If You Build a Political Web Site, Will They Come?

Copyright © 2006, Idea Group Inc. Copying or distributing in print or electronic forms without written permission of Idea GroupInc. is prohibited.

International Journal of Electronic Government Research, 2(2), 1-21, April-July 2006 1

If You Build a Political Web Site,Will They Come?

The Internet and Political Activism in BritainPippa Norris, Harvard University, USA

John Curtice, Strathclyde University, UK

ABSTRACT

This study focuses on the capacity of the Internet for strengthening political activism.The first part summarizes debates about these issues in the previous literature.This study starts from the premise that political activism is a multidimensionalphenomenon and that we need to understand how different channels ofparticipation relate to the social and political characteristics of the onlinepopulation. We predict that certain dimensions of activism will probably bestrengthened by the rise of the knowledge society, particularly cause-orientedforms of political participation, reflecting the prior social and politicalcharacteristics of the online population. By contrast, we expect the Internet tohave far less impact upon conventional channels of political participation,exemplified by election campaigns. The second part summarizes the sources ofdata and the key measures of political activism used in this study, drawing uponthe British Social Attitudes Survey from 2003. The third part examines the evidencefor the relationship between use of the Internet and patterns of civic engagementin the British context. The conclusion summarizes the results and considers theirbroader implications.

Keywords: Please provide

INTRODUCTIONThe rise of knowledge societies rep-

resents one of the most profound trans-formations that have occurred in recentdecades. This phenomenon, characterizedby the widespread diffusion of informa-

tion and communication technologies(ICTs) across society, promises to havemajor consequences by expanding accessto education and training, broadeningchannels of expression and social net-works, as well as revolutionizing the na-

2 International Journal of Electronic Government Research, 2(2), 1-21, April-July 2006


ture of work and the economy. The pri-mary impact of this development has beenevident in affluent nations such as theUnited States, Sweden, and Britain, butthe Internet has also been widely regardedas an important instrument for socialchange in poorer countries with relativelyhigh levels of ICTs, such as Malaysia andBrazil (U.N., 2002; Franda, 2002). Thecore issue for this study concerns the so-cial and political consequences of the riseof knowledge societies, in particular thecapacity of the Internet for strengtheningcivic engagement.

To consider these issues, the firstpart summarizes debates about the impactof the Internet on civic engagement. Thisstudy assumes that political activism is amultidimensional phenomenon. The impactof the Internet on each of these dimen-sions, in turn, is assumed to be heavilydependent upon the social and politicalcharacteristics of Internet users. Given thisframework, the study predicts that theprimary impact of using the Internet willbe upon facilitating cause-oriented formsof political activism, thereby strengthen-ing social movements, voluntary associa-tions, and interest groups, more than uponconventional channels of political partici-pation, exemplified by election campaigns.To test these propositions in the Britishcontext, the second part summarizes thesources of data and the key measures ofpolitical activism used in this study, draw-ing upon the British Social Attitudes Sur-vey of 2003. The third part examines theevidence for the relationship between useof the Internet and indicators of civic en-gagement. The conclusion in the final part

summarizes the empirical results and con-siders their broader implications.

THEORIES OF THE IMPACT OFKNOWLEDGE SOCIETIES ONDEMOCRACY

Multiple theories exist about how thegrowth of knowledge societies could po-tentially influence political participation andcivic engagement in contemporary democ-racies. Three main perspectives can beidentified in the previous literature.

The Internet as a Virtual AgoraThe most positive view is held by

cyber-optimists, who emphasize thePanglossian possibilities of the Internet forthe involvement of ordinary citizens in di-rect, deliberative, or “strong” democracy.Digital technologies are thought to holdpromise as a mechanism facilitating alter-native channels of civic engagement, ex-emplified by political chat-rooms, remoteelectronic voting in elections, referenda,and plebiscites, and the mobilization ofvirtual communities, thereby revitalizinglevels of mass participation in public af-fairs (Barber, 1998; Budge, 1996, Rash,1997; Rheingold, 1993; Schwartz, 1996).This view was popular as the Internet ini-tially rapidly expanded in the United Statesduring the mid-1990s, and this perspec-tive continues to be expressed by enthu-siasts today (Gilder 2000). For example,Hauben and Hauben (1997) argue that bybringing people together, the Internet canhelp rebuild a sense of community and trust(see also Wellman & Guilia, 1999). Em-pirical backing for this view has come fromanalysis of the Pew Internet and Ameri-



can Life Project, which suggests thatInternet users have wider social networksthan non-users (Ranney, 2000; Robinsonet al., 2000; Pew 2001; Uslaner 2004), aresult replicated in Britain (Gardner &Oswald, 2001).

The Knowledge Eliteand Social Inequalities

Yet these claims remain highly con-troversial. Cyber-pessimists regard theknowledge society as a Pandora’s boxreinforcing existing inequalities of powerand wealth, generating deeper divisionsbetween the information rich and poor. Inthis perspective, the global and social di-vides in Internet access mean that, far fromencouraging mass participation, the growthof ICTs will disproportionately benefit themost affluent sectors in the developedworld (Golding, 1996; Hayward, 1995;Murdock & Golding, 1989; Weber,Loumakis, & Bergman, 2003). Observ-ers suggests that traditional interest groups,major parties, and governments have thecapacity to reassert their control in the vir-tual political sphere, just as traditionalmultinational corporations have the abilityto reestablish their predominance in theworld of e-commerce (Hill & Hughes,1998; McChesney 1999; Selnow, 1998;Toulouse & Luke, 1998). If political re-sources on the Internet reflect the voiceand influence of the more affluent sectorsand dominant groups, this could reinforceexisting political disparities and class bi-ases commonly found in political activismwithin democratic societies.

Politics as UsualAn alternative skeptical perspective

suggests that so far the potential of theknowledge society has failed to have adramatic impact on the practical reality of“politics as usual,” for good or ill, even incountries such as the United States at theforefront of digital technologies (Margolis& Resnick, 2000). This view stresses theembedded status quo and the difficultiesof achieving radical change to political sys-tems through technological mechanisms.During the 2000 American election cam-paign, for example, commentators suggestthat George W. Bush and Al Gore usedtheir Web pages essentially as glossyshop-windows, as fundraising tools, andas campaign ads, rather than as interac-tive “bottom up” formats facilitating pub-lic comment and discussion (Foot &Schneider, 2002; Media Metrix, 2000).Elsewhere, content analysis of politicalparty Web sites in countries as diverse asthe UK, France, Mexico, and the Republicof Korea have found that their primarypurpose has been the provision of stan-dard information about party organizationsand policies that was also widely avail-able off-line, providing more of the samerather than anything new, still less interac-tive facilities:

Party presence on the Internet seemsto represent largely an additionalelement to a party’s repertoire of actionalong with more traditionalcommunication forms rather than atransformation of the fundamentalrelationship between political partiesand the public, as some earlier



advocates of cyber democracy hoped.(Gibson, Nixon, & Ward, 2003)

Studies of the contents of govern-ment department Web sites have alsofound that these are often primarily usedfor the dissemination of information andthe provision of routine administrative ser-vices. The Internet thereby serves as anaid to good governance and transparency,but it does not necessarily function so ef-fectively as a medium expanding oppor-tunities for citizen consultation, policy dis-cussion, or other public inputs into thepolicymaking process (Allen, Juillet,Paquet, & Roy, 2001; Chadwick & May,2003; Fountain, 2001; Haque, 2002;Stowers, 1999; Thomas & Streib, 2003).

The Multidimensional Natureof Political Activism

In contrast to these perspectives, thisstudy follows the convention establishedby Verba and his colleagues (Verba, Nie,& Kim, 1978; Verba, Schlozman, &Brady 1995) in assuming that political ac-tivism is multidimensional with many dis-tinct forms of involvement, each associ-ated with differing costs and benefits. Theimpact of the Internet can be expected todiffer in each of these, mainly due to theprior social and political characteristics ofthe online population. Four main catego-ries can be distinguished, each with dif-ferent costs and benefits: voting, cam-paign-oriented activism, cause-orientedactivism, and civic-oriented activism.

Voting in regular elections is one ofthe most ubiquitous forms of citizen-ori-ented participation, requiring some initia-

tive and awareness for an informed choicebut making fairly minimal demands of time,knowledge, and effort. Through the ballotbox, voting exerts diffuse pressure overparties and elected officials, and the out-comes of elections affect all citizens. Par-ticipating at the ballot box is central to citi-zenship in representative democracy butdue to its relatively low costs the act isatypical of other, more demanding formsof participation. The Internet can be ex-pected to encourage voting participationmainly by lowering some of the informa-tion hurdles to making an informed choice,although the provision of remote electronicvoting through a variety of new technolo-gies can be expected to have a more radi-cal impact upon turnout (Tolbert &McNeal, 2003; Norris, 2004).

Campaign-oriented forms of par-ticipation concern acts focused primarilyupon how people can influence parliamentand government in representative democ-racy, primarily through political parties inBritish politics. Verba, Nie, and Kim fo-cus on this aspect when they defined po-litical participation as “those legal activi-ties by private citizens that are more orless directly aimed at influencing the se-lection of governmental personnel and/orthe actions they take” (Verba et al., 1978,p. 46). Work for parties or candidates,including party membership and volunteerwork, election leafleting, financial dona-tions to parties or candidates, attendinglocal party meetings, and get-out-the-votedrives, all typify this category. Parties servemultiple functions in representative democ-racies: notably simplifying and structuringelectoral choices; organizing and mobiliz-



ing campaigns; aggregating disparate in-terests; channeling political debate; select-ing candidates; structuring parliamentarydivisions; acting as policy think tanks; andorganizing government. Not only are par-ties one of the main conduits of politicalparticipation, they also serve to boost andstrengthen electoral turnout. If mass partymembership is under threat, as many indi-cators suggest, this could have seriousimplications for representative democracy(Mair & van Biezen 2001; Scarrow,2001). Campaigning and party work typi-cally generates collective rather than indi-vidual benefits but requires greater initia-tive, time, and effort (and sometimes ex-penditure) than merely casting a ballot.The Internet can be expected to providenew opportunities for activism in partiesand election campaigns, for example,through downloading information, joiningparties or donating funds, or participatingin discussion groups hosted on party orcandidate Web sites (Gibson, Nixon, &Ward, 2003; Hague & Loader, 1999;Norris, 2001). At the same time, the onlinepopulation is usually younger than aver-age, while party members and activists aretypically drawn from middle-aged andolder sectors of society. The social char-acteristics of online users mean that theyare unlikely to be drawn toward partyWeb sites and thus traditional forms ofcampaign activism.

Cause-oriented activities are fo-cused primarily upon influencing specificissues and policies. These acts are exem-plified by whether respondents have ac-tual experience in taking part in demon-strations and protests, signing a petition,

or raising an issue in the news media. Thedistinction is not water-tight; for example,political parties can organize mass dem-onstrations, and social movements oftenadopt mixed action strategies that com-bine traditional repertoires, such as lob-bying representatives, with a variety of al-ternative modes such as online network-ing, street protests, and consumer boy-cotts. Nevertheless, compared with cam-paign-oriented actions, the distinctive as-pect of cause-oriented repertoires is thatthese are most commonly used to pursuespecific issues and policy concerns amongdiverse targets, both within and also wellbeyond the electoral arena. These actsseek to influence representative democ-racies within the nation-state through theconventional channels of contactingelected officials, ministers, civil servants,and government departments, but theirtarget is often broader and more diffuse,possibly in the non-profit or private sec-tors, whether directed at shaping publicopinion and “life-styles,” publicizing cer-tain issues through the news media, mobi-lizing a networked coalition with othergroups or non-profit agencies, influencingthe practices of international bodies suchas the World Trade Organization or theUnited Nations, or impacting public policyin other countries. The proliferation ofcause-oriented Web sites, combined withthe typical age and educational profile ofthe online population, makes this a par-ticularly rich area of activism that can beexpected to be reinforced through theInternet.

