Voters' information behaviour when using political actors' web sites during the 2011 Scottish Parliament election campaign

Introduction

In May 2011, the Scottish National Party (SNP) swept to power in the Scottish Parliament elections, winning 69 of the 129 available seats with a 45.4 per cent share of the constituency votes and a 44 per cent share of the regional votes, thereby forming the first ever Scottish administration with a working majority. This was, as The Scotsman newspaper proclaimed, “an achievement unprecedented in modern Scottish politics” (Anon, 2011). The Scottish electoral system had, after all, been designed specifically to prevent one party winning an outright majority, with its proportional, regional votes approach intended to balance out any gains made from the first-past-the-post constituency seats.

In the immediate aftermath of its victory, the SNP highlighted the crucial role played by its online campaign, including the use of a bespoke voter database, Activate, and a new platform, NationBuilder, which integrated Facebook and Twitter with the party's web site. Candidates and party activists were encouraged to use these tools to identify, contact and interact with potential voters online, and combine this with more traditional face-to-face canvassing work on the street or doorstep. As a result, the SNP claimed that the 2011 campaign was the “first European election where online has swayed the vote” (Gordon, 2011) and that their strategy would now be the “model for political parties all over the world” (Wade, 2011).

If the SNP's assertions are to be believed, then these are at odds with the views of those who claim that the internet is unlikely to ever be a truly revolutionary campaigning tool in the UK. Indeed, following the 2010 UK general election, which had been predicted by many observers to be one on which online technologies would have a significant impact (Swaine, 2010), Williamson (2010a) concluded that “it is time to put aside the idea of an ‘internet election’”. He argued that, although digital media play an increasingly important role in the UK elections, particularly as management tools with which parties organise their candidates and activists, the party-oriented nature of British politics does not lend itself to web-based campaigns that capture the public's imagination. Williamson (2010b) contrasts this with the campaign culture in the USA, where a more personality-led approach provides a powerful means of mobilising public support and raising campaign funds online.

But what did the Scottish electorate think of the online offerings of the SNP and the other parties during the 2011 Scottish Parliament campaign? Were these sites informative, engaging and likely to impact upon their electoral choice? This paper reports the results of a study of voters' online information behaviour, conducted at the height of the campaign in Aberdeen, North-east Scotland, where members of the public were given the opportunity to explore, and provide critical feedback on, the campaign sites of parties and candidates standing for election. More specifically, the study aimed to:

• identify motivations for, and barriers to, the public's use of election campaign sites;

• investigate the types of information, tools and technologies the electorate most values when accessing these sites;

• assess the likelihood of such sites being visited again in the future; and

• explore the extent to which the use of campaign sites might influence voting behaviour.

The paper begins by providing a brief overview of the literature in the field of online electioneering campaigning, focusing on those studies that have explored voters' perceptions of the internet as an electoral campaign tool. It then describes the data collection methodology adopted in the current study, before discussing some of the key themes emerging from the results, in terms of the participants' perceptions of the accessibility, value and impact of online campaign sites. The paper concludes by considering the implications of the research results for those candidates who successfully gained election to the Scottish Parliament, as well as for other political actors and academics internationally.

Literature review

Since the mid-1990s, a significant body of literature has emerged that has discussed the use of the internet as an electoral tool by political actors worldwide. As Ward and Vedel (2006) observe, the early literature heralded a general wave of enthusiasm about the potential impact of the internet on the political sphere, where “mobilisation” or “equalisation” theorists predicted that it would facilitate a new, more participatory style of politics, bringing parties and politicians and an increasingly disaffected electorate closer together, and drawing more people into the democratic process. Within a few years, however, these Utopian claims were being questioned by a second wave of more sceptical voices – “reinforcement” or “normalisation” theorists who argued that the internet, far from revolutionising political communication and participation, simply reflected and reinforced existing patterns of offline behaviour. More recently, a new wave of optimism has begun to emerge, prompted largely by developments in the USA, where, for example, Howard Dean's presidential candidacy campaign in 2004 (Hindman, 2005) and, in particular, Barack Obama's 2008 campaign (Cogburn and Espinoza-Vasquez, 2011) utilised new, more interactive web 2.0 technologies to raise campaign funds and create networks of supporters and volunteers.

