Après un week-end durant lequel Theresa May a tenté de convaincre une énième fois les hard Brexiteers de la soutenir lors d’une rencontre à Chequers Court, la Chambre des Communes a voté à deux reprises lundi en faveur des votes indicatifs. Derrière cette expression se dissimule une reprise en main partielle du processus du Brexit par les députés. L’objectif est de débattre de toutes les options connues pour trouver une issue à l’impasse dans laquelle se trouvent les Britanniques. Pour les présenter, nous avons repris intégralement un texte publié par la Hansard Society qui explique très clairement (en anglais) le contenu des motions qui seront potentiellement discutées et les enjeux procéduraux liés aux votes indicatifs. Pour tout dire, il était temps que ce type de débat ait lieu. Cela fait des mois que nous soutenions cette possibilité afin que les parlementaires constatent par eux-mêmes que le deal, même minoritaire, est peut-être la moins mauvaise des solutions. Néanmoins, c’est sans doute un peu tard en raison de l’opposition durable du Gouvernement contre les votes indicatifs et d’une certaine frilosité des conservateurs pour défier Mme May sur ce point.

Nous souhaitons également souligner un point : si les MPs ne parviennent pas à s’entendre sur telle ou telle option, cela ne signifie pas que, automatiquement, ils adhéreront à l’occasion d’un troisième meaningful vote au projet de la Première ministre. Par conséquent, les débats du 27 mars n’aboutiront pas forcément sur un consensus viable permettant de sortir de l’ornière. Dans de telles circonstances, ce serait un prolongement au-delà du 22 mai qui se dessinerait avec peut-être une démission de Theresa May et des élections anticipées à la clef. Nous avions aussi évoqué il y a bien longtemps cet enchaînement assez logique des événements. Là encore, il est regrettable qu’il ne soit pas survenu plus tôt, ce qui aurait pu éviter que la crise ne se prolonge.

Nous rappelons enfin que ces votes sont bien indicatifs : le Gouvernement peut ne pas les suivre. Le no deal au 12 avril ne doit donc pas être totalement exclu, même s’il est de moins en moins probable dans l’immédiat.

Indicative Votes: Options, voting methods and voting systems

“https://www.hansardsociety.org.uk/publications/briefings/indicative-votes-options-voting-methods-and-voting-systems”

Ruth Fox is Director and Head of Research at the Hansard Society ; Lewis Baston is the former Director of Research at the Electoral Reform Society and research fellow at Democratic Audit. He now works as a consultant and author and is a Fellow of the Royal Society of Arts

---

1). Indicative votes on Brexit: what and why?

When MPs, ministers and officials speak of indicative votes, they generally mean a series of successive votes by MPs on freestanding motions in which Members can vote in favour of as many of the options as they consider acceptable, following a debate on each of the proposals.

However, in the absence of any clarity about what options MPs might actually vote on, using what voting system, and when, ‘indicative votes’ has, amidst the frustration with events, become a catch-all term for a voting process through which a consensus on the way forward might be generated in order to break the current Brexit deadlock.

Both Houses of Parliament have prior experience of indicative votes (see Appendix B), most recently in relation to House of Lords reform in 2003 and 2007. The results – in which no option commanded a majority in 2003, and two contradictory options were supported in 2007 – were not a strong advert for the process.

However, indicative votes do not have to be designed and implemented in the way chosen over a decade ago. A sequenced and structured approach to multi-dimensional decision-making could be established in other ways which might generate a more helpful, informative and productive outcome than has been produced so far using Parliament’s traditional, binary proposition-amendment voting model.

Given the polarised nature of the Brexit debate, there can be no guarantee that anything positive will arise from this process, particularly if MPs are not prepared to compromise. And ultimately, if progress is made and there is an outcome, the government will have to decide whether, politically, it can live with it and the implications that flow from it. The extent to which opinion about the way forward on Brexit now splits across party lines may make it very difficult for this government, or indeed any other potential governing combination in this Parliament, to adopt the outcome of an indicative vote process and hope to command the necessary political support for whatever that may be, on an ongoing basis, in vote after vote, over months if not years.

Nonetheless, on Monday 25 March, MPs voted by 329 to 302 for a cross-party amendment to give precedence to a Brexit-related Business Motion on Wednesday 27 March to introduce indicative votes in the House of Commons.

This paper is designed to help inform MPs as they deliberate on the process and procedure to be used for that indicative vote process.

