Newsletter

MEI Newsletter, No. 3 [PDF]

2023.2.10

Global Distribution of Economic Values of Cross-Border Data Flows

by Wendy C. Y. Li

ChatGPT has disrupted the world and started changing how we innovate and produce. It is pushing firms, especially Big Tech, to reallocate resources to cope with this new technological disruption. As the new generative AI technology is incorporated into firms’ operations, its impacts on firms’ innovation and productivity can be significant. However, the capability of AI technology like ChatGPT hinges on big data – AI is a tool of finding patterns in data, and data determines the performance of AI.

China and the U.S. lead in AI technology, and they have 70 percent of top AI researchers and 94 percent of AI start-up funding in the world during 2016-2020 [1]. While having a strict data policy, China outpaced the U.S. in international data flows in 2014, and the gap has been widening since then. China’s international data flow has the world’s highest economic value (US $201 billion in 2017) and enjoyed a double-digit growth during 2012-2017. In comparison, the economic value of U.S.’s international data flow is less than one half of China’s and has been declining [2]. China’s share of journal citations in AI surpassed that of the U.S. and topped the world in 2020 [3]. Although the U.S. is still leading in basic AI research, China is likely to edge ahead in the AI implementation, especially given the vast amount of data available in China [4].

Data is not only a strategic asset, but its cross-border flows are important for firms’ daily operations and for the development of developing countries. Big Tech and online platform companies, typically multinational firms, can concentrate their operations in one or a handful of countries while providing digital services worldwide, and their daily operations inevitably involve tremendous cross-border data flows. Their data-driven business models allow them to serve the global market without facing traditional physical constraints, to easily scale up their businesses, and to enter adjacent industries and markets, as long as they can access the data needed. Moreover, firms in developing countries can also operate domestically and access to overseas markets through online platforms. All of these activities rely on cross-border data flows.

Countries around the world have been negotiating agreements on data flows, but the economic values associated with cross-border data flows have been an elusive topic, let alone how they change with the fast data growth. Understanding that data is a strategic asset but not knowing how much it is worth, policymakers may simply control or block the data flows, a situation that can result in rising trade costs, restricting data sharing, limiting the potential of data, and hindering productivity growth. Examples of countries that have taken or are considering stricter data regulations such as data localization policies to restrict data flows across borders include Bangladesh, Indonesia, Pakistan, and Vietnam.

It was unclear how to measure the economic values of cross-border data flows until very recently. Through our series of studies on the value of data and data flows [5,6,7,2], we develop a methodology to measure the economic values of data and cross-border data flows around the world. To our knowledge, this is the only method that can measure the value of cross-border data flows. Using this method, we find that the economic value of cross-region data flows in 2020 is several hundred billion dollars, and the estimated global value of data is several trillion dollars [7]. The economic values of data flows, however, are distributed very unevenly. The leading ten countries control over two thirds of international data flows, and that top four countries, namely China (including Hong Kong), the U.S., the U.K., and Taiwan, control over 50%.

How the distribution of international data flows has evolved in the last decade not only reveals the areas of rising digital economy but also hints at the impact of national or regional data policies. With the exception that the data flow between Latin America and US & Canada has grown significantly, most new development in the data infrastructure has been taking place in Europe, Asia, and Middle East. My calculation shows that the international data flows for the countries in these regions can translate to hundreds of billions of U.S. dollars in value. In contrast, the data flow between Europe and US & Canada, once the highest cross-region Internet traffic, has been on the decline. It is likely that Europe’s General Data Protection Regulation (GDPR) played a role in the shift of cross-region data flows in recent years.

The global distribution of the values of cross-border data flows is an important topic for global data governance, trade, investment, development, and tax policies. Given the fact that data can increase the firms’ innovations, productivity, and profitability, controlling data flows can greatly affect the distribution of the benefits of data around the world. Understanding that data are a national strategic asset, a growing number of countries have implemented or are considering more restrictive data policies involving storage requirements and flow prohibitions [8]. In fact, this is likely a wrong direction for small or developing countries. Because international data flows only occupy less than 10% of the global data traffic, most data flows are confined in the borders of largest countries. Smaller countries can access large amounts of data only through international data flows.

