Ask the economists: Internet & development - towards a Wider World Web?



What benefits would the Internet bring to the developing world? And what impact would the arrival of several billion new users have on the Internet as we know it today?

Read below the questions and answers from this online debate hosted by Sam Paltridge from the OECD’s Science, Technology and Industry division  on Thursday 21 February 2008.

Sam Paltridge

Q. How feasible is it for the private sector to provide global internet access in poor countries (beyond local "reseller" ISPs)? Where razor-thin margins from pre-paid plans may be viable to sustain a wireless operator, will the market be sufficient to justify the investment in a fibre link through central Africa or remote parts of Asia to reach the global internet backbone? The geographic -- not just economic -- hurdle seems so high that it may require government funding, akin to a city sewer system or rural electrification. Might state-managed telecoms in this dimension be desirable?
Kenneth Cukier, The Economist

A. State managed telecoms was the predominant model in developed and developing countries for most of the twentieth century. The results, even in many OECD countries, were very poor and for developing countries the system was crippling.  Access growth in most developing countries has only really gathered pace following liberalization which allowed the private sector to invest in much needed infrastructure. Prior to that governments faced the dilemma of where to invest scarce resources (i.e. it was a case of sewers, electrification or telecoms) hence long waiting lists for service.  Even where governments wanted to invest (e.g. to address poor coverage in rural areas) the lack of commercial networks, in good condition or with reasonable geographic reach, meant the costs were often prohibitive.

The margins for pre-paid service may be thin but there are growing number of very profitable private sector operators in Africa and Asia and many operate in and across numerous countries. This has been complemented by a tremendous increase in South-South investment in service provisions as these companies extend successful models (including micro-entrepreneurial resale) in developing countries.  As a result international connectivity becomes a matter for their own commercial judgement when permitted (more 70 countries have international gateway monopolies). As the size of the market grows this allows a commercial case to be made for new international infrastructure and innovation develops at the service level. One example is African mobile companies allowing their users to roam across international borders, in countries where they operate, without incurring additional charges -- something not generally available in OECD countries!

As far as access to global Internet backbones is concerned there has been tremendous change in this market over the past several years.  Indian companies, for example, are the largest owners of undersea cables which span the globe. And due to competition in the Indian market neighboring countries, some of which are landlocked, have access to increasing amounts of capacity at far lower prices. In late 2006, for example, Nepal Telecom announced it could achieve a huge saving on the cost of international bandwidth by routing traffic via Indian fibre optic cables than over satellites.  (USD 1 700 per Mbps per month compared to USD 7 400 per month over satellite).


Q. Is there a specific action towards identifying (and possibly fostering) specific research & development actions and topics that represent roadblocks for innovation towards lowering costs,  in order to get innovative new solutions that could change the outlook for developing regions ability to have better access to ICTs? For instance one barrier to low cost CDNs (Content delivery networks) are patent issues in the domain of wireless Forward Error Correction Algorithms (Turbo codes, FEC). Open source alternatives would allow new kind of low priced services to emerge, but there is no industry interest in funding the neccessary research in this domain. (DIGITALWORLD EU-FP7 research project)
Roland A. Burger, DigitalWorld


A. The OECD has not looked at the specific examples you mention but, as per the earlier question, we do see potential for open source to play a role in bringing low cost solutions to developing countries. The OECD Ministerial on the Future of the Internet Economy will be preceded by three stakeholder forums involving Business, Civil Society/organized labour and the Internet Technical Community. These forums will carry forward messages to the Ministerial on where these communities see gaps or a need for government action or actions by other stakeholders. The Ministerial itself will highlight the role of R&D, innovation and creativity in addressing how to expand support for the global Internet economy.


Q By way of follow up how feasible is it for the private sector to provision global Internet service (ie, connection to the backbone)? Many poor countries, due to geography, are remote or landlocked. Can an economic model really take hold for Internet access for these countries? Am I right in supposing that even if the gateway monopolies were abolished, the cost would still be very high and bandwidth very constrained? Meanwhile, references to the mobile industry do not seem to fly, since the technology follows a vastly different capital-expenditure model (eg, build piecemeal as subscribers grow, and benefit from generous vendor financing). So it seems like providing global Internet access in poor countries is going to be quite hard. Do you agree? And again: How feasible  is it for the private sector to provide global Internet service in poor countries?
Kenneth Cukier, The Economist