The paper is not an exhaustive examination of all the options, permutations, and implications. But it is designed to support the debate by adding some context and detail to the issues under discussion and making some informed suggestions about structure, sequencing, and the rules that will need to be drawn up to govern the process in place of the normal guidance provided by Standing Orders.

The Hansard Society’s research focus is on Parliament, not public policy. In setting out the potential Brexit options upon which MPs may wish to vote, we have adopted the suggestions and terminology widely used by others in this debate. Our aim is to illustrate the range and type of choices that exist in relation to the options, and the sequencing of decision-making that could be adopted in relation to both the immediate decisions required on the next steps and the longer-term end-state destination options. The choices laid out should not be interpreted as an endorsement of any or all of those options by the Hansard Society. Others will undoubtedly suggest alternative options or variations on options. Whatever the approach, further options beyond those we have set out here can be accommodated in the process, providing that due care and attention is given to where in the sequencing of decision-making they most appropriately fall.

What is the House of Commons trying to achieve?

Amidst the clamour for indicative votes, it is clear that different MPs have different objectives for and expectations of this process. This in itself will make it very difficult to draft the rules, choose the options, and determine the voting system in a way that commands broad support – and therefore legitimacy – across the House.

For some MPs, the purpose of indicative votes is merely to test opinion – to take the temperature of the House – on a range of different, non-binding options.

For others, the intention is to determine which one or more of the potential models for the UK’s future relationship with the EU may genuinely command a majority in the House of Commons and therefore offer a way forward which the government of the day might then pursue.

Some MPs want the outcome of this process to be binding on the government; others, including the Prime Minister and other leading government figures, do not.

But being clear about the objective(s) is essential, as there are different species of option under consideration, and there is an important relationship between the purpose of the process, the options to be voted on, and the voting system that is chosen.

2). Choosing the options: comparing ‘apples’ and ‘oranges’?

There has been press speculation in recent days about a range of options under consideration for indicative votes. The list below is based on ideas floated in the media after ministerial discussions on indicative votes were reportedly held in Whitehall on 22 March.

We use this list for illustrative purposes only. It is possible – indeed likely – that MPs will want to assess further options.

The proposed options considered by ministers reportedly included:

• The Prime Minister’s Deal (Withdrawal Agreement and Political Declaration)
• Revoke Article 50
• Second referendum
• Withdrawal Agreement plus Customs Union
• Withdrawal Agreement plus Customs Union and Single Market
• Free Trade Agreement
• Leave with no deal

However, these options mix ‘next step’ actions and longer-term ‘end-state’ destinations.

If the indicative votes process is to deliver any useful outcome, the options involved and the structuring of decision-making must be conceptually coherent and take account of these different types of options.

i). Differentiating between next steps and end-state options

The difference between the next step and the end destination is important. Some options such as revocation and no deal are both, but there are some options which are one or the other. The deal, for instance is a next step, but combined with a customs union or single market it is an end-state.

It is possible to agree with someone about the next step but have different views about the end destination (favouring a long Article 50 extension is compatible both with an integrated customs union/single market agreement and a distanced free trade agreement, for example).

It is also possible to agree with someone about the end destination while having very different views about the next step to achieve it (for example, some people with the ultimate aim of securing a customs union/single market model will support the existing Withdrawal Agreement, but others who object to provisions in the existing Withdrawal Agreement such as the backstop may seek to achieve their long-term aim via a long extension to Article 50).

ii). Is the option aspirational and dependent, at least in part, on the EU, or is it something that can be achieved solely by the UK?

Deal, revocation and no deal are all achievable by the UK alone. Other options are aspirations that need the agreement of the EU27, either directly (through altering the Political Declaration) or because they require a long Article 50 extension period.

Voting for an option that is likely to be ruled out by the EU27 would require the decision process to be repeated. There is nothing wrong with the House of Commons expressing support for an aspiration, to give the government guidance on the next stages, but it is not directly comparable with something that is directly and unilaterally achievable.

iii). Are the options independent of each other

There are several options in the mix which are similar and not mutually exclusive. For instance, a referendum implies two or more options going through to the public vote, and several of the deal-plus-‘soft’-Brexit options are similar.

