Highlights from The Great Convergence: radical change will happen when cost of moving people falls as much as the cost of moving ideas in recent years

Understanding the First Unbundling’s Stylized Facts

Chapter 2 identified five top-line facts that marked globalization’s first unbundling: The North industrialized while the South deindustrialized Trade boomed Growth took off worldwide but sooner and faster in the North than in the South The Great Divergence happened Urbanization accelerated, especially in the North.

Falling trade costs produced industrialization in the North and deindustrialization in the South.1

The curtain goes up on the Krugman-Venables story with industry equally dispersed between the North and South since people everywhere were tied to the land and horrible transportation tied industry to people.

The American colonies—which eventually became the United States—were explicitly forbidden from exporting manufactured goods, but exports of raw materials like cotton and wood were encouraged. Colonies were supposed to supply England with raw materials and buy British manufactured goods. Likewise, the emigration of British skilled manufacturing workers and the export of British textile machinery were forbidden by Acts of Parliament.

Globalization is, I believe, in for a radical new transformation, but it will only happen if the cost of moving people falls in the future as much as the cost of moving ideas has in the recent past. The driving force is simple.

If technology opens a sluicegate that allows these people to offer their labor services in advanced economies without actually being there, the impact on jobs could be shocking. And the necessary technology is, I conjecture, not too far away.

John Naisbitt: “The most reliable way to forecast the future is to try to understand the present.”

The three-cascading-constraints view of globalization rests on bedrock made of three costs: the costs of getting goods, ideas, and people from one place to another. Since the timer on modern globalization started in 1820, these costs have generally been compressed by technological advances. Politics, however, have frequently trumped technology.

The rise of international production networks has deeply changed the politics of protection—at least for the nations that are involved in these networks. When a nation’s factories are crossing borders, closing the borders no longer saves jobs even in the short run.

The second unbundling started when falling oil prices provided a powerful tailwind (Figure 62). In inflation-adjusted prices, a standard barrel of oil halved from $40 in 1990 to $20 in 2000; this made it cheaper to move goods internationally. But the second unbundling continued to power ahead in its second decade of this century despite the strong headwind created by a fivefold rise in oil prices. Plainly, oil prices do affect the cost of moving goods, but they are not determinant.

Communication Costs

The trajectory of communication costs seems to be much easier to calculate. The “laws” driving the ICT revolution—Moore’s, Gilder’s, and Metcalfe’s—are in the rising part of their S-curve (see Chapter 3 for details). This suggests that the cost of moving ideas is likely to continue to fall in the coming years—even without any new Star Trek-like technological breakthrough.

It would take a truly radical closing of ICT borders to stop firms from leveraging their knowledge with low-cost labor abroad. My guess is that, at least in G7 nations, the instinct for an open society is stronger than any protectionist instinct that is likely to arise. It seems likely, therefore, that the cost of moving ideas internationally will continue to fall.

Communication Technology (CT), however, is only one half of the ICT revolution. The other half concerns Information Technology (IT).

When thinking about the future of global supply chains, we must speculate about the possibility of truly revolutionary IT developments. One such possible development relates to computer-integrated manufacturing (CIM). This is not futurology. It has already had a big impact. It has already produced a tectonic shift in manufacturing in high-wage nations—moving manufacturing from a situation where machines helped workers make things to a situation where workers help machines make things. The integration and automation of tasks, however, does not stop at the factory gate. Many design, engineering, and management tasks have been computerized.

Face-to-Face Costs, the Virtual Presence Revolution and Telerobotics

The third separation cost—the cost of face-to-face interaction—is also likely to persist on its downward path. More specifically, really good ICT is creating reasonable substitutes for in-person meetings. This “virtual presence revolution” is based on high-quality video and audio systems on both ends of what can be thought of as “the telephone wire.” It is—in essence—really, really good Skype. An example is Cisco Systems’ TelePresence. This combines full-size images of participants, using three plasma screens, sound channels, high-precision microphones, custom lighting, and high-definition cameras. Audio is arranged such that the voices of the participants on the “left” (who could be in Mumbai) sound like they are coming from the left.

The result is much more information being passed among participants than is possible with audio or even standard video conferencing. High-quality video allows a much better reading of faces. Psychological research shows that “microexpressions”—split-second facial changes lasting only a twenty-fifth of a second—can indicate whether a person is concealing an emotion, consciously or unconsciously. These reactions cannot be perceived over regular video calls or Skype and, indeed, these sorts of nonverbal messages are one of the reasons face-to-face meetings generally lead to better understanding and trust than calls or Skype.

Such systems are already deployed in high-end services sectors. They have reduced the need for face-to-face meetings in businesses such as consulting firms and financial services companies. However, they are still expensive and still limited to fixed facilities. If such systems became much cheaper and more mobile, they could significantly reduce the need for specialists and managers to travel to remote factories and offices.

