Chris Game

In the higher education world June means two things: the assessment season and the publication of university league tables. We had the Guardian University Guide rankings this week and here at the University of Birmingham we were very excited, having apparently shot up in a single year from 30th out of 119 institutions to 15th rankings. We seem to have cracked the system, so watch out, Oxbridge, here we come!

Rigorous assessment, as the Education Secretary regularly reminds us, is a vital part of any educational package, and certainly in HE it’s guaranteed to appeal to fee-conscious parents, if not necessarily to their prospective student offspring. Here in my little corner of the university, therefore, we like to feel we’ve done our bit towards our university’s startlingly improved ranking with our work on rooting out and publicising plagiarism.

Go too deeply into it and plagiarism – appropriating someone else’s work, language, thoughts, ideas, or whatever as your own – can quickly become over-complicated. What counts as a thought? Can you plagiarise yourself? Is wrongful appropriation a crime or an ethical lapse? So we try to keep it simple: it’s cheating, it’s bad, and the more of it you detect, the better.

In what we might call the dark ages, plagiarism rates were the numbers or percentages of all students’ marks that, on investigation, were found to be higher than they should have been, due to the students having cheated. Regrettably, because either our investigations were insufficiently diligent or students weren’t actually cheating that much, the rates were mostly pretty low and risked making us look careless about these matters.

How, then, could these disappointingly low rates be boosted without actually forcing students to cheat? Easy! What we’re really interested in is detecting marks wrongly assigned. But ‘wrongly assigned’ makes it seem as if we, not students, are responsible, so let’s keep the useful term ‘plagiarism’, but extend it to include ALL marks that were found to have been wrongly assigned.

So, if a student falsely claims to have handed work in on time, thereby avoiding a mark penalty for late submission, that’s plagiarism. If our external examiner finds we’ve over-marked an assignment, that’s plagiarism. If the Examinations Office incorrectly records or transcribes a mark, that’s plagiarism. The result: we’ve got stats that plagiarism detection software manufacturers would kill for.

At which point, I should emphasise, at John McEnroe-like volume, I AM NOT BEING SERIOUS!! Or rather, I am being serious, but not about plagiarism. Almost everything in the last few paragraphs – apart from the UoB’s Guardian ranking – derives from my warped imagination.

However, rewrite those paragraphs, replacing plagiarism or intellectual fraud with ‘benefit fraud’, and you’ve got a fair description of how in the real world Ministers and the media manipulate statistics to get them to tell a more convenient story. In brief, if you bundle various unrelated activities together and label them collectively as ‘fraud’, then feed the numbers to the media, you can have voters baying for all the benefit cuts you were going to introduce anyway.

Let’s take the most recent benefit fraud outrage panic, prompted by the release of some latest figures by the Department for Work and Pensions (DWP) hard upon the manslaughter trial revelations of benefit scrounger Mick Philpott. The Daily Telegraph went with “Fraudulent and wrong benefit claims hit £3.5 billion record”. The Daily Express launched a “Call for new blitz on benefits to cut £3.6 billion fraud bill”.

Don’t bother about the slight statistical discrepancy; we’ll come back to it shortly. Besides, what does the odd hundred million matter when most of your readers presumably subscribe broadly to the average view of respondents in a recent YouGov/TUC survey: that 41% of the entire welfare budget goes on benefits to unemployed people, and that 27% of the welfare budget is claimed fraudulently.

In fact, DWP statistics show that of the total 2012/13 welfare budget of £167 billion, nearly half (£80 billion) went on state pensions, 14% on housing benefit, 8% on disability living allowance, 5% on pension credits, and 3% each (just over £5 billion) on jobseeker’s allowance, income support, and council tax benefit. The TUC’s ‘correct’ answer to its ‘benefits to unemployed people’ question is the 3% for jobseeker’s allowance, and even if this does, perhaps misleadingly, ignore other benefits for which the unemployed may qualify, it’s still massively closer to the true picture than 41%.

As for the popular guesstimate of 27% of the welfare budget going in fraudulent claims, the DWP’s more evidence-based estimate is contained in the annual publication referred to above, the clue to which is in its title: Fraud and Error in the Benefit System. Yes, it covers fraud, but its purpose is to measure the total cost to the taxpayer of all incorrect benefit payments made, whatever the cause – and, most importantly, to distinguish between those causes.

The preliminary estimate for 2012/13 is that total over-payment due to fraud and error across all benefits was £3.5 billion, or 2.1% of the total welfare budget. That’s where the Daily Telegraph’s figure came from, except that the paper didn’t bother clarifying that only a third of it was actually the result of deliberate claimant fraud. The other two-thirds were caused by claimant error with no fraudulent intent, and error by DWP or local authority officials.

Moreover, non-fraudulent claimant and official errors were also responsible for under-payments of £1.4 billion, so the net cost to the taxpayer of all benefit fraud and payment error was £2.1 billion or a little under 1.3% of the total budget, with actual fraudulent claims being responsible not for £45 billion, which would be 27% of the budget, nor for £3.5 billion, but £1.2 billion or 0.7%.

Interestingly, the Daily Express recognised that £1.2 billion is the correct fraud figure, but, possibly not to be outdone by the Telegraph, they got their headline by trebling it to £3.6 billion to cover a three-year period.

It’s always tempting, when one knows roughly the respective figures, to compare benefit fraud and tax fraud – but not that easy. First, there’s the terminology. HMRC don’t investigate anything as crude as tax fraud. Rather, they measure tax gaps, between what should be and actually is collected, and talk only of people circumventing or evading paying their taxes. In 2010/11 they measured a tax gap of £32 billion, representing 6.7% of total tax liabilities, and 15 times the net benefit overpayment gap. As usual, by far the biggest gap was in VAT – £9.6 billion, or over 10% of the tax due.

When it comes to differentiating tax-circumventing behaviours, HMRC reckon that unambiguous tax evasion accounts for 46% of the gap, or £14.7 billion – although it seems unlikely that some of the other behaviours, like failure to take reasonable care, would be quite as sympathetically interpreted, were the perpetrators benefit claimants.

Time, I think, to make my own position clear. By trying to identify the most accurate recent measures in what is inevitably an immensely difficult and controversial policy field, I do not diminish in any way the financial, political or moral importance of either benefit or tax fraud. I welcome in particular The Local Government Fraud Strategy, both for the work that it is spearheading and for the increased knowledge and awareness to which it will lead.

My concern is with the deliberate distortion of public knowledge and awareness. Politicians, even Ministers, cannot be held entirely responsible for the public’s misconceptions of how the world works, but they can be criticised when they pander to those misconceptions by falsely presenting or encouraging the false presentation of their own figures. DWP Secretary Iain Duncan Smith – rebuked again only last month by the UK Statistics Authority for distorting his department’s figures on the impact of the benefits cap – and junior minister, Lord Freud, have both had publicly to apologise for exaggerating the extent of benefit fraud, so now they just rely on friendly media to do their dirty work for them. They, above all, should know that the subject’s importance deserves much better.

Chris is a Visiting Lecturer at INLOGOV interested in the politics of local government; local elections, electoral reform and other electoral behaviour; party politics; political leadership and management; member-officer relations; central-local relations; use of consumer and opinion research in local government; the modernisation agenda and the implementation of executive local government.