More on IDS’s fraudulent misuse of data on benefit fraud

Posted on

MSN have just published an opinion from me on Iain Duncan Smith's claims on benefit fraud. Since the version on line is less technically robust than the first draft let me offer that earlier version here, instead:

It’s been widely reported that Iain Duncan Smith, the Work and Pensions Secretary in the Coalition government, has claimed that £10 billion has been lost in fraud and error under the tax credit system put in place by Labour. The result is that je has claimed that the tax credit system is "not fit for purpose".

It’s an interesting claim for a number of reasons. First of all, to come up with his headline number Iain Duncan Smith had to spread his calculations over a period of seven years from 2003 to 2010.

Second, he claims that the loss should be compared with total tax credit payments of £171 billion during that period. You would think as a result that the fraud and error rate was as a result about 5.8%.

That, however, is completely misleading. From 2003 to 2010 total benefit payments in the UK came to £1,089 billion according to HM Treasury Budget documentation for the years in question.

And, according to latest data on the error rate in the benefits system as a whole the cost of fraud is no more than 0.7% a year.   In 2011/12 that meant £1.2 billion was paid because of fraudulent claims. And given that this fraud rate has been remarkably consistent over time, it also implies that in total fraud might have cost over all benefits, including the old age pension, just £7.6 billion during the Labour years Duncan Smith referred to when making his claim.

Of course, error is an issue too, but it’s also not as big as the claim made would suggest, especially when many errors result in underpayments. In fact net errors (that’s over payments less underpayments due to error) amount to just 0.5% a year. Over the Labour years that might, therefore, imply an overspend of £5.4 billion. Put it together and over spending of £1,089 million fraud and error cost a total of just £13 billion in seven years under Labour, which is extraordinarily low. Most complex human systems would expect to see a much higher error rate than that. What is more, it means Duncan Smith’s claim that £10 billion was lost to just £171 billion of tax credit payments just does not stack.

Making that claim does though draw attention away from a much bigger problem. H M Revenue & Customs admit that tax fraud and error costs at least £23 billion a year.  The FT did however, suggest in 2012 that this was “complete guesswork”  suggesting the number may be much higher. I think it is. My estimate of tax fraud in the UK is £70 billion a year.

Now we can disagree about whether the right number is £23 billion of £70 billion or somewhere in between, but however it is looked at the cost of tax fraud is many, many times bigger than the £1.2 billion that benefit fraud will cost in total this year. And that then leads to the real question Iain Duncan Smith and the government need to answer, which is why are they tackling the wrong issue when it comes to fraud, because they very obviously are?