How the model works

Vote source model, not uniform swing.Most local election forecasts apply uniform national swing — if Labour is down 15 points nationally, every ward loses 15 points. This breaks when a party goes from near-zero to 25% in four years. You can't swing from zero. Our model tracks where blocks of voters are moving. Using vote flow rates from the British Election Study Internet Panel — which follows 30,000 voters over time and records exactly which parties they switch between — the model estimates how many 2022 Conservative voters now support Reform, how many 2022 Labour voters have moved to the Greens, and how many have dropped out entirely. These empirically grounded flow rates replace assumptions about uniform movement.

The model applies these flows differently by council type. A London borough, a northern metropolitan borough, a rural county council, and a southern shire district are distinct political environments where the same national polling translates to very different local outcomes. We use five council profiles (London boroughs, metropolitan boroughs, county councils, unitary authorities, and shire districts) with BES-calibrated vote flow rates for each.

Ward-level geographic distribution

Within each council, party vote shares are not applied uniformly. Where 2022 ward-level results exist — covering roughly 75% of wards — the model uses the historical ward-level pattern as a distribution template. A ward where the Greens polled 40% in 2022 gets a larger projected Green share than one where they polled 5%, even when the council-wide average is the same. This preserves the geographic concentration that is critical to seat outcomes under first-past-the-post.

For wards without 2022 baselines — typically wards where a party is standing for the first time — the model applies a discounted estimate based on the council-level projection and, where available, 2024 general election constituency results mapped to ward boundaries. The concentration effect is dampened at 70% rather than fully preserved, allowing for some expansion of support beyond 2022 strongholds while preventing the model from being imprisoned by four-year-old data in a fast-moving political environment.

Reform UK: the geographic anchor problem

Reform contested almost no local wards before 2025. There is no 2022 ward-level baseline to distribute from. Instead, the model uses the general election constituency-level projections from our GE forecasting model as a geographic anchor, mapped to council wards via ward-constituency boundary overlaps. A ward in a constituency where the GE model projects Reform at 35% gets a higher local projection than one where it projects 15%.

This GE anchor is then adjusted: a local discount factor (0.70 in the central scenario) accounts for the established pattern that insurgent party support is lower in local elections than general elections, reflecting lower turnout among non-habitual voters and the absence of the national campaign spotlight. The discount factor varies across the three scenarios.

Multi-member ward allocation

London boroughs and all-up metropolitan boroughs elect three councillors per ward simultaneously. Voters can cast up to three votes, and the dominant pattern is slate voting — people vote for all three candidates from one party. This creates winner-takes-all amplification: the party with the plurality in a ward typically wins all three seats, not one-third of them.

The model replicates this using margin-based seat allocation calibrated against historical multi-member ward results. Where the leading party has a comfortable margin, it sweeps all seats. In close races between two parties, seats split two-to-one. This amplification effect is important: Labour's inner London strongholds deliver far more seats per ward than a proportional model would suggest, and Reform's northern victories are more decisive — winning a ward by even a small margin in Sunderland yields three seats, not one.

Monte Carlo simulation

The model runs 1,000 simulated elections per scenario, varying ward-level vote shares with random noise calibrated to historical election variance. This produces probability distributions rather than point estimates — we report not just “Reform projected 1,569 seats” but the probability that Reform wins the most seats across all simulations. This Monte Carlo approach captures the uncertainty that any single point estimate necessarily hides.

Three scenarios

Local election forecasting in a five-party fragmented system is harder than any previous election in English local government history. Rather than presenting a single projection with false precision, we publish three scenarios reflecting different assumptions about the key parameters.

Scenario A (Mild):BES-anchored parameters with minimal adjustment. Reform's local discount is higher (0.55), meaning their general election support translates less efficiently to local votes. Labour retention of 2022 voters is stronger. Under these parameters, Labour wins the most seats nationally with high probability — their London and metro strongholds hold up despite losses elsewhere.

Scenario B (Central):Parameters adjusted toward evidence from the 2025 local elections and subsequent by-elections, where Reform's conversion of polling support into council seats exceeded historical patterns. The Reform local discount is moderate (0.70) and Con-to-Reform switching rates are calibrated to observed 2025 patterns. This scenario shows Reform narrowly winning the most seats nationally, but with meaningful uncertainty.

Scenario C (Severe):Parameters aligned with the most aggressive external forecasts (particularly Stephen Fisher's projection of Reform +2,260 net gains). Reform's local discount is low (0.80), meaning their GE polling translates almost fully to local votes. This scenario produces a decisive Reform victory.

The truth is probably somewhere in the central scenario's range, but we cannot know until votes are counted. The scenarios exist to make this uncertainty explicit rather than hiding it behind a single number.

Coverage

The model covers all 136 English councils holding elections on 7 May 2026: 32 London boroughs, 32 metropolitan boroughs, 18 unitary authorities, 6 county councils, and 48 district councils. This represents 5,004 contested seats(matching the official count from Rallings and Thrasher within rounding). The projections incorporate confirmed candidate lists from Democracy Club's Statements of Persons Nominated, published 9 April 2026.

Key uncertainties

Reform's local conversion rateis the single largest source of uncertainty. Reform won 41% of seats with 31% of votes in 2025 — a seats-to-votes ratio that had previously only been achieved by established parties with deep local infrastructure. Whether this was a one-off breakthrough or a new normal is genuinely unknown. The three scenarios reflect this uncertainty directly.

Green geographic distributionin London remains difficult to model precisely. The Greens are standing in 90–100% of London wards for the first time, and their national polling surge represents unprecedented territory for a party with historically hyper-local support. Green seats in inner London boroughs such as Lewisham, Hackney, and Lambeth are the most confident Green projections; outer London and boroughs without established Green organisation carry more uncertainty.

Liberal Democrat incumbency effectsare structurally difficult to capture. Lib Dem support is the most geographically concentrated of any party — their councillors dramatically outperform their national polling in wards where they have a sitting incumbent, and dramatically underperform elsewhere. Lib Dem seat counts should be treated as potentially understated.

Independent councillors and local partiesaccount for roughly 3% of projected seats. Candidates with a personal vote can swing individual wards by 20 points, and no national model can see that coming. Tower Hamlets (where Aspire is a major force) is a specific council where the model's projection should be treated with caution.

Data sources

• Polling data: National Westminster voting intention polls from all BPC-registered pollsters, aggregated with recency weighting (10-day half-life), sample size weighting, and pollster accuracy corrections. Updated every 6 hours.
• Vote flow rates: British Election Study Internet Panel (May 2025 wave), providing empirical voter migration data between parties.
• Ward-level baselines: 2022 English local election results from the Open Council Data project, covering ward-level vote shares for all parties.
• Reform geographic anchor: General election constituency-level projections from the Reading Signal GE forecasting model.
• Candidacy data: Democracy Club Statements of Persons Nominated, confirming which parties are standing in each ward.
• Council reference data: Rallings and Thrasher (Exeter University) for seat counts and defending party allocations.
• External benchmarks: Stephen Fisher (Elections Etc, Oxford) for net gain projections; PollCheck for ward-level model comparison.

What changes after May 7

The local election results will provide the first large-scale calibration data for several model parameters that are currently held at assumed values: the Reform local discount factor can be directly measured from actual results; the BES vote flow rates can be validated against real ward-level vote shares; the slate voting amplification thresholds can be calibrated against actual multi-member ward outcomes; and the locals-to-GE reverse translator can be updated with fresh calibration data, improving the general election forecast.

We will publish a post-election calibration analysis comparing projections to results at council and ward level, documenting where the model was right, where it was wrong, and what parameters need adjustment.