UAS Programme Development and Prioritization

This chapter presents and discusses some of the main steps to develop a universal access and service (UAS) programme, including an ICT sector review, demand analysis, subsidy estimation, and also a discussion on measuring economic impact and benefits and how this may help to prioritize programme components and individual projects.

Section 6.1 expands on the sector analysis already outlined in the UAS policy chapter (Section 4.2.1) which helps to ascertain the country’s current UAS status and answer the question: What areas and population groups do not yet have affordable access to ICT services? This results in a detailed identification of service gaps in the country; the section then looks at quantifying the demand for various ICT services in those areas that are unserved, as well as qualifying the demand, e.g., at what price point and for which level of service is there a need or a demand? Both desk-based and field demand study methodologies are presented. The purpose of this step is to:

  • Identify which areas (regions, provinces, districts etc.,) of the country are to be included into the UAS programme;
  • Identify which ICT services are required in each of these areas; and
  • Quantify the demand for those ICT services, to be used to model potential revenues, which in turn helps to prioritize projects within the UAS programme.

With the groundwork laid in identifying unserved communities and quantifying their demand, Section 6.2 then looks at determining how much it will cost to provide UAS services. It further looks at what subsidies are required by comparing revenues against costs. Approaches to modelling costs, assessing viability and estimating subsidy requirements are also presented.

Section 6.3. discusses the status of economic impact analyses, measurements and findings, separate for communications projects and broadband Internet & ICT service development. While these broader economic impact studies may not be well suited for the purpose of UAS programme development or prioritization, their understanding and knowledge of the latest research in this area provides nevertheless crucial information for any UAS programme planner.

Section 6.4 then looks at practical approaches to project prioritization in the context of UAS programme development.

Given the dynamic nature of the communications market, and the increasing number of players, collection and analysis of the data required to perform some of the above analyses may be quite challenging,. A UAS programme developed two years ago, for example, may already be partially obsolete.

It is therefore recommended that regulators have a designated UAS department that regularly monitors, collects and analyze data on an annual basis. In addition, regulators need to maintain close and continued co-operation with the ICT industry in developing and updating UAS programmes.

Reference Documents

4.6.1 ICT Sector Analysis and Assessing Demand

The first step towards developing a universal access and service (UAS) programme is to determine the country’s current UAS status. The country’s unserved communities and/or regions need to be identified and their demand for services assessed. This can be accomplished through an ICT sector review, described in Section 6.1.1, which includes policy issues and market players, stakeholder consultations, near-future plans (e.g., one to three years) with regard to existing infrastructure, and that may be complemented with a demand study. This enables the regulator to identify and design feasible UAS targets that can built upon the existing strength and potential of a country and leverage the capacities and synergies of market players and stakeholders already active in the market. The ICT sector review may allow other development players to identify effective strategies that might not need high-end technology and bandwidth. For example, if a country has an active health-related NGO that wants to reach the rural population, it might use a combination of an FM radio station for information dissemination and add call centre and information retrieval services for call-in and ‘question and answer’ services.

Uganda’s success at lowering the rate of HIV/AIDS infection within the country is partly attributed to a campaign that uses radio to educate and raise awareness of HIV/AIDS issues, particularly with teenagers and other at-risk citizens, via radio-talk shows [1]. When implementing a UAS strategy, governments or regulators like to have a demand study undertaken that identifies consumer needs, preferences and demand for service. This can also be used to highlight the social and economic impacts of new service provision into formerly underserved areas.

For telephony services, the simplest initial desk-based demand estimate can be made in terms of the call-generating or purchasing power of specific areas or administrative units. The most common approach for estimating demand is establishing the individual household expenditure on communications, as described in Section 6.1.2.

In addition, for telephony, there is also demand and revenue potential for incoming calls to as of yet unserved areas which should be included in this desk-based demand estimation, as described in Section 6.1.3.For Internet services, an initial demand scenario can be created based on estimates due to the presence of administrative, business or institutional structures, as illustrated in Section 6.1.4.

The demand scenarios can be based on different assumptions, creating conservative, moderate and optimistic demand scenarios, and include a sensitivity analysis. However, in order to confirm or refine these desk-based demand estimates, it is recommend to undertake field demand studies where actual users and potentianl users of ICT services are interviewed about their needs, affordability and other topics, as explained in Section 6.1.5. ICT sector review

To prepare for an ICT sector review, which is required both for policy and programme development, statistical and other country data (e.g. socio-economic, cultural) needs to be collected and compiled to summarize the country’s geographic, political, economic, social and cultural characteristics, in addition to ICT data. Regional population, geographic, topographic, demographic, and socio-economic indicators that reflect relative wealth, wellbeing and poverty should be tabulated and compared, for the entire country, including served and unserved areas. This data is needed for the ICT sector review, as it joins socio-economic data with ICT data; both are needed in order to develop a universal access and service (UAS) programme and identify the potential role and impact communications may have on development.

Socio-economic data review

The result of gathering this data is a basic desk-based comparative study that tabulates the socio-economic and infrastructural wellbeing or poverty of a country. Ideally, to get the clearest picture of the county’s UAS status, the study should be broken down into lower administrative levels, e.g., province-by-province, district-by-district, etc. Typically, the following socio-economic data are collected and compiled:

  • Population size, density and distribution;
  • Number of towns, villages etc., classified by size;
  • Income levels and distribution;
  • Nature of economic activities and major sources of income;
  • Health & health infrastructure statistics and indicators;
  • Education infrastructure statistics (e.g., number of schools, primary and secondary, tertiary institutions such as universities and colleges etc.) and education levels and enrolment;
  • Composite Human Development Index (HDI); an index combining normalized measures of life expectancy, literacy, education, and GDP per capita for countries worldwide;
  • Commercial indicators (e.g., number of bank branches, businesses, etc.);
  • Local governments and administrative offices, other institutions, NGOs, etc.;
  • Major construction or development activities and programmes in a province or district;
  • Basic infrastructure such as power, transportation (e.g., paved roads, etc.), and postal services;
  • Terrain (e.g., rugged mountains, hills, savannah, crop lowlands, etc.); and
  • Socio-cultural distinctions that may have relevance (e.g., ethnicity, religion, language, minority groups, presence of indigenous people, nomads, etc.).

In addition to the government’s official statistics office, sometimes various ministries (agriculture, local government or rural development, planning, transport, health and education) have useful additional resources and a good understanding of the specific needs of regions and certain population groups. Also, banks that have a national-wide presence, as well as farmers associations, NGOs, development agencies and micro-finance institutions can contribute valuable information. A benefit of this level of research is that it will help the UAS programme identify and connect with other development initiatives.Once data is collected, its comparative analysis provides an understanding and description of the country’s geographic zones, regions, administrative departments, municipal units and other population centres. The output can provide both descriptive material and tabular comparative analyses that can be used to classify the country’s regions for total market capacity (potential revenue), level of development, as well as identify the existence of social and physical infrastructures that create demand for communications services in unserved areas.

ICT data Review

After analysing the socio-economic data, it is necessary to review the existing communications infrastructure throughout the entire country. As discussed previously, understanding the country’s communication infrastructure will:

  • Identify which services and applications the infrastructure can support;
  • Identify the infrastructure gap
  • Identify the next steps for achieving universal access and service (UAS);
  • Identify feasible upgrades and alternatives, as well as any potential for leveraging existing technology for new types of services.

