Our party is proudly committed to the replacement of archaic First Past The Post with Single Transferable Voting, an electoral system which satisfactorily balances proportionality, local representation and voter choice.
However, a major roadblock to the introduction of STV for Westminster and English local elections would be quizzical or apprehensive attitudes amongst a sizable section of the British electorate over its precise technical details, which our party currently fails to address. As a party, we should develop our position on electoral reform by adopting more specific policies regarding technical details for STV.
Firstly, we Liberal Democrats should consider adopting as policy the requirement for a minimum number of candidates to be ranked, as currently mandated in elections for the Australian Senate and the Tasmanian House of Assembly. This would mean that voters would be required to rank at least as many preferences for candidates as there are available seats for their ballots to be deemed valid.
Whilst STV is designed to encourage split-ticket voting and foster political cooperation and moderation, FPTP would likely cast a long shadow for several parliamentary terms after being replaced. Some voters would probably vote solely for candidates from just one party, or even just for one candidate, with FPTP having imprinted upon them either a deeply engrained partisanship or an imperative to vote tactically, both of which are intrinsic components of said system.
Such behaviour could in turn result in perceptively unfair outcomes in seat apportionments as a result of vote exhaustion arising from limited rankings, i.e., voters ranking candidates from just one party and fewer candidates than there are open seats. Taking Shropshire as an example, during the 2019 general election, the Conservatives won 60.3% of the vote, Labour 25.9%, the Liberal Democrats 10.4%, the Greens 2.7% and others a combined 0.7%. If Shropshire had been a five-seat multimember constituency, and those results translated accordingly, it could be considered unfair for us to have won a seat with one-tenth of the vote whilst the Conservatives and Labour would win one seat for every one-fifth of the vote they won respectively.
Minimum ranking requirements might be necessary in order to safeguard the benefits of STV whilst mitigating the caustic legacies of FPTP that would undermine the more proportional system and serve to vindicate its detractors.
Secondly, we should consider adopting as policy a limit on the number of candidates that a given party may field in a given constituency.
There will be the defenders of FPTP who will conjure up the image of a Dutch-style A2-sized ballot paper as a depressingly effective scare tactic against the adoption of a fairer voting system. This is in spite of the unlikelihood of every single party fielding as many candidates as there are available seats, depending on their predicted vote shares, let alone any one party fielding more candidates than there are available seats.
Nonetheless, it would be prudent for there to be a written rule limiting the number of candidates each party could field in a constituency to the maximum number of seats available, if only to reassurance the voting public over ballot paper formatting.
And thirdly, we should consider deciding upon a mechanism for filling vacant seats under STV, if only because this issue may serve to stymie electoral reform by sowing confusion amongst voters.
It would not be difficult to imagine that any method of filling vacant seats without a by-election, namely internal party elections, predetermined orders of successions or vote countbacks, would be unpopular measures, criticised as backroom dealings and fundamental undemocratic.
As by-elections would be the most publicly agreeable method, we should consider whether to endorse single-winner, instant run-off elections as used in Scottish local elections, or possibly a full by-election where all seats are open due to possible perceptions of instant
* Samuel James Jackson is the Chair of the Policy Committee of the Yorkshire and the Humber Liberal Democrats and had served as the Liberal Democratic candidate in Halifax during the 2024 general election.
Follow @libdemvoice.org on Bluesky
Like us on Facebook
Subscribe to our feed
Sign-up for our daily email digest
13 seats change hands in council contests
How four bison brought new life to ancient woodland in Kent
Tom Bailey
Cassie
Iain Donaldson
Peter Hirst
38 Comments
“As a party, we should develop our position on electoral reform by adopting more specific policies regarding technical details for STV.” Agreed.
Thanks for this post, as a strong supporter of PR I feel for it have any chance of success with the electorate it has to be explained and the argument for it put in such a way so as not to put them off with complicated jargon.
This is a good article and I agree that – if (a big if) the party wishes to make seeking electoral reform a priority in its campaigning – the technical details need to be agreed so that questions can be answered properly instead of just saying “but it’s sooooo unfair now” and the party could go into a hypothetical coalition agreement with a clear set of demands.
