Final Report

Final Report

World Telecommunication/ICT Indicators Meeting

Geneva, Switzerland

15-17 January 2003







Introduction

1

The third World Telecommunication/ICT Indicators Meeting—organized by the International Telecommunication

2

Union (ITU)—took place in Geneva, Switzerland, from 15-17 January 2003. There were 140 participants,

including 50 women, from 98 countries. A total of 17 international organizations were represented. The meeting

was divided into nine sessions, each headed by a coordinator.

The purpose of the meeting was to discuss topics related to the identification, definition, collection, processing,

dissemination and use of telecommunication/ICT indicators and to enhance collaboration between the different

parties involved. The meeting brought together entities responsible for telecommunication/ICT statistics and

analysis, including telecommunication ministries, regulators and operators; national statistical offices (NSO);

international organizations and researchers. Presentations were made on the following topics: ITU indicator

activities; international organization indicator activities; mobile indicators; experiences of telecom regulators and

National Statistical Offices; gender; indicator handbook; indicators in analysis; the Millennium Development

Goals (MDG) and the World Summit for the Information Society (WSIS) and measuring the digital divide.

Key issues emerging from the meeting included:

 Need to strengthen telecom/ICT statistics collection in developing countries. In this regard, it was noted

that loans and grants are available; that there are efforts to raise the visibility of indicators among high-

level policy makers; and the assistance of international organization in providing guidelines (e.g., sample

surveys, definitions, etc.) and in some assistance funding.

 Need for identification and definition of key telecom/ICT indicators to improve international comparability.

It was noted that the relevance of indicators might vary depending on the state of a country‘s economic

development. The effort of several international organizations in providing guidelines and definitions was

noted. The meeting also updated the indicators in the ITU Telecommunication/ICT Indicator Handbook

that can assist sector regulators in identifying the key indicators to collect.

 The measurement of the information society and the digital divide, particularly through composite

indexes, is difficult. More attention needs to be paid to the construction of the indexes. It was noted that

an expert group could provide input into the construction of such indexes by the ITU.

 A growing number of telecommunication regulators and national statistical offices are collecting ICT

statistics through industry questionnaires and household surveys.



ITU data processing

As the United Nations specialized agency for

telecommunications, the ITU is responsible for producing

statistics covering its sector. By means of an annual

questionnaire the ITU collects data for around 100 indicators

from more than 200 countries and territories. Data providers

include telecom ministries, regulators and operators. The first

presentation highlighted the reliance of the ITU on national

entities to collect ICT statistics. The presentation emphasized the

main challenges of the ITU‘s statistical work (see figure,

―Challenges‖).

The ITU‘s work further covers the analysis of

telecommunication/ICT developments, including national e-

readiness studies. That research has pointed to the need for additional indicators—beyond traditional telecom



1

Information and Communication Technology.

2

The programme and background documents are available at the following web site: http://www.itu.int/ITU-

D/ict/WICT02/index.html

1

statistics—to measure ICT knowledge. Introducing these four clusters

of indicators (see figure, ―Four groups of indicators‖) would help

identify knowledge-based barriers to Internet use. They could also be

used to estimate the potential market size in terms of Internet users.

Discussion concentrated around the topics of definitions and

comparability of indicators, timely collection and dissemination of

indicators, and cooperation between agencies. The definition and

reliability of indicators was a major issue and it was noted that

organizations involved with telecommunication/ICT indicators need to

agree on basic definitions. Data are sometimes not comparable

because the definitions of the indicators vary across countries. One

example is households, which may have different meanings depending on the country. Similar problems apply to

the definitions of Internet user or newspaper circulation. It was further noted that data on Internet hosts is

unreliable since a host cannot reliably be linked to a country. Traffic measurements need to be revised, since

more traffic is carried using Internet Protocol. Regarding ICT knowledge indicators, the suggested relationship

between Internet penetration and knowledge merits further study and the research should be extended to other

regions. It was further noted that it is difficult to analyze the impact of language on ICT take-up.