Lastly civic-oriented activities, bycontrast, involve membership and work-



ing together in voluntary associations, aswell as collaborating with communitygroups to solve a local problem. The coreclaim of “Toquevillian” theories of socialcapital is that typical face-to-face delib-erative activities and horizontal collabo-ration within voluntary organizations farremoved from the political sphere — ex-emplified by trade unions, social clubs, andphilanthropic groups — promote interper-sonal trust, social tolerance and coopera-tive behavior. In turn, these norms are re-garded as cementing the bonds of sociallife, creating the foundation for buildinglocal communities, civil society, and demo-cratic governance. In a “win-win” situa-tion, participation in associational life isthought to generate individual rewards,such as career opportunities and personalsupport networks, as well as facilitatingcommunity goods, by fostering the capac-ity of people to work together on localproblems. Civic organizations such asunions, churches, and community groups,Putnam suggests, play a vital role in theproduction of social capital where theysucceed in bridging divisive social cleav-ages, integrating people from diversebackgrounds and values, promoting “hab-its of the heart” such as tolerance, coop-eration, and reciprocity, thereby contrib-uting toward a dense, rich, and vibrantsocial infrastructure (Putnam, 1993, 1996,2000, 2002; Pharr & Putnam, 2000). Thisdimension involves direct action within lo-cal communities, such as raising funds fora local hospital or school, where the pre-cise dividing line between the “social” and“political” breaks down. Trade unions andchurches, in particular, have long been re-

garded as central pillars of civic society,which have traditionally served the func-tion of drawing European citizens into pub-lic life. For a variety of reasons, includingthe way that voluntary associations canstrengthen social networks, foster leader-ship skills, heighten political awareness,create party linkages, and facilitate cam-paign work, people affiliated with church-based or union organizations can be ex-pected to participate more fully in publiclife. (Cassel, 1999; Radcliff & Davis,2000) Access to the knowledge societycan be expected to expand social net-works and information, facilitating mem-bership in civic associations and socialgroups, although the evidence whether theInternet strengthens or weakens socialcapital remains under debate (Bimber,1998; Horrigan, Rainie, & Fox 2001).

Therefore the argument developedin this study rejects the view that every-thing will change as the Internet facilitatesradical forms of direct democracy thatcome to replace the traditional channelsof representative governance (as optimistsoriginally hoped). Nor do we accept thatthe digital divide will inevitably reinforceexisting socio-economic and demographicdisparities in political activism (as pessi-mists predicted). The alternative view thatnothing will change as the digital worldmerely replicates “politics as usual” (as theskeptics suggest) also seems implausible.Instead the argument developed here sug-gests that we need to understand the mul-tidimensional nature of political activismand how this interacts with the character-istics of Internet users. We predict thatcertain dimensions of activism will prob-



ably be strengthened by the rise of theknowledge society, particularly cause-ori-ented forms of political participation, re-flecting the prior social and political char-acteristics of the online population.

CONCEPTUAL FRAMEWORK,EVIDENCE, AND SURVEY DATA

Interpretations of the results of theexisting empirical studies about the role ofICTs on civic engagement remain dividedin part because the outcome may be con-ditional upon the characteristics of earlyInternet adopters, the location of individu-als within society, and the particular typesof society under comparison. It remainsunclear how far the patterns uncovered inmuch of the existing U.S. research onthese issues are conditioned by the par-ticular characteristics of American politicsand society, or how far the findings holdmore generally elsewhere. One way toexplore this issue is to consider the evi-dence for the uses and functions of theInternet in Britain. Both Britain andAmerica are affluent post-industrial ser-vice economies, sharing strong culturallinks. Yet significant contrasts are also evi-dent, including in general levels of elec-toral turnout, party activism, and associa-tional membership (Norris 2003).

To explore the impact of Internet useupon political activism in the British con-text, this paper draws upon the results ofa multi-wave research project based on aspecial battery of items, contained in theBritish Social Attitudes (BSA) 2003. TheBSA is a representative national surveythat has been conducted every year since1983. Each survey is conducted using

more than 3,000 interviews with a repre-sentative random sample of people in Brit-ain. The 2003 survey monitored use ofthe Internet, social capital, and civic en-gagement. The core items are being re-peated in two subsequent waves of theBSA surveys, allowing the evolution of theInternet to be monitored over time.

The 2003 survey gauged contempo-rary Internet access at home and at work,as well as measuring a variety of ways ofusing new communication and informationtechnologies. To illustrate the baselinepopulation, Figure 1 shows how far peopleuse the Internet in the British sample. Over-all, almost half (47%) of respondentsnever used the Internet, so that 53% wentonline in Britain, either at home, at work,or somewhere else. This proportion isabout 8-10% less than the comparable fig-ure in America during the equivalent pe-riod, as monitored by the Pew Internet &American Life Project1. Among the onlinepopulation in Britain, just over one quar-ter (28%) had access only at home, whileabout one quarter had access both at homeand at work, and a few (4%) only ac-cessed the Internet at work. These esti-mates confirm, as expected, that Internetaccess currently remains lower in Britainthan in the United States. Nevertheless thePew surveys suggest that for the last fewyears American Internet access appearsto have stabilized at around two-thirds ofthe U.S. population, while by contrast Fig-ure 2 indicates that in Britain personal useof the Internet has not yet hit a ceiling andindeed continues to expand slightly in re-cent years, albeit at a slower rate thanearlier. The 2005 BSA suggests that use



has now spread slightly further, to 56% ofrespondents.

