Power to the people? Assessing the social impact of the People's Network in public libraries in Shropshire.

Adrian Oliver Barlow

University of York

<[email protected]>

Introduction

The People's Network is an attempt to equalise the provision of public access computers across communities, and is meant to have a particularly positive effect upon marginalised categories in society, increasing those categories' usage of computers and the Internet. That is the express aim of the UK Government, as outlined in Framework for the Future, by the Department of Culture, Media and Sport (DCMS, 2003), which states the People's Network 'has a vital role to play in delivering the government's commitment to universal access to the Internet' (Ibid: 9). Assessing the social impact of the People's Network, therefore, is crucial in determining how successful the UK has been in fulfilling those pledges.

There is a dearth of quantitative studies on the People's Network. As Colater (2001) suggests, 'for understandable reasons - lack of resources, lack of expertise, conceptual and methodological problems, a concentration on service delivery - the issue of outcome measurement has not been widely addressed' within the library and information-services sector. Colater’s call for 'outcome measurement' was echoed by Selwyn (2003), who enquired, 'are public ICT centres actually widening levels of access to ICT for those individuals who previously were not using ICT, or are they merely increasing levels of use among existing users?' This particular study is meant to address the concerns outlined by Colater and by Selwyn.

Library users in 9 libraries in the English county of Shropshire were selected as the sample base. Librarians in each of those libraries gave users the option of completing a questionnaire when they came to book a session on the Network, and collected them later. One semi-structured interview was done with a member of the library staff in each location, as well as one interview in a reference library where no questionnaires were issued. In all, 478 questionnaires were distributed across 9 different lending libraries in Shropshire.

The questionnaire asked respondents to list their postcodes, with the assurance that the data would remain confidential. 218 of the 348 respondents (60%) gave their postcodes for research purposes. Those postcodes were then entered into Mosaic, the geo-demographic package of Experian, the credit reference company. Mosaic uses 61 behavioural types and 11 categories in its data analysis, all sorted into discrete categories; and the analysis is informed by a wide range of information, from Census data to market research, incorporating different factors such as age, education, income and accommodation amongst other things.

Mosaic has 11 categories, which run from A to K, and are split into numeric subsections within each letter (such as A1, A2, B1, B2 and so on). Ultimately each overarching category contains properties where the residents are similar. For instance, although there are differences within the subcategories of category K – Rural Isolation – the people classified therein may be expected to have broadly matching consumer tastes, and share approximately similar social backgrounds, as well as life chances on a par with one another.

In the UK, there are approximately 17 households assigned per postcode; Mosaic catalogues and classifies these postcodes, allowing sophisticated insight into areas on a minute level. The in-depth catalogue makes the package very suitable for the analysis of the background of people who responded to the People’s Network questionnaire. It gives texture to the findings, and the study has ‘postcode granularity’: the nature of the software allows fine distinctions to be made, and garners insights unavailable by other means into respondents who use the People’s Network. MOSAIC therefore functions as a proxy for the social background of respondents to the study, and is used as a measure of the social impact of the Network, being an important tool for the overall analysis.

Methodology and Sampling

The empirical research for this study was conducted in Shropshire over March and April 2006. There are 23 libraries under the jurisdiction of the Shropshire County Council, and this study featured 10 of them, so just under half of the county libraries were included, both the bigger towns of the county, and a reasonable cross-section of smaller libraries. The selection of libraries was also meant to cover the county’s geography, from towns to hamlets, and all points of the compass, so it was necessary to include libraries in the remote fringes. One advantage in conducting a county-wide study is that the resulting model accounts for differing sizes and constituencies of libraries, not just focusing on the main lending libraries at the expense of smaller branch libraries.

The nine libraries were weighted within the sample; each one had their sample size determined by how many People's Network computers were in each individual library. Libraries with 1-3 PN computers received (PN x 7) questionnaires, libraries with 4-6 PN received (PN x 8) and libraries with more than 7 PN received (PN x 9). In effect this weighting meant that bigger libraries received more questionnaires than their smaller counterparts: this is meant to ensure that libraries in bigger areas with more footfall are represented adequately in the final tally.

346 out of 478 questionnaires were returned - a response rate of 74%. However, because this study uses cluster sampling and not simple random sampling, it is not possible to give an accurate estimation of the standard error for the cohort as compared to non-respondents – standard error relies on a comparison of the achieved sample against the characteristics of the general population as a whole, and this information was unavailable for comparative purposes. Likewise it is difficult to estimate the percentage of users who completed this questionnaire, as compared to the amount of users who used the People's Network during the time in which the questionnaires were available in each library. No reliable baseline exists within the Council’s records for the number of visitors who used the libraries overall within the sampling period, so comparison is not possible.

