Data, Trade, and Growth

Cross-border data flows are a major aspect of today’s global economy. Moreover, the volume of cross-border data flows is growing at a rapid rate, with data flowing through terrestrial and submarine cables, and to a much lesser extent via satellites. According to TeleGeography, a consulting firm that keeps track of international data flows, demand for international bandwidth increased at a compound annual rate of 49 percent between 2008 and 2012. Cross-border data flows are also increasingly critical in trade negotiations. Cross-border data flows are one of the main subjects of the new round of trade negotiations announced by the United States Trade Representative, Ambassador Ron Kirk in January 2013. Similarly, the European Union is considering new data privacy regulations that would impact flows of data in and out of the EU. Ironically, statistics about the magnitude of cross-border data flows are scarce, despite their growing economic and political importance. The trade report from the Census Bureau and the Bureau of Economic Analysis contains some information about imports and exports of telecommunications services, but we show that these figures likely miss much of the increase in cross-border data traffic because of fundamental changes in the structure of global networks. Similarly, international agencies such as the ITU only collect fragmentary statistics on cross-border data flows, though they are putting more effort into estimating such figures. Moreover, we may not have the right conceptual framework for thinking about the economic impact of cross-border data flows. The usual categories of exports and imports don’t apply very well to cross-border data flows, since it’s not clear that an outflow of data from a country should count as an export. Indeed, long-established conventions treat outgoing international phone calls as imports, even though both the originating network and the receiving network play an equal role in the call. We will see below the assignment of outgoing international phone calls to the import category is essentially an artifact of regulation. This paper is intended as a preliminary exploration of the empirical, conceptual, and policy aspects of cross-border data flows. There are six sections. • An examination of how cross-border data flows are being reported — and not being reported — in the current trade statistics. • A typology of cross-border data flows. • An analysis of the magnitude of cross-border data flows, based on a comparison all internet/IP traffic. We find that for the United States, cross-border flows are roughly 16-25% of all U.S. data traffic, which is very significant. Cross-border data flows between Europe and the rest of the world equal roughly 13-16% of all European data traffic. These estimates, which should be treated as highly imprecise and tentative, suggest that the United States is more interconnected with the rest of the world than Europe. • An initial analysis of the economic benefits of cross-border data flows. In particular, we develop a framework for measuring the impact of data flows that cross borders without leaving a monetary footprint. • We explore the implications of latency as the equivalent of the “cost of shipping” for data. • Finally, we briefly examine the effects of policy on data, trade, and growth. An early draft of this paper was presented at a non-reviewed conference on “Measuring the Effects of Globalization” (Washington DC, February 28-March 1, 2013).