Genuinely successful, high-yield e-government projects are still difficult to find and even more difficult to measure, even five years after Richard Heeks’ classic paper, “Most E-government-for-Development Projects Fail”.[i]  Heeks claimed that 85% of e-government projects in developing countries don’t achieve promised goals, of which 35% are total failures, 50% are partial failures, and only 15% are successes. Soon thereafter, the World Bank reported that its sectoral based projects with information and communication technology (ICT) components had an “alarmingly high failure rate” with 50% suffering disputes and 80% requiring contract amendments[ii]. With a total ICT component portfolio estimated currently at US$7.3 billion and less than half of the projects rated “satisfactory” by the Bank’s own evaluation system, it would seem that many ICT projects can waste precious resources that could be devoted to competing development needs.

Our best guess is that things have not improved for two reasons. First, today’s e-government systems are increasingly dependent upon inter-agency cooperation. “Agency-centric”, or “silo” approaches to systems reform, and “supply and install” projects seldom work. For example, vehicle licensing systems often require the interoperation and cross-agency cooperation of transport, tax, and finance systems, perhaps with links to private sector financial institutions, insurance, motor agencies and the police. Second, many projects depend on trade facilitation systems that involve interoperation between customs and traders for all government agencies and permit-issuing authorities, border control, airports etc. In some countries there may be over 20 agencies involved in permit-issuing and over 30 stakeholder systems involved in implementing an ICT-based “single window” trade facilitation initiative. Increasingly, e-government projects are bumping up against technology policy issues that need to be addressed at the national level. For instance, interoperability policies often need to be addressed in order for proprietary systems at local and national level government units to “talk to each other.”

Secondly, e-government boils down to people using computer-aided business systems in ways that make government more accessible, effective and accountable. Every stakeholder eventually asks either “What’s in it for me?” or “Why should I be involved in this e-government project?” In this article we present three examples of successful e-government deployments that have succeeded because they have addressed policy issues upfront and/or can easily answer the crucial cui bono question. The three cases include: Nangi Village Wireless Outreach (Nepal), Nemmadi (Karnataka, India), and MK Connects (Macedonia).

Case 1: The Nangi Case - If You Build it They Will Come

The deployment of ICT services in the Himalayas is a daunting task, and particularly in the high mountains of western Nepal near the Annapurna range, where the challenges are especially acute. Nevertheless, in September 2003 the village of Nangi, with 800 inhabitants and a tradition of doing almost all chores without automation, was connected successfully to the worldwide Internet[iii]. This was the culmination of a five-year planning process--a cooperative venture among several villages to set up a group of relay stations, starting with Nangi and extending to Pokhara, the nearest Internet hub. While the villages were primitive, the technology used was modern, rugged and reliable. A dozen access points were connected to the dial-up ISP in Pokhara. Gradually, more villages and additional regions have been added to this network and a variety of Internet-based services have been implemented with considerable success: telemedicine, online classrooms, improved communication and many more. The service has been so robust and continuously successful that a member of the Scotch parliament complained that the Nangi Project, developed by local volunteers, was delivering a higher level of service in the mountains of Nepal than that available from British Telecom in the Scotch Highlands.[iv]

Since the original implementation five years ago there has been a surprisingly rapid deployment of additional capabilities and applications. From one village, the number of installations has risen to 22 sites, including ten schools, a library and a hospital. The basic deployment approach has been extended to several other regions within Nepal, and support both from government and external agencies has increased substantially over the years. As the number of sites increases the applications are going up proportionately. A visit to the local homepage of the website “nepalwireless” gives more examples of Nepalese villagers using the local network. For example, Gandaki College of Engineering and Sciences helped develop an e-commerce application using open source software. Villagers use this application to provide information about the products they have for sale. There is also a local bulletin board application which allows the posting of urgent information, news, announcements etc. This bulletin board also has user advertisements, attracting buyers and sellers. Also Kathmandu Engineering College has developed telemedicine applications which have been successfully connected with Dolakha Hospital, located about five hours bus ride away to Eastern Nepal from the Kathmandu Model Hospital. At the same time VOIP (Voice over IP) applications using open source software have also been developed for handling voice traffic, providing significant revenue for the project.

As of late 2008, the list of new applications seems endless – interactive content based on school textbooks and  curricula are being developed in the local language with the help of Open Learning Exchange – Nepal (OLE-Nepal) using open source software for the “One Laptop per Child program” for second and sixth grade students in science, math, and English. Useful applications for villagers in the local language are also being developed by OLE-Nepal and have been made available through e-library. The application is platform-free and can also be used in Microsoft Windows. To provide more services to the rural population and generate revenue, the project is working now as a business partner with thamel.com, an e-commerce business developed by a Nepalese business entrepreneur, to provide money transfer service in the villages for the villagers working abroad and credit card transaction service through virtual ATM machines to the tourists traveling in the Annapurna region. With thamel.com it also has started providing services to Nepalese citizens living abroad to send presents such as flowers, electronic goods, birthday cakes, and many other items to their families living in Pokhara and in its vicinity. The picture below shows the network deployment after the initial installation at Nangi. (Nangi is shown as NGI and Pokhara as PKR.)