Gibson and Ward (2009) provide a concise review of the key studies in the field and, in so doing, note that the literature has been dominated by “supply side” questions, where researchers have quantified the extent to which political actors have adopted online campaigning tools, or where they have conducted content analyses of campaign web sites. In reviewing these studies, they identify a number of broad trends, including that:

• parties and candidates have focused on the one-way dissemination of information to the electorate, rather than utilising the interactive features offered by new technologies;

• where political actors have adopted a more participatory approach, they have tended to offer only “controlled interactivity” via the provision of web-based feedback forms or generic email addresses, rather than providing more open, unstructured opportunities for debate, such as chat rooms and message boards; and

• parties are increasingly using the internet as a means of securing resources, in terms of members, volunteers and financial donations.

In the USA, meanwhile, a series of national telephone surveys by the Pew Research Center have charted the incremental growth in the number of Americans going online to obtain electoral information during successive presidential and mid-term contests (Rainie et al., 2005; Rainie and Horrigan, 2007; Smith, 2009). The most recent Pew study (Smith, 2011) estimated that 54 per cent of American adults used the internet to connect to the political process during the 2010 mid-term campaign; and while all generations were to some extent “online political users”, those most intensely engaged were young adults aged under 30.

There has also been a raft of studies, predominantly conducted in the USA and often based on existing survey data, which have used multiple regression techniques to explore relationships between internet use during elections and citizens' levels of political efficacy, knowledge, trust or engagement, or of their likelihood of voting. The results of these studies have been mixed: while some have found internet use to be an important predictor, or at least correlate, of political knowledge and interest (Drew and Weaver, 2006), others have found more modest associations (Kenski and Stroud, 2006).

A number of more experimental, laboratory-based investigations have also taken place, again largely in the USA and often involving convenience samples of university students. Here, participants have been exposed to candidates' web sites and their attitudes towards political issues and the candidates' characters have then been measured using Likert-type scales in a post-test questionnaire (Hansen and Benoit, 2005; Towner and Dulio, 2011). However, due to the differing approaches and objectives of these studies, no clear patterns have emerged from the results. Some experimental studies have focused on the effects of exposure to interactive features on candidate web sites. Again, though, these have produced contradictory findings. For example, Tedesco (2007) established that exposure to highly interactive features resulted in significant increases in political information efficacy; but Sundar et al. (2003) found that “too much interactivity” interfered with users' recall of policy issues.

Most studies of users of online campaign resources have been largely quantitative in nature; very few researchers have adopted a more qualitative approach. In the USA, Stromer-Galley and Foot (2002) and Wells and Dudash (2007) conducted focus groups with citizens and students, respectively, to explore their perceptions of the engagement opportunities presented by campaign web sites. Meanwhile, away from the campaign trail, Ferguson and Howell (2004) discussed the deliberations of a “blog jury” who monitored the UK political blogs; and Lynch and Hogan (2011) used focus groups to investigate social media use by Irish political parties, from the perspective of young citizens.

This lack of qualitative studies is bemoaned by those observers who point out that “we still know relatively little about the way citizens engage with technology and online information” (Gibson and Ward, 2009). Indeed, Gibson and Römmele (2005) argue that obtaining “a better in-depth understanding of individuals' online election experiences” would assist in better shaping the questions asked in quantitative opinion surveys, allowing more precise causal inferences to be drawn about voters' exposure to campaign sites. The largely qualitative research discussed in this paper is, therefore, particularly timely and potentially significant.

