Modes, Quality and Cost of Internet Access

Chapter 5


‘Recent discussions have focused on a “two-speed” internet. Today the internet may be an internet of 2.9 billion different “speeds”, in terms of how the individual internet user relates to online networks and society as an online citizen,’ (Broadband for All 2014, ITU).

A holistic understanding of the digital divide necessarily involves studying not just access to an internet-accessible device, but also what the access entails in terms of cost, internet speed at which this access is available, frequency of use and the use made of the internet.

This chapter therefore focuses on internet users in the study settlements and details the modes/devices of access, the cost and quality of internet services used in these low-income urban neighbourhoods, as well as the frequency and patterns of use. The chapter is organised in nine sections. Sections 2 and 3 discuss the frequency and purpose of internet use. Sections 4 and 5 describe the points of access and type of services available/preferred. Section 6 describes the modes/devices used. Section 7 focuses on the expenditure incurred on internet services. The quality of internet access is discussed in Section 8. Section 9 concludes the chapter.


Figure 5.1 Frequency of internet use (% of user respondents, n=564)

Figure 5.2 Frequency of internet use by gender (% of user respondents, n=564)

Sixty-six percent of internet users have reported daily use of the internet (Figure 5.1), while 13% of users access the internet only one to three times a month. Around 10% of users access the net once a week and another 9% two to five times a week. Seventy-one percent of male users access the internet every day as against 52% of female users (Figure 5.2). Twenty-five percent of female users have reported access to internet only one to three times a month. In the following paragraphs, the frequency of internet use has been studied with respect to age, occupation and wealth status.

Figure 5.3 Frequency of internet use by age (% of user respondents, n=564)

The largest proportion of users who access the net every day fall into the 31-35 age-group – around 82%. The smallest proportion of daily users is in the 16-20 age-group (Figure 5.3), though more than half of all internet users belong to this age-group (Figure 4.20).

Figure 5.4 Frequency of internet use by occupation (% of user respondents, n=564)

Similarly, although students form the biggest segment of internet users, their frequency of use is the lowest (Figure 5.4). Nearly 20% of students use the internet only one to three times a month. Affordability is the key here. Although the younger age-groups and students are the biggest chunk of users in the study settlements, they may not have the financial resources to own personal devices for access or be able to afford access to the internet every day. Moreover, they may not need to use it every day. Working people, especially those in service (in both the organised and unorganised sectors) are more likely to use the net every day. It is possible that they access the internet at their workplace. However, a large number of daily-wagers and non-working populations also access the internet every day. This is probably for the social networking services that are so popular.

Figure 5.5 Frequency of internet use by wealth quintile (% of user respondents, n=564)

Wealth status shows a direct relationship with frequency of internet use. Everyday use increases with improving economic status. In the lower wealth quintiles, the proportion of those who use the internet only one to three times a month is comparatively higher (Figure 5.5). However, note that 62% of users in the lowest wealth quintile feel the need to use the internet every day even though the cost would be a higher proportion of their income.


The largest proportion of users access the internet for entertainment (80%) and social networking (76%). Fifty percent of users access news and information and 47% use it to look for jobs and education (Figure 5.6). Only 12% reported the use of online services, including applications to government entitlements and services, e-banking and e-commerce.

Figure 5.6 Purpose of internet use (% of user respondents, n=564)

Note: Multiple response

The purpose of internet use across gender shows that more men than women use it for social networking and entertainment (Figure 5.7). A larger proportion of women than men, however, reported that they use the internet for jobs and education. There is no significant difference across gender in use of the internet for information search. A slightly higher proportion of men use the internet for online services and communication.

Figure 5.7 Purpose of internet use by gender (% of user respondents, n=564)

Figure 5.8 Purpose of internet use by age (% of user respondents, n=564)

Note: Multiple response

The purpose of internet use remains more or less standard across age-groups. In all age-groups, the principal use of the internet is reported to be entertainment and social networking. In the 26-30 and 31-35 age-groups (Figure 5.8), the internet is used (comparatively) more for online services and communication (e-mails, Skype, etc). The proportion of people using the internet for information search is also high in these age-groups.

