Posts Tagged ‘elections’

Explaining support for Brexit among would-be MPs in the 2017 general election

Siim Trumm, Caitlin Milazzo, and Joshua Townsley use survey data to explore support for Brexit among 2017 parliamentary candidates. The findings reveal some interesting insights into politicians’ motivations for supporting Brexit. For example, candidates standing in seats with higher Leave support were not significantly more likely to vote Leave themselves. Instead, candidates’ views on immigration and democracy stand out as key determinants of their decision to vote Leave.

The United Kingdom’s decision in 2016 to leave the European Union marked a significant historical juncture and has defined politics since. The referendum result was also somewhat unexpected, leading to several studies exploring the question of who supported Brexit. Most of these have, however, focused on voters.

We conducted a study to explore the extent of support for Brexit among parliamentary candidates and the factors that drive it. What did we expect? Of course, there should be clear differences along party lines, with Conservative and UKIP candidates more likely to support Brexit than their Labour and Liberal Democrat opponents. But beyond this, we might also expect some of the same factors that drive Leave versus Remain divide among voters to be at play among politicians. We might, therefore, expect support for Brexit to be stronger among male candidates, older candidates, and those who have an occupational background in manual work. Brexit support might also be higher among candidates who hold stronger British identity and are more concerned about immigration. We also expected candidates who seek to represent areas with higher levels of support for Brexit to be more likely to support Brexit themselves.

We analysed data from the 2017 British Representation Study to test our expectations. Our analysis includes responses from 872 candidates who stood in the 2017 election. The sample is broadly representative of all 2017 candidates in terms of partisanship, nation, and electoral performance.

We ran a series of logistic regression models to predict Leave support among candidates. Table 1 presents the predicted probabilities (in effect, the likelihood of a candidate supporting Leave based on a particular characteristic, while the other characteristics are held constant) for candidates’ likelihood of voting Leave, to illustrate the findings of our models. The first characteristic, for instance, shows that the probability of voting Leave is 24% among candidates who consider immigration-related issues to be the most important problem facing the country (versus 8.6% among those who do not).

So, what explained candidates’ support for Leave? Interestingly, we find that candidates’ gender, age, occupational background, and incumbency status have no systematic effect on their likelihood of voting Leave or Remain. This appears to contrast with what we know about how voters broke for Leave versus Remain. We also find that, after controlling for all other variables, candidates who stood in constituencies with stronger Leave support were not significantly more likely to support Brexit themselves. This, again, reinforces the narrative that would-be MPs were somewhat ‘out of touch’ with their voters.

Elsewhere, we found similarities. Like voters, candidates were more likely to support Brexit if they considered immigration to be the main problem facing the country and thought negatively about its cultural impact. Candidates who believed that Brexit would be positive for the UK’s economy and democracy, and were more critical about the state of EU democracy, were also more likely to be Leavers than those who did not hold such views.

Interestingly, we can also compare the pre-referendum voting intention and the actual vote choice for 442 candidates by combining survey data from the 2015 and 2017 British Representation Study. Whereas polling of voters suggests that there was quite a bit of fluidity in public opinion during the 12 months leading up to the EU referendum, this was not the case among politicians. Table 2 shows that only 12 (out of 442) candidates changed their mind in the run up to the 2016 referendum, constituting only 2.7% of all candidates we have this information for. An overwhelming majority of candidates (97.3%) remained unchanged in their vote choice. Unsurprisingly, the most common option was candidates who intended to vote Remain indeed casting their ballot for Remain (363 candidates, or 82.1%). Meanwhile, 67 (15.2%) candidates who intended to vote for Leave do so. In other words, there were very few candidates who switched their voting choice in the immediate run up to the EU referendum.

There are three broad conclusions arising from our study. First, the evidence reaffirms the narrative in British politics that politicians were ‘out of touch’ with voters when it came to supporting Brexit in 2016. They were much less supportive of Brexit than the electorate as a whole and this gulf was not just between voters and the political elites at the very top of the political ladder, but between voters and the ‘political class’ more broadly.

Second, support for – and opposition to – Brexit was quite stable among would-be MPs. While support for Leave varied between politicians with contrasting political beliefs and perceptions of immigration, we did not find evidence that many candidates switched sides in the run up to the EU referendum.

Finally, the findings reveal some interesting differences and similarities between the motivations for Leave support among politicians and voters. Contrary to our expectations, we find that some of the factors that drove voter support for Brexit (e.g., age, gender, occupational background) do not have the same effects among candidates. But, just like voters, more optimistic views of how Brexit was expected to impact British democracy and economy, as well as attitudes on immigration, are associated with greater likelihood of voting Leave.


Note: The above draws on the authors’ published work in Political Studies.

About the Authors

Siim Trumm is Assistant Professor in Politics at the University of Nottingham. His areas of research include British politics, party politics, parliaments and legislative behaviour, and electoral campaigns.



Caitlin Milazzo is Associate Professor in the School of Politics and International Relations at the University of Nottingham. Her areas of research include electoral behaviour, British politics, and political campaigning.



Joshua Townsley is a Research Fellow at the University of Nottingham. He researches electoral campaigns and political behaviour.




