Findings (continued)

Putting it right

Definition

2.397 Putting it right means detecting when payments go wrong and taking prompt action to correct them with appropriate penalties to deter a recurrence. It is vital that fraud or error is detected as soon as possible to keep overpayments to a minimum.

2.398 The greater emphasis on securing the gateway to benefit lets fraud investigators concentrate on people who intend to steal. This means:

ensuring that fraud investigators are well-trained
providing high quality referrals to investigators
passing on experience and information to other investigators quickly
publicising the efforts of investigators
providing investigators with sufficient resources
sharing information with other agencies as allowed by the law
recording overpayments and making every effort to recover them
applying varied and effective sanctions.

CAFT organisation

2.399 CAFT was introduced following the internally commissioned Appleby report in 1995 into fraud and corruption in Lambeth. It became fully operational in 1997 and adopted the following mission statement:

To act as an independent corporate unit to protect the staff, finances. clients and residents of the council by implementing a comprehensive, fair, balanced professional strategy, to

develop an anti-fraud culture
deter, prevent, detect and investigate fraud and corruption
seek appropriate sanctions against those who perpetrate or seek to perpetrate fraud and corruption and
obtain redress in respect of any resources involved.

2.400 CAFT has 2 teams who investigate HB, CTB and housing. Their work is based on postcodes in the north and south of the borough. The teams’ complement at the time of our on-site inspection were:

North team – 4 fraud investigators and 3 visiting officers
South team – 5 fraud investigators and 2 visiting officers.

Their own group fraud manager heads both teams. Investigators within these teams are expected to be competent in both housing fraud and HB and CTB fraud. The average length of service of currently employed investigators is 4.5 years.

2.401

Figure 2.60 shows the current CAFT organisation.

Fig. 2.60: CAFT organisation chart

Source: Lambeth

CAFT procedures

2.402 CAFT is currently seeking ISO9000 accreditation. A requirement of this is a comprehensive set of procedures. We found that the procedure documents covered all aspects of the investigative process from receipt of the referral to the decision to carry out a sanction.

2.403 We saw examples of the documents an investigator is likely to use in the course of their enquiries. These are flexible in that the standard version can be tailored to suit the investigator’s needs. We also saw copies of the CAFT mission statement and the prosecution policy.

2.404 CAFT is located in Lambeth Town Hall and has its own discrete offices protected by an intruder alarm system and CCTV. The Head of CAFT reports to the Borough Solicitor.

Case management

2.405 Lambeth uses a bespoke case management system for HB and CTB fraud called FACTS. All fraud referrals are logged onto the database. The system is based on Microsoft Access and is capable of maintaining caseload records and providing statistical returns. It also receives weekly downloads of new RA claims from the Academy system. Evidence from our sampling exercises leads us to question the reliability of the data within the system.

2.406 Record control and maintenance is carried out by the CAFT IT officer. He also maintains the integrity of the system. No deletions, queries or alterations can be made to a case record except by this officer. One weakness of the system is that weekly benefit savings (WBS) for "gone aways" has to be manually calculated

We recommend that Lambeth reviews the requirements of the fraud case management system.

Fraud awareness

2.407 The BFI Good Practice Guide recommends that a comprehensive fraud awareness training programme helps an LA to minimise the risks of fraud entering the system. It also indicates that fraud awareness training increases the likelihood of detecting fraud at the earliest opportunity.

2.408 CAFT has carried out fraud awareness seminars for other Lambeth staff, for Members and for Capita staff. This has been part of an LA-wide seminar programme for raising fraud awareness. Capita HB staff have received training once this year as have the determination officers. Overpayments staff and counter staff have not received this training.

2.409 The seminars consist of an introduction to:

the work of CAFT
the principles of HB and CTB
the types of HB fraud
the proactive work of CAFT in detecting fraud
liaison with Capita
the role of staff in detecting fraud.

2.410 The seminars are tailored for each audience, for example assessment staff, determination officers and estate management boards.