Methodological approach

The research was hosted by the Rosemount Community Centre, an Aberdeen City Council community learning facility located in central Aberdeen. At the time of the study, the centre offered a varied programme of classes, ranging from flower arranging to computing for beginners. It also hosted regular advice surgeries with two of the local government councillors representing the local electoral ward. It was anticipated, therefore, that in conducting the study in the centre it would satisfy the researchers' long-standing philosophy of getting as close as possible to the everyday lives of their research subjects, and in the process would reach a wide cross-section of the local population. The research also took place at a time when there appeared to be a lack of trust in local democracy in Aberdeen, caused largely by the council's response to a public consultation on controversial plans to redevelop a Victorian park located in the heart of the city centre (Ross, 2010). With this in mind, it was hoped that the nature of the research would appeal to potential participants during a period of heightened political awareness locally, particularly as a number of the city's councillors were standing as candidates for election to the Scottish Parliament.

The study took place during the two-week period beginning 18 April 2011, shortly before the election date of 5 May, and the researchers were located in the centre's coffee room, where users congregate before or after classes. The research was promoted by means of posters displayed throughout the centre, as well as articles in the local press. In addition, research team members visited and “canvassed” the numerous small shops and businesses in the vicinity of the centre; and one of the researchers occasionally positioned themselves directly outside the centre's entrance, inviting passing pedestrians to participate. Combined, these approaches were successful in persuading 64 individuals to take part.

The research used the interactive, electronically-assisted interview method – previously developed by the researchers during a study of the British public's use of parliamentary web sites (Marcella et al., 2003) – where the participants were observed and questioned as they searched for and used information on the web sites and social media sites of parties and candidates standing for election to the Scottish Parliament.

Online access was achieved with the use of a laptop computer and mobile broadband “dongle” technology. While interviews were occasionally hampered by weak signals and slow data transmission speeds (particularly when respondents wished to view video clips), the access speeds obtained during the study generally proved acceptable to the interviewees. The interviews varied in length from 13 min to almost 50 min, depending on the information seeking techniques, time availability and personal interest of the research subjects. With the participants' consent, all interviews were audio-recorded digitally and subsequently transcribed. In addition, a software package IE HistoryView was used to log invisibly all URLs visited during the interviews.

The interview schedule comprised four distinct parts:

1. demographic questions exploring age, occupation, education, and electoral constituency;

2. structured questions on voting patterns, political participation, computer use, and past need for campaign information;

3. a free-form period of undirected information seeking on the campaign site(s) of the participant's choice; and

4. post-search, structured questions on the user friendliness of the sites visited, the relevance and comprehensibility of the information found, the likelihood of such sites being revisited, and the extent to which the information obtained may affect electoral choices.

Demographic profile of the study sample

Table I illustrates the key demographic features of the participants. As can be seen, there was a reasonable spread of ages in the sample, although more females than males participated in the research. When compared with recent official statistics (Aberdeen City Council, 2003; Scottish Neighbourhood Statistics, 2010), the sample was, proportionately, better educated and from higher socio-economic classes than the Aberdeen population as a whole. The sample was also slightly over-representative in terms of retired individuals.

Study results

Post-search questions

The periods of information seeking were followed by a series of structured, evaluative questions. 49 (76.6 per cent) of the respondents believed that the types of sites they had just visited were a useful way of providing campaign information, although 15 of these felt that they would be of greater benefit to younger, more computer-literate voters. The internet was generally regarded as a cost-effective way of disseminating up-to-date campaign information to a wide audience, although many respondents felt that the parties and candidates could be doing more to advertise the existence of such sites:

People don't know that these sites exist. When they put leaflets through the door they should mention these sites.

Indeed, some envisaged a future where the online site would replace the campaign leaflet drop. Although others hoped that the internet would not become the main focus of elections, and should instead complement existing campaign methods:

What we wouldn't want is them withdrawing to a darkened room somewhere just putting messages out on the internet. You still need to see your candidates.

In terms of ease of use, the vast majority of participants described the campaign sites as “very” (64.1 per cent) or “quite” easy (32.8 per cent) to use. The content of the sites was also regarded as comprehensible, with most describing it as “very” (70.3 per cent) or “quite” easy (28.1 per cent) to understand. Opinions were more divided, however, on how interesting they had found the content: while 16 (25 per cent) felt it “very interesting”, over half (33; 51.6 per cent) thought it only “quite interesting” and 15 (23.4 per cent) described it as “not interesting at all”.