The ICT sector review also necessitates a thorough review of current policy and regulatory frameworks – including issues such as tariff policy, spectrum allocation, licensing and liberalization, among many others (see also Chapter 2 on regulatory reform and UAS). In addition, developing a clear view of the policy changes likely to be implemented in the short-to-medium term is essential because these changes will determine what infrastructure development and service provisions will be available in the near future, and these, in turn, could be leveraged for Universal Access (UA) and rural ICT projects. A helpful tool for the ICT sector review is to interview various industry players. This may be outsourced to an independent expert entity or accomplished through a qualitative questionnaire. In addition to obtaining detailed information on the existing and planned infrastructure from each operator, questions as to any planned policy initiatives, market development and regulatory constraints may be included. The interview or survey process ideally includes the main telecommunications and Internet infrastructure and service providers, a few of the smaller players, and if possible, potential new entrants and alternative service providers (e.g., railway, power companies). If they exist, public phone operator companies and others related to telecentres or Internet cafes could also be interviewed. This process provides a good understanding of the existing context, demand, constraints and opportunities within the telecommunications/ICT sector in terms of reaching the country’s unserved communities and achieving UAS for both telephony and the Internet. These interviews, crucial to identifying the current status of UAS, will ascertain which areas will be served commercially and which will require intervention. The topics to be discussed with the relevant communications operators and service providers include the following:

  • Network related topics:
  • Network statistics, current coverage, points of presence (POPs), bandwidth capacities;
  • Cost structure, both capex and opex;
  • Network expansion and investment plans;
  • Technologies used and future trends/ preferences;
  • Financial topics:
  • Revenues (average and marginal) for various services;
  • Expectations for financial returns;
  • Market and subscriber issues:
  • Levels/range of services, numbers and types of customers for each service type, etc.;
  • Marketing and distribution;
  • Tariff strategies and prices;
  • Public access:
  • Experience with deployment of phone or Internet public access facilities;
  • Current numbers, types and deployment of public ICT access facilities;
  • UAS issues:
  • Experience with reach into rural areas and underserved communities;
  • Approaches of the operators and ISPs towards reaching low income people and communities; and
  • Strategic ideas for UAS, etc.

The operators and service providers should be asked whether they are planning to serve the unserved areas and population groups in the near future and which areas or target groups will require government intervention. These discussions may be held in private due to the confidential nature of expansion plans. It is also very helpful to include non-ICT players that have either a commercial interest in rural areas (e.g., agricultural suppliers) or in socio-economic development, such as donor and development agencies, and NGOs. Results of the overall ICT sector analysis should be summarized in a report and presented to the telecommunications industry and other relevant stakeholders for validation and refinement. Such presentations also provide opportunities for identifying areas of uncertainty so that further informational needs can be identified and later integrated into the demand study design. Per capita and household expenditure on communications

Universal access and service (UAS) programme developers can build up a desk-based demand model for unserved areas based on data collected during the ICT sector review which may be refined later through a field demand study. The model uses national data on communication expenditure and projects it to unserved areas. It starts out with the country’s total telecommunications revenue divided by its GDP, resulting in the percentage telecom expenditure of GDP. This can be refined with additional data where available, as follows:

  • Percentage of GDP accounting for household income (typically 60-70 percent of total GDP);
  • Typical rural or low-income compared to average income;
  • Regional variation of income;
  • Percentage of telecom revenue from business users (to be subtracted from the total telecom revenue for a closer approximation of household expenditure);
  • Data on telecom expenditure by households; and
  • Household number and size.

By having data on population or household and income in various identified unserved areas, the total revenue generating potential of each unserved area can be calculated on an annual basis, as follows:

(per capita income) x (population) x (the telecom expenditure percentage)or(per household income) x (no. of households) x (the telecom expenditure percentage)

In most countries, the demand for telephone service has been identified at somewhere between 2-5 per cent of a country’s GDP, using ITU’s indicator of telecommunications revenue as percentage of GDP, as can be seen in the Figure below.

Figure: Total Telecom Revenues as % of GDP - 2007

Source: ITU-infoDev ICT Regulation Toolkit – UAS Module based on ITU World Telecommunication/ICT Indicators.

A recent review of several telecoms demand studies in developing countries found a high variability in spending levels per household, as well as significant rise globally of the percentage in household income spent on communications [1]. Field surveys of rural areas in Africa have indicated household expenditure levels of more than 5 per cent in some countries. This makes sense, considering that rural incomes are often lower than average income, thus the expenditure is a higher percentage of the income and also, living in rural and remote parts of the country, the opportunity costs of travelling to communicate are much higher. This makes the argument for communications services all the more attractive. In a large demand study commissioned in 2005, by the Nigerian Communications Commission (NCC), respondents stated they spent on average USD 20.7 per month on the mobile phone (fixed phones are almost nonexistent outside the main cities), which is 7 per cent of monthly household income. In the developed world, phone cost expenditure is typically less than 2% of Gross National Income (GNI) per head, due to a combination of low telephone costs and higher incomes. People want and need to communicate and are prepared to pay for it no matter what their economic status is. The benefits they receive from communications expenditures are now well known as they relate to routine family, business and emergency matters, and to time and expense savings (the opportunity cost) of alternative means of communications, such as the necessity to travel.In NCC’s demand study, over one third of respondents (36 per cent) indicated that they travel to another town to make a telephone call (as shown in the Figure below). 21 per cent of the respondents actually travel to meet in person because they have no access to a phone. A third of respondents send a letter or use a messenger (or a combination of both – i.e., write a letter and then have a messenger deliver it) to communicate. Approximately 8 per cent do none of the above or selected “Other” (methods of communication) in the survey. Less than 1 per cent of respondents stated that they have their own mobile phones and travel to the nearest coverage area.

Figure: Alternative Means of Communication Due to Lack of Phones

Source: NCC Demand study, 2005

On average, respondents stated they spend one hour and 40 minutes travelling to make a phone call and return home. The average total cost of return travel to make a phone call is USD 2.84. This is based on over 5,000 interviews in all six regions of Nigeria, and a sample that included urban, semi-urban and rural areas (but excluding major cities).

Reference Documents Incoming call revenue

When considering the demand and revenue potential of unserved or underserved areas and communities, it is important to include the demand for incoming calls, a significant source of revenue. Adding revenues stemming from calls flowing into rural areas from urban (and international) destinations to revenues from basic service connections and outgoing calls, the total potential revenue from unserved areas and communities could be double what potential users themselves are willing and able to spend. The following illustrates the results of a rural demand study undertaken in 2005 for Mozambique’s Ministry of Transport and Communications. The study of 226 rural households without telephone access in three different districts in the province of Zambezia, revealed that 19 per cent of the respondents stated that they have close family members living abroad, often in Malawi or South Africa. The Figure on the right shows that 24 per cent of respondents stated that they have close family members living in Maputo or another major city outside of Zambezia. These findings predict (incoming) international and national long-distance traffic that might be generated by providing services to those rural areas.

Figure: Family members in Maputo or other major city outside the province

Figure: Family members in Maputo or other major city outside the province

Source: Mozambique Demand study 2005


Further, the study revealed that 92 per cent of respondents would receive incoming calls if their village had phone service. On average, the respondents would expect to receive three incoming calls per week. When asked how many calls they would make on a public phone per week themselves if phones were installed in their village, respondents stated they would make on average four calls. Similarly, in the 2005 study commissioned by the Nigerian Communication Commission (NCC), 39 per cent of respondents stated that they have family members living abroad, as shown in the Figure on the left. This incoming call market (into rural areas) can generate revenues for carriers either in the form of termination charges or urban customer revenue. In many countries, urban callers tend to be more affluent and can afford more calls and lengthier calls to rural relatives and friends.

Figure: Family members living abroad – in %

Figure: Family members living abroad – in %

Source: NCC Demand study 2005


In Latin America, data from Chile demonstrate that rural telecommunications operators earned more than 60 percent of their revenues from incoming call terminating charges. It is remarkable that rural pay phones in Chile report this significant number of incoming calls because typically, stand-alone pay phones are not set-up to receive incoming calls as easily as a phone shop that has an operator in attendance [1]. For pay phones to be effective for incoming calls, they need someone living close to the payphone who is willing to receive incoming calls and relay messages to the called party, or who is willing to set up appointments between the caller and called party. Another example of significant revenue potential made from incoming calls, is Bangladesh’s Grameen’s Village Phone programme, discussed in detail in Section 3.3.1. This well-documented case illustrates how telephone service can be extended to low-income, rural populations. The average usage of the village phones amounts to about 1600 minutes per month, out of which approximately 1000 minutes are from incoming calls [2].In conclusion, when estimating demand, incoming call revenue needs to be included, either through data gained from a demand study or reasonable assumptions based on data from comparable markets.

Practice Notes Demand for Internet services

In developing countries that include access to the Internet as part of their universal access and service (UAS) policy, the target is typically not individual household access but public access to the Internet. Thus, this public usage through Internet cafés and telecentres needs to be forecasted, as well as some private demand for Internet services. This section discusses methods of forecasting Internet demand.