3). Structuring the decision-making process

Looking at the ideas that in recent months have been debated by MPs, appeared in amendments to various government motions, and circulated in the press and social media, there appear to be three distinct categories in play: next step actions; end-state destinations; and further procedures to validate the end-state.

i). Sequencing

With the new default EU exit day looming on 12 April, the most pressing decision relates to the immediate next step that will be taken on or before that date. There are only four viable options:

1. Leave the EU with no deal on 12 April;
2. Revoke Article 50;
3. Seek a long extension of the Article 50 period to facilitate other options/further discussion with the EU;
4. Approve the Withdrawal Agreement and Political Declaration which the government has negotiated and leave the EU on 22 May – Arguably, ‘Withdrawal Agreement plus referendum’ might be regarded as a distinct next step, a fifth option. – Modified versions of the deal, via changes to the Political Declaration, might also be regarded as a distinct next step, but could be contingent on a long extension and might therefore be better seen as related to the third option.

(According to the 22 March European Council decision, accepted by the UK, the Article 50 period is extended by default to 22 May only if the House of Commons approves the Withdrawal Agreement by 29 March. However, if the House of Commons failed to do so, but by 12 April had given its approval or looked likely to do so, this could feed into consideration of a further extension beyond 12 April.)

If Option 1: Leave the EU with no deal on 12 April:

If this option were included in the initial indicative voting and was successful, no further votes would be required – this is both a next step and an end-state option.

The other three options may each involve further decisions:

If Option 2: Revoke Article 50:

The government could choose to do this of its own volition. Many legal commentators believe it would need legislation. Alternatively, an additional procedural step could be included, by seeking the public’s consent through a second referendum (see below).

Option 3: Long extension for negotiation or Option 4: Approve the Withdrawal Agreement with or without changes to the Political Declaration:

These two options, if chosen, would open up a new set of questions about MPs’ preferences, and might therefore require a further set of decisions to be made on options such as:

End-state destinations (this is an indicative rather than exhaustive list):

• Customs Union
• Customs Union and Single Market membership
• Other association models
• Free Trade Agreement

Revocation or leaving without a deal might recur as options later, if option 3 were chosen and the negotiations were to prove unsatisfactory. This might also require further procedural steps.

Procedure(s) to validate the end-state:

• What, if any, procedure(s) should be required to affirm the final outcome? Future procedure can influence whether the next stage is acceptable to the House or not. The most evident example of this is support for the Prime Minister’s agreement, which some MPs are willing to offer conditional either on a future referendum or on guarantees of parliamentary involvement in decisions about the future relationship.
• Another example is the Kyle-Wilson proposal to support the Withdrawal Agreement subject to it being confirmed in a referendum. Here, it is important to distinguish between a referendum to deliver an end-state outcome (e.g. to revoke Article 50) and a referendum as a further step in the decision-making process to validate the end-state that the government chooses, subsequent to any advice given by the House through this indicative votes process.
• A referendum is not the only possible procedural mechanism that could be adopted. Other procedural options to validate later decisions might include requiring further ‘meaningful votes’ on the final agreement(s) with the EU and on other trade agreements; consultation with Parliament during the process of reaching agreement on the future relationship; or options to create standing institutions to achieve cross-party consensus on the future relationship.

ii). How will the wording of the options be determined?

The wording of the propositions to be voted upon is important. The integrity of any process of deciding on options depends on a fair representation of what each option means. This is why the Electoral Commission has a role in assessing which words should appear on referendum ballot papers.

When indicative votes were used in the House of Commons in 2003 and 2007 in relation to the composition of the House of Lords, the options were drawn from a report by the Joint Committee on House of Lords Reform and subsequently a government White Paper.

Without a source such as this to which MPs can have recourse, the problem in the House of Commons is that, on a cross-party basis, there is no particular mechanism or group to whom MPs can turn easily to organise and co-ordinate a process to determine the options, particularly at short notice.

The supporters of each of the potential options ought to be involved in agreeing the wording. Bodies that might contribute to the process might include, for example, the Liaison Committee (the committee comprising the chairs of select committees), or a relevant select committee such as the Procedure Committee or the Brexit Committee. Supervision and oversight of the process might fall under the auspices of the Speaker of the House of Commons. However, if it is proposed that amendments to any of the propositions will be permitted then it may not be appropriate for the Speaker – who will have to select those amendments – also to be involved in helping to determine the main motions. In this scenario, the role could be performed by the Deputy Speakers.

MPs must recognise that amidst the current of opinion there is a near-infinite number of smaller variations and ways of describing a proposition. There needs to be some consolidation and tolerance of variation within each end-state option.

iii). What do the rules need to address?

Changing the method of voting and counting will require a bespoke process in which the rules of engagement are clearly set out. It will be necessary to dispense with some Standing Orders that would normally pertain to debate in the Chamber and replace them with stand-alone provisions for the purpose of this set of votes only.

In drawing up the bespoke set of rules, key questions to be addressed include (but are not limited to):

• Who will decide the options to be voted on, how, and in what order?
• What measures will ensure that the views of smaller Opposition parties are taken into account in the decision-making process? (In any normal vote they have a reasonable expectation of being able to put their views in a motion or amendment, and on some occasions of the House voting on it.)
• When the options are decided, will MPs be permitted to propose amendments to the motions? If so, will the Chair be given the power of selection in the normal way?
• Who will move the motions and open the debate? It has been variously proposed that the motions for indicative votes might be moved either by the government or by a backbencher. We are not aware of any precedent for non-ministerial motions to be moved in this way. Previous motions on indicative votes – including those in 2003 and 2007 on House of Lords reform – were all moved by a minister.
• Who will open and close the debate? (Normally this would be done by the relevant minister.)
• Should the practice of not voting on the same question twice in the same session ( the source of a recent controversial ruling by the Speaker) be disapplied for the purpose of this vote?
• What provisions need to be made in the voting process to accommodate proxy voting?
• If a ballot paper is used for the vote(s), how long will MPs have to cast their vote? And what provisions will be made to publish the votes cast by each Member? How soon after the vote closes will the results be announced? How far, if at all, might the voting process model be that which is used for deferred divisions?

4). Voting methods and criteria

There are two conceptually separate aspects to the choice about how to conduct indicative votes:

1. Voting method – modalities like ballots or lobby counting, private or public, etc.;
2. Voting system – the rules for counting and getting a result.

The two need to be congruent with each other, in that the method has to be suitable for implementing the voting system used, and this is not possible with every combination of system and method.

i). Voting methods

Should the voting method be private or public? This will affect the technique of voting, as options that use the existing machinery of the voting lobbies involve publicly-recorded votes, while those with paper ballots are compatible with both public and private votes. There is also the option of leaving it for individual MPs to choose whether or not to make their vote public.

There are fair arguments on both sides for private and public. The argument for private is that indicative votes are about finding out what the House really believes, as a way of informing the government about which course to take. Private voting is therefore less subject to formal and informal whipping, strategic voting and – sad to say – intimidation. On the other hand, recorded public voting is more compatible with legitimate accountability to the electorate, and with ascertaining what MPs might vote for if and when publicly-recorded voting is required (such as in voting on the approval motion and legislation that are statutory ratification requirements for the Withdrawal Agreement, or in legislating for a referendum). *Whilst private voting might help secure consensus, on an issue of as great public importance as Brexit it is inconceivable that MPs’ decisions would not be publicly recorded. *

Should the vote be serial or one-off? Some voting systems require one (Single Transferable Vote and cumulative voting need a single ballot) or the other (proposition-amendment needs a series of votes). But there are also voting systems that can be implemented either way (Alternative Vote; approval voting). Serial methods are more in line with traditional methods of voting in Parliament, but are also probably more subject to strategic voting.

ii). Voting criteria

Some criteria for a ‘good’ voting system to use for indicative votes will be a matter of consensus, but others will be a matter of choice.

The following are among the criteria on which there is, or can be expected to be, consensus:

The system should not produce a result that is useless for the purposes for which the indicative votes are taking place. Among ‘useless’ results one can list:

• all options being negated;
• the passing of two or more options that combine into asking for the unrealistic/impossible;
• the passing of an option that is inherently unrealistic;
• passing an end-stage result but with no way of getting there.

The system should not incentivise insincere voting which is based on tactical/strategic considerations about future votes, rather than a view on the merits of the proposition being voted upon. It is impossible to devise a system which is completely immune from this, but it can be minimised.

The system should not be affected by accidental factors unrelated to preferences, such as the order in which votes are taken i.e. voting should not stop when you reach a vote with majority support, or privilege one option that is left standing if all others are defeated. This rules out the traditional way in which motion-amendment voting takes place on legislation.

The system should be compatible with an acceptable voting method. There is no point, for instance, in using a system that requires a method that is not acceptable to MPs.

