The discrepancy between unemployment figures and the census could, of
course, be produced by fraud, as it could be produced by many other factors.
However, we know from experience across the country that the census figures
are unreliable. By contrast, fraud is much better understood and constantly
monitored, and the existence of large-scale undetected fraud in an urban
area is much less likely.
Fraud has been the subject of repeated surveys in relation to the DSS/DWP
caseload over a long period of time. Most overpayment in relation to
unemployment takes place because a person's circumstances change without
informing the benefit office, which happens commonly in casualised, low paid
labour markets. The image of the benefits system in the popular press is
that the system is that all claimants have to do is walk into the office,
and they will be greeted by smiling, brainwashed officers throwing money
over them like confetti. The references in this correspondence to benefit
drops and phantom claimants depend on the similar assumption that the
benefit system is particularly vulnerable to deliberate, fraudulent
misrepresentation from large numbers of people. It isn't. I can't claim
the system is tight as a drum, but it doesn't leak like a sieve either.
Currently there are samples taken within each benefit office on a quarterly
basis, involving both examination of case records and personal interviews
with claimants. Irregularities in the pattern of claims in particular areas
are picked up from computer records, and subject to special operations. The
penalties for non-compliance (suspension of benefit and prejudice to future
claims) are severe and, because they do not require external processes for
action, they are virtually immediate. The complaints I've made about the
figures do not relate to the quality of the survey evidence, but to the lack
of relationship between that evidence and the published global figures -
which have since been cut by nearly three-quarters.
There are other reasons to treat the comparison of benefit figures and the
census with caution.
1. We haven't been told the provenance of the figures from which the
comparison is made, but the figures for unemployed people probably come from
sample surveys based on population. Sample surveys generally rely on
estimates of the population to provide the denominators. If you think the
population is larger, which just about everybody did (and does), the sample
will use a greater multiplier to produce an estimate of the total number of
claims or people affected. This is another reason why the failure of the
census is so deeply problematic; it infects much else we do.
2. We should also be cautious about the figures for people in receipt of
benefit. Benefit figures globally are calculated from sample surveys, and
not (as most of us would imagine) from adding up the administrative records.
Government samples, and surveys like the FES which are used to
cross-validate them, have always tended to overrepresent the proportion of
claimants, who are (self-evidently) more willing to disclose details of
income than others who don't claim.
3. The figures may not be strictly comparable. Claims for "unemployed
benefits" (does this mean JSA?) are not the same as people being unemployed.
The benefit figures cannot be tied down to a particular day, and variations
in unemployment in small local areas can be considerable from day to day and
week to week.
Paul Spicker
Professor of Public Policy
Centre for Public Policy and Management
The Robert Gordon University
Garthdee Road
Aberdeen AB10 7QE
Scotland
Tel: + 44 (0) 1224 263120
Fax: + 44 (0) 1224 263434
Website: http://www.rgu.ac.uk/publicpolicy/
******************************************************
Please note that if you press the 'Reply' button your
message will go only to the sender of this message.
If you want to reply to the whole list, use your mailer's
'Reply-to-All' button to send your message automatically
to [log in to unmask]
*******************************************************
|