Figure 5.9 Purpose of internet use by educational status (% of user respondents, n=564)

Note: Multiple response

Entertainment is the main purpose of use for all respondents regardless of educational attainment (Figure 5.9). Social networking is used less by persons with no education or very basic primary education (these users mostly use the internet for audio-visual entertainment). Those who have been educated up to higher secondary/diploma and above also use the internet for educational research, job search and information-seeking. But use of online services does not see appreciable change across education levels, suggesting that e-governance or online facilities are either not available or difficult to use.

Figure 5.10 Purpose of internet use by occupation (% of user respondents, n=564)

Note: Multiple response

Students however report significant internet usage for job/education and information-seeking (Figure 5.10). There is not much variation in purpose of use by wealth quintiles, though the internet is used comparatively more for online services by persons belonging to the higher wealth quintiles (15% of the top quintile as against 5% of the lowest quintile; figures not shown).

Box. 5.1 Purpose of use

Most people here use the internet only for Facebook and WhatsApp. Most of the people use it for ‘timepass’. I use the internet to check official mails. I forward clients’ complaints by e-mail to my seniors. I also use the internet for online banking. I pay online for my post-paid connection and recharge family members’ mobiles online. I make online money transfers and shop online.

(Male internet user, 26, married, college graduate)

I use the internet only for Facebook and WhatsApp… But my lecturer told me about the Maharojgar online site where one can register for government jobs. He told me there is an app that allows anybody to pay their light bill… No, I don’t know about any other e-governance sites and e-seva kendras.

(Male internet user, 17, unmarried, first-year BCom)


Mobile (feature or smart) phones are the major access point for the internet in low-income neighbourhoods. Mobiles are affordable (almost all the households studied had a mobile phone of some sort), in comparison to laptops or computers, and they can be carried anywhere and accessed from anywhere. Mobiles are also easier to use than computers for those with no ICT skills. Additionally, mobiles are the only option for data access for internet users in resource-poor environments (Box 5.2). There is an almost complete absence of wired broadband in these bastis, and cable TV providers do not extend internet services to these areas because the costs of wired broadband (plans offered are at Rs 599-1,100 a month) are much higher than mobile internet, and consequently there is little demand. Increasingly, cybercafés are closing down in the city, partly because they are no longer economically viable given the popularity of mobile internet, and partly because of the cumbersome records they are required to maintain. There are no other public access points.

Figure 5.11 Points of internet access (% of user respondents, n=564)

Of the 564 internet users, 59.2% access the internet on mobile phones only. Nine percent use the internet at cybercafés, 4.8% on computers or laptops, and 3.9% at the workplace, school, college or computer class (Figure 5.11). The proportion of men accessing the internet on the mobile phone is higher than women. More women depend on cybercafés and the workplace for access to the internet.

Figure 5.12 Points of internet access BY age (% of user respondents, n=564)

The younger age-groups make greater use of cybercafés. The older age-groups (above 30) do not use cybercafés at all and a significant proportion of this age-group uses the internet at the workplace (Figure 5.12). These internet users are more likely to be better-educated and in service.

Figure 5.13 Points of internet access by educational STATUS (% of user respondents, n=564)

Interestingly, the use of mobile phones for internet access declines with increasing education (Figure 5.13), when multiple access points increase. The mobile phone is the only access point for users with no education, which is not surprising since it is both more affordable and simpler to use than a computer. Those who are educated up to higher secondary school and above use multiple access points.

Figure 5.14 Points of internet access by occupation (% of user respondents, n=564)

While mobile phones are the major access points for the internet across all occupations, users who have regular jobs in the formal or informal sector (classified as service) and students access the internet from multiple points (Figure 5.14). Students also access the internet at cybercafés.

