Bridging the Broadband Divide:
Access and Demand in the Developing World
Dr. Heather E. Hudson1
Abstract
Wireless growth in developing countries has been explosive in the past five years.
More than half of all telephone subscribers in the developing regions of the Asia/Pacific
and Africa use wireless; for most, their cellphone is their first and only phone. The shift
to IP-based services offers enormous potential for lower cost voice services (using VOIP)
and IP-based video and other multimedia. Broadband is a key requirement for access to
these new services. However, Internet access is still very limited, and broadband is still
unavailable and/or unaffordable in many developing countries in the region. Where
Internet service is available, it is typically priced much higher than in wealthier countries,
making it beyond the means of most potential users. These conditions severely hamper
exploitation of the Internet’s potential for social and economic applications – in
commerce, education, health and other public services.
This paper focuses on strategies to increase Internet, and particularly broadband,
access in low and lower middle income countries (as defined by World Bank indicators).
It includes analysis of the current status of Internet, broadband and wireless (mobile)
access in developing regions, including wireless growth, management data on fixed
networks, and pricing and availability of Internet services. It then examines lessons from
the growth of wireless including the impact of competition on innovative services and
pricing, the enormous pent-up demand for communication services, and the irrelevance
of past regulatory distinctions. The paper then proposes strategies to increase broadband
Internet investment and access through such means as limiting exclusivity periods in
licenses, allowing resale, facilitating use of broadband wireless technologies, reducing
local barriers such as permits and fees, and using incentives and targeted subsidies such
as universal service and access funds.
1. Access in Developing Regions
1.1. Digital Access Index (DAI)
During the past decade, the Internet has changed the way people around the world
communicate, through e-mail and through the Worldwide Web for accessing and sharing
information in the form of text and graphics, and increasingly audio and video. The past
decade has also introduced personal wireless communications, so that there are now more
than 1.3 billion wireless telephone subscribers around the world. Disparities in access to
telecommunications, and particularly to the Internet, have become known as the “digital
divide.”
1
Fulbright Policy Research Chair, Carleton University, Ottawa, Canada (fall 2009) and Incoming Director,
Institute for Social and Economic Research (ISER), University of Alaska Anchorage (as of Jan 2010).
Contact: heatherehudson@gmail.com .
1
The ITU has developed a Digital Access Index (DAI),1 ranking countries by
variables in five categories: availability of infrastructure, affordability of access,
educational level, quality of ICT services, and Internet usage. Chart 1 shows the DAI
index for countries in the Asia-Pacific region. Typically, wealthier countries rank higher
than poorer countries on this index. Within the poorer (low and lower-middle income
countries) there are some exceptions, such as Fiji and the Philippines, where are
education levels are relatively high. The poorest countries (Bhutan, Laos, Cambodia,
Myanmar, Bangladesh and Nepal) rank low on all components of the index.
Chart 1: ITU Digital Access Index: Asia-Pacific
Solomon Islands
0.9
Vanuatu
Papua New Guinea
Samoa
0.8
Fiji
Bhutan
0.7
Lao PDR
Cambodia
0.6
Myanmar
Bangladesh
Nepal
0.5
Pakistan
Vietnam
0.4
India
Indonesia
Mongolia
0.3
Sri Lanka
China
0.2
Maldives
Philippines
0.1
Thailand
Malaysia
Macao
0
Japan
1
Pacific Islands Asia Low /Low er Middle Incom e Asia High Incom e
Singapore
Hong Kong
Taiw an
Korea
2
1.2. Internet and Broadband Access
The variation in Internet access per 100 population is much more extreme than
would be expected from the DAI. In the Asia/Pacific, fewer than one person in 100 in the
poorest countries has access to the Internet, while in the high income countries such as
Singapore and Taiwan, more than 50 people per 100 population access the Internet. See
Chart 2.