A. It is very feasible if they are permitted by the policy/regulatory environment in which they operate.  Some countries do indeed face greater barriers due to geography. The case of Nepal cited earlier is one example where competition in a neighbouring country helped a landlocked country to reduce the cost of international bandwidth.  In respect to price trends once a monopoly is abolished it takes time to roll out new infrastructure.  Prices will remain high while monopoly power can be exerted as has been the case in many countries -- both developed and developing. Why mobile performance is underlined is that it is likely that this will be the first platform for many people in developing countries to gain access to the Internet. Already in many developing countries the number of mobile phone users to fixed users has a ratio of more than 20 to 1. In India, as noted in an earlier question, there are five times as many mobile internet subscribers as fixed Internet subscribers. The capabilities, of course, are different and much more development needs to be undertaken to develop services for wireless users. It is feasible for the private sector to provide global Internet service in poor countries -- they are doing it now.


Q. Why focus on the Internet only? Without diminishing, or even questioning the importance of the Internet for development, it should be recognised that other technologies also have great potential for development. One of these is radio, which has been and continues to be used extensively for development, sometimes in conjunction with the Internet. This technology is particularly suitable where literacy levels are low.
Helena Grunfeld, Centre for Strategic Economic Studies Victoria University

A. You are very correct to point to radio as being an important means of communication and it already plays a very valuable role.  But there are generally less regulatory barriers to using radio for these purposes. 


Q.What role do Internet Exchange Points (IXPs) play in the creation of  Internet service?
Name withheld

A. IXPs are places where different Internet networks can physically interconnect to send and receive traffic between their networks. Following the commercialization of the Internet they rapidly spread around the world to enable service providers to economically and efficiently exchange traffic locally.

In the absence of an IXP, in any country, local traffic between two service providers will by and large be exchanged internationally.  In these cases an email sent from one user to another, in the same country but using different service providers, may be routed via an IXP in New York or Paris rather exchanged domestically. By way of contrast, if that traffic is exchanged locally it can be far less expensive (i.e. avoiding expensive international circuits) and provide better performance for users (e.g. in some countries avoiding satellite circuits with their inherent delays).

Some 90 countries don’t have IXP’s today.  There is a map created by the Packet Clearing House which keeps count of countries with and without an IXP:
IXPs are relatively inexpensive (e.g. less than USD 40 000 per IXP) to establish and can benefit all stakeholders. For less than USD 4 million each country of the world without an IXP could have one.

Such expenditure, for example, in the form of development co-operation, however, would only make sense if the conditions are in place to enable the IXP to operate efficiently and become industry driven.  The main challenge is often is creating awareness of the benefits of IXPs.


Q. In many developing countries, taxes derived from inbound voice telephone calls fund critical national resources like schools and hospitals.  As low-cost Voice-over-IP telephony replaces the high-cost channel-switched telephony of the last century, how should governments adjust?  Can VoIP be taxed, or should governments give up on collecting special surcharges for telephony services and find other  sources of revenue?
Name withheld

A. It was once thought that communication services were a “luxury” that economies and societies could afford once other priorities for scarce public resources (remember providers were mostly state owned and often part of a government department) had been met. The result was under-investment and long waiting lists even for those people that could afford service. The diversion of capital away from the sector, both for worthy and unworthy purposes, effectively locked communications access into a perpetual cycle of underdevelopment. That is why many developing countries struggled to exceed telephone penetration rates of 1% or 2% of their populations for most of the twentieth century.

It is now common to view the use of ICTs to be critical for social and economic development. Policy reform to achieve this goal (e.g. liberalisation and privatization), has enabled the private sector to lead in investing the required capital and growing the size of the market (including for low income users). As the size of the market has grown so too has taxation revenue whether it be routine tax applied to any goods and services (e.g. on telephone calls) or corporate tax paid by service providers. In addition because there are viable commercial networks it has now become affordable for governments to assist in extending service to areas that formerly had no service and to initiate programs for connecting hospital and schools where the cost would have previously been prohibitive.

In respect to VoIP there is certainly some bypass of traditional telephone service at the international level. Not all VoIP, however, bypasses taxes. For example an international Skype-out call or SMS, from an Internet user to a telephone user, generates revenue for providers (and consequently for governments). But it also supports overall economic and social development in ways likely to yield greater benefits than constraining such services.