As Internet use has diffused morewidely in Britain, the social compositionof the online population has become lessdistinctive in its higher educational back-ground and its younger age profile, al-though analysis of the social characteris-tics of the online population by Bromley(2004) confirms that important digital di-vides in Internet access remain in Britain,as commonly found elsewhere (Norris,2001; Wilhelm, 1999). This means thatany multivariate analysis of the impact ofInternet use needs to control for the priorsocial and demographic characteristics ofusers, including their age, sex, race, edu-cation, and social class.

Nevertheless, with only cross-sec-tional survey data it remains difficult to iso-

late and disentangle the impact of accessto the Internet from the social backgroundof users. The BSA-2003 does provide aproxy indicator, however, by monitoringwhen people reported that they first wentonline. This makes it possible to comparethe groups who are and are not online, aswell as to compare the group who onlyrecently started to use the Internet againstthe early adopters, who commenced us-ing the Internet five or more years ago (i.e.,prior to 1998). Any effect from the cumu-lative experience of using the Internetshould be apparent if we find some sig-nificantly different political attitudes andbehavior among these groups, for example,if more experienced users acquire civicskills and social networks online that en-courage them to become more active incommunity affairs. Figure 3 illustrates some

Figure 1. Access to the Internet in Britain, 2003

Source: The British Social Attitudes Survey, 2003

4

22

28

47

0

10

20

30

40

50

Work Only Personal and Work Personal Only Neither



of these common contrasts, for example,how the most experienced users were sig-nificantly more likely to use the Internetfor email, banking, and news. At the sametime caution is needed when drawing in-ferences based on length of Internet ex-perience, because over time the onlinepopulation has gradually “normalized” inits characteristics, becoming more repre-sentative of the general public. As a re-sult, it remains unclear whether any con-trasts between the more and less experi-enced user groups can be attributed toInternet exposure per se, or to the “nor-malization” of the online population.

If we establish that levels of civicengagement are significantly associated

with use of the Internet, we still need toestablish a suitable benchmark to evalu-ate the strength of this relationship. Thisstudy therefore compares Internet use withthe role of regular newspaper readership.Studies have commonly found that thosewho habitually use and pay attention tonewspapers are significantly more knowl-edgeable than the average citizen aboutparty policies, civics, and candidates, aswell as being more interested in public af-fairs and more likely to turnout to vote(Miller, 1991; Newton, 1997; Norris etal., 1999, p. 113). Similar patterns arecommonly found elsewhere, with regularnewspaper readers more informed andengaged than average (Norris 2000). The

Figure 2. Growing home use of the Internet in Britain, 1999-2003

Note: “Do you yourself ever use the Internet or World Wide Web for any reason (other than yourwork)?”Source: The British Social Attitudes surveys, 1999-2003

22

31

4446

51

0

10

20

30

40

50

60

1999 2000 2001 2002 2003

% at home % personal use



models in this study therefore compare therelationship between Internet use and civicengagement with that between regularnewspaper readership and civic engage-ment. Controlling for newspaper reader-ship also provides a way to monitor priorpolitical predispositions.

The design of the items measuringcivic engagement are based upon the as-sumption that involvement is multidimen-sional, with many distinct forms, each as-sociated with differing costs and benefits(Verba et al., 1978, 1995). This studycompares the impact of Internet use andnewspaper readership on the four main

dimensions of activism already discussed,namely: voting, campaign-oriented, cause-oriented, and civic-oriented. These aresummarized into a Political Activism In-dex combining all dimensions2. The basicitems used to develop the measures arelisted in Appendix A.

Voting participation is measured inthe BSA 2003 by whether respondentsreported that they recalled voting in the2001 British general election.

Experience of campaign-orientedactivism is gauged in this study by a three-item scale including whether people aremembers of a party3, whether they ex-

5 years+ 3-5 years 1 to 3 years Under a year

Length of experience of using the Internet

0.20

0.40

0.60

0.80

% O

nlin

e w

ho u

se th

e In

tern

et fo

r...

EmailNews and informationGeneral informationBankingNews and information

Figure 3. Uses of the Internet by length of experience

Source: The British Social Attitudes Survey, 2003



press a “fairly” or “very” strong party iden-tification, and whether they have ever con-tacted their MP or MSP. Admittedly, thesemeasures are far from ideal. Party identi-fication is a psychological attitude, ratherthan form of behavior. Many people con-tact their MP for reasons that have noth-ing to do with campaigning per se. Bettermeasures of campaign activism would alsogauge typical activities, such as canvass-ing, donating funds to parties, or attend-ing party meetings. These will be moni-tored in subsequent BSA surveys, but themeasures used here can be regarded asat least general, if far from perfect, prox-ies for campaign activism that are avail-able in the 2003 survey.

Cause-oriented activism is mea-sured in this study by a seven-battery itemincluding whether, in response to an un-just or harmful government action, peoplehave signed a petition, spoken to an influ-ential person, contacted a governmentdepartment, raised the issue in an organi-zation, formed a group of like-mindedpeople, contacted radio, TV, or a news-paper about the issue, or gone on a pro-test or demonstration.

Lastly, experience of civic activismis gauged here by a 14-point scale sum-marizing membership in a series of differ-ent types of voluntary organization and as-sociations, focusing mainly upon traditionalsectors such as parent-teacher associa-tions, charitable organizations, churchgroups, and social clubs (excluding partymembership), although also including some“new” social movements, exemplified bygroups concerned about the environmentand about international issues.

The summary Political Activism In-dex, providing an overview, is constructedsimply by adding together experience ofeach of these different types of acts (eachcoded 0/1). The study has thereforeformed additive indices of each item withineach of the four groups and also createdan additive index across all four groups. Itshould be noted that the resulting summaryindex is currently heavily dominated by thecivic-oriented scale, as this has the mostitems. It should also be noted that in thisconceptual framework, with the importantexception of partisan identification, thisstudy focuses upon political activity; weare concerned with doing politics ratherthan being attentive to public affairs orhaving psychological attitudes thought con-ducive to civic engagement, such as trustin parliament or a sense of political effi-cacy, explored elsewhere (Curtice &Norris, 2004). The study therefore doesnot regard exposure or attention to masscommunications, exemplified by follow-ing political events in newspapers, as in-dicators of political activism per se. Thesefactors may indeed plausibly contributetoward participation, and thereby helpexplain this phenomenon, as prior pre-conditions, but they are not, in themselves,channels that citizens can use for express-ing political concerns or mobilizing groupinterests.