Shropshire County Council produces statistics on how many people have booked sessions, but these statistics are problematic. The Council statistics measure how many half-hour sessions are booked, not the duration of each session, so that a 2-hour session is recorded in half-hour blocks. The Council statistics do not measure the amount of unique users – if someone used the Network 7 times in a week, this is recorded as 7 separate visits, but no record is kept whether it is the same person each time. The lack of a reliable baseline against which the results can be measured indicates these results should be treated cautiously. Table 1 details the response rate for each library.

Table 1 – Libraries, People's Network computers, questionnaire response rates

There are two categories which, judging by the interview data, are likely to be under-sampled by the questionnaire. This survey was conducted before the peak tourist times in summer, so the time of year was not conducive to gathering large samples of tourist visitors. The second under-sampled category is seasonal workers, who might have been deterred from filling in this survey: a few librarians said that the questionnaire might be too advanced for some of the foreign users. Migrant workers tend to participate in the informal economy, fruit-picking, catering, helping out on farms, and come into the library to use the computers:

'We get a lot of foreign visitors. I say 'visitors' as they're working in the catering industry, packing, that sort of thing, and they use it for e-mails. It's a bit of a social thing as well, as they meet their friends here. It's more than just somewhere they can e-mail from.’ (Library 10)

Literature Review

The proportion of socially excluded people who use libraries as compared to the entire population is still quite low, which is a problem the People’s Network is theoretically meant to address (Loader and Keeble, 2004:42). One case study of a representative sample in South Wales, conducted by Selwyn (2003), found that the People's Network was not relevant to marginalised categories. Public access sites were not especially attractive to many users, having 'only moderate levels of recognition, and even lower levels of usage, amongst the general population' (Selwyn, 2003:15). From the 624 of the 1001 respondents in Selwyn's survey who did use the Internet, only 13% used computers in the library. Its use was marginal for all categories, regardless of age, sex, or social class.

Aside from Selwyn's findings, it has been argued elsewhere that the People's Network could only ever have a marginal effect on digital inequality since different tiers of inequality such as business and capital are determining of where to place new technologies on the basis of an area's status and demographic profile (Graham, 2003). Furthermore, different areas of town and cities are spatially ordered: some areas have considerable numbers of technologically-aware citizens and correspondingly many networks and information nodes, while other areas have few nodes, and this naturally has an impact on different communities' receptivity to ICT (Lash, 2002). Against this backdrop of unequal interest in, and awareness of, ICT within different communities, the People's Network may amerliorate digital inequality to some extent, but never erase it entirely (Loader and Keeble, 2004).

There is an education and class dimension to differential access. Quan-Haase et al (2002), for instance, found that the most important factor for involvement in e-government and civic participation was the educational level of Internet users themselves, not just access to Internet resources. In a UK context, Hawkins et al (2001) reported that social classes D and E were underrepresented in libraries, both in terms of their numbers proportionate to other social classes, but also in terms of the frequency which they borrowed books. In other words, classes D and E, the most disadvantaged of all, were the people least likely to use libraries and the facilities therein.

Nevertheless, Gannon-Leary et al (2003) found that in a sample of over 450 People's Network users in the North-East, 22% were unemployed, and 16% were retired. That over 100 respondents were unemployed is surprisingly high: unfortunately, Gannon-Leary does not cross-reference this figure with the ages of questionnaire respondents, meaning it is impossible to gauge the age range or the duration of unemployment of the unemployed respondents with any accuracy,.

Turner and Kendall (2000) examined the userbase in Chester Library. 178 replies were received to questionnaires and there was good evidence that the People's Network reached the disadvantaged: 22% of respondents were unemployed, and a quarter of the total Internet users were casual visitors from overseas. The majority of users had received no training in using PCs. 80% of users were under 35. Schofield et al (2004) examined a People's Network library – it had been successful in providing access to ICT for people who would otherwise have had no access. There were caveats, however: technology was often old, viewed unfavourably by the users, and online learning opportunities were not taken up to the extent local funding bodies had hoped.

Hillingdon Borough Council's ICT User Survey from 2003 looked at People's Network users' demographics; this survey covered 17 libraries and had 906 respondents. 30% of users were from ethnic minorities, more than double the amount of ethnic minorities in the borough's demographic profile. The People's Network brought new people into the library and persuaded users to embrace ICT within a library setting (Hillingdon Borough Council, 2003).

