Data flows across the world. This is the impact on GDP

In around 25 years, the internet has become an integral part of our daily lives, connecting billions of users and businesses worldwide and leading to an explosion in the volume of cross-border digital flows. This column attempts to measure these flows and their impact on global activity in general. Global flows of goods, services, finance, people, and data have raised world GDP by at least 10% in the past decade, with the contribution to growth of GDP from data flows nearly matching the value of global trade in physical goods and services.

The literature on globalisation goes back a long way and has mostly concentrated on the international trade of physical goods and services. Recently, the literature has seen a renewal of interest in whether more intensive trade of goods and services leads to greater and more inclusive growth in economic activity (see, for example, early work by Rodrik 1997, and evidence from the aftermath of the GATT Uruguay Round in Estevadeordal and Taylor 2013).

There is no doubt that this is a relevant debate, but it may ignore another emerging phenomenon in international flows, namely, the recent explosion in the volume of cross-border digital flows and its impact on world activity at large.

In about 25 years, the internet has become an integral part of our daily lives, connecting billions of users and businesses worldwide. From the first author of this column buying goods online on Amazon.fr which are then shipped internationally to Belgium, to major textile companies purchasing materials out of global Chinese B2B platforms such as Alibaba, it is clear that data flows are exploding internationally. These data flows are more than just transactions; they also include inter-firm information, knowledge, and communications. Rio Tinto, for example, transmits data continuously from its mines, processing plants, and vehicle fleet to ‘excellence centres’ worldwide, where analysts monitor operations in real time and head off production delays. The energy giants BP and Shell are using ‘Internet of Things’ technology to explore reserves and track well production around the world.

The core issue for economists is getting a clear grasp on these data flows. Although a number of researchers have begun to take up to the challenge of measuring cross- border data flows (e.g. Meltzer 2014), the evidence remains at best anecdotal. In recent research from the McKinsey Global Institute (MGI), we took up the challenge and tried to assemble the most comprehensive view on those questions (Mankyika et al. 2016).

The hyper growth of cross-border data flows

We computed a volume measure of all flows, in bits per year, passing either though the public internet or through digital private networks. For this computation, we used the most extensive private data sets captured by Tele-geography, a private firm which tracks capacity and use of the global network of submarine optical fiber cables.

We found that, since 2005, the volume of data flows, measured in terabits per second, has multiplied by a factor of 45 in a decade, to reach an estimated 400 terabits per second by the end of 2016 (Figure 1). This contrasts with the fact that in same period, the traditional value flows of physical goods and services have barely managed to grow at the pace of worldwide nominal GDP. A part of the growth is due to digital information units becoming more and more enriched and shifting to broadband. Yet, the story that flows are growing on a large scale remains valid.

Digital flows intensity depends on the underlying goods and services. From being virtually inexistent 20 years ago, approximately 12% of physical trade of goods is now conducted via international B2C and B2B e-commerce. In China, already close to 20% of imports and exports takes place on digital platforms – approximately double the share in Europe. This digital share grows significantly when the underlying product is digitised – more than 30% of international communication worldwide in 2016 was via Skype, while among the high-profile news and entertainment companies, close to 80% of bits traffic originates internationally for the Financial Times, 60% for BBC, and up to 50% for NewsFeed and Netflix.

Cross-border data flows are not traditional flows

Data flows are different from traditional flows, with three noticeable features:

More global: Data flows spread farther.Contrary to the common statement that the “world is flat”, a great deal of trade has always moved within countries (Valeriano Martínez et al. 2017). This ‘home bias’ is reduced by the internet as far-flung buyers and sellers can easily connect with a few clicks. In our study, we found confirmation of other research that the home bias regarding physical trade is roughly halved for digital commerce (Lendle et al. 2012)More inclusive: SME participation in global trade is more widespread.A large share of physical exchange globalisation is concentrated in the hands of a small number of major multinational companies. The picture is also that a large part of e-commerce is though very large e-commerce sites such as Alibaba, Amazon, eBay, Flipkart, and Rakuten. But what is new is that these companies have also morphed into major marketplaces that are hosting millions of small enterprises around the world and turning them into ‘micro-multinational’ exporters. For example, across a sample of 18 countries, anywhere from 88% to 100% of eBay sellers are exporters, while only one out of ten small firms exports through physical channels (Lendle 2012).Large global dynamics: The cross-border data network remains concentrated but is quickly evolving and shifting to the East. International data flows of physical flows can be analysed as a social network connecting countries. While the cross-border data network is positively correlated with the one built up from the physical trade of goods and services flows, the correlation is weaker than early studies suggest (e.g. Barnett et al. 2001), in fact, weak enough to demonstrate that the global data network obeys its own dynamics.

