Under the Microscope / Prof William Reville: My columns mostly deal with research findings in the natural sciences. But, as Prof Neil Collins, the amiable head of the department of government at UCC reminds me, the social sciences also use scientific methods in their researches. Prof Collins, with two colleagues, Fiona Buckley and Theresa Reidy, recently carried out a study of the influence photographs of election candidates on ballot papers have on determining how voters cast their ballot.
Since 1999, photographs of candidates are placed on ballot papers for local, national and European elections in Ireland. The main reason for this change is to help voters who have literacy difficulties. It should also alleviate the problem of several candidates with the same or similar names appearing on the ballot paper.
However, the introduction of photographs is accompanied by a number of complications. One weakness is the relatively low level of candidate recognition, particularly of MEP candidates, among voters. And a complication arises from the fact that people are willing to make important decisions based only on photographs.
It is well established in social psychology that an attractive appearance elicits a positive response from the beholder. (You will note from the photographs appearing on this page that Dick Ahlstrom and I are particularly blessed in this regard!) Research carried out in 1986 showed that voters are quite prepared to translate faces into votes on the basis of a single photograph. In the words of one researcher (SW Rosenberg) - "a single photograph is sufficient to create a distinct and reliable impression of the person". And in the words of another (HW Singleman) - "in various walks of life, including politics, handsome men are able to take full advantage of the widespread tendency to associate physical attractiveness with positively valued personal traits". Singleman noted that physical attractiveness works better for male than for female candidates.
In an ideal world every voter would make up his/her mind on the basis of a detailed consideration of each candidate's election manifesto. However, in the real word, many voters come to the ballot station poorly informed about the candidates' manifestoes and make their voting decisions based on cognitive shortcuts. These include party affiliation, gender, sporting prowess of the candidate, and so on, and of course, as already mentioned, physical appearance. Position on the ballot paper is also important and several studies have shown that voters tend to favour candidates whose names appear near the top of the ballot paper.
The survey study carried out by Prof Collins and colleagues used the June 2004 local election in Ireland. The dual mandate had recently been abolished removing well known national politicians from the election. Many candidates appeared on the ballot paper for the first time and were relatively unknown.
The survey was administered on election day at three polling stations, two urban and one rural. Turnout exceeded 60 per cent and 650 respondents completed the survey. Two replica ballot papers were used. The first contained only photographs of the candidates, with no names or party identification. The candidates were real candidates from a different electoral area to that surveyed and were unknown to the survey respondents. In the second ballot paper, party affiliations were noted beside the photographs. The photograph only ballot paper was used in the rural polling station and in one urban station. The photograph plus party affiliation ballot paper was used in the other urban polling centre. Twelve candidates, nine men and three women, appeared on the ballot paper. All respondents were actual voters who were approached outside the polling station after they had voted officially, and asked to cast their survey ballot.
The candidates appeared on the ballot papers in the following order from top to bottom: A (male), B (male), C (male), D (female), E (male), F (female), G (female), H (male), I (male), J (male), K (male), L (male). The most notable result of the survey was the similarity between the real election and the survey ballot.
Five candidates were elected in the real election in the following order I, K, A, G, D. Four out of the five candidates elected in the real election were also "elected" in the survey. The survey results for the "photograph only" ballot showed candidates elected in the following order I, G, B, D, K (D and K received equal votes). The survey results for the "photograph plus party affiliation" ballot showed candidates elected in the following order G, F, I, D, A.
The 2004 local election was a "low information" election. The level of candidate recognition by voters was low, general public interest in local elections is not great and the level of voter engagement with candidates' political positions tends to be low. In such a context it would be expected that factors such as photogenicity would play a greater part than they otherwise might. This study showing that, when given only candidate photographs, survey respondents replicated an election result with 80 per cent accuracy, nicely confirms this point.
The introduction of new measures to encourage voting is a good idea. However, the particular device of the photograph introduces an artefactual element and the recent research indicates that this device needs to be thought out more.
The survey carried out by Prof Collins and colleagues also gathered information on the influence of gender on voting patterns and also the effect of the age profile of voters. Further information is available from the department of government, University College Cork (tel: 021-4902941).
William Reville is associate professor of biochemistry and public awareness of science officer at UCC