Taiwan’s data policy provides an interesting contrast. In the last decade, Google built data centers and helped train local AI engineers in Taiwan, and Taiwan shares its data with Google in return. As a result, Taiwan experienced the world’s highest growth in the value of international data flows during 2012-2017, when the average annual growth rate was 154%. The extremely fast growth elevated the amount of Taiwan’s international data flows to the fourth place in the world. This example demonstrates how smaller countries can adopt an open data policy and partner with Big Tech to increase the growth rates of international data flows, consequently their economic values and investments. Big Tech may earn a significant share of the increased value of data in this public-private partnership, but countries with smaller populations may still benefit by adopting a more open data policy to attract investments, incubate local skilled labors, and enjoy the access of data flows.

As generative AI technology disrupts the world, stricter restrictions in cross-border data flows can greatly limit small and developing countries’ innovation and development opportunities, especially when the new technology heavily relies on big data and costly cloud computing infrastructure.

[1]  United Nations Conference on Trade and Development. 2021. Digital Economy Report 2021 – Cross-border Data Flows and Development: From Whom The Data Flow, September.

[2]  Li, W.C.Y. 2023. Global Distribution of Economic Values of Cross-Border Data Flows, Moon Economics Institute Discussion Paper, No. 3, January.

[3]  Stanford Human-Centered Artificial Intelligence, 2021. 2021 AI Index Report.  https://aiindex.stanford.edu/ai-index-report-2021/

[4]  Smith, C.S. 2023. China’s AI Implementation Is Edging Ahead of the US, Forbes, January 14.  https://www.forbes.com/sites/craigsmith/2023/01/14/chinas-ai-implementation-is-edging-ahead-of-the-us/?sh=321b522f2dfb

[5]  Li, W.C.Y., Nirei, M., & Yamana, K. 2019. Value of Data: There’s No Such Thing as a Free Lunch in the Digital Economy, RIETI discussion paper 19-E-022.

[6]  Li, W.C.Y. & Chi, P.J. 2021. Online Platforms’ Creative “Disruption” in Organizational Capital – The Accumulated Information of the Firm, Moon Economics Institute Discussion Paper, No. 1, August.

[7]  Li, W.C.Y. 2022. Economic Values of Data and Data Flows, and Global Minimum Tax, Moon Economics Institute Discussion Paper, No. 2, January.

[8]  Organization for Economic Co-operation and Development. 2022. Cross-Border Data Flows: Taking Stock of Key Policies and Initiatives, October.

Copyright: Wendy C. Y. Li, 2023.

MEI Newsletter, No. 2 [PDF]

2021.12.31

Economic Values of Data and Data Flows, and Global Minimum Tax

by Wendy C. Y. Li

Big Tech and online platform companies can operate mainly in one or a handful of countries but provide digital services worldwide. Under the current production-based tax system, they are known and often criticized for paying little or zero tax in many countries. To ensure those “digital firms” pay their fair shares in the countries where they earn revenues, 136 countries have recently agreed to impose a 15% global minimum tax on any multinational firm with an annual revenue more than 20 billion euros and a profit margin more than 10%. However, as commented by Janet Yellen, the new agreement on the global minimum tax may not be applicable to one of the best-known Big Tech, Amazon, because its reported profit rate in 2020 is only 6.3%.

