The Damn Lies of Statistical Silence



Date of last increase Previous Amount (£) Current Amount (£) Number Receiving Benefit Date of Information
Unemployment supplement paid with:
Pension paid for pneumoconiosis etc.: Nov. 81 780.00 per annum 858.00 per annum 12 Nov. 1980

TNA: BN 69, Social Security Advisory Committee Memorandum 11/82 Appendix 1, 1982.1

Twelve people. Twelve people were unemployed and receiving an addition to their benefit based on the fact that they had contracted a lung disease as a result of their job. Most of the other statistics in this file were rounded to the nearest thousand.

The amount of statistics the DHSS held (and I assume still holds) is quite mind-boggling. Even more interesting is the stuff they they don’t keep track of. Peter Townsend wrote in the 1970s that it was remarkable that the authorities didn’t record certain figures on matters of national importance.

Bureaucracies have vested interests in producing information which, so far as possible, is non-contentious and expressed in forms which are familiar and acceptable to prevailing values. […] [They] also have vested interests in defining problems for which they are largely responsible in forms which, when applied in research, show that these problems are of “manageable” proportions.

When the functions of administration and intelligence are served by the same department, [evaluating the external work of that organisation] will be distorted by the [the internal basis for decision making]. If this were not so, it would be difficult to explain the dearth of information in the annual reports of the [Supplementary Benefits Commission] about the wage-stop variation between areas in the award of discretionary allowances, and the number of unemployed whose benefit is terminated – all these subjects being of considerable public interest in recent years.2

Fraud statistics in today’s benefit system

Today, I can quite easily tell you what the fraud rate is in Incapacity Benefit. It’s 0.3%. £20 million. For years the government claimed that fraud was a major problem in the social security system. Yet their own figures show how TINY a problem fraud really is. 2% of all benefit payments – £3.2 billion – are overpayments due to “fraud and error”. 65% of that is due to error not fraud. 0.5% of all benefits – or a quarter of overpayments – are down to the government’s own errors.

Additionally, £1.3 billion of underpayments are caused by “error”.

The maths is simple. £1.1 billion of overpayments are caused by fraud. And there are £1.3 billion of underpayments.3 So, fraud doesn’t even cover the amount of money that should have been paid in the first place. There’s a shortfall of 200 million quid.

Pat Healy, the Social Security correspondent in The Times wrote:

More public money is lost by civil servancts awarding too much social security than is lost by people deliberately defrauding the system. Far more money is saved in unclaimed benefits than is lost by fraud. Yet public attention is once again being focused on the minority of cheats.4

This was in 1980. The story is eerily familiar today.


There are two criticisms, then, to mop up:

  • “You can’t trust the accuracy of government statistics.”
  • “If fraud were so easy to detect, it would be. The problem is the people slipping under the radar.”

Both can be answered pretty simply.

Historically, the British social security system is one of the tightest in the world. Its anti-fraud measures are considered to be world-class, and many countries take their lead from us. So paranoid were the Thatcher government about “scroungers” that they poured millions into fraud detection and prevention. They spent more than they ever recovered as a matter of principle.5

So, while government stats need to be taken with a pinch of salt – don’t all statistics? – these are probably the most accurate ones you’re going to get.

The point is, we were told that people cheating the system were costing us billions of pounds. So, if out-and-out fraud isn’t the issue, then where is the proof that people are taking money they don’t really need, or money which they are entitled to but that they misspend?

Would it be so difficult to conduct such a survey? If our social security system is at breaking point because of all these work-shy, feckless, breeding, malingering scroungers, then surely it’s in everybody’s best interests to spend a few million to save a few billion?

A response came there none.


It’s quite simple. If the government genuinely believes scroungers are to blame for the size of the welfare bill, it has the resources and technologies prove it. Or, more likely, scroungers are a convenient political scapegoat to deflect attention from the billions missing from the budget due to tax evasion.

You can prove anything with statistics. And as the DHSS proved in the 1980s, you can keep statistics to a startling degree of accuracy on the smallest of issues if there is the political will. That we have no stats on the numbers of people “abusing” the system should tell us all an awful lot about the priorities of government. And the “truth” about benefit claimants.


1 TNA: BN 69, Social Security Advisory Committee Memorandum 11/82 Appendix 1, 1982.
2 Peter Townsend, ‘Reports and Surveys – Politics and the Statistics of Poverty’, Political Quarterly 43(1), 1972, pp. 103-12, esp. p. 104.
3 All of these taken from the government’s own figures. See Department of Work and Pensions, Fraud and Error in the Benefits System: Preliminary 2011/12 Estimates (Great Britain), Revised Edition, (London : DWP, 6 June 2012).
4 Pat Healy, ‘Just where are all the scroungers?’ The Times, 27 February 1980, p. 14.
5 Michael Sullivan, The Development of the British Welfare State Michael Sullivan (Hemel Hempstead : Prentice Hall, 1996), p. 233.

Print Friendly