Friday, 30 August 2013

By-Election Results August 2013

No. of Candidates
+/- July
Average/ contest
+/- July


Plaid Cymru**


* There were no by-elections in Scotland.
** There was only one by-election in Wales, which was defended and held by Plaid Cymru.
*** There were no Independent clashes in any of this month's contests.
**** 'Other' this month consisted of Independent Community and Health Concern (321 votes), English Democrats (twice - 72 and 98 votes), United People's Party (28 votes), and Putting Hartlepool First (194 votes).

Overall, 27,576 votes were cast over 18 individual local (tier one and tier two) authority contests. All percentages are rounded to the nearest single decimal place. For comparison see July's results here.

I have altered the table's attributes slightly so it now measures the percentage change in total vote since the previous month and got rid of the confusing holds/win categories. Seat change can capture the overall state of play better in just a single column.

As noted previously, there are considerable variations month-on-month in by-elections. One month a disproportionate chunk of contests can mushroom in a Labour area and the next find themselves clustered in leafy Tory shires. Therefore huge swings in total and average votes, as seen in pretty much all the parties except for Labour this month, are more likely to reflect geographical peculiarities than real shifts in opinion among the electorate. However, where there is comparatively little swing in a party's vote over a period of time then there is a good chance it is indicative of a mood in the wider population.

Over the period of a quarter the accidents of geography stand a good chance of evening themselves out and representing an accurate - and statistically significant - snapshot of the voting behaviour of the hardest of hardcore voters. This observation however is only limited to quarters three, four, and one in a single year. As elections generally tend to take place in quarter two parties have the tendency to roll up their by-elections to coincide with them. Look at the huge numbers of by-elections that took place at the same time as May's County Council elections. As the counties generally favour the Conservatives this fed through to a large preponderance of elections in Tory-held seats and therefore a "skewing" of the entire quarter's results. I expect the quarter two skew will be tilted toward Labour in 2014 as a number of City Councils are due up.


Gary Elsby said...

No. of candidates means what?

Gary Elsby said...

The chart must be wrong and therefore may give a false (and maybe) advantage to labour.

Phil said...

You're such a side splitter, Gary.

You are of course welcome to peruse the local by-election charts I've been compiling this year. Labour's results, apart from Quarter 2 (as explained above) are more or less in line with poll results. Remember, these are actual votes cast by the hardest of hard core voters in real elections. If Labour's results were skewed in anyway it would show up in wild fluctuations from month to month. As this isn't the case it suggests the results it's getting are approaching the real proportion of people prepared to vote Labour.

67 said...

You do things that may not appear odd at all to you but I don't quite get the chart.

Are you saying that there were 18 local elections and the Tories fielded 18 candidates?

Just answer the questions Phil.

Phil said...

Got it in one. Not difficult.

Gary elsby said...

For fear of being laughed at:

Does the chart show Labour fielding 17 candidates in 18 elections?

If so, please explain this phenomenon.

Phil said...

It does. Sometimes Labour has difficulty finding candidates to stand in true blue Tory country. The Tories sometimes have the same problem when it comes to dyed-in-wool Labour wards and constituencies.

Gary Elsby said...

You would of course like to inform us where this phenomenon took place, Labour not fielding a candidate.

NI possibly, but your chart does not show any of that.

My original point was that your column was to mean what, No. of candidates fielded or No. of winning candidates.
I struggle with the thought of Labour not fielding a candidate (other than NI) in any election as is their duty to do so.
If not, what was Labour's advice?

The % can be skewed if input is not weighted properly.

Phil said...

Perhaps you have problems looking at tables - it's straight forward for someone who has familiarity with GCSE maths and science provides.

How don't know how you can get any plainer than number of candidates = number of candidates who stood in the elections. I would have thought seats with no change (n/c), minuses and pluses would also be self-explanatory for anyone who follows election results. It would appear not.

As for why Labour didn't stand in a particular seat, I don't really know why. It was most likely in a place where party organisation does not exist. It does happen.

Gary Elsby said...

My GSCE maths, science and English (failed) requires the election detail so that I can research it.

What's so hard about it?
You produced a chart based on if FFS!

Phil, we scratched 20 to stand in 20 seats and that was hard work.

We know you then managed to get 44 in one go!

Phil said...

Do you even reread what you write before you click 'publish your comment'. I guess the answer to that would be 'no'.

Gary Elsby said...

Left wing bloggers bulling up Labour at everyone else's expense is not novel by any standards.

Phil said...

I cannot remember. Why don't you find out yourself? I have better things to do than run errands for a clown.

Gary Elsby said...

We thought so all along.