An estimate of the total revenue potential should be made, including both public demand at telecentres and private demand of households, businesses and institutions. The estimate should include private demand in order to assess the feasibility and viability of extending Internet infrastructure and service to a targeted area. Further, demand can be broken down into the various Internet related services (from e-mail via browsing to Internet telephony, e-commerce or e-government services) as well as several pricing models (time-based, data-throughput-based, flat-rate and bundled pricing).Demand obviously has two dimensions – those who want to use Internet services, and those who can afford to use it, e.g., many youths want Internet access but often can’t afford it. Assessing potential demand for Internet in unserved areas starts with interviewing the existing national ISPs and cyber cafés, and collecting other key data on topics such as the following:

  • Internet subscriber penetration (business/ household);
  • Internet user penetration;
  • Socio-demographics of Internet users;
  • Computer penetration (if possible, disaggregated by urban and rural, and business and household);
  • Number of businesses and institutions (per region, administrative units);
  • Typical Average Revenue Per User (ARPU) per standard or low-end user;
  • Typical ARPU of business users;
  • Typical ARPU of specific institutions (e.g., NGOs, local government offices, schools, health centres, etc.);
  • Typical respective data speeds to different types of users; and
  • Data on cybercafé number, usage, revenue, and services that are most in demand.

The above data provide an understanding of how much demand there is for the Internet in rural and underserved areas and should be used to develop a demand model. However, the model may have to use certain assumptions, which need to be conservative and are likely to vary from country to country. Some illustrative and hypothetical assumptions that might be made from the data are:

  • If Internet user penetration is 20 percent (nation-wide), then a plausible assumption could be that demand in unserved areas comes initially from 10 per cent of the population;
  • If the national average for telecom users is USD 10 per month on communications, then an assumption might be made that users in unserved areas might spend USD 5 per month; and
  • If 30 per cent of nation-wide businesses have computers, a starting assumption might be that 15 per cent of businesses in unserved areas have computers;

These assumptions can be tested by the field demand study where actual and potential users are interviewed. By preceding the demand study with a sector review, the field demand study will then focus on indicators that are actually required to model demand and assess feasibility for the UAS programme. This will support the development of an effective UA programme. The demand study makes it possible to develop an estimate of the number of short-to-medium term potential Internet users on a district-by-district basis, and a resulting estimate of potential revenues and subsidies for the provision of an ISP wireless access system. In many developing countries, the degree of confidence will be less than for telephony as data in this emerging Internet market is more transitory and less reliable.

Businesses and institutions as vanguard users of Internet

In most developing countries, local government agencies, social infrastructure and health institutions, schools, NGOs and businesses provide most of the demand for Internet services. They can be considered intermediaries: while the vast majority of these institutions are located in district centres and/or other small towns, the benefits of their access to the Internet can spread to villages. Many of their clients are villagers who may also visit their offices frequently. There is also a growing public demand for the Internet via cybercafés.Overall, the potential user community and demand for Internet services has to be identified in a more consultative fashion than that for telephony. In most developing countries it is less advanced than in the developed world and that has more varied uses and applications than telephony. Estimating Internet demand often involves more than the simple identification of the number of business or administrative units. It includes consideration of how to catalyse partnerships between development, administrative and private agencies to create sustainable and scalable ICT demand. Field demand studies

A demand study carried out in the field (often interviewing rural end-users) is important in the development of a universal access and service (UAS) programme for the following reasons:

  • While modelling using desk-based data research and industry interviews is important, there can be data gaps that need to be filled by a field demand study;
  • Key assumptions of a model can be validated or refined through a field demand study – the demand study is the bottom-up approach interviewing end-users, which complements the modelling top-down approach;
  • A demand study increases the credibility of the UAS programme with industry and other stakeholders as it represents a thorough investigation and not a theoretical modelling approach; and
  • Elements of a potential UAS programme can readily be tested, such as the following questions:
  • What are suitable service levels and required service elements?
  • What are the best public access models in the country’s context?

A side benefit of field demand studies is that it can educate potential future users about ICT services and their applications and benefits. Field demand studies range from sample or pilot field studies to full-scale representative baseline demand studies. These have several levels of depth and accuracy. Demand studies can provide information on the following:

  • Needs for communication of various kinds;
  • Private demand, affordability and willingness of users to pay for telephone, Internet and other ICT services;
  • The numbers, location preferences and mode of public access points;
  • Technology and service preferences (e.g., mobile versus fixed, required Internet speeds, broadband needs); and
  • Knowledge and demand for different value added and non-voice ICT services.

Deciding on the scope of a field study

Various factors – e.g., size of country, expected scale of UAS programme and whether or not a pilot project is planned – will influence the required scope of the demand study. The scope of the field study will also depend upon what data is already available and how reliable the data is. Demand studies require money and time and thus determining the right size of the demand study is important. Key questions to be asked in determining the scope and size of a field study are the following:

  • Is there the political will or the need to include certain regions or populations? E.g., some countries have strong regional or ethnic differences that require field studies to be more inclusive than others;
  • How reliable is the existing available data? E.g., if there is good quality data and research on household income, population and ICT usage then the field study can be smaller and focus more on specific UAS issues;
  • What assumptions industry and policymakers make about the demand for and importance of UAS? E.g., in some countries there might still be a prevailing perception that communications is not as important as other development initiatives. Access to the telephone or the Internet may be considered a luxury in unserved areas where fundamental needs such as food, water and roads are not served. In those cases, a wider demand study can be an educational tool and provide credibility that will assist in changing perspectives among stakeholders;
  • How comfortable are (industry) stakeholders with certain demand model assumptions? E.g., if there is more uncertainty about demand indicators and assumptions, then a field demand study is required to solidify the model; and
  • Is a pilot project planned? E.g., a pilot project can provide hard data on demand, affordability, revenues and sustainability. If a pilot is planned, the demand study may focus mainly on the planned region or area of the pilot.

Elements and process [1]

A field demand study can be designed either for telephony or non-voice ICT, or both. Usually, a demand expert designs the survey with the assistance and active collaboration of a local research institution (e.g., university social research department or institute) or a market-research company. The following diagram shows the various steps of a field study, which ideally, is based on a prior desk study.

Figure: Steps of a Field Study

Source: A Rural ICT Toolkit for Africa, Andrew Dymond, Sonja Oestman, infoDev 2003

For an overview of a demand study design, see the Practice Note A demand study in rural areas in Mozambique.  

Step 1: Study objectives

Study objectives articulate the purpose of the study, outline the services to be investigated, determine the region(s) in which the demand study is to be conducted and identify which target groups are to be interviewed.A UAS demand study that focuses on telecommunications and Internet services in areas that currently don’t have services has the following objectives:

  • Identify and assess the potential demand for both voice and Internet services;
  • Determine user ability to pay for communications services, including priority customers such as business and institutions;
  • Explore needs and preferences of users regarding private and business service, public phones, public access for the Internet, value-added services (VAS) and required Internet speeds; and
  • Explore and validate appropriate UAS targets for desired service levels. This could include travel to public phones, numbers and type of public phones and public access for the Internet.

Step 2: Selection of representative districts and communities

If representative provinces, districts and communities are carefully selected for the study, then a smaller sample may be used. Although statistically representative field studies may be desirable, they are expensive and time consuming, and in most cases their level of accuracy may not be required for the purposes of a UAS programme design. Based on an earlier desk study, the following criteria and questions should guide the selection of a representative sample from which results can be extrapolated nationally:

  • Does the sample cover all typical or key regions and districts?
  • Does the sample cover gender and age groups representation?
  • Does the sample cover typical sizes of villages or towns?
  • Does the sample cover both areas with telephone and ICT services and areas without?
  • Does the sample cover various economic situations (e.g., poor to more affluent areas)? and
  • Does the sample cover different population densities and terrain (e.g., remote and sparsely populated, more densely populated, mountainous or plains, etc.)?