Three major sponsors have looked into the efficacy of the People's Network, and each sponsor has issued a report. For the MLA, Peter Brophy has carried out yearly reports into the progress of the People's Network. Brophy argues (2004:23), 'the People’s Network is becoming an essential infrastructural element both locally and nationally'. Brophy's report has been criticised as insufficiently rigorous, as it examined a 'relatively small sample' of library authorities from which to extrapolate general trends (Pateman, 2003).

The Tavistock Institute's report into the People's Network, Books and Bytes, has a mixed-methods approach, including interviews, case studies, documentary analysis, and regional workshops. Libraries worked hard at 'outreach' programs aimed at socially-excluded populations, and were somewhat successful in encouraging new users to enter into the libraries. But one problem with the methodology was evident: 'most library services were unable to provide data on the profile of their users of open access PN services' (Tavistock Institute, 2004:81). Without this data for profiling, the study's classifications of users was tentative on key variables such as age or gender, which is unfortunate considering the engaging nature of the study and its breadth.

The third document is a DCMS Select Committee report on public libraries which covers the success or otherwise of the People's Network (HMSO, 2005). Drawing on written and oral testimony, they stated 'this introduction to ICT will close the technological illiteracy gap for some by giving people everywhere the chance to learn to use this technological tool'. They found that the presence of People's Network was reversing a general slow decline in public library use within recent memory. However, little qualitative evidence was adduced in support of this claim.

Findings

What is the demographic profile of the userbase of the People's Nework?

The majority, 63% of users, (n=218) were 40 years old or younger. Only 11% of users (n=38) were over the age of 60. However, these raw figures are skewed by the fact that one library only allows users aged under 18 to use the People's Network; when the raw data is filtered to exclude cases from that library, only 48% of users (n=167) are under the age of 40. There were more male respondents than female (54% to 46%). This is surprising, because libraries have been traditionally used more by females. The Public Library User Surveys National Report for 2001-2002 found that nearly 60% of library visitors were female (cited in Blackburn, 2004).

Only 22% of all users surveyed used the Internet by itself in the libraries, meaning 78% of users did partake in other library activities. Of the users who pursued other activities, 22% borrowed books, while 17% borrowed books and CDs together. Interestingly, 17% of users said they used three or more library facilities as well as the Internet. In total, over half the sample also borrowed books. As only a fifth of users solely use the Internet, most of the surveyed People's Network users are highly 'integrated' users of other library offerings. The People's Network does not seem to dominate users' motives for going to the library, but appears to complement libraries' traditional services. There is some evidence from interview data that, '[children] come in to play games, but it has brought a lot of children into the library who otherwise wouldn't be here at all' (Library 9), but such data is not readily quantifiable.

The majority of respondents (64%) were light users (0-2 hours per week). This indicates that the take-up of People's Network computers is reasonably well-dispersed: there are lots of infrequent users. However, 28% of users used the PN for between 3-5 hours per week, indicating that it is from this category where any 'hard core' users might originate. Several librarians attested that the People's Network users were a disparate category, each with their own routines which may influence how long they spend on the Internet in the library:

'We have regulars, people who are here nearly every day, but then at the same time we have people who drop in, perhaps on holiday, we have people who don't have it at home and come in to use our Internet, it's very much a mixture' (Library 3)

In terms of the geo-demographic profile of the userbase, 218 respondents put their postcode details on the questionnaire as requested; exactly 60% of the total sample. Those responses were fed into Mosaic, and a chart was produced (Table 2), with library numbers listed along the horizontal axis, and the Mosaic categories on the vertical axis.

Table 2 – Respondents sorted by Mosaic classification and by library they used (n=218)

As a companion to Table 2, its counterpart Table 3 gives some context to the raw data. Next to the totals for different categories are equivalent numbers for the percentage of people in each, within Shropshire, and also within the general UK population. This data is provided for comparative purposes, to evaluate which categories are particularly over or under-represented in the sample. The Shropshire figures do not total 100% as some houses were unclassified:

Table 3 – The characteristics of the sample, against Shropshire geodemographics (n=218)

Table 3 reveals a number of interesting aspects. Categories where the sample was significantly smaller than the Shropshire average were B, C, and I, but it is worth noting category I only had one person in it. Categories where the sample was significantly larger than the Shropshire average were D, E, F, and J, however since categories E and F had very few people in them, those results should be treated with caution.