With regards to by-elections, under PR in general or STV in particular, would it – or should it – be obligatory for a by-election to be triggered if an MP chooses to leave their party?
I’m not particularly convinced by the excuses for avoiding a by-election under the current system, and the notion that voters choose a person rather than a party seems even weaker under PR.
Scottish local elections work fine under STV, There was a strong demand for it for the Scottish Parliament elections but we had to be content with the top-up list. STV has often helped to prevent the election of even more SNP councillors who would otherwise have been elected with less than 30% of the vote. I don’t, however, understand Mr. Jackson’s point about a “full by-election”. Is he meaning that when a councillor dies or resigns everyone in the ward has to stand down and submit to re-election? Surely not.
I’m a big fan of STV and we should definitely be pushing for it to be the replacement to FPTP at Westminster, and use every reasonable opportunity to explain the advantages.
It would take a few election cycles for everyone to get the hang of its full functionality and there’s a very fair chance that the first time someone votes under STV they’ll simply put their 1 where their X used to go, but it will improve with each election.
I wouldn’t support requiring a minimum number of preferences. If people don’t wish to get the full value from their vote – that’s on them! I do, however, think that more can be done to persuade voters to consider 2nd, 3rd and 4th preferences. When I worked at polling stations my standard patter would start with a reminder that it’s ‘numbers for this one’ followed with a comment on how many councillors will be elected a bit about
how they might as well have their say on all of them. If people don’t know or care about more than a couple of candidates, there’s no point forcing them to give their random 2nd preference to Mr Aardvark.
(If I remember correctly, candidate names are in random, not alphabetical order, as it’s already assumed some voters will 1,2,3 their way down the paper.
“Taking Shropshire as an example, during the 2019 general election, the Conservatives won 60.3% of the vote, Labour 25.9%, the Liberal Democrats 10.4%, the Greens 2.7% and others a combined 0.7%……….and those results translated accordingly….”
The phrase “and those results translated accordingly” is very important. Would they necessarily be repeated in an STV election? In an FPTP election a voter might have to consider not only whether or not their preferred candidate or party has any realistic chance of winning but also who they really don’t want to win and use tactical voting accordingly.
Could it be that voting patterns might be very different under STV?
I’m afraid I disagree with the suggestions made in this article though I understand that the author is seeking to stimulate debate. Unfortunately the proposed solutions are more problematic that the issues he thinks require to be addressed.
Firstly, just as it is perfectly valid for a voter to decide to cast no votes, it should also be perfectly acceptable for a voter to indicate a first preference only. For example, if a voter in Scotland finds himself with 4 candidates for a 3 seat ward and it so pro-independence in outlook that he would never vote for a Unionist politician under any circumstance, are we really saying that that voter’s vote should not be valid if it just indicated a First preference for the SNP candidate? It would be an outrage to require that voter to indicate a Second Preference for a candidate representation get a Unionist Party to allow the First Preference vote to be valid.
Secondly, while it is possible that the Monster Raving Loony Party May have a huge number of candidates in one ward just to make a point – they did this recently already in a FPTP by-election – I don’t see any justification for banning parties from doing so. Maximising voter choice is one of the attractions of STV so we should not be restricting that. The parties themselves will understand that running too many candidates actually makes it less likely that they will get candidates elected. Let the parties decide for themselves on this.
I moved to Ireland because of Brexit and have already voted in three STV elections (local, national and European). One of the many positives of living here is the fact that we have a democratic voting system where most votes cast have value in electing someone. There is one big problem with STV, however, that Samuel doesn’t mention, when parties field more than one candidate.
Some time ago I had a closer look at the Scottish local election results and the candidate appearing highest on the ballot paper outpolled his or her party colleague(s) around 80% of the time. In the absence of any other credible explanation, this would suggest a very significant alphabetical bias – which is fundamentally unfair, especially in cases where a party fields two candidates but is only likely to achieve one quota. I still think STV is the best system, but it is deeply flawed if you insist on listing candidates alphabetically on the ballot paper.
