Sunday, 9 September 2018

Inferences and statistics

Results of a new YouGov poll for the People's Vote lot  in the Observer today, one of a series possibly paid for by Superdry blokey,and clearly designed to influence public opinion.This one is also reckoned to have an influence on upcoming political conferences: 'The poll comes as union delegates gather in Manchester for the annual TUC conference, where Brexit will be debated on Monday, and two weeks before the Labour party conference in Liverpool, where delegates are expected to debate and vote on Brexit policy.'  The Observer argues that the poll shows 'heavy union backing for a second Brexit vote'

The actual poll is described further in an article by ace pollster P Kellner. Rather than details about the size of the poll,other than describing it as 'large' -- n ='more than 2,700',  or its representativeness, there is a general claim that the methods used have been successful in the past, in predicting the election of Corbyn as Labour Leader, specifically (so that's one out of ...?). 

Lest we forget,You Gov did screw up on the actual Brexit vote: 'Joe Twyman, head of political research for YouGov, one of the U.K.’s most prominent polling firms, appeared on [Sky's] set with the results of an online exit poll conducted for Sky. He explained that the firm had been tracking the same voters—and they had moved farther into the Remain camp that day.' One result, according to Bloomberg Businessweek was:


Hedge funds aiming to win big from trades that day had hired YouGov and at least five other polling companies, including Farage's favorite pollster. Their services, on the day and in the days leading up to the vote, varied, but pollsters sold hedge funds critical, advance information, including data that would have been illegal for them to give the public. Some hedge funds gained confidence, through private exit polls, that most Britons had voted to leave the EU, or that the vote was far closer than the public believed—knowledge pollsters provided while voting was still underway and hours ahead of official tallies. These hedge funds were in the perfect position to earn fortunes by short selling the British pound [because the mistaken news of a Remain vote pushed up the pound temporarily] .

OK, so I've used a classic ideological trope myself here, insulting YouGov by associatingit with this dodgy procedure before actually considering their new data, which might be utterly pure and disinterested.

The actual question asked was: 'When the negotiations with the European Union about Brexit are complete, would you support or oppose a public vote on the outcome?'

Kellner adds that:

the new surveys find that union members’ views are much the same, whether they are men or women, young or old, north or south, white-collar or blue-collar. If the data were adjusted to a different demographic profile, the overall percentages would move by no more than one or two points.

As with all single questions, though, there are problems of interpretation. Remainers know that as well as anyone, pointing out that a vote to leave in 2016 could have meant many things (but not apparently recognising that so would any new vote to remain, of course). Kellner seems to have no doubts:



We can therefore be confident that most members of Unite, Unison and GMB (which have a combined membership of more than 3 million people) really do want the UK to stay in the EU, and a chance to vote in a fresh referendum once the negotiations are complete.

Well... they seem to support 'a public vote', not specififcally a fresh referendum. And advocating a public vote might not be exactly the same as wanting to stay in the EU.

Nevertheless, the findings seem clear:

68% of Unison members [and sizeable majorities of members of other unions] said they would vote to remain in the EU in any new referendum [the actual question was 'If there were a referendum today on whether or not the UK should remain a member of the EU, how would you vote?' Nothing about the terms of course]


Their reasons are clear. It’s not that trade union members are indulging in gesture politics or ideological breast-beating. They are worried about the impact of Brexit on jobs, taxes, living standards and the NHS. They fear a Brexit Britain would find it harder to sell products and services abroad...Their attitudes to immigration are especially significant. In the 2016 referendum, one of the arguments for Brexit was that immigrant workers were undercutting the pay of low-paid British workers. Brexit, so the argument ran, would allow Britain to stop this. As a result, there would be more, and better-paid, jobs for British workers.
Many Unite, Unison and GMB members earn below-average wages. They might be expected to support that part of the Brexit agenda. They don’t. Overwhelming majorities, ranging from 74% to 85%, want EU citizens either to have complete freedom of movement to come to the UK, or the freedom to settle here if they have a job or university place lined up [at the risk of being mealy-mouthed, this would still infringe EC rules which do not offer such reservations of their precious 'freedoms'].

I can only confess myself baffled. It might be that TUs are an unusual minority in some way,although Kellner's point about tweaking the demographics seems to rule that out.  Or that the poll distorts the data -- a fair bet but these are large swings? Or that people really have changed their minds? 

Details of the poll are not available so far -- watch this space...




No comments:

Post a Comment