Results and turnout at the May 2017 Wales local elections

Introduction

This report provides analysis on the participation and administration of the local government elections in Wales, held on 4 May 2017. For this, data was collected from Returning Officers and Electoral Registration Officers in Wales. This comprised the Form K ‘Statement as to Postal Ballot Papers’ and an additional data form, which included data relating to electoral registration, turnout, absent voting and rejected ballots.

Local elections were scheduled in 852 wards across all 22 local authority areas in Wales. Of the 1,254 seats available, 1,161 were contested and 92 candidates were returned unopposed. Elections were postponed in two wards and one seat in Powys had no candidates standing for election.1

Electorate  2.3 million
Turnout 

Ballot box: 42.0% (895,943 votes)

Postal vote: 69.7%

In-person: 36.3%
Rejected ballots

Ballots at the count: 0.5%

Postal votes: 3.2%
Proxies appointed

2,873 (0.1% of the electorate)

167 emergency proxies


 

Electorate

The elections gave over 2.1 million registered electors the opportunity to vote.2 Since December 2016, this is an increase in electorate size of 1.0% across Wales, ranging from an increase of 4.2% in Flintshire to a decrease of 3.2% in Wrexham.

More than 20,000 electors (0.94% of the electorate in contested wards) were added to the electoral register during the weeks leading up to the election. As a proportion of the electorate, this was highest in Gwynedd, where additions represented 2.6% of the electorate in contested wards, and lowest in Rhondda Cyon Taf, where additions represented 0.3% of the electorate in contested wards.

A total of 25,587 applications to register were received in the weeks leading up to the election.3 Almost a quarter of these (22.5%) were recorded as duplicates. At a local authority level, duplicates were reportedly as high as 43% in Denbighshire and 41.4% in Swansea.

In addition, 31,003 individuals applied to register to vote after the registration deadline. This is likely to be attributable to the announcement of the general election, which came after the deadline for registration for the council elections. Based on data from 14 local authorities, 187 individuals were recorded as trying to vote on election day despite not being registered.4

Votes were cast at 2,254 polling stations as well as by post.
 

Turnout

The measure of turnout referred to in this report, ‘ballot box turnout’, includes all valid votes cast (‘valid vote turnout’) and votes rejected at the count. ‘Total turnout’ refers to valid votes cast, votes rejected at the count, and those rejected at the postal ballot verification stage before the count. Table 3.1 confirms that only a small fraction of voters try to vote but fail to do so successfully.

Table 3.1: Turnout

  Electorate Valid vote turnout Ballot box turnout Total turnout Difference
valid vote –
total (pp)
2017 2.28 m 41.8% 42.0% 42.4% 0.6
2012/135 2.31 m 38.7% 38.9% 39.4% 0.7
2008 2.11 m 43.7% 44.0% 44.5% 0.8

Ballot box turnout was 42.0%, which is higher than the 38.9% turnout at the 2012/2013 local government elections in Wales but still lower than the 2008 elections. 

Turnout ranged from 36.3% in Caerphilly and Newport to 53.0% in Ceredigion, and increased across all local authorities with the exception of Anglesey. At a ward level, turnout ranged from 17.2% in Treforest (Rhondda Cynon Taff) to 69.6% in Bethel (Gwynedd).

Table 3.2: Turnout 2012/2013 – 2017

Local authority 2012/13 2017 Change (pp)
Blaenau Gwent 37.7% 40.0% 2.3
Bridgend 34.4% 39.9% 5.5
Caerphilly 36.0% 36.3% 0.3
Cardiff 38.0% 42.8% 4.8
Carmarthenshire 44.5% 46.8% 2.3
Ceredigion 46.8% 53.0% 6.2
Conwy 38.0% 41.5% 3.5
Denbighshire 39.1% 42.4% 3.3
Flintshire 36.9% 38.5% 1.6
Gwynedd 48.3% 52.1% 3.8
Isle of Anglesey 50.5% 45.9% -4.6
Merthyr Tydfil 37.3% 37.7% 0.4
Monmouthshire 38.1% 46.7% 8.6
Neath Port Talbot 39.4% 42.3% 2.9
Newport 33.9% 36.3% 2.4
Pembrokeshire 45.8% 47.5% 1.7
Powys 46.7% 47.2% 0.5
Rhondda Cynon Taf 36.1% 39.9% 3.9
Swansea 36.5% 38.1% 1.7
Torfaen 37.2% 37.5% 0.4
Vale of Glamorgan 39.3% 46.1% 6.8
Wrexham 34.2% 40.0% 5.9


 

Rejected ballot papers

The proportion of ballot papers rejected at the count was 0.5%. This compares with 0.6% in the 2012/2013 local government elections.

