London School of Economics: Inside the mind of the voter

The Need

Funded by the European Research Council, ‘Inside the Mind of a Voter’ is a project jointly carried out by Opinium Research and a team of LSE academics on behalf of the ECREP initiative led by Dr Michael Bruter. The research focuses on electoral psychology (the psychological mechanisms at stake when people vote, and notably when they are standing in the polling booth) and electoral ergonomy (impact of electoral procedures on how citizens vote).

The research, currently in its third year, already covers 12 countries and is expanding.

Innovative: The project uses an unprecedented combination of cutting-edge quantitative and qualitative methods. They include multi-year multi-mode surveys, innovative experiments (such as filming the shadow of voters through the polling booth curtain), election diaries, and direct observation in polling stations in collaboration with official observers.

Demonstrable outcomes: ECREP is Europe’s only research initiative dedicated to electoral psychology with results presented widely by our teams, (recently at the Houses of Parliament, European Parliament, EU-Canada summit on Youth Participation). The research is used to inform and advise Governments and Electoral Commissions. It has resulted in countries reforming their electoral procedures to better engage citizens, improve democratic satisfaction, and optimise the organisation of the vote.

Approach & Outcome

Throughout the world, nations face a growing tendency of voters to desert elections, express defiance towards their political systems, and feel alienated by those who govern them. Whilst the most severe financial, economic, and social crisis since WWII is further increasing the gap between citizens and their democracies, it is more crucial than ever that we understand the bases of electoral participation, what elections mean to citizens, and how to optimise the legitimacy of and citizens’ satisfaction with democratic processes.

The ‘Inside the Mind of a Voter’ project directed by Dr Michael Bruter, Dr Sarah Harrison, and the ECREP initiative at LSE and carried out jointly with Opinium Research breaks new analytical and methodological borders in our understanding of the relationship between citizens and their political systems. It also pushes boundaries in opinion research and offers unprecedented ethical care in studying one of the most intimate aspects’ of human behaviour, their electoral choice.

To better understand the psychological mechanisms at stake during one of the most intimate moments of a citizen’s life – the moment they cast their vote,  the project relies on the super-imposition of quantitative and qualitative methods including:

  • multi-wave, multi-year, multi-mode surveys (over 15,000 respondents so far in twelve countries)
  • in-depth narrative interviews,
  • spot-interviews (“in just one sentence, tell us what went through your mind whilst you were in the polling booth”),
  • lab experiments (e-voting, social media campaigning, and an unprecedented visual experiment that filmed the shadow of voters through the curtain of a polling booth),
  • ‘natural’ experiments (on postal voting, advanced voting, etc),
  • election diaries where voters recorded their daily thoughts throughout the election period,
  • direct observation in polling stations by official election observers

Alongside the general overview of voters, we have designed bespoke modules dedicated to specific categories such as pre-voters (aged 16-18) and first-time voters in 7 countries.

Our teams have co-operated on the project design, fieldwork logistics, and follow a stringent ethics protocol aimed at ensuring full anonymity and the secure handling of sensitive data, to ensure absolute methodological transparency.

Key challenges

Considering the subject matter, comparative breadth, and methodological complexity of the project, we were inevitably confronted by some challenging issues, which we aimed to foresee and quickly resolve.

Questionnaire parallelism and translation are critical to ensure comparability across countries. We collaborated with our sample partner network and local researchers, to scrutinize survey questions in their contextual relevance and translation.

With regards to the online survey component, it has been essential to minimise dropout rates across waves. Two elements made this challenging. Firstly, beyond the minimal two waves everywhere, several countries (USA, UK, France, etc) have a multi-year component. Secondly, our questionnaires are complex and relatively long (>10 minutes for the first wave, approximately 5 minutes for subsequent waves). We minimize dropout through explanations to participants, care in the study design and presentation, modulation of the incentives (over-rewarding later waves compared to wave 1), reminders strategy, and quality of the panel and processes. As a result, our re-response rate at wave 2 varies from 74% in the USA to 97% on the Opinium UK panel, and at wave 3 (1 year after wave 1) it averages 76% of the original participants.

Sampling specific groups such as pre-voters and first-time voters is difficult because of limited information on the target population. This makes it harder to conduct sample analysis, measure sampling bias, and create weightings. In response, we combined the data we had on these groups with data on their parents, thus enabling us to optimise the sample based on the representativeness of the parents and accurately assess sampling error.

