Skip to main content


Across the three projects, the research team will integrate different methods and modes of analysis including qualitative ethnographic participant observation, surveys of electors and candidates, automated text analysis and survey experiments. While mixing qualitative and quantitative data may be standard in the social sciences, the innovative elements are the use of comparative ethnographic research in election campaigns and applying computational models to understand sexism across structured and unstructured data. The ethnographic data will chart everyday experiences of candidates not easily captured in a survey including for example the complexities of individuals’ experiences, the unspoken ways candidates navigate their experiences with the public, their party, and the media and the often deeply felt emotion and feelings entwined with experiencing sexism. This rich and detailed ethnographic data on everyday “real life” campaigning allows the researcher to get at the way sexism is experienced and played out in social interaction in both subtle and non-subtle ways that is not captured in the candidate surveys. Likewise, the quantitative voter and candidate survey data, from the ESS and election surveys, will inform the themes that the research team explores in our in-depth interviews with candidates and in informal conversations during fieldwork. 

Reconceptualising Voter Bias

The aim of this project is to understand the dimensions of voter sexism: There is evidence that voter sexism still acts as a barrier to women’s representation. How does voter sexism vary across individuals and countries? What dimensions of sexism are linked to women’s political representation?

A main tool documenting the existence of voter sexism will be a module fielded in Round 11 of the ESS (fieldwork 2023 in over 30 countries) -- “Gender in Contemporary Europe: Rethinking Equality and the Backlash” (Banducci et al., 2020). The bulk of cross-national surveys have placed their focus on acceptance of equal participation of women in society, and in the workforce in particular (such as the items included in the ESS, International Social Survey Programme (ISSP) and World Values Survey (WVS). While the existing surveys capture an important aspect of cultural norms about the position of women, they are the tip of the iceberg, insofar as we are missing the attitudes and psychological predispositions underpinning these norms, namely prejudices.

There are at least two additional sources of data for Project 1: 1) Online panel surveys conducted at the time of national elections in the UK, the Netherlands, Spain and Turkey; 2) existing cross-national survey data (e.g. Comparative Study of Electoral Systems and World Values Survey). From these data, we can construct indicators of sexist attitudes across all participating countries, test through individual and aggregate analysis how these indicators are related to electoral outcomes (e.g. success of women candidates) and media coverage (see below). Our interest will be in accounting for cross-national variation in sexism and its influence on the representation of women.

On the Campaign Trail

The aim of this project is to understand how and under which conditions women candidates experience sexism and violence on the campaign trail. Because the barriers to women’s representation have structural and relational dimensions, examining how candidates experience them in the everyday experiences while campaigning will give a more nuanced understanding of its dynamics. Furthermore, voter sexism and media bias, whether real or perceived, places strategic imperatives on women candidates. This gender penalty may play out differently, in interactions with journalists, talking to voters or Tweeting, during the campaign period. 

To examine these imperatives, the research team will conduct residential ethnographic fieldwork including interviews and participant observation during the campaign period in each of the four selected countries. Central to the fieldwork will be: attendance at local campaign events (e.g. candidate meetings with voter groups, speeches and other campaign events) and shadowing of candidates that will allow for: a) participant observation of interactions between candidates, voters, journalists and campaign workers across multiple setting including public places such as the doorstep, parks, coffee shops to deepen understanding of everyday processes of the campaign; b) informal individual and group conversations with a wide range of people including voters and those involved in the campaign; c) informal conversations with the candidate; d) observing campaign media coverage in real time and e) observing interactions with journalists and news media representatives. During the campaigns, we will also be monitoring the social media and news media coverage of the candidates and their campaigns. These ethnographies will provide a contextualised account of everyday sexism on the campaign trail across electoral system and cultures.  

The Gender Penalty and Working Twice as Hard – Campaign Effort, Media and Voter Bias

Our aim in the third project is to understand how the media shape the “gender penalty” and the “twice as good” effect. In this project brings together the cultures of electoral sexism research from Project 1 and how candidates negotiate the gender penalty to examine the “twice as good” hypothesis. We link data from elections surveys, candidate surveys and quantitative content analysis of the media coverage and social media campaigns in the same four countries of our ethnographic work. Drawing on insights from the ethnographic work in Project 2 and applying them to original and existing surveys of citizens and candidates and large-scale media analysis, I assess two linked mechanisms: First, what is the extent to which candidates anticipate media and voter sexism and how does this influence their campaign effort. Second, I assess how candidate quality and campaign effort influence media coverage, whether media coverage is biased and how this bias influences voter evaluations of candidates. As in project 2, we will map how these relationships vary across contexts as well as across parties and candidates (e.g. candidates of color and younger/older candidates). Our analytical strategy will be to model candidate electoral success as a function of media coverage and voter preferences. Media coverage is considered to be endogenous to the amount of effort made in the campaign whereas voter preferences are influence by media coverage and conditioned by sexist attitudes.