2. Context
What is meant by internet governance?
Internet governance is associated with ‘a vital but relatively narrow set of policy issues related to the global coordination of internet domain names and addresses’.1 These are managed by ICANN through the Internet Assigned Numbers Authority (IANA) function. ICANN’s bylaws expressly limit its areas of responsibility and prevent it from engaging in issues outside its scope, such as content issues, or the impact of artificial intelligence on societies. Additionally, the internet’s landscape evolves in response to new technologies and innovations throughout the ecosystem, which limits the effectiveness of a narrow governance approach.
In 2005, the Working Group on Internet Governance defined internet governance more broadly, as ‘the development and application by Governments, the private sector and civil society, in their respective roles, of shared principles, norms, rules, decision-making procedures, and programmes that shape the evolution and use of the Internet’.2 The Internet Governance Forum (IGF), created by the World Summit on the Information Society,3 has adopted an inclusive attitude towards internet governance topics, and this flexible approach has enabled the evolving dialogue to reflect emerging issues.4
Domain names and Internet Protocol (IP) addresses are fundamental to the internet’s core architecture, giving each device connected to the web an identifier that is globally unique and universally accepted.5 Both domain names and IP addresses enable this and have remained relevant in the face of remarkable technological change over the past 20 years. That said, there are signs of a shift taking place. Since 2005, the number of domains per 100 internet users has declined from a highpoint of 12 (2008) to 9 (2018).6 The past decade has seen several substitutes for domain names gaining market share, notably social media accounts for individuals and businesses and apps for the mobile market, particularly in emerging economies.7 New internet protocols that support basic operations of the internet, like domain name system (DNS) over HTTPS (DoH) released by the Internet Engineering Task Force (IETF), risk splitting the internet’s DNS root zone. It also remains unclear what role, if any, domains and IP addresses will play as the Internet of Things (IoT) develops into a globally interoperable network. Meanwhile, slow adoption of IPv68 and increased use of network address translation9 continue to risk fragmentation of the IP address space. As of September 2019, only 24 per cent of the world’s internet traffic was IPv6 preferred.10
Figure 1: Domains per internet user 2005–18
— Source: Compiled by the authors from Verisign Domain Name Industry Brief and ITU Key 2005-2018 ICT data.
Figure 1: Domains per internet user 2005–18
Whether or not domain names and IP addresses have as much relevance in the future, an essential component of any internet governance strategy is the continuing focus on the system’s unique identifiers. But this narrow focus is insufficient to tackle widely recognized urgent internet governance issues, which currently have no internationally agreed upon multi-stakeholder home. These issues include AI, big data, IoT and its applicable technical standards, oversight of new recursive resolvers (i.e. using DoH), as well as privacy, cybersecurity and the role of states in cyberspace.
The changing nature of internet governance and strategic risks
O’Hara and Hall’s ‘Four Internets’ paper, describes four competing visions for the future of internet governance: the ‘open internet’ favoured by the US; the ‘bourgeois’ internet of the EU ‘where trolling and bad behavior are minimized and privacy protected, possibly at the cost of innovation’; China’s ‘authoritarian’ model; and the ‘spoiler’ internet for states such as Russia and North Korea, which exploit its open standards for strategic gain.11 All four approaches have differences that could characterize the future shape of the internet, or even fragment it. However, there are greater commonalities between the EU and US versus Chinese and Russian approaches – accentuating a potential East–West divide.
As the field of internet governance necessarily evolves to meet the challenges of technological and societal change, ideological fault lines between the EU and US are likely to emerge.
The relevance of the four internets model to this paper is that while internet naming and addressing (the traditional core of internet governance) have not to date posed major ideological differences between the EU and US, continued agreement cannot be taken for granted. As the field of internet governance necessarily evolves to meet the challenges of technological and societal change, ideological fault lines between the EU and US are likely to emerge. The future of a free, open internet in which human rights and the rule of law are respected is not guaranteed.12 There is a strategic imperative for the EU and US to emphasize areas of common ground in order to prevail against the emergence of an internet whose fundamental values differ from those upon which the network was founded. It is, therefore, important for the EU and US to work together to encourage participation of moderate, like-minded stakeholders in internet governance processes. The authors recommend that the EU Delegation seek to establish a taskforce for this purpose.
