The Greens’ path to breakthrough runs through the Labour Party

The Greens’ path to breakthrough runs through the Labour Party

Since the 2019 election, the Greens have been quietly creeping up in polls. Now sat at 5-7% in averages, the party is at its highest point since summer 2019, which was itself the Green’s best polling since before 2015. 5-7% might not seem like a lot, but in a FPTP election with dozens of seats on a knife edge, it could make all the difference.

The problem for Greens, like all smaller parties, is how to turn this support into MPs. Unlike the Liberal Democrats or Plaid Cymru, there are no historic strongholds to concentrate campaigning and a much smaller (but growing) presence in local elections. Without the PR European Elections, too, there are no national elections in England where the Greens can expect to win a proportional number of seats.

Historically, the Greens (at least in England and Wales) have struggled with General Election targeting. Outside Brighton Pavilion, the party has sensibly prioritised Bristol West since 2010 and made significant inroads, but there are virtually no other breakthrough opportunities.

In 2015 in particular, newfound support was difficult to coordinate geographically. The party targeted 11 seats alongside Brighton Pavilion. In three of these (Solihull, Reading East and York Central) the Greens did not even get their deposit back, while in a fourth (Norwich South) the party did worse than 2010. This illustrates the challenge of targeting seats when starting from virtually (electorally) nowhere.

One particular problem for the party is the tension between local and general election coalitions. In some areas, the Greens have been very successful at picking up unexpected wards. This includes significant Councillor groups in rural areas such as Suffolk, Gloucestershire, Herefordshire and Mendips, some working class, leave-voting areas in Labour heartlands, and a few random ones, such as being the official opposition in Solihull since 2014.

There is an illusion that local election victories will naturally transfer to national election votes. This is perfectly illustrated by Solihull, where the Greens routinely win 25% of the vote in local elections but have never won more than 5% in either of the parliamentary constituencies. On the flip side, some of the Green-est constituencies in the 2019 European Elections (according to Chris Hanretty’s estimates here), have no elected Green Councillors at all (Bath, Leeds North West, Manchester Withington, Lewisham Deptford etc).

So, as the Green Party of England and Wales embarks on a leadership election and the Scottish Greens join a government for the first time, where might a Green breakthrough at Westminster come?

Green-Curious Voters

To consider where future Green gains might come, we have to study a measure of support for the party other than previous election results – we are interested in those people who have not voted Green before, not those who have.

In June 2020, the British Election Study (BES) asked respondents “How likely is it that you would ever vote for each of the following parties?”, including the Greens, on a scale from 0 (very unlikely) to 10 (very likely). Although this is not a perfect measure of possible future support, it allows respondents to indicate a level of openness to voting Green in future. The BES is a high quality survey of over 30,000 adults.

On this scale, 34% said it was ‘Very Unlikely’ they would ever vote Green, and a further 11.8% said they “Don’t Know”. This leaves 54.2% of respondents with some possibility of voting Green in future, 33.6% with an answer of 5 or higher and 4.1% giving the highest score of 10 (Very likely).

British Election Study (June 2020)

So, what sort of people are most likely to say they will vote Green in future? The graph below shows the average response from each demographic/political group. As one might guess, Green 2019 voters are the most likely to say they will vote Green in future, followed by Labour 2019 voters, age 18-24, mixed race, SNP 2019, white other (not British), Remain 2016, postgraduates and 2019 Liberal Democrats.

On the flip side, 2019 Conservative voters have the lowest average score for any group. This was a slight surprise for me, as I thought the Greens (at least in England and Wales) had some appeal among Conservatives. This group is followed by those with no qualifications, Leave 2016, Brexit Party 2019, age 65+.

The 3 point difference between Conservative and Labour 2019 voters should put to bed the idea that the Greens draw equally from the two parties.

Responses are also correlated with previous Green election results. I found that this correlation is strongest with 2019 European Election results (again, based on Chris Hanretty’s estimates here), probably due to the fact that the party had significantly higher support nationwide and was on the ballot in every constituency.

Coloured by 2019 General Election vote

Green Target Seats

So, given what we know about Green-curious voters, how should that transfer into target seats?

To try and estimate this, I use Multi-Level Regression and Poststratifictaion (MRP) to model the distribution of respondents in each constituency, based on census data. MRP allows you to model the effect of individual demographic data (in our case age, education and gender) as well as data about the constituency (using Pippa Norris’s combined election and census results data here). After estimating the responses from each demographic group, I multiply this by the number of people from that demographic group in the constituency to find a constituency-level prediction.

To improve the model’s ability to consider the effect of 2019 vote, I also use an auxiliary MRP model to estimate parties’ support among demographic groups at the 2019 election. This means that in poststratification, we can estimate the percentage of each party’s 2019 vote in a constituency that would respond in a certain way (from 0 – very unlikely to 10 – very likely).

