TPRC46: Research Conference on Communications, Information and Internet Policy

Abstract
Digital trade and protectionism have become major topics in communications policy. But how much do we really know about digital trade?

Essentially, standard trade paradigms don’t apply to the digital economy very well; they track the monetary value of bilateral transactions between a buyer in one country and a seller in another. Transactions only get counted when they pass a customs barrier, which doesn’t happen on the Internet. Many digital services are “free.” Trade statistics do not capture the way data flows and digital platforms support other forms of trade, both digital and non-digital. Further, statistics regarding the services of digital platforms are scattered across a set of categories that were created for pre-Internet industrial forms. These categories sometimes conflate digital and non-digital forms of trade, or arbitrarily splinter digital forms of exchange into multiple categories.

This paper takes a fresh look at digital trade by studying data flows and their relationship to other mobile factors of production, such as capital and goods. We use Telegeography statistics regarding traffic of the top 50 websites in a country. The Telegeography data tracks the “balance” of digital traffic, because it shows how much of a country’s web site traffic is requesting sites in another country. We will correlate these digital balances of trade with payment statistics from information services-related trade and foreign direct investment between countries. We will show how agglomeration economies and access to large markets play a huge role in shaping the balance of digital data.

In classic Heckscher-Olin trade theory, factor mobility is a substitute for trading goods across borders. If factors such as capital and labor can move freely and at low cost, they will disperse geographically to minimize transportation costs. Factor mobility also plays an important role in the reaction of firms to trade barriers. National policies that impose high tariffs while permitting capital mobility, for example, might induce FDI to build production facilities in the protected country, so that foreign firms can avoid tariffs but continue selling into that market. The new trade theory, in contrast, emphasizes the “centripetal” pull of scale economies, agglomeration economies, and access to large markets. These centripetal forces promote concentration of economic activity in a single provider that interacts with the rest of the world through trade rather than FDI and local production. (Krugman, 1993) The newer trade theory also observes that as countries become more similar in size, relative factor endowments and technical efficiency, international economic activity will become dominated by multinational firms, which are “exporters of the services of firm-specific assets.” (Markusen and Venables, 1995).

In both the old and the new theories, any attempt to understand trade patterns and assess the impact of national trade policies would require an analysis of the interdependent flow of all relevant factors of production. Yet too many approaches to digital trade only attempt to measure cross-border payments. Also, standard trade theory only treats labor and capital as the mobile factors of production. While the new trade theory sometimes engages with “knowledge capital” and the advantages of firm internalization, data flows are not typically treated as a factor of production. Data is also a mobile factor of production. This is the basic insight that underlies our research.