Northernheckler's Blog

A Yorkshireman's adventures in the big Smoke

At least a third of statistics are made up

Just a quick one. I’ll add links later – I have real work to do.
This morning’s Metro carries a huge “every-one-on-the-train-can- read-it” headline proclaiming

‘Labour loses a third of voters”

Not based on a real election obviously (which does beg a major issue) but on last night’s Harris poll which shows Labour on 29%. (Note that the more interesting aspect of this poll is that it shows all three major parties losing support).

Thing is though, Labour’s share of the popular vote at the General Election was 35.3%

The poll asks people whether they’re likely to vote for the same party as they did last time, and uses the figure for who people in the poll said they voted for, rather than the 2005 General Election stats. This however gives a barely altered figure of 36%.

So if Labour had “lost” a third of its votes (and remember it’s a poll not an election) it’s standing would be 24% according to the Harris Poll, or 23.5% if we went with the actual voting figures.

It’s not though – the figure they give for the poll is 29% for Labour – or a loss of around 18% – between a 6th and a 5th of voters – in other words a little over half the fraction that the Metro is claiming they have lost.

Other recent polls have shown Labour support as high as 35% – barely a loss of votes at all; and averages of polls show Labour’s current showing at 32%

Which would show that Labour have “lost” around 9% of their vote.

I concede that this is hardly a ringing endorsement.

But they have NOT lost a third of their vote.

To suggest this is complete and utter fabrication. The Metro should be ashamed of this blatant falsehood


March 10, 2010 - Posted by | politics | , ,


  1. The accompanying projection of seats is even funnier. It says 39 seats would go to a party called ‘Wouldn’t vote’ and 110 for a party called ‘Undecided’.

    Technically, Labour *could* lose a third of its vote and still be on 29%, as long as it was recruiting lots of voters from elsewhere.

    Comment by Lewis Baston | March 10, 2010 | Reply

    • And of course a sampled poll can’t easily predict anything about actual numbers of votes – the vote could actually increase – and the share of vote drop by a third. Lies damn Lies & all that

      Comment by northernheckler | March 11, 2010 | Reply

  2. Social comments and analytics for this post…

    This post was mentioned on Twitter by northernheckler: At least a third of statistics are made up:

    Trackback by uberVU - social comments | March 10, 2010 | Reply

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: