Back to Blog

Posted: 17.01.24

Polls Apart

Our Managing Director, Insight Nick Bowes looks at a busy week on the polling and electoral analysis front and what it might mean for London.

The Government have faced a few bruising weeks in recent months, so in that respect last week’s struggles with the Rwanda Bill feels all too familiar. But we also saw some important polling and the publication of crucial analysis which shed more light on the coming General Election, the scale of the task ahead for Labour if they’re to win and the stubborn refusal of the Conservative’s unpopularity to budge.

1.     Knowing Your Boundaries

First up, academics Colin Rallings and Michael Thrasher published their analysis of the new parliamentary constituency boundaries. In political circles, Rallings and Thrasher are about as close as we get to a celebrity duo – their work will be regularly cited on the run in to the General Election.

The Boundary Commission have undertaken an extensive exercise to redraw the parliamentary constituency map. Populations change, and periodically a rejigging of boundaries is required to keep them broadly similar in size[1]. While the overall number of MPs isn’t increasing from the current 650, there are changes across the country – some regions and nations increase their share, others see a fall.

What Rallings and Thrasher do is analyse the outcome of the 2019 General Election result but on the new boundaries – to give notional majorities – and it provides a glimpse of the huge task ahead if Labour want to win the next General Election.

A Monumental Task for Labour

Here are four scenarios to consider:

  • For Labour to simply deprive the Tories of a majority (e.g. lose 47 seats) will require a 4.2% swing .
  • For Labour to be the biggest party in a hung parliament (e.g. win at least 81 seats), the swing will need to be 8.2% .
  • However, if Keir Starmer is to gain an overall majority of just one, he’ll need a swing of 12.7% (e.g. win at least 125 seats)
  • And for a decent workable majority of over 30 the swing increases further to 13.8%.

To put this last stat in context, the biggest swing since the war was the 10.2% Tony Blair’s Labour achieved in 1997. Margaret Thatcher achieved a swing of 5.3% on her way to victory in 1979, David Cameron 5.1% in 2010 and Clement Attlee 11.8% in the 1945 Labour landslide – so the mountain Labour need to climb is truly unprecedented this side of the Second World War.

London’s New Political Map

London is one region that is seeing an increase in the number of MPs it elects – up two from 2019[2] (75 from 73).This has led to quite substantial redrawing of the electoral map – some constituencies are unchanged or only modestly effected. In other cases, whole new constituencies are created – such as Croydon East, Edmonton & Winchmore Hill, Queen’s Park & Maida Vale. In my own corner of south-east London, the shuffling of the political map sees me move out of Lewisham Deptford into a new constituency of Lewisham West and East Dulwich – Rallings & Thrasher predict this new seat has a notional 2019 Labour majority of 23,823. For those interested, you can put your postcode in here and find what changes are taking place where you live.

In London, if the 2019 election had been fought on the new boundaries, Labour would have won three more MPs, and the Tories one fewer. And we can see how the redrawn boundaries result in some interesting new battlegrounds.

  • The new seat of Beckenham & Penge in Bromley is deemed marginally Labour – with the Tory’s chances not helped by long-standing local MP Bob Stewart not seeking re-election. So this will be one to watch.
  • Not far away, Rallings and Thrasher’s analysis identifies the new constituency of Eltham and Chislehurst as marginally Conservative. Labour’s Clive Efford – MP for the old constituency of Eltham since 1997 and with a high profile locally – will be fighting this new seat.
  • The old constituency of Kensington – currently represented by Tory Felicity Buchan with the city’s smallest majority of just 150 – becomes Kensington & Bayswater, with Rallings & Thrasher deeming this new seat as narrowly favouring Labour.
  • Crucial on election night will be the seat of Chelsea & Fulham, home to Tory Chairman Greg Hands – this is the key seat Labour will need to capture to become the largest party.

It is important to stress again that Rallings & Thrasher’s analysis is based on the 2019 election result – it does not take into consideration current opinion polls. So there’s much yet up for grabs, although with the scale of the Labour lead seen over the past year, all notional Tory majorities up to 10,000 and more look vulnerable to falling to Labour.

The Two Latest Polls

In the days after Rallings & Thrasher published their analysis, we’ve seen two really important polls which confirmed the trends we’ve seen, and will have put wind in Labour’s sales, but left Rishi Sunak and the Tories pretty deflated.

First was a YouGov MRP[3] poll commissioned by the previously unknown group, the Conservative Britain Alliance[4]. The poll predicts that Keir Starmer is on course to win a majority of 120, with Labour on 385 seats and the Tories down at 169 – just four more than John Major’s Tory Party won after Blair’s landslide in 1997. A slug of cabinet ministers would lose their seats, including Jeremy Hunt, Penny Mordaunt and Grant Shapps.

If the YouGov MRP poll were played out at the General Election in London, the Conservatives would be left with just six MPs out of 75[5]. Inner London would no longer have a single Tory MP and big names like Iain Duncan Smith would lose their seats. Labour would even capture Uxbridge and South Ruislip – the seat which they narrowly failed to win at the 2023 by-election.

A second YouGov poll published later last week saw Labour extend their lead over the Tories to 27 percentage points – 47% v 20% - the highest lead since Liz Truss was Prime Minister. Interestingly – and of concern to the Conservatives given the particular threat to their vote - Reform are currently at 12% - ahead of the Lib Dems and the Greens – the highest vote share for them recorded to date.

Labour is also ahead in all age brackets except over 65s – and remarkably just 4% of those under 24 plan to vote Tory. It’s only slightly higher at 12% for those aged 25-49. And as some have pointed out, Labour’s lead at this point in the election cycle is now larger than in 1997 – and while Labour’s lead was shrinking by this point on the run in to 1997, ominously for the Tories this time it is increasing.

There are many bumps in the road ahead – there is considerable uncertainty in the world, and the situation in the Middle East looks particularly worrying. But in the Sky News poll of polls – which takes an average across a range of different opinion polls – Labour head towards February with a twenty point lead over the Tories. It is a long way back for Rishi Sunak and, despite an improving economy, falling inflation and possible tax cuts in the March Budget, narrowing this gap any time soon looks difficult, meaning any lingering prospect of a Spring election must be receding fast.




[1] While the Boundary Commission are independent, Parliament legislates the framework in which they operate – all (except a handful) of constituencies must be within 5% of an average electorate size of 73,392. Note this is ‘electorate’ – so those on the electoral register – not population size, or all those eligible to vote.

[2] As an aside, up to 1997, Greater London elected 84 MPs, representing a population of around 6.8million yet fast forward to 2024 and just 75 MPs will be elected for a population of just shy of 9 million.

[3] MRP – short for multi-level regression and post-stratification – is a kind of sophisticated statistical methodology which, as well as asking voters how they intend to vote, also captures other important data which is used to calculate how various groups of people are likely to vote. They are a more recent invention but came into their own in the 2017 General Election when successfully predicting some constituency results that traditional polling methods failed to predict.

[4] Some have speculated at what the motives of this new group were, publishing a poll this hostile to the Government just before a crucial vote on the Rwanda Bill.

[5] Holding Ruislip, Northwood & Pinner; Finchley & Golders Green; Upminster; Old Bexley & Sidcup; Orpington; and Carshalton & Wallington.