Two of the biggest categories in the study were Category J – Grey Perspectives – and Category K – Rural Isolation. Grey Perspectives is made up of areas with many quite affluent elderly people; their high level of representation may be expected bearing in mind that “in comparison with the national average, Shropshire is weighted towards the older age categories. Shropshire has a greater proportion of population in all the age categories above and inclusive of 45-49 year olds compared to England” (Shropshire County Council, 2006). Nonetheless, the number of respondents from the Grey Perspectives category is still over-represented compared to the Shropshire average.

The users in category K – Rural Isolation - were the biggest category, and this category was comprised of people from market towns and small outlying villages, and also demarcates people who live in deep countryside, including farmers and those who work on the land. While Rural Isolation is the 3rd wealthest category of all 11 in the Mosaic categories overall, this headline number can be misleading. Much of the wealth within Rural Isolation is largely to do with the influx of incomers from cities and towns moving into the countryside, and is concentrated in their hands. This is particularly prononunced in the sub-section K57, known as Summer Playgrounds, “where urban people own many second homes and where bed and breakfasts and other agro-tourism enterprises provide important sources of seasonal income” (Nomad Plus, 2006). Residents who fit into the K57 category are likely to have second homes or holiday homes in the vicinity, and their wealth increases the overall finanical score of category K.

Similarly, another sub-category within the Rural Isolation category, K58, known as Greenbelt Guardians, “contains farming communities set in areas of high landscape value which, on account of their accessibility to towns, attract a minority of very wealthy households” (Nomad Plus, 2006b). In contrast to K57 and K58, many of the other subsections in category K – K59, K60 and K61 - are comprised of long-standing dwellers in communities, people born and bred in the area, whose incomes are not on a par with those who come from outside the area. The breakdown of Category K within the People’s Network sample is set out in Table 4.

Table 4 – Category K – Rural Isolation – within this People’s Network sample (n=62)

Moving on to other topics, the categories which had significantly greater representation in the sample than the Shropshire average were Category D, named Industrial Grit. This is a predominantly urban category with skilled manual workers, who mostly live in terraced houses in close-knit industrial areas. Category H was also represented, and people in category H seem to originate in approximately the same strata as Category D. Category H - known as Blue Collar Enterprise - includes people in manufacturing or service-sector work who may have bought their own council houses. Both categories are what might be called the 'traditional working classes' as such, working in manual or low-grade service occupations, and households may run a car and have some disposable income. It was also noticeable that category D spanned 8 libraries from the entire sample, and category H spanned 7 libraries from the entire sample – these categories were not confined to libraries within the bigger towns of Shropshire, despite their superficially urban nature.

The three categories which are financially poorest and which live in the areas where people have the worst life chances – Categories F, G and I – are almost completely unrepresented in the Mosaic data, being only 5% (n=10) of all respondents. However, many of the respondents are from non-middle class areas – Categories D and H, and also the bulk of Category K as well – so in that sense the People’s Network does belie the perception that libraries are mostly populated by the middle-classes. Even in Category K, the wealthest two segments – K57 and K58 – only comprise 33% of all Category K users: it appears that the People’s Network is more likely to be used by those who are poorer than the norm in this case.

This profile of users of the People's Network seems to contradict the picture of the 'traditional library user' to some extent. The traditional user, as Blackburn remarks in his synopsis of previous research into library demographics, 'would appear to be AB, 55+, female, white, British, retired and metropolitan' (Blackburn, 2004). Contrary to Blackburn’s findings for library users as a whole, the traditional professional middle-classes – most likely to be found in Mosaic categories A, B, and C – are not especially big users of the People’s Network in Shropshire, and between them accounted for only 17% of all Mosaic postcodes in the study. The data would appear to indicate that the mass of users do not come from either extreme of the social spectrum – neither very rich nor very poor – but instead seem to cluster around the lower-middle to the middle of the spectrum.

There was little correlation between the areas which users came from, and whether they had access to the Internet anywhere else except at the library itself. It might be expected, the richer an area where someone lives, the more likely it is that they will have Internet access, because 'there is a clear relationship between economic status and Internet use. More users come from higher economic brackets' (Dutton et al, 2005: 116). Insofar as only 10 out of 218 respondents lived in the three most deprived areas, that statement is true (see Table 5). Having said that, there is seemingly no relationship between wealth of users and their access elsewhere. Despite living in well-off areas, some of the users who lived in wealthier areas only had Internet access in their local library - the reason behind this is unclear, but librarians suggested it may be related to geographical isolation meaning a lack of broadband connections for private homes in the more remote parts of the county.

