April 20, 2014


The Internet “Access Trap” in Developing Countries

Three of five people in the world  still do not have access to the Internet.  From the perspective of standard economic models, this is puzzling. The supply of international connectivity has expanded dramatically since 2009, when several submarine fiber cables came online connecting even the poorest countries in Africa to the global Internet. Also, with only a few exceptions, nearly every developing country now has some form of competitive market for broadband services.

Despite this, few of these countries are close to achieving the UN Broadband Commission’s goal of entry-level broadband services priced at less than 5 percent of average monthly income. The Affordability Report, released last month by the Alliance for Affordable Internet Internet (A4AI), a consortium of private companies and public sector organizations dedicated to bringing Internet costs down through policy change, found that in at least 46 countries “the cost of entry-level broadband services exceeds 40 percent of monthly income for people living under $2/day, and in many countries exceeds 80 percent or even 100 percent of monthly income” (I co-authored the Affordability Report with Sonia Jorge).

                                                                           source: A4AI Affordability Report

One of the most interesting findings in the report is that at the global level, the majority of people for whom broadband is unaffordable live not in the poorest countries, but in larger (lower) middle-income countries with high income inequality, such as China, India and Brazil. We found that many of these countries serve high-end broadband customers in urban areas quite well.  However, poorer communities in urban and rural areas remain underserved because of seemingly weak demand, giving network operators limited incentive to invest in these markets. These mechanisms reinforce one another, creating an “access trap” by further limiting demand and discouraging new market entrants.

A4AI’s Policy & Regulatory Best Practices are the start of a consensus about how countries escape this access trap, but coordinating multiple efforts towards a beneficial public outcome remains a challenge. For example, policy makers can drive demand by making broadband relevant to people living in poor communities. Perhaps the best way to achieve this is to update the governance of critical public services, such as health, education and water, for the mobile broadband era.  Cloud-based solutions such as Form Hub can help teams more effectively deliver clean water and health services working across massive geographical areas. As public services drive people to adopt mobile broadband, the private sector will likely develop and offer services to meet the needs of new users, including poorer communities.

Further, policy makers can take steps to lower the cost, and thus the  risk, of investing in under-served communities.  Google’s Project Link is providing an open access fiber-optic network around Kampala, Uganda, to help Internet service providers reach end users with faster speeds at lower prices.  Policy makers can play a similar role by  building the Internet into other basic infrastructure. For example, fiber ducts can be built into roads, easing negotiations with local authorities for advanced services such as fiber to the home. Many developing countries also have extensive under-utilized spectrum, which can lead to much faster, much cheaper mobile broadband in rural communities.

We still have much to learn about which policies are most effective at which stages of a country’s Internet infrastructure development.  However, we know the stakes couldn’t be higher. McKinsey recently found that the Internet could contribute $300 billion to Africa’s economy by 2025.  The A4AI Affordability Report makes it clear that many countries still have a long way to go to realize these social and economic gains, but that governments can make decisions now to ensure a broadband-enabled future comes much more quickly.



  1. GeoTel says:

    Income causes a huge disparity when it comes to Internet service. This is one place where speed an issue; it is simply about being able to connect. As more communities have affordable connectivity on a global level, it’ll have a direct impact on the economy, business growth and all around efficiency and connectivity.

  2. paul says:

    I wonder how interesting it is that mobile broadband is cheaper than fixed almost everywhere.

  3. Eric Chiang says:

    Very interesting analysis Joshua. You make several good points.

    I would like to emphasize that that there are prerequisites that need to be in place before people can fully benefit from internet access. For example, you point out that most people who cannot afford broadband access are in China, India, and Brazil. But those are also the regions in the world where most illiterate adults are found. Not being able to read or write (or type) severely limits the benefits one would gain from internet access (ignoring for the moment communication through only pictures, or voice-recognition systems and text to speech software). I suppose there could be indirect benefits, such as if internet access facilitates the delivery of public services or maybe the presence of the internet could “motivate” illiterate adults to learn to read, but for the most part I would argue that improving literacy in these areas is more paramount than internet access.

    Similarly, there may be families and communities where food and medicine is more important than internet access (ideally you can have all of them, but there is always the issue of opportunity cost). You do acknowledge that internet access could result in better delivery of health, education, and water but I am skeptical how reliably this will actually be the case.

    Paul mentions the intriguing case of mobile broadband. There are several cases where existing infrastructure actually serves as an impediment to development – in the U.S. for example, sunk infrastructure costs in phones, gasoline-based cars, and credit cards have significantly hindered adoption of newer technologies. So I think that instead of replicating the broadband system that developed countries use, it is better to come up with more custom-tailored solutions, like the ones mentioned in Joshua’s second-to-last paragraph involving Project Link.

    Finally, one very narrow exception that should be kept in mind are specific cultural or religious groups who do NOT want internet access. In the U.S. the classic example is the Amish, who shun many (but not all) forms of technology. Practices may also vary for different Amish communities (some may allow the use of internet but not ownership of a computer; many allow cell phones but are concerned about internet access through cells). On the extreme side you also have uncontacted tribes (such as in India and South America) who do not want any contact with the outside (modern) world. Obviously, internet access for these groups would be unnecessary and unwarranted. But these are all corner-case exceptions and certainly not the norm.