You are here

Examining the Role of the Internet in Korean, Australian, and Danish Elections

We're starting the last day of this very enjoyable AoIR 2008 conference already. This one is going to busy for me, as three of my papers are scheduled for today - two of them, in fact, in competing sessions (but luckily my colleagues Lars Kirchhoff and Thomas Nicolai, who are the lead authors, are able to present one of them). This morning, we're starting with a session on the online dimensions of national elections across a number of countries.

The first presenter is Yeon-Ok Lee, whose focus is on last December's presidential election in South Korea. The previous election to this, of course, was won by a small margin by the liberal underdog Rooh Moo-Hyun, due in good part to the activism of Korea's Netizens and to coverage by citizen journalism site OhmyNews. This made the 2007 election a particularly interesting case for further research.

There is a history of literature on e-campaigning, and there are questions about whether online campaigning necessarily levels the playing field and mobilises support for minor candidates particularly well, or whether by now a normalisation has taken place in Korea, so that the primacy of major candidates is reinforced. The present research is also interested in classifying campaign Websites according to their approach to informing, involving, connecting, or mobilising citizens. Further, it is also important to keep in mind contextual factors affecting e-campaigning, of course.

The project began by collecting as much Web data (or Webometrics) as possible - identifying the respective sizes of the 20 official Websites for the 12 candidates and parties, counting the frequency of entries appearig in search engine indices when searching for presidential candidates, identifying the number of incoming links to the Websites, and estimating Web traffic to the sites. Additionally, the project utilised a tool called LexiURL for co-link analysis, in order to analyse the extent to which linking sites see individual competing candidates as more or less closely related to one another. Further, some qualitative research examining Website content was also conducted.

Perhaps unsurprisingly, online attention was focussed especially on the major parties and candidates. Frequency counts and Web traffic showed the conservatives ahead (indeed, in terms of frequency in search indices, very well ahead) of the liberals, while site sizes and the number of incoming links were larger for liberal and labour sites than for the conservative GNP site (incoming links to labour sites and candidates were substantially stronger than for the GNP). This may also be related to the fact that the GNP candidate was a front-runner throughout the entire campaign, and did not feel obliged to engage in as significant an amount of e-campaigning as the other candidates and parties.

Co-link maps are indicative of the relative level of public awareness and the ideological orientation of candidates; a single incoming link indicates symbolic affiliation, for example. What is interesting in the map of connections between candidate Websites which emerge is that the eventual conservative victor is a relative outsider on the margins of the network.

Overall, then, there was no significant effect of online campaigning this time around; online campaigns reproduced offline sentiment relatively directly. There were no significant e-campaign innovations in this campaign. Contextual factors, by contrast, were highly important; the conservatives made significant gains from Korea's worsening economy and the resulting animosity towards the outgoing government, and online campaigning did not have a significant impact on public perceptions.

My own paper is a further update on our collaborative research on blog mapping, with Lars Kirchhoff, Thomas Nicolai, Jason Wilson, and Barry Saunders. The slides are below, the full paper is online, and I'll add the audio as soon as I can the audio is now online, too.

Jens Hoff is the third presenter, discussing the Danish parliamentary election which took place in late 2007 as well. What, if any, was the impact of the Internet on election outcomes - do political uses of the Net affect users politically? This study worked with a panel of some 5080 voters whose entire online activity was logged during November 2007. The results were weighted so that they were representative of the overall Danish Internet population. Additionally, a follow-up survey was answered by 980 respondents.

Political use of the Internet was varied amongst this group - mainstream media Websites and election quizzes were most popular, followed at some distance by party and candidate Websites. Candidate blogs and debate sites were substantially less popular. There were no gender differences in such use; younger cohorts were more active than older users; differences between more or less educated users were minor; and the lowest (often students) and highest income groups were most active.

From the survey, there was only a very limited influence of Net use on voting intentions, and only marginally more users agreed that the Net had changed attitudes on key issues, influenced opinions on politics and candidates, etc. This connects well with theories which suggest that the mainstream media set political agendas; there are also some differences here which depend on the personal characteristics of users, but (other than the level of interest in politics) such socio-demographic characteristics were not significant.

Based on the data, Jens suggests that the extent of Internet usage determines in part how strongly users are influenced politically by it; influence also depends on the type of Websites used; and less politically interested users are more likely to be influenced than more strongly politically invested users. Additionally, he distinguishes between 'Net activists' and more passive 'information seekers' - and here, too, Net activists are less influenced politically by Internet use than are information seekers.

Technorati : , , , , , , , , , , : , , , , , , , , , ,


Han Woo Park has kindly let me know that the slides for the South Korea paper are now online. More information about this and related research is at Han's Website.