Q. Are there specific measures OECD members can take to see that more people get online in the next decade? Are these measure different for the 2nd billion people online, compared to the 3rd billion cohort, or the 4th billion cohort?  What is your long-term view of the requirements for Internet development, as measured by the number of users?
Irene Wu, Ph.D., Yahoo! Fellow in Residence Georgetown University

A. For Internet access to be extended to the next several billion users it has to be made increasingly accessible, affordable and applicable. Each of these challenges service providers to extend the reach of networks, lower costs and tailor services to meet the demands of new users (in the case of the latter both in relevance of content and suitability for the devices they will use).

Domestic policy is the most critical factor for developing all three but governments from developed countries and their agencies can play a positive role in capacity building. Specific to the communications sector this could include, for example, sharing experience in building institutions, such as regulatory agencies. OECD governments also need to play a role in keeping the Internet open to innovation and ensuring their own markets are open to competition.


Q.  The issue of the “next billion users” and access is likely to be a big topic at the upcoming  Internet Governance Forum (IGF) in India towards the close of 2008. Can you see any tie-in with your work and this conference? Secondly, what and how much of a role do you think governments will have to play in bringing the Internet to those not currently online?
Kieren McCarthy, ICANN

A. The OECD participated in the first (Athens, 2006) and second (Rio 2007) IGFs, including in the main panel session dealing with access and several related workshops. In 2008 we look forward to sharing the outcomes of the OECD Ministerial on the Future of the Internet Economy with the next IGF. On the second question all stakeholders can play a role in making the opportunities of economic and social development enabled by the Internet available to the next several billion users. Governments which act to give the private sector the lead in developing access will see the fastest growth but there are range of complementary measures addressed in previous questions across a range of policy areas.


Q. The Digital Divide seems can only get wider, and the prospect of having ‘the Internet’ get to all the planets population seems to pose a rather difficult operational task for those in the industries in charge. Given the large lack of education and access to technology, by the unconnected masses, and the rate at which technology is developing; it could lead to labour mobility issues, closing of developed economy borders to skilled workers, possibly social tensions in the long-term. Failing world economic could drastically slow the drive to get access to all, and capital return requirements may find it difficult to get this access realised.

Would you think this line of argument is reasonable? How would you suggest a solution? What would be the stance of OECD members with limited capabilities in a contracting economy?
Vernon K Padayachee, Atos Orgin, Spain


A. Not all aspects of the digital divide get wider. Mobile telephony is one example where the gap can be said to be closing. In respect to labour mobility improved communications can help in numerous and new ways. There is a growing use, for example, of Internet cafes and mobile telephones for international remittances by foreign workers to their families in their country or origin. Not only can such services be less expensive they can also be more secure.  In the reverse direction improved communications can help people from developed countries working in developing countries and it may be harder to attract such people to assist in economic and social development without such facilities.  Improved communications can also create employment opportunities for people in developing countries.


Q. How is the impact of the Internet on economic development in poor countries modelled and subsequently measured?
Antti Savilaakso, Sustainability Analyst, Dexia Asset Management

A. Unsatisfactorily!  Policy makers would like better data to help inform this question. OECD work has documented the positive relationship between ICTs and economic growth, in developed countries, and we are beginning to get better data that will help econometricians to tackle these questions more systematically compared to a decade ago.  Less work has been undertaken on the impact of ICTs in developing countries.  Some studies have been done on the relationship between mobile telephones and economic growth generally because of their greater diffusion to date than the Internet.  One initiative to develop better data in developing countries is joint initiative between several international organizations.  Information on The Partnership on Measuring ICT for Development  can be found here:


Q. The paper describes a number of innovative approaches used to market ICT services to the poor, the most notable of which is Grameen phone, and ascribes these to a deregulated environment. Why is it argued that these models could not have been developed in monopoly environments, when in fact they did? The Village Phone Ladies are the mobile equivalent of the privately owned public call offices, quite common in many developing countries well before the advent of competition.
Helena Grunfeld, Victoria University

A. Perhaps the innovation which has made the most difference to developing countries has been pre-paid cards. This was an innovation that came from competitive markets, first in OECD countries, but increasingly adopted in developing countries.  It is not that services can’t be developed in monopoly markets – they can be and were – but they may be held back or be unresponsive to customer demand.  ISDN is an example of a technology which came from a monopoly environment.  While there was plenty of supply push there was little customer demand in most countries (Germany and Japan being among a small group of exceptions).  After investing a great deal in the technology incumbent telcos were then reluctant to roll out  ADSL quickly so as not to cannibalize other parts of their business revenue – even though there was patent demand for high speed access to the Internet.

In the case of Grameen Phone there is a good report by TeleCommons Development Group(TDG) for the Canadian International Development Agency that well describes the challenges they faced in establishing service that was fiercely resisted by the state owned  incumbent state owned (BTTB).