As some of our surveys were used to conduct natural field experiments (i.e comparing in-station voters and postal voters in the UK, the vote of citizens in US counties offering postal voting vs advanced voting), we had multiple layers of sample testing, enabling us to recreate representative sub-samples to complement the representative global samples.

Finally, the multidimensional nature of the research presented a significant yet extremely exciting challenge in the question of how could we combine the various quantitative and qualitative methodologies. Our double response to this challenge represents one of the key innovations of this project. Firstly, we create ‘bridges’ between the various research components (e.g. the surveys include some open ended questions that correspond to the themes of the qualitative interviews, the election diaries include questionnaires items from the representative surveys, which allows us to estimate bias in the sample of diary keepers, same thing for the experiments…). Secondly, we use innovative coding to triangulate results using data squaring and cubing across the research components.

Key findings:

The project is ongoing but has already led to a wealth of findings published academically, attracting intense media interest, and shared with practitioners. Here are some key findings from the electoral psychology and ergonomy sides of our research:


  • Between 20 and 30% of voters change their minds within a week of any major election. In some low salience referenda this proportion can double;
  • 25% of Americans have already cried because of an election and 63% say that election night makes them feel emotional;
  • 29% of Americans and 40% of French have already changed their mind on the day of the vote in at least one major election, over half of them did so at the polling station;
  • 56% of Brits and 58% of Swedes remember going to a polling station with their parents as a child; respectively 73% and 80% remember the first election in which they voted;
  • 63% of voters say that they feel happy when they are in the polling booth, 60% feel excited, 74% proud, and 79% say that it is an important moment for them;


  • 18-25 year old British voters who voted by post in 2010 were twice more likely to vote for the extreme right than those who went to a polling station even controlling for their voting intention 3 weeks before;
  • When using a typical electronic voting machine, voters spend an average 20 seconds thinking before casting their vote against 30 seconds when using a typical UK paper ballot, this increases to 30 seconds, and a full minute with a typical French paper ballot;
  • 16-18 year olds invited to vote in their first election in a polling station have a turnout of 36.9%. When invited to do the same using e-voting, turnout plummets to 17.4%;
  • While a majority of young citizens want parties to use social media campaigning, when they do, young voters feel even more negative about politicians than when they use traditional campaigns;
  • Voters who vote in a polling station as opposed to by post or electronically end up feeling more efficacious, more positive about democracy, and are more likely to vote according to what they think is best for their country rather than for themselves.

Dissemination and social and political impact

Even though our project is still ongoing, it has already had a significant impact.

Firstly, we have multiplied prominent public presentations in recent months:

  • May 2012 – Opinium and the LSE presented a research report at the Houses of Parliament. Some of our key findings were consequently discussed by Peers in the House of Lords debate on the EU bill that month;
  • May 2012 – Dr Bruter and Dr Harrison presented a report on youth participation at a meeting in the European Parliament;
  • October 2011 – Dr Bruter highlighted data from the research in a keynote speech at the EU-Canada Meeting on Youth Participation;
  • July 2012 – Dr Bruter detailed the implications on the likely effect of lowering the voting age to 16;
  • November 2011, February, May, July 2013 – Dr Bruter and Dr Harrison presented findings to 4 different conferences of worldwide Electoral Commissioners.

Secondly, our project has also had a significant impact on policy making. Dr Bruter and Dr Harrison have conducted work on behalf of the European Commission and the European Audiovisual and Culture Executive Agency (EACEA). They have also advised several Governments and Electoral Commissions on electoral system change, understanding the psychology and motivations of voters, and optimising the introduction of new voting technology and procedures for specific population groups.

Thanks to our project findings, we advise interested stakeholders on issues such as voters’ education, e-voting, strengthening participation, first-time-voters care, reform of electoral procedures and understanding issues of trust and democratic legitimacy.

Altogether, our project is innovative and unique in numerous ways: we combine unprecedented cutting-edge methodological approaches, we foster an exciting collaboration between a pioneering academic team and an innovative opinion company, we place research ethics at the heart of our fieldwork, whilst designing sophisticated solutions to problems of comparative comparability, sampling, data integration, validity, and inference. Whilst the project is ongoing, we have already exposed some of the exciting findings that have had an almost immediate and direct impact on a broad range of academic, social, and political users.