I’ve been readying some work about benefit takeup, and in the course of doing it I’ve been looking again at figures on error – the mistakes made by claimants and officials. There’s an overlap with fraud, but there is some room to disentangle them. This is what the figures look like.
Estimates of errors in the benefit system (percentages of value)
|Fraud||Claimant error||Official error||Total|
|Council Tax Benefit|
|Underpaid||no data||no data||no data||0.7|
Viewed overall, Pension Credit has the worst record – more than one assessment in 12 is wrong – followed by Tax Credits and Housing Benefit. If we look only at the figures for underpayments, it seems that DLA has the worst record, followed by Pensions Credit. At the other end, the benefits with the lowest proportion of problems are Retirement Pension and JSA.
The most common reasons for underpayments are
- overestimating earnings;
- understating household composition;
- overstating the worth of Tax Credits;
- failing to claim premiums for special needs; and
- undeclared housing costs.
The benefits system is heavily dependent on people being able to report what is going on, how much and when. The problems this causes say a lot about the complexity of the benefit system, and the kind of thing that we ought to avoid when designing systems.