Psephology 5: Is there a way to tell if a ward is marginal?

There are all sorts of factors involved in answering this question, and naturally, I'm nosy enough to explore the question. I'll look at a few of them, because some LibDem somewhere might find it useful. I've been looking on the Internet, and very few people seem to be doing this; playing around with the electoral data in various ways. I've seen stuff on the LSE site and Plymouth University, but few grassroots party supporters seem to be looking at their ward data. I know I'm a political anorak, but it's fun! Really! Anyway, Mark Pack's book says go to the data.

Firstly, which wards could you target your efforts towards?  How about the half dozen that have the highest historical voting percentage for the LibDems. There are no LibDem councillors in Sunderland at the moment, so if you're an optimist like me, the world is your oyster.

Ward	Average over time	Ward	Average over time
St Chad's	4.6%	St Michaels	12.5%
Castle	6.9%	Hetton	12.6%
Ryhope	7.2%	St Peters	12.9%
Copt Hill	7.5%	Sandhill	12.9%
Redhill	8.2%	Pallion	13.%
Southwick	10.0%	Washington North	13.6%
Doxford	10.4%	Barnes	14.9%
Fulwell	11.2%	Shiney Row	14.5%
St Anne's	11.5%	Washington South	16.1%
Washington East	11.8%	Hendon	16.2%
Silksworth	11.9%	Washington West	16.9%
Houghton	12.1%	Millfield	34.9%
Washington Central	12.4%

I've ranked the wards in low-high order of the average LibDem vote, 2004-2012, which gives a good idea of the wards with the highest support for the party. As might be expected, Millfield, historically our greatest success, has the best result. And wouldn't you know it, my ward is lowest of the lot! Most wards are trending down in 2012, but if you start from a higher position...

One could rank any of the main parties by average ward voting %, so this is the Conservative result:-

Ward	average for wards	Ward	average for wards
Castle	11.09%	Shiney Row	22.82%
Redhill	11.63%	Silksworth	23.98%
Houghton	12.53%	Wash Central	29.13%
Hetton	13.94%	Doxford	30.04%
Millfield	14.86%	Ryhope	30.91%
Wash North	15.02%	Wash South	31.93%
Wash West	17.58%	Wash East	33.96%
St. Anne's	18.24%	Barnes	39.91%
Copt Hill	18.88%	St. Peter's	43.37%
Pallion	20.71%	St. Chad's	47.23%
Hendon	21.23%	St.Michael's	51.94%
Southwick	21.91%	Fulwell	52.38%
Sandhill	22.44%

Conservatives do very well in Fulwell and St Michael's, which is exactly where they elected councillors in 2012. St Chad's has historically been a Conservative stronghold for many years. Conversely, Castle ward, with the lowest Tory vote, elected a Labour councillor with an 86.11% share of the vote! I actually know this guy, a popular local activist who's done a great deal for his area, one might say in spite of Labour factionalism, deselection and reselection after popular protest, etc. Oooh! Handbags!

http://www.sunderlandecho.com/news/local/all-news/labour-party-pledge-over-sunderland-city-council-candidates-selection-1-3857025

What about the Labour Party. Well, here goes:-

Ward	average % vote	Ward	average % vote
St.Michael's	28.6%	Pallion	53.8%
Fulwell	31.8%	Ryhope	54.1%
Barnes	34.5%	Southwick	55.1%
Millfield	37.4%	Shiney Row	57.3%
St. Peter's	38.3%	Washington West	57.8%
St. Chad's	44.0%	Silksworth	58.3%
Washington South	44.9%	St. Anne's	58.9%
Copt Hill	46.3%	Hetton	59.3%
Washington East	46.7%	Sandhill	61.7%
Doxford	49.4%	Castle	62.1%
Hendon	50.5%	Washington North	64.0%
Houghton	52.0%	Redhill	65.6%
Washington Central	53.2%

Interesting results. There's a smaller core Labour vote in some areas than you might expect. In Barnes there's an average Labour vote of  34.5%. In 2012 it was 55.5%, so are these newcomers swing voters or issue voters, or as I see it, opportunities. In most wards, the trend is the same, except for Copt Hill, Houghton and Hetton.

This highlights the perils of individualism - in Sunderland there are a number of strong local characters who win the popular vote through force of personality. For example, in Houghton and Copt Hill wards,(see psephology blog 4) the issue of the huge local landfill and its possible contamination of the water supply has dominated the minds of the electorate, and so we have single issue independent candidates who have been very successful in recent years (51% and 55% of the vote in 2012). You could say that there was little point in putting up a candidate against people who have so energised a worried electorate in their favour, unless you were strong and very vocal on Green issues.

