Proponents of OLPC as a solution for underdevelopment seem to be operating under two main fallacies. The first is what we might call, drawing from Atanu Dey, the arithmetic fallacy -- i.e., they are overlooking the most basic arithmetic which shows that purchasing one laptop per child is way beyond the means for developing countries (and, even more beyond the means, when factoring in expenses like electricity, bandwidth, repairs or replacement, etc.)
There is a second fallacy, which is also very important, which we might call the miracle transformation fallacy -- i.e., the notion that, if we could get little green laptops into children's hands, it would miraculously transform their lives. This fallacy falls within an approach known as "media determinism," the notion that a particular media or technology will automatically have a certain effect no matter what context it is deployed in.
However, a long history of experience with all media indicates that they are heavily influenced by the context of their use.
One good example of this is the TV show, Sesame Street, which was designed to foster greater social and educational equity in the U.S. by bringing quality educational programming into the lives of young children who lacked other educational resources. However, subsequent research showed that rich kids benefited much more from Sesame Street than poor kids, in part because they had the parents and other family members who sat and watched with them and discussed the show in ways that maximized what children got out of it.
This kind of "Sesame Street effect," in which the rich disproportionately benefit from reforms targeted at the poor, is also seen in educational computing. A national study by Wenglinsky in the U.S. indicates that, overall, there is a consistently negative interaction between frequency of technology use and test score outcomes in mathematics (at both the fourth and eighth grade), science (at both the fourth and eighth grade), and reading (at the eighth grade). However, this is not true for all students. He found that the single strongest factor determining whether students received a positive benefit from school use of technology was the children's socioeconomic status (SES), with low-SES children tending to see test scores go down with computer use and high-SES children tending to see them go up. Wenglinsky's data suggests that this is in part due to the differential ways computers are used by teachers of high- and low-SES students.
Well, then, what if we bypass the teachers and just put the computers directly into children's hands? There is also some research on this question as well--and it's equally depressing. Though there are conflicting reports on whether the overall effect on children's academic achievement of access to a home computer and Internet is positive or negative, there is unanimity that any effect is highly differentiated by race and SES, with white and high-SES children getting the most benefit from a home computer, and Black and low-SES children either getting little academic benefit or even suffering lower academic outcomes after getting access to home computers and the Internet (see, e.g., this study at Duke University.) Again, the reasons likely are related to the types of parental or family support for using a computer as a learning tool, rather than as a distraction from study. A study in Romania also confirmed negative effects on academic achievement of home access to computers.
In spite of these disconcerting research findings, I remain optimistic about the potential benefits of laptop use by children. However, I do not think these benefits will come about simply by placing laptops in childrens' hands. Rather, laptop programs need to be part of complex educational reform efforts, involving teacher training, curriculum development, pedagogical support, etc. I also believe that it is the least realistic to carry out these broad reform efforts in the world's poorest nations, which lack the other resources necessary for their success (electricity, bandwidth, teachers who are skilled at using technology, etc.) Yes, we can and will get there eventually, and we need to keep that on our agenda, but we can't leap there suddenly.