Despite not being overly enthused about the content, around two-thirds (44; 68.8 per cent) indicated that they might look at campaign sites again, either further exploring the site(s) they had just seen, or extending their search to other sites “to see what the other parties have got on there”. Of the 20 participants who felt it unlikely that they would ever visit such sites again, most indicated that they would continue to rely solely on more traditional sources of campaign information:

I'll probably just stick to the TV and newspapers.

Finally, participants were asked if the information they had just seen had helped them in deciding which way they were likely to vote. There was little evidence here of any significant impact on voting behaviour. One respondent indicated that seeing the Greens' web site had persuaded him to give them his second, regional list vote; while a 40-year-old female interviewee suggested that exploring the campaign sites had kindled an interest in voting for the very first time. Meanwhile, two participants who had examined the far-right parties' sites both indicated that this had merely reinforced their opinion that they would never consider voting for these parties.

For the vast majority (60 of the 64 participants), the interactive sessions had had no influence on their democratic choice. Some noted that their decision had already been made, based on information obtained via traditional media. While some indicated that they had a long-established allegiance to a specific party, which was unlikely to be affected by receiving campaign information, either online or offline:

No, I've got my own political views, and I don't think a web site's going to change that.

Some others felt that they had neither the time nor the patience to visit a range of campaign sites. Indeed, a small number expressed a desire to see some form of central, politically neutral site, where voters could readily compare and contrast the profiles and policy aims of all local parties and candidates:

You could do it all within the same web site and have the different manifestos. I guess they could all be condensed and put into bullet-point form on Facebook somewhere.

Conclusions and further research

This study, using the interactive, electronically-assisted interview method, was successful in eliciting a rich stream of data from voters about the use and potential impact of online election campaign sites. Above all else, it revealed the dichotomy that appears to exist between the online electoral information needs of the public in North-east Scotland, and the nature and the extent of the information actually provided by parties and candidates.

Perhaps, the most dominant theme to emerge was that of a need for brevity and clarity in the presentation of policy information. Very few participants were prepared to spend time perusing lengthy, verbose party manifestos. Instead, many expressed a need for short, sharp, “bite-size” policy statements that can be read and digested with relative ease. While some Scottish parties are already taking steps to present their policy goals in a more concise, user-friendly fashion, there is clearly much to be done in this regard.

The study also highlighted a need for currency in the information provided online. While the party sites were generally regarded favourably for their provision of up-to-date election news, voters were bemused by those candidate sites where the content remained static throughout the crucial campaign period. Indeed, this lack of online activity understandably led participants to question the rationale for these candidates establishing a campaign site. As one asked succinctly, “why have it?”

A clear need was also demonstrated for policy statements and commentary relating specifically to local constituency issues. In this respect, the parties' sites were perceived as being wanting: although some sub-national policy information was present, it tended to become “lost” amongst other site content. As national party sites appear to face challenges in being “all things to all people”, the current authors would argue that it is even more incumbent upon the individual candidates and their campaign teams to provide online information on specific local questions, or on how national agendas might affect their potential constituents. Yet, several of the candidate sites explored in this study were lacking any significant policy commentary: the contestants in North-east Scotland often appeared more interested in discussing the weather rather than the issues being raised by voters. Indeed, in a complementary study of the content of all Scottish candidates' social media sites during the 2011 campaign (Baxter and Marcella, 2013), the current authors established that this pattern was replicated across the country. In terms of local policy information provision, we would also suggest that a more prominent role be afforded to the online sites of local branches of political parties. The larger parties in Scotland have a number of regional associations, many of which have their own sites containing political information and news with a more local slant. However, their existence is barely acknowledged on national party web sites, with only the Conservatives and the Greens currently providing relatively prominent links to their regional sites.