The solution is very simple. The preponderance of first votes for the higher candidate suggests that many voters are effectively ‘opting out’ of the candidate choice and simply expressing a party one. In view of this, it would be legitimate to allow parties to specify to the returning officer the order in which they want their candidates to appear – as already happens in list elections. You could make things even easier by grouping party candidates together and listing the parties alphabetically.
@Brad Barrows
“Firstly, just as it is perfectly valid for a voter to decide to cast no votes, it should also be perfectly acceptable for a voter to indicate a first preference only……….”
Valid point. One major benefit of STV is it maximises voter power at the expense of party power. We shouldn’t be trying to dictate to voters how they should use their vote(s).
@Simon Horder – are you implying that alphabetical bias doesn’t happen in STV elections in Ireland? What happens about order of candidates on the ballot paper?
@Simon Horner, rotating the order of candidates for each party is a better solution, so if someone just wants to vote for whoever is listed first, it will be a random chance as to which candidate that is. Such an approach removes the alphabetical issue completely and is easy to implement with modern printing technology.
@Brad Barrows
No, my observation applied to all STV elections where candidates are listed alphabetically – which includes here in Ireland.
Now that you have raised the point, I’ll have a closer look at the last Dáil results (although the sample could be on the small side as there are only 39 constituencies and in many places, parties only put up one candidate).
I suspect there will be some bias, but possibly less pronounced. Dáil members seem to have a higher profile as ‘individuals’ here. I live in the Taoiseach’s constituency and was quite surprised at the last election when I was canvassed by one of his own party activists. I was urged to give my first preference to the other Fianna Fail candidate and there was no mention of putting a ‘2’ in Micheál Martin’s box!
Requiring a full by-election of all MPs for a seat when any one MP resigns is a complete non-starter because it could so be easily abused for party advantage. Imagine a seat has two Labour and three Tory MPs. Labour starts doing well in the polls, and it looks like if there was an election today, they’d get three of the MPs for that seat. So they have one of their MPs resign, which forces all 5 MPs for that seat to face an unwanted byelection, purely in order to try to capture one of the Tory seats.
You’d have by-elections basically happening all the time as parties trying to use them to get rid of opposing MPs. It wouldn’t be at all workable.
@Simon Horner
I didn’t comment on the issue you raised about the order in which candidates are listed but will address it now. I agree that a significant number of voters are more concerned by the Party than the candidate and will happily vote for the first named candidate of their party on the ballot paper, hence leading to an advantage based on alphabetical surname when a party has more than one candidate. Of course it is in the interest of parties to try to get as equal a vote share for each candidate they are standing to try to maximise the chances of getting more elected – Sinn Fein are excellent at doing this in Northern Ireland – so it should really be left to the parties to deal with this issue themselves. That said, randomising the order on the ballot paper rather than arranging alphabetically would merely replace a ‘name’ advantage with an advantage that is based on random chance.
@Simon Horner
“Dáil members seem to have a higher profile as ‘individuals’ here.”
Perhaps because under STV elections they might have to do more campaigning as individuals – since the voters have the opportunity to rank candidates – so even if a voter confines their preferences to one party the actual preferences are still important for the would-be member?
That ballot form printed in the article is enough to put anyone off voting. Like Brad Barrows I reckon the anyone should feel perfectly comfortable in voting for only one candidate, or indeed none at all, which I have done on several occasions in the past few years.
Do we really want to base our calculations exclusively on the need for the elector to identify directly with their MP when, whether we purists like it or not, most people who vote tend to look for the party on the ballot paper first rather than the candidate? I still find the German system, where about half of its MPs are elected directly by FPTP and the rest by PR from party lists to be worth a punt. That’s a bit like the late Lord Jenkins’ AV Plus, isn’t it?
We have had three sets of local elections in Scotland now with STV, with a fourth coming up in May, and people seem well used to it.
The alphabetic ordering is surely as big an issue with multi-member FPTP as practised in local elections in England and Wales?
Why would a party want to stand more candidates than there are places? This would simply split their first preference votes and lead to early elimination. My understanding is that STV rewards parties that stand a realistic number of candidates and then keep in touch with their supporters to ensure a favourable distribution of first preference votes.
