further.

Given Spain have very low levels of productivity, very low levels of internationalization, very low levels of innovation, and all these things that need to change. Given the barriers within Europe and how bad the situation is and how policymakers have an inability to figure out the other kinds of barriers, besides aid or administration, I worry a great deal about what will happen to the eurozone.

Citing Knowledge@Wharton


For Personal use:

Please use the following citations to quote for personal use:


"Not That Flat: Pankaj Ghemawat Challenges Globalization’s Adherents." Knowledge@Wharton. The Wharton School, University of Pennsylvania, 04 September, 2012. Web. 26 April, 2017 <>


Not That Flat: Pankaj Ghemawat Challenges Globalization’s Adherents. Knowledge@Wharton (2012, September 04). Retrieved from


"Not That Flat: Pankaj Ghemawat Challenges Globalization’s Adherents" Knowledge@Wharton, September 04, 2012,
accessed April 26, 2017.

For Educational/Business use:

Please contact us for repurposing articles, podcasts, or videos using our content licensing contact form.