Virtual Society? logo



They came, they surfed, they went back to the beach: why some people stop using the internet


Related Story: Slump in the Number of US Internet Users (Cyber Dialogue 30th November 1999)


Sally Wyatt
They came, they surfed, they went back to the beach: why some people stop using the internet
(Prepared for the Society for Social Studies of Science conference, San Diego, October 1999)



In many discussions about the future of the internet, it is assumed that once people have been exposed to its wonders, they will embrace it wholeheartedly. Thus, data about the growth of numbers of users are extrapolated to demonstrate even greater future growth. Problems of non-access are associated with various forms of social exclusion; the possibility of voluntary non-access is rarely acknowledged. Even more rare is the recognition of the possibility that people might make an informed choice not to continue to use the internet.

Within the technology and innovation studies literature, there is discussion of public acceptance of new technology, public resistance (often assumed to be based on ignorance and fear), barriers to use and how to overcome them. Such literature usually assumes access to technology is necessarily desirable, and the question becomes one of how to increase access. Informed, voluntary rejection of technology is not mentioned. I will return to this later, but I want to argue that this invisibility reflects the continued dominance of the virtues of technological progress, not only amongst policy-makers but also amongst the STS academic community, who would probably reject such a charge.

This presentation draws on some of the data being analysed within the Virtual Society? Project, From the net to the web and beyond: actors and interests in the construction of the internet being conducted together with two colleagues from the University of East London, Tiziana Terranova and Graham Thomas, who do not necessarily share the views I am about to express. At best, this is work-in-progress, however, it is more of an initial exploration of some ideas preparatory to developing further research. Identifying this invisible group raises again the questions surrounding the ANT dictum, ‘follow the actors’, which have already been raised by Martin & Scott 1992, Russell 1986, Bijsterveld 1991. How do you do this when some actors are completely invisible?

The presentation has three aims:

  1. To review the available data and literature about internet rejectors
  2. To discuss the policy implications of internet rejection and non-use
  3. To explore the implications for technology studies of the absence of technology rejectors

1. Available data and literature about internet rejectors

This will be a very short discussion – there isn’t much. I did a literature search on various combinations of internet, computers, information technology, technology on the one hand and rejection, dropout, non-use, barriers, have-nots on the other, and came up with very little, and most often with no hits at all. ‘Barriers’ yielded most, but most of that was about national level adoption or education. ‘Drop-outs’ also provided quite a few, including some interesting material about young people who dropped out of school or university as a result of doing too much internet surfing. My absolute favourite was entitled, ‘treating technophobia – a longitudinal evaluation of the computerphobia reduction program’. Sounds very painful!

The only work I’m aware of is by two Americans – James Katz and Philip Aspden in a paper which appeared last year in Telecommunications Policy. Their analysis of ‘internet dropouts’ was a side-effect of some research about barriers to Internet use in the US. They candidly admit they included the category of ‘former user’ in their surveys only for logical completeness. They conducted a national random telephone survey to investigate users’ and non-users’ perceptions of the Internet. They were surprised to discover in October 1995 that ex-users and current users each accounted for about 8% of the sample. They did another survey in November 1996, by which time the proportion of current users had more than doubled to 19% of the sample, the proportion of ex-users had also increased, but by less, to 11%

Some interesting results:

Another tiny bit of data concerns the use of the Amsterdam Digital City – one of the early digital cities. In the past couple of years, many previously regular users have stopped visiting. In this case, they have not become ex-users of the Internet, but as service providers have proliferated, they have more choices. Until recently, the Amsterdam Digital City was one of the few sources of email and an easily accessible host for web pages. The community building aspirations of the ADC were not compelling for many of their early users. This also has important implications for those committed to the community building potential of the Internet.