There’s no reason so specify how many preferences an individual voter must express either. If there are, say, 5 places, it just may be that a voter’s 8th preference is the one that elects a candidate, if all the top 7 are eliminated. More likely it will be one of the top 3 that counts.
I have checked the 2020 Dáil results (rapidly so there might be some error) and in the 72 cases where parties fielded more than one candidate – nearly all Fianna Fail and Fine Gael – 42 of them were first alphabetically which is 58%. I think that might be statistically significant, but only just given the small sample.
I agree that rotating the order of candidates on ballot papers is probably a better solution but I doubt if you would find a returning officer keen on the idea. Printing of multiple ballot papers would be more expensive and making sure they are evenly distributed across the electorate would be very complicated. The counting machines would also have to be a lot more sophisticated and any hand-counts would inevitably involving a much bigger error rate.
Electoral system geeks have come up with the idea of a circular ballot paper with the names emanating from the middle like spokes. This is probably the ultimate ‘fair’ ballot paper but I can’t see it ever catching on!
@Alan Jelfs “The alphabetic ordering is surely as big an issue with multi-member FPTP as practised in local elections in England and Wales?”
In our local (English) election, I wanted to cast a vote for an enthusiastic independent which left me with two votes for a party (won’t say which 😉 ) with three candidates. It was an arbitrary selection of two of them (don’t remember if I just picked the first two). On the bright side, the independent successfully displaced one of the “safe as houses” Tory councillors, but the losing councillor may well have been the victim of random or alphabetical choice as well.
@Peter Watson: Voting is likely to be more personalised under STV than under FPTP, because voters can choose between candidates of the same party. Requiring a representative who switch parties to fight a by-election under their new colours is a bad idea anyway, as it gives too much power to party whips. Just think about it -— a serial rebel in a (reasonably) safe seat could have the whip withdrawn, resulting in a by-election which the original party would have a good chance of winning. Having said that, voters tend to notice when they are being ‘played’ by party machines and react against it (see Leyton 1965). This would be especially true in the “full” by-election scenario.
And actually, if you think about it logically, the implication that representatives are merely party ciphers means that there shouldnt be by-elections at all. If we decide to institutionalise the idea that people really vote for parties, then casual vacancies should just be filled by nomination by the representative’s original party. So any switcher would just be replaced without election by a party placeman. This would be a scary scenario that would dampen any prospects for independence of our elected representatives.
I respect the level of technical knowledge on display here, but all this thread (and many others like it) show is that there is considerable disagreement even within the ‘PR Community’ about what to propose. PR will only ever come about through the assistance of one the major parties, so our position on it – whatever that is – will have to be modified again during the negotiations. Why not just say that we have a general commitment to reform the voting system to make it more accurately reflect people’s opinions and leave it at that.
@James Fowler
“but all this thread (and many others like it) show is that there is considerable disagreement even within the ‘PR Community’ about what to propose”
The thread isn’t about PR – it’s about one form of PR – STV. The debate (not sure it’s a disagreement) concerns details of STV.
“PR will only ever come about through the assistance of one the major parties”
There you have a point – because the 2 major parties have serious authoritarian tendencies – not suited to PR which usually leads to sharing power and compromising.
As good a reason as any for trying to do more to promote PR ourselves (alongside values and policies), irrespective of which form of PR. Promoting it as “Power to the People”.
Yes, by all means ‘promote’ PR; but for goodness sake KEEP IT SIMPLE! We could start by dusting off a tape of the video that John Cleese made for the SDP in 1987. There’s much to learn from applying a bit of humour!
I suggest that the first task is to persuade people that having a more representative parliament is important. If people were persuaded of this we could sell the virtues of whatever system we prefer.
We need to remember we are up against well-funded opponents who will cite irrelevant results in other countries – we need to cite successes and benefits of the system where it is used.
We also need to find a way of fighting an establishment with access to the funds needed to spread the sort of disinformationmation that we saw during the EU referendum.
@Tom Harney “I suggest that the first task is to persuade people that having a more representative parliament is important.”
