Damned Lying Statistics, or What passes for Guardian journalism these days?
If anything, things are getting worse.
The culprit this time is Julian Glover, who (if today’s Guardian leader is anything to go by) seems to have taken Mark Twain’s claim that
There are three kinds of lies: lies, damned lies, and statistics.
and adopted it as a professional mantra.
The first sign of trouble is the headlines. Given an uncompromising headline on the front page of the print edition (“Tories in poll lead but losing battle over class”) and an even more strident line on the Guardian’s homepage (“Tories losing battle over class”), the article itself bears the more nuanced “Guardian/ICM poll: Conservatives show vulnerability in class battle“.
At the heart of Glover’s claim is a result from the latest Guardian/ICM poll. Glover writes:
The Conservatives are losing the battle over class, according to a Guardian/ICM poll published today, which shows a third of voters see the Tories as the party of the upper classes.
But there are signs that Labour’s attack on Tory toffs is sticking with a substantial minority identifying the party with the upper classes, even though almost no one in Britain admits belonging to this group.
ICM asked 1000 people whether they felt Labour and Conservatives most stands up for people from the upper, middle, or working classes, or for all people equally. It is perfectly correct that the 33% of respondents said that the Conservatives mostly stand up for the upper classes, but 11% of respondents gave the same answer about the Labour party. It is strange that Glover could not spare the words to contextualise his eye catching claim with this perhaps more surprising result.
The most strikingly dishonest part of Glover’s piece, however, is the claim that the Tories are “…losing the battle over class”. ICM did not ask these questions in preceding monthly polls (December, November, October), so neither we, nor Glover, has any way of knowing what effect Labour’s attempts to make class an issue are having. In fact, it may have had either no effect, or a negative effect – the situation may have improved for the Conservatives – and we could not establish this from the data.
One strong point Glover might have raised – but either failed to identify or else simply ignored – is the correlation between intending voting for a party other than the Conservatives, and believing the Tories are for the upper class. 55% of Labour, 47% of Liberal Democrat, and 39% of “other” voters held that view, compared to 33% overall and just 8% among Conservatives. This finding is particularly interesting because there is no such clear division between voters in Labour’s equivalent measure – 34% of Conservative, 37% of Labour, and 34% of Liberal Democrat voters believe Labour most stands up for the working class. If a Labour core vote strategy is to be pursued, this is the kind of result that would reinforce the belief that focusing class can be effective.
There are also detailed findings which should cause Labour some concern. Overall, 30% feel Labour stands up most for the working class, but that figure drops to 23% in Social Class D & E, and 16% of DE respondents felt Labour stands up most strongly for the upper class (compared to 11% overall).
Interestingly, the ICM poll shows that substantially more people see the Tories as standing up for either the middle classes, or everyone equally, than hold that opinion about Labour (58% versus 47%). If previous research is correct in showing an increasing tendency to self-report as middle class, it is this division between the parties – not the question of the upper classes – which could be most significant.
More tellingly, of course, Labour’s share of the vote declined by 2 points in this poll, to just 29%. If this is what happens when – as Julian Glover believes – the Tories are losing the battle on class – one can only wonder what apocalyptic fate awaits the Labour party if the Tories begin to win the argument.