Table 5 – Wealth of users in overall Mosaic categories by net access (n=218)

How far can it be said that the People’s Network is accessed by those from socially marginalized categories? The Mosaic data is unpromising in this regard, but other sources of data from the questionnaire are more forthcoming. Certainly, one aspect to consider in relation to access by people who are socially marginalized is that a very significant minority of users, that is 46% in the entire sample including people who did not give their postcode for Mosaic analysis (n=161), have no Internet access elsewhere. Furthermore, 9% of questionnaire respondents (n=31) mentioned as a reason for using the People's Network that they didn't have a computer, or have the Internet at home. For this cohort, the People's Network provides a direct reason to visit library premises.

Aside from the fact that many users have no access elsewhere, the main pull factor is that the Network is free – 82% of people (n=288) people ticked that option. Convenience and/or ease of access was very important: 72% of people (n=252) cited either or both of those factors as being influential in choosing to visit their library. It should be pointed out that none of the libraries were physically distant from the centre of the communities they serve, meaning that many users of the service may incorporate their visit into daily routines, perhaps dropping into the library after shopping, school, or work. Several of the libraries, moreover, were co-located in buildings which had cafes, community centres, or drop-in points nearby. The People's Network does not exist as a stand-alone system for library users, but is interconnected with their library habits, their urban geographies, and their socialising patterns to some extent.

For a significant proportion of users, the fact that they can use the Internet whilst being sociable is a big pull factor. Many questionnaire respondents use the library effectively as a social hub, a 'third space' between home and work (Buschman, 2005), and thus using the library was a way that users participated in their local communities on the ground. 24% (n=84) used the Internet for reasons of being sociable or getting out of the house. This was the case whether users had Internet access elsewhere or not: 40 users who ticked the 'sociable' box had access elsewhere, while 44 users did not, so this was not a substantial relationship. Even some users who could use the Internet elsewhere if they wished still found the social side of the People's Network to be very meaningful. Interview data seems to confirm many users of the People's Network seemed to know other users through their own social networks, and used the library facilities to socialise before or after a session:

'[Some users here] tend to meet people, sit next to each other. Quite often when you go to book somebody on, they know the person who's coming off, so they're chatting away together. They come in the queue, and talk, so it does give it a bit of an atmosphere you know.' (Library 10)

From the analysis of the data, it appears that the People’s Network does have an impact on some types of respondents more than others. The typical user would appear likelier to be male than female, over 40 years old, most probably from a rural background, or else from a working-class urban area, and slightly more likely to have Internet access elsewhere than not. This profile is not similar to the stereotypical library user mentioned by Blackburn (2004), so one thing that can be said is that the People’s Network in Shropshire is less socially exclusive than the library service nationally as a whole.

Furthermore, the categories likeliest to use the Internet most frequently in this study are broadly in line with those in the Tavistock Institute’s Book and Bytes evaluation. In that study, the People’s Network monitoring data identified most of the users as being either very young or fairly old, male rather than female, and as being fairly likely to come from one of the following categories: Job-seekers, retirees, students, or migrants/seasonal workers (Tavistock Institute, 2004:80). All of these categories were found to be fairly regular users of the People’s Network facilities in Shropshire, whether via questionnaire returns or via interviews with librarians, so this data does provide some quantitative support for their qualitative approach.

Discussion: The relationship between Mosaic, take-up of the People’s Network, and Bourdieu’s habitus

The categories most tempted into libraries by the People's Network, or who use libraries more often than before, seems to be those of the skilled working-classes. However, there are many categories who didn't seem to be using the People's Network, or hardly at all: these are the unskilled, the unqualified, and the marginalised in general. There are two levels to this non-usage by the socially marginalized: micro and macro.

On the micro level, there is a need for individuals to be comfortable with, and confident on, the machines: IT-literacy was a theme which recurred consistently in interviews. Academically-marginalised categories may not have the comprehension skills to use the Internet to its full capacity without some kind of training (Pawley, 2003), which was something that several librarians brought up unprompted in discussion. Adults who have low IT-literacy, linked with a lack of reading literacy, are undeniably disadvantaged when using computers (Garrod, 2003). In response to user demand for assistance, many of the libraries had embarked on a series of formal linkups involving colleges, commercial enterprises, and libraries to provide skills training:

'We've had Shropshire Training do a series of, they've had a tutor here since the beginning of the year, in fact since the beginning of 2005, it ends this Friday, whereby a tutor has come in each week to help... that has gone done very well here, it has been less successful in other venues. We gave it a fair amount of publicity, we have regulars now who come in on a Friday because someone is here.' (Library 3)

