Chapter 9


Five computers connected to the Web 24X7 were stationed in a booth in a Delhi slum in 2000, their monitors protruding through holes in the wall, along with joysticks and buttons that substituted for keyboard and mouse. Children interacted with the computers without any instructor or mentor. The process allowed children to learn at their own pace and speed. They quickly learnt to click, drag, cut, paste, launch some programmes, and get onto the internet.

The project came to be known as Hole in the Wall, and was celebrated worldwide as a groundbreaking model of digital inclusion.

It quickly became apparent, however, that the children were spending most of their time drawing or playing computer games. There was little interaction with the internet because it functioned poorly, no computer/internet literacy programs had been made available, and little content was available in the only language the children knew – Hindi. There was no community involvement in the running of the booth either. Parents began to question the value of the experiment because it seemed to end at entertainment for the children (Warschaeur, 2003).

Hole in the Wall illustrated that reducing inequalities in access to and use of digital technologies is not just about the provisioning of hardware and software. Digital inequality is embedded in a complex array of factors encompassing physical, digital, human and social resources and relationships. The technological context cannot be looked at in isolation.

The lessons of Hole in the Wall are relevant to the task of internet inclusion in India today when the focus of the state, policymakers and the telecom industry appears to be on macro-level issues of technology access, and when the increasing number of consumers with devices and data connections is being seen as internet inclusion.

This chapter discusses the need for a different approach to the digital divide, and makes some recommendations for the internet inclusion of marginalised communities in urban India.


‘Internet in India has now become inclusive, which augurs well for the industry and society at large,’ stated the representative body of internet and mobile value-added services, Internet and Mobile Association of India (IAMAI) in September 2015 (IAMAI, 2015). IAMAI put the number of internet users in India at 352 million by June 30, 2015, more than the entire population of the US (PTI, 2015), and declared that the country would have 500 million internet users by 2017 (Tech Desk, 2015).

This conflation of number of internet users with internet inclusion needs to be questioned.

Our study, one of the first to examine the extent of the digital divide in a fast-growing Indian metropolis, provides a reality check. It points to a digitally divided city.

The CCDS study reveals that if at least five out of 10 residents of Pune city as a whole are online, as suggested by industry estimates (IMRB-IAMAI, 2014, Mumbai), less than two out of 10 individuals in the low-income areas of the city surveyed use the internet. This low level of penetration is despite keeping the definition of internet user wide enough to include not just data subscribers but everybody who has used the net anywhere, on any device, over the last three months.

Our study has also pointed to the ways in which digital inequality overlays economic and socio-cultural exclusion in the low-income, resource-poor and diversely-literate communities under study:

  • The marginalised geographical areas under study pose infrastructural constraints. Network connectivity and coverage in these slum settlements is poor. No wired broadband services are offered. No public Wi-Fi networks or public access points exist within these areas.
  • There is a clear correlation between economic deprivation and internet exclusion. Households at the lower end of the wealth index are much less likely to be online than those at the higher end. Nearly 55% of internet users belong to the upper fourth and fifth wealth quintiles. Households that can afford enabling infrastructure including computers, dongles, smartphones and feature phones are far more likely to be online. The absence of an internet-ready device is reported as a major barrier to access, especially for women, the most disadvantaged group in these socially-excluded communities. Economic constraints make it difficult for users in the study locations to afford computers, and those who do have computers cannot get wired broadband connections, even if they could afford them.
  • Education is also strongly correlated with internet use. Those who have no education, or only primary education, are least likely to be internet users. Households that have even one member who is in the final year of school or has completed schooling are much more likely to be online.
  • Education is also strongly correlated with patterns of use: users with higher levels of education report use of the internet for purposes other than entertainment and social networking, while the handful of respondents who are users despite having no education mainly use the internet for entertainment and social networking, which can be managed with audio-visual, rather than text-based content.
  • Lack of skills to use computers and other digital devices is reported by non-users as a major barrier to internet use. And it is the poorest, the least-educated, the older age-groups and women who are most likely to lack the opportunities to acquire ICT skills.
  • Women, who face discrimination in the male-dominated environments under study, lag far behind men in ownership and use of computers, mobile phones and internet. Women who do use the internet don’t have the autonomy to log on when, where and how they want to. Infrastructural barriers and economic barriers also hit them hardest, because they neither have the financial capital to buy their own devices nor the freedom to visit public access points outside the settlement.
  • There are major age barriers, with only 7% of users above the age of 35. The few users above 35 tend to be those who have stable livelihoods.
  • Finally, and significantly, there are the attitudinal barriers in these socially-excluded environments. Many non-users, who have always suffered information poverty, believe the internet offers nothing that can be relevant to their restricted lives.

