A Introduction

The founders of Stitch Fix and Strava understood something basic about people. Humans like to use data to connect with other people and to compare with their peers. Based on those insights, these entrepreneurs were able to build two new digital service companies. Both Stitch Fix (a clothing service) and Strava (a social network) rely on personal data to provide services to their customers. Stitch Fix clients first answer a detailed questionnaire about their clothing likes and dislikes. In return, these customers receive clothes and style recommendations designed by stylists and artificial intelligence (AI) to help them look and feel better about themselves.Footnote ¹ Meanwhile, runners, cyclists and triathletes turn to Strava to measure their performance and instantly compare it to others around the world.Footnote ² The two companies could not succeed without the relatively free flow of data across borders. Data flows move across borders when individuals, companies or governments authorize data to be transferred from one country (the source of data) to another country where the data may be processed or used.Footnote ³

Firms have long relied on data to improve the efficiency and quality of goods and services. However, today market actors also utilize data to create entirely new services, such as personalized healthcare, and sectors such as apps, Internet-connected devices (Internet of Things, IoT), cloud service providers and AI. These sectors are the foundation of the data-driven economy: an economy built around the collection, preservation, protection, implementation and understanding of many different types of data. Although no one has exact figures, a significant portion of the data-driven economy is built on personal data – that is, data by and about people or a person.Footnote ⁴

The data-driven economy portends major changes for the ability of individuals to shape their destiny and autonomy. Firms active in the data-driven economy are dependent upon data, much of which is personal data. According to the US National Institute of Standards (NIST), in the past, personal data was something that researchers had to ask for, store, analyze. Because it was not easy to collect personal data, scholars struggled to get sufficient information to do a full analysis. But today almost all our daily activities are data-collection opportunities, thanks to the mobile Internet, the IoT, and other data-driven technologies. Moreover, in the past, people could control their data to some extent because researchers, whether firms or individuals, had to obtain, or at least go through the motions of obtaining, consent. However, with the data-driven economy, people whose data is collected and used have provided their personal data without fully informed consent. To put it differently, despite mechanisms to opt in or out of data collection, people do not understand that in return for providing data that firms then monetize, they receive the many free services presented by digital technologies.Footnote ⁵ In this sense, while the mission of data-driven firms, such as Stitch Fix and Strava may be to help customers, their strategy for so doing may also conflict with long-accepted ideas about autonomy.Footnote ⁶

Compared to Alibaba or Google, Stitch Fix and Strava are small players in the data-driven economy, but they are not atypical. Many of these firms see providing data services as akin to providing a public good. For example, Google’s corporate mission is ‘to organize the world’s information and make it universally accessible and useful’.Footnote ⁷ Not surprisingly, researchers and policymakers now believe that data is the most traded good or service. In 2016, the McKinsey Global Institute asserted that the value of data flows has overtaken the value of global trade in physical goods.Footnote ⁸ According to the World Economic Forum, ‘the world produces 2.5 quintillion bytes a day, and 90 per cent of all data has been produced in just the last two years’.Footnote ⁹

To succeed in the data-driven economy, companies and researchers need access to significant amounts of data – what economists term ‘economies of scale’. Policymakers in many countries want to encourage these scale economies with shared norms and rules, but they also want these norms and rules to explicitly limit trade in some types of data to ensure the safety and privacy of their citizens. In elaborating this rule framework decision-makers must develop a process that reassures their citizens that the rules-based system is transparent, accountable and open to citizens’ input.Footnote ¹⁰ With shared norms and rules, the Internet would be less likely to fragment; more people would have greater access to information; and individuals could create and share more information.Footnote ¹¹ Individuals might also be better able to obtain rents from their personal data and have some modicum of control over its use. However, policymakers around the world disagree on how and where to develop such shared rules.Footnote ¹²

Many executives and policymakers argue that trade agreements are the appropriate venue in which to govern cross-border data flows, because they believe that when information flows across borders, these flows are essentially traded.Footnote ¹³ They have negotiated e-commerce and digital trade chapters for this purpose. Herein, we distinguish between e-commerce (goods and services delivered via the Internet and associated with a transaction) and ‘digital trade’, which includes ‘e-commerce’ as well as new data-based services, such as Stitch Fix, or social platforms, such as Twitter.Footnote ¹⁴