There was a clear, sustainable desire for the tuition. Librarians often commented that some sort of training would be helpful for users to profit on the Internet, for 'there is a plethora of information, and so problems with evaluating what is useful and what is accurate' (Library 1). This sentiment was often echoed: 'much of the information is incredibly good, but then there is an awful lot of dross as well' (Library 3). A clear relationship exists between IT-literacy and the ability to find information on the Internet: Hargittai (2002) found experience with the technology to be 'positively related to online skill.' Indeed, several librarians made comments of this ilk, mentioning that lack of IT-literacy often correlates with a lack of functional literacy:

'You get people who can't really use them, after leaving school, you'd expect them to be able to read but some find it difficult, and the same with computers, so they can ask for a lot of assistance and take up a lot of time'. (Library 10)

The micro level of an individual’s skills and literacy is profoundly influenced by the macro level: the influence of class, social background, and education upon that same individual’s life chances. Educational attainment varies according to social background, and the disparity in life chances and in educational attainment is clear enough from the analysis in Mosaic. Webber (2005) has shown empirically in a longitudinal study that children in Mosaic categories F, G, and I score particularly low on mean GCSE scores when compared to their peers in other categories. Of 24 Mosiac categories in Webber's study, individual sub-sections of the aggregate categories F, G, and I occupied seven out of the bottom eight placings. Categories F, G, and I are the categories which have the worst life chances, which achieve least in formal schooling, and which are noticeable by their absence from the postcode returns to this library survey. They are the three least wealthy Mosaic categories, and therefore not especially likely to afford expensive computer equipment: they seem to have a nexus of disadvantage, where one aspect of disadvantage reinforces another.

In a follow-up study conducted by Webber and Butler, also using Mosaic as the analytic tool, aside from the performance of the pupil at KS2 tests at the age of 11, the type of neighbourhood in which the pupil lived was a more reliable predictor of the pupil’s GCSE performance than any other information about the pupil. Another key finding from the study was that the performance of pupils was influenced by the neighbourhoods from which the other pupils in the school were drawn, this being only marginally less significant than their own social background (Webber and Butler, 2005).

It is therefore unsurprising that one of the markers of low educational attainment - the inability to read or write enough to function in everyday life, not being functionally literate – is most common amongst the socially marginalised. A report by the Department for Education and Employment, DfEE (1999), estimated that one adult in five in this country is not functionally literate, a number which is higher amongst the least well-off in society. There is therefore a clear structural barrier for the children and adults in the socially disadvantaged categories F, G and I, which may impede their desire or ability to use the People’s Network.

A key concept which may inform discussion at this point is habitus—Bourdieu's word for the milieu in which a person lives, makes sense of their surroundings, and undertakes a range of activities in daily life. This habitus, and accompanying material conditions, exerts an effect on the aspirations and attitudes of the individual agent, their dispositions and actions (Bourdieu, 1986). The habitus is formed by, and affects, a person’s experiences throughout life. But the habitus is not neccessarily a consciously willed process, it is a set of beliefs or ways of thinking which are taken for granted by the individual concerned: ‘the habitus - embodied history, internalized as a second nature and so forgotten as history - is the active presence of the whole past of which it is the product’ (Bourdieu and Wacquant, 1992).

Differential access to educational, cultural and social capital will naturally affect the habitus, and thereby affect people's conception of the merits of ICT: people who may have high levels of cultural and educational capital might take to the Internet quite happily, but people who lack such capital may have many more reservations or suspicions about using the Internet (Kvasny, 2005). For instance, one aspect of the People's Network which may deter people who have low levels of educational, social, or cultural capital from using the Internet is unhappiness with the milieu - for some socially excluded categories, the library environment is unwelcoming (Pateman, 1999). That people may find the library environment unwelcoming is itself an example of habitus: if someone is unused to the environment, or dislikes it through (perhaps) associations with bad experiences at school, they may feel ill-at-ease there, and thus reluctant to use the Internet in such surroundings. It makes intuitive sense, then, that segments of people in Mosaic categories F, G, and I might be particularly unlikely or unable to engage with the Internet in library settings.