2.411 We were told by the Borough Solicitor that the fraud awareness sessions had been ineffective and are now limited to 2 sessions a year for all Lambeth employees, none specifically for HB. We were told that the high turnover of contractor staff has meant that the HB sessions had had a negligible impact.

We recommend that Lambeth reinstates frequent and regular fraud awareness sessions for all of Capita’s benefits staff.

2.412 This will have the benefit of:

minimising the risks of fraud entering the system
detecting fraud at the earliest opportunity
increasing the number and quality of fraud referrals.

Fraud publicity and hotline

2.413 A fraud hotline is a good way of encouraging fraud referrals from the public and from staff. Lambeth has operated a fraud hotline since August 1998.

2.414 The hotline number is advertised on 4 different posters, produced by CAFT. The design of the posters is eye catching and original.

2.415 We were told that the posters had been displayed in:

the Town Hall
OMH
the Citizens Advice Bureau (CAB)
Brixton tube station
NHOs.

At the time of the inspection we found that no posters were displayed in the public areas of OMH.

2.416 Lambeth received 96 referrals from the fraud hotline in 1999/2000 (8% of total referrals). Of these, 5 resulted in Lambeth claiming that a fraud had been established. However, Lambeth was unable to provide details of how many led to a WBS claim.

2.417 All calls are logged on a database. Each call is allocated a sequential reference number, which is given to the caller. This facilitates cross-referencing when callers call later with further information. We consider this good practice.

2.418 Information received from callers is recorded on a Lambeth report form. We consider that this form could be improved by a clearer layout, and by the inclusion of prompts to elicit further information that would prove useful to an investigation. Suggested prompts include:

the description of an alleged partner
full details of any alleged employment
whether collusion is suspected.

2.419 In our sample of 70 fraud investigations, there were 3 cases that had originated from calls to the hotline (4% of the total). However, none of these had led to a WBS being claimed.

2.420 We consider that the effectiveness of the hotline is limited because:

the number printed on the posters is out of date following changes in London telephone codes
it is not a Freefone number
there is no dedicated extension allowing only incoming calls
the hotline is not advertised on the claim forms
callers reporting suspected frauds to the Coventry call centre are told to phone the hotline number instead
the report form does not contain sufficient prompts to obtain all relevant information.

2.421 While not always concerned with benefit fraud, the work of CAFT is often covered in the local press when it is involved in a high-profile investigation or as the result of a court case. CAFT has also produced a periodic newsletter that is distributed to all Lambeth staff and its work is also covered in the Lambeth corporate magazine, which is issued to the public.

We recommend that Lambeth improves the effectiveness of fraud publicity and the hotline by:

ensuring that hotline posters display the correct telephone number
use of a Freefone number
displaying the hotline posters at OMH
revising the format of the anonymous telephone call stencil to collect all relevant information
providing a stock of report forms and fraud awareness training to the staff at the Coventry call centre
monitoring the outcome of hotline referrals.

Fraud investigation training

2.422 We consider it good practice that fraud investigators are well trained and that they receive regular refresher training. A comprehensive training programme will:
increase staff knowledge and awareness of procedures
reduce errors and inconsistencies
ensure that staff keep up to date with changes
develop and better motivate staff.

2.423 We were told that Lambeth had provided an adequate training budget for CAFT staff. The costs are shown in Figure 2.61

Fig. 2.61: CAFT Training costs
Year	Allocated budget	Actual spend
1997/98	£15,000	£14,062
1998/99	£20,000	£43,591
1999/2000	£20,000	£16,997

Source: Lambeth

2.424 Figure 2.62 shows the numbers of days training that investigators have received in the past 3 years. We found that external training has decreased in 1999/2000, as all investigators now have accredited investigator status and no new staff have been recruited.

Fig. 2.62: Training days received by CAFT investigators
Year	Training from BFIS	Other external fraud training
1997/98	35	70
1998/99	20	94
1999/2000	35	74

Source: Lambeth

2.425 We found that CAFT staff had received training from a number of sources in the past 3 years. This included training on the following from BFIS investigators:

use of administrative penalties
inspectors’ powers
use of the document scanner
surveillance
use of the RAT
distance learning on the SSAA 1992.