The next step is “holographic telepresence.” This projects real-time, three-dimensional holographic images of people (along with audio) in a way that makes it seem as if the remote person is right next to you. This allows the participants to gauge each other’s full “body language” in an interactive way. This is the stuff of science fiction, but it is not unimaginable. Cisco has already demonstrated a beta version. Interested readers can find videos on this by browsing for “Holographic Video Conferencing.” Telerobotics is another important trend. After all, moving people is not just about people-to-people meetings, it is also about people-to-machine interactions. Keeping a complex production process running usually involves specialists manually engaging with various forms of hardware.  If virtual presence technology were combined with human-controlled robots of the type used in operating rooms, technicians could conduct inspections or undertake repairs from remote locations.

As with telepresence, the widespread use of telerobotics is constrained by high costs. But if it is possible to develop systems that allow surgeons to fix people at a distance, surely it is possible to develop systems that allow technicians sitting in Stuttgart to fix machinery in Brazil. Given the falling cost of manufacturing things, it would seem to be only a matter of time before face-to-face and person-to-machine constraints are relaxed.

An important complement of this trend is the rapid development of computerized translation. When it comes to translating written words, amazing strides have been made in the last ten years or so. Google Translate, for example, was the source of great mirth to bilingual people in 2000. Now it is really quite good and getting better all the time. Even more recently, Apple released “iTranslate,” which translates voice to and from dozens of languages. The language barrier, which has been an important separating force throughout human history, may soon be lowered or even dismantled.

The Future of Production Unbundling

The impact of lower trade costs and lower face-to-face cost quite clearly make it easier to separate production processes into finer stages while simultaneously making it easier to transfer more stages overseas. These trends thus suggest that the unbundling of G7 factories and the offshoring of an ever wider range of production stages is likely to continue.

The push of high Northern wages and the pull of low Southern wages will continue to drain manufacturing jobs from the G7 to an ever widening circle of developing nations.

Some types of ICT—especially better communications technology (CT)—reduce the cost of specialization, in the sense that they make it easier to coordinate a finer division of labor. This pushes toward more fractionalization. Other types of ICT—especially robotics and computerization—reduce the benefit of specialization because they make it easier for one worker to deal with a wider range of tasks. In short, CT is pro-fractionalization, whereas IT is anti-fractionalization. Mobile, always-on, virtual presence would be an extreme example of better communication technology that pushes firms toward an ever finer division of labor.

The decision to relocate production stages abroad turns on the costs and benefits of doing so. Generally speaking, lowering any of the separation costs makes offshoring more attractive given the very large wage differences that exist around the world despite the Great Convergence.

There is always a tendency to locate production near consumers.

As the Great Convergence proceeds, the force of the argument switches from anti-offshoring to pro-offshoring. The number of customers with the means to buy products is, after all, rising faster in the developing world than it is in the developed world.

By lowering face-to-face and face-to-machine interactions, such a revolution would make the geography of international production networks less sensitive to cartographical distance, and thus easier to spread to a wider range of developing nations.

Extreme specialization that allows firms to become the low-cost supplier in particular varieties of intermediate goods (see Chapter 6). In other words, such trade is based on firm-level excellence rather than on wage gaps. This tendency to shift from cost-competitiveness based on low-wage to cost-competitiveness based on firm-level excellence is already underway. Developing nations like China are producing more sophisticated intermediate goods domestically—parts that previously would have been imported. China, for example, is a major supplier of intermediate goods to nations around the world.

As routine, low-skill, and repetitive tasks are easier to computerize and automate, advancing IT is likely to continue eliminating occupations that involves such tasks. At the same time, the more intensive use of sophisticated production machines will make the remaining jobs more skill, capital, and technology intensive. This leads to a polarization of stages in terms of skill content. Routine low-skill tasks are bundled into high-skill occupations, while the remaining low-skill tasks will typically be highly labor intensive but less routine. The resulting broader stages will involve more capital-intensive, more technology-intensive, and more skill-intensive processes. This tends to favor their location in developed nations. Putting it crudely, this trend suggests that there will be factory jobs in G7 nations for high-skill workers and robots. Low- and medium-skill workers will see their jobs eliminated or offshored.

Deep down, this polarization stems from the fact that computers were substitutes for some workers but complements for others, as David Autor, Larry Katz, and Melissa Kearney pointed out in their 2006 paper, “The Polarization of the U.S. Labor Market.”

One of this book’s central premises is that understanding globalization requires a sharp distinction between three types of “separation costs”—trade costs, communication costs, and face-to-face costs.

  • Globalization’s first acceleration—or first unbundling—came when the cost of moving goods plummeted in the nineteenth century.
  • Globalization’s second unbundling came when the cost of moving ideas plummeted in the late twentieth century.
  • A third unbundling is likely to happen if the cost of moving people plummets.

I am talking about technologies that would create very close substitutes to being there in person. The two breakthroughs discussed might accomplish this. The first is really good substitutes for people traveling to be in the same room to exchange brain services (telepresence). The second is really good substitutes for people traveling to provide manual services (telerobotics).

