Does Cyber‐Campaigning Win Votes? Online Communication in the 2004 Australian Election

Abstract

Studies of candidates’ and parties’ web use during elections have focused primarily on the contents of their websites and the reasons behind the adoption of the technology. Evaluations of the electoral impact of web campaigns on voters have been limited and inconclusive. This study uses data from the 2004 Australian Candidate and Election studies to investigate the factors determining candidates’ use of web campaigning and its effect on the vote. The findings show that enthusiasm for web campaigning is strangest among established party members. More significantly, we show that despite being linked with traditional campaigning techniques, web campaigning exerts an independent and significant impact on the level of electoral support that a candidate receives. We investigate how this effect may be taking place and conclude that web campaigning, at least under present electoral conditions, constitutes an important component of a winning election strategy.

Acknowledgements

The 2001 and 2004 Australian Election Studies and Australian Candidate Studies were collected by Rachel Gibson, Ian McAllister, Clive Bean and David Gow and funded by the Australian Research Council. An earlier version of the paper was presented at the Center for the Study of Democracy, University of California Irvine, July 2005. Our thanks to seminar participants and to two anonymous reviewers from this journal for their constructive and valuable comments.

Notes

1. Rebecca Fairley Raney, “Experts want to dissect McCain’s internet fundraising”, The New York Times on the Web, 18 February 2000, available online at: ⟨http://www.ntytimes.com/li…ch/00/02/cyber/articles/18campaign.html⟩ (accessed 19 February 2000). Rebecca Fairley Raney, “Volunteers’ actions lead skeptics to question McCain’s online donations”, The New York Times on the Web, 12 February 2000, available online at: ⟨http://www.ntytimes.com/li…ch/00/02/cyber/articles/12campaign.html⟩ (accessed 14 February 2000).

2. A profile that was no doubt further enhanced by the US Federal Election Commission ruling that online credit card donations were eligible for matching funds (von Sternberg, 1999).

3. Roland Watson, “Against all odds”, The Times, 26 November, Section 2, p. 7; Ryan Lizza, “Dean.com”, The New Republic Online, 23 May 2003, available online at: ⟨http://www.tnr.com/doc.mhtml?i=20030602&s=lizza060203⟩ (accessed 4 June 2003).

4. E. J. Dionne Jr., “Dean’s grass‐roots cash cow”, Washingtonpost, available online at: ⟨http://www.washingtonpost.com/wp‐dyn/articles/A23937‐2003Jul7.html?referrer=email⟩ (accessed 20 July 2003).

5. “Youth vote influence by online information”, Pew Internet and American Life Project, 2000, available online at: ⟨http://www.pewinternet.org/reports/reports.asp?Report=27&Section=ReportLevel1&Field=Level1ID&ID=94⟩ (accessed 12 January 2001). These findings are supported by a survey of American adults conducted in November 2000 for the Democracy Online Project at George Washington University that examined use of the internet for election‐related information. See “Post‐election 2000 survey on internet use for civics and politics”, December 2000, The Democracy Online Project, George Washington University, available online at: ⟨http://democracyonline.org/databank/dec2000survey.shtml⟩ (accessed 7 July 2001).

6. James Crabtree, “Whatever happened to the e‐lection”, The Work Foundation, 2001, available online at: ⟨http://www.theworkfoundation.com/pdf/Election percent20designed1.pdf⟩ (accessed December 2001).

7. All results reported here are for candidates in elections to the federal House of Representatives (lower house) only. The upper house, the Senate, has a PR STV electoral system that operates in a similar way to a party list, and votes are cast on a state or territory‐wide basis. Combining the two for the voting analysis reported in Table 5 would obviously be problematic, and for consistency purposes, therefore, Senate candidates are excluded from all of the analyses reported in the paper.

8. Earlier analyses, combining Senate and House of Representatives candidates, included incumbency as an independent variable, since we would expect incumbents to have better political resources and campaigning skills compared to challengers. However, in restricting the analyses to lower house candidates the inclusion of incumbency creates difficulties in the measurement of party membership, since none of the minor parties had incumbent lower house candidates. For that reason we have excluded incumbency. Analyses using incumbency instead of party come to the same substantive conclusions.

9. The AES is comparable with, albeit slightly higher than, the most recent publicly available official estimates from the Australian Bureau of Statistics. A report issued in December 2005 revealed that 63% of adults aged 18 years or over had accessed the internet from any site during 2004–2005. See pub. 8146.0 Household Use of Information Technology, Australia, 2004–2005, available online at: ⟨http://www.abs.gov.au/Ausstats/[email protected]/0/acc2d18cc958bc7bca2568a9001393ae?OpenDocument⟩ (accessed 10 April 2006).

10. Preliminary analyses identified these activities separately, but since they were not significant, they have been combined in the interests of parsimony.

11. The question asked was: “During the 2004 election campaign did you visit any of the following websites? (Circle as many as apply)”. The results were “Party site” 5%; “own MP’s site” 2%; “Candidate in my constituency” 2%.

12. We were unable to examine those who accessed candidate and party sites because the sample was too small for statistical analysis (n < 30).

13. Given the rich and extensive range of measures for campaign quality we were able to include, we would consider the significance of any such candidate effects to be limited.