JohnC said:
Lab 493 59.1% (+16.1%)
Con 172 20.6% (+1.2%)
BNP 93 11.2% (+11.2%)
Ind 76 9.1% (-8.2%)
Green 0 (-8.3%)
Socialist Alliance 0 (-12.%)
Does that make sense? What's the point in including parties that didn't stand? And why show the BNP as +11.2% when in fact their vote dropped from the last time they stood by 4.7%? Or is that just from the last full cycle of council elections?
Good question. You are right to be confused as the comparisons in brackets are rubbish. Election Statistics can be manipulated and that is what has happened with the figures in brackets.
Explanation
The last full election was fought in 2003 when there were three seats up for grabs. Only Labour stood a full slate of three candidates all of whom won; Tories and Socialist Alliance stood only one candidate apiece, Greens two candidates and there were also two independent candidates. You cannot compare performance for a by-election with only one seat, with a multi-vote election for three seats, especially when parties stand different numbers of candidates.
For example, in 2003, the Socialist Alliance candidate got 183 votes, which is only 5.9% of the total votes cast of 3,119; the three Labour candidates got 1,896 votes or 60.8% of the votes cast (ten times the SA vote and almost the same proportion as in the recent by-election), while the two Greens totalled 236 votes (full results link below). However there is an established convention when calculating % swings in multi-member seats, to calculate the average vote for each party. This is the same as the % of votes when each party stands a full slate of candidates but can produce bizarre and misleading figures when they do not.
On this basis the SA got an 'average' of 183 votes for their one candidate, the two Greens averaged 118, the sole Tory got an average of 297, the three Labour candidates averaged 632 votes. Bizarely the two independents are lumped together, despite the fact that they might have completely contradictory platforms, with an average vote of 254 votes. This gives the percentages that have been used in the figures quoted. While there is a case for saying 12% of electors chose to put
one vote for the SA, their other
two votes could have gone anywhere - two Labour; two Green; one Labour, one Green; two independents; one independent, one Green; etc etc.
There is no reason to assume that in a one-seat by-election, the calculated "comparison" votes from 2003 - ie "12.3%" vote for the sole SA or "42.6%" for Labour etc, have any meaning at all. Barry B is therefore wrong to say that this was a "respectable vote of 12%" for the SA, as their one candidate was standing alongside two Greens, two independents and three Labour. and supporters of any one of those parties/independents could have cast a majority (two votes) of their votes for their 'preferred party' and the SA vote could represent a minor protest. This is different to getting 12% in single seat election or in a three-member seat where every party stands three candidates. In fact even 12% of the vote for the SA, far from being respectable, would have been well below the average vote for Respect in the May 2006 elections, where in 78 wards across England, Respect had an average vote of 19.6%, Respect got more than 12% in over two-thirds (53) of these wards.
There was a by-election for Skerton West earlier in the year (June 2006) for one of the other (Lab) seats in this ward, which provides a better basis for comparison. However even this has to be used with caution, as there was no independent in this by-election, only Lab, BNP, Tory.
However, this comparison would be:
Lab 493 59.1%
(-0.7%)
Con 172 20.6%
(-3.7%)
BNP 93 11.2%
(-4.7%)
Ind 76 9.1%
(+9.1%)
Sources:
Dec 2006 by-election:
http://committeeadmin.lancaster.gov.uk/mgElectionAreaResults.asp?XXR=0&ID=66&RPID=3700&J=7
June 2006 by-election
http://committeeadmin.lancaster.gov.uk/mgElectionAreaResults.asp?XXR=0&ID=30&RPID=3697&J=4
May 2003 Full Election
http://committeeadmin.lancaster.gov.uk/mgElectionAreaResults.asp?XXR=0&ID=57&RPID=3705&J=12