2.426 We analysed the fraud training provided in 1999/2000 by other external providers and the results of our analysis are shown in Figure 2.63.

Fig. 2.63: Training days received by CAFT investigators from other external fraud providers
Course	Number of investigators	Number of days	Total number of days
Investigative interviewing	2	5	10
Criminal Procedure and Investigations Act	2	5	10
Surveillance	3	5	15
IRRV	1	2	2
WBS	1	2	2
BSI	1	3	3
Financial awareness	4	1	4
UV scanner	5	1	5
RATS	8	1	8
Prosecution workshop	15	1	15
Total	42	26	74

Source: Lambeth

2.427 CAFT has invested heavily in the provision of training. However, we did not see any evidence to show that managers had evaluated the effectiveness of the training received. From the evidence we have seen we question whether it has received value for money when judged by the quality of the investigation found in our sampling. For example, an IUC being conducted was abandoned because it was taking too long.

Fraud referrals

2.428 Good quality referrals are important to the success of any fraud investigation section. We would expect the majority of the referrals to originate from the benefits section but in Lambeth we found that the rate of referrals from benefits assessors is in decline.

2.429 CAFT receives referrals from a variety of other sources apart from the benefits section. Figure 2.64 shows the numbers of these referrals over the past 3 years. CAFT has done proactive exercises in low earners, landlord and hostel initiatives, student awards, and Do Not Redirect and electoral roll reconciliation. The data on these is included at other specific initiatives.

Fig. 2.64: Sources of referrals – 1997/98 to 1999/2000
Source	1997/98	1998/99	1999/2000
Benefits section	1,018	583	201
CAFT	323	621	321
Other internal	6	176	62
BA/BFIS	0	37	45
Data matching – National Fraud Initiative (NFI) and Housing Benefit Matching Service (HBMS)	159	590	217
Fraud hotline	0	51	96
Other external referrals	45	142	76
Other specific initiatives	165	230	119
Total	1,716	2,430	1,137

Source: Lambeth

2.430 We found that the backlog of outstanding work in the benefits section at the assessment stage is having a negative impact on the level of fraud referrals. Referrals from this source have declined from approximately 63% of all referrals to less than 17%.

2.431 We found that the reasons for this drastic decline are the:

low priority that we were told Capita gives to possible fraudulent claims
restrictive procedure HB staff believe is in place for making a referral
inexperience and poor training of assessment staff
high turnover of assessment staff.

2.432 We found that the number of referrals compared to caseload was low. This is shown in Figure 2.65. In our sample of new and renewal claims we found that 5 (15%) of new claims and 4 (12%) of renewal claims should have been referred for investigation. In the past 3 years the percentage of referrals from the benefits section as a proportion of total caseload has declined from 1.89% in 1997/98 to 0.5% in 1999/2000. This is worrying in view of Lambeth’s failure to secure the gateway to the benefits system.

Fig. 2.65: Benefits section referrals and fraud established as a percentage of caseload – 1997/98 to 1999/2000
	1997/98	1998/99	1999/2000
HB caseload	53,806	46,836	37,357
Number of referrals	1,716	2,430	1,137
Number of referrals as a % of caseload	3	5	3
Number of referrals where fraud was established	539	1,131	517
Fraud established as a % of caseload	1	2.5	1.3

Source: BFI analysis

2.433 We tested our sample of 70 fraud files for source and outcome of referrals. The outcome is shown in Figure 2.66. We found 4 cases where it was not possible to identify the source of the referral. This shows that the fraud case management system information may not be reliable. Although no referrals from BA and BFIS were found in the sample, we were told that 30 were referred in 1999/2000. The low level of referrals reflects the fragmented interface that Lambeth has with 5 different BA fraud sectors making day-to-day liaison difficult. Most of the referrals in our sample came from the VOs, who were involved in proactive visits to claimants. Only 11% of referrals in the sample were from benefits assessors. The poor quality of the referrals from this source is also highlighted by the fact that none resulted in a fraud being established.