Suggestions were made on how the ITU might enhance the timeliness of the indicators. It was suggested that

instead of sending out one long questionnaire covering a range of topics, the ITU could collect its data in stages,

by the type of indicator. Dividing its questionnaire into topics would allow national counterparts to forward it to

the appropriate entity in charge of the data. It was also suggested that the ITU should quickly disseminate a

limited number of key indicators even if the data is not available for all countries. The limited number of

indicators would reduce the time needed to collect and disseminate the data and would allow the ITU to

continuously provide a set of key, up-to-date indicators. It was also suggested that the ITU could develop a

generic questionnaire or set of indicators that countries could compile.

It was noted that international agencies collecting statistics typically rely on national counterparts to provide the

necessary data. In some countries the data requested is not compiled. It is important for national agencies to

incorporate key ICT indicators to existing national surveys and questionnaires. Such indicators would include PC

and Internet penetration and use. It was also noted that statistical capacity in developing countries needs

strengthening in order to improve the availability of ICT data.



Cooperation

The session included three presentations by international and regional agencies on the overall theme of

coordination of international and regional statistics on telecommunication and ICT.

The World Bank, while not a specialist agency, nevertheless maintains

a key interest in ICT issues and policies. The presentation focused on

two key areas: the compilation and dissemination of secondary data on

key aspects of ICT and the use of the data to support the Bank‘s work

program and policy dialogue with member countries. The Bank does

not collect ICT data itself, rather it uses a number of international

sources, especially the ITU, to compile internationally comparable

indicators. These are published in a variety of forms, in particular as

part of the World

Development Indicators

data set. On the policy side,

the Global Information and

Communication

Technologies Department (GICT) brings together the public sector

advisory role and support for private sector investment. They carry out

research and analysis and as such are major users of ICT data.

Eurostat, the statistical agency of the European Commission,

provides support to member states and works to develop common

approaches to the collection and dissemination of data in many areas,

including ICT. Most recently they have been working to develop new

instruments to collect data on the information society. Data concerns include structural indicators as well as



2

developing a better understanding of the whole sector through national enterprise and household surveys.

These are carried out by national statistical agencies with technical support and funding from Eurostat. These

surveys are still quite new and methodological work is continuing. Current concerns include: survey methods;

specific aspects such as e-skills and e-security; measuring ICT expenditure and investment; and measurement

of broadband networks. Eurostat works closely with a number of partners in this area, especially ITU, OECD and

national statistical agencies.

OECD, in collaboration with its member states and other agencies has been involved in developing good

practice on the measurement of ICT and the information economy.

Through the Working Party on Indicators of the Information Society,

work has been done and published on defining the ICT sector,

measuring access to and use of information technologies, electronic

commerce and ICT in education and government. Model survey

modules have been developed for measuring ICT use in households

and by individuals and for measuring ICT use in businesses. These

modules are being used in a number of member states and OECD is

also involved in outreach with other countries.

The discussion following the three presentations covered a number of

different points, but three main themes emerged. The first was

concerned with the need to establish clear and consistent definitions

of terms and coverage, if the usefulness of international statistics is to be improved. A number of delegates were

concerned about the problems of translating and applying core concepts from one language into another and the

need for the international agencies to both recognize and address this issue. Another issue was the need for

agencies to adopt common terms and classifications, so that users could more easily compare data from one

source with another. The importance of this was recognized, although it was also pointed out that agencies have

their own internal demands for data and these have to be based on current structures and organization.

The second theme was concerned with capacity building and the importance of using the skills and expertise in

these agencies to help strengthen statistical systems and ICT reporting in developing countries. In this regard,

agencies have an important role to play, both in the provision of technical advice as well as helping to finance

capacity building programs.

The third set of questions focused on the use of ICT data and in particular the links between growth of the use of

information technology and economic and social development. All agencies need to justify the use of scarce

resources, and countries need support to make the case for governments and others to invest in better ICT data

systems.