Figure 5.15 Points of internet access by wealth quintile (% of user respondents, n=564)

Affordability is clearly a determining factor in access. Those in the lower wealth quintiles are restricted to data access on mobile phones and cybercafés while the upper wealth quintiles have a larger share of users accessing the internet through computers and laptops. Those in the fifth quintile (31.9%) have the capacity to access the internet from multiple points (Figure 5.15).

Box. 5.2 Why the mobile phone is the preferred access point

I feel that mobile internet gives one access to information exactly when it is needed. Otherwise one would have to visit the cybercafé and finish all one’s work before leaving the place. So mobile internet is a better way to stay connected.

(Female internet user, 28, married, educated up to Standard 10, homemaker)

I prefer to use the internet on a touchscreen mobile. It’s easier to use than laptops and PCs. Cybercafés are about half-a-km away from the settlement and they do not allow for privacy of use. So the mobile is the best option.

(Male internet user, 26, married, college graduate, service job)

Box. 5.3 The privacy problem

Nineteen-year-old Satish B, a Standard 11 student from a low-income settlement in Pune’s Aundh area, uploaded pictures of himself and his girlfriend on Facebook as part of a private album. Some of these pictures found their way onto his friends’ timelines.

Satish generally accesses the internet from his friends’ mobile phones, since he cannot afford a smartphone himself.

‘The pictures were not visible to anyone but me. I use my friends’ mobiles to access Facebook (Facebook karna, in colloquial terms), and I think I may have forgotten to log out from my account on one such occasion,’ he says.

The relationship between the public and private is fragile for those accessing the internet in low-income settlements. Many people without economic and technological resources rely on others’ phones to go online, and privacy can easily be breached, for ‘fun’, revenge or one-upmanship.

Kishore is a 17-year-old school drop-out who works at his uncle’s mobile repair shop in a PCMC settlement. He downloads content from the internet and copies it onto clients’ phones for a fee of Rs 30-50.

‘I knew nothing about the internet one-and-a-half years ago,’ he says. ‘My uncle, who runs the shop, taught me the basics.’

Kishore has set up and installed Facebook accounts on the mobiles of many residents of his settlement, charging up to Rs 100 for the service. His clients are mostly first-timers on the internet, with little or no knowledge of English.

The arrangement that allows Kishore to earn a livelihood also allows privacy to be compromised. It is unclear how many of his clients figure out how to change their passwords. As the creator of his clients’ accounts and passwords, Kishore could, if he wished to, play around with their private data.

Sanjeev is a part-time mobile phone repairer in the same settlement. ‘People come to me because I am good at repairing mobiles, and I charge Rs 100 for a job that would cost at least Rs 300 at a proper shop,’ he says. Sanjeev learnt the basics of mobile phone repairing while studying at an industrial training institute in 2010, though he dropped out of the course after six months.

But Sanjeev’s clients may pay a price for this low-cost repair. ‘Whenever I have to repair a mobile phone, I tell clients that it cannot happen without the memory card. I tell them I need the memory to check for viruses. They willingly agree to leave the card with me, inside the phone,’ he says. ‘I love finding out what is in people’s memory cards… I browse through and copy whatever interests me. Mostly, it is images, wallpapers and videos,’ he says.

~ Aritra Bhattacharya,, February 2015


52.3% of users access the internet on mobiles with 2G services, 24.5% on mobiles with 3G services. Nine percent use cybercafés and only 7.3% use wired broadband (Figure 5.16), this last most likely at work, school/college since, as we have seen, there is no wired broadband in the bastis. The mobile phone, though convenient, has several limitations and constraints including poor connectivity, slow speeds, small screens and limited data storage. Those in the upper wealth quintiles rely more on dongles and 3G services.