Chart 2: Internet Access per 100 Inhabitants
Vanuatu
70
Samoa
Micronesia
Cambodia
60
Myanmar
Bangladesh
Lao PDR
Bhutan
50
Nepal
Maldives
Indonesia
40
Sri Lanka
India
Pakistan
30
Philippines
Mongolia
Vietnam
20
Thailand
China
Malaysia
10
Macao
Korea
Japan
Hong Kong
0
1
Pacific Islands Asia Low /Low er Middle Incom e Asia High Incom e
Singapore
Taiw an
Globally, in the poorest countries, only about 3 people in 100 use the Internet,
primarily in urban areas. Internet access in rural areas is generally limited, and broadband
virtually nonexistent. In contrast, more than 40 people per 100 in the Americas, Europe,
and Australia and New Zealand are online. The contrast in broadband access is even
more stark, with only one person in 500 subscribing to broadband in the poorest
countries. See Chart 3.2
3
Chart 3: Internet and Broadband Usage
50.0
44.8
45.0
42.8
41.4
40.0
35.0
31.7
30.0
Least Developed Countries
25.0
Africa
22.0
Asia
20.4
20.0
Americas
17.4
16.6
Oceania
14.0
15.0
Europe
12.8
World
10.3
10.0
8.5
6.9
5.4
5.0
5.3
3.4
3.1
1.2
0.4
0.20.2
0.0
Internet
subscribers/100
Internet Users/100
Broadband
subscribers/100
1.3. Internet and Broadband Pricing
One of the reasons for lower Internet use is likely to be the price of Internet
access. In the low income countries of the Asia/Pacific, broadband (where available)
costs about 170 percent of GNI per capita, while in subSaharan Africa, broadband
subscriptions cost about 450 percent of GNI per capita. Yet broadband access is not only
absolutely cheaper (about 1/10 the price) in the wealthiest Asian countries, but
dramatically more affordable; in Japan, Taiwan, Singapore and Korea, broadband access
costs less than 1 percent of per capita monthly income. See Chart 4.
4
Chart 4:
Broadband Pricing and Affordability:
Asia-Pacific and Africa
400
500
450
350
400
USD per month
350
250
300
250
200
200
150
150
Percent of GNI per capita
300
100
100
50
50
0
0
Asia-Pac
Low Incom e
Asia-Pac
Low er
Middle
Asia-Pac
UpperMiddle and
High Incom e
North Africa Sub Saharan
Africa
2. International Bandwidth
Bandwidth is limited not only within most developing countries, but among
countries in their subregion and linking them with the rest of the world. The recent
investment in submarine fiber networks has resulted in international connectivity of more
than 2.3 megabits per person in wealthier countries of the Asia/Pacific, compared to
about 41 bits per person in poor countries in that region, and only 9 bits per person in
subSaharan Africa. See Chart 5. Thus, in these countries, access to the Internet is much
more difficult and generally much more costly than for most industrialized nations.
(Connectivity in east Africa is likely to improve significantly with the newly installed
SEACOM cable, and with other submarine cables to be completed in the next few years.)
5
Chart 5:
International Bandwidth: Asia-Pacific and Africa
2500
2351.6
2000
Bits per Inhabitant
Asia-Pac Low Income
1500
Asia-Pac Lower Middle
Asia-Pac Upper-Middle
and High Income
North Africa
1000
Sub Saharan Africa
500
208.4
155.6
40.9
9.4
0
1
3. The Growth of Wireless
With the introduction of competition in mobile services in developing countries
during the past decade, access to voice telephony has increased exponentially: “The first
billion telephone subscriptions, reached by the end of 1997, took more than a century to
accomplish. It took just four years to add the second billion, three years for the third
billion and two years for the fourth billion.”3 In 1997, the teledensity for low income
countries was just 1.5 subscriptions per 100 inhabitants; by 2007, teledensity had
increased to 23.9 per 100 inhabitants, of which more than 95 percent were mobile
subscribers (some with multiple subscriptions or SIM cards). The OECD estimates that
between 1996 and 2006 annual investment in telecommunication infrastructure more than
6
doubled in low income countries, from $4.4 billion to $9.6 billion per annum, while
equipment costs were falling and network infrastructure was shifting to less expensive
wireless-based systems. 4
For most subscribers in these countries, their cellphone is their first and only
phone; in 22 Asian developing countries, more than 70 percent of all phone subscribers
are mobile subscribers. See chart 7. The contribution of wireless to voice access can be
seen in Chart 8, which shows the percentage of wireless subscribers of all subscribers in
low and lower middle income countries in the region. In more than half of these
countries, more than 50 percent of all subscribers are wireless customers. In Thailand,
the Philippines and Bangladesh, and several subSaharan African countries, they are more
than 90 percent.
Chart 6:
Cambodia
Mobile Subscribers/All Subscribers: Asia
Bangladesh
Philippines
Lao P.D.R.
Oman
120
Maldives
Kuw ait
Malaysia
Bahrain
Indonesia
United Arab Emirates
100
Mongolia
Saudi Arabia
Qatar
Palestine
Macao, China
Sri Lanka
80
Israel
Jordan
Yemen
Pakistan
Singapore
Thailand
60
Hong Kong, China
Georgia
Azerbaijan
Kazakhstan
Taiw an, China
Korea (Rep.)