While countries have begun to build a regulatory environment for e-commerce, it is unclear how to build an effective enabling environment for data. Many developing countries are not yet ready for such rule-making. After all, the bulk of firms like Strava and Stitch Fix are being created in middle-income and wealthy countries.Footnote ¹⁵ In many developing countries, business people are hobbled by obstacles such as unstable Internet connections, limited funding, inadequate numbers of researchers, and complementary policies, and infrastructure.Footnote ¹⁶ Moreover, while many countries have open data strategies for government-funded or public data, others have not yet figured out how to ensure that when firms mine the personal data of their users, they protect it from misuse, theft, or human rights violations. Firms that do not adequately protect the data that they collect, monetize, and share could lead users to experience problems such as identity theft, manipulative marketing or discrimination.Footnote ¹⁷ Users deserve a chance to shape new rules and to influence how firms use data.Footnote ¹⁸

This chapter examines the new role of data in trade and explores how trade in data differs from trade in goods and services. Clearly, data is different and may need a distinct set of rules. Although there are six different types of data, we focus on two types: public data and personal data (information that relates to an identified or identifiable individual). We then examine several analogies used by analysts to describe data as an input, which can help us understand how data could be regulated. Next, we discuss how trade policymakers are regulating trade in data and how these efforts have created a regulatory patchwork. Finally, we suggest an alternative approach noting that any agreement must be built by and for the people whose data serve as its foundation. Before trade negotiators try to develop rules regarding cross-border data flows, they must acknowledge the special character of data and focus first on creating an effective enabling environment, then built trust in that new economy by empowering people around the world to control their data.

B The Peculiarities of Data and the Role of Data in Trade

Data and information have long been a key component of trade, but as noted earlier, data has also created new forms of trade. However, cross-border data flows are quite different from trade in goods or other types of services for many reasons: First, many services from payroll to data analytics rely on access to cross-border data flows. These data flows may yield a good or a service, or both.Footnote ¹⁹ Second, trade in digital services differs from trade in other services because suppliers and consumers do not need to be in the same physical location for a transaction to occur. Third, trade in data is fluid and frequent, and location is hard to determine on the borderless network. Trade in the same set of data can occur repeatedly in nanoseconds – for instance, when millions of people download Drake’s latest song. As a result, researchers and policymakers may find it hard to determine what is an import or export. They may also struggle to ascertain when data or data sets are subject to domestic law (such as intellectual property law) and what type of trans-border enforcement is appropriate.Footnote ²⁰ Fourth, when data flows across borders, it may or may not be affiliated with a transaction. Hence, it is hard to describe some of these flows as ‘traded’.Footnote ²¹ Fifth, economists generally agree that many types of data are public goods, which governments should provide and regulate effectively. Furthermore, when states restrict the free flow of data, they reduce access to information, which in turn can diminish economic growth, productivity and innovation domestically and globally.Footnote ²² Such restrictions can also affect the functioning of the Internet.Footnote ²³ Sixth, trade in data occurs on a shared platform (the Internet) that is held in common. Seventh, and as earlier mentioned, much of the data flowing across borders and powering new sectors is personal data – digital data created by and about people. While they may benefit from services built on that data, the people who are the source of it do not control it. Data is their asset, yet they cannot manage, exchange and account for it.Footnote ²⁴

Recent surveys show that people around the world are increasingly concerned about how firms use, protect, control and trade personal data. A 2018 poll of 25,262 Internet users in twenty-five countries found that half of Internet users surveyed are more concerned about their online privacy than they were a year ago, reflecting growing concern around the world about online privacy and the power of social media platforms.Footnote ²⁵ Citizens want their governments to strengthen data protection laws and to beef up enforcement. In 2017, the Australian government stated that ‘governments that ignore potential gains through consumer data rights will make the task of garnering social license needed for other data reforms more difficult’.Footnote ²⁶

In sum, cross-border data flows may not fit the traditional definition of trade. Policymakers should thus at least question whether the traditional model of trade rules needs reforms to reflect the specificities of data.