Big Tech have been reshaping how we live and how firms produce and the pace has been accelerating during the Covid pandemic when many more activities are moved to online. The core of those companies relies on artificial intelligence (AI) algorithms and data. As AI algorithms become more affordable and adaptable, data become the key that determines the accuracy of the algorithms. In other words, data is the heart of Big Tech’s competitiveness in the digital era. As a firm’s key input, data are crucial for firms to innovate, produce and compete in the digital era. Because the network effect of online platforms and the data network effect, Big Tech has been growing rapidly and can internalize the benefits of the externalities derived from data. Their data-driven business model allows them to serve the global market without facing traditional physical constraints, to easily scale up their businesses, and to enter adjacent industries, as long as they can access the data needed. For example, Booking.com serves 137,791 destinations, with 28.9 million properties in 229 countries [1]. However, most of its operations are conducted in its Amsterdam headquarters. Another example is Oracle’s recent purchase of Cerner, a large electronic health record vendor, by US $28.3 billion in cash to enter the healthcare service sector.

When Big Tech and online platforms concentrate their operations in one or few countries, their daily operations inevitably involve tremendous cross-border data flows. This is one of the main reasons why many countries with net data export assert that Big Tech should pay for the benefits deriving from the data collected in their countries [2]. Data flows are important to digital trade, and therefore estimating the economic value of cross-border data flows is critical in providing useful information for global policymakers to devise and implement the new global minimum tax, a new tax system that will undoubtedly affect trade between countries.

With the law of Big Tech’s value of data [3] and our recent advances in understanding data valuation (see Moon Economics Institute Newsletter No.1, August 2021), we can now estimate the value associated with cross-border data flows and how it distributes around the world, a topic of direct relevance to data governance, trade, and tax policies. The published data for global Internet traffic have shown that cross-region data flows increased substantially from 2015 to 2020 (UNCTAD, 2021 and the references therein). Our calculations indicate that these cross-region data flows can translate to hundreds of billion U.S. dollars of value of data per year (see the figure below).

This economic value of cross-border data flow based on the combined values for two-way internet traffic effectively is data trade. It is worth noting that the magnitude of this data trade today is nontrivial when comparing it with the size of traditional trade goods and services (for example, the traditional trade goods and services between the US and Europe were worth $1.1 trillion in 2019 according the Office of the U.S. Trade Representative). Therefore, it is helpful for policymakers to consider the economic value of cross-border data flows to and from their regions when formulating the data governance policy. In particular, when imposing mandates for data localization, one should not lose sight of the economic value of the cross-border data flows that may be interrupted and the ensuing extra transaction costs that may incur for businesses. From the business perspectives, understanding the economic value of cross-border data flows can aid in the evaluation of the impacts of the global minimum tax and data localization policy.

Albeit a crucial asset to Big Tech and online platform companies, data has not been incorporated in a firm’s income statement based on the current accounting standard. Instead, all investments in data are categorized as expenditures, a practice that can significantly underestimate the profit rates of digital firms. The lack of knowledge about the value of data associated with cross-border data flows is also an immediate obstacle to understand data trade. As a result, it has been challenging for governments to use existing measures and address new issues introduced by firms that are data-intensive and have data-driven business models.

This situation is understandable because the value of data and data flow was hard to measure until very recently. Through our series of studies on the value of data, we have found that the value of the tons of data collected from consumers and third-party sellers by Amazon’s online retail markets is enormous [1]. We also find that capitalizing data will change the profit rate of Big Tech substantially [4]. For example, capitalizing Amazon’s value of data during 2002-2017 can increase its average profitability by 17% with an annual growth rate of 12.2%. For the Big Tech as a whole, the corresponding profitability during the same period of time increases 11.4% with an annual growth rate of 2.8%. These results show that Big Tech’s profitability when including the value of data can easily meet the criterion of 10% profit margin.

Last but not least, we find that, unlike the law of Big Tech’s value of data, the growth in the global value of data does not necessarily follow the growth in the Internet traffic and can be affected by the degree of diversification in data use. During 2014-2017, the estimated global value of data has already shown a slight declining trend due to Big Tech’s rapid gain in shares. Although this trend might have changed in 2020-2021 during which the COVID-19 pandemic forced a dramatic increase in online activities, the implications from the analysis based on pre-pandemic data still hold. When the data access and utilization is heavily concentrated and controlled by Big Tech and is limited for other firms, the growth in the global value of data created can be suppressed and even become negative.

