03 December 2019

Polling in the Margins

Paula Surridge looks at Labour’s election chances in Hastings and Rye and Chipping Barnet

At the start of an election campaign it is usual for lists of ‘target’ seats for each party to appear: the seats the party needs to win over in order to head into Downing street on the day after polling day. While the 2019 election campaign is a little different, as each of the major parties has reason to look closely at how to defend the seats it currently holds as well as those on its ‘most wanted’ list.

To aid in this task the parties are blessed with an ever-growing number of sites giving detailed breakdowns and advice on a constituency by constituency basis, as well as the proliferation of ‘headline’ vote intention polls. How can activists and voters make use of the array of polling data to work out what might be happening in their local constituencies? While the various MRP models, tactical voting sites and on-line calculators give ‘now casts’ for constituencies they are often not easily adapted when headline polling changes or to the ‘local’ sense being gleaned by campaigners out in the constituencies. For this reason, when asked about any constituency I prefer to look at likely scenarios to think through how the results of a seat might look under different types of polling movements.

In this piece based on the my chapter from Corbynism from Below, I set out the logic behind my starting points in a constituency. In a shifting electoral landscape it is inevitable that some things have moved on in the last few weeks. Nigel Farage’s decision to stand down Brexit Party candidates in seats the Conservatives already hold changes the dynamic not only in those seats but also makes it less likely that the Brexit Party will be able to generate the same level of effect elsewhere where they do stand. My ‘best guess’ currently is to put in a Brexit party vote share a little lower than the 2015 UKIP share in constituencies where they are still standing.

Liberal Democrat support has also fallen since the start of the campaign but remains well above 2017 levels. It also appears to be more concentrated than in 2010 and so my starting position now would be to put at 2010 levels in constituencies where the proportion with degrees is above average but a below 2010 polling elsewhere.

Another development since I wrote the original piece is this website by Christabel Cooper which brings together almost all the data I recommend using and removes the need to visit Wikipedia, Democratic Dashboard and Constituency Explorer (though the 2010 election result is still worth digging out in most places).

Using the publicly available data and the idea of scenarios rather than predictions, we can compare what might happen in two of Labour’s target seats – Hastings and Rye and Chipping Barnet (target seats 7 and 8 respectively and targets 3 and 4 in England).

These two seats while almost identical in the swing needed to turn ‘red’ on December 13th have very different political contexts. Hastings is a ‘Leave’ leaning seat, with slightly below average levels of population with a degree and with almost 95% of the electorate of white British ethnicity. In contrast Chipping Barnet is a ‘remain’ leaning area, with higher than average education levels and a more diverse electorate, with 72% of white ethnicity. Based on how the profile of Labour voters changed between 2010 and 2017, we would expect that Chipping Barnet would be a more ‘natural’ Labour seat in 2019. Both seats have relatively weak challenges from the Liberal Democrats, with the LibDems in 3rd place in both seats in 2010, while as both seats were held by the Conservatives in 2017, neither are contested by the Brexit party at this election. As UKIP did not stand in Chipping Barnet in 2017, this doesn’t change the context here very much, but in Hastings UKIP had a notable presence in 2015 (with 13% of the vote) and retained almost 3% of votes in 2017.

Using this political context as the starting point, the scenarios in the two constituencies look a little different. Beginning with Chipping Barnet – in this seat the performance of the Liberal Democrats (and which of the major parties lose more votes to them) will be critical. The Conservative vote share fell here in 2017, despite UKIP not standing and we might expect at least a small further decline as some Conservative voters move to the LibDems. However, there is also potential here for the LibDems to also take votes from Labour. At this stage in the campaign, it looks as though many of the remain leaning Labour voters will stay with the Labour party, to win Chipping Barnet the party needs to ensure it holds on to those remain voters as tightly as possible. The current best estimates have this seat as a tie and so the need to further squeeze the LibDem vote in this seat is urgent for the Labour party. It would be a very poor night for the party if it failed to win over a target so high up the list with favourable demographics.

Hastings is rather different in its demographic profile, in some respects looking more akin to seats Labour are defending in the North and Midlands than surrounding seats in the South East. At the simple aggregate level, the collapse of UKIP between 2015 and 2017 seems to have benefited Labour to a greater degree than the Conservatives. As there will be no Brexit Party candidate in this seat, the danger for Labour is that leave leaning voters switch directly to the Conservatives this time round. This is a seat where Labour may be treading a tight rope in trying to hold on to voters from both sides of the referendum debate. The LibDems are likely to pick up at least some of the remain voters here and exactly how many could be critical for whether Labour can win this seat. Even a small resurgence of the Liberal Democrats here could put this seat out of reach for Labour given that there is also likely to be some direct switching of leave voters to the Conservatives. The best estimates currently suggest Labour are much further behind here than in Chipping Barnet, and it seems unlikely that the party can squeeze quite enough votes from the LibDems here to win. A pattern that may be repeated elsewhere in the country where Labour is under attack on both its leave and remain flanks.

How constituencies will vote is never certain, and perhaps less so in 2019 than in previous elections, despite our wealth of data and models. Local campaigning and events along the way may matter and move things decisively one way or another. In less marginal constituencies, different scenarios deliver the same result (for example Bethnal Green and Bow as shown here) but in many seats across the country different outcomes are within what is plausible given the current data. While voters are themselves volatile and unpredictable; it is the range of plausible outcomes in a large number of seats that makes the shape of the House of Commons on Boxing day so difficult to predict.

Corbynism from Below is now 30% off in our end of year sale.