When considering the issues of using personal computers and underdeveloped countries, it is necessary to consider costs, access, usage and definitions, and one might also consider the most appropriate form of provision of access: personal ownership, schools/colleges or community telecentres.
When considering the issues arising in relation to personal computers and underdeveloped countries, what are the assumptions? The first question that might be asked is, what countries are under discussion? (Is the criterion total economic output, average salary, percentage of individuals in absolute poverty, or in relative poverty, or one of many other possible criteria? Are variations between urban and rural ICT access relevant? Different researchers may produce a variety of answers to these questions.)
Distinctions between external provision of hardware (and possibly some training and support) via political or aid organisations versus self-sustaining community schemes have been made. Cawther suggests the latter may have superior outcomes, resulting perhaps from greater ‘ownership’ of the schemes.
How far does cost affect usage rates of PCs in underdeveloped countries? In urban areas some researchers have found high levels of PC access. However Cawther's review puts internet access in Africa in 1999 at 0.1%. Cawther suggests that a fundamental problem is underuse of equipment provided and discusses approaches that may alleviate this problem and thereby reduce unit costs by increasing access.
What is the most appropriate form of ICT access e.g. individual ownership, school/college provision or community telecentres? Community telecentres and facilitation of community access to school computing equipment are two ways suggested to promote usage and access and drive down unit costs.
What are the real costs of ICT provision to underdeveloped countries? Joris Komen points out in a mailing list debate that initial costs only (perhaps plus three years upkeep) tend to be taken into account regarding aid agency provision of new ICT equipment, rather than total costs over a realistic term of use. (This being in response to suggestions that bulk purchase of new ICT equipment may prove to be more cost-effective than refurbishing of discarded equipment.) Power, connectivity, maintenance, training and support all need to be factored into any cost equations (see Cawther).