Internet: The Great Divider

On the 26th of January 1999 Sugata Mitra and some associates installed a computer in the wall of an urban slum in India to see what the local children would make of it. Soon children were teaching themselves how to work the computer, with no training or assistance. A further experiment in a rural setting had children teaching themselves English through using a similar computer. These “Hole in the Wall” experiments added to the already burgeoning literature proclaiming computers and the internet as the “great equaliser” with the implicit acceptance that information and communication technology is inherently equalising through its impact on income mobility.

Sadly, in the decade that followed this utopian ideal has not been realised, and might even be argued to seem further off than it did then. While rich countries bought into the new technology and have reaped the productivity and other benefits that came with it, access in poor countries has remained a reality only for a small part of the population. Personally I think that  the purchasing decision of internet access acts to bias the distribution of access even further, discrediting the idea that the internet can act as an equalising force if left to market forces.

Purchasing Decision of Internet Access

I think it is obvious that merely having access to a computer with internet can be far removed from actual participation in the information society: Having a library in a town does not mean that everyone in the town will read the books, just so access to an internet cafe does not mean that a person will enjoy the benefits of internet use that will be discussed below.

Further, for a large part of the population having access to the internet is not determined by the digital divide, but rather by the digital decision. While internet cafes and community internet kiosks or internet access at an educational institution surely play a vital role, it cannot be compared to having broadband access at home, a luxury that only some can afford.  In the same way, while having internet access on a mobile phone can surely augment the internet participation of an individual that is already online, I would argue that the access it provides does not make it sufficient as a means to participate as fully in the internet society. Having some internet access is better than having no access, just as having some education is better than no education, but in the globalised market of the internet society, not having the same access could lead to stagnated development instead of catch up growth.

Uncertainty Regarding Expected Benefits

As with all assets, the possible purchase and use is determinant on the cost of the asset and the perceived benefit thereof. The problem with the perceived potential benefits of participation is that it is hard to measure it from the ‘outside’. To use an analogous example: it is argued that a parent’s education is a strong influence on a child’s education, since the parent can explain the benefit of having an education. In the same way there are strong peer effects and spillovers with regards to who participates in the internet, leading to further clustering along income segments. Rich schools, and universities in particular, with computer facilities are attended by the richer students in South Africa, furthering the disparities between rich and poor by introducing non-users from rich homes into environments where users can disseminate information regarding the benefit of having access to the internet.

The Cost of Internet

While the benefits of internet access might not be apparent to the non-user, he or she is faced with some very real costs. The most obvious of these is the monetary costs, firstly of the hardware with which to access the internet, and secondly for the internet access itself.

The increase in processing power of computers has followed Moore’s Law for the last 30 years, which translates into the cost per megahertz dropping significantly over the years. For the most part, however, this did not translate into cheaper computers, seeing as new hardware received new software. The Windows 8, released in 2012, requires one gigahertz processing power, two gigabytes of RAM and at least 20 gigabytes storage space. This compares to 223 megahertz, 64 megabytes of RAM and 1.5 gigabytes of memory for Windows XP that was sold preinstalled on computers until 2010. While there are operating systems that are far less resource intensive, this is merely to serve as proxy for the expectation of what a personal computer should be able to handle. Most computer stores will not sell first hand computers that cannot handle these expected characteristics, which means that a consumer will have to pay at least R2000 for a desktop computer, judging from some online stores I browsed.

Other than the actual hardware, a broadband internet connection is also a costly running expense. According to a report by Ovum, a consultancy, broadband access in South Africa is expensive when compared to other emerging markets; with prices almost double that of developed countries.

Other than these costs, one cannot overlook the non-trivial time cost of becoming comfortable with the internet and staying integrated. The fact that poor people face longer commute times and are more likely to spend time on things like gathering fire-wood means that a seemingly triviality like exploring opportunities on the internet might not be a consideration. On the other hand rich children and middle class individuals with short commutes and comfortable lives are able to expend a part of their time on such entertainment.

The Impact of Access

For those like us with access to the internet society there are a myriad of realised benefits discussed in the literature. Two of the main themes of these benefits, and two of the most relevant to the generation of monetary income, are access to information and social networking.

Access to Information

The first channel through which the greater access to information impacts individuals is through its positive impact on educational achievement. Fairlie, Beltran and Das found that the effect of having a home computer after controlling for various factors was about the same as having a college educated mother on the GPA scores of an American student. Nicholas Spaull found a large and significant effect of computer use on reading and mathematics performance for South African students from both rich and poor schools. In South Africa, where teacher quality cannot be taken as a given, having the resources available on the internet might have an even greater impact than it would have otherwise.

The second channel through which the access to information can impact the earning potential of individuals is through providing market information. Robert Jensen found that the introduction of only mobile phones in the fishing market of Kerala, India, was able to increase market efficiency to such an extent that the mean coefficient of dispersion of prices across markets was reduced from 70% to less than 15%. This almost eliminated the typical five to eight per cent wastage that occurred due to sellers not being matched with buyers. A comparable improvement was found in the agricultural market after the introduction of mobile phones in Niger by Jenny Aker.