Analysis of the ResultsWe can start by examining the simple

bivariate relationships between Internet useand these multidimensional indicators ofpolitical activism, with the important pro-viso that the background of online users



in Britain continues to be skewed towardthe well-educated and more affluent so-cial sectors, which are both resourcesclosely associated with political activism,so that multivariate analysis is requiredcontrolling for these factors.

Table 1 presents the mean score onthese indicators for the group of all Internetusers compared with all non-users, thestandard deviation, and the significance ofthe difference (estimated by ANOVA). Inline with our initial expectations, the re-sults confirm that Internet users in Britainproved significantly more politically en-gaged than non-users across the indica-

tors of cause-oriented and civic-orientedparticipation, as well as in the total activ-ism scale. This pattern suggests that theforms of political involvement that are mostlikely to benefit through the developmentof the Internet are through single-issuepolitics, voluntary associations, and com-munity groups, as expected given the priorcharacteristics of the online population. Bycontrast, if we compare more traditionalforms of engagement, Internet users wereslightly less likely to vote than non-users,and there were no significant differencesbetween users and non-users in their lev-els of campaign activism.

Table 1. Mean scores on the indicator of political activism scales by experienceof Internet use, without any controls

Notes: The significance of the mean difference between Internet users and non-users is measuredby ANOVA. See the text for details of the construction of the scales.Source: The British Social Attitudes Survey, 2003

Uses Internet Political activism

scale

Voted Campaign-oriented activism

Cause-oriented activism

Civic-oriented activism

Non-users 2.15 .69 .48 .43 .55

All Internet users 2.65 .65 .51 .65 .83

Difference +0.50 -.04 +.03 +.22 +.28

Sig. .000 .004 .149 .000 .000

Eta .117 .044 .022 .129 .129

SD Non-users 1.83 .46 .48 .74 .55

SD all Internet users 2.36 .48 .51 1.01 1.21

Used the Internet...

Under a year 2.79 .66 .49 .75 .93

1-3 years 2.94 .67 .56 .74 .98

3-5 years 3.39 .63 .66 .92 1.14

5 years+ 3.74 .65 .65 1.07 1.39

Difference +0.95 -0.01 +0.16 +0.32 0.46

Sig. .000 .486 .009 .000 .000

Eta .147 .039 .084 .126 .133

Total sample 2.38 .67 .50 .53 .68



The comparison of activism by lengthof Internet use displays a slightly differentpattern, illustrated in Figure 4, where earlyadopters with the longest experience ofgoing online proved significantly morepolitically active across all dimensions ex-cept voting turnout (where there was nodifference) when compared against thosewho had first ventured online more re-cently. Yet without any controls it is notpossible to determine whether this patternis due to the effects of length of exposureto the Internet per se or whether this re-

flected the gradual changes in the socialcomposition and political attitudes of theInternet population.

The multivariate regression analysispresented in Table 2 displays the resultsof the full model predicting overall levelsof political activism, including the range ofsocial and demographic controls as wellas the impact of regular newspaper read-ing and use of the Internet. The resultsshow that, as expected, age had a signifi-cant and strong effect upon political par-ticipation, with people becoming more

Table 2. The impact of Internet use on the overall political activism scale, withcontrols (full model)

Notes: The model presents the results of OLS regression analysis where the dependent variableis the overall Political Activism scale. Class is categorized by the respondent’s occupationwhere the “working class” functions as the contrast category in the model. The significantvariables are highlighted in bold.Source: The British Social Attitudes Survey, 2003

Unstandardized Coefficients

Standardized Coefficients

Sig.

B Std. Error Beta P.

(Constant) -.580 .137 .000

DEMOGRAPHIC

Age (in years) .032 .002 .271 .000

Sex (Male=1/Female=0) -.008 .062 -.002 .902

Ethnic minority (1) .000 .116 .000 .998

SOCIO-ECONOMIC

Education (Highest qualification on 7-point scale)

.242 .019 .251 .000

Managerial .433 .088 .097 .000

Lower white collar -.007 .052 -.002 .895

Petit bourgeoisie .037 .041 .014 .360

Foremen and technicians .032 .025 .020 .199

MEDIA USE

Regular newspaper reader (Normally reads paper at least 3 times a week)

.273 .061 .065 .000

Uses Internet (Yes=1/No=0) .323 .074 .076 .000

Adjusted R2 .135



active as they enter middle age, with aslight fall in a curvilinear pattern in the over-70s. Interestingly, gender and race appearto be insignificant predictors of activism inthis model, although women and men inBritain have been found to have differentpatterns of participation in other studies(Norris, Lovenduski, & Campbell 2004).As numerous studies have reported, byproviding civic skills and boosting a senseof internal efficacy, education is one of thestrongest influences upon activism (Verbaet al., 1978, 1995). Graduates and thosewith higher educational qualifications areconsistently the most politically engaged.

Social class also plays a significant role inparticipation, with managers and profes-sionals the most engaged, in part becauseoccupational status is so closely relatedto educational qualifications. Even with thisbattery of controls, both regular newspa-per readership and Internet use remain sig-nificant predictors of political activism.This relationship may obviously be recip-rocal, with knowledge and engagementencouraging media habits, as well as viceversa (Norris, 2000). The way in whichthe Internet serves as a source of generalinformation about news, current affairs,and political events may help provide the

Figure 4. Civic engagement by length of experience of using the Internet

Source: The British Social Attitudes surveys, 1999-2003

Never Last 6 months

6 months to 1 year

1 to 3 years 3 to 5 years 5 years+

When R started using the Internet? Q301

0.40

0.60

0.80

1.00

1.20

1.40

Mea

n

Campaign-oriented activismCivic-oriented activismCause-oriented activismVoted

Campaign

Voted

Civic

Cause



knowledge and confidence that arestrongly associated with active engagementin public affairs. And patterns of activismthat develop may well encourage greateruse of the Internet as a way to find outabout current affairs, government services,or events in the news. In this regard, theInternet may appear to function in a simi-lar way to newspapers, while also pro-viding social networks and reinforcingcontacts that can help mobilize citizens inthe public sphere.