Figure 5.16 Medium of internet access (% of user respondents, n=564)

Box. 5.4 Medium of internet access

I had a dongle which I stopped using. It was too expensive. The cybercafé is there but there is no privacy to use the internet. So the mobile is the best option.

(Male internet user, 26, married, college graduate, service job)

The cybercafés don’t have sufficient computers and often they would be crowded, so we had to wait to get our turn. The speed would be slow; the PCs were old, pages didn’t load. Sometimes half-an-hour would be wasted like this.

(Female internet user, 23, single, college graduate, service job)



Ownership or non-ownership of internet-enabled devices is an important determinant of access to the internet. As we have seen earlier, 96.5% of households in the study locations own at least one mobile phone (Table 5.1); 34.5% of households own basic mobile phones capable of voice communication/SMS only, while 62% have feature phones on which the internet in some form can be enabled, often limited to certain internet functions that are bundled in by the service provider. Only 30% have smartphones which allow fuller use of the internet than feature phones.

Table 5.1 Ownership of devices that can connect to internet (% of households)

Note: Mobile phone includes all phones including basic phones and internet-accessible phones. The responses are multiple and do not total 100

Internet packs

Internet data packs are offered and chosen according to the purpose of use. They are of two kinds. One categorised by the duration or validity and the other by volume of data transfer. The use of both these has been analysed in the following paragraphs.

Internet packs by data transfer

These internet packs vary from up to 300 MB to above 1 GB. The lower data packs primarily cater to messaging services and social media. Limited data and information download would be possible in data packs up to 300 MB. More than one-third of internet users in the six study locations used internet recharge packs that allow them data transfer of up to 300 MB only (Figure 6.10). Internet users who used data transfer above 1 GB form a small percentage (8.5%) of users. Nearly 30% of users were not aware of the details of the internet packs they bought, indicating that for the bulk of consumers, cost/affordability determines the purchase of data packs. The limited data transfers also suggest that these packs are used mainly for social media and messaging services, and to a limited extent for entertainment. Even entertainment involves download of audio/video and calls for higher data transfers.

Figure 5.17 Data packs by gender (data transfer limit)

Note: 49.6% of females and 22.5% of males reported ‘don’t know’

A large proportion of the female users are unaware of details about the data packs available. The graph shows only those who have reported the use of data packs (n=390). 32% of male users report using data packs of 300 MB-1GB (Figure 5.17). No gender difference is observed among the small number of internet users who use data packs above 1 GB.

Figure 5.18 Data packs by age (data transfer limit)

Internet users in the 20-36 age-group show consumption of higher data packs (Figure 5.18). It is possible that this age-group is working and able to afford more data transfers. In the 16-20 age-group, 44% of users use packs of less than 300 MB which suffice for social networking and entertainment.

Figure 5.19 Data packs by occupation

With respect to data usage across occupations, self-employed workers use the largest (above 1 GB) data packs (Figure 5.19). Those in service also use higher data packs. Daily-wagers and students use small data packs up to 300 MB. This could be due to both the cost factor as well as the limited purpose of use.

Internet packs by period of validity

Figure 5.20 Type of internet packages (days’ validity)

Note: Not applicable/Don’t know=113 cases

Fifty-one percent of users prefer 1-7-day packs (Figure 5.20). Affordability seems to be the primary reason for this choice. Lower data validity also implies lower data transfer limits. These packs are primarily used for social networking. Around 44% of users select packs of over 20 days’ validity. These cost more and also give higher data transfers.

Figure 5.21 Data packs by gender

Women are seen to be using data packs with shorter validity. In male-dominated households, women are likely to have little disposable income. In addition, when cost is a factor, male family members seem be the ones spending more on data packs. Nearly 47% of men use packs of more than 20 days’ validity (Figure 5.21). Young internet users in the 16-20 age-group buy data packs of shorter validity (Figure 5.22). There could be several reasons for this but the most important is that they have limited amounts of money to spend. In contrast, older users opt for packs with longer validity.