The term multi-stakeholder has become ubiquitous when discussing internet governance processes over the past 20 years, particularly those originating from the US. What does multi-stakeholder mean, and why has it become a charged term in the context of the four competing visions for the future of internet governance?
What does multi-stakeholder mean?
The term ‘multi-stakeholder’ was first coined in the 1990s as a way of extending the types of actors involved in policy and corporate decision-making. The starting point is a ‘complex, controversial issue on an international scale’,13 not unlike climate change.
In theory, multi-stakeholder governance decentralizes and democratizes decision-making. In internet policy, it is usually associated with ‘bottom-up’, ‘rough consensus’ policies developed by all stakeholders on an equal footing. For the majority of Western commentators this is viewed as the appropriate model for governing the internet,14 on the basis that involving diverse stakeholders leads to better policy outcomes.15 Multi-stakeholder is viewed as an alternative approach to intergovernmental, multilateral processes (typified by the International Telecommunication Union (ITU) and favoured by authoritarian regimes such as Russia, China and Iran). In this sense both have become charged expressions that import their advocates’ contrasting visions for the future of the internet – divided along the East–West or ‘four internets’ axis.
ICANN is above all the poster child of multi-stakeholder internet governance. Unfortunately, gaps between theory and ICANN’s practice threaten to become a strategic risk for both the EU and US visions for the future of internet governance. As practised within ICANN, multi-stakeholder governance gives equivalency to all voices, no matter how radical or self-interested, and there are underdeveloped mechanisms for breaking a deadlock or asserting the public interest.
The role of governments remains a contentious point, illustrating that, behind EU and US support for multi-stakeholder governance, there may not be a consistent, shared understanding of what ‘multi-stakeholder’ actually means. The structure of ICANN illustrates how these divergent views can impede closer cooperation in practice.
From ICANN’s inception in 1998, it was intended by the US government that the ‘private sector… take leadership for domain name system (DNS) management’.16 A foundational principle for ICANN – and what marks it out as different from the multilateral ITU – is that it is not government-led. This was originally articulated as ‘private-sector led’17 and, following the conclusion of the World Summit on the Information Society 2003–2015 (WSIS), morphed into ‘a bottom-up consensus-based multi-stakeholder process’.18 The ICANN structure has two types of stakeholder groups: the ‘supporting organizations’ – the Generic Names Supporting Organization (GNSO), the Address Supporting Organization (ASO) and the country code Names Supporting Organization (ccNSO) – which are focused on making policy for domain names and IP addresses; and the ‘advisory committees’ – the Governmental Advisory Committee (GAC), the Security and Stability Advisory Committee (SSAC), the Root Server System Advisory Committee (RSSAC), and the At Large Advisory Committee (ALAC) – which provide advice to the ICANN board. The advice of governments and other advisory committees is not binding on the ICANN directors, but over time the bylaws have strengthened the board’s obligations to provide reasons for not following GAC advice.19
Throughout ICANN’s numerous transitional periods, the US government has increasingly tried to stay at arms-length from ICANN to ensure its independence from government oversight. A respected advisory committee role within the community and acknowledgment of government’s role in internet governance suited this aim. Additionally, the US is traditionally a free market-driven economy with a light regulatory touch and decentralized governance, which is both an alternative structure compared to the EU and provides private industry a strong position in internet governance. With respect to multi-stakeholderism, the US vocally supports inclusion of all stakeholders, such as academia and civil society, but the US style of multi-stakeholderism can seem from the outside to be market-led.20
The US is traditionally a free market-driven economy with a light regulatory touch and decentralized governance, which is both an alternative structure compared to the EU and provides private industry a strong position in internet governance.
European policymakers would likely have set up ICANN in a different manner, for example, possibly including more accountability to the public to counterbalance corporate interests.21 To European policymakers, and representatives of other governments (especially authoritarian governments such as China, which would favour a multilateral solution to internet governance22), it is conceptually troubling to have sovereign states in a capacity that may be perceived as lesser than or secondary to the directors of a private corporation. In reality, the board’s powers to reject the policy recommendations of supporting organizations have always been limited.23
However, differences in status of the various stakeholder groups and advisory committees within ICANN are, for the most part, more problematic in concept than in practice. Over time, the powers and influence of ‘advisory committees’ has grown (for example, the SSAC is one of the most influential organs of the ICANN community), and successive changes to the bylaws have gradually strengthened the role of governments within ICANN.
