Race to the top: using experiments to understand gender bias towards female politicians

ABSTRACT

Gender balance in political representation is an important goal of governments. In this paper, we ask: Do voters judge female politicians less favourably than male politicians, when given an otherwise identical set of information about their backgrounds? We employ an innovative online experiment (N = 1933) to measure Australians’ attitudes towards female politicians and examine a series of hypotheses. We find voters see female candidates as more capable and are more likely to vote for them, but they are less likely to expect them to win. Female candidates are seen as more capable in their military and healthcare roles, but gender is perceived to be a major barrier to a female candidate’s success. Women and those aligning with the Labor/Green parties are more supportive of a female candidate, but we find limited evidence that those aligning with the Liberal/Nationals are openly hostile to a female candidate.

政治代表中的性别平衡是政府的一个重要目标。本文设问：假定其他背景信息相同，选举人会视女政治家不如男政治家么？我们采用了一种创新式在线试验（数目=1933），来测量澳大利亚人对于女政治家的态度，并检测一系列的假设。我们发现，选举人认为女候选人更有能力，更愿意把票投给他们，但又不大会期望她们赢。他们认为女性候选人更长于斗争及健康护理角色，性别被看作女候选人成功的一个主要障碍。妇女以及站队工党/绿党的人更支持女性候选人，工党/绿党的支持者公开仇视女性候选人的情况有，但不多。

KEYWORDS:

• Gender
• voter behaviour
• gender bias
• experimental research methods
• Australian politics

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Andrea Carson is a political scientist and Associate Professor in the Department of Politics, Media and Philosophy at LaTrobe University.

Leah Ruppanner is an Associate Professor of Sociology at the University of Melbourne. She is currently the co-director of The Policy Lab and her work focuses on the intersection of family, work, gender and policy.

Jenny M. Lewis is Professor of Public Policy in the School of Social and Political Sciences, and was the founding Director of The Policy Lab at The University of Melbourne, Australia. She is Associate Dean Research for the Faculty of Arts. Jenny is the current President of the International Research Society for Public Management and a past president of the Australian Political Studies Association. She is the author of 6 books, and more than 70 journal articles and book chapters.

ORCID

Andrea Carson http://orcid.org/0000-0001-5503-3334

Leah Ruppanner http://orcid.org/0000-0002-6111-1914

Jenny M. Lewis http://orcid.org/0000-0001-5858-8634

Notes

1 No women sat in the lower house of Australian parliament in 1977, and as a result little research has occurred on how voters evaluate female politicians in Australia (Renfrow 2003). Some have studied if female politicians face obstacles that men do not (Studlar and McAllister 1991; Denemark, Ward, and Bean 2012).

2 Respondents are anonymous but had to be over 18 years old, have access to the internet, and able to read English in order to participate. Participants received monetary compensation if they fully completed the survey. These payment rates are set by Qualtrics. We restricted our sample to those who reported their gender to provide direct comparisons which reduced our sample by 180 respondents.

3 It is possible that respondents still had inherent bias about our hypothetical candidates’ party affiliation which may impose some bias into the estimates. Specifically, respondents may think the female candidate is more likely to come from the Labor/Green party than the Liberal/National party. This would mean that those from the Liberal/Nationals would be less supportive of the female candidate because they anticipate she is from the other party than because she is a woman. However, our results do not support the claim as we find the Liberal/Nationals are equally supportive of Jennifer as a candidate as James (see Table 3). Thus, our results are illustrative for voters who do not obviously rely on any other cues besides gender.

4 This was done to ensure we captured Australian voters and to be compliant with our University-approved human research ethics committee approval.

5 We ran t-tests across these measures to determine whether our respondents were distributed differently by education, partisanship, gender and rural/urban location into each of our treatments. We found no significant difference in the likelihood of being assigned to Jennifer versus James across these demographics meaning our randomisation was successful.

Additional information

Funding

This work was supported by Australian Research Council: [grant number DP180101711 and DP180101217]; Melbourne School of Government: [grant number Faculty of Arts Collaboration Grant].