Internet Governance/Internet Governance and Development

What is the digital divide and why does it matter?
The digital divide is the term used to refer to the gap between those who have access to the Internet and its associated technologies and services, and those who do not. Although not specifically about the gap between rich and poor countries – a digital divide can exist within a country between urban and rural areas, for example, or between socio-economic groups – the term is frequently used to refer to the access gap between developed and developing countries.

As Table 1 indicates, the gap is striking, and Internet penetration rates are closely connected to the wealth of a country. Such numbers matter for at least two reasons. First, because in today’s global economy, lack of access to the Internet means lack of access to world markets. The digital divide between rich and poor countries prevents the latter from selling their products on a global scale, and restricts the choice of services and goods available.

In addition, the digital divide is important because the lack of computers and Internet sites in poor countries means that these countries have only a limited presence on one of the most important (and certainly most dispersed) forms of media ever invented. Thus the digital divide is limiting cultural and regional diversity on the Internet.

Table 1: GNI and Internet Penetration

Source: Internet World Stats (http://internetworldstats.com, Sept 2005).

What is the relationship between Internet governance and the digital divide?
Decisions made in the realm of Internet governance can play a significant role in alleviating or exacerbating the digital divide.36 Although many of these decisions may appear purely technical, they have important social, political and economic consequences. The areas of governance that impact the digital divide include (but are by no means limited to) the following:

Internationalized Domain Names

Currently, all domain names must be entered in standard ASCII (American Standard Code for Information Interchange) characters, which are designed to support the Latin alphabet. This means that diacritical marks, as well as Asian or other international characters, are not supported. Many developing countries feel that the exclusion of their languages from domain names limits Internet access. Users who are not familiar with English have a difficult time accessing English-language URLs; in addition, the lack of foreign-script support makes it difficult for indigenous businesses and entities to be represented on the Internet.

Implementing IDNs is not a simple task. The IETF has developed standards for non-Latin scripts, but complex technical and governance issues have prevented their adoption. ICANN has also been criticized (perhaps not always fairly) for acting slowly on the issue. Adopting IDNs will require greater coordination between national governments and the technical community. It will also require coordination between nations that share a common script (e.g., China, Korea and Japan) so that appropriate standards can be developed. Private entities will also have to be brought on board. Most existing versions of Microsoft Windows, for instance, do not support existing IDN technical standards without substantial updates (though future versions are said to include support), making it difficult to implement IDNs even if standards were agreed upon. The matter is made even more complex by the fact that domain names are used in contexts that employ ASCII encodings—e.g., in URLs, SMTP for email and so on. Thus, the ease of entering a domain name in a natural language script is made more difficult when it is embedded in a URL. In addition, the “. ” is not usually a punctuation character in all scripts.

Country Code Top-Level Domain Names

The use of country-level domain names (e.g., .in, .jp, .sg) is another issue that can determine access in developing countries. In many cases, governments have mismanaged these valuable resources. For example, a recent study suggests that management of Cambodia’s ccTLD became less efficient and more expensive when the government assumed control from an NGO. Bangladesh is another country where concerns have been expressed. Poor management of ccTLDs can limit a country’s visibility on the Internet. In India, relatively few companies had registered “.in” domain names until recently due to requirements that sites be hosted only on servers within the country; under the new system, registrations in the .in ccTLD are growing rapidly.

Poor management can also represent missed revenue opportunities. Operators in some small developing countries have successfully marketed their ccTLDs as alternatives to ICANN’s top-level domain names (Table 2).37 However, when such domain names are not owned by States and marketed by private companies, the misuse can be considered a violation of a country’s collective intellectual property resources.

Table 2: ccTLDs as Alternative Top-level Domain Names

Source: Gelbstein and Kurbalija (2005,44)

Standards

Decisions on technical standards also play an important role in shaping the digital divide. Standards that are proprietary (i.e., whose intellectual property is owned by private entities) can make the costs of access to technology prohibitive by requiring expensive royalty payments. In this context, it is important to remember that the global reach of the Internet depends on the use of open standards, notably TCP/IP, XML and other standards for email, voice over the network, streaming audio and video, collaboration and peer-to-peer exchanges. Such standards have not only made the global network connect seamlessly, they have also made it possible to connect affordably. Problems arise, however, when private standards are used, as developing countries then have to pay for the right to use companies’ intellectual property rights.

In addition to affordability, open standards are also important because they allow developing countries to modify and enhance technology for their particular needs. Open source software, for example, can enable adaptation of operating systems and software packages to local languages or conditions. Indeed, the future of access in the developing world may depend to a significant extent on such open, modifiable code. A number of low-cost access devices (e.g., the Simputer, a handheld developed in India, or Nivo, a thin-client developed for African users), rely for their affordability and usability on open source software. More promisingly, Linux and other open source software, which can be modified freely, are being used with growing frequency by governments, companies and individual users in the developing world as alternatives to more expensive proprietary software. A key governance strategy to bridge the digital divide could therefore include steps to encourage FOSS – for instance, by providing training or other forms of support to organizations that use open source software, or by requiring government agencies to adopt operating systems like Linux.