A more practical way of thinking about multi-stakeholder processes is to adopt a flexible, results-orientated approach, driven by considerations of required expertise, which caters to those who are currently not being heard.24
What are the key internet governance institutions?
An exhaustive review of the institutions involved in internet governance is outside the scope of this paper. Table 1 summarizes the characteristics of key institutions active in the internet governance landscape. There are others, such as technical standards bodies, the regional and national domain name registries, registrars, and the organizations that manage the distribution of IP addresses. The intention is to focus on areas of interest and possible influence for the EU Delegation.
Table 1 shows the institution’s characteristics, openness (how easy it is for stakeholder groups to participate), what the forum’s decision-making process is (if there is one), whether government plays a formal role, primary participating stakeholders, the oversight body or convener of the forum, and the type of formal and informal outputs by the forum.
Table 1: Internet governance forums
|
Characteristics |
IGF |
National/Regional IGF (NRIs) |
IETF |
ITU |
UN GGE |
|---|---|---|---|---|---|
|
Openness25 |
Open |
Open |
Open |
Closed |
Closed |
|
Decision-making process |
n/a |
n/a |
Rough consensus |
Consensus among member states |
Consensus within the group |
|
Formal government role |
No |
No |
No |
Yes |
Yes |
|
Participation26 |
|
|
|
|
|
|
Oversight/Convener |
UN |
National & regional bodies |
Internet Society (ISOC) |
Members (states, industry, some academia) |
UN |
|
Outputs |
|
|
|
|
|
Source: Compiled by the authors.
To identify potential gaps in the thematic coverage of internet governance institutions, Figure 2 summarizes the institutions involved in the internet governance space and their areas of activity. The numbers assigned for each activity are an estimate of the intensity of each institution’s involvement in different areas. The analysis focuses on international processes and omits several institutions, which play important roles in shaping policy and standards at the national or regional levels, such as courts, sector regulation and laws, regional internet registries, and national and regional standards bodies such as the US National Institute for Science and Technology (NIST) and the European Technical Standards Institute (ETSI).
Figure 2: The internet governance radar
— Source: Compiled by the authors.
Note: The intensity of activity and selection of organizations reflect the authors’ view and may not reflect the views of the relevant institutions or processes.
Note: The intensity of activity and selection of organizations reflect the authors’ view and may not reflect the views of the relevant institutions or processes.
Figure 2: The internet governance radar
Note: The intensity of activity and selection of organizations reflect the authors’ view and may not reflect the views of the relevant institutions or processes.
Viewed through a thematic lens, the analysis reveals gaps in the international coordination of policy issues that have a major societal impact. For example, much of the policy on AI and big data is being handled de facto by a small number of private platforms as part of product development. Likewise, the IoT lacks an international policy space and key aspects of the technical standards are also occurring within technical companies or opaquely through the ITU.
Environmental analysis (far environment)
The far environment
To analyse drivers in the far environment, which may affect the EU–US relationship, this paper uses the familiar ‘PEST’ environmental scanning framework to review Political, Economic, Sociological and Technological factors relevant to internet governance. This analysis does not attempt to be exhaustive. Instead, the tool is used to give a high-level view of the drivers of change in the far environment that are likely to have an impact on internet governance processes.
Political
- China’s rise as a global superpower and its increasing influence in intergovernmental organizations such as the UN;
- Resurgence of Russia and the ‘Axis of Incivility’;27
- ‘Swing states’ (e.g. in the Gulf and in Africa) attracted by authoritarian and state-control solutions for communications technology;
- Loss of influence (US);
- Isolationist policies pursued by the current US administration are straining international cooperation;28 and
- Loss of confidence in democratic processes in liberal democracies and the rise of populism.
These drivers may undermine international support for multi-stakeholder internet governance, which is characterized by its critics as being favoured and dominated by US interests.
Economic
- Increased influence and coordination of the rest of the world, notably by China and Russia, whose visions for internet governance are incompatible with those of the EU and US;
- China and Russia are now competing in terms of investment and know-how with the US, in ways that were inconceivable a short time ago;
- Russia’s new ‘sovereign internet’ law and reported ability to ‘disconnect’ from the global internet’;29
- China has developed home-grown internet and technology giants (such as Alibaba and Huawei) that are competing internationally;30 and
- China’s Belt and Road Initiative, which so far spans 70 countries31 and aims to support technical infrastructure development in sub-Saharan Africa for decades.32
The impact on internet governance is seen through the growth in support from ‘swing states’ (states that are undecided on the future direction of internet governance) for the ITU, a government-led standardization body, to have a role in internet governance – an approach favoured by authoritarian states such as China.