Criteria which may be considered but on which different views might be found, include:

In voting theory the idea of a ‘Condorcet winner’ is significant. This is a choice which, in a two-choice vote compared to any other available option, is preferred by a majority (the members of the majority will probably be different on different votes). A Condorcet winner might have little support as a first choice but be an acceptable fall-back for a majority, which gives it a form of democratic legitimacy despite its low initial support. There might not be a Condorcet winner at all, but it is a significant piece of information. At the very least, a Condorcet loser – which would be defeated by any other option in a one-to-one vote – should not be the winner.

The system should produce information not only about an option that is capable of majority support, but also about the relative attractiveness of all options that are being considered. The more information the indicative votes generate about the wishes of the House, and what these might be when presented with narrower choices in the future, the more useful the votes will be to government and policy-makers. Against this, it may be argued that the system should provide clarity; and that the more nuanced the information arising, the less clear the outcome of the vote will be.

The system should provide a single determinate result. The extent to which this is important differs at each stage. For the next step, a single result from the menu of four or five options would seem to be very desirable, given that these are more precisely defined and that the time for decision is short. For end-states, this is less important, and for the House to provide a range of acceptable outcomes might be useful to the government in its future negotiating strategy. Therefore, the best system might not be the same for voting on the next steps and on the final outcome.

5). Voting system: the options

A). Proposition-negation

This is the standard system for parliamentary voting. The vote is between a proposition and its negation, rather than directly between different propositions. Votes would be taken in turn on each of the indicative vote options. This is the procedure that was used in the indicative votes on House of Lords reform in 2003 and 2007.

Advantages

• This method is familiar and well-understood by MPs.
• It gives a comparative measure of how much absolute support there is for each option and (for end-states) how well such an option might weather the actual parliamentary process when it comes to it.
• It does not matter if several options are similar to each other.

Disadvantages

• There is a very high risk of the method misfiring, as shown in the House of Lords reform votes, and failing to produce a useful result.
• As previous experience shows, it is also vulnerable to strategic voting.

B). Alternative Vote (AV)

This system is a preferential vote used to select a single winner. The principle is essentially that the winner has to have majority support over the combined vote for all the other contenders still in the count.

If there is no winner (i.e. no option with at least 50 per cent-plus-one support) then the lowest-ranked candidate is eliminated from the count.

It is possible to conduct an AV count either in a single ballot (like the Presidency of Ireland) or in a succession of ballots (basically the voting system used for ‘The X Factor’). It is used to choose the Speaker in both Houses – using a single ballot in the Lords, and sequential votes in the Commons. Select Committee Chairs are also elected using a single-ballot AV count.

Voting in a single ballot involves the voter designating their first-choice option with a ‘1’. Their next-favourite option is given a ‘2’ and so on, until the voter is indifferent between the remaining options.

Voting in successive ballots can be done using a single cross for one option, as in a parliamentary election. A whole new ballot takes place at each stage of the count, with the removal of the option of voting for the proposition that came last at the previous count.

Each option has somewhat different properties. Voting among several options in this context is probably best done by a single AV count, as it is less vulnerable to strategic voting.

Advantages

• The advantage of an AV ballot is that it will produce a single option that, at face value, commands a majority in the House of Commons. If the government seeks definitive guidance from the House of Commons, then this procedure will give an instruction.
• In an AV count, it does not matter much if several options are similar to each other (although it is possible that in a repeated, exhaustive – ‘X Factor’-style – ballot, opponents will vote strategically to ensure the weakest variant gets through).

Disadvantages

• The main drawback of AV is that, particularly in a situation with polarised views, it can prevent the emergence of compromise options. It may eliminate a ‘Condorcet winner’ – i.e. an option that a majority prefers to any other option – at an early stage of the count because it has insufficient first preferences.
• Particularly when it is taken in successive (exhaustive) ballots, AV is vulnerable to insincere voting in order to eliminate strong competitors.

C). Single Transferable Vote (STV)

STV is the multiple-winner version of preferential voting. It uses a single, preferential ballot as in the single-ballot version of AV. The rules for counting are different, but based on the same principle. For instance, in choosing among three options, there is a quota of 25% of the vote. If an option has more than enough support to be elected, a fraction of its votes is transferred to the next-favourite option of the voters. If this process does not choose another winner, lowest-placed candidates are eliminated and their preferences redistributed, as under AV.

