We’re now exactly three months since May’s local elections, and I am delighted (and exhausted) to say that I’ve finally finished compiling all of the detailed data from all 32 councils. You can find what is hands-down the most comprehensive look at an election in Scottish history via the links on the general LE22 landing page here. Having pulled all that data together at the local level, there are some interesting stories it can be used to tell at the national level too.
One of the defining features of the Single Transferable Vote system used for local elections is, very obviously, that it’s preferential. Given how different it is to FPTP, it’s often explained to voters via a simple line to the effect of “STV, it’s as easy as 1, 2, 3!” What has become clear through my data collation process is that in Scotland, it might actually only be 1, 2, 3.
Quite early in the process, I’d spotted that on the rare occasion a party stood three candidates, the rate at which their voters then transferred to any other party cratered. I did a little thread about it on Twitter, during which I realised what I was really finding was how few people were using a fourth preference at all. That prompted me to start collecting the data for preference use as standard.
Overall Preference Use
With every single ward neatly in my spreadsheets, I could check how many voters who’d had a given preference available to them used it – e.g. wards with only 6 candidates are ignored for working out how many people used a 7th preference. That looks like this:
It turns out that preference use drops sharply right from the get go, with nearly 13% of voters not even bothering with a second preference. Third preference use falls a few percentage points short of two-thirds of valid votes. After that, it falls right off a cliff edge, which I’ve highlighted by shifting to the pink tone. The overwhelming majority of voters who had the option to do so did not go beyond their third preference – it’s not that much more than a quarter that use a fourth, only a smidge more than a sixth use a fifth, and so on.
Interestingly enough, that does suggest greater use of preferences than the local poll I commissioned before the elections suggested. That had found 76% of people would use a second preference, and only 55% a third. That said, we know that the headline figures for the SNP and Independents in particular were some way off from that poll, so the later preferences could similarly have been impacted by the fact it just seems to be hard to poll for local elections.
Preference use matters because there are plenty of instances where a very small number of voters just going that little bit farther down their preference listing could have changed the result. A particularly dramatic one comes from Highland’s Inverness West ward, where Labour end up 0.3 of a vote behind the Conservatives, dropping out of the race and leading to a Green councillor. Had that been the other way around, Conservative transfers would easily have seen Labour elected instead. The Greens have something of a trend with 0.3 votes this year, as the closest result of any ward was in Argyll and Bute’s Oban South and the Isles, where they fell short at the final stage by that number.
Given how important preferences are to the system, what’s going on that is leading to people using so few of them?
No Numbers Please, We're Scottish
One possibility at the forefront of my mind is quite simply that voters still don’t entirely grasp how the system works. Referring back to my poll, I had asked if people felt they understood Scotland’s various voting systems. A total of 58% had claimed to have a good understanding of STV (of which 14% claimed a very good understanding), but they may have been a more engaged segment of the electorate. Also, sometimes people think they understand something but haven’t actually fully grasped it.
I’ve said before I think the fact that Scotland uses a different voting system every time we go to the polls is confusing for people. If you’re only using a system once every five years, barring by-elections, you could be forgiven for not being entirely up on how it works. STV is also the most “alien” of the voting systems we use, requiring voters to mark multiple candidates by number rather than a simple cross in one box. That may also be why the rate of spoiled ballots is abnormally high at 1.85% this year, versus 0.37% (constituency) and 0.19% (list) at Holyrood last year.
Voters may also only be half-reading the ballot paper instructions (as in the example ballot below, from Wikipedia). Although these do tell people they can use as many or as few preferences as they like, they also take the same “1, 2, 3 and so on” approach so many of us do to explaining STV. It’s possible that a lot of voters giving the instructions a quick glance end up only absorbing the 1,2, 3 of it.
I don’t think this can possibly be the whole of the explanation, but I also don’t think it’s a coincidence just how steep the drop in usage is between third and fourth preferences. At no other point on the chart do we see a drop of over half of the people who had used the previous preference choosing not to do so. A more “natural” feeling distribution might have went via the low 40’s, or at least the mid 30’s.
It's Already Giein Me the Boke
Moving to another possible contributing factor, the lack of preference might suggest that the “Vote ’til You Boke” messaging we’ve increasingly been borrowing from Northern Ireland didn’t have much cut through. That shouldn’t be too much of a surprise though, as ordinary voters don’t live in the social media spheres this stuff comes up in. However, it may also show that Scots were pretty quick to boke!
Plenty of voters absolutely will have made an active choice to stop where they did, even knowing full well how the voting system works. I myself was one of them, effectively taking advantage of my BBS knowledge as to how my ward would go to only mark a very limited selection of preferences out of sheer grumpiness. I also lacked truly “objectionable” fringe parties on my ballot, whereas their presence can be the kind of thing that makes voters keep marking preferences for other, less objectionable, candidates out of principle.
Indeed, in many wards people didn’t even have had a full suite of even non-objectionable parties to transfer to. Only just under half of wards (175 of 355) had all five Holyrood parties present, for example, with the Lib Dems and Greens having quite a wide spread of absences. Especially given that the country remains deeply constitutionally split and that seeps into everything, you can imagine things like a Conservative voter with no Lib Dem option stopping after they’d marked all the Conservative and Labour candidates – though a follow up piece tomorrow will look at how plenty of transfers did still go cross-constitutional.
Regardless of whether it’s by accident or design, it’s clear that voters in Scotland are not yet taking full advantage of STV’s transferrable nature. A lot of the discourse after the election focussed on the number of spoiled ballots and whether the alphabet effect, whereby same-party candidates higher up the alphabet are more likely to be elected, was a problem. Not much was said about how much use voters were making of preferences, perhaps because it’s harder to get to that detail.
However, perhaps we should be talking about this. Given that transferability is one of the key arguments for STV, there are questions to be asked of whether that’s actually proving as advantageous or worthwhile as the system’s strongest proponents blithely claim. As ever, I emphasise STV is vastly preferable on a democratic basis to FPTP, but this adds to my general belief that it’s not quite as wonderous as often claimed.
If you find this or other Ballot Box Scotland output useful and/or interesting, and you can afford to do so, please consider donating to support my work. I love doing this, but it’s a one-man project and takes a lot of time and effort. All donations, no matter how small, are greatly appreciated and extremely helpful.