Sociological
- The future of work,33 combined with ageing34 and declining populations in the developed world,35 will result in decreasing tax revenues;
- While several emerging economies (e.g. China, Russia and Brazil) have below replacement-level fertility, India and many African countries have large youth populations;36 and
- Decisions over how new technologies like 5G and protocols like DoH are adopted at the national and regional levels are likely to deepen the digital divides, as are the handling of other policy issues such as data protection, surveillance, access to services, and the right to disconnect.
These factors will inhibit the ability of developed world governments (especially the US and EU) to effect change or sustain a multi-stakeholder vision for internet governance without the support of other stakeholders.
Technological
- Forces for a ‘splinternet’ are strengthening37 – not only risking alternative DNS roots, but other technological splits (e.g. in 5G with a Chinese Huawei internet vs a Western internet);
- The future of the internet will not be web- or email-based; and
- China has emerged as a global leader in technology development and deployment. How far will Chinese technologies embed an alternative, authoritarian, approach to governance, human rights (e.g. privacy), and surveillance.
With the roll-out of smart cities, the IoT and advances in artificial intelligence, governing tomorrow’s internet will be more complex than coordinating domain names and IP addresses, and will involve politically, ethically and culturally challenging issues.
1 Mueller, M. (2010), Networks and States, Cambridge, MA, MIT Press, p. 9.
2 Working Group on Internet Governance (2005), ‘Report of the Working Group on Internet Governance’, p. 4, http://www.wgig.org/docs/WGIGREPORT.pdf (accessed 4 Oct. 2019).
3 International Telecommunications Union (ITU) (2005), Tunis Agenda for the Information Society, 18 November 2005, para 72, https://www.itu.int/net/wsis/docs2/tunis/off/6rev1.html (accessed 4 Oct. 2019).
4 At the first IGF meeting in 2006, the main session topics were openness, access, security, diversity and emerging issues, see Doria, A. and Kleinwachter, W. (eds) (2008), Internet Governance Forum: The First Two Years, Internet Governance Forum, http://www.intgovforum.org/multilingual/filedepot_download/3367/5 (accessed 14 Oct. 2019). By 2018, the core topic of the IGF meeting was ‘The Internet of Trust’ and the main sessions were on cybersecurity, trust and privacy; the evolution of internet governance; development, innovation & economic issues; and human rights, gender and youth, see IGF (2018), ‘The Internet of Trust’, conference schedule, https://www.intgovforum.org/multilingual/index.php?q=filedepot_download/6785/1342 (accessed 14 Oct. 2019).
5 O’Hara, K. and Hall, W. (2018), ‘Four Internets: The Geopolitics of Digital Governance’, CIGI Papers No. 206, December 2018, https://www.cigionline.org/publications/four-internets-geopolitics-digital-governance (accessed 4 Oct. 2019).
6 Authors’ analysis based on the following sources: Verisign (n.d.), ‘Verisign Domain Name Industry Brief Archive (2009–2018)’, https://www.verisign.com/en_US/domain-names/dnib/domain-name-industry-brief-reports/index.xhtml#2019 (accessed 4 Oct. 2019); ITU (2018), ‘Key ICT indicators for developed and developing countries and the world (totals and penetration rates)’, https://www.itu.int/en/ITU-D/Statistics/Documents/statistics/2018/ITU_Key_2005-2018_ICT_data_with%20LDCs_rev27Nov2018.xls (accessed 4 Oct. 2019).
7 Oxford Information Labs, EURid, Emily Taylor Consultancy and Abu-Ghazeleh Intellectual Property (2016), Middle East and Adjoining Countries DNS Marketplace Study, ICANN 2016, https://www.icann.org/en/system/files/files/meac-dns-study-26feb16-en.pdf (accessed 4 Oct. 2019); Oxford Information Labs, LACTLD, EURid and InterConnect Communications (2017), Latin American and Caribbean DNS Marketplace Study, ICANN, 13 March 2017, https://www.icann.org/en/system/files/files/lac-dns-marketplace-study-13mar17-en.pdf (accessed 4 Oct. 2019).