Similar models were run with thebattery of controls for all the four sepa-rate indicators of political activism. Theresults in Table 3 confirm that Internet us-ers remain more active than non-users incause-oriented and civic-oriented formsof participation, suggesting that this is notsimply a product of their distinctive social

profile in terms of their age, gender, race,education, and class. The differenceamong users and non-users remains in-significant in the more traditional campaign-oriented forms of activism. And contraryto expectations, after applying these con-trols, Internet users became significantlymore likely to vote, not less. It may bethat the ubiquity and particular character-istics of this activity, with the lowest de-mands of time and energy, mean that vot-ing participation is associated with bothmedia. When the effects of Internet useare compared with those associated withregular newspaper readership, the patternsshow the strongest contrast between tra-ditional forms of campaign activism (whichare significantly related to regular news-paper readership) and cause-oriented ac-tivism (which are significantly related to

Table 3. Summary models of the impact of media use on the activism indicators,with demographic and social controls (not presented)

Notes: The table presents the results of regression analysis models, including the unstandardizedbetas (B), the standard error (s.e.), and their significance, where the dependent variables arethe indicators of the four main dimensions of political activism. The full model presented inTable 2 is used, controlling for the respondent’s age, sex, race, education, and occupationalclass, although these coefficients are not reported here. A binary logistic model is used forvoting participation and OLS linear models for the other scales. The significant variables arehighlighted in bold.Source: The British Social Attitudes Survey, 2003

Voted Campaign-oriented activism



B se sig B se sig B se sig B se sig

Regular newspaper reader (Normally reads paper at least 3 times a week)

.280 .020 .000 .117 .020 .000 .036 .027 .172 .067 .032 .036

Uses Internet (Yes=1/No=0)

.216 .086 .012 .043 .024 .071 .073 .033 .026 .158 .039 .000



Internet use). The implications of thesepatterns for the political participation indemocracy are considered next.

CONCLUSIONAND DISCUSSION

The rise of the Internet has gener-ated considerable interest and concernabout its possible consequences for gov-ernment and democracy. The bursting ofthe Internet economic bubble dampenedthe more utopian political hopes as well,and the conventional wisdom shifted in amore skeptical direction. Yet in reality boththe many hopes and fears may well proveto have been exaggerated, although thisdoes not mean that there are no signifi-cant political consequences flowing fromthe development of new communicationand information technologies.

The conclusion from this study is thatany analysis of the impact of using theInternet needs to take into account thedistinct dimensions by which people chan-nel their activism into public affairs. Weneed to understand how the types ofInternet activism interact with the socialprofile of the online population. This ac-count suggests that the most popular formsof online activism are likely to reflect thepreponderance of younger and well-edu-cated populations using the Internet, inBritain and elsewhere, until such a time asthe online population eventually “normal-izes” to reflect a cross-section of the gen-eral electorate.

The conclusion from the British sur-vey evidence is that the potential impactof the Internet on democratic participa-tion depends heavily upon the type of ac-

tivism under comparison. The online popu-lation is most predisposed to engage incause-oriented forms of activism, charac-teristic of petitioning, demonstrating, andcontacting the media over single-issuepolitics and civic-oriented activities, suchas belonging to voluntary associations andcommunity organizations. By contrast, tra-ditional campaign-oriented forms of po-litical activism are associated morestrongly with newspaper readership. Thepatterns by voting participation suggest thatwithout any controls, Internet users areless likely to turnout than non-users, al-though this pattern is reversed once con-trols are introduced for the age, educa-tion, gender, and class of the online popu-lation. Subsequent surveys will monitorhow far this pattern persists or evolveswith newer developments in the Internet.The implications of these findings are notsimply about whether use of the Internetwill mobilize citizens at individual level butalso for the type of political practices thatthe rise of the Internet might encourage inthe political system. What seems appar-ent is that use of the Internet by politicalparties seems unlikely to stem any ero-sion in traditional campaign-oriented ac-tivities. At the same time, the new tech-nologies will probably prove to be of great-est benefit to engaging supporters in so-cial movements, transnational policy net-works, and single-issue causes, encour-aging their expansion in many democra-cies.

ACKNOWLEDGMENTSThe special battery of items on

Internet use was part of the E-society re-



search program funded by the Economicand Social Research Council of the UK.The authors are most grateful to the ESRCfor supporting this project and to all thestaff of The National Centre for SocialResearch (NatCen) for administering andconducting the BSA 2003 survey. Formore details, see http://www.natcen.ac.uk.

REFERENCESAllen, B. A., Juillet, L., Paquet, G., & Roy,

J. (2001). E-governance & governmenton-line in Canada: Partnerships, people& prospects. Government Informa-tion Quarterly, 18(2), 93-104.

Barber, B. R. (1998). Three scenarios forthe future of technology and strong de-mocracy. Political Science Quarterly,113(4), 573-589.

Bimber, B. (1998). The Internet and po-litical transformation: populism, commu-nity and accelerated pluralism. Polity,XXXI(1), 133-160.

Bimber, B. (2000). The study of informa-tion technology and civic engagement.Political Communication, 17(4),329-333.

Bimber, B. (2001). Information and po-litical engagement in America: Thesearch for effects of information tech-nology at the individual level. PoliticalResearch Quarterly, 54(1), 53-67.

Boas, T. C. (2000). The dictator’s di-lemma? The Internet and U.S. policytoward Cuba. The Washington Quar-terly, 23(3), 57-67.