Network Proliferation after Nangi Initial Installation

There are several aspects which make the Nangi case unusual. First of all, it was not initiated through government, NGO or multilateral interventions. Second, it was managed entirely by Nepalese local leaders, with assistance provided through skillful selection of outside helpers, mostly volunteers, from developed nations. There was no role of the USAID, UNDP or World Bank in the project. Third, the hardware used was top-of-the-line, and not a compromise among cheap or aging technologies. Fourth, the cost of the project was relatively modest, since there were no “middle-man” charges and because several equipment suppliers were so impressed with the Nangi concept that they gave large price concessions. Perhaps the most unusual aspect of the case is that the entire effort was not driven by an application, like licensing, taxation or finance. Instead, the idea for Nangi was, “if you build it they will come”, like the theme from the film “Field of Dreams”. The connection was established, and the applications followed. The leader of the original project, Mahabir Pun, received the prestigious Ramon Magsaysay Award, sometimes dubbed the “Asian Nobel Prize”, for Community Leadership last year in a ceremony in Manila in August 2007.[v]

Case 2: The Nemmadi Case - For-profit Rural Business Centers Diversify into Service-led Employment and Village BPOs

The Microsoft Schools of Government project teaches that the main aspects of e-government are: service delivery that can be addressed through technology are: inter-agency collaboration; the strategic use of technology for a killer application; security and privacy; data sharing; governance of cross-organizational initiatives; document and data management; and standardization of internal systems.[vi] The next example, from India, is a composite of these aspects. The rural business kiosk project, known as “Nemmadi” in Karnataka state of India has been successful for two reasons: (i) it is providing financially viable rural kiosks that are realizing “development” benefits in terms of delivering government-to-citizen services (G2C) and business-to-citizen services (B2C), and (ii) Nemmadi is increasing employability prospects through education and rural Business Process Outsourcing ( BPO ) services. The project was introduced in 2004, so it is too early for a comprehensive impact analysis of the Nemmadi model, yet early indications are that this public/private partnership may distinguish itself in bringing network infrastructure, services and job opportunities to “the bottom of the pyramid.”

Nemmadi Kiosk Delivering G2C and B2C Services 

The Nemmadi services are based on transactions per use, which allows a business based on volume rather than the high up-front costs that other telecenter models have difficulties supporting. The picture below is of the Nemmadi center at Indalawadi in Karnataka, an example of the Rural Business Center (RBC) servicing local needs through access to e-Literacy programs.

Nemmadi Center Providing Access to e-Literacy Programs

The Karnataka government facilitates the Nemmadi project as part of a build-own-operate (BOO) model. “Nemmadi,” meaning “peace of mind,” in Kannada language, was coined by the Government of Karnataka. The role of the public sector is limited to providing data and strictly enforcing a competitively bid, consortium-based service level agreement (SLA) that has been established between the state government and three Indian private sector companies—Comat Technologies, 3i Infotech, and n-Logue Communication Ltd. The SLA specifies daily hours of operation, maximum wait time for services, and other metrics.

The business model for Nemmadi is based on a combination of support for scaling up the number of people reached through telecenters—a quantitative scale-up—increasing the scope of activities—a functional scale-up—and improving the organization’s effectiveness and efficiency of its core activities—an organizational scale-up. One of the lessons of Nemmadi is that the one without the others is not sustainable. A company or consortium that has strong values—e.g., to provide universal access or e-literacy and e-governance services for people in rural areas—but is badly run, without proper attention to translating values into profits, will plainly not do well. A combination of a firm commitment to providing multiple services and strong commercial competence gives a good chance of success.

Offering multiple services allows capital and operating expenses to be amortized and spread across different services. The rural business centers are basically a network that delivers G2C services during the day, and B2B and B2C services after-hours in education, and BPO services. This is the functional scaling up. Comat is now also providing convenience services such as travel ticket booking, mobile top up and utility bill payments.

The G2C services include accessing the Record of Rights, Tenancy, and Crop Inspection (RTC, land record) data online and printing out the RTC , food coupons, issuing caste and income certificates, updating crop data, and utility payments. Fees are derived from a complex revenue model in which government has fixed different charges for different services, and these vary depending on the geographical location, season, service and extent of access to the back office databases. Issuing the RTC electronically has eliminated the village accountant who created, modified, and supervised handwritten manual records and the associated bribes.

What makes Nemmadi different from most models that currently exist is that its services are based on transactions/per use, which allows Comat to build a business based on volume rather than the high up-front costs that other models have difficulties supporting. To illustrate: 20 million land records divided by 167 taluka offices are 120,000 records per office. Priced at Rs 15 (US0.37) per RTC certificate, this averages Rs 1,800,000 (US$44,830) per office in revenues. This is sufficient over time to cover operating costs, provide a modest return to state revenues, ensure good service levels and result in a profit-making proposition for Comat. For other private services, such as ticketing services and topping up pre-paid mobile phones, the revenue model includes origination fees and transaction costs.

