Pundits regularly invoke the notion of a world economy that is either “shrinking” or becoming “flat.” Explanations of this alleged flattening include technological innovations in transportation and communication that have enabled goods and ideas to flow more freely. The offshoring of service jobs, particularly call centers and computer software in India, has grabbed recent media attention. In his bestseller The World is Flat, New York Times columnist Thomas Friedman (2005) wrote of how he had “interviewed Indian entrepreneurs who wanted to prepare my taxes from Bangalore, read my X-rays from Bangalore, trace my lost luggage from Bangalore and write my new software from Bangalore.”
Most economists, cognizant of the gains from trade, do not view a “flat” world as an alarming prospect. As the 2004 Economic Report of the President remarked sanguinely, “When a good or service is produced more cheaply abroad, it makes more sense to import it than to make or provide it domestically” (p. 229). In a press conference after the release of the report, the Chair of the Council of Economic Advisors at that time, Gregory Mankiw, elaborated on the remarks in the report saying, “Whether things of value, whether imports from abroad, come over the Internet or come on ships, the basic economic forces are the same.” As Mankiw and Swagel (2006) describe in their insideraccount, these seemingly innocuous remarks about outsourcing managed to arouse a controversy. This was partly because of election-year sentiments but also because the threat of service offshoring raises serious concerns for many onlookers. How will we maintain our standard of living, people wonder, when we have to compete with highly skilled foreigners who are willing to do our jobs for a fraction of our wages?
Just as mainstream trade theory identifies gains from trade, it also shows that real wages of some workers tend to fall as a consequence of freer trade. Mankiw and Swagel (2006) respond to these concerns by arguing that the accumulated statistical evidence on the offshoring of services demonstrates that the magnitudes remain quite small compared to the size of the labour market. This case for complacency recalls a debate between Leamer and Krugman over whether rising imports from low-wage manufacturers were responsible for rising wage inequality in the US. Leamer (2000) pointed out that prices are determined on the margin and the volume of trade is irrelevant for wage determination. Krugman (2000) argued that, on the contrary, trade volumes were crucial evidence on the changes in factor prices that can be attributed to trade. However, if recent growth of service imports continues unabated, the current trickle of offshoring could turn into a flood. Mankiw and Swagel’s argument would be more persuasive if there were strong reasons to believe that economic impediments to offshoring will curb its future growth.
Our research investigates whether geographic separation limits offshoring trade, thereby shielding domestic workers from direct competition with their foreign counterparts. We develop a model that envisions employers searching globally for the most suitable workers for any given task and posits that distance raises the costs of using foreign workers. These higher costs reflect travel, training, or translation time associated with using workers that reside far from where their services will be consumed. Firms choose workers that offer the lowest costs after adjusting wages for productivity and distance-based service delivery costs.
We use data on bilateral trade in services for a large sample of countries to infer the extent to which distance increases the relative costs of using remote workers. We focus on the service category called Other Commercial Services (OCS) that includes service categories that are subject to the offshoring debate—professional services and call centers. Distance effects are conventionally estimated by applying the “gravity equation” to bilateral trade. This method, which has been widely used to study goods trade, posits that exports from an origin country to a destination country are proportional to the economic sizes of each partner country and inversely related to the distance between them.
The scatter plot in Figure 1 of British imports of OCS (averaged over 2000–2004) provides intuition on how the gravity equation is used to estimate distance effects. The vertical axis shows British imports as a share of origin-country GDP and the horizontal axis represents the distance from Britain. Origin countries are identified with standard codes and the data is plotted on a log scale. A symbol system (specified in the legend) indicates whether the origin country has linkages with Britain in terms of a colonial relationship, a common language, or a shared legal system. The figure shows a strong negative relationship between trade and distance, something that we would not observe if services could be delivered costlessly across space.
Figure 1 shows that the univariate regression line through this data has an absolute slope—what we call the “distance effect”—of 0.64. However, the figure also makes it clear that distance is not the only bilateral variable that influences trade. When the influence of colonial ties is taken into account, for example, the slope with respect to distance increases to 0.87. Our regressions control for all the linkages shown in the figure. They also control for time zone differences that are thought by many to be important considerations for offshoring. In keeping with recent theoretical developments and our own model, we include importer and exporter indicator variables.
Figure 1: The impact of distance on British imports of Other Commercial Services (OCS), 2000–2004 averages
Figure 2 shows annual distance effects for OCS and goods estimated for the last decade of available data. Distance effects for OCS are initially higher than those for goods (for the same set of country pairs) but become lower in later years. In 2004, distance effects for OCS are 1.24, implying that a 10% increase in distance reduces bilateral trade by 12.4%.
Using theory and estimated distance effects, we are able to measure the extent to which geographic separation insulates local workers from foreign competition. The calculations reveal that, from the point of view of a London service purchaser, workers in Oxford can be paid 99% to 373% more than workers in Bangalore in productivity-adjusted wages and yet still be more attractive, once service-delivery costs are taken into account. This is because the Bangalore workers are 100 times more distant from London than the Oxford workers.
Figure 2: Estimated distance effects for Other Commercial Services (OCS) and goods, 1995–2004
Those results indicate that geographic barriers offer high-wage workers substantial insulation from low-wage competitors based in remote countries. Distance has long acted as a serious impediment to international transactions. Unfortunately, most of what we know about the effects of distance on international transactions is based on studies of trade in goods. A consensus appears to be forming that freight costs cannot explain the strength and functional form of the distance effect for goods.1 Instead, physical distance seems to be picking up some combination of the barriers imposed by cultural differences, the continued desire for face-to-face communication, and the geographically-biased structure of social and business networks. These factors apply to services as well as goods. Thus, there are two senses in which Mankiw is right to say it does not matter whether imports arrive by ship or by broadband. First, there will be potential gains from trade in either case. Second, there will be distance costs that limit the realisation of those gains.
How much should high-wage workers fear competition from much lower paid workers in India and China? Our findings suggest that distance still provides signification protection. Since these estimates reflect averages across a range of services, there are surely services where competition is especially acute. Moreover, service delivery costs associated with distance appear to have fallen over the last decade to a level that is slightly below the level estimated for goods. Unfortunately, the data do not clearly indicate whether distance costs for services will continue their downward trend or level off. We suspect that persistent cultural differences, as well as locally-biased social networks, will maintain distance costs at a high enough level to forestall the small, flat world envisioned by some journalistic accounts.
This commentary is based on “How Remote is the Offshoring Threat?” by the same authors available as CEPR Discussion Paper No. 6542 at http://www.cepr.org/pubs/ new-dps/dplist.asp?dpno=6542
Grossman, Gene M., 1998, Comment, in Frankel J.A. (ed), The Regionalization of the World Economy, NBER Project Report, The University of Chicago Press.
Leamer, Edward E., 2000, “What’s the use of factor contents?,” Journal of International Economics 50(1), 17–49.
Krugman, Paul R., 2000, “Technology, trade and factor prices,” Journal of International Economics 50(1), 51–71.
Mankiw, N. Gregory and Phillip Swagel, 2006, “The politics and economics of offshore outsourcing,” Journal of Monetary Economics 53(5), 1027–1056.
1 For an exposition of the reasoning, see Grossman (1998).