I think the first task is persuading voters that Lib Dems in government or influencing it is a good idea, then they’ll be more amenable to an electoral system which delivers that. The party needs to know what sort of electoral reform it will advocate if given the opportunity, but much of the debate seems to be putting the cart before the horse by prioritising that over clarifying and selling the party’s other policies.
‘the requirement for a minimum number of candidates to be ranked, as currently mandated in elections ……….This would mean that voters would be required to rank at least as many preferences for candidates as there are available seats for their ballots to be deemed valid.’
Err, no.
One of the advantages of STV is that it allows to voter to vote for the candidates as they wish. If they want to vote for only one party and leave the rest blank – a sort of FPTP – that is their choice. If having ranked candidates to 1 to 5, they think candidates 6 to 10 are equally undeserving, that is their choice. We could seek to persuade that they are not taking full advantage of their right to vote, but that’s as far as it goes.
Once you’ve been through an hour or more trying to rank 30 candidates for 11 seats (some internal party elections), carefully voting all the way down to ensure that the worst candidates occupy places 27 to 30, you don’t want to inflict that process to someone who already has doubts about PR.
The beauty of STV is its resonace with what ought to happen. Anyone who is curious about electoral systems will soon realise that its methodology is designed to accomodate voter choice, proportionality and retaining a constituency link. It’s working perfectly well in Scotland and Ireland without the complexities you mention though dealing with by-elections does require some thought.
If you have to spend time explaining the system to the public you have already lost the arguement. Keep things as simple as possible. First reform Local Elections not Westminster. Just propose STV voting with all existing ward boundaries unchanged. Replacing an X with 1,2,3 (Which is as easy as A,B,C – the slogan used in Northern Ireland, I believe). The discussion for politics nerds, not the public, then becomes how to change the wards, not the voting system. Once the local election system is accepted and working only then move on to Westminster elections.
MMP is more proportional AND people can understand it more easily than STV. Just saying.
Backing up my previous post, here’s an interesting debate from ERS. 2 sensible countries use MMP, namely Germany and NZ. Interestingly, MMP has become more popular in NZ since it was implemented. The comments are especially interesting. In deciding between STV and MMP it would be worth asking the question: “Which system is easier to explain to the public?” Although AV is far preferable to FPTP, I’d suggest that one of the main reasons AV lost is that the public wasn’t prepared to spend 2 minutes to understand it. You’ve got an even bigger hurdle with STV!
https://www.electoral-reform.org.uk/new-zealands-mmp-electoral-system-how-does-it-work/
@Russell Simpson
You only have to look at Scotland to see why MMP is open to being fixed by two parties colluding together. Say you have one party (call them the SNP) which knows it will win most of the constituency seats and so will not benefit from top-ups from the list vote, and another party (call them the Scottish Green Party) which is unlikely to win any constituencies but will win top-up seats. Now let the leaders of the two parties say to their supporters: vote SNP in the constituencies and Scottish Green Party on the list. In effect, this is what Alex Salmond tried to do with his Alba Party. On the actual vote shares in the Holyrood 2021 elections, this would have given the SNP and Scottish Green Party a 2/3 super-majority on a combined 48-49% of the vote. As it happened, even without any prompting the SNP won 48% in the constituencies, but only 40% on the list vote, while the Scottish Green Party won 1% in the constituencies and 8% on the list vote.
@ Lawrence Cox
I’m glad you raised the situation in Scotland. MMP as implemented in Scotland is not proportional (eg in 2011 the SNP got 54% of the seats from 44% of the vote) for 2 reasons. Firstly, breaking Scotland into regions dilutes the proportionality (NZ/Germany don’t do this). Secondly, in Scotland they ignored the overhang which not only further dilutes the proportionality but also permits the “fixing” you refer to. Once again, NZ and Germany allowed for this. Why Scotland didn’t is beyond me. The Scottish system has made it more difficult to wean the British public off FPTP, which is a great shame.
I’m against MMP too, because it’s a crude attempt at keeping FPTP then trying to ‘fix’ it with list PR, which limits voter choice, and if the list areas too large – limits accountability.