It’s a vicious circle.

The inequalities in society cause an unequal distribution of resources, including digital resources. So, those with the least income, education and autonomy have the poorest access to digital technologies. This lack of access to an essential resource for modern life causes information poverty, which leads to unequal participation in society. Unequal participation in society reinforces social inequalities (van Dijk, 2013, p 33).

Over the two years of our study we have observed the diffusion of internet in the study settlements accelerating. The point this research flags is: greater internet penetration and use will not automatically ameliorate digital inequality.

The CCDS survey on barriers to internet access reveals two levels of digital inequality:

  1. The first-level digital divide between the internet haves and have-nots, users and non-users.
  2. The second-level digital divide, also called the capability divide, which affects those who are users, but with differential skill levels and use of internet applications.

The first-level digital divide results mainly from lack of awareness about the internet, absence of the technological capital to connect (computers, mobile devices capable of internet access, public access points within settlements), economic barriers that make it difficult to afford hardware and data services, skill barriers with low levels of education and absence of ICT training and support, and socio-cultural barriers that limit women’s access.

The capability divide is more complex, and often overlooked.

Our study reveals that while internet users have the operational skills to connect on low-end mobile devices, at cybercafés or workplaces, the majority use the internet for entertainment and social networking, and to a lesser extent for purposes such as educational research, searching and applying for jobs, or searching and applying for government services and entitlements. Many of these users lack the autonomy to access the internet whenever and wherever they choose. They also lack social support in getting connected or trouble-shooting. The absence of social support makes it difficult for them to learn to use the internet for more than entertainment, social networking and simple searches.

These are the users who, wooed by low-cost offerings by telecom hardware and service providers, and aggressive marketing by e-businesses, are amongst those swelling the ranks of internet users in India. Telecom companies are tying up with internet giants like Facebook to offer free bundles of apps and internet services to mobile telephony subscribers. Though technically all these subscribers have access to the internet, whether they can be included amongst the information haves – and beneficiaries of the internet ‘revolution’ – is debatable. They may more appropriately be termed the ‘have-less’, a term first used to describe low-income, migrant, working-class or unemployed urban populations in China (Cartier et al, 2005). These internet users may spend several hours a week on online entertainment and social networking but have no knowledge of browsers, e-mail accounts, or other diverse uses of the internet. Qiu, who first coined the term ‘have-less’, believes it would be a serious mistake to equate this group with the haves (Qiu, 2009).

Any long-term and meaningful digital inclusion policy then must aim to provide equal internet access to all.


Digital inclusion policy is generally focused on overcoming uneven access to technology. But if ICTs are to be seen as a tool for the empowerment of people, a way to bridge the other divides of society, then digital inclusion policy must be organised around access as well as adoption and application (Becker et al, 2012, p 9). ‘Access principles address the infrastructure a community needs to have in place in order to provide opportunities to benefit from digital life. Adoption principles look to overcome individual barriers that make use of broadband technology less likely, even when access is available. Finally, the application principles look at specific purpose areas where the thoughtful deployment of broadband technologies can enhance the economic success of communities and the lives of their residents.’

While the equality-of-access orientation centres around physical access to digital devices, the equity-orientation focuses more on increasing the likelihood that all users, especially the poor and disadvantaged, can take advantage of the information available online to improve their day-to-day lives. ‘For example, an equality orientation might be satisfied with the presence of computers in all public libraries and public schools. An equity orientation, however, focuses attention on increasing awareness, amongst all members of the population, that information is available, where the information exists, and how to navigate websites to find information when it is needed,’ (Gomez, 2015).

In India, where, as this survey shows, physical access to technology is still a major hurdle, we will need a combination of the equality-of-access and the equity-orientation.

India has no explicit policy for bridging the digital divide. But the central government’s Digital India programme promises ‘Digital Infrastructure as a Utility to Every Citizen, Governance & Services on Demand and Digital Empowerment of Citizens’ (Kumar, 2015), suggesting that both access and adoption issues are on the radar for the federal government.

On the access side the Digital India programme focuses on rural India, promising wired broadband connections for 250,000 gram panchayats (funded from the Universal Service Obligation Fund1), mobile connectivity to 42,300 villages, internet-linked common service centres for all villages, and conversion of 150,000 post offices into multi-service centres (Thakore, 2015).

