Imagine a large room, filled with boxes containing machines that are switched on once every two years to perform a single function. They are single function machines, used to count votes and produce an accurate election result within hours of the polls closing. Florida? Washington DC? No, these relatively expensive pieces of technology are housed in a large office in the Regional Electoral Tribunal of Parana, Brazil. Every six months officials are charged with taking the machines out of their boxes and turning them on to make sure they still work, before turning them off again and carefully replacing them in boxes on the shelf. Given the uncertainties of the electricity supply, a back up of car batteries is used if the supply fails on election night, roughly every two years.

The officials who show visitors around are rightly proud of a system that can reduce voter fraud and produce an accurate count within hours of the polls closing, particular in a country with a huge electorate, where voting is compulsory and electoral democracy re-established within recent memory. But the skeptical observer is forced to wonder if electronic counting of votes is really the answer to the issues of corruption, vote rigging and buying that plague Brazilian and other elections. Surely the problems lie much earlier in the process, when the votes of poor or illiterate people are ‘bought’ – not at the polling station when people turn up to cast their ballot.

At a time when the Internet backlash is still underway, the dot-com boom and bust has been and gone and even PC penetration at home in the UK seems to be peaking, public and third sector faith in the transforming power of Information and Communications Technologies (ICTs) seems undimmed. And nowhere is this faith more profound than in those concerned with developing countries. The United Nations Development Programme, in common with most donor organizations, now has an explicit commitment to the ‘digital agenda’ and to bridging the North-South, ‘‘digital divide.’’

But the divide between the North and South is no more ‘digital,’ than the divide between poor residents of East London and the wealthy inhabitants of Surrey is ‘digital.’ One group may have access to the Internet at home and work and the other group may not, but what divide them are income, education and life chances. ICTs undoubtedly have a role to play in closing all those gaps – both in inner London and the developing world, but they are not a total solution, not a substitute for other interventions and not a way to let donor organisations or governments ‘off the hook’ of more intractable problems.

The current mania for ICT-led development is often branded ‘e-governance’ – a complex process of institutional restructuring and transformation, involving government and its partners. What we often see however is less transformation and more automation, less process reorganisation and more putting a shiny front end on long-term systematic failure.

So why is ICT-led development so fashionable and what are its abiding myths? One is that ICT offers substantial savings in government processes. This is not necessarily true even in developed countries, where the automation of individual processes from parking fines to benefit applications has cut some costs, but no-one has yet produced an accurate, overall figure for ‘e-government’ savings in the round. In countries where labour is very cheap and technology sometimes expensive, this even less likely to be the case.

The maxim “Information is power” is often blithely quoted in discussions of the “digital divide”. But access to information should not be confused with access to power. The potential for ICTs to help create a better informed and educated populace, who in turn become a more demanding and scrutinising electorate, is undoubted. Examples of individual successes abound – for instance the email campaign that helped topple Philippine President Joseph Estrada. But for the majority of the world’s poor the ‘information’ contained on the Internet is difficult to understand (not just because most of it is in English), irrelevant to their daily lives and expensive to obtain.

The possibilities of technological ‘leapfrogging’ for developing countries are also a staple of speeches by Western politicians. It is certainly the case that developing countries can often bypass failed technologies or forge ahead with new ones such as wireless communication in rural areas or e-learning (the six largest distance learning universities in the world are located in developing countries). But ‘leapfrogging’ suggests moving past an incumbent and the world rankings of GDP are proving remarkably resistant to rapid change. All this is not to suggest that there is no role for ICTs either in economic development or in institutional reform in developing countries. There is clearly an important role for both. But if the experience of e-governance in the UK has anything to teach us, it is that technology has to be integrated into broader policy areas, it does not work as an add-on. In other words, there is no e-government, just good government enabled by ICTs, no e-democracy without an underlying healthy democracy and so on.

We need better, more systematic evaluation of e-governance projects that attempt to capture both costs and benefits in the round and, in the case of developing countries, do not start from the assumption that a technological solution is always cheaper and more effective than a purely human one. We need projects that reflect the priorities of areas where they are deployed, not those of donors. Too many e-governance projects are funding-driven and represent off-the-peg solutions developed for different circumstances. Electronic service delivery (ESD) should not necessarily be a priority in countries with very low Internet penetration for example.

Finally, we must not kid ourselves that closing the ‘digital divide’ is synonymous with closing the health, wealth or education divides. If we continue to push technology at people who lack for the most basic essentials, a backlash is inevitable. As Kunda Dixit, editor of the Nepali Times said at a conference on technology and development earlier this year, “Before sticking a computer into a school, how about building a roof over it, staff it with competent teachers who are not absent half the year, ensure there are more girls in the classrooms, make sure the children are adequately nourished and not physically and mentally stunted because they don’t have enough to eat.”