AMS – Additional Member System, used for the Scottish Parliament, a mixture of FPTP and PR
FPTP – First Past the Post, used for Westminster and Scottish Parliament Constituency elections
PR – Proportional Representation
STV – Single Transferable Vote, current policy, and used for council elections
How would the proposed electoral system work?
My motion effectively calls for us to adopt a form of Open List Proportional Representation modelled on the systems used in Scandinavian countries. There are differences to exact operation in each country, but the basic features are;
- Most seats are directly elected from multi-member constituencies
- A portion of seats are elected on the basis of the overall vote share, but allocated to the constituencies
- To be elected to seats based on the overall vote share parties need to cross a % threshold
- Party lists are open, allowing voters to specifically vote for individual candidates rather than simply electing candidates in party-defined order from the list
These systems are therefore highly proportional, whist giving voters substantially more say over who is elected than the closed list system we currently use as part of AMS in the Scottish Parliament. An example that approximates the 2016 Scottish Parliament election to a Scandi-style system is below, though it’s important to emphasise the word “example” there – the policy motion is not as specific as to call for this exact set of boundaries.
The focus of explanation in this post is Scottish Parliament Elections, but the basic principles are just as applicable to Westminster. Local, rather than parliamentary, elections are a slightly different issue. Local elections in Scandinavian countries generally use a much simpler system of “at-large” lists – just treating a council area as one big constituency and not dividing it up. That works there partly because their councils are much smaller than ours (they do local democracy properly) and because they don’t have a culture of expecting hyper-local representatives within those councils. Rather than confuse matters at this point by including that caveat in the policy, I opted just to suggest the parliamentary Scandi-style as a blanket system for clarity.
Something not outlined in the policy due to complexity is the possibility to have more local areas represented underneath the formal constituencies. Denmark has 10 constituencies for its parliamentary elections, but around 90 “nominating districts”. Parties can and do submit lists with candidates in different orders for each nominating district within a constituency, giving their “lead” candidate for that district a sense of being particularly local to that area. So even though the formal constituencies may seem big, there’s a lot of flexibility at the back-end of the system to deliver more local representation.
The policy also does not state what formula we’d want to use for allocating seats – D’Hondt, Sainte-Lague and Largest Remainder being some commonly used ones. I view that as a “legislation not policy” level of detail that wasn’t necessary to specify and can instead be left up to our representatives’ judgement in any reform process.
Finally, the focus of the policy is on delivering proportionality between parties, and it is therefore silent on Independent candidates. Again this was primarily a matter of keeping policy simple rather than trying to draft legislation that covers all eventualities. An easy way to make the system Independent-friendly would be to allow them to bypass the percentage threshold.
Why replace our current policy?
As with most supporters of PR in the UK, our current policy supports STV. Arguments in support of PR often point to our European neighbours as an example – yet only two of them (Ireland and Malta) use STV. List based systems are used in the overwhelming majority of European countries – and indeed in most countries with PR worldwide. We should be following that general trend rather than advocating a system that is better suited to elections within organisations than to public elections.
Since STV is the most commonly argued for PR system in the UK, it’s often assumed it is the most proportional system. It’s also regularly claimed to eliminate wasted votes, and to give voters the most choice over who is elected. Whilst it fares better on all of these fronts than FPTP, these claims are all overblown, and the proposed policy is better than STV on all three of these measures.
More Proportional than STV
STV actually suffers from the same problem that FPTP and AMS – if you only use the results in small chunks of an overall area, a small disproportionality in each one can add up to a large disproportionality overall. In PR systems STV is often the worst for this, given the relatively small number of people elected per area. Our current policy calls for 5 member STV, where the quota to win a seat is around 16.7%. A party can easily win 5% or even 10% of the vote under such a system, but no seats.
As one of the smaller parties in parliament, how well we’d do under any system can be a useful indication of how proportional it is. In 2016, we got 6.6% of votes – but for those same votes under STV, we would likely have won between 2 and 5 MSPs (1.6% – 3.9%). That’s fewer than the 6 (4.7%) we elected with AMS.
As a Scandi-style system uses the overall vote share to allocate some seats, it’s much more proportional. We’d have won 9 seats (7.0%) in 2016 with such a system.
Fewer Wasted Votes than STV
STV has far fewer wasted votes than FPTP, but it still has a large number just as a consequence of how it works. Remember a wasted vote is any that doesn’t count towards electing a candidate. This can be because they got more votes than they needed to win, or because they weren’t elected. STV does allow votes to transfer in these scenarios, but many are still wasted due to the quota.
Since the quota for a seat in 5 member STV is just under 16.7%, only around 83% of votes are needed to fill all 5 seats. The remaining almost 16.7% of votes will always be wasted, regardless of what’s on them. Even more votes can be wasted if many voters use few preferences. Although electing more people from one area reduces the quota and thus wasted votes, that erodes some of the other stated advantages of STV, such as that it allows local representation.
Again, as a Scandi-style system uses the overall vote share to allocate some seats, it will often do a far better job of minimising waste than STV. Looking at the 2016 election again, only 5% of votes would have went to waste.
More Candidate Choice than STV
STV is more candidate-centred than the list systems we currently use in Scotland, but it still has major constraints on how much influence voters have over who is elected. Parties usually only stand as many candidates as they think they can win, as standing too many risks accidentally splitting their vote too widely in the early stages of a count.
For a small party like us that usually only stands one candidate, if a voter doesn’t like that candidate, their choice is exactly the same as it is with FPTP – knowingly vote for a bad candidate, or risk their preferred party not winning enough seats. Even for parties standing multiple candidates, a similar pressure applies. If you want a Labour government and they are standing two candidates, choosing not to preference one you think is bad will damage their chances.
The open lists in Scandi-style systems allow voters are able to specifically choose between candidates within a party’s list. Unlike STV, there’s no risk that a party can stand “too many” candidates and damage their own chances. In fact it’s wise to stand more candidates than you think you will elect just in case you have a very good election or an elected candidate vacates their seat later. This gives voters about as much choice as it is possible to do in an electoral system.
I oppose Amendment 1, which has very flawed logic. It relies on there being an inconsistency between electoral reform “not be to the advantage of any single party” whilst differentiating between parliamentary and non-parliamentary parties for ballot access purposes. This isn’t the case – “any single party” is intended to be taken quite literally. A provision which would treat (currently) 6 parties the same clearly isn’t designed to advantage any single one of them.
It’s common practice in other European countries for parliamentary parties to be entitled to re-stand, but for new parties to have to go through a registration process (I note again though a “policy not legislation” vagueness in when signature collection needs to be done and for how long registration lasts). Parties that hold seats have already been effectively approved of by voters. Rather than effectively requiring voters to register that approval all over again for parliamentary parties, the most sensible “treat all parties the same” measure would be to remove ballot access requirements entirely.
I’d therefore advise voting against this amendment, allowing standard European practice to be added to our policy, and if anyone feels strongly about it, to bring a motion next year which modifies the policy to state we want to abolish deposits and any other barriers to ballot access.
I’m happy with Amendment 2, as the ranked list of preferred voting systems was only included as a totally cynical measure to try and get more people to vote for the motion by still mentioning STV… What? I’m being honest!