Some of the problems with the proportionality in Scotland could be improved by having more List MSPs, or having one big list, but that is not attractive for voters.
But fundamentally a big problem with MMP, or AMS, is the ability to game the system. You can try to make it harder, but it’s problem enough that people will try to game it, and time spent by voters trying to work out everyone else is voting so they can tactically vote to maximise personal influence, is not time spent scrutinising the candidates, parties or policies. Which suits parties that don’t want their actions scrutinised.
@Fiona
I have no idea what you mean by the constituency vote in MMP being crude. The party vote top up is only a fix in the sense that it adjusts for the lack of proportionality, making it proportional. Surely the constituency vote being FPTP is a benefit rather than a hindrance, given UK voters know how that works and it also preserves the Constituency MP.
Regarding your second paragraph: as I said, to make the Scottish system proportional you need to deal with the overhang, which is when the number of MPs from the constituency vote is more than the party vote qualifies.
The reason why breaking the party vote into 8 regions reduces the proportionality is that if the same party just missed out on the next MP in each region they could be as many as 4 MPs short.
Regarding “gaming” MMP (as Salmond tried to in 2021). As I said, this can only happen because Scotland did not allow for the possibility of an overhang. Which is really bizarre. I can’t imagine what they were thinking.
There are pros and cons with both STV and MMP but in my mind by far the biggest problem with either system is the fact that STV is complicated to explain in 2 minutes (and that’s all you’ve got for the average punter) and is surely the reason why Samuel felt it was necessary to post this piece.
As far as replacing resigning candidates goes, I don’t think STV provides any sensible options:
– it doesn’t require party affiliations, so the “next candidate on the list” option of party list can’t easily apply, and “party appoints successor” is even worse
– it’s non-monotonic and non-ordering so any attempt to recount the original could lead to some very weird results (not to mention requiring a lot of ballot papers storing)
– a single AV runoff would break proportionality, which means that some resignations are much more impactful than others (especially those of the smaller parties).
– ERS have suggested “just leave the place vacant, you’ve got plenty of others” which works fine for annually-elected positions but would I think be considered unacceptable for MPs.
This is definitely one area that a party list system – open rather than closed, for preference – has advantages over STV.
It would be fantastic to have a choice of MP’s to go to for representation, especially where that included a choice of several parties as would normally be the case. There was a precedent for multi member constituencies, as they used to exist into the 20th C under FPTP.
There might be a case for a party top up list under STV to make it more proportional. It could still operate under a 5% entry threshold as in Germany, setup to avoid fascist fringe parties getting elected.
@John
There may be good reasons for not admitting parties with less than 5% (eg Germany/NZ) of vote but I’m not keen on the reason being that they are not parties I agree with (assuming they are not breaking the law). I heard people say that it was good when UKIP got no seats despite getting over 10% of the votes. I’d say, let UKIP have their 75 MPs. I’m sure it was partly their frustration with the electoral system that allowed Brexit to happen. Maybe it’s better to allow fair representation so their views can be argued against. Taking away their representation won’t make their ideas go away.
@cim
The fact that STV enables voters to spread their preferences around different parties’ candidates and independents is I would think one reason why the tories and labour don’t like it. It takes power away from the parties and gives it to the voters. It does away with safe seats – all the candidates have to work at their campaigns.
@Russell Simpson
If a system has ‘perfect proportionality’ – i.e. even with a tiny percentage of the vote a party can win seats – there might be a great many fringe parties in the parliament. I was aware that for the Knesset (120 members) the whole of Israel is one constituency. I had thought there was no threshold but according to https://www.gov.il/apps/elections/elections-knesset 19/eng/about/ElectoralSystem_eng.html there’s a 2% one.
If the system needs a threshold then a very low one resulting in candidates from a lot of parties being elected it might lead to some strange bedfellows struggling to form a government.
STV doesn’t need a threshold. If the country is broken down to a large number of multi-member constituencies each having a small – say 3-6 – MPs then a candidate with a tiny % of first preferences is unlikely to pick up enough transfers to be elected. Seemples as those meerkats would say….or used to say…