Recently, it has become evident that developing countries (and others) can suffer not just through the overt adoption of proprietary standards, but also through a more subtle process by which private entities “hijack” open standards and, through modifications, turn them into de facto proprietary standards. This problem has been apparent, for example, in the regular addition of so-called enhancements by companies to HTML and XML. The result is that many features on certain web pages are only fully accessible, for example, on Microsoft’s Internet Explorer.38 A similar situation has been evident, too, with JavaScript, the programming language that powers many websites. Through a steady stream of additions and modifications by various private companies, this ostensibly open standard has been balkanized into several competing versions.

IP Address Allocation

As noted in Section II, the shortage of IP numbers has been a matter requiring governance solutions. Developing countries, in particular, have been concerned as over 80 percent of existing IP numbers have been allocated to organizations in North America. This anomalous situation largely exists due to the historical development of the Internet, and the fact that early IP allocations were made on a first-comefirst-served basis. Since the mid-1990s, however, IP numbers have been managed by RIRs, and have been allocated on the basis of demonstrable need. Figure 2 includes the list of RIRs and the percentage of IP numbers allocated by them since 1999. It shows that the historical anomaly has somewhat righted itself, although concerns for inequalities persist in IP allocation.

The shortage of IP space is likely to be substantially eased with the introduction of a new IP known as IPv6. Concerns, however, persist about methods of allocation that will be adopted for this new protocol: it is imperative for developing countries – especially those currently lacking the technical know-how – to ensure that they receive their equitable share.39

39It should also be noted that equitable is not necessarily equal, and that actual use needs to be taken into account. Poorly crafted assignment policies can lead to explosion of routing tables that could interfere with the operation of the Internet.

Cost of Connection

Figure 2: IP Addresses Allocated Since 1999 (percentage by RIR)

Figure 3: Internet costs and GDP/capita (monthly basis)

As Figure 3 suggests, the cost of Internet access is often substantially higher in developing countries than in developed countries. As explained earlier, there are many reasons for this disparity. The ambiguities and disparities in international traffic charges, discussed in Section II, are one reason. Equally important reasons include a lack of infrastructure, inadequate competition policies, under-developed markets, and predatory pricing by quasi (or actual) monopolies in the ISP market. Addressing all these issues requires a multi-pronged approach that operates at the national, regional, and international levels.

At the national level, the priority should be to foster the development of competitive and open markets for ISPs. One important component of such a strategy is to lower licensing fees and norms; this ensures that new players can enter the market without undue burden, and fosters healthy competition. In addition, ISPs in developing countries often rely on a dominant (sometimes State-owned) telco for access to international bandwidth; this monopoly access to international bandwidth can artificially inflate access prices. Finally, national policies and regulations on interconnection can play a key role. While interconnection terms and prices between network providers are generally best left to the market, in some cases (e.g., when there exists a dominant ISP or telco that can determine interconnection fees) states may need to impose regulation to ensure a more equitable relationship that permits smaller ISPs to connect at fair prices with larger players. The resulting lower costs, of course, should then be passed on to consumers.

In addition to these national policies aimed at fostering competition, the ambiguities and inequities in the international settlement regime, highlighted by WGIG, are also essential to address. As mentioned, one of the main stumbling blocks is the lack of an appropriate global forum where a fair and equitable system for international rates can be negotiated; identifying such a forum is therefore a critical need.

In the absence of such a forum, many developing countries have taken matters into their own hands by setting up regional Internet Exchange Points (IXPs). Such IXPs keep regional traffic regional – i.e., by routing traffic locally, they reduce the amount of international bandwidth that must be used, and as such also reduce payments to international operators. (Without IXPs, even an email sent from one city to another within a given country could be routed via the United States). Over 150 such IXPs now exist around the world, and the Asia-Pacific region was one of the first parts of the developing world to begin routing a substantial portion of traffic through a regional IXP.

Finally, reducing international access charges will depend on reducing users’ dependence on international content. Ultimately, it is the lack of local content, stored on local servers, that increases the consumption of international bandwidth. It is thus critical for governments (and other entities) to make it easier to host content locally, as well as to encourage the development of local content. In many developing countries with non-Latin alphabets, this in turn also requires support for local language fonts and local-language software.

What is the current status of developing country participation in Internet governance?
The preceding list of issues makes it clear that Internet governance impacts developing countries in a variety of ways. Calls for greater participation by the developing world have therefore become an increasingly common feature of debates and discussions over Internet governance. These calls have been accompanied by a growing recognition that, for the most part, developing countries are not adequately represented in most governance fora and, when they are represented, often do not have adequate technical capacity or resources to participate on equal terms.