Step 3: Survey instruments and questionnaires

There are several general considerations to take into account when developing questionnaires. These are as follows:

  • The length of the interview – respondents can become impatient or bored if the questionnaire is too long resulting in poor quality of responses. If possible, limit questionnaires to 30 minutes;
  • Self-administered versus interviewer-administered questionnaires – the latter is the preferred approach in underserved areas, which are often rural and where literacy might be an issue. Interviewer-administered questionnaires allow more control, consistency (as questions can be explained in case of misunderstandings) and higher completion rates;
  • Open and closed questions – this depends on the objective of the study; open-ended questions require a more complex analysis and classification process, and should be used sparingly. On the other hand, closed questions are prone to confirm preconceived expectations. Multiple choice questions are an effective compromise between open and closed questions and ideally are tested or developed through initial pilot surveys with focus groups and with more open-ended questions.

Typical UAS demand survey instruments include questionnaires for private, business, community leaders and informants, and might have separate surveys for telephony and Internet. A rapid community assessment checklist, in which the researchers are requested to record a set of community data for each locality targeted, is also useful to aid later analysis of survey results. The methodology and survey instruments of a field study could include a mix of the following:

  • Focus group discussions. Depending on the objective, several open interviews can be conducted for purposes of generating a discussion around some of the key hypotheses of the survey and the planned UAS policy or programme. Focus groups can be used to develop the questionnaire design and to tailor multiple-choice questions. Special attention should be paid to variations in the socio-economic characteristics of participants and actual access and utilization of telecommunications and Internet services. Examples of topics to discuss might include:
    • The location of public telephones and/or telecentres;
    • The distance people are willing to travel;
    • The main use of the phone and/or the Internet; and
    • The amount they would be willing to spend.
  • Key informants. A few key informants, (ideally in each district surveyed) selected on the basis of their overall knowledge of the area and use of ICT in the community, could be interviewed. Key informants can provide considerable qualitative information about the area, economy, local village life, etc. They can include district administration officials; representatives of health centres, schools, community leaders, business people, phone operators, NGOs, head of farmer associations, local bank branches, etc.;
  • Household & small business survey. At the heart of the field demand study is a stratified random sample of a significant number of households and small businesses in each community. An interviewer-administered questionnaire with both closed and open-ended questions is typically used to solicit responses from male or female heads of households; and
  • Rapid community assessment. This assessment profiles the community being surveyed and details the number and type of key administrative and social or public infrastructure institutions, businesses, households, market(s), and other key social collection points, transportation facilities, etc. that generate demand or locations for public access.

Control groups

A further distinction is required between existing users in areas where people have access to telecommunications or the Internet, and potential users where there is no access. Demand studies typically need to include at least one control group from areas that have access to the telephone or other ICT services. By doing this, the study can measure the actual existing demand, usage and willingness to pay.  

Overview of key elements of various questionnaires

The questionnaire typically covers, as a minimum, the following areas:

Figure: Overview of Key Elements of Various Questionnaires

Examples of different types of questionnaires used in a rural demand study in Mozambique are provided as Reference documents.  

Step 4:Training and pilot survey

 The actual training of the research team helps to deepen their understanding of the field survey objectives, and role-playing the interview process among the researchers has been found particularly helpful. An initial pilot survey is recommended to give the survey team a test-run and to refine the methodology and survey instruments before beginning the main research.  

Step 5: Analysis of demand study findings

The results of the field demand study will inform and refine the UAS programme design. Demand studies also help to adjust previous estimates of cost, revenue and commercial viability, and the amount of subsidy likely to be required per area or project. Typical results of demand studies indicate the following:

  • The current accessibility of telecommunications and ICT networks and services, both in terms of how much they cost and how far end users need to travel in order to use them;
  • Potential penetration of private telephony and Internet service (fixed or wireless) amongst businesses, schools, clinics, government offices, households, etc.;
  • Which technology is requested (e.g., fixed versus mobile, high-speed up to 256 Kbps or broadband Internet connection);
  • Identifying the optimal number of public phones or telecentres per community or neighbourhood, as well as the population size to support each public phone or telecentre, best location, preference for mode of public access facilities (e.g., manned or unmanned, coin pay, card phone, etc.);
  • Interest in Internet and potential for public access to telecentre services;
  • Private Internet demand;
  • Interest in voice messaging, SMS (text), information services, e-mail, fax, etc.;
  • Readiness of schools, institutions, and businesses for Internet or participation in ICT programmes; and
  • Potential partnerships that could be forged to stimulate sustainable and scalable demand for Internet and ICT services.

Caution about household income data

While field studies are infinitely more valuable than simple desk-based modelling in many regards, there are still some challenges. For example, obtaining data on household income through questionnaires is notoriously difficult, especially in developing countries. The main challenges in determining income are as follows:

  • People usually do not like to disclose their income level and may under-report it;
  • A larger number of people, especially in rural areas, have incomes that vary from month-to-month. Some may not know, or be able to predict, what their monthly household income actually is (e.g., shopkeepers, self-employed, and especially farmers or labourers whose income is seasonal), and while they may know their seasonal incomes, they man not be able to calculate a monthly average household income;
  • People’s income is often derived from numerous sources, e.g., their shop, from breeding and selling animals on the side, occasional jobs, etc. Again, in many cases people do not know how their income from various sources averages out over the year, and are not able to state a monthly household income;
  • Income sources that are non-cash, contribute to the wealth of households. Examples of this are barter, trade and subsistence farming;
  • Several household members may contribute to the household through various activities and employment (e.g., the grandfather tends animals, the father has a shop, the mother does some sewing, and the oldest son works part-time at a garage and might give his parents part of that money);
  • In some countries a considerable number of people derive income from the grey or black market [2] which they are hesitant to report.

Demand related to Internet services is in many cases even more challenging to assess as there numerous Internet applications and services as presented in Section 6.1.4. Therefore, in general, results from desk or demand studies are still approximations to actual demand, but they are likely to provide the best information that is available. It is very helpful to disseminate the field demand findings to the sector and other stakeholders and make it available to the public e.g., through publishing it on the regulators web-page. The survey output can specifically be used for the next step of UAS programme development which is the modelling costs, viability and subsidy analysis described in Section 6.2.

Practice Notes

Reference Documents

4.6.2 Modelling costs, viability and subsidy analysis

Once the ICT sector analysis and assessment of demand are conducted, the costing aspect needs to be addressed in order to develop a universal access and service (UAS) programme:

    • How much will the UAS programme cost and how much will individual projects cost;
    • What is the commercial viability and sustainability of the UAS programme and projects; and
    • What is the estimated financial shortfall between the cost and revenues through the provision of UAS services, and therefore how much subsidy is required to finance UAS.

The following methodology on modelling costs and viability and subsidy analysis is designed to assist UAS programme development in the context of a Universal Access and Service Fund (UASF) using competitive mechanisms to allocate subsidies. The typical process to competitively allocate UASF subsidies for UAS service projects is as follows:

    • Identify how much subsidy an operator or service provider needs in order to construct the stipulated level of infrastructure and to provide UAS service to the target area, population group or project, using current least-cost technology that meets the required quality and level of service; then
    • Submit the specified area to be supplied with UAS services to tender and accept the lowest bid commensurate with meeting the published technical, corporate and operational pre-qualification requirements, provided the required subsidy is less than or equal to the set maximum subsidy.

The methodology for calculating the amounts of subsidy to be offered is designed to achieve, with the subsidy, commercial sector investment and sustainable operations. The purpose of establishing maximum subsidy levels is to set a realistic ceiling. It is not necessary to establish exact costs and a detailed engineering study by the regulator is not required. An illustration of the methodology used to estimate the maximum subsidy is described in Section 6.2.1 for a public access telephony project and in Section 6.2.2 for a public access Internet project This methodology is chosen, among other reasons, because the regulator has less information than telecommunications operators and service providers have about the costs of providing UAS services. Operators and service providers also consider their financial and non-financial benefits they may stand to gain from providing UAS. Some of the benefits of being a UAS provider may include:

    • Increased brand recognition;
    • Good public relations;
    • Additional private revenue; and
    • More traffic due to increased national network, etc.

With a maximum subsidy amount in place, the competition among operators and service providers ensures that costs are not inflated, and that the operators include their tangible and intangible benefits from being a UAS provider into their considerations for the subsidy request. Revenue and cost estimation for public access telephony

For every universal access (UA) region or project, the expected communication service revenue and costs to supply the service can be estimated and submitted to a viability test. This viability test will establish if a certain project needs a subsidy in the first place, and if so, will calculate the likely maximum one-time subsidy that an operator or service provider would require to provide service in a commercially sustainable manner.  


