miss_s_b: (Politics: Democracy)
[personal profile] miss_s_b
I am utterly sick of people attempting to read in the entrails of the local election results what might happen in the general election in 2015. Here is a (non-exhaustive) list of why such efforts are fruitless bullshit:
  1. People vote differently in local and general elections. Recognising that the local council has very little power any more, people take council elections far less seriously. This means that turnout is derisory and the result is thus distorted by the people who DO turnout being either seriously committed to one party or another, or wanting to "send a message" by spoiling their ballot or voting for a party they know cannot win.

  2. Local elections were not held in every parliamentary constituency. This means that any extrapolation to a national result is "projection" - i.e. guesswork

  3. Time will pass between now and a general election. Political change tends to be evolutionary rather than revolutionary, but sometimes there IS a revolutionary change, and none of us knows what will happen in the next two years. The old adage that a week is a long time in politics still holds true in many cases.

  4. Data which applies to our electoral system is incomplete, verging on non-existent. The only polling data we have which is reliable at constituency level is from previous general elections. No polling company holds/collects data at constituency level, and most of them hold/collect data at a national level. This is USELESS for predicting the results of a general election under the first past the post electoral system, where every single constituency has its own idiosyncracies. We will never get a British Nate Silver or change the woeful quality of political analysis in the media unless this changes.

  5. The pundits in the media are crap. Relying on the proclamations of media pundits is an exercise in futility when none of them even acknowledge that they are guessing from incomplete data.
I'm sure there are other reasons that could be added to this list, but you get the idea. Basically, if anyone tells you that they can predict what will happen in May 2015 after this week's local election results, your bullshit detector should be pinging off the top of the scale.

Date: Saturday, May 4th, 2013 04:00 pm (UTC)
matgb: Artwork of 19th century upper class anarchist, text: MatGB (Default)
From: [personal profile] matgb
Local elections were not held in every parliamentary constituency. This means that any extrapolation to a national result is "projection" - i.e. guesswork

Nice.

Estimation would be fair, but guesswork? No. Not in the slightest. The projections of, say, Rallings and Thrasher, which is what the BBC use, have perfectly valid methodology which has been improved since I spent, oh, two years studying it as part of my degree.

Scientifically valid sampling is used to identify key sample wards, sometimes even individual polling stations, that can be used to extrapolate because they're representative. Basically the same methodology as the exit poll they use for the General Election-the one that almost got the LD seat numbers perfect when every pundit was expecting a massive breakthrough.

Slightly disappointed to see the word 'projection' for this sort of analysis so utterly dismissed, it's used in a fair few fields of science afterall-and while the UK is definitely not as advanced as some in the US and Canada at applying proper scientific methodology to Political Science, it doesn't mean those that are doing a bloody good job at it should be dismissed as mere guessers.

Date: Saturday, May 4th, 2013 04:02 pm (UTC)
matgb: Artwork of 19th century upper class anarchist, text: MatGB (Default)
From: [personal profile] matgb
Oh, the rest of your points are still valid of course.

Date: Saturday, May 4th, 2013 04:13 pm (UTC)
matgb: Artwork of 19th century upper class anarchist, text: MatGB (Default)
From: [personal profile] matgb
Yes and no-looking back at pre 2010 LEs, 2009 was anomalous as UKIP were on a ride then and weird stuff happened becuase the Euros were the same day, but 2008 was a lot closer to 2010 GE.

But because of the other factors, and of course campaigning and similar, Labour climbed backa bit, LDs dropped 2 points, Tories dropped a few more.. We all know locals and generals aren't the same-Calder Valley would be a LD/Tory marginal on the local results held the same day as the general, but in aggregate the key ward analysis can be a lot more predictive-it's one of the better opinion polls available.

It's a damn site more accurate than, say, YouGov, for example.

But punching those numbers into, say, Baxter, and giving a GE seat prediction is slightly futile, we know things liek tactical voting, individual candidates, etc make a massive difference.

Date: Sunday, May 5th, 2013 01:15 am (UTC)
lilacsigil: 12 Apostles rocks, text "Rock On" (12 Apostles)
From: [personal profile] lilacsigil
Wow, I didn't know that the data-gathering at constituency level was so poor. That wouldn't matter so much in most electorates in my country (we have preferential voting) but for the UK this is a major problem.