Offshoring, in other words, is a means of arbitraging international wage differences.

Telerobotics, Telepresence, and “Virtual Immigration”

Such arbitrage via offshoring is not possible for all activities. For the offshoring option to work, the firm needs some way of getting Mexican labor services out of Mexico. For many types of manufactured goods, this is easy since the labor services, as mentioned, are added to goods that are then exported. For many other types of activities—especially service activities—labor services cannot be separated from the laborers. For example, the only way to use Mexican labor services to tend to a U.S. garden is to have Mexicans in the garden. Telerobotics could change all this for manual workers. It would allow workers based in developing nations to provide labor services inside developed nations without actually being there. Call it “virtual immigration” or telecommuting for manual workers. Hotel rooms in Oslo could be cleaned by maids sitting in Manila—or more precisely by robots in Oslo that were controlled by Philippine-based workers.

The remote provision of labor services is likely to flow both ways. The general trend would be for low-skilled workers from developing nations to telecommute to rich nations, and high-skilled workers from rich nations to telecommute to developing nations. For example, experienced German technicians could fix German-made capital equipment in China by controlling sophisticated robots placed in Chinese factories.

Telepresence could do the same for brain workers living in developing nations. When telepresence meeting facilities are cheap and portable, and holographic telepresence is widespread, the need for face-to-face meetings will be greatly reduced, even if the need is not eliminated. This will make it much easier to coordinate the provision of brain power at great distances. Given the vast North-South salary differences that exist for engineers, designers, accountants, lawyers, publishers (and let us not forget professors of economics), the ability to fractionalize the production of business services could lead to a great deal of “virtual… Some highlights have been hidden or truncated due to export limits.

This would be nothing more than an amplification of what is already happening. “Microwork” or “micro-outsourcing” is the ability to get individuals to perform small, disjointed tasks as part of a larger project with all the work taking place over the Web. Virtual presence will make the fractionalization and offshoring much easier to coordinate.

All sorts of back-office tasks have been offshored or outsourced already. This could go much further.

In a nutshell, the next radical change in globalization is likely to involve workers in one nation undertaking service tasks in another nation—tasks that today require physical presence. Or to use the unbundling theme, globalization’s third unbundling is likely to mean that labor services are physically unbundled from laborers. Consequences Relaxation of the face-to-face constraint via telepresence and telerobotics would make it much easier to separate the physical application of labor services from the physical presence of laborers. This is likely to produce two monumental changes. The first would stem from developed nation workers and managers applying their talents inside a wider range of developing nations without actually traveling to the nations.

The North-South imbalance in knowledge-per-worker is still quite extreme. Opportunities for arbitraging this imbalance are abundant.

If the geographic extent of the GVC revolution does widen, more developing nations could join the rapid-industrialization parade. This might reignite the commodity super-cycle and the Great Convergence would continue apace. The second set of monumental changes would come from poor nation workers applying their skills in rich nations without leaving home.

For manufacturing sectors, this would be an evolution—a continuation of the unbundling and offshoring trend. But rather than sending production stages abroad to take advantage of lower cost labor, the labor would telecommute to factors that remained in advance economies.

Virtual immigration. For service sectors, the impact is likely to be more revolutionary. Many service sectors were only indirectly affected by the first and second unbundlings since they sold services that were essentially un-tradable. The heart of the un-tradability was the necessity for service providers and service buyers to be physically in the same place at the same time. Really cheap, reliable, and ubiquitous virtual presence technology and telerobotics would break the necessity. Nontraded services would become tradable.  In short, the third unbundling could do to the service sector what the second unbundling already did to the manufacturing sector.

Because something like two-thirds of all jobs are in service sectors, the impact could be historic. In a broad swath of service sectors, rich nation workers could find themselves in direct wage competition with poor nation workers providing their labor services remotely.

According to Erik Brynjolfsson and Andrew McAfee, authors of The Second Machine Age, the near future will be marked by a very systematic use of AI to operate robots that replace humans in high-wage nations.6 The authors point out that this would have large effects for workers ranging from truck drivers to investment managers. I would suggest that “Remote Intelligence” (RI) could end up as at least as transformative.  After all, why go for computer operators when remote human operators would be so much more responsive… Start thinking ahead about hte impact of RI not just AI.

“Things have changed so much that not even the future is what it used to be.” I hope this book serves as a reminder that today’s globalization does not resemble your parents’ globalization. And tomorrow’s globalization is very likely to be quite different from today’s. The baseline reason is that the driving forces changed. Until the late twentieth century, the main driver was a massive cut in the cost of moving goods, which was ultimately triggered by the steam revolution.  The main driver switched to phenomenal drops in the cost of moving ideas when the ICT revolution came along. In the future, the main driver may be transformative reductions in the cost of telepresence and telerobotics triggered by the virtual presence revolution. If I am right, it will be important for governments and businesses to start rethinking globalization.