Fig. 2.66: Source and outcome of referrals in our sample
Source	Number	Fraud established	% of sample
CAFT	23	9	13
Other LA staff	12	6	9
HBMS	9	8	11
NFI	3	3	4
Benefits assessors	6	0	0
Do Not Redirect	3	3	4
Other (including fraud hot line)	10	3	4
Source not known	4	1	1
Total			47

Source: BFI analysis

2.434 CAFT uses a standard referral form BF1. This is used for all internal referrals except for those raised by the VOs who use a multi-purpose control sheet. The form requests the following information:

a copy of the HB form
any additional evidence
any other relevant information.

2.435 Once an investigation has been completed a form BF3 is sent to the referrer. We were told that in the case of the benefits assessor, this was more likely to be the section than the individual who made the referral.

2.436 In the 70 sample fraud cases we found:

4 cases with no fraud file, only a form showing the result of the investigation
4 cases where the origin of the referral was not immediately apparent
evidence of a failure to systematically prioritise cases.

2.437 We were told that there were occasions where a telephone referral taken by an investigator may not always be logged and that the Group Fraud Manager (GFM) may not be informed. The risks of this are that cases are not properly risk assessed and that cases that could warrant investigation are not followed up and management information will not be recorded. This is particularly true of informal overload cases from BFIS where an investigator may take a referral over the telephone and make a decision whether or not to investigate it without recording it.

2.438 At the time of the inspection Lambeth was not carrying out any formal risk assessment of referrals. While on site we were shown a draft referral risk assessment form to aid investigators in prioritising their work. This was to be introduced immediately.

2.439 We analysed the source of referrals in the past 3 years against the number that resulted in a fraud being established. The results are shown in Figure 2.67. We found that the most fruitful referrals in terms of a fraud being established were from data matching and other specific initiatives. On average 74% of data matching referrals and 71% of other specific initiatives led to a fraud being established in the period 1997 to 2000 and 71% of other specific initiatives. In contrast, for the same period only 15% of referrals from the benefits section led to a fraud being established. This raises the issue not only of the volume of referrals from the benefits section, but also the quality. CAFT undertakes no analysis of the failed referrals.

Fig. 2.67: Source of referrals and fraud established 1999/2000

We recommend that Lambeth obtains better quality referrals from the benefits section by:

increasing the number of fraud awareness seminars for assessment and counter staff
simplifying the procedures for benefits staff to make a referral requiring less information
considering permanently locating CAFT staff in the benefits section.

We recommend that:

all referrals are systematically sifted and prioritised by using risk analysis techniques
formal feedback of outcomes is given to individual referrers in the benefits section
all referrals are logged on to the FACTS database
fraud files are created for all accepted referrals
all referrals are risk assessed to aid prioritisation
rejected referrals are analysed to inform future fraud awareness training for benefits staff.

Data matching referrals

Fig. 2.68: HBMS referrals against fraud established – 1997/98 to 1999/2000

Source: BFI inspection

2.440 Lambeth has successfully participated in the HBMS programme since it started. Since 1997 HBMS has proved to be a valuable source of referrals with an average of 74% leading to a fraud being claimed by Lambeth, as shown in Figure 2.68. In our sample of 70 fraud files we found 9 that had originated from this source, of which 8 had led to a fraud being established.

2.441 Lambeth also carries out internal data matching by participating in the National Fraud Initiative. In our sample, we found 3 cases that had originated from this source, all of which led to a fraud investigation. However, possible prosecution action in 2 of the cases was compromised by delays in investigation.

2.442 EA commented in the management letter of December 1999 that there was a delay in investigating a significant number of matches. It found that 1,945 pension matches and 1,209 invalid NINO matches were still to be investigated. Although these cases may not have included HB or CTB claimants, Lambeth’s failure to act promptly could have resulted in potential fraud cases not being detected. EA also stated that out of 1,242 HB matches 105 had not been investigated. Those that had been investigated had resulted in WBS of £23,000. However, CAFT informed us that all matches had now been investigated. We saw the returns showing that all data had been submitted for NFI 2000. The exercise yielded WBS of £50,000 and fraud overpayments amounting to £59,000.

