BOLD 2003: Development and the Internet

Edited transcript of Andrew McLaughlin, posing a hypothetical argument about the Digital Divide

Let me start off today’s session by making the case against caring about the digital divide, at least in that sense. As you probably will guess, all of us that are up here think something different than what I am about to say, but let me start off by making this case.

The fundamental argument I want to pose is that the cost of providing universal Internet access is not worth the benefit. Africa, to take one case, is desperately poor. GDP comparisons only begin to capture the disparity. In the U.S., gross domestic product is something like US$33,000 a person. In sub-Saharan Africa, which means everything below the band of Morocco to Egypt, gross domestic product is less than US$500 per person. And if you don’t count South Africa, it’s about US$300 per person per year. Connectivity costs, by contrast, are the highest in developing countries. Much of the population in those countries is rural; electricity is unreliable and expensive; it’s very hard to get communications to people.

Let’s take an example: At an off-grid, Costa Rica tele-center—in other words one that has to rely on satellite connectivity to get its Internet access—the annual cost is roughly US$10,000 per Internet-enabled computer per year, in order to keep that tele-center hooked up. This is from a World Bank study, by the way, or, I should say, a paper by a World Bank economist. Compare that cost per connected computer to the fact that the average person living on a US$1 a day—and there’s 1.2 to 1.5 billion people in that economic band, depending on who’s counting—that average person living on US$365 a year or less in income, will, if they have access, spend, on average, something like $10 a year on communications. So if you were to divide this out and try to figure out what it would cost to give that average dollar-a-day person an hour of Internet access a week at a tele-center, if you do some math (and I’ll warn you in advance that you can fight over these numbers, and I think I would fight over them, but anyway this is one relatively reputable economist’s set of calculations) it’s about $50 a person a year to provide somebody at an off-grid tele-center an hour a week of Internet connectivity.

How much is US$50 per person per year in low income countries, the least developed nations? US$50 per person per year is about 10 times typical public spending on health per person. In the cluster of least developed countries health expenditures generally run about US$5 per person per year. Also, US$50 is about 10 times the typical amount of discretionary education spending per primary student. So it’s a lot more money than gets spent on health or education for the average citizen or the average primary student in these least developed countries.

Question: Is it worth it? If you’ve got US$50 per person per year to spend, is it really the right thing to do to give them an hour a week of Internet access at a tele-center, or should you instead be boosting spending on health care, spending on basic services—sanitation, primary education? The barriers to Internet use of course go far beyond connectivity. A lack of education is a fundamental barrier. For example in Ethiopia, 65% of the population itself is illiterate, as are most of the people who are earning a dollar a day or less, worldwide. For those 65%, unless you teach them how to read, Internet access is kind of a meaningless thing anyway. Moreover, there’s a lack of technical skills to maintain Internet infrastructure and connectivity. It is a sustainability problem.

There are language problems. Many of the very poor don’t speak a major global language that’s widely represented on the Internet. For example in Nigeria, there are 17 million people who speak only Igbo. I bet many of you have never even heard of Igbo. There are 17 million people for whom that is their language. And somebody at one of these international institutions snooped around for a while to try to find as many websites in Igbo as possible, and they came up with five websites, one of which was some translated religious tracts, one of them was a two-page, very crude Igbo-to-English dictionary, and the other three were random pages that people had put up about…I can’t even remember what…in Igbo. So that’s a whole other problem that Internet connectivity, communications connectivity, doesn’t begin to address.

And finally, of course, there’s a lack of payment mechanisms, not just to buy things from Amazon, but just to pay for connectivity itself. Typically people aren’t carrying around credit cards.

So wouldn’t it be a misallocation of resources to pursue universal Internet access, as the G8 called for, and as the UN General Assembly has also called for? If money is available, isn’t it better to allocate it to nutrition, basic health care, clean water and sanitation, education, roads, and so forth?