International trade in goods and services that use lots of data has the potential to boost economic growth across the world, but sharing data across borders is sensitive to laws, culture, geographic distance, digital connectivity and a range of institutional questions. Research on how these issues should be considered by policy-makers as they create trade-boosting policies is nascent, and here we relay notable literature and findings as we contribute to the international policy discussion.
International flows of data have expanded rapidly in recent years, with cross-border bandwidth growing 45 times between 2006 and 2016, and as much as 2 billion people now using cloud computing to store digital information, according to Cisco. This is the result of cross-border use of goods and services across many sectors, from the 15% of total downstream internet traffic accounted for by Netflix, as estimated by Sandvine, to the digital monitoring of jet engines around the world by companies such as Rolls Royce. This puts data at the centre of how countries trade and relate with each other across the world, but we don’t know as much about how policy affects data flows in economic exchange between countries as you might expect.
The Web changed international trade
The Web was developed with much more than international trade in mind, but has a strong bias towards unencumbered economic exchange. When ODI co-founder Sir Tim Berners-Lee wrote his proposal for the Web, he rejected systems for organising the world’s electronic information in one, grand, compartmentalised way, and instead believed that some simple bits of technology – identifiers and hypertext – could link pages of information into a network that would leave internet users to innovate in ways that they wanted, while also being able to connect to others. It’s that thought that spread the internet around the world, invited unimagined innovation, and connected billions of people in ways that they weren’t before.
Over the 30 years since the Web was invented and the passing of some of the more idealistic visions of the changes it would bring, many countries have found it necessary to add friction into data exchange. Governments around the world have been updating their data protection laws in order to protect their citizens from harm, and the European Union’s General Data Protection Regulation (GDPR), and the Cross-Border Privacy Rules adopted by the Asia-Pacific Economic Cooperation forum, are just two recent answers to how to coordinate international standards. Some of this may have been driven by the large body of support in many countries – above 80 percent in Indonesia, China, India, and Mexico, according to a survey by the Centre for International Governance Innovation – for data about citizens to be stored domestically. Whatever the reason, there have also been reactions by countries such as China, which has long had its Great Firewall, while the Russian Duma rubber-stamped a bill earlier this year that allows the government to cut off internet traffic in the country from foreign servers.
Data infrastructure as trade competitiveness
Our primer on data-enabled international trade looks at the cross-border data flows from the perspective of trade competitiveness: the ability of a country to create goods and services domestically with data, and sell them to foreign consumers. We suggest that the quality of national data infrastructure will affect how much countries will be able to compete in the industries of the future, such as artificial intelligence, and the extent to which they could collaborate with others around the world and sell services that often include sensitive data about users. This raises numerous policy questions about how to develop data infrastructure and the international means of exchange.
Analysing trade competitiveness means looking at many of the ‘behind the border’ conditions in a country, such as the degree of competition and availability of skills, but there are two questions for cross-border data exchange that involve looking to other countries: one, market access and the ease with which companies in one country can sell to consumers in another; two, factor conditions and the costs faced by domestic firms in getting access to the resources they need for production. The former is a classic question of trade agreements and tariffs, while the latter is particularly acute in data exchange questions because the nature of the Web means that flows of data – the resource that enables digital services – should face much less friction than the exchange of more traditional raw materials, like oil or timber. This leaves us with a broad research question that we are considering in our work: what makes it easier for firms to export data-enabled products from one country to another?
It’s all about the institutions
The simple answer to this is ‘trust and institutions.’ The economist, Douglass North, won the Nobel Prize for his work on explaining the importance of institutions – norms, rules, and public bodies – to economic development, and the extent to which they reduce transaction costs, allow more exchange, and widen the range of things that a society can produce. But that’s easier said than done in one country, let alone a group of them like those in the European Union, or between nations on different sides of the world. According to Apolitical, there is a lot of detail in how we might develop institutions to facilitate data-enabled trade, and thankfully there is some research – discussed below – on what affects international trade in digital services that we can use to think about the policy landscape for data questions.
World Bank research has shown that trade and services are heavily dependent on institutions, and that matters because services are becoming more tradeable and the complexity that countries can achieve in their services exports is correlated with how much they grow. This supports the argument of the Atlas for Economic Complexity with regard to goods, which says that countries are likely to become richer as they are able to produce more complicated things. Economic strategy questions then arise for developing countries, who might traditionally have favoured growth through manufacturing, but the International Monetary Fund’s research on India suggests that reform to services can raise domestic productivity and exports.