C New Uses for Data Require New Ways of Thinking about Data

When individuals try to describe how firms are using data to reorder markets, they often compare data to other longstanding inputs to the provision of goods and services. In so doing, they hope to create greater understanding of the import and value of data. As an example, the World Economic Forum describes data as the oxygen of digital life.Footnote ²⁷ In contrast, The Economist describes data as a new type of raw material, such as oil, on par with capital and labour.Footnote ²⁸ However, law professor Lauren Scholz notes that this analogy is not helpful because the supply of oil is limited and only one actor can use a given portion of oil at one time. However, if you have access to data, then you can use it to create information and value.Footnote ²⁹ Other analysts describe data as a form of capital, which can be shared and leveraged within and between organizations.Footnote ³⁰ They note that the big data firms, such as Google, Facebook, Amazon, Uber, Stitch Fix and Strava, commodify and monetize data, creating new revenues and/or functions for the company.Footnote ³¹

Meanwhile, some other scholars posit that we should think about data as labour, as in the early phases of the industrial revolution. We provide our data for free to firms that turn around and monetize this information. But you and I, like the workers of yore, lack bargaining power and are unable to meaningfully negotiate over payments for our data. Most of us are not sufficiently protected from misuse of our personal data or violations of our privacy. In this way, we are denied a share of the economic value of our data, just as workers in the early industrial age. We are facilitating a massive transfer of wealth from ordinary people to the tech titans.Footnote ³² In search of evidence, two scholars traced the AI supply chain and found invisible labour, outsourced or crowdsourced, hidden behind interfaces and camouflaged within algorithmic processes. They note ‘[s]ometimes this labor is entirely unpaid, as in the case of the Google’s reCAPTCHA. In a paradox that many of us have experienced, to prove that you are not an artificial agent, you are forced to train Google’s image recognition AI system for free, by selecting multiple boxes that contain street numbers, or cars, or houses’.Footnote ³³ Moreover, these scholars note that treating data like capital exacerbates inequality and limits the productivity gains from big data and AI. They suggest that we should organize collectively to form a ‘data labor union’ that would bargain for fees for assessing our data. The union could certify data quality and guide ‘users to develop their earning potential’. Meanwhile, data collectors ‘must allow users to understand, withdraw, and transfer their data across competitors’.Footnote ³⁴ Only by organizing collectively can we control how our data are used.

Still other scholars argue that personal data is a form of property that individuals can assert rights to control and to access.Footnote ³⁵ This concept underpins the European Union’s General Data Protection Regulation (GDPR). The notion that data is a form of personal property that people should be able to control also undergirds other countries’ approaches, such as those of Brazil and China.Footnote ³⁶

If we view data as property, then corporations would have to pay the data generators (you and I) for permission, collection and use of data. The big firms would probably not offer services for free if we had to pay. Moreover, firms would then have an incentive to keep data accurate and carefully stored.Footnote ³⁷ But law professor Lisa Austin warns that if you think about data as property, you have to balance the ownership claims of the owners of personal data with those of the firms processing and monitoring that data.Footnote ³⁸ Nor can we ensure that our private information is not misused. As law professor Teresa Scassa has noted, privacy laws are ill fitted to a context in which data is a key economic asset.Footnote ³⁹

Finally, the UK government has introduced the notion that data is similar to infrastructure. In a paper prepared for the National Infrastructure Commission, the authors noted ‘the managed and built environments increasingly depend upon data in real time. New mechanisms for the assembly, management and processing of data provide a new impetus for thinking how the data is best managed so that society can best utilize its resources, solve the most problems and provide the most social good for most people’.Footnote ⁴⁰ In this view, government plays an important role providing and regulating data and promoting its sharing and consumption.Footnote ⁴¹

Except for data as property, these analogies have not significantly influenced national and international regulations. Moreover, these analogies miss an important aspect of the nature of personal data. It is a by-product of our thinking, actions and simply living. It is not one thing, and thus, we should not simply view it as a resource, or as our property, capital, labour, or infrastructure.