For urban India, the programme mandates public Wi-Fi spots for 1 million-plus cities and tourist destinations, communication infrastructure in all new urban development/construction and licensing of ‘Virtual Network Operators’2 to bring down costs and take broadband access to remote areas.

Public policy makes no mention of infrastructure-provisioning for underserved and marginalised urban areas, including the low-income settlements and slums that house the urban underclass.

While privatisation has been an important component in the extension of telecommunications access to low-income consumers, our study reveals the poor and uneven quality of internet access provided to the base of pyramid users. The poor quality of access also limits patterns of use. For more intensive use of the internet, such as filling in and submitting forms for admission to educational institutions and jobs, or applying for government entitlements, respondents report that they are largely dependent on commercially-run cybercafés. At the time this study was conducted there were no free – or even subsidised – high-speed internet services at public libraries or other public spaces that could be easily accessed by the study populations.

While ICT provisioning need not be entirely government-owned, it is important to recognise that ‘the diffusion of any technology is a site of struggle, with access policy reflecting broader issues of political, social and economic power’ (Warschauer, 2003, p 35). Any technology that overcomes – or compounds – disadvantages related to poverty or social exclusion cannot be left entirely to market forces, lest the market continue to develop services for those with greater economic resources, resulting in a two-speed internet where economically stronger users move on to faster and faster broadband superhighways with full access to everything the internet has to offer, while the economically weaker users struggle along with low-end mobile devices, slow services and limited baskets of internet applications offered by telecom players who are expanding their markets in the name of ‘connecting the poor’.

Support and incentives, particularly from the state, will be required to meet the needs of low-income populations and level the playing field.

On the adoption side, Digital India reaffirms the National Digital Literacy Mission target of one digitally literate person in every household. But there is no policy or programme aimed at bringing digital empowerment to the most digitally excluded groups, including women and the poor.

Maharashtra’s state initiatives (discussed in the previous chapter) including the MS-CIT digital literacy trainings and ICT training as part of school curriculum from Class 1, address the adoption and capacity-building requirements of the urban poor more specifically, with fee waivers for low-income households, while its common service centres help the urban underclass apply for and receive government services and entitlements. While this is a good beginning, the state must focus attention on providing equal access to women, the disabled and other particularly vulnerable groups within the urban underclass, and on building their capacity to use the internet to the fullest.


Reframing the digital divide debate

At the outset, this research recommends a change in the way the digital divide debate is framed. From a focus on technology per se and getting people connected – anywhere, anyhow – to a focus on the social contexts of technology.

The overall challenge is not how to overcome the digital divide but how to expand access to and use of ICTs to promote social inclusion.

At the policy level, it is important for governments, urban planners and civil society in India to realise that ‘where and how information is or is not accessible can reinforce privilege and limit opportunities for individuals in underserved and marginalised communities’ (Gomez, 2015).

Research on the barriers to digital equality and needs assessments of particularly vulnerable populations must feed into and inform public programmes such as Digital India.

E-governance and social inclusion

The purpose of e-governance is to bring government to the doorstep of citizens.

But governance does not become inclusive by simply putting government resources and services onto online systems and adding an ‘e’ before governance.

Setting up systems for equal access to information, and equal opportunities to use that information, must become an integral part of e-governance.

Lack of literacy, lack of access to media, and social isolation all serve to distance the poor from government resources and information. This lack of access to government further marginalises them. Only 12% of users in the study settlements had used the internet to apply for or find out about government services and entitlements, or to conduct any online transaction. Even users from the upper wealth quintiles did not report greater use of the internet for government services or online transactions, suggesting that the process doesn’t work, is user-unfriendly, or is not addressing the particular needs of the poor.

E-governance programmes must therefore be carefully designed to meet the needs of the poorest and most marginalised. Simpler user interfaces, minimal text, and multilingual options need to be developed. Mechanisms to help the marginalised negotiate e-governance systems must be put in place. The helpdesks set up to assist the poor in Pune to make their online applications for admission to schools under the 25% reservation for economically-weaker sections are a good example of such mechanisms. The PCMC’s Sarathi programme for grievance redressal is another, where complaints can be lodged online as well as by telephone, and soon via an app. Such systems encourage and aid communication from the bottom up.