Bonfadelli, H. (2002). The Internet andknowledge gaps: a theoretical and em-pirical investigation. European Journalof Communication, 17(1), 65-84.

Bromley, C. (2004). Can Britain close thedigital divide? In A. Park (Ed.), Britishsocial attitudes, 21st report. London:Sage.

Budge, I. (1996). The new challenge ofdirect democracy. Oxford: PolityPress.

Carpini, M. X. D. (2000). Gen.com:Youth, civic engagement, and the newinformation environment. PoliticalCommunication, 17(4), 341-349.

Cassel, C. A. (1999). Voluntary associa-tions, churches, and social participationtheories of turnout. Social ScienceQuarterly, 80(3), 504-517.

Chadwick, A., & May, C. (2003). Inter-actions between states and citizens inthe age of the Internet: ‘E-government’in the United States, Britain and the Eu-ropean Union. Governance, 16(2),271-300.

David, R. (1999). The web of politics.Oxford: Oxford University Press.

Davis, R., & Owen, D. (1998). Newmedia and American politics. NewYork: Oxford University Press.

Drake, W. J., Kalathil, S., & Boas, T. C.(2000). Dictatorships in the digital age:Some considerations on the Internet inChina and Cuba. iMP: The Magazineon Information Impacts. RetrievedOctober from www.cisp.org/imp

Foot, K. A., & Schneider, S. M. (2002).Online action in campaign 2000: Anexploratory analysis of the US politicalweb sphere. Journal of Broadcasting& Electronic Media, 46(2), 222-244.

Fountain, J. E. (2001). Building the vir-tual state: Information technologyand institutional change. Washington,



DC: Brookings Institution Press.Franda, M. (2002). Launching into

cyberspace: Internet developmentand politics in five world regions.Boulder, CO: Lynne Rienner.

Gibson, R., Nixon, P., & Ward, S. (Eds.).(2003). Political parties and theInternet: Net gain? London:Routledge.

Gilder, G. (2000). Telecosm: How infi-nite bandwidth will revolutionize ourworld. New York: Free Press.

Golding, P. (1996). World wide wedge:division and contradiction in the globalinformation infrastructure. MonthlyReview, 48(3), 70-85.

Hague, B. N., & Loader, B. D. (Eds.).(1999). Digital democracy: Dis-course and decision making in theinformation age. New York:Routledge.

Haque, M. S. (2002). E-governance inIndia: Its impacts on relations amongcitizens, politicians and public servants.International Review of Administra-tive Sciences, 68(2), 231-250.

Hayward, T. (1995). Info-rich, info-poor: Access and exchange in theglobal information society. K.G. Saur.

Hill, K. A., & Hughes, J. E. (1998).Cyberpolitics: Citizen activism in theage of the Internet. Lanham, MD:Rowan & Littlefield.

Horrigan, J., Rainie, L., & Fox, S. (2001).Online communities: Networks thatnurture long-distance relationshipsand local ties. Pew Internet & Ameri-can Life Project. Retrieved from http://www.pewinternet.org

Jennings, M. K., & Zeitner, V. (2003).

Internet use and civic engagement —A longitudinal analysis. Public Opin-ion Quarterly, 67(3), 311-334.

Johnson, T. J., & Kaye, B. K. (2003).Around the World Wide Web in 80ways — How motives for going onlineare linked to Internet activities amongpolitically interested Internet users. So-cial Science Computer Review,21(3), 304-325.

Kalathil, S., & Boas, T. C. (2003). Opennetworks closed regimes: The impactof the Internet on authoritarian rule.Washington, D.C.: Carnegie Endow-ment for International Peace.

Kamarck, E. C., & Nye, J. S., Jr. (1999).Democracy.com? Governance in anetworked world. Hollis, NH: HollisPublishing.

Keck, M. E., & Sikkink, K. (1998). Ac-tivists beyond borders — Advocacynetworks in international politics.Ithaca, NY: Cornell University Press.

Kent Jennings, M., & Zeitner, V. (2003).Internet use and civic engagement —A longitudinal analysis. Public Opin-ion Quarterly, 67(3), 311-334.

Mair, P., & van Biezen, I. (2001). Partymembership in twenty European de-mocracies 1980-2000. Party Politics,7(1), 7-22.

Margolis, M., & Resnick, D. (2000).Politics as usual: The cyberspace‘revolution’. Thousand Oaks, CA:Sage.

McChesney, R. W. (1999). Rich media,poor democracy. Champaign, IL: Uni-versity of Illinois Press.

Media Metrix. (October 2000). Cam-paign 2000: Party politics on the



World Wide Web. Retrieved from http://www.mediametrix.com

Murdock, G., & Golding, P. (1989). In-formation poverty and political inequal-ity: Citizenship in the age of privatisedcommunications. Journal of Commu-nication, 39, 180-195.

Norris, P. (2000). A virtuous circle. NewYork: Cambridge University Press.

Norris, P. (2001). Digital divide. NewYork: Cambridge University Press.

Norris, P. (2002). Democratic phoenix:Reinventing political activism. NewYork: Cambridge University Press.

Norris, P. (2003). The bridging and bond-ing role of online communities. In P. N.Howard, & S. Jones (Eds.), Societyonline: The Internet in context. Thou-sand Oaks, CA: Sage.

Norris, P. (2004). Will new technologyboost turnout? In N. Kersting, & H.Baldersheim (Eds.), Electronic votingand democracy: A comparativeanalysis (pp. 193-225), London:Palgrave.

Norris, P. (2005). The impact of theInternet on political activism: Evidencefrom Europe. International Journal ofElectronic Governance, 1(1), XX-XX.

Norris, P., Lovenduski, J., & Campbell,R. (2004). Gender and political par-ticipation (Report for the UK Elec-toral Commission).

Norris, P., & Sanders, D. (2003). Me-dium or message? Political Commu-nications.

Pharr, S., & Putnam, R. (Eds.). (2000).Disaffected democracies: What’stroubling the trilateral countries?