In Figure 2, we report our estimates of both the level per capita, of cross-border data flows per country – which measures a country’s relative participation in global flows – as well the eigenvector of country data flows – which measures how central a country is in the participation of cross-border data flows. In general, cross-border data per capita is growing at an average of 50% per year per country, confirming the tremendous momentum of data flows for globalisation. The variance per country is also relatively large, with the amount of terabytes per capita in the Netherlands being twice as large the level of the US or the UK, and reflecting the position of the Netherlands as an international B2B data hub. Further, only a few countries have a very high centrality measure, pointing to the fact that the cross-border data network is still more concentrated than that of traditional physical flows of goods and services (Hausman and Hidalgo 2011). It is still also rather unbalanced, reflecting the head start by the US in adopting and using internet technologies.

Figure 2 Cross-border data flow network dynamics

Yet this network is changing relatively quickly. Relative to 15 years ago, US centrality is decreasing. The centre of gravity has been shifting ‘East’ to large European countries such as Germany and the UK. In Europe, a new order is also starting to develop; we are witnessing the strong participation of the Netherlands, Southern Europe merely catching up, and small Northern countries such as Belgium moving more and more to the periphery of the cross-border data network. Finally, also of note is the rise of Asian countries, notably China/Hong Kong and Singapore. This correlates well with the rise of Asia in worldwide digital commerce.

Large value derived from cross-border data flows

This torrent of data may be changing the nature of globalisation – and it is thus crucial to estimate its contribution to economic growth. As a first attempt, we estimated a simple production model where a country’s GDP growth is related to its stock of capital and labour, augmented by a latent total productivity growth variable which we assume to be positively related to all types of international flows. The idea behind this assumption is that globalisation acts as a major efficiency channel for countries participating in international flows, for example, through better resources allocation, or still better scale.

We compiled data for more than 150 countries for 20 years regarding six types of cross-border flows: physical flows of goods and services, FDI flows, financial flows, labour migration flows, and data flows measured in bits. As flows are most likely correlated to each other, we first resorted to a principal component analysis of flows and found that the largest factor accounted for up to 60% of the variance among flows, with all flows being positively correlated to the factor. Among the factors, this primary factor was also the only one to be statistically (and, as expected, positively) associated with a country’s economic growth.

Estimating a pooled cross-section, time series co-integration model of country GDP growth, we find that, together, global flows of goods, services, finance, people, and data have raised world GDP by at least 10% in the past decade, adding US$8 trillion of GDP by 2015. More crucially, and in part driven by the material growth in cross-border data bits internationally, the value of data flows has nearly matched the value of global trade in physical goods. By 2014, cross-border data flows accounted for $2.3 trillion of this value, or roughly 3.5% of total world GDP.

This estimate is only a first benchmark which will require further verification. But it underscores the importance of global data flows for economies at large. It also highlights new elements of consideration for economists, for policymakers, and for business. Given the significant contribution to GDP, governments must address pending issues such as free flows of data, cybersecurity, and privacy. They must also harness flows better through international standardisation of single payment systems, standardisation of internet of things protocols, coordination of tax issues, and integrated logistics. On the business side, the world’s biggest digital platforms – from e-commerce marketplaces to social media network – have become global in a matter of a few years, but though their concentration may be a concern, they have also amassed hundreds of millions of companies that can benefit from improved export opportunities and achieve major productivity gains.

Furthermore, the international flow of information facilitated by these digital technologies is a powerful driver of new performance for global firms, for example in optimising distributed R&D and innovation. Ultimately, everyone will need to go with the flow.

References

Barnett, G A, B Chon and D Rosen (2001), “The structure of the Internet flows in cyberspace”, Networks and Communication Studies 15(1–2), 61–80.

Estevadeordal, A and A M Taylor (2013), “Is the Washington consensus dead? Growth, openness and the Great liberalization, 1970s–2000s”, Review of Economics and Statistics 95(5), 1669-1690.

Gomez-Herrera, E, B Martens and G Turlea (2014) “The drivers and impediments for cross-border e-commerce in the EU”, Information Economics and Policy, 28, 83-96.

Hausmann, R, and C A. Hidalgo (2011), “The network structure of economic output,” Journal of Economic Growth 16(4).

Lendle, A, M Olarreaga, S Schropp and P-L Vézina (2012), “There Goes Gravity: How eBay Reduces Trade Costs.” World Bank Working Paper.

Mankiya, J, S Lund, J Bughin, J Woetzel, K Stamenov and D Dhingra (2016), Digital Globalization: The New Era of Global Flows, McKinsey Global Institute.

Meltzer, J ( 2014), “The Importance of the Internet and Transatlantic Data Flows for U.S. and EU Trade and Investment”, Brookings Institution.

Rodrik, D (1997) Has globalization gone too far?, Washington, DC: Institute for International Economics.

Valeriano Martínez, S, I Mateo-Mantecón and R Sainz-González. (2017), “Intra-national home bias: New evidence from the United States commodity flow survey”, Economics Letters 151, 4-9.