This page shows the detail of the calculations performed to estimate the general election result for 2010 had the new boundaries for Washington and Sunderland West been in force at that time. The basic idea of the calculation is to look at the district council wards which make up the new seat, and estimate how they voted in 2010. This estimate is based on the recent local election results in those wards, with adjustments made to allow for different turnouts and different voting patterns for local and general elections.
Due to changes in ward boundaries, some new wards are divided among several old constituencies (and sometimes in Scotland between two new constituencies). Such wards are shown divided into several parts, called , , and so on.
Election results from a recent local election are given. This is usually the local election closest to 2010 from the period 2007-2011. For multiple-member wards, the votes shown are the sum of the votes cast for all candidates of each party. A negative number indicates candidate(s) elected unopposed. Where zero votes are shown, the ward results have not yet been acquired, and the implied result is still provisional and will be improved later.
There are two problems with using the raw (actual) local election results to imply general election results ward-by-ward. Firstly, turnout can be different between local and general elections, which means that the total number of local votes cast does not equal the number of votes cast in each old seat at the general election. To correct for this, we adjust the local votes to match the general election turnout. Each ward's result is scaled, whilst keeping constant the percentage support for each party, so that the total turnout adds up to the old seat's general election turnout. All wards in the old seat are assumed to have the same percentage turnout.
Secondly, even though we now have the correct number of votes in the old seat, the party totals will not match the general election result. Some parties do better at the local election than the general election, and others do vice versa. We call these strong and weak parties respectively. The next step is to transfer votes from strong parties to weak parties. This is done by taking votes proportionally away from strong parties and putting them in a transfer pool. For instance if a party got 12,500 local votes, but only 10,000 general election votes, it will lose 20% of its votes from each ward. These votes in each ward will be put in the ward's transfer pool and allocated to the weak parties. Weak parties will get votes transferred to them in proportion to the votes they need over the old seat. For instance, if two weak parties need 4,000 and 1,000 votes respectively, the first party will get 80% of each ward's transfer pool and the second party will get 20% of each ward's transfer pool. The transfer of votes between parties is done on a ward-by-ward basis using these common percentages.
More details of the calculation formulas are available.
The old seat(s) needed are:
|New seat: Washington and Sunderland West|
And these are the implied results for 2010 for the new seat of Washington and Sunderland West.