Although interest in the Internet is probably linked to habitus, that has not deterred libraries from attempting to draw in new users. All the libraries put 'press adverts in the Shropshire Star' (Library 5) - in smaller villages, librarians put notices in newsletters, and circulated leaflets. The People's Network was also part of the “introduction to the library” package for new members, so 'everytime we've joined someone they get a "welcome to the library" leaflet which explains things to the people at the desk' (Library 8). Even though the Network has been running for a few years now, some librarians were still unsure as to how far knowledge about the service had penetrated local communities: 'It still amazes me how many people will still come in here and say, "have you got Internet access?"' (Library 5). One library, however, employed an innovative way of spreading the news about the People's Network:

'We do various promotions at different times. We've done leaflets which we deliberately aimed at getting out at the communities, we've left them in doctors' surgeries, vets, sport centres... to get at people who might otherwise might not know what is available. We've had bookmarks done in the past. And certainly, if any of us go and do talks to outside categories, we always plug it as well. Trying to spread the word.'(Library 2)

This particular library had vouchers which revealed where users had found the publicity material. It was effective in tracking the efficacy of the scheme: all the non-traditional places had yielded new users to varying degrees. But, amidst this talk of broadening access to the Internet via the People's Network, it is important to remember that people may have valid, clear reasons for not engaging with ICT in the first place. Education may be a barrier, but aside from that, engagement depends upon individuals having compelling motivations behind using the Internet: the individual needs to believe it will add something to life which wasn't there before.

Adoption of the Internet into daily rounties is unlikely to occur without an impetus for action, such as the individual believing the Internet will be useful for their purposes (Wyatt et al, 2002). Again, though, that impetus is linked to the habitus – if someone must ‘believe the Intenet will add something that wasn’t there before’, they need to conceptualise why the Internet will be of benefit to them; how it can broaden their horizons. The concept of habitus gives some insight into the differences between categories, why uptake in some areas is greater than others, without being too deterministic. Those in Mosaic categories F, G, and I seem to reside in areas where expectations and horizons in many aspects of life are lower than those of their counterparts in other categories; hence the disparity in participation rates with regard to the People’s Network may be partially explained by these differences.

Conclusion – How successful has the People's Network been in its core mission?

The demographic profile of users drawn from the questionnaire indicates evidence that the People’s Network is reaching people, as 46% of the achieved sample had no access elsewhere. Most users who gave postcode information (about 60% of the sample) are from three main categories. The first set is skilled manual workers, the second is older retirees, and the third category is people from relatively isolated countryside areas, depending on how rural the library was.

On the one hand, a more complicated picture exists as well; Mosaic reveals that respondents who gave their details live in reasonably affluent areas, with very few from truly impoverished areas. This is in contrast to several other People’s Network studies discussed in the Literature Review, wherein the users were often socially-disadvantaged, but it is true that a small portion of users in Shropshire wrote that they were “job-hunting” or searching for work. Nevertheless, in Shropshire as elsewhere, the People’s Network operates within an environment of spatially-ordered differential access and capability with IT.

On the other hand, public libraries serve a remit to provide the Internet to all at no cost at the point of use. Many users (80%) found the free nature of the service enticing. Indeed, by remaining free from charges, it is arguable that the People's Network performs a public service in aid of the information-poor and/or the disadvantaged. People who do not have the Internet at home are more likely to be poorer than those who do have it at home; there is a persistent correlation between household wealth and the likelihood of owning a home computer (see Oxford Internet Institute, 2005). By not charging fees for using the Internet, the People’s Network does not financially impose upon users without Internet access.

Despite the fact that the service is free at the point of use, the People's Network seemingly passes unnoticed by potential users in several Mosaic categories, in the sense that publicity does not seem to reach them. Problems concerning achieving access from underrepresented categories are potentially due to a library environment being off-putting to some poorer or less bookish users. It is hard to know how to induce these into the library setting in the first place, or if it would be worthwhile, considering that the factors which may deter them – such as low levels of literacy, a dislike of computers, or an unwillingness to go into such an environment – are difficult to remedy. The factors which lie behind differential interest in the People’s Network, as outlined in the section referring to the habitus, appears not amenable to quick fixes or solutions.

Lastly, with regards to Shropshire, it is worth reiterating the conclusion that the People’s Network appears to be attracting people who have no access elsewhere, even if those people don't appear to be socially underprivileged in many other aspects. There is convincing evidence that most users use the People’s Network alongside other library services, not in place of them. Moreover, there is no relationship between the demographic area where a user comes from, and their likelihood of having access to the Internet elsewhere than the library, which indicates that the People’s Network is attracting a wide and mixed userbase, within those parameters already outlined.

References

Blackburn, J. (2004) English Public Libraries: Research Review and Initial Strategic Recommendations, prepared for Andrew Stephens of the MLA

Bourdieu, P. (1986) The Forms of Capital, in Baron, S, Field, J, and Schuller T, (eds.), Social Capital - Critical Perspectives, Oxford: OUP

Bourdieu, P., & Waquant, L. 1992. An Invitation to Reflexive Sociology. Chicago:University of Chicago Press.

