To understand voting behaviour, we must consider voters' emotions and their interaction with electoral arrangements and the complex functions elections serve in democracies. We can then optimize voting via electoral ergonomics — the design of electoral arrangements that consider voters' bodies and minds.
Within the space of five months in 2016, two consecutive electoral earthquakes shook the Western world: the victory of the ‘leave’ (or ‘Brexit’) camp in the 23 June referendum on Britain's membership of the European Union, and that of Donald Trump in the US presidential election of 8 November. Both outcomes seemed to contradict the predictions of opinion polls as well as electoral stock markets, and meant camps that unprecedented national and international stakeholders had warned against ended up winning. The emotional tension of both votes led to questions about what had changed in the political psychology of voters, and indeed of nations, to lead to outcomes that many had thought — or wished — to be impossible.
The emotional act of voting
The mass media often suggests that people do not care about elections, that many abstain because they are not interested, but we repeatedly find that elections are actually a time of heightened emotion for many people. In many elections, 20 to 30% of voters either make up or change their minds within a week of the vote, about half of them on election day itself1, a figure which commercial surveys confirmed yet again during the 23 June referendum in the UK. This finding obliges us to consider more carefully what comes to disrupt known patterns of electoral decision-making, which invariably implies understanding more about the emotional act of voting and how voters' emotions interact with institutional design (for example, whether citizens have to vote in polling stations or can do so from home, how ballot papers are designed, or how young people's first election is organized)2.
In the aftermath of Britain's EU membership referendum, we conducted the third wave of a large panel study of British citizens. One of the questions that we asked pertained to the emotions that respondents felt as they discovered the referendum result. 32% said that they were ready to cry when they did, a figure that was even higher (46%) amongst some categories such as young voters aged 18 to 24 years.
This emotionality has political, societal, and of course academic consequences that we cannot ignore. Often, it is expressed quite positively. When we ask young people how they feel when they vote, their most frequent answers are that it makes them feel part of their community, gives them a heightened sense of responsibility, but also makes them feel proud, excited, and happy3. Indeed, despite popular belief, young people showed that they cared enough about EU citizenship and how it affects their identity to be more likely to vote in the referendum than in recent general elections.
However, while many will feel those empowering emotions as a simple result of standing in the polling booth, elections can also be associated with more negative and potentially consequential emotions, prone to continue to affect voters well after a vote. Another finding of our referendum study was that 51% of voters mentioned that they felt anger towards people who had voted differently from them, and 46% even experienced some level of disgust. Even before the vote, however, we found that the emotionality of voters had projective consequences, with people opposed to Brexit (and young people in particular) imagining their own future social, economic, and political behaviour to be severely affected by the perspective of a Brexit, for instance in terms of their likelihood to buy a home (15% more saw themselves as investing in a home in the context of a ‘remain’ victory than a ‘leave’ one), move abroad (+14%), or start or extend their family.
We need to better understand how and when this range of emotions occur, and what makes a given voter more or less likely to experience them — be it because of their personal characteristics, the way they interact with the electoral process, or whether they voted or not (and if so for the winning or losing party). Understanding how those positive and negative reactions are triggered is crucial because they could have serious consequences for democracy. On the one hand, feeling closer to one's community and a sense of responsibility for its future may lead to greater legitimation of the system, acceptance of its outputs, and civically respectful behaviour. On the other hand, emotional dislike towards another electoral camp, whilst seemingly very rare in the past, seems to have become the name of the game in an increasing number of cases — not least after the 2016 UK referendum and US presidential election, whereby our findings suggest that many citizens have moved on from resenting politicians and the political system to actually targeting their anger at fellow voters themselves. This has been expressed with mass demonstrations across British and US cities, notably by large numbers of young people, some of whom believe that the two electoral results stemmed from irreparable generational cleavages, cries that electoral results should just not be accepted, and expressions of anger and frustration towards opposite voters. This may be a key mechanism towards the creation of long-lasting cleavages and durable polarization. Indeed, if the same people repeatedly experience those negative emotions over time, there is a serious risk that they may quickly come to feel irreparably alienated by their political system.