8 See, for example, InterConnect Communications (2012), MC/111 Internet Protocol Version 6 Deployment Study, Ofcom, https://www.ofcom.org.uk/__data/assets/pdf_file/0028/55891/internet-protocol.pdf (accessed 4 Oct. 2019).
9 See, for example, InterConnect Communications (2013), MC/159 Report on the Implications of Carrier Grade Network Address Translators, Ofcom, https://www.ofcom.org.uk/__data/assets/pdf_file/0020/37802/cgnat.pdf (accessed 4 Oct. 2019).
10 Asia-Pacific Network Information Centre (APNIC) (n.d.), ‘IPv6 Capable Rate by country (%)’, https://stats.labs.apnic.net/ipv6 (accessed 4 Oct. 2019).
11 O’Hara, K. and Hall, W. (2018), Four Internets: The Geopolitics of Digital Governance, CIGI Papers No. 206, December 2018, https://www.cigionline.org/publications/four-internets-geopolitics-digital-governance (accessed 4 Oct. 2019).
12 Ibid.; Global Commission on Internet Governance (2016), ‘One Internet’, https://www.cigionline.org/sites/default/files/gcig_final_report_-_with_cover.pdf (accessed 15 Oct. 2019); Hoffmann, S., Bradshaw, S., and Taylor, E. (forthcoming 2019), ‘Networks and geopolitics: how great power rivalries infected 5G’, CIGI.
13 Hofmann, J. (2016), ‘Multi-stakeholderism in Internet governance: putting a fiction into practice’, Journal of Cyber Policy, 1(1), pp. 29–49, https://doi.org/10.1080/23738871.2016.1158303.
14 For example, OECD (2014), ‘Principles for Internet Policy Making’, http://www.oecd.org/internet/ieconomy/oecd-principles-for-internet-
policy-making.pdf (accessed 15 Oct. 2019); Global Commission on Internet Governance (2016), ‘One Internet’.
15 Wentworth, S. (2017), ‘Internet multi-stakeholder governance’, Journal of Cyber Policy, 2(3), pp. 318–322, https://www.tandfonline.com/doi/full/10.1080/23738871.2017.1400574 (accessed 15 Oct. 2019).
16 United States Department of Commerce (1998), ‘White Paper on the Management of Internet Names and Addresses’, https://www.icann.org/resources/unthemed-pages/white-paper-2012-02-25-en (accessed 4 Oct. 2019).
17 National Telecommunications and Information Administration (NTIA) (1998), ‘Statement of Policy on the Management of Internet Names and Addresses, https://www.ntia.doc.gov/federal-register-notice/1998/statement-policy-management-internet-names-and-addresses (accessed 4 Oct. 2019).
18 The term ‘multi-stakeholder’ first appeared in ICANN’s bylaws in October 2016, following extensive revisions associated with the IANA transition. At the time of the WSIS, ICANN’s core values included ‘While remaining rooted in the private sector’; compare ICANN (2005), ‘Bylaws for Internet Corporation for Assigned Names and Numbers’, https://www.icann.org/resources/pages/bylaws-2005-04-08-en (accessed 4 Oct. 2019); and ICANN (2016), ‘Bylaws for Internet Corporation for Assigned Names and Numbers’, https://www.icann.org/resources/pages/bylaws-2016-09-30-en (accessed 4 Oct. 2019).
19 Compare Section 3(a) of the original 1998 bylaws, which requires the board to ‘consider’ government advice, to the most recent bylaws (section 12.2(a)(x)), which adopt a ‘comply or explain’ formula, and sets a minimum threshold of 60 per cent vote of the board to reject GAC consensus advice, ICANN (1998), ‘Bylaws for Internet Corporation for Assigned Names and Numbers, https://www.icann.org/resources/unthemed-pages/bylaws-1998-11-23-en (accessed 4 Oct. 2019); and ICANN (2018), ‘Bylaws for Internet Corporation for Assigned Names and Numbers’, https://www.icann.org/resources/pages/governance/bylaws-en (accessed 4 Oct. 2019).