The revenue estimate typically takes the following into consideration:

  • Regional/rural population (p);
  • Regional/rural per capita income (i), estimated either as:
  • the average of the lower deciles of national income distribution – corresponding to the proportion of the population who are rural (this assumes that the rural population is generally at a lower income level than urban), or
  • an average for the region (e.g. district, provincial) GDP (for example household income from UNDP studies); and
  • An expenditure factor (e), which is the percentage of income spent on telecommunications. This can be the national average, or a more conservative regional (e.g., district, provincial) estimate, or findings from demand studies, if available.

Revenues can thus be estimated as follows:

R = p x i x e (Revenues = Population x rural per capita income x expenditure factor)

Findings from demand studies are helpful to refine the input for (i) – the income of rural people, and (e) – the communications expenditure factor. While the installed public access phones will not necessarily be able to capture all potential demand, it is likely, if wireless technology is used, that the potential service provider will also be able to capture some private demand, including business and institutions, as well as revenue from urban customers travelling to those rural areas.


The capital and operating expenditure costs of supplying service can be estimated using the pre-dominant technology (usually wireless) utilised by existing operators and service providers in the market. This assumes that existing operators will be interested in expanding into the unserved areas targeted by the UAS policy. Using the predominant technology will represent the cost ceiling, since more cost-efficient alternative and newer technologies, whether selected by the major operators or new entrants, would generally need to be less expensive to be considered competitive solutions.

Maximum subsidy estimate

The maximum subsidy requirement may be calculated in one of two ways:

  • Construct a 10-year cash flow for each project and area, showing revenue (with an allowance for growth over the cash flow period), capital and operating costs, and calculate the net present value (NPV) using the operator’s assumed cost of capital as the discount rate. If the NPV is negative, this amount will represent the maximum subsidy required by the operator to provide service. If the NPV is positive, this indicates that no subsidy may be required, although it will be necessary to investigate alternative revenue and cost assumptions to determine the robustness of the calculation; or
  • Make a basic benchmark calculation that assumes a standard pay-back period of capital cost from revenues of perhaps three years, which is typical for private telecommunications investments. If the projected revenues are greater than 33 per cent of the capital costs (meaning it takes less than three years to achieve pay-back), the project is viable. If they are less than 33 percent of capital cost, the project has negative viability. This will indicate whether, and by how much, the revenues fall short of providing the operator with an acceptable rate of return.

Whatever method is used, both models and their assumptions should be discussed with operators and service providers. Either of the above methodologies should safely estimate the maximum subsidy required for the following reasons:

  • The assumption that public service points will be the prime instrument for securing potential revenue is conservative. In the event an operator is also able to use the same infrastructure to provide private services (e.g., mobile services to individuals that possess a mobile phone), more revenue could be available from some private customers. In this event, the actual revenue would be higher than that calculated by this methodology; and
  • The revenues are based only on local (rural) affordability. For example, if revenue from incoming calls is significant, and is also encouraged through a good interconnection arrangement (such as asymmetric interconnection) and revenue share with a village phone operator, it is possible that operators’ incomes could improve significantly due to incoming calls from urban areas. Revenue and cost estimation for public access internet

In many developing countries, there is limited information regarding the cost and demand for semi-urban and rural Internet provision. While field demand studies are very helpful overall and in some countries provide the very first survey-based data from rural areas, they often still yield insufficient solid data on Internet demand in some countries where many people have not used the Internet before. A potential guide for estimating cost and demand is through examining the experience of other universal access and service (UAS) programme designs and tenders. In 2004 in Uganda, district level Internet POPs have been subsidized between USD 10,000 and USD 60,000 and public Internet centres for between USD 15,000 and USD 25,000, however, data evaluating their sustainability are not yet available. Subsidy estimates include the capital, the cost of the digital bandwidth leased from terrestrial service providers, and the initial promotion, marketing and basic training. For estimates of potential subsidy requirements for an Internet programme, it is usually recommended to offer a value equal to almost the entire capital costs of a district centre POP and a public access centre. The reason for this is that demand for Internet is typically less immediate and takes more time to build up, and operating, maintenance and staff costs in districts are considerably higher than in cities. Offering the entire capital costs as the maximum subsidy, and then letting the market determine the real subsidy through competitive bidding, ensures that it will not be under-estimated. As an example, capital costs for the Internet component of a pilot in Mozambique are estimated at USD 75,000. This included VSAT backhaul as the worst-case scenario. Details of the cost components are shown in the Table below:

Source: Mozambique Universal Access Internet pilot programme 2007

Source: Mozambique Universal Access Internet pilot programme 2007

4.6.3 Economic impact of UAS projects

As discussed in Chapter 1 of this module, countries develop universal access and service (UAS) policies based on the premise that providing access to basic and advanced telecommunications and ICT services have a wide-ranging socio-economic rationale. There are general studies and analyses, sometimes conducted by academics, that address broader questions of economic impact of communications, ICT and broadband Internet. Their understanding and latest knowledge is an important background for UAS policymakers and UAS program planners, as these studies provide the general rationale and justification. Section 6.3.1 outlines the impacts of communications and Section 6.3.2 discusses the status of research for broadband Internet and ICT. In the case of telecommunications projects, it can often be demonstrated that there is significant benefit in the form of ‘consumer surpluses’, over and above the price paid for the service. These include items such as the following:

    • Businesses (small or large) often report that the money they save due to greater efficiency and saved personal travel time related to stock control, delivery co-ordination, following up sales opportunities, maintenance calls, etc. amount to several times the cost of the telephone rental and calls they make.
    • Farmers and micro-business proprieters often report that the phone enables them to gain timely and geographically-specific information on urban market prices that increase their bargaining power with ‘middlemen’ and enable them to earn more for their product or secure a better price for their inputs.
    • A third of personal calls typically represent personal or family emergencies that would again require travel or other costs if the call was not made. Sometimes people report their benefit in terms of lower health risk, a life saved, better family relationships, more opportunities.
    • Institutions and government agencies – schools, clinics, local administrations, NGOs and other development agencies - similarly report increased efficiencies and the ability to deliver services in a more timely, effective and less wasteful manner through use of the telephone.

The benefits of broadband Internet and ICT are less direct and more difficult to quantify, though, there is general consensus that critical macro-economic value is gained from the provision of broadband access to ICT services. Research and analysis on measuring broadband impact are only beginning. What is clear is that the impact of broadband is highly dependent on framework conditions within the country and also within the sector which is to benefit from broadband access. An example to illustrate this is the following: a health project plans broadband connections to be provided to rural district hospitals to facilitate remote diagnosis, consultation, transmission of imagery and data, and video-conferencing between the rural hospital staff and specialist doctors in the urban hospitals. In order for the health sector to reap the benefits, certain conditions need to be in place, such as:

    • Trained staff at rural hospital to operate and maintain necessary ICT facilities and equipment;
    • Increased staff at the urban consultation hospitals to be able to accommodate increased demands through rural hospitals;
    • Privacy regulation on patients records which are electronically submitted; and
    • Cost accounting and financial incentives for urban hospitals to provide remote consultation to rural hospitals.

Successful case examples of how broadband and ICT is addressing crucial development issues need to be studied in terms of there required pre-conditions to be transferable to other countries. Another important aspect is their sustainability. Piloting broadband networks and ICT projects maybe a helpful tool to adapt and refine the models to be used for a particular country and assess the required inputs, outputs, outcomes and project efficiencies. While the economic impact provides the broad rationale for UAS programmes and projects, nowadays UAS programme development typically requires less extensive detailed economic analysis, as it is more widely accepted that ICT are crucial for socio-economic development. In UAS programme development it is about making choices and decisions on how to prioritize UAS projects among available options. Section 6.4 presents practical approaches, considerations and methods that are used to help to prioritize UAS projects, some based on financial analysis and others considering benefits.