Mobile Indicators

The session on mobile indicators shifted the focus of the conference from issues of data collection and

coordination to one of the specific information and communication

technologies being measured, namely mobile communications

networks and services. The ITU presented the theme ―Mobile

overtakes fixed: what consequence for telecommunication indicators‖.

It was noted that, during 2001, the number of mobile users worldwide

had overtaken the number of fixed lines. In some economies, such as

DPR Congo or Cambodia, mobile users already outnumber fixed lines

by a factor or up to eight. In the future, it might be useful to report the

indicator ―effective teledensity‖ which may be defined as mobile users

or fixed-lines, whichever is higher, per 100 inhabitants. Future

challenges for statisticians in this area would be to produce better

measures of mobile data users, mobile coverage, average revenue

per user and the mobile information society.

The ITU‘s Mobile/Internet Index was also presented, which had been published for the first time in the Internet

for a Mobile Generation report (September, 2002). The index covers 26 variables divided between three clusters

of infrastructure, market structure and usage. The index is intended to measure each economy‘s current status

of development in mobile communications and the Internet, and its likely future ability to adopt the ―mobile

Internet‖. As the industry matures, it should be possible to introduce more measures of actual mobile data use.





3

In discussion, it was proposed that the next edition of the index be

extended to include:

 New variables, such as broadband Internet charges,

percentage of Mobile Internet subscribers to the population,

volume of SMS/MMS traffic, charges for access to content etc;

 Better methodologies for dealing with missing data and for

developing weights to be assigned to clusters;

 Ways of distinguishing different quality (speed) of Internet

access.

Some of the existing variables in the index might be dropped or

substituted by others. Some argued that it would be better to increase the number of variables included in the

index, including using measures applicable to individual countries or regions, while others argued that the index

should be kept simple so that it would be comprehensible to the general public and could cover as many

economies as possible.

Other issues that arose in discussion included the likely impact of third generation (3G) mobile services and the

continuing rapid pace of technological change in the sector. Overall, it was concluded that continuing

benchmarking of mobile services would be essential to measuring the future competitiveness of nations.



Regulators and National Statistical Offices

Six representatives of different economies made presentations in this session: United Kingdom, Latvia, Togo,

Spain, Japan and Hong Kong, China. The representatives included telecommunication regulating agencies,

telecommunication operators, national statistical offices and ministries responsible for telecommunications or

statistics.

In their presentations, the representatives referred to a variety of issues on the subject of the session:

administrative and regulatory framework in the matter of collecting ICT data; collection, processing and diffusion

of ICT statistical information; division of these tasks between the different responsible organisms; and the

problems which they face in their work.

The presentations covered a variety of different ways in which ICT

indicators are collected. In the UK, the telecom regulator (OFTEL)

collects and disseminates quarterly reports on the mobile and fixed

line market. Although it has formal powers to request data from

operators, it has found that a collaborative approach works best.

OFTEL also carries out consumer market research that helps provide

demand side information but it is difficult to agree on consistent

consumer definitions. OFTEL also noted that the national statistical

agency also collects data such as turnover, employment and ISP

activity. Coordination within the country is important and there are

ways of improving data collection between OFTEL and the national

statistical agency.

In Latvia, the Central Statistical Bureau is responsible for telecom statistics. Market liberalization has made data

collection more difficult since operators consider their data to be

confidential so much depends on cooperation. Another issue is that

there are no uniform European-wide guidelines on methods and

variables. The CSB has carried out surveys on ICT usage in

enterprises and computerization in schools.

In Togo, the telecommunication operator collects and disseminates

telecommunication/ICT indicators. Although there is an industry

regulator and national statistical office, they lack the necessary

resources to carry out the task. It was noted that in the absence of

formal surveys on Internet use, the operator regularly checks with the

many Internet cafes to derive estimates regarding the number of

users in the country.





