Countries can generate significant returns from incorporating ICTs into government reform programs. Singapore's government estimates that every dollar it has spent on ICTs has generated $2.70 in returns. But it isn't easy to do, and for every success there are examples of expensive disappointment. Richard Heeks suggests that failure rates for IT projects in developing countries may be as high as 80 percent. Introducing e-government without corresponding institutional reform seems to be a leading cause of breakdown.
E-government applications which involve direct interaction with citizens or widespread access may be particularly ill-suited to success given the broader environment in developing countries. Take education programs. Computer labs in developing countries carry annual costs of around $78-$104 per student at the level of one computer per twenty students. Compare that to annual discretionary budgets (what is left over after paying for salaries, needed to cover items such as chalk, books and buildings) of $5 per student per year for primary schools in low income countries. Beyond expense, returns don't appear to be dramatic enough to justify diverting resources from alternate, more suitable interventions.
Indeed, public Internet access programs for kids in the developing World like the 'Hole in the Wall' in India get plenty of media attention, what is less clear is that such programs generate significant returns in terms of educational outcomes. As one parent of a child using the hole in the wall kiosk complained:
My son used to be doing very well in school, he used to concentrate on his homework, but now he spends all of his free time playing computer games at the kiosk and his schoolwork is suffering...
Adults, meanwhile, often don't see the need to use the Internet at all --in contrast to other ICTs. There are a number of rural areas in poor developing countries that are under the mobile footprint or have a public phone box and are also near public Internet access points. A survey of three such areas in Gujarat, Mozambique and Tanzania found that seventy percent of people had used the telephone at least five times a year. In contrast, and in spite of the availability of Internet facilities in local towns, less that two percent of people had ever used them. This suggests perhaps questionable returns to public subsidy or provision of Internet access points in such communities. The big problem isn't access, it is finding useful things to do on line in the face of educational, language, financial and institutional barriers.