The existing literature suggests that the physical interaction between voters and electoral arrangements may have important implications for electoral behaviour, and in our recent research, we have coined the concept of ‘electoral ergonomics’ to refer to the way in which electoral arrangements interact with citizens' psychology and optimize their experience given the possible functions of elections2.
Beyond the impact of electoral systems on voters' behaviour, growing attention has been devoted to the impact of polling station location on voting. For example, one study found that voting in a church made people more likely to vote in favour of conservative candidates and against same-sex marriage compared with voting in schools4. Similarly, after looking into the psychology of extreme right voters in the context of the 2010 UK general election, we found that young voters aged 18–24 could be nearly twice as likely to vote for extreme right parties when using postal voting compared with people voting at polling stations, but that the difference was only 24% for people aged 25–44, and that the tendency is reversed for voters aged 45 and above5 (Fig. 1). This is probably due to the higher sense of duty and responsibility that voters in general and young voters in particular feel when they vote in polling stations. Similarly, in the context of the 2012 US presidential election, allowing advance voting (the right to vote in specific polling stations in advance of election day) led to significantly more sociotropic behaviour than allowing postal or internet voting (people voting in advance from home)6. Finally, among young people, the use of internet voting leads to lower satisfaction and lower willingness to vote again than actually going to a polling station to participate in an election3, underscoring the importance of elections as a ritual7 (Fig. 2).
Breaking methodological frontiers
One difficulty faced by those who want to research the effects of emotions and ergonomics on electoral behaviour, is that the traditional methods used in the social sciences do not fully equip us to study phenomena that are dynamic, complex, and largely subconscious. To break new frontiers, we therefore need to innovate in terms of the methods that we use. A critical stumbling block is the over-reliance on self-declaration when it comes to understanding the motivations behind one's vote, despite the fact that decades of psychological research have revealed the vast preponderance of subconscious evaluations in what we think of as ‘reason’. This knowledge legitimizes the efforts of scholars to use implicit measures to understand human attitudes (for example, ref. 8). It is also why, alongside traditional qualitative and survey evidence, increasing efforts have been made to use innovative experiments in behavioural research. This trend, however, is often criticized for relying on artificial settings and overly controlled models.
Still, it is notable that some scholars have now started to use experiments based on biological analysis, such as French et al.9 who studied cortisol levels among people who have just voted in order to assess the impact of voting on stress levels. Similarly, our team has piloted new visual experiments2 where we partnered with professional film-makers to capture the shadow of voters and analyse the emotions that they display, using kinesics, which aims to decipher the 70% of human communication which is non-verbal.
Emotions and ergonomics are both extremely complex objects to study and we need to continue to explore new methodological avenues to capture and model them. At the same time, we also need to combine them with some of our more traditional and narrative instruments, because very often, emotions which are descriptively opposed (stress and excitement, sense of duty and sense of responsibility) can largely overlap physiologically. As a research agenda, methods are thus at the centre of our task. This includes looking into what other disciplines — not only neighbouring ones but more remote and unusual partners from neurology to architecture and from anthropology to the creative arts — have to tell us about human emotions and their interaction with systemic designs.
If we fail to account for the role of emotions and ergonomics in the vote, we will continue to predict the way in which people ‘intend to vote’ rather than the ballot that they actually cast on the day. On the face of it, particularly at the aggregate level, the error caused by this discrepancy may not be obvious, but it will be there nonetheless and constitute a consistent dark shadow in our understanding of political behaviour. We will also fail to see how details which we always assume to be neutral — such as the design of ballot paper or the facilitation of electoral procedures through internet voting — actually affect whether people participate in elections or not, the way they vote, and the way they feel towards their electoral democracies. Instead, we need to understand how the way elections are organized and administered as well as their specific atmosphere trigger specific emotions, which may consciously or subconsciously lead voters to behave in ways that they may not even have been able to predict themselves a week before election day. This means opening new theoretical avenues of research, but also new methods and new bridges and collaborations to match the extraordinary mysteries that human behaviour still holds in the field of elections.