4

In Spain, the national statistics office recently completed a household ICT survey. The regulator as well as

EUROSTAT funded the survey. It was noted that there would soon be

new regulation that will oblige all EU countries to follow the same

rules for surveys, enhancing comparability. The survey included

questions about Internet use in the last 3 months, why people did not

use the Internet and what they used it for.

In Japan, there is a decentralized system for statistical activities with

a government wide coordination organ. Data are collected through

administrative records, surveys and information compiled by private

associations. Surveys have been carried out on ICT facilities, e-

commerce, ICT goods and services, time spent and purpose for the

use of the Internet, Internet penetration by age group and sex, etc.

In Hong Kong, China the national statistical offices carries out annual

surveys on household and business use of ICT. New questions for

2002 include security, the duration of playing online games and PC

knowledge. There are also monthly surveys of online retail sales,

household purchases on the Internet, and bi-annual IT manpower

survey. They have adopted broad definition on e-commerce provided

by the OECD.

During the discussion, differences in the compilation and statistical

treatment in the different countries were noted, making special

reference to the importance of the harmonization of indicators and their

definitions. The harmonization of ICT indicators and corresponding

definitions will allow a common methodology to be established to

obtain better comparisons about the situation of ICT in the world.

Also, it is necessary to show the different existing approaches in the collection of ICT data. One of the main

missions of telecom regulators is to analyze the ICT market and not in

principle to disseminate statistical information. However, the mandate

of the national statistical offices is to conduct statistical operations that

allow the diffusion and knowledge in general form and the detailed

disaggregating of this information. Nevertheless, as the regulating

agencies require statistical information for their task of market

analysis, they usually obtain from the operators. For that reason, it is

important to compare data gathered by the different organisms to

avoid discrepancies in the information. It would be advisable to

establish the necessary mechanisms for greater coordination and

collaboration between all the national and international bodies and

competent organisms in the elaboration of ICT indicators. Cooperation

among regulators in Europe was also mentioned, including the role of the European Body of Regulators in

harmonizing the information collected.



ICT and Gender

The collection of telecommunication/ICT indicators disaggregated by

sex is of great importance to document the digital divide by gender

and substantiate it for policy makers. This is especially important in

light of the Millennium Development Goals and the World Summit on

the Information Society, with its emphasis on bringing the benefits of

the information age to all, and the link between women‘s

empowerment and poverty alleviation. However, to date very few

telecom/ICT statistics are disaggregated by sex. In general, the

available gender statistics come from developed countries, which

have better gender equity ratings and from market research firms,

which are not always reliable or comparable. Additionally, the market

research firms are interested largely in middle and high-income countries. It is very difficult to obtain statistics



5

from developing countries where the gender digital divide is felt to be

most profound. Some suggestions were made on indicators that

would be desirable to collect (see slide, ‗Which gender statistics

needed?) and examples presented of good national practices in this

area.

Gender statistics are especially difficult to obtain from countries in

view of the fact that many of them do not even collect ICT statistics.

The ITU‘s only gendered statistic is national telecommunications staff

and here the results have not been entirely successful (see slide, ‗‘ITU

experience‘). The ITU was asked to take the lead in requesting

countries to provide ICT statistics disaggregated by sex.

In the discussion several countries felt that available gender statistics did not accurately represent the situation

of women in ICT and telecommunication in their countries. It was suggested that data may be available in

household surveys even though it is not always published.



Telecom Indicator Handbook

The rise of mobile and the Internet and the convergence of broadcasting, computing and telecommunications

call for the introduction of new indicators. The ITU described the

Telecommunication Indicator Handbook, which identifies and

provides definitions for key indicators (see slide). A revised list of

globally comparable telecommunication/ICT indicators was presented

and discussed at length. The rationale behind this exercise is the

reality that the ITU cannot collect every possible ICT indicator and

there is a need for a globally agreed list and definitions to enhance

comparability and guide data collectors. Special emphasis was

placed on the development of community access indicators and

household-level statistics. A supporting document was also circulated

by the ITU, providing examples of how specific countries have used

some of the new indicators and how they are defined. A number of

new indicators were adopted and several existing indicators deleted.

