22

The following graphic is doing the rounds on Facebook:

3250: number of staff at DWP investigating benefit fraud of £1.2bn, 300: number of staff at HMRC investigating tax evasion of £70bn+

Is it true?

Does the graphic give an accurate representation of the comparative headcount/effort/resources that the UK government (as a whole) puts into investigating these two things?

To what extent are the two things - benefit fraud and tax evasion - directly comparable?

Background info:

  • 'benefits' (BrEng) = 'welfare' (AmEng)

  • DWP = Department of Work and Pensions (the central government department that pays benefits)

  • HMRC = Her Majesty's Revenue and Customs (the central government department that collects taxes)

The source is given in the graphic as:

https://twitter.com/nw_nicholas/status/420485474139246592

which is a twitter post by Nicholas Wilson saying:

Mr Ethical
‏@nw_nicholas

.@UKuncut DWP has 3000+ staff for benefit fraud of £1.2bn; HMRC has 300 staff for tax evasion of £70bn+

Sections of text which are stated to be from DWP and HMRC

The image tweeted by ‏nw_nicholas gives the source for the first section of text as ‏"From the DWP" with a blt.ly shortened URL, which resolves to:

http://compass.port.ac.uk/UoP/file/1f9193d1-7dc7-455a-a7a0-23441e2ac44e/1/Fraud_Criminal_Justice_IMSLRN.zip/page_02.htm

which is a page on "Fraud and the Criminal Justice System", (c) University of Portsmouth, 2012.

I don't know who is the author of the text on that page (which ‏@nw_nicholas cites in the image on twitter), although a Google search for "the 7500 overestimates the actual number of fraud" turns up no results other than that page itself.

Portsmouth University has an Institute of Criminal Justice Studies, perhaps that's related to the document somehow.

The second document cited by Nicholas Wilson is this PDF published by HMRC:
"Closing in HMRC’s approach to tax evasion" and the quote he gives is accurate (p11), although of course "more than 300" isn't exactly the same thing as 300 (but surely if the number were much higher than 300 then the HMRC document would use that higher number).

Community
  • 1
A E
  • 7,319
  • 6
  • 30
  • 50
  • 6
    What does "fair picture" mean? Maybe instead of 1.2bn and 70bn, we should use "number of cases of tax evasion" and "number of cases of benefit fraud". – GEdgar Feb 16 '15 at 15:00
  • 1
    @GEdgar, what I was trying to get at with "does it give a fair picture ..." was, maybe there's some other parts of the government investigating one or other of those things. Or other enforcement actions going on that aren't captured in the graphic. I'll amend the qn a little bit... – A E Feb 16 '15 at 15:04
  • 11
    For what it's worth, HMRC's tax evasion work is far more automated than DWP's benefit fraud work. Tax evasion can be found by computers crunching bank transfers/payments etc, while benefit fraud usually requires manual investigation. The graphic appears to be correct, but irrelevant. – Jon Story Feb 17 '15 at 12:45
  • @JonStory, that's interesting. Do you know of any sources for that? I'm wondering to what extent that's the way they choose to investigate it versus being intrinsic in the nature of the work itself. I can imagine how one might use manpower-heavy methods to investigate tax evasion - e.g. surveillance, informants - and how one might use automated methods to detect benefits fraud. So to the extent that it's a *choice* of investigatory tactics, that's interesting in itself. – A E Feb 17 '15 at 14:19
  • 1
    Unfortunately not, my information is from "reliable friends who work for the agencies", hence posting it as a comment rather than a proper verifiable answer. I'm sure there are more manpower heavy ways, but as the above graphic shows, this way is (personnel wise) cheaper – Jon Story Feb 17 '15 at 14:21
  • @JonStory, right, I thought that might be the case. As to whether it's cheaper, I guess that depends on the relative effectiveness of the two methods in recovering money which otherwise would have been lost. – A E Feb 17 '15 at 14:27
  • The question **assumes** that we have reliable estimates of the extent of tax evasion and benefit fraud. That is not clear at all. Perhaps it deserves a separate question. – matt_black Feb 17 '15 at 20:48
  • @matt_black, true. There probably is an official estimate - or a choice of them from different sources - but I guess anything unrecorded is by definition hard to estimate accurately. – A E Feb 18 '15 at 20:38
  • 2
    Considering the fact that there is likely do be orders of magnitude **less instances** of meaningful tax evasion (e.g. those that materially contribute to 70B instead of being rounding error), this might be even less relevant that Jon Story's comment suggests, even if true. – user5341 Mar 18 '15 at 16:55
  • @DVK, I'm not clear what you mean. Why would there necessarily be orders of magnitude less instances of meaningful tax evasion? What's '70B'? – A E Mar 18 '15 at 17:57
  • 2
    @AE - making up example #s, if we have 699 companies evading 100Million pounds each (given the scale of Apple and such, not unrealistic), you'd only need to investigate 699 entities for tax evasion to capture nearly 100% of your stated 70 Billion pounds tax evasion total - more precisely, all but 100 million pounds). Whereas benefit fraud is smaller piecemail scale, so you need to investigate tens/hundreds of thousands of people, to recover 1.2 billion pounds. So you can get away with SIGNIFICANTLY less personel for tax evasion, and recover more money. – user5341 Mar 18 '15 at 18:17
  • @DVK, oh, I understand what you're saying. If I understand the source document [HMRC: Closing in on tax evasion](http://webarchive.nationalarchives.gov.uk/20121204171507/http://www.hmrc.gov.uk/budget-updates/march2012/tax-evasion-report.pdf) correctly, the 300 staff at HMRC are investigating the evasion of personal tax - tax evasion by natural people, living individuals - rather than corporations. That only accounts for £32bn of the tax gap though, I have no clear idea where the graphic is getting the '£70bn+' figure from. The apparent omission of staff investigating corporate .... – A E Mar 18 '15 at 19:04
  • ... tax fraud (there must be some staff investigating that somewhere in government, surely) seems like one way in which the graphic is maybe inaccurate. – A E Mar 18 '15 at 19:05

1 Answers1

8

The figure of £1.2bn for benefit fraud appears to be accurate: Fraud and Error in the Benefit System 2013/14

The figure of £70bn seems rather high; HMRC estimates a figure of £34 billion in total for 2012/3: Measuring tax gaps 2014 edition. Moreover, that is the total for all taxes, including excise duties, corporation tax and the like. There are unofficial figures which are higher but the claim that the official figures are inaccurate is a debateable point.

The figure of 3250 for DWP investigators seems reasonably sourced.

The figure of 300 is the most misleading; it is as the quote says for those investigating 'affluent tax evaders'. This is a group targeted on the rich but not super rich (roughly the top 500 000 tax payers, excluding the top ~5 000). The Affluent Compliance Team. This is < 2% of taxpayers.

In conclusion, the figure cannot be regarded as accurate because it compares a highball figure for the total tax evasion against the staff within a subgroup of the personal tax team.

richardb
  • 2,405
  • 1
  • 17
  • 17
  • I suggest you should basically include the entire Inland Revenue Service which about 29,000 people in tax enforcement https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/631033/HMRC_WMI_headcount_and_payroll_data_June_2017.csv/preview – user2617804 Aug 03 '17 at 06:01
  • Your answer could mention there were [20 million DWP benefit claimants at May 2017](https://www.ons.gov.uk/employmentandlabourmarket/peoplenotinwork/outofworkbenefits). That is 40 times the number of rich/affluent quoted above. That could be an argument for 40x as many fraud investigators if you ignore average complexity of cases or other factors. So 11 times as many might not be disproportionate. – RedGrittyBrick Feb 06 '18 at 15:29