Brophy, P. (2004) The People’s Network: Moving Forward, MLA http://www.mla.gov.uk/resources/assets//I/id1414rep_pdf_4287.pdf

Buschman, J. (2005) On Libraries and the Public Sphere, Library Philosophy and Practice, 7(2)

Colater, F. (2001) Realising The Potential of Cultural Services: The Case for Libraries, Local Government Association Research Briefing 12.1

DCMS, (2003) Framework for the Future: Libraries Learning and Information in the Next Decade

Dutton, W.H et al (2005) The Internet in Britain: The Oxford Internet Survey (OxIS), May 2005

Gannon-Leary, P., Banwell, L., and Parker, S. (2003) An evaluation of the development of the People’s Network in the North East, Library and Information Research, 27(87)

Garrod, P. (2003) Framework for the Future: Access to digital skills and services, Ariadne, 36

Graham, S. (2003) The software-sorted city: Rethinking the ‘digital divide', in Graham, S, (Ed.) The Cybercities Reader, London: Routledge

Hargittai, E. (2002) Second-Level Digital Divide: Differences in people's online skills, First Monday, 7(4)

Hawkins, M., Morris, A., and Sumison, J. (2001) Socio-Economic Features of UK Public Library Users, Library Management, 22(6 and 7)

Hillingdon Borough Council, (2003) People's Network ICT User Survey Summary Report

House of Commons Select Committee on Culture, Public Libraries, London: HMSO, 24 February 2005, Media and Sport Third Report, HC 81-I.

Kvasny, L. (2005) The role of the habitus in shaping discourses about the digital divide. Journal of Computer-Mediated Communication, 10(2)

Lash, S. (2002) Critique of Information, London: Sage

Loader, B., and Keeble, L. (2004) Challenging The Digital Divide? York: Joseph Rowntree Foundation

DfEE (1999) A Fresh Start: Improving Literacy and Numeracy, London

Nomad Plus. (2006) Summer Playgrounds, http://open.nottinghamcity.gov.uk/reports/Intelligence/Community%20and%20living/Mosaic/TypeK57_Summer%20Playgrounds.pdf

Nomad Plus. (2006b) Greenbelt Guardians, http://open.nottinghamcity.gov.uk/reports%5CIntelligence%5CCommunity%20and%20living%5CMosaic%5CTypeK58_Greenbelt%20Guardians.pdf

Pateman, J. (1999) Public Libraries, Social Exclusion, and Social Class, Information for Social Change Journal, 10

Pateman, J. (2003) The People's Network: A Turning Point for Public Libraries, Information for Social Change Journal, 17

Pawley, C. (2003). Information literacy: a contradictory coupling, Library Quarterly, 73(4)

Proctor, R., Lee, H.. and Reilly, R. (1998) Access to Public Libraries: The Impact of Opening Hours, Hours Reductions and Closures, 1986-1997, British Library Research and Innovation Centre

Quan-Haase, A., Wellman, B., Witte, J., and Hampton, K. (2002) Capitalizing on the Internet: Network Capital, Participatory Capital, and Sense of Community, in Wellman, B. and Haythornthwaite, C. (ed.) The Internet in Everyday Life, Oxford: Blackwell

Schofield, F., McMenemy, D. and Henderson, K. (2004) People's network libraries: comparative case studies of old and new ICT learning centres, Library Review, 53(3)

Selwyn, N. (2003) ICT For All? Access and use of Public ICT Sites in the UK, Information, Communication, and Society, 6(3) (September 2003)

Shropshire County Council. (2006) Census - 2001 - Population & Age Structure, http://www.shropshire.gov.uk/factsfigures.nsf/open/2446F06FDB25620980256CC400419A93

Tavistock Institute, (2004) Books and Bytes: New Service Paradigms for the 21st Century Library, London: Big Lottery Fund

Turner, K., and Kendall, M. (2000) Public use of the Internet at Chester library, UK, Information Research, 5(3)

Webber, R . (2005) Geodemographic classifications and the analysis of public datasets: adding value to PLASC, presented at the Census: Past and Future conference at the University of Leicester

Webber, R., and Butler T. (2005) Classifying pupils by where they live: how well does this predict variations in their GCSE results, CASA Working Paper No 99 (Dec 2005) http://www.casa.ucl.ac.uk/working_papers/paper99.pdf

Wyatt, S, Thomas, G, and Terranova, T. (2002) They Came, They Surfed, They Went Back to the Beach: Conceptualizing Use and Non-Use of the Internet, in Woolgar, S (ed.) Virtual Society?, Oxford: OUPAddams, J. (1930). The second twenty years at Hull-House: September 1909 to September 1929. New York: Macmillan.