Lauri Rapeli, Mikko Mattila, and Achillefs Papageorgiou combine two panel surveys, conducted in the UK between 1991 and 2017, to examine the impact of unemployment, retirement, changes in partnership status, moving and disability on voting. They find that turnout declines with divorce; for other life events, the impacts diverge across the voter groups they identify.
Studying why some people vote while others do not has traditionally been one of the most intensely examined topics of political behaviour. Some researchers have emphasised the importance of contextual determinants, such as the competitiveness of elections, while others have focused on individual-level predictors of voting.
When it comes to the individual-level determinants, researchers have offered many different explanations. Among the most influential are the rational choice model, the psychological model, and the resource model. In the rational choice model, voters seek to maximise personal economic benefits. The psychological model sees voters as driven by emotions, such as strong identification with a certain political party. The resource model emphasises the importance of civic skills, time, and economic and social resources as key factors, which lower the threshold of voting.
Alongside these models, many scholars agree that voting is also a habit. As argued by theories of political socialisation, people develop a stable habit of (non)voting during late adolescence and early adulthood. There is plenty of evidence, mainly from experimental research, suggesting that this voting habit does not easily change. Once developed, it mostly stays the same during the entire life course.
In our study, published in Political Studies, we test the persistence of voting habits by examining whether various life events affect turnout. Our analysis uses two panel surveys, conducted in the UK between 1991 and 2017: the Understanding Society: the UK Household Longitudinal Study (UKHLS) and its predecessor the British Household Panel Study (BHPS). The period covers 25 years and seven British general elections.
This timeframe allows us to connect a long history of electoral behaviour with many events that occur in different stages of life: changes in partnership status, such as entering or exiting marriage or cohabitation or becoming a widow(er), becoming unemployed or retiring from workforce, moving or experiencing health problems. This is not an all-encompassing list of important life events, but it offers a more nuanced picture of the impact of various life changes on turnout than what previous scholarship has done.
Theoretically, apart from entering a steady relationship, and perhaps retiring, all life events included in our analysis should decrease turnout. Losing a partner, exiting working life, moving to a new neighbourhood or suffering from poor health are all likely to have an adverse effect on social connections, and also to cause general stress. Major life changes require attention and energy, which leaves less resources for other activities, such as politics. Profound changes in a person’s life conditions might also alter political priorities, which could have consequences for behavioural patterns, such as voting. Similarly, entering a relationship might invigorate a person’s social life, provide new ways of political mobilisation and therefore activate a passive voting habit.
In order to examine how the various life events affect people with different voting habits, we divided our sample into three groups: 1) respondents who voted in previous two elections (habitual voters, 68 % of observations); 2) respondents who voted in only one of the previous two elections (occasional voters, 18 %); 3) respondents who voted in neither of the previous two elections (habitual non-voters, 14 %). For each of these three voter groups, we then calculated the probabilities of turning out to vote depending on whether a person has experienced a certain life event or not in the period leading up to the election.
We found many differences across the three groups in terms of how life events influence voting. Starting cohabitation with a partner, for example, has a clear positive impact on habitual non-voters and occasional voters, while it has a small negative impact on habitual voters. All groups are heavily affected by divorce. The negative impact (-26.5 percentage points) is largest for occasional voters, but divorce also diminishes the probability of voting among habitual voters by over 16 percentage points. This can be considered to be a substantial effect, especially since it concerns people who have a habit of voting. In terms of moving, occasional voters and habitual voters are affected about equally, with a negative impact of a little over ten percentage points. Becoming disabled only lowers voting propensity among habitual voters. Retirement, on the other hand, increases voting only among habitual voters, although only by about four percentage points.
Overall, the results suggest two conclusions with particular relevance to the literature on turnout. Firstly, it seems clear that although turning out remains very high among habitual voters regardless of life events, the voting habit is by no means resistant to changes in life circumstances. We would therefore argue that, although our analysis does not suggest overthrowing the voting habit thesis altogether, it seems that future scholarship should do better in acknowledging the malleability of voting habits. In our analysis, habitual voting does not seem to be quite as resistant to change as many previous analyses have assumed. Secondly, all types of changes in partnership status as well as moving to a new address have the largest effects on turnout. This implies that broken social connections are the concrete mechanism underlying the findings. People are, first and foremost, social creatures and this shows also in their voting behaviour. Perhaps, as we are tempted to speculate, political scientists should start to give more emphasis to social needs as the primary driving force behind political behaviour, rather than ideological commitment or maximisation of personal economic gain? Perhaps the key to reinvigorating democracy, or at least turnout rates, in the end lies in strengthening the social fabric among democratic publics?
Note: the above draws on the authors’ published work in Political Studies.
Lauri Rapeli is Head of Research in the Social Science Research Institute at the Åbo Akademi University, Finland.
Mikko Mattila is Professor in the Faculty of Social Sciences at the University of Helsinki, Finland.
Achillefs Papageorgiou is Researcher at the University of Helsinki, Finland.