Comments on: Campaigning is Dead, Long Live Campaigning!

By: Steve

Steve — Mon, 27 Jul 2009 10:22:18 +0000

There seems to be one thing wrong with your analysis. You want to compare the votes for a single party in the same circumstances with and without leaflets. This is not easy without for example delivering leaflets to only half of the constituency. You could compare adjacent constituencies but this introduces more confounding factors that quite plausibly could swamp a significant leaflet effect.

Your method includes huge compounding factors and seems to me to be unlikely to provide reliable conclusions.

The best approach would seem to be a survey of a large number of constituencies coupled with some sort of factor analysis. But then there are technical problems with factor analysis.

By: Ed Butt

Ed Butt — Sun, 26 Jul 2009 15:46:07 +0000

Campaigns have little influence now, I believe it’s how people come across on television that really matters. So:
Conservatives; Cameron = Inspector Linley, William Hague =- Jim Bullseye Bowen
Boris Johnson = Worzel Gummidge.

Labour: Brown=Bluto,
Mandy = Dick Dastardly
Harriet harman = Cruella DeVil
and
Alan Sugar = Alan Sugar

Who’s going to be most popular

By: Stu

Stu — Sun, 26 Jul 2009 09:30:35 +0000

Parasite, it doesn’t. The ‘equation’ is (an attempt to be) satirising those who look upon quantity of leaflets as a measure of a campaign (cf. Nich Starling’s otherwise fantastic election coverage, and Costigan’s quoted remark), where really they’ve made nary a difference to the outcome.

As for how I got the figures to make the three pie charts, it was an extremely technical scientific process of multiplying opinion poll result by the number of leaflets delivered by the party divided by the total number of leaflets delivered by all the parties.

The entire point I was making is that you can’t do that, because leaflets hardly seem to affect the outcome at all.

they want a political party to stick up for them, fix cracked pavements etc

Yes! Exactly! NOT a party who puts a million leaflets through their door. Hence the movie analogy – you can do all the advertising you like, but unless you come through, you won’t win out in the long term.

I suppose the logical conclusion of that is for parties to start focusing more on delivering whatever they can whenever they can deliver it, rather than campaigning. Maybe.

By: Parasite

Parasite — Sun, 26 Jul 2009 08:57:25 +0000

Do not see at all how the existing vote share plus leaflet share equals projected result, eg. how a Tory 35% + 34% equates to a 52% projection.

I hate to be the stick in the mud but leaflets and also canvassing are a proven method, particularly if done over a long space of time not just in a four week by-election campaign (Chloe Smith having been in place since early 2008 probably helped here). It seems to work in my patch a damn sight better than the public agreeing with one’s ideas: they want a political party to stick up for them, fix cracked pavements etc, not fight the ideological war by proxy on their doorstep.

By: Stu

Stu — Sun, 26 Jul 2009 07:36:54 +0000

Richard, the point was that the leaflets don’t have an effect either way. I think you’re misinterpreting my tongue-in-cheek attempt to show that Costigan was wrong to suggest that ‘putting out a load of leaflets’ was a tactic which made a difference in the by election as some kind of deep, serious statistical analysis (which, for the record, it isn’t).

I created the scatter graph you suggest, with ‘percentage change’ rather than ‘points change’. Suffice to say it looks pretty much the same except that UKIP go off the scale

As for quality/quantity, do you really think the UKIP leaflets were that much better than all the others?

By: Richard

Richard — Sun, 26 Jul 2009 01:19:17 +0000

It must be said that your statistics are awful. Your conclusion might be right, but UKIP’s share of the vote did not rise by 7%, it rose by 140%, because it went from 5% to 12%. Conservative gained 14%, Labour lost 42%, Greens lost 29% and the LibDems lost 7%.

Using the number of percentage points gained, rather than percentage of the forecast voting share gained, renders your statistics meaningless, as it takes no account of inherent likelihood of people wanting to, feeling able to, or feeling it is meaningful to, vote for a party. Using the correct figures the graph shows no greater correlation though, unless you are the BBC (i.e. you pretend UKIP doesn’t exist).

Labour and LibDem, were criticised quite widely for putting out misleading leaflets. Both lost share. So maybe with leaflets quality is more important than quantity.

By: Quaequam Blog! » The trouble with Rennardism (clue: it isn’t the leaflets)

Sun, 26 Jul 2009 01:17:49 +0000

[…] most attractive candidate to the electorate nearly always wins irrespective of resources.” Stuart Sharpe seeks to prove this with graphs. But I have to say that when it comes to winning tactics, I’d […]

By: Charlotte Gore

Charlotte Gore — Sat, 25 Jul 2009 21:43:19 +0000

Stu, this is brilliant work.

By: First Class posts on Saturday | Letters From A Tory

First Class posts on Saturday | Letters From A Tory — Sat, 25 Jul 2009 20:47:45 +0000

[…] Sharpe’s Opinion say leafletting before elections is […]

By: Stu

Stu — Sat, 25 Jul 2009 17:54:29 +0000

Oh gosh no – don’t get me wrong! Good ideas don’t win campaigns! Popular ones do. And you can’t tell what’s going to be popular in advance, popular opinion is always wrong, and what the people want is not change, per se, but moderate centrism. (To be fair, that’s an easy logical point to prove – if nobody wanted centrism, then centrism would be extremist. By definition.)

Of course, I’m being defeatist – but then again that’s why I’m not a party member, of any party.

By: Costigan Quist

Costigan Quist — Sat, 25 Jul 2009 17:04:30 +0000

Nice piece. I agree completely that leaflets are only one element of winning support and, of course, all the evidence tells us that the voting intentions people have going into any election campaign are by far the best predictor of what happens at the end.

But unfortunately, I’ve seen little evidence to support your theory that having the good ideas wins campaigns either. I wish they did; I’d love to be in a world where they did; but I’ve not seen evidence that good ideas are especially more effective than lots of leaflets or anything else.

By: Stu

Stu — Sat, 25 Jul 2009 11:38:45 +0000

Says the person who won’t put a name against their comment? Interesting…

By: Anon

Anon — Sat, 25 Jul 2009 11:06:56 +0000

“the Libertarian Party… don’t ‘campaign’ so much in the traditional sense of the term, they have well-read blogs and send out emails and use Twitter to get their point across, and look what happened to them – 36 votes.”

People like to blow off steam on blogs, so it’s no surprise that the bolshy libertarian crowd get a fair on-line audience. But the party… what emails? Their communication abilities are zero. Can’t even organise the on-line constituency where they are best represented. Truly, truly pathetic.