Perhaps because of its centrality to the governance process, ICANN has come in for particular criticism over its perceived exclusion of developing countries. However, to a greater or lesser degree, such concerns have been expressed with regard to a variety of bodies related to Internet governance. In 2002, the Commonwealth Telecommunications Organization (CTO) and Panos, a non-profit organization, issued a landmark study, Louder Voices,40 in which they listed a number of steps that could be taken to enhance developing country participation in technical governance (see below). With regard to current levels of participation, the study reached three general conclusions:


 * First, the study found that developing countries are for the most part represented in intergovernmental organizations like ITU and WTO, but that such organizations frequently pay scant attention to the connection between communications policy and development. The study therefore identified a “missing link” between technology policy and development in many important decision-making bodies.


 * Second, the study found that developing countries were generally under-represented in nontraditional decision-making venues, such as the standards-setting bodies, ICANN and other technical groups. Given the centrality of such groups to the management of the Internet, this represents a serious handicap on developing country participation in Internet governance. Combined with the previous observation, it also explains why many developing countries have been pushing for inter-governmental bodies to play a greater role in Internet governance.


 * Third, the study found that, when it comes to governance decisions led by the market, developing countries have virtually no representation at all. This is an important shortcoming for many Internet governance decisions are determined by market-driven processes that result in de facto standards. Developing country exclusion from such processes is, of course, simply a reflection of their more general exclusion from global markets.

How can barriers to developing country participation be overcome?
The CTO/Panos study clearly illustrated shortcomings in current systems of Internet governance. That study, as well as a report submitted to the Digital Opportunity Task Force (DOT-Force), a group set up by the G8 at its Genoa Summit in 2001,41 identified several key steps that could be taken to increase developing country participation. Five of the most important include:

Increase policy awareness: One of the chief obstacles to effective developing country representation is a lack of awareness about policy, policy venues, and even the basic need or relevance of a governance process. Indeed, the CTO/Panos study found a striking lack of awareness when it came to the social and economic impact ICTs can have on development.

Several steps can be taken to increase awareness. One important way is by developing global ICT policy information resources which would provide, among other things, information on the relevance of ICT activities, as well as dates and information on policy venues. Possible vehicles for this information include a web or email-based newsletter, an annual summary of important issues, a research team to field questions from members, and conferences on particularly significant issues. An e-library, containing relevant information on policy, could also be considered.

Build technical and policy capacity: Developing countries’ inadequate technical and policy capacity is a further fundamental barrier to their participation. The CTO/Panos study concluded that the lack of technical capacity is particularly problematic with regard to emerging issues such as migration to IP-based networks, implementation of third-generation mobile communication systems, and e-commerce applications.

Overcoming such barriers is difficult, given the years of education and experience required. However, capacity building measures do exist. One important mechanism suggested by the study is the establishment of a global network of public and policy research institutes, with offices or “nodes” in developing countries, to work on ICT related issues. These nodes would function as training institutes, and help build policy capacity. In order to ensure the effectiveness of this network, of course, financial support would be required.

Provide financial support: In order to increase developing country representation, a wide variety of financial support mechanisms are required. Funding for the above-mentioned capacity-building network is one possible mechanism. Providing travel fellowships and other means for developing country representatives to attend policy venues (often held in expensive first-world locations) is another.

However, financial support does not simply involve throwing more money at the problem. According to the CTO/Panos study, financial hurdles often tend to involve ineffective use of resources rather than resource deficiency. Resources are often poorly allocated, for example, with inappropriate people being sponsored to represent stakeholder interests. To address such issues, developing countries need to evaluate current practices and redesign them so as to ensure greater efficiency. It might be helpful, for instance, to develop a code of practice or other accountability mechanisms to ensure efficient use of resources.

Strengthen national policy institutions and processes: Although establishing a global network is important, it is also essential to strengthen capacity at home. National and regional institutions in developing countries are frequently weak on several levels. At the national level, political leadership is often lacking, national ICT strategies are often absent, and coordination between government departments and agencies is often inadequate. At the regional level, coordination between governments and user groups sharing common interests is often lacking. Further weaknesses at this level include poor preparation for international meetings and ineffective use of human and financial resources.

One remedial step is to improve information flows and policy coordination between government departments and agencies. Another is to promote sharing of experience and expertise at all levels, including sub-regional and regional. Indeed, such collaboration is perhaps one of the most important steps developing countries can take, as it permits a “sharing of bargaining power” that increases their clout. Importantly, in order for such collectives to be truly effective, they should not be limited to just one sector (e.g., inter-governmental alliances), but be true multi-stakeholder alliances.

Facilitate participation in international policy fora: Finally, if developing country representation is to increase in international institutions, it is essential that such institutions specifically consider developing countries in their organizational structure and processes. For example, meetings can be held in developing regions in order to minimize attendance costs. Scheduling of events should also take into account other international events that might require the participation of the limited talent pool available in developing countries. And finally, internal governance structures (e.g., the allocation of seats or voices on committees, the balance of power between different sub-committees) can be designed in a way to enhance developing country interests. It could be possible, for example, to create advisory committees whose specific goals are to provide input on developing country needs.