40
Japan
India
Kyrgyzstan
Bhutan
China
Syria
20
Lebanon
Tajikistan
Viet Nam
Nepal
Armenia
Uzbekistan
0
Myanmar
Countries
Iran (I.R.)
Turkmenistan
7
Yet these wireless services provide very limited bandwidth, typically a maximum
of 9.6 kilobits per second, and often as little as 2.4 kilobits per second. They can be used
for text messaging and simple e-mail, but are not really suitable for Internet access. They
may, however, provide lessons about how to extend broadband services in developing
regions.
4. Lessons from the Wireless Explosion
4.1. The Impact of Competition
What has driven the explosive growth of wireless in developing regions, despite
the anemic growth of fixed networks? The key driver is competition. Wireless
competition has resulted in innovative pricing and service offerings. Rechargeable smart
cards make phone service accessible to people without bank accounts or credit histories.
Cheap messaging can substitute for many e-mail functions; for example, Filipino wireless
subscribers send more SMS (short message service) messages than texters in any other
country. Low and lower middle income countries with at least two wireless providers
have significantly higher wireless growth rates than countries with a wireless monopoly.
Countries with wireless competition are not wealthier than low growth rate countries; in
fact, both categories include some of the world’s poorest countries.5 See Chart 7.
Chart 7: Wireless Growth Rates in Low and Lower Middle Income Countries:
Monopoly vs. Competition
Competitive Wireless (n=79)
Monopoly Wireless (n=15)
4,000
7,000
3,500
6,000
3,000
2,500
GDP per cap $US$
GDP/cap US$
5,000
2,000
4,000
3,000
1,500
2,000
1,000
1,000
500
0
0
0
10
20
30
40
50
60
0
mobile subs/100
5
10
15
Mobile subs/100
8
20
25
4.2. Demand Greater that Assumed
In designing networks and projecting revenues, planners often assume that there is
little demand for telecommunications in developing regions. The take-up of wireless
services in many developing countries has also demonstrated that there is significant
pent-up demand for telecommunications services, even among relatively low income
users. In Rwanda, people spend up to 12 percent of their income on communications; in
Tanzania, they average 29 percent.6 See Chart 8. While demand for broadband is likely to
be more limited, it is certainly not negligible. For example, entrepreneurs may want
Internet access to order parts and supplies, check on international prices, and arrange
transport of their produce to foreign markets. Some artists and craftspeople use websites
to market their wares globally; SMEs and NGOs use VOIP to communicate with clients
and staff. There may also be significant demand from government agencies and NGOs
operating in rural areas to administer health care services, schools, other social services,
and development projects, particularly in rural areas.
Chart 8:
GDP per capita and spending on
telecommunications
14%
12%
As a perc entage of
inc ome (US$ PPP)
10%
Monthly GDP per
c apita (US$ PPP)
900
800
12%
700
600
8%
8%
500
6%
400
6%
4%
4%
300
4%
4%
2%
200
100
0%
0
Botswana Ethiopia
Namibia
Rwanda
South
Afric a
Uganda
Source: ITU
4.3. Old Distinctions No Longer Relevant
Classifications and distinctions that once were useful may no longer be relevant.
Regulators typically issue separate licenses and approve separate tariff structures for
fixed and mobile services, yet these distinctions have become blurred. Mobile telephone
service was originally designed for communication while in vehicles; however, mobile
systems are now personal, and often be considered a substitute for fixed network
9
connections. As noted above, in many developing countries, wireless has become the first
and only service for many customers who never before had access to a telephone.
Eliminating these licensing distinctions may accelerate access.
As with fixed and mobile, the distinction between voice and data no longer makes
sense; bits are bits, and can be used to transmit anything. Yet in many developing
countries, voice communication is still considered a monopoly service. Incumbent
operators may see VOIP as a threat to their commercial viability, but lack of VOIP and
other cheap broadband services may be a threat to the competitiveness of several
developing countries in the region. India has become a leader in offshoring information
work not only because it has an educated English-speaking workforce, but because the
extremely low cost of communicating between Indian-based businesses and their
customers in the industrialized world. China has built its own domestic IP networks,
which customers can access from any telephone using a prepaid card. Even in Bhutan,
entrepreneurs are starting call centers to serve European clients using VOIP.7
5. Strategies to Increase Investment and Growth
5.1. Limiting Periods of Exclusivity
In a liberalized environment, the length and terms of operator licenses can impact
the pace of growth of networks and services. Regulators typically face choices
concerning how long to protect incumbents to enable them to prepare for competition,
and how long to grant periods of exclusivity or other concessions to new operators to
minimize investment risk. Yet exclusivity and long time periods may be the wrong
variables to focus on if the goal is to increase availability and affordability of
telecommunications services. Instead, investors cite a transparent regulatory environment
with a “level playing field” for all competitors and enforcement of the rules as key to
their assessment of risk. It is highly unlikely that fixed line providers will have an
incentive to roll out broadband services beyond large businesses and some upscale
residential areas if they see no near term threat to their monopoly. Some jurisdictions8
have negotiated terminations of exclusivity periods with monopoly operators in order to
enable their economies to benefit from competition in the telecommunications sector.