21 Anonymous interviews for this paper, November 2018; O’Hara and Hall (2018), Four Internets: The Geopolitics of Digital Governance
22 Ibid.
23 See, for example, ICANN’s original bylaws, section 1(c) ‘The Board shall accept the recommendations of a Supporting Organization if the Board finds that the recommended action, policy or procedure (1) complies with the Articles and Bylaws, (2) was arrived at through fair and open processes (including permitting participation by representatives of other Supporting Organizations if requested), (3) is not reasonably opposed by any of the other Supporting Organizations, and (4) furthers the purposes of, and is in the best interest of, the Corporation’, ICANN (1998), ‘Bylaws for Internet Corporation for Assigned Names and Numbers, https://www.icann.org/resources/unthemed-pages/bylaws-1998-11-06-en#VI (accessed 15 Oct. 2019).
24 Wentworth, S. (2017), ‘Internet multi-stakeholder governance’.
25 Openness is assessed on the ease with which stakeholders can engage meaningfully in the forum. This may include attendance, ability to observe processes, participation in discussions and output creation, participation in decision-making processes, access to meeting notes or summaries, financial barriers to participation, etc.
26 Multi-stakeholders were divided into four groups (government, academia, civil society, and industry). Other groups, such as technical experts, could be included in any of these four groups. The order of the groups is intended to describe the relative importance of that group to the forum but is not definitive.
27 O’Hara and Hall (2018), ‘Four Internets: The Geopolitics of Digital Governance’.
28 Volcovici, V. (2017), ‘U.S. submits formal notice of withdrawal from Paris climate pact’, Reuters, 4 August 2017, https://www.reuters.com/article/us-un-climate-usa-paris/u-s-submits-formal-notice-of-withdrawal-from-paris-climate-pact-idUSKBN1AK2FM (accessed 16 Oct. 2019); BBC News (2018), ‘EU tariffs on US goods come into force’, https://www.bbc.co.uk/news/business-44567636 (accessed 16 Oct. 2019); Lockie, A. (2018), ‘Trump torches allies, threatens NATO pullout after tense WWI memorial trip to Paris’, Business Insider, http://uk.business
insider.com/trump-slams-allies-threatens-nato-pullout-after-wwi-paris-trip-2018-11?r=US&IR=T (accessed 16 Oct. 2019).
29 Jee, C. (2019), ‘Russia wants to cut itself off from the global internet. Here’s what that really means’, MIT Technology Review, 21 Mar 2019, https://www.technologyreview.com/s/613138/russia-wants-to-cut-itself-off-from-the-global-internet-heres-what-that-really-means/ (accessed 4 Oct. 2019).
30 See, for example, Mourdoukoutas, P. (2017), ‘Alibaba beats Amazon’, Forbes, 22 August 2017, https://www.forbes.com/sites/panos
mourdoukoutas/2017/08/22/alibaba-beats-amazon/#488706773f97 (accessed 16 Oct. 2019).
31 Hillman, J. E. (2018), ‘China’s Belt and Road Initiative: Five Years Later’, Center for Strategic and International Studies, https://www.csis.org/analysis/chinas-belt-and-road-initiative-five-years-later-0 (accessed 4 Oct. 2019).
32 See, for example, World Bank (2008), ‘Building Bridges’, https://siteresources.worldbank.org/INTAFRICA/Resources/Building_Bridges_Master_Version_wo-Embg_with_cover.pdf (accessed 16 Oct. 2019); The Infrastructure Consortium for Africa (2015), ‘Africa’s ICT sector and China’, https://www.icafrica.org/en/topics-programmes/ict/africa%E2%80%99s-ict-sector-and-china/ (accessed 16 Oct. 2019).
33 World Development Report 2016 (2016), ‘Digital Dividends’, Figure O.17, p. 22 https://openknowledge.worldbank.org/bitstream/handle/10986/23347/9781464806711.pdf (accessed 16 Oct. 2019).
34 World Population Prospects: the 2017 Revision indicates that the number of older persons aged 60 or over is expected to double by 2050 and to more than triple by 2100, United Nations (n.d.), ‘Ageing’, http://www.un.org/en/sections/issues-depth/ageing/ (accessed 16 Oct. 2019).
35 See United Nations (2015), ‘World Fertility Patterns 2015’, p. 3, http://www.un.org/en/development/desa/population/publications/pdf/fertility/world-fertility-patterns-2015.pdf (accessed 4 Oct. 2019).
36 See, for example, UN (2017), ‘World Population Prospects, The 2017 Revision: Key Findings and Advance Tables’, https://population.un.org/wpp/Publications/Files/WPP2017_KeyFindings.pdf (accessed 4 Oct. 2019).
37 Global Commission on Internet Governance (2016), ‘One Internet’.