Here are 3 different ways of ranking wards with reference to the Labour Party. (Average % vote; performance statistic showing whether ward is above or below city average vote%; increase/decrease in vote since last election). We can tabulate this data side by side for comparison.

Ward (ranked)	average % vote	Ward (ranked)	Performance +/- average	Ward (ranked)	2012 % increase of vote
St.Michael's	28.6	St Michaels	-34.2	Hetton	-17.4
Fulwell	31.8	Fulwell	-31.8	Houghton	8.8
Barnes	34.5	Barnes	-24.1	Copt Hill	9.6
Millfield	37.4	Millfield	-22.0	St Michaels	12.7
St. Peter's	38.3	St Peters	-20.4	W East	16.2
St. Chad's	44.0	Copt Hill	-6.3	Silksworth	18.0
W South	44.9	St Chad's	-6.0	St Chad's	18.8
Copt Hill	46.3	W South	-4.9	Fulwell	19.1
W East	46.7	W East	-4.5	Millfield	21.2
Doxford	49.4	Doxford	-3.5	Pallion	22.4
Hendon	50.5	Hendon	1.0	Southwick	22.9
Houghton	52.0	Houghton	3.6	Shiney Row	25.1
W Central	53.2	W Central	5.6	W Central	25.5
Pallion	53.8	Pallion	6.7	St Peters	25.7
Ryhope	54.1	Southwick	6.7	St Anne's	26.7
Southwick	55.1	Ryhope	7.2	W North	27.0
Shiney Row	57.3	Shiney Row	13.2	W South	27.5
W West	57.8	St Anne's	13.2	Redhill	29.1
Silksworth	58.3	W West	13.8	Hendon	30.2
St. Anne's	58.9	Silksworth	14.6	Barnes	31.1
Hetton	59.3	Hetton	16.4	W West	31.4
Sandhill	61.7	Sandhill	20.6	Doxford	32.1
Castle	62.1	W North	24.6	Castle	33.3
W North	64.0	Redhill	27.3	Ryhope	40.1
Redhill	65.6	Castle	28.2	Sandhill	43.4

This is an interesting measure of whether a political party is achieving well in a particular ward. The LibDem stronghold, Millfield, is a low achieving ward for Labour whichever measure you use. As expected, the Independent and Conservative wards also give a poor score for them. Unexpectedly, Washington East, South and Barnes are also in the top ten potential targets, although Barnes' rising  vote % shows that Labour are targeting this ward. If you collate all three ways of ranking a ward, some interesting results emerge: In order, the hardest wards for Labour to win are:-

1. St Michael's

2. Fulwell

3. Millfield !!!

4. Copt Hill

5. St Chad's

So in conclusion, to find vulnerable wards, we can use data from:-

1. Average and last voting percentages, and their trends.

2. Ward/candidate performance statistics (whether they are above or below the average for the city).

3. Turnout data (more about this next time).

4. The historical favourite, percentage increase in the vote since the last election.

5. Data from ward surveys etc. I haven't got access to this.
6. Study the candidates. If I can find any promotional speeches from the last election I can do some discourse analysis.

7. I would have said the ward data relating to PCC elections which is the latest voting data but Sunderland council seems remarkably closed mouthed about this. Perhaps they didn't save the data. One has to suppose they've at least seen it when compiling the results, but they won't tell me. The word obtuse comes to mind.

Correspondence:-

I refer to your recent request for information under the Freedom of Information Act 2000.  I can confirm that the Council does hold this information, as detailed below.

Please can you tell me the results of the Police and Crime Commissioner Election by ward, for the 25 Sunderland wards. I wish to know: the number of registered voters in each ward at the time of the election. - The vote count for each candidate in each of the 25 wards. - The percentage turnout in each ward.

This information is exempt from disclosure under Section 21 of the Act - Information Accessible to Applicant by Other Means.  The information you have requested is available online via the Council's website on the following link http://www.sunderland.gov.uk/index.aspx?articleid=7188 .

Please note that the Police and Crime Commissioner Elections were conducted on police force areas and the Northumbria police area comprised of the following local authority voting areas Gateshead, Newcastle, Northumberland, North Tyneside, South Tyneside and Sunderland.  Each local authority was required to conduct their own election which included the verification and count. 

Except that this isn't the ward data!!! It's just the same old whole city data. I could do a lot with that information; test my predictions for one thing, but it isn't to be. It doesn't sit well with my LibDem sense of fairness.

Next blog on the usefulness of Turnout data, how many people you need to win, and what percentage of the electorate are  politically active.