Voting patterns are usually analysed using survey data obtained by interviewing individual respondents. Relatively little attention has been paid to the role that within-household relationships have on how people vote. However, the Understanding Society survey undertaken by the Institute for Social and Economic Research at the University of Essex interviews all members of the households in its sample Using an early release version of the data the survey collected throughout 2016, the year of the EU referendum, we have used the survey to analyse the extent to which (a) couples and (b) parents and children share the same attitudes towards the UK’s membership of the EU, and whether the level of agreement between them is greater or less than in a general election.
In our analysis, published today, we show that couples were much more likely to share the same views about EU membership than would be the case if we were simply to pick two people out of the electorate at random. As many as 79% held the same view, compared with the 51% who would have done so by chance. Meanwhile parents and children also shared the same opinion more often than we would expect by chance, although, at 71%, the level of agreement was rather less than in the case of couples.
However, after taking into account the fact that voters face a wider array of options in a general election than the binary choice in the EU referendum between Remain and Leave, the levels of agreement were lower in the EU Referendum than in the 2015 election. For example, when we restricted the analysis of voting in the election to a binary choice between Conservative and Labour, as many as nine in ten (89%) couples voted the same way, as did 86% of parents and children. It seems that the fact that in the referendum voters were being faced with a new question rather than a familiar one meant that the various processes that foster intra-household agreement about politics had less impact on the pattern of people’s choices.
For the most part, the level of agreement – both between couples and between parents and children and in both the EU referendum and the 2015 election – did not vary according to whether they had the same or different demographic characteristics. However, couples were consistently more likely to agree with each other if only one of them was working, suggesting that financial dependency makes agreement more probable. In addition, older couples are more likely than younger couples to vote the same way, which is consistent with the idea that agreement between couples is partly a result of the discussion and persuasion that takes place between them over the years. There is, though, very little evidence that the demographic background of parents and children makes any difference to the level of agreement between them – a result that is not surprising if agreement between them is primarily the product of children being socialised by their parents.
By Alina Carabat
Alina is a Statistician based in the Statistics team at NatCen, responsible for survey sampling, weighting and statistical modelling.