Perhaps we need to turn to some other technologies. For example, there is some recent work by Louis Leung and Ran Wei about mobile phone use/non-use in Hong Kong. Mobile phones have a much longer history than the internet as a consumer technology. Following the work of Everett Rogers in the 1980s about adoption of communication technologies, they identify four groups of factors important in determining whether someone adopts a technology or not:

  1. Individual personality traits (not included in this study)
  2. Socio-economic characteristics
  3. Interpersonal communication influence (including use by individuals’ social networks, and mass media, but not advertising)
  4. Perceived attributes of the innovation itself

Age, income, gender and education all work in expected ways. However, age dominates – if you’re older (unspecified), having more money and more education doesn’t make much difference. Income levels are declining in significance – suggesting the good old theory, beloved by free marketeers, of ‘trickle down’ works – mobile phones are no longer perceived as the preserve of young men in suits. Intensity of use of mass media is not significant, but belonging to social groups which use mobile phones is. Equally unsurprising is the finding that non-users perceive the technology to be unnecessary because they have an alternative or because they find mobile phones complex to use (including pricing structures) or intrusive. Leung and Wei’s results confirm a growing gap between communication rich and poor, with mobile phone users more likely to possess a range of alternative and complimentary forms of telecommunication – pagers, answering machines, etc; whereas non-users simply had one reasonable alternative. [Note: this accepts the premise that more communication devices = good; only one adequate communication device = bad]

This work on mobile phones isn’t very surprising – people don’t use mobile phones if they have alternatives, think they’re intrusive and/ or expensive. By extension, maybe some people don’t use the internet because they have alternative sources of information and forms of communication which are appropriate to their needs, or because they think it is cumbersome and expensive.

2. Policy implications of internet rejection and non-use

The question of dropouts may only be a transient issue – all dropouts may eventually return to the fold. On the other hand, the internet may follow the model of CB radio – explosive growth followed by collapse. It’s still too soon to say. In any event, in the US alone, there are literally millions of former users about whom very little is known. They may be a source of important information for subsequent developments. Even within the paradigm of increasing access, it is important to know why such people leave and what could be done to lure them back. Rather than denying the possibility of their existence, internet service and content providers as well as policy-makers might all have much to learn from this group. Steve Woolgar tells me that when he told the industrial member of the Virtual Society? advisory group who is a member of the World Wide Web consortium about this data regarding the existence of former users, the response is that Steve is ‘completely bonkers’. Why do actors find voluntary rejection so hard to contemplate?

What categories of non-use can we identify?

  1. Never used – because don’t want to
  2. Never used – because can’t get access, for variety of reasons
  3. Stopped using – voluntarily (boring, alternatives, cost, etc.)
  4. Stopped using – involuntarily (cost, loss of institutional access, etc.)

Policy implications are different for the different groups – for 1&3 it might be appropriate to develop new services to attract them [or, learning from the Minitel experience – get rid of some old ones so people have to switch]; traditional access issues related to cost, skill and location might be important for 2&4.

Once one has made the step of including ‘former user’, as well as ‘current user’ and ‘never a user’, it is not too much more of a leap to begin to take apart the notion of ‘user’. What exactly does it mean to be a user? How is it defined? A recent survey in the UK (NOP early summer 1999) suggests that 26% of users didn’t access the internet at all in the week preceding the survey, and a further 20% only accessed it once or twice. [This survey illustrates another curious feature of internet usage data: when users themselves are asked, no one admits to looking for pornography. A rather different picture emerges when one looks at provider data. I realise I have presented the frequency of use data as if it is unproblematic whereas the point of this comment about the nature of use is to call into question the reliability of the data.] The point I’m trying to make is that we need to treat the notion of internet usage in a rather more nuanced way, distinguishing between complete ‘surfies’, and those people who don’t like to get their hair wet and only venture into the water occasionally.

3. Implications for technology studies of the absence of non-users

It’s time to declare my personal interest in this topic. Of course I use the internet, though I don’t have home access. I could not have written this without it – both to check sources and to discuss some of the ideas with friends and colleagues in different parts of the world. However, there is another major 20th century technology which I don’t use – a car, the ‘machine that changed the world’. I’m very well qualified – passed my test first time in Toronto when I was 16 – took me two tries in England when I was 25. I’ve never really driven much – twice I think since I passed my British test 15 years ago, and I have never owned a car. Even amongst critically and environmentally aware STS scholars, at least those in the UK and US, this is regarded as a deviant and bizarre choice. One of my closest friends thinks it reflects a failure to grow up on my part: ‘real adults drive cars’.