There are no reliable statistics about the types, value and amounts of data exchanged across borders and what percentage of cross-border data flow consists of personal data. Both CanadaFootnote ⁴² and in the United States,Footnote ⁴³ are trying to estimate the value of these flows. Despite the lack of exact numbers, we can hypothesize that a significant portion of the data exchanged across borders is personal data. People’s ability to control their data, like other issues of autonomy, is becoming a civil rights issue.Footnote ⁴⁴ According to Ravi Naik, individuals’ rights to data protection ‘have too often been ignored, and it is taking a groundswell of citizen activism to flip the script and hold power to account by individuals asking for their data and determining its use. We are at a watershed moment of a citizen-led demand for data rights, with the hallmarks of a new civil rights movement enmeshed within it’.Footnote ⁴⁵ Some countries, such as Chile, Colombia, Mexico, Turkey and Ecuador, are making personal data protection a constitutional right, although they differ as to the efficacy of enforcement.Footnote ⁴⁶

Truth is, these analogies can only go so far in guiding public policy because the new economy is behaving in ways that few of us understand. For example, the market for data is opaque: we really do not know how firms use our data. In these conditions, data holders/gatherers can deny or grant access to data; they do not have to let people know what data they have collected, whether it is accurate, how they use it and if they sell it.Footnote ⁴⁷ In opaque markets, policymakers should develop policies that facilitate transparency and accountability, as counterweights to opacity. Breznitz argues in this sense that governments must establish the market for data and set the rules for how data are gathered and used.Footnote ⁴⁸ Meanwhile, the Australian Productivity Commission says that governments must move markets from a system based on risk aversion and avoidance (which is not working) to one based on transparency and confidence in data processes.Footnote ⁴⁹

Despite their flaws, two of these analogies may be useful to trade policymakers, as they seek to develop rules governing cross-border exchange of data. First, at the national level, developing country policymakers who see data as a form of basic infrastructure could be more willing to establish data plans. Smart management of all types of data will enable more people to benefit from such data and to create new data-driven services attuned to specific economies and cultures. In contrast, the data as labour analogy might help trade policymakers as they attempt to bridge national strategies and create international rules governing data. In the late nineteenth century, many industrializing states developed national regulations to improve work conditions and protect workers from the vagaries of globalization. These regulations helped raise wages, which in turn led to improvements in labour productivity and greater trade. But not all states adopted such worker protections and trade policymakers feared a race to the bottom among states competing for lower wages and working conditions. The members of the League of Nations established an International Labour Organization (ILO) with rules that would help them find common ground to improve workplace conditions, facilitate peace and encourage trade.Footnote ⁵⁰ We may need a similar organization to help mitigate the differences among national data approaches, if not the WTO.

D The Current State of Rules Governing Cross-Border Data and the Rise of Data Realms

Policymakers have been trying for years to create global rules to govern cross-border data flows both at the World Trade Organization (WTO) and in bilateral and regional trade agreements. The multilateral trade forum of the WTO includes several agreements that address issues affecting data and digital trade. They include the Information Technology Agreement; the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS); and the General Agreement on Trade in Services (GATS). The GATS is the most relevant to the new data-driven services; it has chapters on financial services, telecommunications, computer and media services. But the GATS predates the invention of the Internet and World Wide Web and says nothing explicitly about cross-border data flows. Nonetheless, the WTO panels and the Appellate Body have interpreted the agreement as applying to various online services. While they acknowledge that the agreement is technically neutral – that it was written to apply to changing technologies – academics, business leaders and various governments, including the United States, have argued that the WTO’s rules need both amplification and clarification to apply to new data-driven services, such as those provided by Stitch Fix and Strava.Footnote ⁵¹ Meanwhile, WTO members established a work programme on e-commerce in 1998 and have agreed to waive customs duties on electronic transmissions. They also appear to have made progress on negotiations on data, as a leaked text reveals.Footnote ⁵²