Date: Sunday, May 5th, 2013 07:09 am (UTC)
sashajwolf: photo of Blake with text: "reality is a dangerous concept" (Default)
From: [personal profile] sashajwolf
Angus Reid collected data at constituency level during the 2010 GE campaign, which I suspect is one of the reasons that, despite being new entrants, they tied with Harris for most accurate prediction of the result. You're right that it is a serious problem in general, though.

(Fwiw, AR's most recent UK poll gives Lab 39, Con 27, UKIP 16, LD 8, SNP 3, Green 3, BNP 2; but that one didn't look at the constituency level.)

Date: Sunday, May 5th, 2013 09:36 am (UTC)
sashajwolf: photo of Blake with text: "reality is a dangerous concept" (Default)
From: [personal profile] sashajwolf
Indeed. (I don't have a party any more, though; I didn't renew my Green membership because I didn't like the way they operate locally, and also because it wasn't feeling like the most productive use of limited time. I've been doing some small-scale environmental activism, and not much else. I will probably still vote Green in most elections, though.)

Date: Sunday, May 5th, 2013 09:40 am (UTC)
sashajwolf: photo of Blake with text: "reality is a dangerous concept" (Default)
From: [personal profile] sashajwolf
Thanks, but I doubt it. I did a decent amount of leafletting for them in the Mayoral elections, but nothing after that. I tried, but it was triggering my anxiety disorder, and I didn't feel I was going to achieve enough to justify that,

Date: Sunday, May 5th, 2013 09:46 am (UTC)
sashajwolf: photo of Blake with text: "reality is a dangerous concept" (Default)
From: [personal profile] sashajwolf
Thank you. I do find it frustrating, because clearly Something Needs To Be Done about the state of politics in this country for the sake of all the vulnerable minorities (some of which I belong to) that apparently can't get decent consideration under the current system, but I can't figure out what that Something is or how I can usefully contribute to it. So I retreat to doing what I can to fix other problems instead.

Date: Sunday, May 5th, 2013 09:52 am (UTC)
sashajwolf: photo of Blake with text: "reality is a dangerous concept" (Default)
From: [personal profile] sashajwolf
*nods*

Date: Sunday, May 5th, 2013 06:52 pm (UTC)
matgb: Artwork of 19th century upper class anarchist, text: MatGB (Default)
From: [personal profile] matgb
I'm increasing inclined to put what time I have to this sort of thing towards Unlock Democracy stuff, they're the least awful of the reform campaign groups and are showing impetus in setting up new groups and similar, that two friends, independent of each other, both work there is probably a partial bias but not a major one.

Partly-formed notes

Date: Sunday, May 5th, 2013 10:00 am (UTC)
daweaver:   (compute)
From: [personal profile] daweaver
Some thoughts from a partly-complete post of my own.

The only vote with a serious chance to alter the result of the next Westminster election is Scottish independence, on 18 September 2014.

UIP's strength was concentrated in those areas where, for decades, it's been Conservatives or a protest party. They didn't succeed in Somerset and Devon where it was Conservatives or Lib Dems; they didn't succeed in electoral districts that were young and socially tolerant.

Local elections help to set the mood music, they endorse a narrative of winners and losers. Journalists - print, broadcast, online - have a regrettable tendency to put out something now rather than something considered, and publishing now relies on stereotypes. Local elections are the best opportunity to update those stereotypes. I could argue that voters do use them to send a message - not to the politicians, but to the journalists and broadcasters.

About This Blog

Hello! I'm Jennie (known to many as SB, due to my handle, or The Yorksher Gob because of my old blog's name). This blog is my public face; click here for a list of all the other places you can find me on t'interwebs.



==================
Charities I support:

The Survivors' Trust - donate here
DogsTrust - donate here
CAB - donate here

==================


Creative Commons License
Miss SB by Jennie Rigg is licensed under a Creative Commons Attribution-Non-Commercial-No Derivative Works 2.0 UK: England & Wales License.
Based on a work at miss-s-b.dreamwidth.org.

Please note that any and all opinions expressed in this blog are subject to random change at whim my own, and not necessarily representative of my party, or any of the constituent parts thereof (except myself, obviously).

Printed by Dreamwidth Studios, Maryland USA. Promoted by Jennie Rigg, of Brighouse, West Yorkshire.

Most Popular Tags