Monday 29 March 2021

Gaming the system


One of the myths on which British semi-democracy is built is that voters elect individuals, not parties. Most of us know that it isn’t true, as old stories about donkeys wearing the right colour rosette suggest, but the myth stems from the distant past when MPs formed or joined parties only after being elected. Myths have consequences though, and one of the consequences of this particular myth is that we have two votes in elections for the Senedd. The d’Hondt system of proportional representation, which is intended to use list members as a means of partly correcting the inevitably unrepresentative outcome of using first-past-the-post in the constituency section, doesn’t actually require that we all vote a second time for list members. It would work equally well (and in some respects, rather better) if the votes by party in the constituency section were simply tallied up and used to allocate the list members. The fact that it isn’t used that way is largely because the constituency votes are regarded as having been cast primarily for individuals and only in a secondary sense for parties.

One of the results of basing the list membership on casting a second vote is that it allows – or even encourages – people and parties to try and game the system. Parties can win list seats without even contesting constituencies, something which can hardly be considered to be a ‘correction’ to their under-representation under FPTP. It can also encourage parties to talk about ‘wasted’ second votes, because a dominant party (such as the SNP currently in Scotland, or the Labour Party historically in Wales) in the constituency section is unlikely to win many seats in the list section. In practice, it’s an extremely difficult calculation for an individual elector to make; it’s impossible to be certain about the results of the constituency vote, whatever the polls might say. And tactical use of the second vote can work both ways – it can never be limited to supporters of only one side in a debate. Results of elections to date do show differences in voting patterns between the two votes, but the degree to which the net figures are different is limited. Some of the difference will be accounted for by people voting for their first-choice party on the list in constituencies where that party doesn’t stand a candidate – and some tactical votes will simply cancel each other out.

The announcement by Alex Salmond that he has formed a new party with the express intention of only standing candidates in the list section, whilst encouraging voters to support the SNP in the constituency section, is an attempt to win what he calls a ‘supermajority’ for independence in the next Scottish parliament. Whether it works or not depends on a range of factors, including his own personal popularity (which some polls suggest may not be as high as it has been at times in the past). It’s a gamble (but then he’s always liked a bit of a flutter on the horses). At one extreme, he might just pull it off, but at the other, splitting the pro-independence vote in the list section might result in fewer pro-independence MSPs overall, and end up sabotaging, or at the least delaying, the independence project by denying seats to the pro-Indy Green Party. In the one case it would look like a triumph for a master strategist, at the other like nothing more than a failed vanity project.

Let us assume, for a moment, that it works – i.e. that the SNP pick up most, or even all, of the constituency seats on around 50-53% of the vote, and that all or most of those voters switch to Alba for the second vote, giving that party most of the list seats. Independentistas might well be delighted – but would such an outcome be fair? Winning a supermajority of 80 – 90% of the seats in the parliament on the basis of 50-53% of the vote in each of the two sections merely ends up replicating the problem that d’Hondt was supposed to solve, and disenfranchises many of the 47 – 50% who do not vote for pro-independence parties. It’s perhaps a little unfair to criticise politicians for using the rules as they stand to their advantage, and there’s a rather delightful irony about independentistas using rules imposed by Westminster against continued Westminster rule, but that doesn’t make it fair or reasonable.

Let’s put it another way. It’s an unlikely scenario, but suppose for a moment that the Tories and Lib Dems decided not to contest constituencies, encouraged their supporters to vote for Labour instead, and then carved up the regional lists between them. A movement of just a few percentage points in the polls could then turn a supermajority for independence into a supermajority against. How would independentistas feel about that, equally unrepresentative, result? As I said, it’s a highly unlikely scenario and depends on an assumption that the Tories and Lib Dems would elevate the desire to maintain the union above all else (although it doesn’t even require the complicity of Labour). The circumstances in which parties can effectively game the system to produce an unrepresentative result are rare, but they arguably now exist in Scotland, if not currently in Wales. The whole Alba Party strategy depends on that.

I certainly want to see a large pro-independence majority in the Scottish parliament – and in the Senedd – but I want that to be a reflection of a large pro-independence majority amongst the electorate, not a reflection of a flawed electoral system. Winning independence is about winning people over to the concept, not about tactical manoeuvres with election rules. If the rules can produce a flawed result, it’s time to review the rules, and either make the allocation of list seats directly reflective of the constituency votes or else, even better, replace the list with a proper system of proportional representation, such as STV. Either way, freeing the political process from the silly myths of the past is long overdue.


CapM said...

I think it most likely that it's a Salmond vanity project, succeed or fail.
Given the way the numbers stack up in Scotland at the moment I don't see much of a potential gain to be made by those who want independence whether they support the SNP or not as a result of Salmond playing the system.

I wonder what the effect of a similar initiative in Cymru would be. As you point out both sides can play the system. Have the Conservatives and UKip already done so?

Jonathan said...

CapM "effect of a similar initiative in Cymru?" - We do have something similar - Neil McEvoy and Propel. Same logic exactly. List seats add to the total of constituency seats. But Neil is not Alex, obviously. So we'll see.....

John Dixon said...


There are some superficial similarities, certainly, not least the apparent belief in the power of personality over party, but Propel are standing candidates in constituencies as well as on the list, and a strategy which might work (and only 'might', even then) in a situation where the electorate is highly polarised into two camps (pro and anti Independence) with a narrow margin of separation looks doomed to fail in a situation where pro-Indy parties remain very much a minority. I suspect that the circumstances in which such a strategy can work depend on there being:

1. a single clear issue (and, generalising, it doesn't have to be about independence) where two parties, or groups of parties, take a very different stance;
2. an electorate which largely or entirely rates that issue as the single most important determinant of their vote;
3. a willingness amongst that electorate to switch en masse from one party in one vote to another in the second vote.

It is arguable that such a situation might currently exist in Scotland (although I'm far from certain, personally); it is a very long way from being the case in Wales.