The list of indicators will be circulated to participants who will have an opportunity to provide feedback and

propose additional indicators before the handbook is finalized.



Research and analysis

One of the main purposes for the collection of telecommunication/ICT

indicators is their importance for research and analyses to understand

underlying relationships and trends. This session provided some

examples of work from infrastructure to application to analysis.

Statistics Canada uses data on telecommunications revenues to

examine market concentration in the context of a changing regulatory

environment. Findings were presented based on indicators of

concentration (Hirfindahl-Hirschman Indices – HHIs) for the local

telephony, long distance, data and private line services markets (see

slide on ‗Indicators of

concentration). Markets

were found to be least

concentrated in mobile

services and most concentrated in local wireline services, where

deregulation occurred only very recently. Canada‘s three largest

provinces have moved the farthest away from monopoly, while

smaller provinces were found to be the most highly concentrated.

The impacts of ICT developments on consumers were presented on

the topic of ‗measuring the affordability of telecommunications‘.

Aspects of the definition of affordability, as well as the importance of

measuring affordability were introduced. The presentation highlighted

the ongoing tensions between access and affordability, particularly

6

since the liberalization of telecommunications services. An income-based approach to measuring affordability

could determine how much of their income consumers have to pay for telecommunications usage. Some

common patterns of household expenditures on telecommunications were observed across various countries

(see slide), which could be applied to countries with little or no data in

this area. Indicators of teleaffordability were proposed, in order to

allow for measures of relative affordability between countries (ratio of

telecommunications expenditures as a percentage of GDP and

teledensity). Long-term objectives could include examining

teledensity and the percentage of household expenditures on

telecommunications by household income quintiles, as well as

regional patterns.

A third presentation covered the use of indicators to forecast potential

demand for telecommunication services. Models for evaluating

potential subscribers, the impact of mobile cellular and the

development of rural networks were presented. The use of these

indicators to evaluate the overall potential of subscribers and networks is critical if the development of the

infrastructure and the corresponding network planning activities are to be assessed.



Millennium Development Goals/World Summit of the Information Society

Representatives from the UNESCO‘s Institute for Statistics (UIS) outlined the institute‘s main work and the

barriers to collecting quality data. The communications statistics

programme was presented, including its project B@bel, which is

about promoting multilingualism on the Internet. The UIS is

currently preparing a statistical report for the WSIS where the

main focus will be on developing countries. Emphasis will be

given to areas where UNESCO can offer added valve with respect

to what other agencies are doing such as in areas of access and

content as well as benefits of ICTs – through considering issues

related to language, freedom of expression and preservation of

cultural diversity.

The ITU summarized its contributions towards monitoring the UN

Millennium Development Goals (MDG). Given its specific area of

competence, the ITU‘s focus is mainly on Goal 8: ―Develop a global partnership for development‖, which in turn

has eight targets. One of these targets (# 18) focuses on making

available the benefits on new technologies, particularly ICTs. With a

view to monitoring this target, ITU provides three indicators that it

collects: total telephone penetration, Internet penetration and PC

penetration. The indicators are analyzed through the broad

classification of developed and developing countries (see slide). The

scope of ITU indicators being used to monitor the MDGs may be

narrow, but the number was limited to three. It should be noted that

the goals and targets were determined by a group of international

bodies. Indicators were chosen based on the ability of agencies to

provide a long time series for most countries.

A possible way of measuring the state of and potential for an

information society was presented by the ITU‘s Strategy and Policy Unit (SPU), through their information society

index. The index uses existing related indices (see slide), such as the

UNDP‘s Human Poverty Index and has identified five broad

categories, which would contribute to the index. These go beyond

basic infrastructure and connectivity and cover the areas usage and

access, market and policy, social and education, and economic. One

of the suggestions made was to include cyber security and e-

commerce indicators, important aspects of an information society.