At a local authority level, this ranged from 0.2% in Newport to 0.8% in Neath Port Talbot. At a ward level, two wards rejected more than 3% of ballots at the count and 52 wards rejected none.

The vast majority of ballots (70.6%) were rejected for reasons of being unmarked or wholly void for uncertainty. More than a quarter (26.5%) were rejected due to voting for more than one candidate. 
 

Postal voting 

The total number of postal votes issued for these elections was 387,909, representing 18.2% of all electors with a contested election in their ward. This compares with 17.4% at the 2012/2013 elections. At a ward level, this ranged from 6.3% in Cathays (Cardiff) and Treforest (Rhondda Cynon Taff) to 31.5% in Llandrindod East/Llandrindod West (Powys).

The proportion of postal voters returning their ballot papers always exceeds the turnout among ‘in person’ voters. This year, 69.7% postal electors used their postal vote compared with 36.3% who turned up to vote in person.

Table 4.1: Postal vote turnout vs in-person turnout

  Postal vote In-person
2017 69.7% 36.3%
2012/13 68.3% 33.3%
2008 71.4% 40.0%

Postal votes accounted for 29.2% of all votes included at the count. This compares with 29.3% in 2012/13 and 22.2% in 2008.
 

Rejected postal ballots

Postal voting packs require voters to provide their signature and date of birth. These identifiers are then matched against those provided at the time of application. If the signature or date of birth does not match, the postal vote is rejected and is not included at the count.

Since 2014, Electoral Registration Officers (EROs) have been required to notify electors if their postal vote has been rejected. They can also request the elector to provide an up-to-date signature. 

Table 4.2: Rejected postal 

Year Rejected ballots as % of envelopes returned
2017 3.2%
2012/13 4.4%
2008 4.8%

A total of 8,695 postal votes were rejected by Returning Officers across Wales. This represents a rejection rate of 3.2%, down from 4.4% in 2012/13 and 4.8% in 2008. The fall in the levels of rejected postal votes suggests that the new policy may be having a positive impact.

The most common reason for postal vote rejection, accounting for 43.3% of rejected postal ballots, is mismatched information. In almost a third of cases of rejection (31.8%), voters returned their postal voting envelopes but failed to include either the ballot paper itself or the verification statement or both.

Table 4.3: Reasons for postal vote statement rejection

    2017 2012/13 Change (pp)
Missing information Signature 9.4% 5.2% 4.3
Date of birth 9.6% 4.3% 5.3
Both 5.9% 15.7% -9.8
Mismatched information Signature 14.5% 19.1% -4.6
Date of birth 22.5% 7.1% 15.4
Both 6.4% 8.5% -2.1
Missing forms Ballot paper 14.4% 13.4% 0.9
PV statement 17.4% 26.8% -9.4

Proxies and waivers

The number of electors in Wales who appointed a proxy was 2,873 (0.1% of the electorate). This is consistent with 2012/13 and 2008.
The number of emergency proxies issued was 167. This is more than five times higher than in 2012/13 when 30 emergency proxies were issued.

A concession granted under the terms of the EAA 2006 was that postal electors who either had a disability, or were illiterate, or were unable to furnish a consistent signature could apply for a waiver to use their date of birth as their sole identifier.

A total of 2,717 voters were granted such a waiver for this election, representing 0.7% of postal electors. 
 

There remain inconsistencies in the ways in which local authorities record and report information.

We continue to notice differences in the coding of information by different electoral management software. For example, customers of one
electoral management software supplier consistently report a higher proportion of ‘mismatching’ than others.

Inconsistencies relating to the reasons for and total number of postal vote rejections on the Form K appear to result from the potential for differences in interpretation and treatment. For example, in the treatment of the numbers of covering envelopes and ballot papers returned, covering envelopes may be sent in without the A envelope or postal voting statement enclosed, while the missing document may or may not be sent in a separate covering envelope later, or multiple ballots may be returned in one envelope.

When local authorities are contacted about such anomalies they are often unable to provide revised figures or clarify why the data were coded in
that way. In practice, we use a calculation of field B6 minus field C18 as a surrogate for the total number of postal votes rejected regardless of whether or not it is the same as recorded in field C19.