[I don’t want to give the impression of being a neo-Heideggerian – I do own other contemporary IT-based products, and travel frequently by plane. I enjoy many of the features of late 20th century life.]

To what extent is not driving analagous to not using the internet? Given the prevalence of ‘superhighway’ metaphors in some policy discussions, particularly here in the US, is there anything to be learned from non-drivers? Or are they another invisible group? (Yes – according to my recent literature search.) Especially non-drivers like me who have a choice. I could afford a car – most of my British colleagues have one. Not having one is usually regarded as a sign of deprivation. At the very least, the existence of people like me means we have to be careful about how we interpret data about non-ownership/use of particular technologies. Maybe non-use of the internet reflects some other phenomenon – like it’s boring, there are easier alternatives for obtaining the same information, people would rather spend their time and money doing other things, or people have spatially close social relationships.

Thinking about cars also highlights the connections between the online world and the offline world. I simultaneously inhabit the same world as car drivers and a different one. My life is affected by cars – as a pedestrian and more recently a cyclist, I have to watch out for them; as an organism that needs oxygen, my quality of air is affected; as a bus user, they slow me down. I am constrained by the reach of public transport in where I can live and visit. My knowledge of London is very much based on public transport routes convenient to where I lived. Just as there are different maps of the physical world, so there are of the internet. The London underground map is a better and more useful representation of my experience than a topographical one. There is also a parallel universe which I don’t visit much. Recently I was driven from London to Amsterdam. I was fascinated – to the amusement of my driver friend - by this alien world of motorways, petrol stations, motorway services, drive-on ferries. There are advantages to not driving – saves me lots of money, time and stress, reduces my chances of being killed or killing others, and in these post-Rio, post-Montreal days of greater environmental awareness, provides me with a tremendous feeling of self-righteousness.

Will the cyberworld come to dominate the real world to anything like the same extent? Is it possible to turn off the machine? Or, will everyone’s choices come to be shaped by the net?

1. Technology as cultural symbol

All technologies are imbued with cultural significance. This is especically true for the car, a paradigm case of a symbol of modernity in the 20th century. To many people cars reflect wealth, power, virility and freedom. The internet promises many of the same attributes, on an even larger scale, with its promise of global reach. While this simply might not be appropriate for many individuals and organisations for whom time is not short and for whom geographical distance between self and significant others is not great, the symbolic importance remains profound.

2. Technology as control and surveillance

You need a license to drive a car, though you didn’t during its early days [in Belgium they became compulsory only on 1 January 1967; anyone over 18 on that date received a license automatically, email: Jacques Berleur, ‘funny question’, 20 October 1999]. In many parts of the world, a driver’s license serves as official identification. In the UK, you are legally obliged to inform the DVLC (Driver Vehicle License Centre) of a change of address. Most people are not aware of it, but it effectively serves as a population register, and is used as such by the police. These days, much car use is subject to enormous levels of surveillance, through the spread of cctv (closed circuit television) [discussions in Netherlands about electronic tracking of all cars]. Many of us choose to ignore the surveillance capacities of the internet in our daily practices, hoping that the sheer volume of traffic will serve to protect our privacy, but all of our activities on the internet are pretty transparent. If you wish to escape the surveillance capabilities of modern societies, avoid modern technologies.

3. Network technologies as totalising externalities

Cars are not simply wheels, engines and steel: they include test centres for drivers and vehicles, motorways, garages, petro-chemical industry, drive-in movies, out-of-town shopping centres. The more people use cars, the greater the infrastructure to support them, and the lessening of car-free space. Similarly, the internet is not just web content: it includes computers, telecommunication links, routers, servers and other applications. The more people use it, the more pressure there is to develop user-friendly interfaces and provide greater bandwidth. But there is a paradox here: the technologies become more useful if more people use them but both tarmac highways and electronic superhighways become more congested. There have been several ads recently for cars on British, Dutch and Canadian television in which the ideal presented is of a single car on a remote, non-urban road – no other cars, no other people. This resonates with Wired’s pictorial representation of the future [nature – no machines, no people – use OHP, january 98 issue].