This included BTTB’s refusal to interconnect, limiting where Grameen could build its network, not passing on international settlement revenue and so forth.  While BTTB could have offered service (e.g. public phones) in areas where Grameen was permitted to operate they did not because they judged the service to be uneconomic or not a priority.  Without liberalisation Grameen would not have had the chance to try its model.

Q. "According to a new OECD report, most of them will connect via wireless networks and many will be poor."  How big a factor is the affordability of Internet access for these poor as opposed to other factors (such as don't see the value)?  Is a subsidy contemplated? If so, how will it be calculated?
Name withheld


A. We now have more than three billion mobile phone users in the world and most of the users now joining such networks have very low incomes. They do so because, in their judgment, the service has a utility for some aspect of their lives. Moreover they are prepared, in relative terms, to devote a larger part of their incomes to communications than users in developed countries. One of the main reasons cited is security (e.g. emergency calls) but also because it empowers or enhances their lives in other ways (e.g. employment, social interaction). The challenge as you rightly point out is to make services relevant and readily accessible. Wireless will play a large role simply because it will be more accessible to many users. In India, for example, there are already five times as many number of subscribers accessing the Internet over wireless than fixed lines.

The question on subsidies will vary country by country depending on their circumstances. In contrast to times past, however, the starting point will be that the next several billion users represent a commercial opportunity and that the private sector should take the lead in meeting demand. Once commercial networks have been deployed it then becomes more feasible for governments to consider funding services to areas not addressed by the market whether that be a service for a rural village, school or hospital and so forth.


Q. How much consideration has been given to Internet content with efforts to connect the world's poor?  Currently, a majority of Internet content is not written in the languages of those in developing countries.  Further, even if language is not an issue, current Internet content is often irrelevant to these new users' information needs.  Are policy makers including strategies to create locally relevant content, both in language and scope?
Scott Dalessandro, University of Washington

A. From Nepal to Nigeria local content is generally the most popular with users when the barriers to its creation are removed or the necessary tools become available. In Nepal, following the creation of that country’s first IXP, the largest traffic spike for that year occurred with the online release of the local school results.  In Nigeria, the falling cost of digital cameras and computer editing has enabled that country to become the third largest producer of films.

Making content and services relevant to low income users is certainly prerequisite for making the Internet attractive to the next several billion users. In a competitive environment service providers have a tremendous incentive to address this issue in response to customer demand. Some of this will, of course, be of a popular nature (e.g. sport, entertainment) but also information that can economically empower users such as local market prices or employment opportunities.

It is also the case that governments can play a key role by making ‘public sector information’, in fields as diverse as health and education to meteorological, mapping or cultural heritage, available online.  Public policies can encourage a creative environment that stimulates market and non-market digital content creation, dissemination, and preservation of all kinds, and can be particularly helpful when they are coupled with initiatives aimed at addressing shortages in skills, training, education and human resource development for the creation, distribution and use of innovative digital content.


Q. In addition to making 'Internet access affordable' in developing countries, I would like to know what role do you think that open software could play in the development of the Information Society in developing countries to make 'Internet use also affordable' (i.e. thinking on both applications for companies as for the user digital literacy (open office and free internet resources and social web).
Mar Negreiro


A. Open source software can indeed play a role in making Internet access and use more affordable. Much of the Internet’s success to date can be attributed to its openness at different layers. This enables the creation of many applications and services that compete with existing services (both online and offline) as well as providing a platform for creativity and collaboration. In addition to the examples you mention open source could play a role in providing security options (e.g. software aimed at preventing malware) for low income users in developing countries, otherwise unable to afford such services.


Q. For those countries whose population are not in the 1st billion people online, (a) what can they do or (b) what do they need, to be part of the 2nd billion people who go online?

Irene Wu, Ph.D., Yahoo! Fellow in Residence Georgetown University


A. Initiatives to develop greater use of the Internet, and related ICTs, need to be considered as mainstream across all areas of economic and social development. An example would be developing the skills necessary to take advantage of the opportunities the Internet creates. More specifically, in relation to communication policy, the introduction of competition, separating the function of policy making from operational responsibilities and the creation of an independent regulator, with the power to enforce appropriate regulatory safeguards, are all necessary steps to drive access growth.

This topic is going to one of the main themes at the forthcoming OECD Ministerial on the Future of the Internet Economy (Seoul, Korea, 17-18 June 2008).


Background reading


Related Documents