In the absence of such mechanisms, e-governance will only end up creating one more barrier between government and the marginalised citizen, a barrier that puts them at the mercy of touts and middlemen who will help them negotiate – at a cost – the digital interface.


Raising awareness of the internet and its benefits

Our data have revealed that 41% of non-users in low-income settlements are still not aware of the internet (many users too are only partially informed, confusing the internet with Facebook or WhatsApp). To develop the motivation to go online, communities must understand the benefits and relevance of the internet.

Beyond general awareness, concrete examples of use should be developed for particular communities, which will be more successful in demonstrating the value of the internet for them.

Building relevant (local) content and services

Amongst non-users who are aware of the internet, aspiration to go online is high. 73% said they would like to go online in the future, and 78% said that the internet was as important as any other basic amenity. Though this might indicate that high levels of motivation already exist amongst non-users who know what the internet is, a sizeable 22.5% of non-users said they didn’t think they needed the internet, indicating that the internet may not have much that is relevant to them.

Equally important, therefore, will be the building of local content that is relevant to marginalised communities, helping them access information related to livelihoods, health, housing, education, personal enrichment, public amenities and entitlements, in languages they are comfortable in.

The absence of content in the local languages is a deterrent to more diverse uses of the internet by diversely-literate populations. The PMC’s own website is available in English only (with not even a Google translate function), though the PCMC offers both English and Marathi interfaces. Several documents on the PMC site are uploaded in pdf format in Marathi. But they are available at links that must be accessed in English.

Language is a particular barrier for the over-35 age-groups who are less proficient in the English language than younger people who study English in school, either as the primary medium of instruction or as a second language. Younger users have worked around the language barrier by becoming proficient in the use of the Roman script to text and search in Marathi, Hindi and other languages.

Addressing social acceptability

46.4% of non-users said the internet could have a negative effect. An even higher 61.9% of users agreed that the internet could have negative effects. While it is important that users continue to be aware of the potential for misuse and misinformation on the internet, the pervasive mistrust of the internet in the study settlements and fears about its dangers, especially for women, contribute to the digital divide and the yawning gender gaps in access.

To build social support for use of the internet, much greater awareness will need to be built in local communities on online safety and responsible use. Family and community support is an important driver of internet use. Social support for internet use will encourage more women, in particular, to use the internet. It will also widen the patterns of use.


Access involves provisioning of devices and conduits.


While the prices of computers and smartphones are dropping, our data show that the most marginalised – particularly women and those in the lowest wealth quintiles – still cannot afford even feature phones that are capable of connecting to the internet.

One way to tackle this problem could be to introduce learn-and-earn programmes in which low-income participants who successfully complete several levels of ICT training are offered a free/subsidised device.

Enabling access to text-to-speech and voice recognition software for the visually disabled as well as for those with little or no education could help.

A more holistic solution to the problem, however, is to set up public access points within the settlements. These centres would offer high-speed internet access on computers at subsidised rates. Such centres would address the needs of these communities at several levels:

  1. Offering access at affordable rates to those who are offline because they do not have the infrastructural and economic capital to go online;
  2. Providing access for women, in particular, who tend to have neither the infrastructure to connect nor the freedom/financial resources to visit commercially-operated public access points at faraway locations;
  3. Allowing interactive functions such as filling in and uploading of complex forms and user interfaces, downloading of data, and secure online transactions, for which these communities are dependent on cybercafés;
  4. Providing experience of the full and free internet, rather than the limited basket of services that telecom providers bundle in with low-cost data subscriptions; and
  5. Mentoring new users who are finding their way around the internet, and helping citizens negotiate e-governance services themselves rather than going to a common service centre.

Such centres can be operated by commercial organisations, government, educational institutions or civil society organisations. The Directorate of Libraries is one organisation in Maharashtra that could expand the number of public libraries, add digital infrastructure, and offer digital literacy courses and programmes. The location of public access points within the settlement is key to their success, not only because it makes it easier for women to access the internet but also to increase the sense of community ownership and social support for the adoption of internet technologies. The presence of trainers and mentors to help community members locate and access the information and services they need is also important for their success. These public access centres could over time become hubs for community media production.


Concrete plans need to be made at the city level to extend high-quality broadband infrastructure to low-income areas. Diversely-literate and multilingual communities in fact need higher bandwidth and speeds, because much of their interactions on the internet tend to be audio-visual, not textual.

State, city administrations, telecom service providers, corporate houses and civil society organisations could be roped in to support and facilitate high-speed broadband coverage of these areas. For instance, subsidies for cable TV providers could enable them to offer wired broadband services at cheaper monthly rates to the low-income households they already provide cable TV to.