At the Eleventh WTO Ministerial Conference in Buenos Aires in December 2017, Australia, Japan and Singapore, with the support of sixty-seven other WTO members, launched the E-Commerce Joint Statement Initiative. They hoped to encourage a consensus on what members should negotiate and how.Footnote ⁵³ To further that effort, countries issued proposals and background papers. A group of African countries, also supported by India, advocated keeping the discussions within the WTO’s current exploratory work programme, which has conducted work on e-commerce-related topics within various WTO bodies, such as its Council for Trade in Services and Council for Trade in Goods.Footnote ⁵⁴ Overall, not only is there a lack of consensus on e-commerce issues among the members but also it is often apparent that many of the members do not understand the differences, nor do they clearly distinguish between e-commerce and the provision of data-driven services.Footnote ⁵⁵

Despite this, on 25 January 2019, some seventy-six WTO members agreed to commence dedicated e-commerce talks. The announcement of this initiative was not greeted with universal acclaim. While business groups lauded it, civil society organizations and international labour groups came out against the talks and argued that a new agreement could threaten jobs, privacy and data security.Footnote ⁵⁶ The members of the WTO did not only disagree about whether or not these talks should proceed, they also disagreed about the scope of the talks.Footnote ⁵⁷ Many states – including the United States, Canada, China, Japan, the EU, Australia, Brunei, Hong Kong, Kazakhstan, Korea, Mongolia, New Zealand, Singapore, Chinese Taipei, Thailand, Georgia, Iceland, Liechtenstein, Moldova, Montenegro, Norway, Russia, Switzerland, Macedonia and the Ukraine – are keen to move the talks forward. With regard to data flows in particular, while the United States, Canada, the EU and Brazil generally want to create interoperable and universal rules and limit barriers to cross-border data flows, Russia and China are more concerned with maintaining internal social and political stability and are more open to using domestic regulation to limit such flows.Footnote ⁵⁸ Developing countries are also divided. Policymakers and business leaders in most countries acknowledge that traditional e-commerce could help their farmers and firms trade directly with consumers around the world.Footnote ⁵⁹ So, they are willing to negotiate ‘e-commerce’, but many are leery of negotiating data-driven services, given that they may lack domestic data-driven firms.

Meanwhile, the United States, the EU, Australia, Canada and other nations have placed language governing cross-border data flows in e-commerce chapters of their free trade agreements. As the data-driven economy has expanded in importance, the US, Mexico, Canada, the EU and Japan have recently renamed the newer versions of these chapters ‘digital trade’ chapters. Nations are also negotiating and agreed to digital economy specific agreements such as the Digital Economy Agreement of Australia and Singapore, US Japan Digital Economy Agreement, and the Digital Economy Partnership of Chile, New Zealand, and Singapore.Footnote ⁶⁰

The first agreement, the Comprehensive and Progressive Trans-Pacific Partnership (CPTPP) went into effect in 2019 among eleven nations bordering the Pacific including Australia, Japan, Mexico, Chile and Canada. These nations agreed to the free flow of data across borders as a default, with limited exceptions. All signatories also must adopt a minimum level of privacy regulation. In contrast, the EU–Japan Free Trade Agreement (FTA), which also went into effect in 2019, puts personal data protection at its core. The EU–Japan Free Trade Agreement is the first FTA of the EU that includes rules on data but it also ensures that personal data is adequately protected not only under the agreement but additionally through an adequacy decision of the European Commission – the first such decision under the GDPR heightened standards of data protection.Footnote ⁶¹

The US government next used CPTPP, whose e-commerce chapter is identical to that negotiated under the Transpacific Partnership Agreement (TPP) as a building block for the renegotiation of the North American Free Trade Agreement (NAFTA). NAFTA 2.0, now called the United States–Mexico–Canada Agreement (USMCA), has several interesting elements designed to promote data-driven economic growth. It seems designed to promote AI and other data-driven services. First, the USMCA contains a proper chapter on ‘digital trade’ (chapter 19), rather than one on e-commerce. Secondly, like CPTPP, it bans mandated disclosure of source code. But differently from the CPTPP, it also promotes AI by encouraging the parties to provide public information (information developed or provided to public entities) in a machine-readable and open format that can be ‘searched retrieved, used, reused, and redistributed’.Footnote ⁶²

