Location map of Southampton Test on the UK map

General election forecast

Southampton Test

Predicted winner: Reform UK with 28.1% of the vote, a 0.3pp margin over Labour.

england · south east

Predicted vote share

How the model expects this constituency to vote

Reform UK
28.1% +13.4pp
Labour
27.9% +27.9pp
Green Party
16.3% +6.3pp
Conservative
14.4% +14.4pp
Lib Dem
9.3% +0.2pp
Workers Party
2.0% -0.2pp
Trade Unionist and Socialist Coalition
1.0% -0.0pp
SNP
0.7% +0.7pp
Plaid Cymru
0.3% +0.3pp
Other
0.0% 0.0pp

Coloured number is the swing from the 2024 general election result.

Sitting MP

Satvir Kaur

See also: TheyWorkForYou

2024 general election result

What happened on 4 July 2024

CandidatePartyVotesShare
Satvir Kaur Labour Party 15,945 44.5%
Ben Burcombe-Filer Conservative and Unionist Party 6,612 18.5%
John Peter Edwards Reform UK 5,261 14.7%
Katherine Jane Barbour Green Party 3,594 10.0%
Thomas Gravatt Lib Dem 3,252 9.1%
Wajahat Shaukat Workers Party 775 2.2%
Maggie Fricker Trade Unionist and Socialist Coalition 366 1.0%

Source: Returning officer declaration

How accurate is the model here?

Our 2024 backtest for this constituency

We replayed the same model architecture against the 4 July 2024 general election using only the data a forecaster running the model in May 2024 would have known. Here's what it would have predicted for Southampton Test, and what actually happened.

✓ Winner called correctly , predicted Labour, actual Labour. Major-party MAE: 3.11pp, the average gap between our predicted vote share and the actual share, per major party.

Predicted

Labour
52.7%
Conservative
17.3%
Reform UK
13.6%
Green Party
6.3%
Lib Dem
7.8%
Workers Party
0.5%
Trade Unionist and Socialist Coalition
0.0%
Independent
1.3%

Actual on 4 July 2024

Labour
44.5% +8.2pp
Conservative
18.5% -1.2pp
Reform UK
14.7% -1.1pp
Green Party
10.0% -3.7pp
Lib Dem
9.1% -1.3pp
Workers Party
2.2% -1.6pp
Trade Unionist and Socialist Coalition
1.0% -1.0pp
Independent
0.0% +1.3pp

Positive gap = we predicted that party would win more share than they did. See the full GE backtest across all 650 seats for the model's national accuracy.

How this forecast was built
  1. GE2024 Baseline. Actual GE2024 result in Southampton Test
  2. National Swing (Strong Transition Model). Multiplicative bounded swing, losers shed in proportion to local share, gainers absorb pro-rata national gain. Replaces additive UNS.
  3. BES MRP Prior. Baseline blended with BES Wave 1-30 south_east prior at weight 5%
  4. English-identity floor. English 24.5% → Reform +1.0pp
  5. Normalise. Shares scaled to 100%

Read the full general election methodology →