Reference Documents Economic impacts of communications

Since development planners first recognized the immense socio-economic impact of communications, general hypotheses about impacts with regard to rural communications systems have evolved, and have been demonstrated to various degrees, in a wide range of studies since the early 1980’s. In 2005, the study The Impact of Telecoms on Economic Growth in Developing Countries, has added to this long history of previous research, conducted over many years, to demonstrate that telecommunications has a significant impact on economic growth. This research has highlighted the particular impact of mobile communications penetration on economic growth. The socio-economic impact of communications can be summarized as follows:

  • General regional integration:
      • Areas with communications are less isolated economically and socially, are better able to enter the market system, and will experience improved political administration and social services;
      • The benefits can be described as macro-economic and structural, e.g., development of market system, and the enablement of the services and information sectors; and
      • The level of impact is increasing and becoming more widespread the more mobile the communications medium is becoming [1].
  • Market & social infrastructure:
      • Typically 25-40 per cent of calls made from rural public phones or payphones in developing countries can usually be classed as related to administration, business or financial matters [2];
      • Related to commercial markets, better communications provide both the means and sources of information regarding the price of rural products, typically yielding fairer market relations and more efficient operation of the market system;
      • Improved organizational management is possible – most organizations (e.g., government, health, education and transportation) and businesses run more efficiently as communications services improve, resulting in better coordination, stock ordering & replenishment, more timeliness, and quicker response to operational and maintenance needs; and
      • New market and employment opportunities – businesses can organize better outreach, market access and market expansion.
  • Personal urgent need to communicate:
      • Typically 10-15 per cent of calls made from rural public phones or payphones can usually be classed as ‘personal urgent’. These typically relate to health, family emergencies or other matters considered urgent enough that some other form of communication – often personal travel to deliver a message – would have been necessary; and
      • Urgent matters often include notification of family or social events, coordination of travel arrangements, or such matters as school exam results, deadline related matters (e.g., financial enquiries, school entry applications), etc.;
  • Personal non-urgent communication:
      • Up to 50 per cent of all calls are typically classed as personal and non-urgent, but nonetheless important enough that people are willing to spend 3-5 per cent of personal incomes on them; and
      • The benefit enjoyed is the reduction of isolation for family members living elsewhere, especially with younger generation members studying or working in the capital city.

A compendium of cost-benefit results illustrating many of the above benefits are documented and discussed in the reference document Methodology for Economic Analysis of Telecommunications Projects. The expected economic impacts and targets of a universal access and service (UAS) programme should be stated as specifically as possible, in order to assist with final selection or prioritization of projects. The Practice Note Specific regional and poverty reduction impacts in Mongolia provides an example for this.

Practice Notes

Reference Documents Economic impacts of broadband Internet & ICT service deployment

With the proliferation of broadband connections worldwide, the Internet is increasingly described as being always on and with a minimum speed of 256 kbps. This broadband Internet moves well beyond traditional dial-up and mid-speed Internet (e.g., less than 256 kbps) in terms of the applications it supports and the value it offers. While dial-up Internet offers some additional value over telephone communication, in terms of basic Internet browsing, e-mail, and simple document transmission, most agree that the Internet’s full capacity is only realized through broadband Internet that allows the use of multiple ICT applications and services.

Assessing the economic benefits and impact of broadband Internet is fundamentally different to assessing the impact of telephone communications, for the following two reasons:

  • A broadband Internet connection does not in itself provide any value or service, thus contrasting with a traditional telephone connection that provides an instant use and benefit through immediate and direct verbal communication (saving travel costs and time of alternative means to communicate). Only the use of Internet services and applications made possible through a broadband connection create benefits for the user, such as remote network access, VoIP services, video-conferencing, online-banking etc.; and
  • The realization of benefits from broadband for a country is strongly dependent on overall conditions such as the: regulatory framework (e.g., security for e-commerce transactions, laws for on-line banking, etc.); business environment (e.g., computer penetration, ability to develop skills, capacity for organizational change); supporting infrastructure (e.g., reliable electricity); and ICT literacy among the population, among others.

Like electricity, ICT may be considered a general purpose technology, which is characterized through its pervasiveness throughout the economy and society, its constant evolution and improvement, and its capacity to spawn new innovations.

Measuring economic benefits of broadband

Due to broadband Internet’s multitude of applications and its recent national spread in certain markets only (in many developing countries it is limited to the main urban areas only), it is more challenging to measure its economic impact than traditional telephony. As a consequence, the situation for measuring the impact of broadband can be summarized as follows:

  • It is still in a nascent stage – the first attempts at gathering empirical evidence were made in 2003 in Ontario, Canada (see box below), and in 2005 with national scale data from the United States [1];
  • It is often restricted to the developed and most advanced nations; and
  • There is little quantifiable proof and no internationally comparable data of value [2].

Also, in developed countries the impact of broadband is measured at the individual subscriber level, while most developing countries are aiming at providing broadband at a community level.

A method to measure the benefits from services used via a broadband connection in the developed world is to make the simple assumption that the direct benefit must be higher than the cost of the broadband connection to the subscriber, otherwise they would not subscribe. The assumption is that the subscriber has conducted a personal (or household) cost/benefit analysis and has determined that the net benefit outweighs the cost. This is a reasonable assumption since the market is the final arbiter of value, and demand is the indicator of benefit.

In general, a useful approach to assessing the impacts of ICT is as a system that looks at inputs, outputs, outcomes and efficiencies, as follows:

  • The inputs are mainly the costs of providing broadband to a market including any required complementary investments;
  • The outputs are direct results of the inputs, e.g., a certain number of broadband connections within that market, number of schools or hospitals connected, number of ICT services used, number of on-line training courses provided, etc.;
  • The outcomes are measured through the impacts of the outputs such as: number of school-children now ICT literate, number of hospital staff trained via on-line courses etc.; and
  • The efficiencies of the investment which concerns the cost to produce each unit of output, e.g., USD 5,000 per school to provide broadband connection, USD 500 per on-line course per person, etc.

ChileCompra, the Chilean government’s online procurement system launched in 2000, has saved over USD 70 million in 3 years over the placement of notices in publications, and by increasing efficiency. ChileCompra has increased transparency by giving open access to details of all contracts for good and services procured by the public sector. The system also has a distinct advantage for the private sector since it provides an equal opportunity to all companies to obtain contracts and has helped to balance the distribution of these contracts. While in the overall Chilean economy 80 percent of all business sales are generated by large companies (and only three percent by micro companies and 17 percent by small and medium enterprises), 53 percent of sales through ChileCompra went to large companies and 35 percent went to small and medium enterprises. Twelve percent of contracts were signed by micro enterprises.

Box 1 ChileCompra - Levelling the Playing Field

Source: ITU adapted from EU and ChileCompra

Findings of studies so far

There are studies that have attempted to analyze and measure the economic impacts and benefits of broadband Internet. In general, these studies found evidence supporting the following impacts:

  • ICT sector growth;
  • Productivity gains;
  • Transformation of how individuals, business, government and other parts of the society work, communicate and interact – transforming economic relationships and processes in the private and public sector; and
  • Reduction in pollution (due to reduced travel).

In particular, a study in 2004 [3] analyzed direct benefits to subscribers, the benefits to providers of services, and indirect benefits arising to others as a result of broadband across Europe, and concluded that the potential economic impact of broadband is very significant but that it varies between countries. The variation is mainly dependent on three factors – the size of country (the greater the number of subscribers, the greater the benefit), the cost of transportation (the higher the cost of transportation, the higher the benefits of reducing travel through ICT) and the value of time (the more time is valued, the higher the benefits as ICT saves time).

Another study from 2006 [4], analysing data from the United States on the effect of broadband on several economic indicators, stated that the analysis supports the conclusion that broadband positively affects economic activity, and in particular more rapid growth in employment, the overall number of businesses, and business in IT-intensive sectors.

An illustrative example of benefits of broadband in Canada can be found in the Practice Note South Dundas Township and their broadband experience.

Implications for developing countriesWhile specific academic detail and quantifiable data on the impact of broadband Internet is still being gathered, it is nevertheless evident that broadband development does have significant macro-economic impact.

However, “it is not clear to what extent ICTs have helped to directly reduce major development concerns and particular those of the Millennium Development Goals (MDG), such as poverty, hunger or sickness.”[5] The existing studies have shown that the economic impacts and benefits are variable between countries and dependent on framework conditions. Therefore experiences and models, especially from the most advanced world, are unlikely to be directly transferable to the developing country context.

