Thursday, 16 June 2011

Case #4: The League of Shameless

Not only have they started conflating benefit fraud with fraud throughout the public sector after being caught conflating it with error, it hasn't made them apologise, it hasn't made them accept responsibility. It just gave them another excuse to cherry-pick something else, which in this case is to make an issue out of customer error costing more than fraud or official error.
Not both together you understand, that would be something tangibly real to be concerned about. No, it's higher than them both....separately.

So what's their solution to this non-problem? Instant fines. I don't recall any talk of DWP decision makers being punished over wrongly instated sanctions. So how bad of a non-problem is it? The figures:

£1.3 billion has been lost to customer error
£1.2 billion has been lost to fraud
£0.9 billion has been lost to official error
Not one to trust a press release initiated by either Miller, Grayling, Duncan-Smith or Freud any more. I went to actually look at the report and behold in Table A for Estimated Overpayments in 2010/11 we see what really matters: the percentages of total expenditure curiously missing from the Freud/DWP press release.

Fraud: £1.2 billion, 0.8% of expenditure.
Customer Error: £1.3 billion, 0.8% of expenditure.
Official Error: £0.9 billion, 0.6% of expenditure.
So the solution to the non-problem is being pursued on the basis of what amounts to a 0.2% difference. So no wonder they're focusing instead on the flat numbers rather than percentage; they can claim it has risen unacceptably.
At least if the reports for 2008 and 2009 where on pages 8 and 9 respectively we can see the tables for those years.

Our survey says:

Fraud: £1 billion, 0.8% of expenditure.
Customer Error: £1.1 billion, 0.8% of expenditure.
Official Error: £0.8 billion, 0.6% of expenditure.

Fraud: £1.1 billion, 0.8% of expenditure.
Customer Error: £1.2 billion, 0.8% of expenditure.
Official Error: £1 billion, 0.6% of expenditure.
The highlighted percentages haven't changed. What has changed? There are more benefit claimants. EDIT: or the slight difference in the totals is because of chance; the report says the figures are not 'statistically significant. That's why the numbers went up but the percentages didn't; customer error is no more of a problem now than it was before. Although the figures are roughly rounded it does seem that official error did actually rise the most, just not enough to change the percentage.

Just for fun: guess which are the only two benefits where the official error is higher than either customer error or fraud for the last three years?

Incapacity Benefit and Disability Living Allowance. In fact, official error for IB in 2010 was almost four times higher than the fraud prevalence for it: £20 million (0.3%) fraud compared with £70 million (1.2%) in official error. In fact for most of the time the official error is higher than the customer error in nearly all of the benefits in the table; it's only a few complicated ones that pushes the total for all above the total official error cost.


  1. Quote-"The penalty will encourage people to be more responsible in keeping their claims correct"

    Or in other words .We need to deter even more people not to even claim what they may be entitled to .If people actually get what they should we are screwed.This is from people who have no idea about the basics of the various benefits/allowances.Fifty pound fine for any untruth emanating from Ministers concerning "welfare"-better for the tax-payer,better for justice.

  2. I also note that"it is not statistically significant" appears when the "numbers" do not suit the agenda but when this is blatently true as you say,it suddenly comes a cause for concern for "action".Thanks for your analysis.

  3. In the field of statistics, the term 'significant' has a very specific meaning.

    All it means is that an event is unlikely to have happened by chance. When the document says the figures aren't statistically significant, it is saying that it can not rule out that chance explains the changes.

    I should note that this affects my claim that the rising number of claimants is the reason why the total figures for the various kinds of overpayments has risen. Whilst it doesn't affect my overall point, I'm putting a correction in.

  4. Just to also note that the reason why I cranked this blog post out so quickly was because the press release went up on the 16th. The timing is suspicious because the Welfare Reform bill has passed the Third Reading and now goes to the Lords, plus it was Thursday.

    Certain misleading tabloid non-stories that had come out on Mondays were preceded by DWP press releases on the Thursday or Friday before the weekend. Those newspapers are then given quotes to add in from 'sources within' etc which funny enough only get sent to those same newspapers.

    I think there's a good chance that sometime this week this will be made into a big deal in those same newspapers and it could be tomorrow. Readers of this blog will be already prepared for it.