Advantages

• STV is one of the hardest systems to manipulate, so people will tend to give their sincere first preferences and not vote strategically.
• It does not matter whether there are several similar options under consideration – STV will not harm that ‘family’ of solutions because votes will be transferred.
• It will identify the options with the broadest support within each camp and also those which are capable of attracting transferred votes and therefore forming the basis for a consensus.
• It can narrow-down the options for consideration – such as the two choices going forward to a public referendum –which the government could then prepare as detailed propositions.
• STV is flexible in that it can be used to choose any number of winners, as considered appropriate. There is even a counting method (the Rosenstiel method) which can use an STV vote to calculate a ranked order of preferences.

Disadvantages

• The STV counting method is complex and may be difficult for the public to comprehend.
• It is not suitable for producing one definitive result (that is AV).
• There are conceptual difficulties in this case with the justification for transfers from options that have a surplus i.e. those with more support than necessary to ensure consideration. Transfers might be seen as supporters of a majority position getting ‘two bites at the cherry’ or helping to select what goes forward as the representation of the minority (problems that do not occur when electing a group of candidates).

D). Borda Count

Instead of using elimination to try to get to a winner, as in AV, the Borda count aims to sum up the overall strength of feeling about the whole range of options.

The Borda count is another preferential voting system, in that the voters rate the various options in order, as in AV. However, counting is different. Each preference given is converted during the count into a number of points – maximum points for a first preference, fewer points for a second one, and so on. The total points for each option would be added up to give an overall score for the option – it would generate a winner, runner-up, and winner within each broad family of views.

The Eurovision Song Contest voting system is a variant of the Borda system.

Advantages

• A Borda count would establish both which are the leading options and which options have the most general consent.
• It would allow people who have nuanced views to express them in their vote.

Disadvantages

• The big problem with Borda counts is that they are easily gamed. Voters have an obvious incentive to give insincerely low rankings to the strong options on the ‘other side’ and therefore the count is distorted.
• Another big problem is that if there are several similar options in the ballot, all of their ratings are inflated because their supporters will tend to give high ratings to each other (e.g. supporters of deal + CU + single market would be likely to give high ratings to deal + CU as well). This is the familiar Balkan or Scandinavian group voting phenomenon in Eurovision.
• Borda counts make implicit assumptions about how strongly people prefer one option to their next preference.

E). Approval Voting

Approval voting is a system where people can vote in favour of as many options as they like, without any preferential element. The voter simply marks an ‘X’ for each option that is acceptable. No negative votes are taken. The voter can vote for as many or as few of the options as they like. It can produce simultaneous majorities for different propositions – although, it is to be hoped, not inconsistent ones if voters are playing straight.

Approval voting might conflict with the assumptions that MPs bring to the act of voting, particularly if it is implemented by the method of voting in lobbies. Under approval voting, an indicative vote for an option is a measure of consent to keep that option on the table rather than an active expression of support for it, but in practice MPs might vote as if it meant something ‘harder’ than it does, so using paper ballots is recommended in this instance.

Advantages

• It means voting for things, rather than against them.
• Approval voting is simple but unfamiliar, which probably makes it hard to manipulate.
• Approval voting is good at finding the lowest-common-denominator option which attracts the widest measure of non-committal consent.
• It identifies no-hope options so that they can be discarded.

Disadvantages

• It is not good at measuring how intensely people prefer their favourite options.
• Even with paper ballots, MPs might vote only for strongly favoured options because opinion is polarised.
• It may well not produce a single clear winner.

F). Cumulative Voting

Cumulative voting is where the voter is given a ‘budget’ of a number of votes that they can cast among the various options. The distinctive feature of this system is that, if people so choose, they can use more than one of their votes for the same option.

There are further choices in terms of the number of votes each voter has, and whether there is any restriction on how the votes can be distributed (i.e. is there a maximum number of votes that can be given for the voter’s favourite option?)

Advantages

• Cumulative voting gives a lot of information about the strength of preferences and about options that could command general assent.
• It can pick up second-choice options that have wide consent.

Disadvantages

• In contrast to Borda counts, where including options that are similar to each other helps all of them, schools of thought are disadvantaged by having more than one option in the count because votes end up split between them.
• Cumulative voting is bad at producing an unchallenged winner.

G). Pairwise

This voting system indirectly establishes a collective preference order. The principle is that it puts up each of the viable options against each other and asks people to choose between them – as if it were the final round of an AV count.

Advantages

• It is easily done using the standard Westminster voting method.
• If a Condorcet winner exists, pairwise voting will establish what it is.
• It gives the government lots of information about what happens when push comes to shove.

Disadvantages

• The obvious problem is that if there are multiple options, the system involves holding a lot of votes. With three choices (3 ballots) or four choices (6 ballots) it is manageable, but with seven choices MPs would face up to 21 ballots.
• It is also vulnerable to feedback if the votes are taken sequentially – people may start to vote and abstain strategically.
• It involves a complex and unfamiliar matrix ballot paper if the pairwise choices are all made at the same time to avoid feedback.

Which system?

Depending on the sequencing and structuring of the options – the next step action, end-state destination and consideration of further procedures – at least two and potentially more sorts of decisions are being made. Unsurprisingly, the best voting system might be different for each one.

Next stage: the decision here requires a system that is suitable for a single winner. The risk of a subtler form of vote, which gives levels of qualified approval to different options without choosing between them, is that it will not move the decision forward. Opinion on the next step options is also more polarised than on end-states, and the options are more independent of each other.

The Alternative Vote and Pairwise voting systems are thus the most likely to produce clear-cut choices between next stage options.

Some of the possible next step options render further discussion of end-states irrelevant.

End-state destination: to garner information about what MPs would find to be acceptable end-states (most of which are subject to further definition and negotiation), there is an argument for a voting system that produces a ‘fuzzier’ result, indicating relative preferences between possibilities that need not be entirely independent of each other.

Systems such as cumulative and approval voting may generate the sort of indicative results that are useful for deciding a negotiating approach once the next stage issue has been cleared, or for deciding options to be put to a referendum.

Appendix A: The background to indicative votes on Brexit, 2018-19

The possibility of holding indicative House of Commons votes in the Brexit context was first considered by the Procedure Committee in autumn 2018, as part of its response to a request from the Exiting the EU Committee to consider the procedures that might be used for the House’s statutory approval of the Withdrawal Agreement under s. 13(1) of the EU (Withdrawal) Act (EU(W)A).

The option of holding indicative votes was suggested in evidence to the Procedure Committee’s inquiry by the backbench MPs Dominic Grieve and Sir Oliver Letwin. Mr Grieve and Sir Oliver wanted the House to decide on a set of free-standing motions, before deciding on the Withdrawal Agreement approval motion under the EU(W)A.

In its report, published in November 2018, the Procedure Committee considered that Mr Grieve and Sir Oliver’s proposed procedure would allow a clear decision on the s.13 approval motion, a full discussion of a range of considerations, and the testing of the opinion of the House without undermining the integrity of the s. 13 approval motion. However, the Committee also considered that the proposed procedure would deny the House the certainty of the amendable vote indicated by ministers; risk the House’s reputation by potentially allowing it to back mutually contradictory positions, or none; confine the House’s role with respect to the Withdrawal Agreement to a veto; be difficult for the public to understand; and be without precedent in terms of the relationship between indicative votes and a decision with statutory effect. The Committee did not recommend that indicative votes go ahead.

The Brexit Committee itself first recommended that the House hold a series of indicative votes in a report published in mid-January, immediately after the House had declined for the first time to approve the Withdrawal Agreement and Political Declaration. In its January report, the Brexit Committee recommended that the four options before the House in any such indicative votes should be: a further approval vote on the Withdrawal Agreement and Political Declaration; a no-deal Brexit; renegotiation with the EU, covering a range of possible end-states; and a second referendum.

Since the Brexit Committee’s January report, the Committee and its Chair, the Labour MP Hilary Benn, have been at the forefront of the campaign for indicative votes.

The Brexit Committee reiterated its recommendation in a subsequent report published on 13 March, immediately after the House had rejected the Withdrawal Agreement and Political Declaration for a second time. The Committee said that the House “must be given an opportunity to identify what new approach might secure a majority in the House of Commons”, and that provision for a series of indicative votes should therefore “happen without delay so as to begin the process by which a clear plan can be identified”.

Supporters of indicative votes have also tabled successive amendments to Brexit-related motions moved by the government, as follows:

• 29 January: Amendment (f), Hilary Benn and others; not selected.
• 14 February: Amendment (c), Kenneth Clarke and others; not selected.
• 14 February: Amendment (f), Dr Sarah Wollaston and others, referring explicitly to the Brexit Committee’s January report; not selected.
• 27 February: Amendment (d), Dame Caroline Spelman and others, setting aside SO No. 14(1) to give precedence to motions in the name of the Brexit Committee Chair; not selected.
• 14 March: Amendment (i), Hilary Benn and others, setting aside SO No. 14(1) to give precedence to a cross-party motion; selected and negatived on division by 314 to 312.

Appendix B: Parliament’s previous use of indicative votes

Listed below are examples we are aware of where a variation on indicative or multi-option votes has been used in one or both Houses of Parliament.

Sunday Trading Bill 1993

Amidst accusations that the ban on Sunday trading in England and Wales was no longer fit for purpose, the government offered votes on a range of alternative options from ‘keep Sunday special’ to normal weekly trading hours. The Bill was considered in Committee of the Whole House with five proposals put as principal amendments to which related amendments could be made. Three principal amendments were tabled by the government, and two by backbench MPs. The third amendment, in favour of de-regulation, was passed, and a subsequent option consequently fell and could not be voted on. In the House of Lords, Peers also voted on the options (in March 1994), albeit that these were structured differently than in the Commons and the matter was considered in Committee. The relevant clause of the Bill was not amended as a result of the votes.

Hunting with Dogs, 17 January 2001 and 18 March 2002

The government Bill to regulate hunting with dogs included three parallel clauses setting out alternative approaches to achieve the policy objective: self-regulation, licensing and a total ban. MPs voted on these options at the first day of Committee stage and chose a ban. Peers also voted on the options when the Bill reached the House of Lords; they opted to reject the ban and licensing but voted in favour of self-regulation. The Bill fell when the 2001 general election was called, but the matter was revisited in 2002 with debates in both Houses on successive days, and votes on three free-standing motions afterwards. The results were the same as previously.

Human Fertilisation and Embryology Bill 2008

The government proposed to amend the Abortion Act 1967 to change the law providing for medical termination of a pregnancy. However, MPs across all parties held widely differing views, driven by ethical and moral concerns. A series of propositions were therefore put to the House, each one suggesting a new weekly term limit to replace the twenty-fourth week provision in the original legislation. MPs voted at Committee stage on parallel clauses providing for medical termination of pregnancy at 12, 14, 16, 18, 20, 22 and 24 weeks. The amendments agreed to were then reported to the whole House for further consideration.

House of Lords reform: February 2003 and March 2007

Both Houses of Parliament voted on multiple options for the composition of a reformed House of Lords. It is this experience of indicative votes with which many MPs are familiar, and which largely informs their understanding of and current thinking about the process today.

In 2003, the procedure and options were proposed by the Joint Committee on House of Lords Reform. Seven options were set out, ranging from a fully appointed House to a fully elected House with a range of hybrid options in-between. Both Houses voted on 4 February 2003. In the House of Lords, a clear outcome emerged: Peers voted for the all-appointed option and rejected all other options. In the House of Commons, none of the seven options commanded majority support. One option was defeated by just three votes; three options were rejected without division. An amendment was permitted to the first motion enabling MPs to vote on abolition of the House of Lords, but it too was defeated. The status quo, an unelected House of Lords, was the fourth most popular option but was retained because of the lack of clear support for any other alternative.

In 2007, both Houses repeated the process, utilising options set out in the government’s White Paper on House of Lords Reform. This time the votes in each House were held on separate days, the House of Commons voting first on 7 March, and Peers voting a week later. In the Commons, a preliminary motion – “That this House supports the principle of a bicameral Parliament” – was moved first. The seven options were moved in a different order than in 2003. A final motion was also moved that “this House is of the opinion that the remaining retained places for Peers whose membership is based on the hereditary principle should be removed”, to which an opposition amendment was selected. These last two motions were not moved in the House of Lords.

The White Paper proposed that the votes should also be held using a ballot paper and the Alternative Vote system, the idea being to thwart the perceived gamesmanship exhibited in 2003 when the order of the ballots helped to determine the outcome. Following the declaration of the result, the votes of each MP were then to be made public in the normal way. Facing considerable opposition, however, the proposal for an AV ballot was withdrawn and the process reverted back to a traditional yes/no vote rather than a preferential system. MPs voted for two motions: an 80% elected House (by 305 to 267) and a 100% elected House (336 to 224). Peers again voted for a fully appointed House and rejected all the other options.

Authors

Ruth Fox is Director and Head of Research at the Hansard Society.

Lewis Baston is the former Director of Research at the Electoral Reform Society and research fellow at Democratic Audit. He now works as a consultant and author and is a Fellow of the Royal Society of Arts. 

For further information please email: contact@hansardsociety.org.uk

Published by the Hansard Society.
5th Floor, 9 King Street, London EC2V 8EA