While the United States and Canada have made regulating barriers to cross-border data flows a priority, the EU has made personal data protection a priority. The EU will only sign FTAs that contain language regarding the free flow of data if its FTA partner(s) adequately protect personal data. These nations must go through a process of becoming ‘adequate’. Specifically, these states must create independent government data protection agencies, register databases with those agencies and, in some instances, obtain prior approval from the European Commission before personal data processing may begin.Footnote ⁶³ This process is both time-consuming and expensive, as the EU’s digital trade partners must devote resources to data protection, a difficult choice for nations with limited governance expertise or funds.

Meanwhile, policymakers in China restrict the free flow of data and information not only across borders but also within China. In so doing, Chinese officials maintain social stability and the power of the Communist Party.Footnote ⁶⁴ However, China participated in the negotiation of Regional Comprehensive Economic Partnership (RCEP), a mega-regional trade agreement. RCEP includes Australia, Indian, Japan, South Korea and New Zealand as well as the nations of the Association of Southeast Asian Nations (ASEAN).Footnote ⁶⁵ The RCEP' allows member states to impose whatever national regulatory restrictions they wish, as long as they are applied in a non-discriminatory way (are applied equally to domestic and foreign businesses).The provisions are not disputable.Footnote ⁶⁶

Thus, the three big digital markets – the United States, EU and China – have taken different approaches to cross-border data flows. This patchwork approach is causing another problem for many nations. Nations, such as Canada, Mexico and Australia, that have or seek to build strong trade relationships with the big three must choose which approach they would follow.Footnote ⁶⁷ Countries that choose more than one such market will face high regulatory costs, as their costs of compliance would rise, given different standards.Footnote ⁶⁸

In a recent scholarly study, the WTO secretariat confirmed this patchwork of rules. It examined regional trade agreements that have incorporated specific provisions related to e-commerce. They found significant heterogeneity among the seventy-five chapters that explicitly address e-commerce. For example, these FTAs have different objectives, scope and definitions. The FTAs also define and limit different barriers to trade, and most importantly, some thirty-eight of the seventy-five have different provisions related to the domestic legal framework in which e-commerce takes place. Finally, some forty-four of the seventy-five include language on personal data protection but again with very different definitions and obligations.Footnote ⁶⁹

Developing countries are likely to have the most problems adapting to the data-driven economy. These countries will be customers of AI and other data-driven sectors, rather than producers. According to Kai-Fu Lee, a venture capitalist and former computer scientist, the bulk of profit from the data-driven economy and particularly AI will go to the United States and China: ‘AI is an industry in which strength begets strength: The more data you have, the better your product; the better your product, the more data you can collect; the more data you can collect, the more talent you can attract; the more talent you can attract, the better your product. It’s a virtuous circle, and the USA and China have already amassed the talent, market share and data to set it in motion’.Footnote ⁷⁰

Finally, many developing countries have not yet adopted effective rules protecting personal data online or established rules for the use of public data. Based on data from 2017 the UNCTAD reports that 57 per cent of all countries (some 107 countries of which 66 were developing or transition economies) have put in place legislation to secure the protection of data and privacy. In this area, Asia and Africa show a similar level of adoption, with less than 40 per cent of countries having a law in place. Some 21 per cent of countries have no law at all; and 10 per cent are in the process of drafting legislation.Footnote ⁷¹ Moreover, most of adopted legislation contains rules that are not consistent with either the OECD Guidelines for the Protection of Personal Information and Transborder Data FlowsFootnote ⁷² or EU’s GDPR.Footnote ⁷³

Moreover, some countries hoard and refuse to share publicly held data with their citizenry.Footnote ⁷⁴ In general, data gains value as it is shared, but it has little value if governments hoard it. While there is little empirical proof, open data appears to have important spillover effects including increasing civil discourse, improved public welfare and a more efficient use of public resources. But many states lack right to information laws or do not allow their citizens to view or comment on the data they hold.Footnote ⁷⁵ So not only is there a patchwork for FTAs but there is also a patchwork of approaches to governing various types of data as well.