Once the responses are predicted, to see how an increasing Green vote might look, I take the most likely Green voters from other parties (starting from 10, 9, 8 etc) and add it to the 2019 Green vote. The effect is to see where the Green vote might be expected to grow fastest, as well as the biggest losers from a Green increase in votes. For the purposes of this analysis I have not included 2019 non-voters.

For example, this is what happens in my home constituency of Oxford East if you take the most likely Green voters from other parties. Labour falls very quickly and when the Greens reach approximately 18.4% nationwide, they overtake Labour and win the seat.

This all assumes that the Greens are standing and campaigning in every GB constituency, which is unlikely, but the exercise is more about the geographic spread of potential Green voters than predicting the actual vote share the Greens would need to win more seats.

Leadership candidate Amelia Womack recently suggested the party should be aiming for 8 to 10 MPs by 2030 – so where might these be? Below are the 9 constituencies plus Brighton Pavilion where the ‘most likely’ voters need to switch for the seat to change hands:

ConstituencyGreen Nationwide Vote
Brighton Pavilion2.9% (current)
Bristol West9.2%
Sheffield Central15.3%
Hornsey and Wood Green16.3%
Hampstead and Kilburn18.1%
Dulwich and West Norwood18.3%

All 9 gains are currently held by the Labour Party. As expected, Bristol West still comes out top as the most likely future gain, but interestingly other areas for Green optimism have fallen much further down the list. Bury St Edmunds, where the Greens got 15.7% in 2019 as part of ‘Unite to Remain‘, is at 429. Stroud, where Molly Scott Cato ran in 2019, is at 166. The Isle of Wight, where the party had a surprisingly good result in 2017, is at 342.

The reason these areas are so hard to win is not that they do not have a significant number of ‘Green-curious’ voters, but that there are too many Conservatives. Although the percentage of Conservative 2019 voters who say they are “Very Unlikely” to consider the Greens varies massively (from 40% in Sheffield Central to 84% in Perth and North Perthshire), that number is still significantly higher than amongst Labour voters in every constituency. For somewhere like Stroud, even though Labour and Conservative voters are more likely to consider voting Green than their counterparts elsewhere, the size of the Conservative vote is difficult to overcome.

RankConstituencyGreen Nationwide Vote
78Cities of London and Westminster26.2%
131Finchley and Golders Green30.5%
158Sutton and Cheam32.2%
161Truro and Falmouth32.5%
162Carshalton and Wallington32.6%
173Chipping Barnet33.2%
Top 10 Target Seats for the Greens amongst Conservative-held seats

This where I should admit I was surprised – I knew the Greens drew disproportionately from Labour, but I thought there would be a significantly smaller gap between Labour and Conservatives in Remain areas, and a significantly bigger gap between Labour voters in Remain and Leave area. I previously said I thought the best strategy for the Greens would be to try get 2nd place in Conservative seats, but I now think this is wrong:

Despite the huge Labour majorities in many of these seats, voters are far far more likely to express an openness to voting Green in future. The first 25 most likely Green gains, in my model, are all held by Labour. The 26th is SNP (Glasgow Central), although the smaller sample size in Scotland means the Scottish seats might be more variable, and the 31st is Lib Dem (Bath). Based on this BES question, it seems likely that were the Green Party to win a majority (obviously not happening any time soon…) it would be on the basis of Labour losing every single MP.

The Risk to Labour

However, under FPTP the risk to Labour does not just come when the Greens win more seats. With the Greens drawing disproportionately from Labour in England and Wales, dozens of seats could be lost if the Greens gained just a few percentage points.

The clear beneficiaries of an increase in the Green vote is the Conservatives. If the most likely Green voters from other parties defected in 2019, so that the Greens were on 6% (around where they are currently polling), the Conservatives would gain 8 MPs, Labour would lose 9 MPs, and the Liberal Democrats would gain Sheffield Hallam from Labour and East Dunbartonshire from the SNP. If the Green vote was 10%, where YouGov had the party in a recent poll, the Greens would win Bristol West while the Conservatives gain 20 MPs, Labour lose 20, the Liberal Democrats net plus one, and the SNP lose two.

This depends on both the Greens standing and campaigning in every constituency and all other parties’ votes staying the same, which is obviously unrealistic, so this is not a prediction/projection. It does, however, illustrate the risk to Labour if the Greens continue to build support.

Labour voters, particularly young graduates, are much more volatile than before. Rather than a particular loyalty to Labour, they view a number of parties, especially the Greens, positively. They may have voted Green in the 2019 European Elections or the 2021 Local Elections, further straining any remaining loyalty to Labour. And they are more likely to view voting as expressive or transactional, rather than solely anti-Tory, expecting either representation of their values or a substantial policy offer from the party they support.

Labour cannot win the next election solely by winning back Labour to Conservative switchers, it also needs to hold together and turn out its fragile base.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Create a website or blog at

%d bloggers like this: