Publication - Research and analysis

Electoral reform: consultation analysis

Published: 12 Oct 2018

Analysis of Electoral Reform consultation.

Electoral reform: consultation analysis
5. Ballot papers (Q8 to Q11)

5. Ballot papers (Q8 to Q11)

5.1 The consultation paper discussed two issues related to ballot papers: the inclusion of candidate addresses on local government ballot papers (Chapter 3, Section 3 of the consultation paper), and the list ordering of candidates (Chapter 3, Section 4). This chapter looks at respondent views on both of these issues.

Including candidate addresses on ballot papers (Q8 and Q9)

5.2 The consultation paper put forward proposals for removing the current requirement (as set out in the Scottish Local Government Elections Order 2011) for candidate addresses to appear on local government ballot papers and electoral notices. Returning officers (ROs) would instead be responsible for ensuring candidates include an address within the local authority area on nomination papers, and for publishing a statement to this effect. It was noted that candidate addresses were no longer included on ballot papers for other elections in Scotland (to assist with ballot paper design in the case of Scottish Parliament elections, and to protect candidate security in the case of UK Parliament elections). Thus, the proposed change would be consistent with practice in other elections. Views were sought as follows:

Question 8: Do you agree that candidates’ addresses should not be required to appear on ballot papers for local government elections? [Yes / No]

Question 9: Do you have any other comments to make on this issue?

5.3 Altogether, 741 respondents answered the tick-box part of Question 8. Table 5.1 shows that two-thirds of respondents (64%) agreed that candidate addresses should not be required to appear on ballot papers for local government elections, while a third disagreed (36%). Organisations were more likely than individuals to agree (85% compared to 63%).

Table 5.1: Q8 – Do you agree that candidates’ addresses should not be required to appear on ballot papers for local government elections?

Respondent type Yes No Total
n % n % n %
Organisations 22 85% 4 15% 26 100%
Individuals 453 63% 262 37% 715 100%
Total 475 64% 266 36% 741 100%

5.4 A total of 259 respondents – 32 organisations and 227 individuals – made comments at Questions 8 and / or Question 9. The following sections look at views on the inclusion or exclusion of address information on ballot papers, before looking at related views on qualification criteria for candidates.

5.5 In considering the analysis presented, two points should be noted:

  • Many respondents discussed their views in a general way, commenting on the inclusion of address information on all types of ballot papers, rather than on local election ballot papers in particular. Comments also suggested that not everyone appreciated that there were different rules for different elections.
  • In a relatively small number of cases, individuals who answered ‘yes’ at Question 8 went on to give comments which suggested that they thought addresses should be included on ballot papers, and individuals who answered ‘no’ gave comments which suggested that they did not think that addresses should be included. This situation may have arisen as a result of the question wording (i.e. asking people if they agree that something should not be done). These responses to the tick-box question have not been altered in the analysis.

5.6 The comments provided by respondents indicated three main views on whether candidate addresses should appear on ballot papers: some thought they should be included, some thought they should be excluded, while a third group favoured the inclusion of some limited information to indicate place of residence. This third group of respondents included both those who answered ‘yes’ and ‘no’ at Question 8. This section explores views related to each of the three broad positions.

Views in favour of the inclusion of candidate addresses

5.7 Those who thought candidate addresses should continue to be shown on ballot papers for local elections generally thought that this was important for (local) democracy and accountability, and highlighted the following:

  • A local address demonstrated a connection, and a commitment to an area – respondents thought that this was important to voters and allowed them to make an informed choice in casting their vote.
  • People had a right to know about the background of those standing for public office, and candidates should be willing to be open about their circumstances.
  • Address information was important in allowing people to make contact with candidates or elected representatives.

5.8 For these reasons, respondents thought it was in the public interest, and in the interests of voters, for address information to be included on ballot papers. Some were clear that this should be a home address; others suggested that a work or office address would suffice.

5.9 Some also suggested the inclusion of address information stopped non-local candidates being ‘parachuted’ into an area. In such cases, respondents may not have been fully aware that candidates in local elections must live (or work) within the area they wish to represent, or they may have been expressing concerns about this practice within the context of national elections.

5.10 Other arguments put forward less often by respondents focused on the following:

  • Security of candidates: Some respondents stated that they were not persuaded by the argument that including address information posed a risk to the personal safety of candidates. Such respondents said: (i) they were unconvinced about the presence of a risk or did not think evidence of this had been provided, or they thought that any risk was minimal and outweighed by the public interest in making address information available; (ii) that the move was of limited value in protecting individuals as there were many other ways that people could find addresses; or (iii) that any threat to personal security could be dealt with adequately by the police.
  • Differentiating between candidates: Respondents thought that including address information was helpful in differentiating between candidates with similar / same names, especially with regard to independent candidates.

5.11 There was also a view that address information should be included on ballot papers but should not be made public in any other way.

Views in favour of excluding candidate addresses

5.12 Those who thought addresses should not be shown on ballot papers for local elections most often saw this as an issue of privacy and personal safety. They thought that the removal of address information was important to protect the privacy of candidates and their families, and to protect them from harm and abuse. Several respondents recounted experiences or concerns of candidates (potential and actual) linked to the publication of addresses. It was suggested that this was a particular issue in local elections where candidates were likely to live in close proximity to voters.

5.13 This group of respondents also put forward the following arguments:

  • The current requirement regarding the inclusion of address information represented a barrier to seeking election for some people, including those from groups who were vulnerable to abuse, and may, therefore, have particular concerns about making their home address available on ballot papers (e.g. women, ethnic minorities, those with disabilities, or those from LGBTI communities).
  • The inclusion of candidate addresses served no useful purpose in facilitating contact between voters and candidates / elected representatives. Contact could be made via other routes – e.g. party or constituency offices and ROs – and successful candidates could make contact details available after the election.
  • A local address did not necessarily make an individual a good representative for an area. It was suggested that it was more important to get to know candidates, and that campaign literature could be used to demonstrate a local connection or a commitment to an area. Some argued that there was no evidence of voters being influenced by the inclusion of addresses on ballot papers (and thus there was no reason to include this information), while others expressed concern that voters may be influenced by such information and wished to avoid that.
  • Address information was only needed to confirm qualification to stand and including it on ballot papers represented an unnecessary sharing of personal information (and was thus contrary to good practice, and to the ‘spirit’ of data protection legislation).

5.14 Respondents also thought that removing addresses from ballot papers offered the benefit of consistency with procedure in other elections (some in favour of including address information also argued for consistency across elections); it was also suggested that excluding addresses may offer some scope for reducing the size of ballot papers.

5.15 However, respondents also noted that a move to exclude candidate addresses from ballot papers would have some implications for other aspects of electoral procedures which would have to be given further thought. These included:

  • The need to find alternative ways of distinguishing between candidates with identical or similar names – the use of titles, photos or the inclusion of brief biographical details were suggested; some thought decisions about this should be made by the local RO
  • The need to consider the implications for including address information (of candidates, agents, and candidates acting as their own agent) on other electoral documents, and the extent to which that was available for others to see – specific reference was made to the publication of the ‘notice of election agents appointed’ which includes agent addresses.

5.16 In a few cases respondents who agreed that candidate addresses should not appear on ballot papers nevertheless acknowledged they had some mixed views on the issue:

  • Some thought that including addresses on ballot papers did serve a useful purpose (as discussed by those who supported its retention), but, on balance, believed that candidate safety should be the priority.
  • Some thought that it was right that addresses should not be on ballot papers, but also acknowledged that, in practice, such information was usually widely available or easily obtained anyway and thus removing addresses from ballot papers would not solve the issue of candidate safety.

5.17 Finally, one organisation queried whether inclusion of addresses was actually required under the current law and suggested that clarification was needed on this point.

Inclusion of limited address information

5.18 A substantial group of respondents indicated support for providing some limited residence information for candidates on ballot papers. This group thought it was in the public interest to provide such information as this was relevant in judging a candidate’s connection and commitment to a local area, with some thinking this was relevant to all elections (local and national). However, they also thought that providing full address details was not necessary and / or that privacy and personal safety should take priority. Instead, these respondents suggested other ways of indicating residency on ballot papers, for example, by including:

  • Partial address information such as (part) postcode or ward / area / town of residence (the latter being appropriate in parliamentary elections with larger constituencies)
  • An individual statement as to whether or not each candidate lives within the ward / constituency
  • A general statement confirming that each candidate meets the required residency (or other qualification) criterion.

5.19 An alternative suggestion was that inclusion of address information might be made optional for candidates or left to the discretion of ROs. However, others suggested that those who chose not to include their address might be disadvantaged.

Views on qualification requirements for local elections

5.20 Some respondents discussed the general issue of qualification criteria for candidates in local elections. Regardless of whether respondents agreed or disagreed that address information should continue to be included on local election ballot papers, there was a general view that a residency qualification for candidates was relevant and important. (In most cases, in making their comments respondents did not refer to the other possible ways for candidates to meet the qualification criteria – i.e. place of work, business or property ownership in the local authority area). Respondents argued that this ensured that candidates had a commitment to the area and were knowledgeable about local issues. Thus, some respondents were opposed to non-local candidates in principle; others, however, were concerned about the implications of non-local candidates for travel and relocation expenses. Comments from some respondents suggested that they thought such criteria should apply to all elections.

5.21 There was broad agreement that a robust and transparent system of verification and confirmation (along with appropriate sanctions for breaching the rules) was important to the integrity of the electoral system. However, some of those who favoured removing addresses from ballot papers stressed the distinction between requiring the submission of address information for nomination and verification purposes, and whether that information should be (or needed to be) made available to others.

5.22 In a few cases, respondents argued for qualification criteria to be tightened up by extending the minimum time period required to meet the various criteria (time periods up to two electoral cycles were suggested), by applying a strict ‘full-time address’ definition, or by introducing robust procedures for proving or verifying residency status.

5.23 Other qualification criteria favoured by respondents in this group included requirements to be on the electoral roll in an area; not to be in debt to the local council; and to gather nomination signatures from one per cent of the local population.

5.24 A less common view was that demonstrating a local connection was not important. Some thought a local connection was of limited value in indicating if someone would be an effective representative, while others argued that such a requirement discriminated against groups such as homeless people and travellers.

List order of candidate names (Q10a, 10b, 11)

5.25 Scotland’s local council elections are conducted using a Single Transferrable Vote (STV) voting system. This type of system requires the voter to rank candidates in order of preference – unlike Scottish Parliamentary elections in which the voter simply marks an ‘X’ against the name of their preferred candidate. The consultation paper noted that, in an STV system, the ‘list order effect’ can result in candidates who are listed higher on the ballot paper being selected over those who are lower on the list – thus candidates who are further down the list are at a disadvantage.

5.26 At present, by law, candidates in local council elections are listed alphabetically (by surname) on the ballot paper. The consultation paper discussed different options for ordering candidate names, and asked respondents for their views:

Question 10a: Do you agree that, in order to counteract the list order effect, a change should be made to the way in which candidates are listed on election ballot papers? [Yes / No]

Question 10b: If so, what form of new system would you favour? [Rotation / Randomisation / Alphabetical–reverse alphabetical / Any other (please specify)]

Question 11: Do you have any other comments to make on this issue?

5.27 Question 10a asked respondents if they thought a change should be made in the way candidates are listed on election ballot papers to counteract the ‘list order effect’.

5.28 A total of 732 respondents answered this question. Table 5.2 shows that organisations were divided in their views – with 50% answering ‘yes’ and 50% answering ‘no’ – whereas individuals were generally in favour – with 82% answering ‘yes’.

Table 5.2: Q10a – Do you agree that, in order to counteract the list order effect, a change should be made in the way in which candidates are listed on election ballot papers?

Respondent type Yes No Total
n % n % n %
Organisations 10 50% 10 50% 20 100%
Individuals 586 82% 126 18% 712 100%
Total 596 81% 136 19% 732 100%

5.29 Respondents who answered ‘yes’ to Question 10a were then asked for their views on the type of new system they would prefer (Question 10b). The consultation paper suggested three possibilities: (i) rotation (where candidates’ names are ordered differently on each version), (ii) randomisation (where candidates have their position on the ballot paper determined by lottery) and (iii) alphabetical-reverse alphabetical (where half the ballot papers are printed in alphabetical order and half in reverse alphabetical order). Table 5.3 shows that, among those who replied (n=579), almost two-thirds (64%) were in favour of candidate names being ordered in a random fashion on ballot papers. Organisations often provided comments without expressing a view at either Question 10a or 10b.

Table 5.3: Q10b – If so, what form of new system would you favour?

Respondent type Organisations Individuals Total
n % n % n %
Rotation 1 10% 51 9% 52 9%
Randomisation 5 50% 365 64% 370 64%
Alphabetical–reverse alphabetical 0% 67 12% 67 12%
Other* 3 30% 50 9% 53 9%
More than 1 choice selected 1 10% 36 6% 37 6%
Total 10 100% 569 100% 579 100%

The table includes only those who answered ‘yes’ to Question 10a above.

* Respondents who indicated a preference for one of the first three options but commented that they wanted candidate names grouped by political party have been recoded into the ‘other’ category above.

Percentages may not total 100% due to rounding.

5.30 Altogether, 253 respondents (36 organisations and 217 individuals) made additional comments in relation to these questions – at Question 10b and / or Question 11. Respondents commented on:

  • The status quo: Respondents who answered ‘yes’ at Question 10a explained why they thought a change was needed. Those who answered ‘no’ explained their reasons for preferring the status quo. Respondents who did not answer the tick-box question (including electoral bodies and most public sector organisations) identified challenges and potential risks in changing the current alphabetical listing; these were closely related to the reasons given by respondents who answered ‘no’.
  • The three options discussed in the consultation paper: Respondents who answered ‘yes’ at Question 10a, and then selected one of the three choices at Question 10b (rotation, randomisation, alphabetical–reverse alphabetical) often explained the reasons for their choice and / or gave reasons for ruling out the other options. Those who selected ‘no’ at Question 10a often explained why (in their view) none of the options offered were appropriate.
  • Other options for counteracting the list order effect: Respondents who selected ‘Other’ as their choice at 10b suggested a small number of alternative ways of listing candidate names.

5.31 Each of these topics is discussed further below, and some respondents (including those who did not answer the tick-box question) also made general points which are discussed at the end of this section.

Views regarding the status quo

5.32 There were two main perspectives on the (current) alphabetical listing of candidates’ names on ballot papers: the first was that this was these arrangements should change because they are unfair; the second was that there were likely to be significant risks and substantial costs in any other method of listing candidates’ names.

Views of those favouring change

5.33 Respondents answering ‘yes’ to Question 10a generally reiterated the points made in the consultation paper that candidates nearer the top of the list on ballot paper have an unfair advantage over those nearer the bottom of the list. Some pointed to specific cases where the list order effect had resulted in an inexperienced candidate, higher on the list, unseating an experienced incumbent candidate in the same party, lower on the list. These respondents thought that changing the way names are listed on the ballot paper would make it fairer for all candidates. Some expressed the view that this issue was very important.

Views of those favouring the status quo, and those highlighting the risks of change

5.34 Respondents who answered ‘no’ to Question 10a made a range of points. Similar issues were raised by those who did not answer the tick-box question.

5.35 While some respondents queried whether the list order effect did in fact exist, most acknowledged that it did. However, on closer examination it appeared that some believed that the list order effect was mainly a feature in relation to the selection of candidates from within a particular party where more than one candidate was on the ballot paper (i.e. they did not think that the list order effect necessarily applied in the case where there was only one candidate from a particular party since, in that case, the voter would simply opt for the party of their choice, irrespective of where the candidate’s name appeared in the alphabetical listing). These respondents therefore believed that the list order effect was primarily a problem for candidates, not for voters. This group argued that, from the perspective of voters, the scale and significance of this problem was small, whereas the potential costs and risks of addressing it were likely to be great. Those who answered ‘no’ to Question 10a thought that these costs and risks could not be justified. There was a widespread view, both among those who answered ‘no’ and among those who did not answer the tick-box question, that the interests of voters (rather than candidates) must take priority in this matter.

5.36 Respondents in this group pointed out that voters are familiar with alphabetical ordering of names on ballot papers. They thought that any other way of ordering was likely to cause confusion – particularly if there were multiple versions of a ballot paper within the same ward. (The example of an elderly couple receiving different versions of a postal ballot paper was highlighted.) In addition, any change in the alphabetical listing would only apply to local government elections; candidate names would still be in alphabetical order on ballot papers for Scottish and UK Parliament elections, and respondents suggested that this would cause confusion among voters.

5.37 Other potential risks included:

  • Problems related to printing and checking ballot papers: There could be an increased risk of errors in printing; and there is likely to be increased the pressure on the administrative processes for ensuring ballot accuracy. In addition, a change to a randomised listing would limit print suppliers to those with the necessary digital facilities capable of producing sets of ballot papers with different ordering, which could increase costs substantially.
  • Problems with the count: This included confusion among count staff and an increased risk of errors in counting. It was suggested that a manual count of a randomised ballot paper would be extremely challenging to carry out, thus requiring an electronic count for all elections including by-elections.
  • Adverse impacts on voters with disabilities: Repeated concerns were voiced that any system of randomisation (in particular) would cause problems for groups with disabilities (e.g. dyslexia, learning disabilities, cognitive impairments and visual impairments), who were likely to find it more challenging to use a ballot paper with a long list of candidates in random order. Any changes to the alphabetical ordering of names would need to consider the implications for large print and handheld ballot papers, tactile voting devices, etc., to ensure that people with disabilities are not disadvantaged.

5.38 Some respondents who favoured the status quo and some third sector organisations representing groups with disabilities emphasised that, if there was nevertheless a move towards a random listing of candidates, then to avoid confusion, there should be only one version of the ballot paper for each ward.

Views regarding the three suggested options

5.39 In their comments on the three options discussed in the consultation paper – randomisation, alphabetical–reverse alphabetical and rotation – respondents generally highlighted the perceived advantages of their preferred option and / or the disadvantages of the other two.


5.40 As shown in Table 5.3 above, among the respondents supporting a new system of ordering candidate names, nearly two-thirds (64%) indicated that they supported some form of randomisation. Randomisation was described as the ‘fairest’ way to list candidate names.

5.41 However, among those who supported randomisation, there was disagreement about whether there should be one version of a ballot paper in each ward (with names listed in a random order), or multiple versions. The former, it was suggested, would help to keep costs down and allow voters (particularly those with disabilities) to become familiar with the ballot paper prior to the vote. The latter, it was suggested, was most likely to counteract the list order effect, but also had the greatest risks and costs (see again paragraph 5.37 above).

5.42 Some respondents thought that cost should not be a factor in the decision. However, it was more common for respondents to express concern about the potential costs and / or to suggest ways of minimising costs. These generally involved limiting the number of versions of the ballot paper printed – to one (as discussed above) or a maximum of two.

5.43 In relation to the randomisation process, respondents suggested that the list order on the ballot could be decided by lot or by listing candidates in the order in which their nominations were received.