But with the huge potential of broadband Internet for economic growth and development, developing countries can hardly afford to wait until there are more studies showing clear evidence, as they are already lagging behind the advanced nations in regards to ICT. Most developing countries recognize this reality and are keen to promote broadband Internet development, and many are developing national broadband strategies or policies.

In terms of reaping the benefits of broadband in rural areas and among poorer population groups - which is the concern of a universal access and service policy (UAS), it is not necessarily clear how to proceed, how much to invest, what the required inputs are, what the results are and more importantly, what the required additional framework conditions are.

Nevertheless, some guidance can be given in regards to how each country can tailor the broadband and ICT strategy to their objectives and situation, as follows:

Piloting of broadband ICT projectsPiloting projects is a good approach that allows for the testing of working hypotheses of required input, output, outcome, efficiencies and required additional components. Consider complementary activitiesAny UAS broadband strategy should coordinate with other government ministries and non-government activities and programmes, which can bring a range of complementary activities. These include:

  • Public awareness, training and skills development;
  • ICT deployment in the education;
  • Health sector
  • e-Government;
  • Environment and emissions reduction;
  • Regulatory improvements;
  • Development and expansion of ICT business opportunities; and
  • Expanded ICT and electricity infrastructure.

Pre-conditions for benefits from ICT developmentA certain level of complementary activity and investment should be considered as essential pre-conditions to the effectiveness of ICT and broadband development. For example:

  • Reliable 24 hour commercial power supply is generally required for the use of computers and Internet access to be prevalent and beneficial;
  • School Internet access projects generally need the pre-condition of the national or provincial education authorities to develop a school computerization and networking programme, a computer lab, teacher training, and the specific institution of an IT curriculum, and relevant national content;
  • Promotion of computerization and Internet skills within all government levels, especially including local government and local public health and education institutions;
  • Promotion of tele-working, e-government, e-banking, and conducting other activities on-line to offset emissions generated through travel.

Commercial reality and targeting of growth areas firstUAS deployment strategies for ICT services need to be carefully tailored, with the application of smart subsidy, to harness the potential for commercial sustainability and beneficial impact in critical growth areas. This might entail the following components:

  • Internet points of presence (POPs) in district centres and population concentrations which have administrative, health, non-government, community broadcast and educational institutions, as well as businesses where sufficient demand will exist to sustain ISPs’ business development;
  • Public access centres (cyber cafés, telecentres or information centres) in those same localities, to ensure access to the services for these populations and those of the nearby surrounding communities; and
  • Subsidized Internet access for schools and other vanguard institutions that serve and train the coming generation of active ICT users.

Practice Notes

Reference Documents

4.6.4 Prioritization of USA projects

There are several approaches to assessing or analysing a programme’s socio-economic impact or to prioritizing between several universal access and service (UAS) projects. They are as follows:

    • Qualitative approach;
    • Quantitative approach, using Net-Present Value (NPV) or Internal Rate of Return (IRR) analysis;
    • Comparative subsidy analysis; and
    • Subsidy cost per beneficiary comparisons.

Each of these approaches lends itself to certain types of projects. The qualitative approach might be best employed for pilot projects or slightly larger ICT projects for which it is difficult to estimate hard market data such as costs and revenues. This approach requires some field investigations, and also benefits from using data or experiences, gained in similar projects in comparable countries. While a qualitative assessment will not readily lend itself to a simple ranking, it can help to prioritize alternatives projects, based on an assessment of whether and how closely they meet government and developmental goals. The qualitative approach is presented in Section 6.4.1.The quantitative approach, using economic valuation, outlined in Section 6.4.2, is an approach which has been widely used in telecommunications development projects in the past, for example by Chile’s Fondo de Telecomunicaciones (FDT). However, it is now better understood that UAS telecommunications projects require less economic justification or analysis, but rather a financial estimation of commercial viability and subsidy requirement. A simpler methodology that uses subsidy analysis is described in Section 6.4.3, and a subsidy cost per beneficiary comparison is illustrated in Section 6.4.4. In the light of backbone and broadband infrastructure projects, the NPV analysis might regain some importance, though the precise means of applying quantitative economic valuations to either kind of project is not clear at this time. In some cases it will be most sensible to prioritize projects in order of size of population reached, or required subsidy per targeted person, with the lowest subsidy per person having highest priority. This yields maximum impact at minimum cost. Qualitative approach

The qualitative analysis is best used for ICT services projects where cost and revenues are not easily known or predictable, and other methods are therefore difficult to apply. Nevertheless, it ideally combines a market (financial and commercial) and the socio-economic perspective, i.e., both user needs and demand as well as developmental rationale and impact. Projects that have these two aspects in balance are the most likely to succeed. On the other hand, projects that skip the market side and seek justification based solely on socio-economic and developmental cases can face sustainability problems.

    • The basic issues to be addressed qualitatively are summarised by the following two tables:
    • Table A provides a checklist to define the project’s core ideas, values and approach in a descriptive manner, covering the overall developmental rationale and justification.
    • Table B captures a more market-oriented view of the project’s deliverables, their viability, and the hurdles that may be faced either at implementation or with long term sustainability.

The combination of the qualitative data summarized through these checklists allow an understanding of the project’s impacts.

Source: A rural ICT toolkit for Africa, Andrew Dymond, Sonja Oestmann, infoDev 2003 (updated)

Source: A rural ICT toolkit for Africa, Andrew Dymond, Sonja Oestmann, infoDev 2003 (updated)

/action/image/show?image_id=109 Quantitative socio-economic analysis using NPV

A quantitative socio-economic analysis using net present value (NPV) requires the analyst to:

  • Carry out a normal cash flow analysis using capital and operating costs and revenues, as described in Section 6.2. to calculate a NPV or financial internal rate of return (IRR) in the normal way;
  • Adjust the revenue flows to ‘economic values’ by:
  • estimating the economic benefits received by recipients of the project’s output/ services over and above the price they pay for the services;
  • using this information to derive an ‘economic valuation factor (EVF)’, and
  • using the EVF as a multiplier to convert the project’s financial revenue streams into economic benefit streams. The methodology for calculating EVFs can be found in the Practice Note Economic Evaluation Factor.
  • Adjust the costs using ‘shadow prices’ which reflect the economic value of the various cost items, such as skilled and unskilled labour, imported technology, etc., and eliminate taxes (which are not a cost to the economy) [1];
  • Recalculate the project’s NPV or IRR using the adjusted economic values to calculate the economic performance – e.g., an economic NPV or IRR.

Table C provides a detailed checklist for assessing benefits for projects, if the project lends itself to such an analysis, and assisting with deriving the required inputs for the economic evaluation above. It requires on-the-ground investigations, which can be a component of a demand study. Only the questions relevant to the case under consideration need be answered.

A rural ICT toolkit for Africa, Andrew Dymond, Sonja Oestmann, infoDev 2003 (updated)

In principle, UAS projects with a positive economic NPV but which have a negative financial NPV (i.e., needing subsidy for financial sustainability) should be added to the pool of projects to be offered for subsidy. They may be arranged in order of benefit to cost (benefit to subsidy) ratio or economic NPV, or in any other strategic way.Projects with negative economic NPV at the desired minimum rate of return (i.e., the cost of capital) should not be undertaken, as their returns in terms of development and economic benefits to the wider economy are likely to be marginal.

Practice Notes

Reference Documents Ranking using the findings of a financial subsidy analysis

An increasingly practical way to prioritize projects is through the level of expected private investment compared to subsidy. Some projects require less than 30 per cent subsidization and are therefore likely to generate as least twice the amount of subsidy in capital investment. Whereas universal access and service (UAS) programmes are geared towards extending the reach of markets into areas where service providers may not reach without subsidy, the principle of smart subsidy requires consideration of service sustainability in the medium to long run [1]. Project viability in the context of subsidy requirements can often be summarized as shown in the following table.

Classification of Project Viability

Intelecon Research & Consultancy Ltd.