From a few experimental pilots, the project is increasing the number of kiosks and  people reached through the RBCs. As a measure of the success of the Nemmadi model, Comat, in collaboration with other state governments is rolling out another 5,000 RBCs in 2008 that will serve a population of approximately 1 million people.

Case 3: MKConnects - Macedonia School Internet Connectivity Case

The third example is the MK Connects case, linking hundreds schools and other institutions to stable Internet service in Macedonia. Like many successful IT projects it evolved easily across goals, technologies, and sponsors. The initial impetus came from a visit by the president of Macedonia to the People’s Republic of China (PRC) in 2002 and a subsequent offer from PRC of broad-based computer assistance for the school system in Macedonia. With the early help of the PRC and Microsoft, other interested donors became major participants like: Government of Macedonia (GoM), its Ministry of Education and Science (MoES), and the U.S. Agency for International Development (USAID)[vii].  

The key questions addressed from the outset were: (i) What are ways to provide low-cost and sustainable broadband access to elementary and secondary schools dispersed throughout a developing country?  (ii) How can policy and regulatory changes allow for use of new wireless broadband technology solutions and provide broadband connectivity not only schools but as an initial footprint for connectivity to a larger client base?

Hampered by a telecommunications monopoly, 47% owned by the Macedonian government and 53% by a Hungarian corporation, and laws which inhibited the “legal” growth of the Internet market and Internet Service Providers, the challenge in Macedonia was to develop an Request for proposal (RFP) that would not outright deny the participation of larger providers but would encourage vendors to be innovative in their approach and minimize the participation of MakTel[viii] in the overall scheme of broadband connectivity.

The winning bidder proposed a broadband solution for 430 selected schools. The underlying premise was that the eventual broadband pricing to those schools will be tied to the lowest cost broadband solution commercially available when the project ends. This meant that if a 256K broadband solution is available from another provider for $75 USD a month (unsubsidized by any government entity) then the schools would pay a similar price. This forced the winning bidder to place an emphasis on commercial, governmental and home users in order to balance out expenses. In other words, while the schools may provide 40% of income and utilize 60% of the resources, the other markets may provide 60% of the income and utilize 40% of the resources. The vendor building out the network was also required to add three rural schools for every one urban school location. Based on cost averaging, once the vendor builds the infrastructure to a distant location in Macedonia it would be no more expensive to deliver a packet from that point than from an urban center.

The cost of this project was substantial, and the results excellent. To leverage the computer donations from PRC and Microsoft, USAID spent an additional $1.3 million to install 5,300 computers and 300 printers and another $400,000 for training 4,000 teachers. This amount does not include administrative time for any of the participating donors. In addition, it became clear that security insurance, and maintenance costs needed to be considered more carefully, leading to a more sophisticated approach to gaining agreement at each school for responsibility and for upkeep and administration of the network.

Leveraging private sector investment, The Academy for Educational Development/dot-ORG, the contractor selected in 2005 to manage the project for USAID, helped create the world’s first national wireless broadband network. The MK Connects project officially ended on December 31, 2007. In November 2007, MK Connects was named a finalist of the third annual W2i Wireless Communities Best Practices Award. Other measures of success were: dramatic increases in overall Internet use (now at 47%), household computers (32%) and household connections to the Internet (17%). Over half of the current Internet users in Macedonia, many of them young people, signed on during the past two years.

A key advantage of this successful deployment of Internet to schools in Macedonia is its scalability. The aim was to empower both urban and rural areas so the resulting numbers of users throughout the country increased the probability of sustaining the initial systems. With hundreds of systems in place it is easier to trouble-shoot and add features. The contractor for the project, On.Net, reduced the unit costs significantly.

Summary: Why These Projects Were Successful

It is regrettable that the percentage of successful E -Government projects worldwide is not higher. Heeks’ estimate of only 15 percent may be too severe. But the examples we describe here are in that special group. What characteristics can be found in these three successful projects that distinguish them from the rest? First, as mentioned in the introduction, it is crucial to achieve interoperability - both technical and administrative. The computers have to talk intelligently to each other and so do the affected agencies and stakeholders. Even in the rural Nangi case, it was essential to have interoperability at the hub station as well as the dozens of small villages connected to the wireless network. Second, each of the three cases had a very specific focus with respect to clients, region, and technology use. Third, there was an organized, cooperating cadre of donors which were able to coordinate their activities toward the major goal. Fourth, there was good publicity in various international media outlets throughout the life of the projects, not only increasing their credibility, but also ensuring greater donor support. Fifth, there was recognition that equipment donations involve added  costs that must be planned for. In the Macedonia case about $2 million additional investment was needed to leverage the generous donation of computers and software from PRC and Microsoft. Sixth, in each of these examples it was possible to measure results with reasonable clarity and accuracy. In the Nepal case, arguably the simplest of the three, the number of villages connected and the hours of use was a simple but realistic metric. Finally, as mentioned in the introduction, it is crucial for each stakeholder to appreciate the cui bono question. Who benefits? In the three cases we have cited the beneficiaries are citizens who are assisted by the systems, not bureaucrats or faceless agencies.