Figure 5.22 Data packs by age

In terms of occupation too, those in service and the self-employed opt for internet packs with higher validity (Figure 5.23). This can be explained both on the basis of need and affordability. Those in service, as well as the self-employed, might need the internet for work; they are also likely to be able to afford to spend more on their data packs. Daily-wagers, students and non-working respondents buy only short-duration packs.

Figure 5.23 Data packs by occupation


Box 5.5 Internet package rates of a popular private telecom service provider in mid-2015

Figure 5.24 Expenditure (Rs./MONTH) on the internet

Nearly 45% of the users spend less than Rs 100 per month on the internet (Figure 5.24). This correlates well with the above analysis. Users buy basic internet packs with lower data transfers and validity. Box 5.5 shows the cost of data packs provided by a private telecom service provider whose services are extremely popular in the study settlements, largely because of their affordability. An unlimited WhatsApp pack for a day can cost as little as Re 1. Unlimited Facebook access for a month costs Rs 18. Such packs suffice for users who cannot afford to pay more and whose internet use is in any case limited to social networking and entertainment. The data packs serve to expose users to the internet, but of course the intention of the service provider is to give new entrants a taste of the internet and hook them in the long term.

Expenditure on internet by family income

Table 5.2 Household expenditure on internet by family income (% of households)

Internet users belonging to households with higher family incomes (above Rs 10,000 per month) report higher expenditure on the internet (Table 5.2). A larger proportion of users in the lower income categories spend either less than Rs 50 or a maximum of Rs 100. Users in the higher income groups also show a large proportion spending nothing on the internet. This is possibly because these users access the net at their workplace, or because the service is paid for by elders in the family.

Socioeconomic characteristics

Figure 5.25 Expenditure (Rs./month) on the internet by gender (% of male and female user respondents)

Male users spend more on internet access than female users (Figure 5.25). Almost a quarter of the female users do not incur any expenditure on accessing the internet. This could be because quite a few women access the internet on a family member’s mobile phone. More than half the female users (57%) spend Rs 100 or less per month.

Table 5.3 Household monthly expenditure on internet by age (% of households)

Users in the youngest age-group (16-20 years) who are more likely to be students spend up to Rs 100 per month (Table 5.3). Basic data packs suffice for their social networking and entertainment needs. Users aged 21-35, who are likely to be earning, spend more (up to Rs 300 per month) on internet.

Table 5.4 Household monthly expenditure on the internet by education (% of households)

Those who have no formal education or only primary education are not engaged in occupations where they would have access to computers (Table 5.4). They are thus required to spend for internet access and 45% of them spend less than Rs 100. A small proportion (1.6%) of users who have reported spending more than Rs 500 per month on internet access are those who are better educated (higher secondary or above).

Sixty-eight percent of users, across all occupational categories, spend less than Rs 200 on the internet. Roughly three-fourths of daily-wagers, the self-employed and students, and 65% of those not working, spend Rs 1-200 on the net.

Table 5.5 Household monthly expenditure on internet by occupation

No daily-wagers spend more than Rs 200. The smallest percentage of people spending no money on access are the self-employed; in other words they must spend to use the net. The largest percentage of people spending no money are daily-wagers and those in service. While those in service may use the internet at the workplace, it is possible that daily-wagers use the internet on borrowed devices.

Less than 2% spend more than Rs 500 on the net; almost 5% of those in service spend more than Rs 500 monthly.

Box. 5.6 Choice of internet data packs – negotiating costs

I have to manage the cost… I choose data packages consciously so that I get maximum data transfers in that many days of validity… I compare service providers and look for the most affordable options… I change providers… I prefer a prepaid pack as I have all the information about it from the service provider or the shop beforehand. And I can frequently change plans on prepaid.