Therefore, a more visionary topic for policymakers is to explore possible institutions and rewarding mechanisms for incubating a data sharing ecosystem in order to further unlock the power of data. As Hal Varian (Chief Economist of Google) rightly pointed out at the 2021 ASSA Meeting, data is a club goods which is excludable but non-rival. Because of the data network effect and the network effect of online platforms, it is hard for non-Big Tech firms to overcome their disadvantage in data by collecting data alone. The current discussions in global data governance policies such as data localization are focused mainly on privacy and data security. These are important issues, but it is also necessary to avoid the pitfalls if the actual measures for these policies in fact reinforce the data advantage of Big Tech that controls the gateway of data and has more resources to deal with the compliance costs, consequently worsening the data inequality among firms and between countries.

1. Li, W.C.Y. et al. (2019). Value of Data: There’s No Such Thing as a Free Lunch in the Digital Economy, VOX CEPR Policy Portal column article, July 23rd.

2. Coyle, D. & Li, W.C.Y. (2021). The Data Economy: Market Size and Global Trade, ESCoE Discussion Paper, September.

3. Li, Wendy C. Y., and P. J. Chi (2021), Online Platforms’ Creative “Disruption” in Organizational Capital — The Accumulated Information of the Firm, Moon Economics Institute Discussion Paper Series, No.1, August 2021. [PDF]

4. Li, W.C.Y. (2022). The Data Economy: Economic Value of Data and Data Flow, Profitability, and Global Minimum Tax, Moon Economics Institute Working Paper.

Copyright: Wendy C. Y. Li, 2021.

MEI Newsletter, No. 1 [PDF]

2021.8.22

Law of Big Tech’s Value of Data and Its Implications for Global Minimum Tax

by Wendy C. Y. Li

As the world has entered the era of 5G and the Internet of Things (IoT), there is no question that data is a key digital input to a firm’s production. Firms that use data to organize production enjoy higher productivity and market valuations. How to assess the value of data, however, has been a challenging topic that puzzles economists. Several recent studies [2,3] have recognized the conundrum of linking the exponential growth of global data flow to an economic value.

Our recent study [1] has provided the first and so far the only solution to this conundrum. It also presents new evidence that, in the digital era, Big Tech has been aggressively investing in organizational capital to grasp the great economic opportunities created by the explosive growth in global data. Moreover, our study finds that when the global data volume increases by five folds, the Big Tech’s combined value of data doubles, a clear empirical relationship we have termed Li’s law of value of data (See Figures below).

Another new study under preparation by us (Li, 2021) finds that data can have tremendous value and increase Big Tech’s profitability significantly. For example, we estimate that Amazon’s value of data is around US $125 billion, which accounts for 22.2% of its market valuation in 2017. Capitalizing on Amazon’s value of data can increase its average profitability by 17% with an annual growth rate of 12.2% during the period of 2010 to 2017.

As policymakers around the world are debating the global minimum tax proposal, one critical issue is that the value of data has not been capitalized into the firm’s financial statements. This missing piece of information can lead to substantial undervaluation of Big Tech’s profits. Our results on the profitability of Big Tech arising from the value of data can aid in the consideration of the global minimum tax.

1. Li, Wendy C. Y., and P. J. Chi (2021), Online Platforms’ Creative “Disruption” in Organizational Capital — The Accumulated Information of the Firm, Moon Economics Institute Discussion Paper Series, No.1, August 2021. [PDF]

2. Coyle, D.,  Diepeveen, S., Wdowin, J., Kay, L., & Tennison , J. 2020. The Value of Data: Policy

   Implications, February.

3. Tomiura, E., Ito, B., & Kang, B. 2020. Cross-border Data Transfers under New Regulations:

   Findings from A Survey of Japanese Firms, VOX CEPR Policy Portal column article, March 14th.

Copyright: Wendy C. Y. Li, 2021.