Social Networking

Granovetter makes a distinction between strong and weak social ties. Weak ties can act as bridges between clusters of strong-tie social networks. He argues that the fewer weak ties, or indirect contacts, a person has, the “more encapsulated he will be in terms of knowledge and the worlds beyond his own friendship circle.” He takes the argument further to show that these weak ties effectively lead to more diverse information, including on job opportunities or market information, as discussed above. The internet can serve to strengthen existing strong ties, but is particularly good at maintaining and fostering new weak-tie relationships.

In South Africa, where social networks are artificially segmented as a result of the Apartheid legacy, a bridging technology like the internet would possibly be able to have a large impact on inequality. Typically, merely having a wage income would be able to move an individual out of the lower income deciles. As shown in the table below (Based on data from the National Income Dynamic Study), social networks are the most popular and effective means of finding a job.


While this logic is typically applied to lower income deciles, it just as much applies to the higher deciles. Educated professionals are able to seek international parity compensation all over the world with little effort or cost, as we have seen with some of the Media Lab alumni. Since we are aware of international prices for our skills, our bargaining power increases, leading to higher wages.

Further, social networks can serve individuals by both social learning and social influence. With regards in particular to poor household where the parents of children are uneducated, the broader social community on the internet might be able to get children out of this poverty trap by influencing them to seek formal education. This could potentially lead to greater intergenerational income mobility, a big problem in South Africa.

Policy Implications

It is clear internet access can have a big positive impact on income mobility through the channels discussed above, in particular for poor individuals. That said: it is also clear that the current situation is such that the poor are not participating in the internet society, and will not soon be joining it. If the market keeps functioning as it is at the moment, the internet will stay a tool for the rich only.

If this is to be changed, policy makers have two options: market based policies, or non-market based policies. If having a computer in a household can lead to better educational achievement, greater productivity and greater income mobility, it stands to reason that computers should be subsidised or at least not be subject to import duties and taxes. With regards to the cost of broadband, government might play a stronger role in providing ICT infrastructure so as to reduce the cost of internet, especially in areas that do not already have competition amongst suppliers. The building of the privately funded Seacom undersea cable alone was able to lower bandwidth costs by as much as 83% in two years.

To mitigate the uncertainty regarding the benefits, policy makers can act to increase information to consumers through various channels. The education system and government broadcaster can potentially be used to disseminate this information.

While these market interventions could shift some marginal consumers to access, it will still not help the poor to get on the bandwagon. To expand the effects to these individuals’ non-market interventions is required, for example the provision of free internet access and hardware, in particular to children. The One Laptop per Child initiative is the archetypical example mentioned in this regard, which has already distributed more than 2.4 million custom made laptops to children. While there are some that consider this initiative a failure, it has managed a lot. Another such initiative, supported by the Indian government, is the Aakash tablet computer costing less than $50 USD, which the government intends to supply to college and university students.

In essence, the main policy recommendation I would venture to make is a change in the perspective of policy makers of what ICTs are. The famous economist and philosopher Amatya Sen argues that the capability to “participate in public life” and the “power to participate in the social life of the community” are human rights. In a modern world where “public life” and the “social life of the community” are more and more prone to be partly online, having access should start to be seen as human right. Access to the internet should rather be compared to access to a public square than to access to a theme park. In a country where the results of limiting the access and capabilities of some are still all too apparent, South Africa should be a strong proponent of seeing internet access as a basic right to all.

The current disparities between rich and poor will not be fixed overnight by internet access as some pundits seem to suggest. Internet access is not the panacea that policy makers so desire to fix all the social problems that their countries might face. What the internet is, is merely a tool – a potentially very powerful tool – that might serve some individuals to create better lives for themselves if given the opportunity.


  1. 1
    Barend Lutz on Thursday 08 August, 18:30 PM #

    Great article. Especially the policy advice! I agree with your premise of a computer literate equal world built on technology and great minds ;)

    Your policies could be a step in the right direction, but with the deep structural problems in the country, I wonder if there are not more prevalent concerns to take care of first. I find it hard to imagine a townships – flooding each winter, with all the toilet problems and crime; each house now equipped with subsidised electricity broadband and laptop/tablet.

    I think reduced internet and computer costs will only add a few more to the middle class computer user, but will not add the person who has to gather firewood. But perhaps I am being pessimistic and that new middle class user will create more jobs -> profit -> more consumption -> economic growth.

    • 2
      Jean-Luc Koenig on Friday 09 May, 19:34 PM #

      I beg to differ to the above comment.
      The governement funds more than 14million grants and absurd sums of money are mismanaged if not wasted. We have the money to adopt a far more aggressive ICT policy that has the potential to drastically uplift everyone, especially the fire-wood-carrier with his blackberry in his back pocket;)

      Another good article that is far too relevant

Leave a comment

Leave a Reply