Research

Partygate is one of the most significant political scandals to have taken place in the United Kingdom, with then Prime Minister Boris Johnson becoming the first to break the law while in office. However, the consequences of Partygate for British politics remain under-explored in political science research. In this paper, I shed light on the impact of Partygate on identification with the Conservative Party through an Unexpected Event during Survey Design. As news of the scandals broke out among the British public in November 2021, collection of Wave 21 of the British Election Study Internet Panel was under way. This created a natural experiment where some respondents completed the survey “before” Partygate while others completed the survey “after”, thus being as-if randomly assigned to a control and treatment group. Using a difference-in-differences estimator, I find that Partygate caused an immediate decline in identification with the Conservative Party. The results are robust to alternative estimation approaches, including regression discontinuity in time. These findings bear important implications for the study of political behaviour: they contradict claims of increasing stability of party identification and may shed light on democratic accountability in the United Kingdom.

The question of why individuals harbour a given set of political attitudes is central in political behaviour research. One common answer is that political attitudes are developed through early political socialisation, but our understanding of this process remains limited. To improve theoretical clarity, this paper proposes a model of pre-adult socialisation of party identity within the family closely based on social learning theory. The proposed model is tested on data from the British Household Panel Study (1991-2008) and the UK Household Longitudinal Survey (2009-2021). The empirical analyses provide support for some parental characteristics -- party identification strength and the number of parents -- enhancing the likelihood of political socialisation, as well as for the communication mechanism. No evidence is found in support of the contact mechanism or affective relationships as moderators. Rival explanations of parent-child attitude congruence that draw from political economy and behavioural genetics are also rejected. Taken together, these results suggests that political socialisation remains a strong candidate to explain the formation of political attitudes, but social learning theory may not be how it occurs.