(Female internet user, 23, unmarried, college graduate, service)

When I put in my SIM, I get a message saying, ‘You have taken so and so rupees pack… you will get so many GB…’ We can’t afford postpaid plans, so I use a prepaid plan. I recently got a Rs 150 recharge for 1GB data, now they are charging Rs 300 for 1GB so my brother said we don’t want this, I will use the internet in the office and whenever you want to use it, you go to the cybercafé.

(Female internet user, 20, unmarried, 1st-year Master’s in Computer Management student)


Getting connected

Our mapping of the study sites revealed that there were no public access points in the settlements. The majority of users reported that they faced no difficulty at all in getting internet connections on their mobile phones. Those who reported difficulty in getting connected said they were unable to understand the application procedure. Others said they did not have the necessary documents. Still others reported the reasons as a) not having a good enough device, b) unaffordable rates, c) poor network coverage, and d) refusal to provide wired connection or postpaid mobile internet connections in the settlement.

Perceptions: Quality of service and affordability

Figure 5.26 Users’ perceptions of quality of service (% of internet user respondents, n=564)

User perceptions about the quality of connectivity and speed show that roughly 50% rate these as good (Figure 5.26) and around 40% rate them average. It is important to remember, however, that these users have no benchmark for high-speed and uninterrupted connectivity. Most have never used wired broadband, except perhaps at cybercafés where bandwidth is distributed amongst several computers and consequently speeds are slower.

The CCDS team conducted actual speed tests (on smartphones and tabs with 3G connections) at different locations in Janata Vasahat, Laxmi Nagar, Ambedkar Nagar, Anand Nagar and Mahatma Phule Nagar. (Only one round of tests was conducted at each settlement.) The highest download speed was 9.41 mbps and lowest was 0.04 mbps. At some locations the internet speed was so slow that the speed test yielded no results. Most locations had speeds lower than the minimum acceptable 512 kbps.

Figure 5.27 Users’ perceptions of affordability of service (% of internet user respondents, n=564)

Figure 5.27, similarly, shows a high level of user satisfaction with the rates offered for internet access. But it is important to underline that at present most users are not moving much beyond Facebook, YouTube, WhatsApp and simple search functions. The rates offered for these

services are in fact very affordable. But affordability remains a barrier both in terms of the cost of enabling infrastructure (computers, smartphones, tablets) as well as quality internet access in these settlements. As we have seen, cable TV providers who have the capacity to offer broadband in these settlements find there is no demand since the costs of wired broadband are upwards of Rs 600 a month.


Infrastructure and cost are significant barriers to access. Very few households own computers with dongles. There are no public access points within the bastis. There is little demand for wired broadband, because the majority of residents have only mobile devices that can connect to the net and because wired broadband services cost more than most people are able to pay in these settlements.

Because of the lack of demand, there is no supply of wired broadband services. Internet users in these settlements therefore connect mainly via their mobile devices, and some use cybercafés and internet at workplaces.

Only 30% of users have smartphones. The rest access the internet on feature phones with limited internet functions.

Internet-enabling infrastructure is clearly correlated with wealth. Families in the higher wealth quintiles are more likely to own devices that can connect to the internet.

More than 50% of those who connect on their mobiles use slow 2G services, while about a quarter use 3G. Data packs consumed are of limited validity and data transfer. Consequently their costs are low.

Consumers rate the affordability and quality of service as good or average. But this is because they have no benchmark for higher speeds and better connectivity.

Infrastructure and cost barriers also limit frequency of use. Only 66% of users can get online on a daily basis.

Users are accessing the net mainly for social networking and entertainment, though 47% also use it for education and jobs. The low-cost, limited-data-transfer, short-validity internet packs consumed by the majority of users are designed mainly for social networking. As users graduate to more diverse uses of the internet, more sophisticated modes of access and superior quality of service will be called for.

Despite the government’s big push towards e-governance, only 12% of internet users in these settlements use the internet to apply for entitlements or to conduct online transactions.

Infrastructure and cost barriers are thus also constraining patterns of use of the internet.