5.2. Extending Access through Resale
Authorization of resale of local as well as long distance and other services can
create incentives to meet pent-up demand even if network competition has not yet been
introduced. Franchised payphones can be introduced in developing countries in order to
involve entrepreneurs where the operator has not yet been privatized and/or liberalized.
Indonesia’s franchised call offices known as Wartels (Warung Telekomunikasi), operated
by small entrepreneurs, generate more than $9,000 per line, about 10 times more than
Telkom’s average revenue per line. In Bangladesh, Grameen Phone has rented cellphones
to rural women who provide portable payphone service on foot or bicycle to their
communities. Franchised telephone booths operate in several African countries; in
Senegal, private phone shops average four times the revenue of those operated by the
national carrier. In the Philippines, entrepreneurial wireless subscribers resell small
10
increments of time transferring them to “pay as you go” customers who cannot afford the
price of a phone card.9
Resale of network services can also reduce prices to customers. Most
interexchange carriers in industrialized countries are actually resellers that lease capacity
in bulk from facilities-based providers and repackage for individual and business
customers, offering discounts based on calling volume, communities of interest, time of
day and other calling variables. Similar strategies can be used to resell broadband when
networks that are upgradeable (such as for DSL) or that have excess capacity (such as
optical fiber or satellites) are available.
5.3. Legalizing Bypass
Facilities that are not offered by an incumbent wireline operator may be the least
expensive means to extend access. For example, a VSAT may be an ideal solution to
bring high speed Internet access to a rural school or telecenter. A WiFi “hotspot” may be
a low cost means of providing broadband to a village or neighborhood. However, in some
developing countries, even if the wireline provider does not provide broadband services
in the area, or possibly does not even serve the area, such as VSAT connection would be
considered illegal bypass.
Many monopoly operators claim that bypassing their networks effectively siphons
off revenues that they need to expand their networks, which would also probably create
more jobs. However, the relationship is not so simple. As noted above, without
competition, there is likely to be little incentive to roll out broadband, to choose the most
cost-effective technologies where broadband is deployed, and to price broadband services
reasonably. Thus, policy makers will not further the goal of extending access to
affordable broadband by preserving wireline monopolies.
Protecting dominant carriers may also hinder economic growth. For example, a
West African Internet service provider pointed out that he needed relatively inexpensive
international connection to the Internet in order to provide affordable Internet access for
his customers. By using bypass, he created new jobs in value-added services as an
Internet provider, as well as providing an important information resource for economic
development of the country.10
Prohibiting VOIP traffic may also hinder economic growth. Countries which
allow VOIP make it possible for users to make digital voice calls at a small fraction of
tariffed rates. The benefits can be significant for developing countries seeking to attract
information services. As noted above, new call center businesses in Bhutan are creating
jobs because they can use VOIP over leased circuits instead of retail services on the
incumbent’s network. However, in several countries in the region including Indonesia,
Laos, Nepal11 and Thailand individual VOIP calls are not legal.12
5.4. Reducing Local Barriers
Some developing countries set high duties on imported equipment including
computers and components for telecommunications networks. These duties may generate
revenue for the government, but they increase the cost of network facilities and end user
11
equipment, and thus increase the cost of access for its citizens. For example, in Tanzania,
fewer than 1 person in 100 has a personal computer.13 The economic benefits of having
available and affordable access are likely to outweigh substantially the value of such fees
and duties.
Governments may also inhibit network build out by making it difficult for
operators to secure permits for rights of way or use of existing poles or conduits, or by
charging fees for such permits or other services that place a significant financial burden
on the operator. While such fees may also be attractive sources of revenue for the
government, they may have the effect of delaying access to the Internet for its residents.
5.5. Universal Service and Access Funds
Funds to subsidize installation of telecommunications infrastructure and in some
cases, utilization of telecommunications services, have been adopted in many developing
countries. The goal is to extend coverage or accelerate investment in infrastructure in
areas that are considered unattractive for commercial operators. While universal service
funds generally prioritize coverage for regions and/or populations without access to voice
telephony (traditionally provided over fixed lines and public pay phones),14 many
countries are now expanding definitions to include mobile telephony and Internet access.
Examples include Chile, Ghana, India, Indonesia, Mongolia, Morocco, Mozambique,
Nicaragua, Nigeria, Paraguay, Romania, South Africa, Venezuela, Vietnam.15
6. Community Access
The concept of universal access continues to evolve, both in terms of services that
should be universally included and in our understanding of access, which includes
availability, affordability and reliability. Universal access should therefore be considered
a dynamic concept with a set of moving targets.
Community access was a benchmark commonly applied for voice services in
developing countries: publicly available payphones were to be available in each village or
neighborhood of towns and cities. This approach is now being extended to the Internet, as
many countries are extending public access to the Internet through telecentres, libraries,
post offices, and kiosks.
As demand grows, entrepreneurs, small businesses, NGOs, and some residents
may want direct Internet access. One solution is for the community access point, such as
a telecentre, to become an ISP. The telecentre in Timbuktu, Mali, became an ISP to serve
NGOs, entrepreneurs, and local government agencies that could afford a computer and
access charges.16 However, in some countries, it is not legal for a telecentre to also
operate as an ISP.
Community wireless is another means of extending access to neighborhoods,
campuses and villages. In Alaska, the FCC has issued a waiver to allow the subsidized
Internet service for schools to be accessible to the community. One of the long distance
carriers is now providing Internet coverage to villages using WiFi, extending service
from the village satellite earth station. The price is guaranteed not to exceed what would
be charged for broadband access in urban Alaska.17 Other towns and cities are
12
encouraging installation of municipal wireless systems that can provide free or very low
cost access to broadband services. These approaches could be models for other rural and
developing regions where commercial ISP services are not available.18
7. Conclusion: Getting to Broadband in the Developing World
To compete in the global economy, developing regions will need affordable and
reliable communications services, with sufficient bandwidth to provide Internet access
and other IP-based services. The above lessons from the wireless experience and from the
expansion of telecommunications networks in general could help to unlock the potential
benefits of broadband and IP services for the developing world.
References
1
“ITU Digital Access Index: World’s First Global ICT Ranking.” See http://www.itu.int/ITUD/ict/dai/index.html.
2
Data for this and following charts are from ITU, World Telecommunication Development
Report 2008, Geneva, December 2008.
Paltridge, Sam. “Network Externality Premiums and International Telecommunication
Traffic Exchange.” DSTI/ICCP/CISP(2008)4/FINAL. Paris: OECD, April 15, 2009, p. 8.
3
4
. Paltridge, OECD, 2009. Data derived from the ITU and UNCTAD.
5
See Hudson, Heather E. From Rural Village to Global Village: Telecommunications for
Development in the Information Age. Mahwah, NJ: Erlbaum, 2006.
Hudson, Heather E. “Using Data for Policy Development: Designing Universal Service Fund
Implementation for Tanzania.” Paper presented at the New America Foundation, Washington,
DC, September 2009.
6
7
Interview with TST Ltd. Managers by author in Paro, Bhutan, June 2007. See also
http://www.kuenselonline.com/modules.php?name=News&file=article&sid=8433
8
For example, Hong Kong, Singapore, and India.
13
9
See www.smart.com.ph/smart
10
Personal communication.
Presentation by the author to Nepal Telecommunications Authority, Prime Minister’s
Commission on IT, and telecommunications operators, Kathmandu, Nepal, June 2007.
11
12
Source: ITU Regulatory Database. See www.itu.int/itu-D.
Hudson, Heather E. “Using Data for Policy Development: Designing Universal Service Fund
Implementation for Tanzania.” Paper presented at the New America Foundation, Washington,
DC, September 2009.
13
InfoDev. “Universal Access and Service.” Module 4 of ICT Regulation Toolkit, December
2008. See www.ictregulationtoolkit.org.
14
Hudson, Heather E. “Universal Service Funds: Accelerators or Anachronisms?” Paper
presented at the Telecommunications Policy Research Conference, Arlington, VA, September
2009.
15
16
Field research by the author in Timbuktu, Mali. See also Hudson, Heather E. From Rural
Village to Global Village: Telecommunications for Development in the Information Age.
Mahwah, NJ: Erlbaum, 2006.
Hudson, Heather E. “From Northern Village to Global Village: Lessons in E-Health and EEducation from the Arctic.” Paper presented at the World Summit on the Information Society,
Tunis, November 2005.
17
See Hudson, Heather E. “Municipal Wireless Broadband: Lessons from San Francisco and
Silicon Valley.” Telematics and Informatics, Februrary 2009.
18
14