Some Idle Thoughts About Voting

So apparently there is some kind of election thingy happening on 6th May.

So apparently there is some kind of election thingy happening on 6th May.

“Many forms of Government have been tried, and will be tried in this world of sin and woe. No one pretends that democracy is perfect or all-wise. Indeed, it has been said that democracy is the worst form of government except all those other forms that have been tried from time to time.”
– Sir Winston Churchill, 1947

I’ve been reading thoughts from Peter Kellner and Mat Bowles on the issue of polling and turnout.

First, I have to admit I find the concept of ‘swing’ rather discordant.  Kellner analyses the race in these terms -the election is an iterative equation and not an isolated event. I am reminded of the Monty Hall ‘Game Show’ thought experiment, where there is a prize-goat behind one of three doors.1 After Monty, the host, has revealed one of the bogus doors, the chances of the goat being behind the door you did not choose is 2/3… or so the Mathematicians say, ignoring the fact that something has happened in between. Likewise with elections – the concept of ‘swing’ suggests that this election is merely a function of the previous one.

Yes, yes, I know: Elections are a function of previous outcomes. Voters have an after-the-fact loyalty to the person they voted for last time, for example. The memory of the brutal Thatcher years, or even The Winter of Discontent, still has influence in 2010 when many (if not most) of the voters don’t remember them first hand. Still, like voting along ethnic lines, the fact that the starting positions on the electoral board are skewed doesn’t seem like the ideal of democracy.

All this means that the contest is already over in 382 seats, according to the Electoral Reform Society.  This inspires people in those constituencies to stay at home.  Worse, it nudges the politicians in those constituencies – both those destined for victory, and those who know they will lose – to campaign elsewhere.  So begins a vicious cycle of disenchantment for the electors, and a disconnect between them and their MPs.  Add to that the Heisenberg effect of polling (i.e. measuring voter intentions might actually alter voter intentions) and you can make a strong case that all these pesky statistics actually serve to discourage voting.

How does PR or STV change the equations?  Or must we, in the end, return to Churchill’s quote about democracy and make the best of a bad job?


1.  On reflection, I think in the original formulation of the puzzle, the prize is a car and the goat is the bogey-prize.  I prefer my version though.  A goat in the back garden is a handy alternative to lawmower, and we all like feta, don’t we?

3 thoughts on “Some Idle Thoughts About Voting”

  1. Just to say – if “many if not most” voters don’t remember the Thatcher years first-hand, that would mean that there’d be more voters who were born between 1979 and 1992, say, than there are who were born in all the years between, say, 1910 and 1979 put together. I don’t think that adds up, Rob, even accounting for dementia, repression, and post-traumatic amnesia in the latter group.

    Also, I think the notion of swing simply implies that elections are a function of the period between elections, rather than the previous election. As indeed they should be.

  2. How do STV or PR change things? Simples…

    It’s a well observed truism that Liberal Democrats do terribly in PR based elections (European, London Assembly, Scotland and Ireland). This is because (and I’m arguing this internally very strongly) they continue to concentrate their campaigning resources in their target areas.

    I was asked to leave my part of Yorkshire to go campaign in various target seats as that’s where we’d pick up votes.

    How utterly utterly wrong headed. I refused.

    In an FPTP election, it makes sense, on a limited budget, to abandon support in areas you’ve little chance to concentrate on the seats you need to gain to get in.

    That means politicians chase the median voters in swing seats, almost to the exclusion of all else.

    But in a PR election, especially in STV, every single area has the potential to gain you seats, especially when you take into account the palpable fact that many of your supporters in non-target areas will go and vote tactically. The Lib Dems do this to supporters of other parties all the time with their squeeze message bar charts, other parties do it now as well.

    It will take time to see the full effect (about a decade, at least two elections, is normal), but when you change a voting system voters and parties adapt and campaign differently.

    In PR, I need to identify as many of my supporters as possible across the whole area, and get them out to vote.

    In FPTP, I need to persuade as many voters as possible within the small district that a) I’m one of the top two candidates and therefore in with a chance and b) I’m a lot better than the other guy.

    So I chase the median and run negative attacks. STV should (theoretically) increase turnout over time and decrease negativity over time.

    But the changes only really work if all elections are run on a similar system, while the most important election is FPTP, it distorts everythine else (hence the godawful LD campaigns mentioned earlier).

    Did I say simples? Um, yeah, that’s the brief summary.

    (BTW, can you change the link to point to the version of the post on Dreamwidth? I’m using the old LJ blog as a backup these days, no longer trust the owners or business model)

  3. Rob, the game is: You are offered three doors. Behind one is a sports car, the others a goat. You select your door. The host opens another door revealing a goat. Is it better to keep the same door or open the other? The way to think about it is to imagine a million doors and 9,999,998 are opened after your initial choice. /EndPedant

Leave a Reply

Your email address will not be published. Required fields are marked *