Princeton, NJ: Princeton UniversityPress.

Putnam, R. D. (1993). Making democ-racy work: Civic traditions in mod-ern Italy. Princeton, NJ: Princeton Uni-versity Press.

Putnam, R. D. (1996). The strange dis-appearance of civic America. TheAmerican Prospect, 24.

Putnam, R. D. (2000). Bowling alone:The collapse and revival of Ameri-can community. New York: Simon andSchuster.

Putnam, R. D. (Ed.). (2002). Democra-cies in flux. Oxford: Oxford Univer-sity Press.

Radcliff, B., & Davis, P. (2000). Labororganization and electoral participationin industrial democracies. AmericanJournal of Political Science, 44(1),132-141.

Rash, W., Jr. (1997). Politics on the net:Wiring the political process. NewYork: W.H. Freeman.

Rheingold, H. (1993). The virtual com-munity: Homesteading on the elec-tronic frontier. Reading, MA: AddisonWesley.

Scammell, M. (2000). The Internet andcivic engagement: The age of the citi-zen-consumer. Political Communica-tion, 17(4), 351-355.

Scarrow, S. (2001). Parties without mem-bers? In R. Dalton, & M. Wattenberg(Eds.), Parties without partisans.New York: Oxford University Press.

Schwartz, E. (1996). Netactivism: Howcitizens use the Internet. Sebastapol,CA: Songline Studios.

Selnow, G. W. (1998). Electronic



whistle-stops: The impact of theInternet on American politics.Westport, CT: Praeger.

Shah, D. V., Kwak, N., & Holbert, R. L.(2001). “Connecting” and “disconnect-ing” with civic life: Patterns of Internetuse and the production of social capi-tal. Political Communication, 18(2),141-162.

Shah, D. V., McLeod, J. M., & Yoon, S.H. (2001). Communication, context,and community: An exploration of print,broadcast, and Internet influences.Communication Research, 28(4),464-506.

Stowers, G. N. L. (1999). Becomingcyberactive: State and local govern-ments on the World Wide Web. Gov-ernment Information Quarterly,16(2), 111-127.

Thomas, J. C., & Streib, G. (2003). Thenew face of government: Citizen-initi-ated contacts in the era of e-govern-ment. Journal of Public Administra-tion Research and Theory, 13(1), 83-101.

Tolbert, C. J., & McNeal, R. S. (2003).Unraveling the effects of the Internet onpolitical participation? Political Re-search Quarterly, 56(2), 175-185.

Toulouse, C., & Luke, T. W. (Eds.).(1998). The politics of cyberspace.London: Routledge.

United Nations/ American Society forPublic Administration. (2002).Benchmarking e-government: A glo-bal perspective. New York: UnitedNations/DPEPA.

Uslaner, E. M. (2004). Trust, civic en-gagement, and the Internet. PoliticalCommunication, 21(2), 223-242.

Verba, S., Nie, N., & Kim, J. (1978).Participation and political equality:A seven-nation comparison. NewYork: Cambridge University Press.

Verba, S., Schlozman, K., & Brady, H.E. (1995). Voice and equality: Civicvoluntarism in American politics.Cambridge, MA: Harvard UniversityPress.

Weber, L. M., Loumakis, A., & Bergman,J. (2003). Who participates and why?An analysis of citizens on the Internetand the mass public. Social ScienceComputer Review, 21(1), 26-42.

Wilhelm, A. (2000). Democracy in thedigital age: Challenges to politicallife in cyberspace. New York:Routledge.

ENDNOTES1 See www.pewinternet.org.2 Since the dimensions are theoretically

defined and constructed, based on un-derstanding the role of different formsof participation in representative de-mocracy, the study did not use factoranalysis to generate the classification ormeasurement.

3 It should be noted that the BSA surveymonitored “party and trade union”membership, but the latter was alsomeasured separately, so in the recodedmeasure, the residual group remainsonly the party members.



APPENDIX A

Scales Question wording Voted May I just check, thinking back to the last general election — that is the one in 2001 — do

you remember which party you voted for then, or perhaps you didn't vote in that election? Yes/No

Campaign-oriented activism

And have you ever done any of the things on this card about a government action which you thought was unjust and harmful? ... • Contact my MP or MSP

Are you currently a member of, or do you regularly join in the activities of, any of the organizations on this card? • Political parties or trade unions (inc student unions)

Do you think of yourself as a little closer to one political party than to the others? (IF ‘yes’), Would you call yourself very strong (party), fairly strong, or not very strong? (‘very’+‘fairly’).


And have you ever done any of the things on this card about a government action which you thought was unjust and harmful? • Contact a government department • Contact radio, TV, or a newspaper • Sign a petition • Raise the issue in an organization I already belong to • Go on a protest or demonstration • Form a group of like-minded people


Are you currently a member of, or do you regularly join in the activities of, any of the organizations on this card? • An environmental or conservation group • A pressure group or campaigning organization • Parent-teachers/school parents association/Board of Governors, etc. • Youth groups (e.g., scouts, guides, youth clubs, etc.) • Education, arts, drama, reading, or music group / evening class • Religious group or church organization • A sports or recreation club • Tenants’/residents’ group/neighborhood watch • Social club/working men’s club • Women's group/Women’s Institute • Group for older people (e.g., lunch clubs) • Local groups which raise money for charity (e.g., The Rotary Club) • Other local community or voluntary group • Other national or international group

Pippa Norris is the McGuire lecturer in comparative politics at HarvardUniversity’s Kennedy School of Government. Her work compares politicalcommunications, elections and public opinion, and gender politics. Recent booksfor Cambridge University Press include Sacred and Secular (2004), ElectoralEngineering (2004), and Radical Right (2005).

John Curtice is professor of government and director of the Social StatisticsLaboratory at the University of Strathclyde, and a research consultant to theNational Centre for Social Research. He has written extensively about trends inpolitical participation in Britain in recent years.

If You Build a Political Web Site, Will They Come?

Documents