The methodology for making these assessments is described in Section 6.2 Modelling costs, viability and subsidy requirements. In general, projects and service locations in categories 3 and 4 are the most attractive for UAS programmes since they address areas that are not likely to be served in the short to medium term commercially, yet have a good chance of becoming sustainable and commercially viable in the long run after application of a smart subsidy. They also successfully leverage private investment beyond the amount of subsidy offered. This principle can and should be considered as the most desirable, whether the project is for telephony into new areas, ICT service development or broadband backbone infrastructure, as the principle of commercial viability is generally the only that can guarantee sustainability. Communications projects in these categories also typically carry socio-economic benefits that exceed the level of subsidy provided. The figure below illustrates how financial viability and socio-economic benefit viability can be compared when considering project priorities.

Project Priorities

Intelecon Research & Consultancy Ltd.

Projects in categories 1 and 2 usually do not need subsidy. Even if the geographical areas in this category are not presently covered, they may be contiguous with existing service areas or represent the next logical step in national infrastructure build-out. They will be reached by the market, sooner or later. Projects in category 5 (low financial return and socio-economically weak because they might reach so few people) carry with them the risk that even after receipt of a one-time subsidy, the service provider may not be able to continue profitably in the long run, as the operating costs may be too high and revenues too low because of low population. On balance, if the usage is low, the socio-economic benefits may be considerably below the level of subsidy required, which means they may be hard to justify. However, since this category may apply to the last 3 to 5 per cent of the population but sometimes up to 25 or 30 per cent of a country’s land area, they could justify the government deciding to subsidize them for political reasons, or they could become viable at a later stage in the UAS programme, after the more viable areas have been reached. In the case of telephony access, some areas in category 5 can also be packaged with service areas conforming to category 3 or 4, in order to increase coverage to less viable communities.The final selection of and prioritization of projects, as well as subsidy requirements, will usually involve a blend of qualitative and quantitative regional ranking as well as financial analysis to determine apparent commercial viability and to estimate the financial gap, which is the one-time subsidy required to entice operators to meet UAS targets. The Practice Note Malawi Pilot – Prioritizing and selecting districts for UAS project provides an insight into the use of the above methodology for regional prioritization.

Practice Notes Subsidy cost per beneficiary

Having evaluated the amount of subsidy required (as per the methodology in Section 6.2) in the various projects that make up a universal access and service (UAS) programme, one way to assess, evaluate and rank them is to identify the costs (in terms of required subsidy) per beneficiary or cost per community served. It is appropriate to make cross-country and, more importantly, internal provincial or district comparisons when evaluating UAS costs, in order to help decide how reasonable a programme appears to be.  

Telephony fixed line costs

Although less relevant as a guide for UA telephony competitions today, the table in Section 3.2.1 described the cost of the universal access (UA) programmes in the three leading Latin American countries in terms of subsidy per locality for both public village telephones, telecentres and school Internet programmes.A recent study preparing the first UA pilot project for Mongolia compared Mongolia’s costs with the Latin American examples (since the country’s geography is similar to the Latin American examples in terms of remote regions) and at least one of Mongolia’s programmes involved primarily UA in remote areas. The table summarizes the approximate projected subsidy per person for Mongolia’s 2006 public access network deployment to remote nomadic herder areas. The expected subsidy costs per community were projected to range between USD 5,100 and USD 7,200 per station. The table shows an overall expected average of approximately USD 7.69 per person, though ranging from USD 5.28 to USD 11.86 in the various locations.

Mongolia Subsidy Analysis

Cost-benefit analysis on Mongolia OBA Pilot Program of the Universal Access Strategy, Report submitted to the Global Partnership on Output-based Aid and World Bank, January 2006, Intelecon Research & Consultancy Ltd.

These estimates compared favourably with other fixed network competitions in similarly remote and mountainous areas of the world, of which Chile and Peru were the prime examples. The Chilean competitions awarded subsidies in the range USD 2,256 to USD 12,727 per locality and between USD 2.88 and USD 45.50 per inhabitant (the first round of competition was for terrestrial wireless, while the later rounds included more satellite solutions and were comparable with Mongolia’s).

Experience of Chile\'s UA Subsidy Competitions

Cost-benefit analysis on Mongolia OBA Pilot Program of the Universal Access Strategy, Report submitted to the Global Partnership on Output-based Aid (GPOBA) and World Bank, January 2006, Intelecon Research & Consultancy Ltd.


The following table summarizes the universal access programme of the Telecommunications Investment Fund (FITEL) in Peru. Each project had slightly different features and characteristics. FITEL summarized its projects in terms of the population served and the reduction in average access distance for the populace to reach a public telephone.

Summary of the FITEL Programme

Source: Universal Access Program Assessment Report: Private sector provision of telecommunications services in rural and peri-urban areas in Peru, submitted to World Bank Public-private Advisory Facility (PPIAF), Sept 2003, Intelecon Research & Consultancy Ltd.

Overall, the Peruvian subsidies per site ranged from about USD 5,600 to USD 12,000, excluding FITEL I [1]. These values are comparable with the Chilean subsidies awarded per locality during the same period and the FITEL II and FITEL III projects illustrated the same effect as seen in Chile; the subsidy rising with time, going from more competitive to less competitive conditions and more remote, high cost and poorer areas. The FITEL IV Programme did not have the same access distance reduction effect as the earlier projects because it was designed to provide an additional public phone in locations that already had a public phone.  

Mobile costs

Uganda was the first country to use competitive tendering for technology neutral solutions, won by mobile operators, through its Uganda’s Rural Communications Development Fund (RCDF). The table illustrates that the subsidy amounts in its first country-wide competition were between USD 2.64 and USD 4.29 per person served. The final competitive bids and cost per person were approximately 40 per cent below these projections. Each public access phone served approximately 2,500 inhabitants, while several thousand rural private users were also served on the same networks, which explains the much lower cost per person and illustrates the recent trend towards universal access and service (UAS) since the ascendancy of mobile.

Uganda RCDF Telephony Subsidy Amounts

Cost-benefit analysis on Mongolia OBA Pilot Program of the Universal Access Strategy, Report submitted to the Global Partnership on Output-based Aid and World Bank, January 2006, Intelecon Research & Consultancy Ltd.

The first pilot projects of Nigeria’s Universal Service Provision Fund (USPF) also saw relatively low subsidy costs per person as shown in the following table. These costs included the provision of Internet POPs and public Internet cafés at the Local Government Authority (LGA) headquarters.

Nigeria USPF Pilot Subsidy Estimations

Cost-benefit analysis on Mongolia OBA Pilot Program of the Universal Access Strategy, Report submitted to the Global Partnership on Output-based Aid and World Bank, January 2006, Intelecon Research & Consultancy Ltd.

Internet POP and ICT programme costs

 Internet POP and public access projects in district centres which already have digital backbones serving base station towers, typically cost in the range USD 2 to USD 20 per local inhabitant. The Uganda RCDF has implemented Internet POPs at district centres. These were designed to provide vanguard institutions (e.g., schools, colleges, hospitals, NGOs), government offices and businesses with high-speed wireless service on a commercial demand basis. The average subsidy per POP was around USD 30,000, serving a typical district centre population of 15,000. In addition, one telecentre per district is being funded to a maximum subsidy of approximately USD 25,000. The total average Internet and telecentre subsidy under the RCDF, therefore, averages USD 3 to USD 4 per inhabitant, though the most remote and sparsely populated district centres required much higher subsidies ranging up to USD 20 per inhabitant. Initial costs for remote district centre communications in the Mongolian universal access (UA) pilot programme offered joint subsidies for mobile telephony, Internet POP, public access and free access for the school for three years. Subsidies ranged from USD 16,000 to USD 40,000 for population centres of only 1,000 people. However, from the pilot experience described in the reference documents Output-Based Aid in Mongolia: Expanding telecommunications services to rural areas and Cost-benefit analysis on Mongolia OBA Pilot Program of the Universal Access Strategy (Section 6.3.3), telephony and Internet/ICT competitions are now being bid separately, with the Internet POPs attempted only for the largest and most well-connected centres. The subsidies are expected to range between USD 40,000 to USD 80,000 per centre, i.e., ranging between USD 30 to USD 80 per inhabitant, including all Internet and ICT components. As can be seen in all these examples above, subsidy cost per person and or per locality can vary considerably. These indicators are valuable tools to prioritize provinces or districts or projects, and help to decide how reasonable a programme appears to be.

Reference Documents

Next: Ranking using the findings of a financial subsidy analysis