The Peterson Institute for International Economics has suggested that trade in services could be as high as 20% of international transactions, and that’s despite it traditionally being thought of as a harder type of exchange than that for physical goods. In both instances, international buyers and sellers often have to engage in a process of searching that is more like a network than a market – according to economist James E Rauch – meaning that other than for simple commodities like wheat or timber there is a necessary process of increasing mutual understanding that develops social capital and facilitates trade. And the more complex the thing being exchanged becomes – such as a digital service that uses data about people – the harder it is to build that capital. This in turn leads to the likelihood of trade being boosted by mutual language – countries with fewer speakers of foreign languages face higher hurdles to trading internationally, although the use of machine translation has been shown to increase exports by as much as 17.5% in some circumstances – and geographic proximity, as academic analysis has shown.
Getting close, digitally
Research has revealed that some countries that are far away from each have a high degree of ‘virtual proximity’ – the number of bilateral links between web pages in one country and another – that helps them overcome geographic constraints, perhaps because they share similar tastes and language connections. The spread of the internet has increased trade, as has the domestic use of it in countries, with notable benefits for poor countries. But internet use in some countries actually seems to raise the amount of consumption from domestic providers – according to research for the European Commission – instead of that from foreign suppliers, although consumers also appear to search further afield for unusual items. This might be because local providers are better able to satisfy local tastes and internet use just raises demand and competition. This is supported by the findings of academics at the University of Toronto that music and games are much less likely to be consumed further away than websites for software and financial products.
The performance of countries in international services trade differs considerably. In countries such as France, Germany, Italy, Portugal, Spain and the UK, domestic providers of online digital services dominate, while the United States has been able to establish leading positions in more foreign markets than any other. That might be the result of domestic conditions in the US producing more competitive products or familiarity among foreign consumers of US products. But if we turn the question around and ask about the distance of foreign suppliers from US consumers, we see that even a small increase in physical distance from the US means that a foreign website for services that involves a financial transaction sees a 2.7% fall in its likelihood of being visited by an American consumer.
There’s lots of research and policy work to do
Given these factors – distance, digital connectivity, language – what can policymakers do to edge towards an international policy regime that creates trust through institutions and builds the foundations for a future of dynamic, data-enabled trade? Research into the comparative effects of data privacy regimes such as GDPR and the APEC rules is nascent. Work by academics Avi Goldfarb and Catherine Tucker has shown that privacy regulations in the European Union have hindered the effectiveness of online advertising, but that the effects of such policies on technology use and innovation can vary considerably across sectors, suggesting that trade-offs between privacy preferences and economic aims are hard to satisfy with uniform policy.
On the other hand, the Netherlands Bureau for Economic Policy Analysis has shown that regulatory differences raise the legal and compliance costs for exporting firms, especially services ones, and have been shown to reduce the number of companies that will export from one country to others while also increasing the size of the ones that do. Large technology firms such as Facebook and Google are often thought to have reached their size because of network effects and other explanations from the economics of information, but their size may also be explained by the fact that they have been better at navigating regulatory and cultural demands across many jurisdictions at once. In its recent review of the possible effects of the US-Mexico-Canada Trade Agreement, the United States International Trade Commission argued that significant benefits would flow from a reduction in uncertainty over data policy in the signatory countries, although this is debatable.
Untangling all of this is going to take years of research. The European Centre for International Political Economy has been pioneering in the field, ranking 64 countries on the Digital Trade Restrictiveness Index, discussing how restrictions on data exchange and other aspects of digital exchange restrict innovation and growth, and showing that more restrictive data policies reduce firm productivity in data-intensive sectors. They found similar results for the effects of restricted data flows on the import of data-intensive services. But much more needs to be done, not least on understanding which questions of international data exchange are best answered with economics, and which speak to other values which are more appropriately considered by ethicists and others. And this is on top of the deep problem of how to measure the flow and value of data – the OECD’s Working Party on International Trade in Goods and Services Statistics recently said that its efforts in this area were ‘best […] described as embryonic’.
We are going to build on our report into the links between data infrastructure and trade competitiveness by developing a methodology for how countries can be assessed for the ability to create and export data intensive goods and services. And as we do this we’ll need to take into account all of the issues above, but it will be only one part of research and policy literature that needs to be written in the years to come.