5.44 Those who were not in favour of randomisation highlighted the risk that a random ordering of a long list of candidates would cause confusion among voters, make it more difficult for people to identify the candidate(s) they want to vote for, and make it harder for some people to participate in the vote.

Alphabetical–reverse alphabetical

5.45 Twelve percent (12%) of those who supported a change in the existing system were in favour of an alphabetical–reverse alphabetical listing of candidate names (see again Table 5.3 above). Those preferring this option thought it would be ‘fair’ and ‘simple’ – the cost and complication of the other options was a factor for some who chose this option.

5.46 Those who gave reasons for not supporting this option suggested that, in a long list of candidates, those in the middle of the list would still be disadvantaged whether the list ran alphabetically or reverse alphabetically.


5.47 Nine percent (9%) of those who wanted to change the current system expressed a preference for a rotation system (Table 5.3). Respondents saw this option as addressing the ‘bias’ of an alphabetical ordering since it would ensure that all candidates appear an equal number of times in each position on the ballot paper. It was also seen to be less complicated than randomisation, and less costly. Some also suggested that the use of a rotational system with alphabetical ordering (referred to as ‘Robson Rotation’ in the consultation paper) would assist voters in identifying their preferred candidate(s) on the list – unlike a completely randomised ordering. It was pointed out that, in order to fully counteract the list order effect, it would be necessary to ensure that each version of the ballot paper was distributed randomly in equal numbers.

5.48 The disadvantage of a rotation system was seen to be that it still retained the alphabetical ordering of names, thus some respondents thought it was unlikely to fully solve the problem of the list order effect – unless it was combined with a form of randomisation.

Other suggestions for counteracting the list order effect

5.49 Respondents who favoured some ‘other’ system (not one of the three choices in the consultation questionnaire) generally wanted candidates to be listed by party. However, those who suggested this had different views about: (i) whether the party names should be listed alphabetically or randomly (e.g. decided by lottery); (ii) whether the candidate names under each party name should be decided by the party itself (the most common view), listed randomly or listed alphabetically; and (iii) where independent candidates would appear in the list (alphabetised under ‘I’ for independent; randomly or alphabetically ordered at the top of the list; interspersed randomly among the party lists; etc.).

5.50 However, some respondents argued against listing candidates by party. This group included some who chose one (or none) of the other options offered in the consultation questionnaire. They thought a listing by party would give too much importance to party affiliation, rather than the qualities / attributes of candidates. It would also, essentially, create two list order effects – first of party names, then of candidate names. The process of attempting to address these would (it was thought) be extremely complex and potentially controversial.

5.51 Occasionally, respondents offered other suggestions for addressing the list order effect – for example, ordering candidate names in a chequerboard format, or displaying them in a circle, rather than as a list. Some also suggested ways of making ballot papers more user-friendly – particularly if candidate names were ordered randomly (e.g. using colour-coding with candidate names printed on their party’s colour; having photos of candidates next to their name, etc.). Such suggestions were generally offered by just one or two respondents.

General points

5.52 Some respondents (including electoral bodies and most public sector organisations) made more general points. This group repeatedly emphasised that any change in ballot paper design must be extensively tested and assessed for its impact on voters before any change in the law is made to ensure there are no unintended consequences. Respondents also offered some suggestions for going forward:

  • Any change made to the design of ballot papers should be assessed against the design principles set out in the Electoral Commission’s good practice guide for electoral administrators.[7]
  • Any change should be as simple as possible to avoid excessive costs and unnecessary complexity.
  • Any change should be workable in terms of electoral administration.
  • The process and outcome of making any changes must be transparent.
  • Future technology needs and possible changes to the voting system should also be taken into consideration.

5.53 Some respondents also commented that the problem of the list order effect is a sign that candidates are not well known by their electorate, and that this problem is magnified by low voter turnout. It was argued that effective and clear campaigning by candidates should be the first step in addressing the list order effect.