Gorton & Denton By-Election Explorer

An interactive scenario explorer for the Gorton & Denton by-election on 26 February 2026. It lets you explore how different turnout patterns, demographic mobilisation, and tactical voting could change the result. It is not a prediction.

No. By-elections are inherently unpredictable - they have low turnout, high protest voting, and unique local dynamics that models struggle with. The baseline is our best estimate given available data, but all scenarios are hypothetical. Actual results may differ substantially.

The baseline blends three sources:
1. Seat projection model: Our GB-wide demographic model estimates how national polling translates to local vote shares.
2. Constituency polling: Two local polls (Opinium and Omnisis) are blended with the model to ground the estimate in real local data.
3. By-election adjustments: We account for lower turnout, the incumbent party penalty, and the absence of the Workers Party candidate (10.3% in 2024) - their vote is redistributed based on the profile of WPB voters and the Muslim Vote endorsement of the Green candidate.

We split the constituency estimate across the 7 wards using Census 2021 demographics and blend with ward-level data from both polls. Ward estimates are indicative only - treat them as showing relative patterns between areas, not precise numbers.

Win probabilities come from 10,000 simulated elections that account for polling error, turnout uncertainty, and voter behaviour. When you move sliders, probabilities update to reflect the new scenario. They show how likely each party is to win given the current settings and the inherent uncertainty in the estimates.

It adjusts overall expected turnout (baseline: 38%, typical for a by-election). Higher turnout tends to boost Green and Reform, as the additional voters in a by-election are more likely to be protest voters. Lower turnout concentrates the vote among habitual voters where Labour is relatively stronger.

This constituency spans two council areas: the Manchester wards (Burnage, Gorton, Levenshulme, Longsight) and the Tameside wards (Denton NE, Denton S, Denton W). Green leads in most Manchester wards while Reform leads in all three Denton wards. The slider adjusts relative turnout between these areas - higher Tameside turnout means Reform's strongholds contribute more votes to the constituency total, shifting the overall result toward Reform.

These model what happens if specific demographic groups vote in higher or lower numbers than expected. Each slider has two effects: it shifts vote shares within each ward (e.g. more young voters turning out boosts Green), and it changes how much each ward contributes to the total. Effects are calibrated from both the local polling cross-tabs and national demographic patterns.

The baseline already includes tactical voting captured in both polls - some Labour and Lib Dem voters switching to Green, and some Conservative voters switching to Reform. The sliders let you model additional tactical switching on polling day beyond what's already in the data.

This shows the range of results we'd consider realistic, based on 10,000 simulated elections with different assumptions about polling error, turnout patterns, and voter behaviour. 9 in 10 simulations fell within this range. It does not update when you move sliders - it reflects baseline uncertainty only.

The dashed mark shows each party's GE2024 vote share for comparison. The solid bar shows the current scenario estimate. This makes it easy to see how much each party has gained or lost since the general election.

Each ward's weight is its share of the constituency total vote. At baseline this is based on GE2024 vote counts. When you adjust the area differential or demographic turnout sliders, ward weights shift - a ward whose target demographic turns out more heavily contributes more to the final result.