Others suggested to include broadcasting indicators to reflect modes

in developing countries.





7

The discussion further highlighted the weaknesses of indexes,

including the fact that rankings tend to be closely linked to

GDP/income rankings. Rather than presenting absolute rankings for

countries, indices could rank countries compared to their level of

income/GDP to highlight those that are doing particularly well or bad

within their income category. This would increase the significance of

any given index and uncover strengths and weaknesses of a country.

Indexes that are based on existing data or that use existing indexes,

risk to simply repeat the same results without actually providing any

new conclusions. This was highlighted in a presentation on the

general shortcomings of indices and their often subjective nature.

This would include the choice of indicators used, for example the

exclusion of a key indicator such as broadband subscription.

Including income levels per capita is similarly distorting since it would result in punishing countries that are doing

well in spite of their low incomes. The Republic of Korea and Switzerland were used as examples to show that

single composite indices can be misleading, in terms of both the indicators that are included and those that are

left out. This also highlighted several other problem related to key indicators:

 Using Internet hosts as an indicators is increasingly meaningless since a host cannot always be linked to

a geographic region

 It is important to distinguish between post-paid and pre-paid mobile subscribers since pre-paid cards are

a relatively unreliable indicator

 When comparing countries, per capita income should be adjusted to purchasing power parity (PPP)

 Data on the number of main telephone lines are not collected in the same way across countries and may

result in misleading interpretation



Measuring the Digital Divide

The focus on the digital divide has called upon researchers to find methods

to measure and define the divide. One method is the Mosaic Group‘s

Global Diffusion of the Internet framework. This framework can be used to

evaluate and to study the evolution of the Internet in a country over time.

The Mosaic framework is based on six different dimensions and is flexible

in considering the early stages of diffusion of the Internet. It has been

applied in many low and middle-income countries. The evolution of the

pervasiveness and geographic dispersion dimensions in particular highlight

the balance of supply and demand factors influencing the state of the

Internet. Although the framework needs to be updated, more data needs to

be made available, and more case studies need to be carried out, it has

revealed interesting results with regards to the development of the Internet.

Its qualitative methodology reveals the causes for the dimensions values in

order to make country-specific recommendations.

Another effort to monitor the digital divide is being carried out through the

Orbicom project, based on the framework of Infostate, which is measured

by looking at a country‘s level of Info-use (consumption) and Info-density

(productive capacity). The pilot project, which was carried out in nine

countries (see slide, ―Monitoring the Digital Divide‖) is to be extended to

more than 50 countries, the results of which will be presented at WSIS, at

the end of 2003. So far the results show that the digital divide, although it

is narrowing, is great. More data needs to be collected, especially in the

area of skills.

The United Nations Conference for Trade and Development (UNCTAD)

work on measuring the digital divide are based on Gini coefficients of

inequality. Their findings reveal high levels of inequality in the distribution

of ICT hardware and access to the Internet that are twice as high as average levels of income inequality. The

Gini coefficients reflect the relative age of the technologies, with greater inequality observed for more recently



8

introduced technologies. Mobiles are a notable exception to this rule. The analysis of whether the digital divide is

growing or shrinking is complex. UNCTDAD research suggests relatively

little change over time in the levels of inequality of these distributions,

with small, incremental reductions from their highly unequal levels.

A few participants expressed their concern regarding the emphasis on

the Internet and related applications and technologies. Throughout the

meeting it became clear that the different level of ICT development

across countries impacts the type of indicators that are considered to be

important. While some countries stress the importance of measuring

newer technologies, including broadband, mobile Internet and security

issues, other countries emphasized the importance of broadcasting. The

limited scientific precision of indices and the need to put them into

perspective were highlighted. It was proposed that a Group of Experts be

created to examine the methodology behind proposed ITU ICT indexes.









9