The different breakdown of reasons for rejection collected on the additional data form does not always match this B6 minus C18 calculation. However, the discrepancies seem less severe and the categories have the advantage of being embedded in software and of having greater ‘common sense’ meaning. It would seem sensible to consider replacing fields B15-17 on Form K with three other aggregated categories: rejections for mismatching; rejections for missing identifier information; rejections for absent documentation. As it is, field B15 is largely redundant given that almost all authorities now verify 100% of postal vote returns.

There is no field that captures the number of postal voting statements received by the Returning Officer or at a polling station before the close of
poll. In practice, we use field B6, ‘Number of covering envelopes received by the Returning Officer or at a polling station before the close of poll’ as a surrogate but we know that, as mentioned, electors can return multiple postal ballots in one envelope or return envelopes without any ballots. 

Form K: STATEMENT AS TO POSTAL BALLOT PAPERS
LOCAL GOVERNMENT ELECTION
A. Issue of postal ballot papers
1. Total number of postal ballot papers issued under regulation 71
2. Total number of postal ballot papers issued under regulation 77 (spoilt and returned for cancellation), regulation 78 (lost or not received) and regulation 78A (cancelled due to change of address)
3. Total number of postal ballot papers cancelled under regulation 86A (where the first ballot paper was cancelled and retrieved)
4. Total number of postal ballot papers issued (1 to 3)
5. Total number of ballot papers cancelled under regulation 78A
B (1). Receipt of and replacement postal ballot papers
6. Number of covering envelopes received by the Returning Officer or at a polling station before the close of poll (excluding any undelivered or returned under regulation 77(1) (spoilt), regulation 78(1) (lost) and regulation 86A (cancelled ballot papers)
7. Number of covering envelopes received by the returning officer after the close of poll, excluding any returned as undelivered
8. Number of postal ballot papers returned spoilt for cancellation in time for another ballot paper to be issued
9. Number of postal ballot papers identified as lost or not received in time for another ballot paper to be issued
10. Number of ballot papers cancelled and retrieved in time for another ballot paper to be issued
11.Number of postal ballot papers returned as spoilt too late for another
ballot paper to be issued
12.Number of covering envelopes returned as undelivered
(up to the 25th day after the date of poll)
13. Number of covering envelopes not received by the Returning Officer
(by the 25th day after the date of poll)
14. Total numbers 6 to 13 (this should be the same as that in 4 above)
B (2). Receipt of postal ballot papers – Personal Identifiers
15. Number of covering envelopes set aside for the verification of personal identifiers on postal voting statements
16. Number of postal voting statements subject to verification procedure
rejected as not completed (excluding prior cancellations)
17. Number of postal voting statements rejected following verification procedures due to the personal identifiers on the postal voting statement not matching those in the personal identifiers record (excluding prior cancellations)
C. Count of postal ballot papers
18. Number of ballot papers returned by postal voters which were included in the count of ballot papers
19. Number of cases in which a covering envelope or its contents were
marked “Rejected” (cancellations under regulations 77, 78,78A and 86A are not rejections and should be included in items 2, 3, 5, 8, 9 and 10 above)

 

Additional data form
1) How many proxies were appointed for these elections?
2) How many emergency proxies were appointed for these elections?
3) How many waivers were granted for these elections?
4) How many postal votes were rejected for:
a) Want of a signature
b) Want of a date of birth
c) Want of both
d) Mismatched signature
e) Mismatched DoB
f) Both mismatched
g) Ballot paper unreturned
h) Postal voting statement unreturned
5) What was the total number of polling stations used?
6) How many covering envelopes were returned on polling day before 10pm?
7) How many covering envelopes were returned on the day after polling day before 10pm?
8) Total number of new electors added to the register after the publication of the revised register up to and including those added via the first interim notice of alteration
9) Total number of new electors added to the register via the second and final notices of alteration
10) Total number of applications to register received after the registration deadline.
11) Total number of applications received between the last date for applications to be included on the December 2016 revised register and the last date for applications for the first interim notice of alteration
12) Total number of duplicate applications received between the last date for applications to be included on the December 2016 revised register and the registration deadline for the first interim notice of alteration
13) Total number of applications received between the day after the last date for applications for the first interim notice of alteration and last date for applications for the final notice of alteration
14) Total number of duplicate applications received between the day after the last date for applications for the first notice of alteration and last date for applications for the final notice of alteration
15) How many people tried to vote on polling day and were found not to be registered?