Latest post on Left Futures

Why Lord Ashcroft’s polling on Unite is worthless and should be discounted

Michael AshcroftLord Ashcroft published a poll of Unite members yesterday which made uncomfortable reading for Unite’s leadership – “just under half (49%) of Unite members said they would vote Labour in an election tomorrow; 23% would vote Tory,” revealed Lord Ashcroft for example.

It also gave some encouragement to Ed Miliband’s plan to restructure the party-union link — 57% of the 712 people polled by Lord Ashcroft “said the rules should be changed so members would be asked if they wanted some of their membership fee to go towards the fund, rather than having to opt out“. Mind you, this compares with 87.4% of the 241,249 Unite members who voted in the real poll required by law who voted to retain the opt-out fund as it is now. Answers do depend on how you ask the question!

But Ed will also be less keen on the fact that Ashcroft’s poll showed him only 6 points ahead of David Cameron as “the best Prime Minister” or that the balance of opinion amongst Unite members was that Labour was not “doing a good job of representing the interests of ordinary working people in Britain.”

However, a close examination of the detailed results reveals that the poll is absolutely unrepresentative of the membership of Unite and therefore all the results are essentially worthless:

Forty-two percent of Lord Ashcroft’s sample are classified in social groups AB (higher or intermediate managerial, administrative or professional grades) whilst only ten percent as DE (Semi and unskilled manual workers, and casual or lowest grade workers, pensioners, and others who depend on the welfare state for their income). This is grossly unrepresentative of Unite’s membership and indicate a colossal failure to correctly weight the data (indeed there is no indication in the tables of any weighting at all).

Semi and unskilled workers accounted for a considerable proportion of the former TGWU membership and a smaller but still very significant proportion of Unite.  Pensioners also numbered over a quarter of a million in 2012 (about 18% of the total membership) when they were given both increased rights and benefits in the union (all can vote in the election of the general secretary) and there are also increasing numbers of community members.

Managerial grades and professional staff are of course well represented in the union and accounted for most of the membership of some of the predecessor unions that now form part of Unite, notably ASTMS and TASS, but nothing like 42% of the total membership. Their gross over-representation also accounts for the fact that 38% of respondents were educated to degree level or equivalent professional/managerial qualification, and a further 18% to higher degree, doctorate, or MBA or equivalent professional/managerial qualification.

Given the extreme bias of the sample, it is not worth criticising at length other aspects of the survey such as leading questions like:

Some people think that the rules should be changed so that members have to opt IN to the political fund if they want some of their membership fee to go towards political campaigns and parties, rather than requiring members to opt OUT of the fund if they don’t want this to happen. Which system would you prefer?”

We certainly should not set any story by the responses.

Mark Ferguson at LabourList says the survey “has a terrible whiff of political hit job about it. Especially when Unite are the current Tory attack line du jour.” I agree. Though Lord Ashcroft’s surveys are often interesting and are of a size which enables detailed analysis that simply isn’t possible with most other polling, he doesn’t spend his money without a political justification. The bias, however, is probably unintended as Ferguson reports that over 15,000 interviews were conducted online to reach the 712 ultimate respondents. Shame his pollsters weren’t more careful about the weighting.

 

Comments are closed.

© 2024 Left Futures | Powered by WordPress | theme originated from PrimePress by Ravi Varma