Does political partisanship affect the likelihood of an individual choosing to get vaccinated against Covid-19? Using data from constituencies across England, Margaryta Klymak and Tim Vlandas show that Conservative voters and those who live in Conservative constituencies are more likely to be vaccinated than Labour supporters.
The UK has over 10 million confirmed cases and more than 145,000 deaths, making it one of the most affected countries by the Covid-19 pandemic. Similar to other governments across the world, the UK had initially attempted to contain Covid-19 with a national lockdown.
These measures have been found to be effective at reducing transmission, but vaccination is now the safest and most effective way of containing the pandemic while keeping the economy open. Recently, the UK’s Prime Minister, Boris Johnson, announced a new focus on providing ‘booster vaccines’ to as many people as possible before the end of the year.
Yet, over 20% of the population is not yet fully vaccinated and there is substantial variation across the country. It is therefore crucial to understand the drivers of this variation: partly to ensure that places where two doses have not been sufficiently administered can receive the attention they require; partly to ensure that existing gaps do not get reproduced with the booster vaccine.
Partisan polarisation: what effect on vaccination?
One noteworthy potential source of variation concerns partisan affiliation which maps onto other ideological and material cleavages. Political science literature has long documented increased partisan polarisation with voters of different political parties thinking and behaving differently. However, we still know comparatively less about whether and how public health responses may be undermined by partisan divides. While there is emerging evidence on how partisanship may have influenced individuals’ perceptions of the pandemic and national lockdowns, less is known about how partisanship affects vaccination.
If vaccination rates interact strongly with pre-existing party preference divisions in societies, this might undermine the ability of governments to effectively address the pandemic: certain voters may hold systematically distinct beliefs making them reluctant to be vaccinated, precisely at a time when concerted and decisive collective action is most needed. As a result, there is a risk that partisanship undermines our public health response to Covid-19, by making certain individuals sceptical of the need to get vaccinated or worried about the risks of doing so.
Partisanship and vaccination rates across English constituencies
In recent research, we seek to address this gap by examining the extent to which partisanship is associated with variation in vaccination rates. To do so, we analysed partisan differences in vaccination rates using two separate datasets.
The first dataset contains constituency level vaccination data in October 2021 that is matched with the latest general election results in England. Figure 1 below shows the vote share across 532 constituencies (panel A) and the variation in vaccination rates (panel B). Red and to a lesser extent light blue constituencies shown on this map show places where Conservative vote shares were lower. Comparing patterns of votes to vaccination rates shows that constituencies with higher Conservative support also tended to have higher vaccination rates as of October 2021.
Figure 1: Vaccination and partisanship across English constituencies
Note: This figure shows the vote share (left hand side) and vaccination rate (right hand side), both expressed as percentage points, in different English constituencies. Vaccination rates were calculated by dividing the number of first doses in a constituency by the total population over 18 years old of that constituency and then multiplying this ratio by 100.
Figure 2 plots the vote share (horizontal axis) against the vaccination rate (vertical axis), both expressed as percentage points, in different English constituencies. A regression line displays the strong positive association: the reported coefficient (0.295 – with 95% confidence interval between 0.266 and 0.325) suggests that one percentage point increase in conservative vote share is associated with 0.295 percentage point higher vaccination rates (second doses). Differences in Conservative vote share account for 42% of variation in vaccination rates in second doses in England.
Figure 2: Scatter plot of vote share and vaccination rates across English constituencies
Note: This figure shows the vote share (horizontal axis) and vaccination rate (vertical axis), both expressed as percentage points, in different English constituencies.
How much of this partisan gap is driven by confounding differences in socio-demographic and economic differences across constituencies? We find that this partisan effect holds when controlling for differences in house prices and wages, population composition, the health and deprivation of the constituency and past austerity.
In addition, our results are also robust to rerunning our analysis on vaccination rates for specific age groups, thereby ruling out a confounding effect of demographic differences between Conservative and Labour voters. Conservative votes are associated with higher vaccination rates among all age groups, except those under 18 and over 75 for the first dose only.
The second dataset is a survey carried out by YouGov comprising five waves collected weekly between October and early November 2021. Each wave has between 1,648 and 1,729 respondents and the survey includes variables about past votes in the 2019 general election, how respondents voted in the Brexit referendum, age, number of children, gender, social class, and region location. Controlling for all these confounding effects, we still find that Conservative voters are more likely to be vaccinated than their Labour counterparts.
Our findings point to the importance of partisanship in understanding the patterns in vaccination rates across England. Conservative voters and those who live in Conservative constituencies both appear more likely to be vaccinated. How much of this gap is due to material, rather than ideological, differences? We find that at the constituency and individual level a large part of the differences between Conservative and Labour is absorbed by socio-demographic and economic differences between the two groups. We interpret the remaining statistically significant difference as evidence of some ideological differences but can only speculate about its sources: one possibility is that Labour voters are more likely to distrust the Conservative government; another is a differential vaccine availability or ease of access to Labour voters.
Regardless of its source, the direction of the partisan effect contrasts markedly with recent studies about attitudes towards Covid-19, as well as with the US case, where Republicans are less likely to follow social distancing rules, wear masks and to get vaccinated than Democrats. Thus, the UK government seems to be more effective at convincing people in Conservative constituencies to get vaccinated. But paradoxically, it now faces the reverse problem: in the US, the Democrat led government must convince Republicans to get vaccinated, whereas the UK Conservative government will now need to improve vaccination rates in Labour constituencies and among Labour voters.
Margaryta Klymak is a Fellow and Tutor in Economics at Somerville College, University of Oxford.