Sunday, February 17, 2013

Creative accountancy on Twitter

Spotted on Twitter earlier this morning -

Katherine O'Donnell ‏: Comparison with Nov 2012 Ipsos MORI poll suggest Scots keener than the English on EU - 50% of English wanted out, 53% of Scots want to stay

Brian M K Watson ‏: Meaning 47% of Scots wants to leave EU. 3% is nothing. Scots just as pro/anti EU as majority of England.

Which of course is just plain wrong. Here are the actual figures from Ipsos-Mori -

Scotland :

53% would vote to stay in the EU
34% would vote to leave the EU

England :

42% would vote to stay in the EU
50% would vote to leave the EU


Mr Watson is making the all-too-common schoolboy error of lumping in the 'don't knows' with the supporters of his own preferred position. Of course it's understandable that some random bloke on Twitter might make that mistake, but what is truly unforgivable is that the mainstream media do exactly the same thing, over and over again. How many times have we all seen the ludicrous claim that "two-thirds of Scots oppose independence" paraded as fact? Given that no Yes/No poll on independence has produced anything close to a 66% No vote, that figure can only be arrived at by counting as "anti-independence" those who tell pollsters that they don't yet know whether they will vote Yes or No to independence, or that they won't vote at all.

'Misleading' doesn't even begin to cover it.

The most irritating example was surely when Michael White (if ever a man was so condescending with so little justification I'd like to know about it) informed Joan McAlpine that the fact that 40% of Scots in a poll supported independence meant that she still needed to persuade another 10% to vote Yes. Er, no, Michael. George Cunningham rules do not apply here, and abstainers do not count as No voters.

* * *

On the subject of incredibly daft things said on Twitter, Duncan Hothersall reacted with utter fury to my blogpost on Friday pointing out that Labour's vote had fallen by 4% in the Rutherglen South by-election. He started by describing my entirely factual post as "idiotic", and then went all Alice in Wonderland on us as he manfully struggled to explain how a 4% drop in Labour support was not in fact a 4% drop in Labour support. His first claim was that the result of the by-election could not be compared to last year's election in the same ward, as different electoral systems - AV and STV respectively - had been used. (On this point he was egged on by Douglas McLellan, who as a Lib Dem really ought to know better.)

In fact, of course, the exact same electoral system was used in both elections. It's true that STV works in a strikingly similar way to AV when there is only one seat to be filled (as is almost always the case in by-elections), but that doesn't change the fact that it's still STV - complete with a quota, which is not a feature of AV.

Duncan's next tack was to claim that Labour's 39.9% share of the vote in the by-election did not represent a 4% drop on their 43.9% share in last year's election in the same ward, because (and I quote) the 43.9% figure was "not real". From what I can gather, this means that people who voted Labour last year were not "really" voting Labour because they were voting for two different Labour candidates, not one. But hang on just a minute here - does that mean, for example, that we can't say that 43.2% of the British electorate voted Labour in 1997, because those people were voting for 639 different Labour candidates, rather than just one? Incredibly, it appears that Duncan's answer to that question can only be 'yes' -

"This isn't a vote for a party slate though, is it. You're getting confused with AMS."

So there you have it, folks. If you're not voting for a Labour "party slate", you're not voting Labour, and your vote should not be added to the Labour tally. As there are no party slates in UK general elections, that means there is no such thing as "Conservative votes" or "Labour votes" - there are just "votes".

I have grave fears that this revelation will make John Curtice's head explode.

3 comments:

  1. 'I have grave fears that this revelation will make John Curtice's head explode.'

    One can but hope...

    ReplyDelete
  2. You've got to hand it to Duncan, he's got an awesome talent for spin. Very tricky to catch out, but always possible.

    ReplyDelete
  3. Maybe he was tired after spending 24 hours every day on twitter, and doesn't understand 1st preference votes are still 1st preference no matter if it is AV or STV.

    ReplyDelete