Without sufficient understanding and interaction with data-driven firms and their customers, developing country policymakers may struggle to effectively advocate for their short- and long-term interests in the data-driven economy. Zimbabwe provides an example: the government signed a strategic cooperation framework agreement with a Chinese start-up, CloudWalk Technology, for a large-scale facial recognition programme. Zimbabwe will export a database of their citizens’ faces to China, allowing CloudWalk to improve their underlying algorithms with more data. The government allegedly agreed to the system to improve public safety, while the company wanted to improve the accuracy of its facial recognition system which was based on Chinese faces and needed a wider range of facial types. However, the government of Zimbabwe could use this system to more closely monitor its citizens, which could undermine social stability and trust.Footnote ⁷⁶ While such a situation may be rare, it provides a strong rationale for Zimbabwe and other countries to develop and debate a strategy for protecting personal data.

E A Path Forward

Humans have long exchanged data between borders, but never have they traded so much data or benefited from so many new services built on data. These new services may make us smarter, richer, more flexible and more efficient. But not all countries or people are ready to participate in this brave new world. The OECD recently noted that ‘governments and stakeholders have a responsibility to shape a common digital future that improves peoples’ lives and boost economic growth for countries at all levels of development, while ensuring that nobody is left behind’.Footnote ⁷⁷ However, for governance to succeed and be trusted, it needs to be built on shared norms and rules.

Policymakers should first work at the national level to develop a national strategy for data and then move towards interoperability of approaches rather than harmonization. They must find a way to conduct discussions on data governance that build public trust, consistent with the multi-stakeholder processes embedded in other forms of Internet governance. Against this backdrop, this chapter suggests five steps for moving forward, summarized below.

Step 3: Clarify the Rules and Exceptions to the Rules, So Nations Do Not Restrict Cross-Border Data Flows More Frequently or Broadly than Necessary

Like other treaties, a data-driven economy agreement should include exceptions to the rules. Nations can use the exceptions to ‘excuse’ violations to the agreement when they pursue other important policy objectives. (Figure 16.1 shows that governments have a wide range of reasons to restrict cross-border data flows.) Countries can only use these exceptions, however, if they do so in the least trade distorting manner. Yet, so far, there is no clear model that policymakers can follow to distinguish between legitimate and trade-distorting data flow regulation. The current language in trade agreements is vague and states must rely on trade disputes to develop clarity and some degree of legal certainty. However, there have been few disputes to provide guidance and policymakers have not yet agreed on updating the WTO law language with regard to the general exception clauses or other specific exceptions.

Figure 16.1. Why and how do governments restrict cross-border information flows? Prepared by Caitlyn Leong

Policymakers should begin by delineating how and when nations can use the exceptions to limit cross-border flows in the name of protecting domestic security or cybersecurity. For example, some governments, such as India, Brazil, the United States and the UK, have called on companies to provide backdoors to encrypted communications to help law enforcement. However, such an encryption backdoor would undermine trust and the effectiveness of encryption as a tool for keeping individuals, firms and governments safe online.

F Conclusion

The world is awash with data and there is no consensus on how to regulate it. The five outlined steps can help nations prepare for future negotiations and build value from data. These ideas will not address all the issues that arise in regulating cross-border data flows, and any new approach is likely to face many challenges, especially from those vested in the existing organizations and approaches to governing data. But clearly, we are stuck in a rut on trade and must creatively address the trade and non-trade dimensions of cross-border data flows. Policymakers from a wide range of countries may be more willing to compromise if they see that their citizens will benefit from clear, interoperable rules and from receiving funds for their data. Moreover, this approach could help firms accommodate national differences regarding ethics of data usage, disinformation and other upcoming regulatory issues. It could also give developing countries greater leverage in the discussions on data flows, where they would ordinarily be ‘rule takers’.Footnote ⁹⁵ Finally, these ideas could help more countries better integrate data-driven firms and their traditional firms. By collaborating and rethinking the process of global rule-making on data, we will be better able to achieve the change we wish to see in the world – where people have greater autonomy and control over their data and data drives equitable growth.