The local municipal administrations could encourage the public sector telecom organisations BSNL3, BBNL4 and RailTel5 to extend their networks into low-income settlements by charging lower reinstatement charges for services provided to marginalised areas and subsidising the monthly internet subscriptions of low-income households.

Free or subsidised Wi-Fi networks such as the one set up in Phule Nagar (see box in Chapter 8) could be set up in other settlements to introduce low-income communities to the internet. The presence of a computer and internet literacy centre within settlements providing Wi-Fi coverage is important. The success of this model is visible. Three months after the Wi-Fi network was introduced, women who had never used the internet before had taken the training courses offered and become so familiar with the internet that they were able to shop online and help their children with internet research for their school projects. However, the downside of Wi-Fi technologies has also become apparent: users report that speeds are low, connectivity is not uniform all over the settlement, power outages disrupt the network, and use of a public network hampers privacy.

As the TRAI says in its recommendations to the Department of Telecom on the National Broadband Plan (May 4, 2011), an effective broadband network would be a combination of wireless and wireline technologies, with the emphasis on fibre to the home in the metros and large cities. ‘Wireless in access network may provide wider coverage but lacks high capacity bandwidth support… It is difficult to provide high speed broadband with good quality of service using only wireless technology.’

User capability

It’s simplistic – and wishful thinking – to assume that once connected, citizens will be empowered to use the internet in useful and meaningful ways.

‘If, as Castells suggests, “the internet is the fabric of our lives”, and if those living within this fabric are to have the freedom to achieve the lifestyles they desire, then they must be able to acquire new media literacies,’ (Mansell, 2002).

These new media literacies entail more than just the ability to read and write. Yes, apps, touchscreens, stripped-down websites with graphic prompts for navigation and greater use of audio-visual content are making it easier for those with minimal education to use the internet, but differing levels of ICT skills and education will determine, in the future, how people use the internet and what benefit they derive from it. Focusing on the technology of access in the absence of an equal if not greater focus on user capability will result in a further stratification where the have-less will not use the internet to its full potential, while the ‘creamy layer’ of the knowledge economy will use it to get even further ahead by seeking and creating new knowledge.

Absence of ICT skills and education poses one of the biggest barriers to internet access for the marginalised. Twenty-seven percent of non-users in this survey said they were not online because they didn’t know how to use the internet. Another 16.2% of non-users cited lack of education. Although those who already use the internet have not explicitly cited lack of ICT skills as a barrier, the data show that users with a higher level of education are far more likely to use the internet for purposes beyond social networking and entertainment, such as education and job search.

The one-size-fits-all three-month ICT training offered in Maharashtra via MS-CIT is not enough. Shorter workshops and trainings, with flexible timings and audio-visual instruction, offered to the socially excluded in their own neighbourhoods, will encourage more people to begin using the internet. Workshops must be pitched at various levels, beginning with basic workshops to overcome the discomfort with computers and culminating in workshops that help users evaluate the credibility and reliability of the information they source on the internet, and also help them use multimedia to create their own content.

The quality of ICT education at the school level must be improved. The infrastructure must be up-to-date and fully functional; children should be allowed hands-on training from Standard 1; instructors should be trained in imparting ICT skills. There is a marked difference in the knowledge of ICTs between children in the government-run e-learning academies and the other schools. This difference is also visible in government schools where civil society organisations offer e-learning and ICT trainings. The transformation of all schools into e-learning academies must be hastened, otherwise children from low-income families will continue to be at a digital disadvantage, lacking any exposure to ICTs at home, and getting only cursory knowledge at school.


1. The Universal Service Obligation Fund of India aims to provide widespread and non-discriminatory access to quality ICT services at affordable prices to all people in rural and remote areas (

2. Virtual Network Operators (VNO) are telecom service providers that rely on the network of other telecom companies to provide services to consumers. A VNO buys bulk talktime and bandwidth from an operator and then sells it to users. It can provide any or all the services that are being provided by the network operator

3. Bharat Sanchar Nigam Limited is one of the largest and leading public sector units providing a range of telecom services in India

4. Bharat Broadband Network Limited set up under the Companies Act by Government of India has been mandated to create the National Optical Fibre Network (NOFN) in India

5. RailTel Corporation of India Limited is a public sector undertaking which extends broadband and application services to the masses through the RailWire platform