All articles posted on this blog give the views of the author(s), and not the position of LSE British Politics and Policy, nor of the London School of Economics and Political Science. Featured image credit: by Parker Johnson on Unsplash.

How data-driven democracy both helps and hinders politics

Much data relating to parliament is now being collected and made available for anyone to access. Does this monitoring mean more democracy? Ben Worthy and Stefani Langehennig argue that the resulting numbers often lack context and so feed into subjective narratives.

We live in a data-driven democracy, never more than now. Our eyes are locked on graphs, measures and comparisons, and an unending stream of data offers ways of making sense of and quantifying the political world around us. Yet our faith in the ‘transparency of numbers’ might be misplaced. Layers of biases, uncertainties and inequalities lurk beneath the clarity they offer.

One target of this data-gathering is Westminster, which is what our new Leverhume Trust project looks at. The project combines analysis of media stories and social media with case studies and surveys to map out who is using all of this data and what impact it is having, both on those being watched and those doing the watching.

You can now easily measure, analyse and compare what MPs and peers are doing. For any curious citizen, there is now a whole raft of data sources about how elected and unelected members vote, how they use their expenses, or what they do or don’t do outside their political role. Postcode look ups, where you can find which way your MP went in key votes, are now a regular feature. New sites analyse and update members ‘expenses and register of interests data.

The data creates a set of benchmarks and measures, and a kind of ecosystem of political monitoring. They are often used as a shortcut to tell us about individual politicians we don’t know, and to get a sense of where they stand, or stood, on various issues. It creates a trail of accountability that can come to haunt politicians, as seen recently with controversy over who voted against pay rises for the public sector in 2017. It also opens up the group dynamics of blocs of MPs and can tell us, for example, who blocked Brexit (not who you think).

But how much transparency does it really offer? The first problem is that numbers over-simplify a complicated world. Quantification, measures and numbers offer us the illusion of certainty. Voting for or against something can be far from simple. The website that provides this data itself warns that ‘bit more subjectivity comes into play’ in interpreting voting decisions. Take Huw Merriman’s explanation of the contortions members faced in the Brexit vote in April 2019:

I passionately believe that we have to follow the 2016 referendum result, even though I voted remain. I voted for the triggering of article 50, to keep no deal on the table, against a second referendum and against a long delay to our exit date. My voting record in Parliament reflects the will of the British people…anything else would lead to huge mistrust in our political system.

Much of what MPs do is hidden or very difficult to quantify. There is no comparable data, for example, on how many constituency surgeries MPs hold, and we can only see on social media or the local press what they do in their communities. Even in Westminster, valuable work in committees or bending ministerial ears is necessarily out of sight.

The second problem is that objective numbers lead to subjective judgment. Data makes lists and comparisons easy – almost too easy. Any benchmark risks slipping from describing something to making a moral judgement about it. Jeremy Corbyn’s voting record either proves he was ‘on the right side of history’ or would ‘make Thatcher proud’, depending on your taste. Data on expenses can tell us who is the ‘worst abuser’ or who claimed the least (22p for a banana, before you ask). But what is a good level of expenses use and what does that tell us? Even the public sector pay vote is more complex than it looks and, as Full Fact pointed out ‘needs some context’ as did the £10,000 pay rise that was for office costs.

The judgment itself is skewed. Female MPs suffered more from the expenses crisis. The Sun had to apologise to one female MP on its list of ‘lazy’ MPs. The House of Commons was concerned that well-off MPs could afford not to claim any expenses, while less wealthy had to – and risk criticism. Even more complex is the House of Lords, where everything the data reveals is bound up in the slow-burn question of the ‘other places’ legitimacy, role and reform.

The final problem is that these measures become ‘Engines of Anxiety‘ for those being watched, who then react accordingly. Can you game the numbers? It’s often said Written Question numbers shot up when TheyWorkForYou used it as measure of activity. Nick De Bois MP pointed out how speaking in the House was done:

Sometimes…so you can enlighten constituents on your position on any given issue. Either that, or because it’s not a good thing to have against your name ‘Below-average number of speeches in the House of Commons’ on that pesky ‘They Work for You’ website, which relentlessly measures how active you are in the chamber.

The danger is that these new tools help and hinder politics, a problem of numbers versus narrative. Taken together, this data creates a kind of continuous scrutiny, which constantly expands. Beyond expenses you can see, for example, which former MPs still have Parliamentary passes or even if someone with a Parliament IP address has made changes to Wikipedia.

The numbers help the public to know and understand more, more simply and make it easier to hold politicians to account. Yet the numbers themselves are trapped in a narrative, a familiar story of expenses, interests and black and white views. The danger is that our new data-driven democracy reinforces an age-old tale about politicians.


Note: the project on which the above draws is funded by the Leverhulme Trust. If you have used Parliament data, please help with the project survey here.

About the Authors

Ben Worthy is Senior Lecturer in Politics at Birkbeck, University of London.




Stefani Langehennig isPostdoctoral Researcher at Birkbeck, University of London.



All articles posted on this blog give the views of the author(s), and not the position of LSE British Politics and Policy, nor of the London School of Economics and Political Science. Featured image credit: by Nikolay Tarashchenko on Unsplash.

Powered by WordPress | Designed by: index backlink | Thanks to insanity workout, car insurance and cyber security