A desire to see more online engagement between politicians and voters was also evident in the study. Here, however, the participants' comments were somewhat contradictory, in that they expressed disappointment with the current levels of debate on the social media sites of parties and candidates, yet were themselves reluctant to enter into any form of public dialogue with political actors or other voters. The findings of the authors' complementary study (Baxter and Marcella, 2013) would suggest that this reticence, on the part of both politicians and voters, was a national rather than just a North-east Scotland phenomenon.

Overall, then, while online campaign sites were generally regarded as serving a useful purpose, as being easy to use and understand, relatively interesting and likely to be visited again, there was very little evidence in this study to indicate that they had any significant impact on voting behaviour during the 2011 Scottish campaign. Rather, it would suggest that more traditional information sources, particularly print and broadcast media, coupled with long-established campaign techniques, such as leaflet deliveries and door knocking, continue to be more influential in determining voters' democratic choices. The current authors would therefore challenge the SNP's assertion that its online campaign had “swayed the vote”. While it is acknowledged that elements of the SNP's internal strategy, particularly its voter database, will have played a vital role in informing and organising the party's activists, we would question the effects of the public face of the SNP's online campaign, namely the web sites and social media sites of the party and its candidates. In line with other commentators (Barnes, 2011; Taylor, 2011) we would argue that a range of other factors were far more influential in the party's electoral success, including: the perceived charisma of its leader, Alex Salmond; the inept campaign of its main rival, the Labour Party; and the collapse of the Liberal Democrat vote, due to its coalition partnership with the Conservative Party at the UK Government level. With these points in mind, we believe that the first true “internet election” in the UK has yet to materialise.

The implications of the results of this study might be considered in two different ways; first, for those candidates who were successfully elected to the Scottish Parliament in 2011. When the parliament was being established in the late 1990s, an Expert Panel on Information and Communications Technologies (1998) recommended that it should focus upon the contribution that new technologies might make in enabling greater openness and transparency and in assisting the democratic process. The panel believed that the parliament should “aspire to be an example of best practice in parliamentary information systems”. Since then, studies by Smith and Webster (2004, 2008) have established that members of the Scottish Parliament (MSPs) are “intensive and competent users” of new information and communications technologies (ICTs), and that ICTs “have become a cultural norm of contemporary parliamentary life”. Smith and Webster's studies took place before the emergence of Facebook and Twitter; but the potential value of these social media as information provision and communication tools has since become recognised within the Scottish Parliament, where it is acknowledged that they “can increase the accessibility of MSPs and offer new ways in which to engage constituents, stakeholders and the wider public” (Scottish Parliament Standards, Procedures and Public Appointments Committee, 2012). However, as this study revealed, those individuals seeking election to the Scottish Parliament frequently failed to provide up-to-date, local policy information on their campaign sites, or enter into any meaningful online engagement with potential constituents. We would suggest, therefore, that many of the successful candidates will have to significantly alter their information practices on entering the Parliament if they are to become fully integrated into its community of modern, online parliamentarians.

The present authors' research on online electioneering, conducted over the last decade, has concentrated on the state of affairs in Scotland, and will continue to do so. Indeed, the forthcoming referendum on Scottish independence, to be held in September 2014, will offer a unique opportunity to explore the extent and impact of online campaigning in an entirely different, and less party-focused, polling situation. For example, it is anticipated that the electronically-assisted interview method will be used to investigate Scottish citizens' perceptions and use of any online sites created by the respective “yes” and “no” campaign groups.

Despite this regional focus on the Scottish scene, the current authors feel that the implications of their work might also be considered in the context of the wider, international political arena. The study discussed in this paper has added to what is a rather sparse body of qualitative work on voters' online election information needs, and the ways in which citizens engage with campaign web sites. And while we would lay no claims to the results of the study being applicable outwith the Scottish setting, we firmly believe that the voter-centred methodological approach described here might be used or adapted by colleagues in the political and information research communities worldwide, in order to further shape our understanding of the use and the true democratic impact of online campaign tools.