[Technology vs. Nature]

[Is rejection of technology, whether cars or the internet, an indication of a deep green agenda? Is it possible to occupy a more ambivalent position, or is Donna Haraway right, that we’re all cyborgs now? – need to think about this some more, won’t mention it]

4. Specific technology as leitmotif for social science research

In the 1970s and throughout part of the 1980s, even for researchers not particularly involved in STS, the car and its industry were the site of much empirical research. Moreover, the auto worker was the prototypical industrial worker. [Today, is the prototypical worker the code slave, the call centre operator or the supercool multimedia designer?] For better or worse, the car industry became the symbol of industrial society, and much effort was expended in understanding the dynamics of that industry. There were problems with this including the promulgation of the idea of the skilled male worker as the norm, and the generalisation of a set of industrial relations and working practices to other sectors. Social theory focused on questions of alienation and massification, extending them, not always appropriately, to other areas of social life.

In terms of social science research agendas, we may be witnessing something similar today, possibly even on a grander scale. Just as countries have programmes to promote ICT R&D, to encourage the location of ICT production and the adoption of ICTs by industry, lots of social science research programmes are devoted to studying the re-configuration of the social-technical divide, using ICTs in general and the internet in particular as exemplars. For example, the Virtual Society? Programme asks,

(ESRC 1999: 3)

Perfectly good questions, but even though they allow for negative answers and the question mark illustrates the programme’s commitment to scepticism, there is a danger that we may repeat past mistakes – to totalise; to take the use of ICT – by individuals, organisations and countries - as the norm and non-use as a deficiency to be remedied. I am suggesting that one way of avoiding these problems is to take seriously non-users and former users as relevant social groups, as actors who might influence the shape of the world. Once we’ve solved the problem of how to research the invisible, maybe it will be time to stop researching ICT altogether.


The work on which this is based is supported by the Virtual Society? Programme ( of the Economic and Social Research Council, grant no. L132251050. Tiziana Terranova and Graham Thomas, with whom I am working on this project, almost certainly do not share the views expressed here.


Bijsterveld, Karen (1991) ‘The nature of aging. Some problems of an "insider’s perspective" illustrated by dutch debates aboug aging (1945-1982)’ Paper presented at the Society for Social Studies of Science meeting, cambridge, MA. November.

Burrows, Roger et al (1999) ‘Virtual community care? Social policy and the emergence of computer mediated social support’ paper prepared for submission to Information, communication and society.

ESRC (1999) Virtual Society? The social science of electronic technologies, Profile ’99, Swindon: ESRC.

Katz, James E and Aspden, Philip (1998) ‘Internet dropouts in the USA’ Telecommunications Policy, 22, 4/5, pp.327-39.

Leung, Louis and Wei, Ran (1999) ‘Who are the mobile phone have-nots? Influences and consequences’ New Media & Society, 1,2, pp.209-26.

Martin, Brian and Scott, Pam (1992) ‘Automatic vehicle identification: A test of theories of technology’ Science, Technology and Human Values, 17,4, pp.485-505.

NOP (1999) ‘Internet user profile study’ confidential report prepared in September for syndicate members.

Russell, Stewart (1986) ‘The social construction of artefacts: A response to Pinch and Bijker’ Social Studies of Science, 16, pp.331-46.

Email comments from Flis Henwood, Helen Kennedy, Tim Jordan, Ian Miles, Lera Miles, Nod Miller, Dave O’Reilly, Graham Thomas; pencil comments from Hans Radder.

Sally Wyatt
University of Amsterdam


Related Story: Slump in the Number of US Internet Users (Cyber Dialogue 30th November 1999)



Return to top of page
Switch to text only version

Page developed by Christine Hine
Page maintained by Marike van Harskamp
Contents current at 5th January 2000