2.443 CAFT also has CAFIS, its own internal data matching system which is able to undertake internal data matching against the following records:

Academy benefits and CT systems
electoral roll
education awards
payroll
housing accounts
market traders
parking permits.

2.444 We consider this to be good practice. However, we did not see any data showing how effective these matches were.

Do Not Redirect

2.445 LAs were given powers in 1999 to require the Royal Mail to return benefit post even where a paid for redirection notice is in force. Section 182A of the SSAA 1992 requires the Royal Mail to return benefits post to the LA provided that it has been sent in special envelopes marked Do not redirect. Used effectively, Do Not Redirect is a valuable weapon in preventing and detecting fraud.

2.446 Lambeth did not implement Do Not Redirect until November 1999 and XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXX
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX X
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX X

2.447 Cheques and giros returned by the Royal Mail are referred to the Payments Manager. However, we found that no record is kept of the number or value of payments that are returned, nor the reasons for return. By not recording this information, Lambeth is missing an opportunity to measure the success of the Do Not Redirect scheme.

2.448 Cheques returned under this scheme were not routinely being forwarded to CAFT for investigation. We were told CAFT had asked for referrals to stop because the backlog of work outstanding on the payments section was leading to significant delays in forwarding the details for investigation. We consider Lambeth to be missing an opportunity to detect fraud.

2.449 In our sample of 70 cases we found 3 cases investigated as a result of a Do Not Redirect referral within our sample. All led to the establishment of a 75% WBS.

We recommend that Lambeth:

make better use of the Do Not Redirect mail scheme, by extending its coverage to include all outgoing benefits correspondence
refer all cases where mail has been returned to CAFT for investigation
monitor the number and value of all mail returned.

Quality of fraud cases

Delays in investigation

2.450 We examined our sample of fraud investigations to establish the average time taken to complete an investigation. We found evidence of delay at all stages of investigation. Figure 2.69 shows the average time taken for each stage:

Fig. 2.69: Analysis of time taken to complete investigations in our sample cases
Stage of investigation	Average time taken (calendar days)
From receipt of referral to raising a fraud file	17 days
From raising a fraud file to first investigative action	9 days
From first investigative action to closure	78 days
Total time taken from receipt of referral to closure	104 days

Source: BFI analysis

2.451 In 12 cases there were significant delays in excess of 4 months during the investigation stage, including:

one case where no action had been taken on a suspected "gone away" 12 months after the date of referral
one case where no action had been taken on a hotline referral 13 months after the date of referral
one case where no action had been taken on a suspected living together 16 months after the date of referral.

2.452 In particular, we found a case where excessive delay during the course of the investigation had compromised prosecution action and had led to a substantial LA error overpayment. Figure 2.70 provides details.

Fig. 2.70: Case study of a delayed fraud investigation
Date	Action taken
September 1998	Data match of HB claims to LB Sutton payroll records, as part of NFI shows claimant to be employed. Claimant also in receipt of IS.
25 April 1999	Evidence obtained from LB Sutton confirming start date and basic weekly wage. HB not suspended, BA not informed of suspected fraud.
13 October 1999	Fraud file raised. HB not suspended, BA not informed of suspected fraud.
17 and 23 November 1999	Ineffective visits carried out at claimant’s address. Benefit not suspended, BA not informed of suspected fraud.
7 December 1999	Letter sent to claimant requesting they attend for IUC.
16 December 1999	Claimant interviewed under caution. Interview abandoned as taking too long. Claimant admitted defrauding Lambeth. HB and CTB WBS claimed at £63 per week.
January – February 2000	CAFT notified BA of the fraud and the outcome of interview by SLA3. Fraud overpayment created from 28 September 1998 to 25 April 1999. LA error overpayment created from 26 April 1999 until date of IUC, because of delay by CAFT in acting upon the information held. BA confirmed IS overpayment of £6,000. IS WBS at £100 a week. BFIS advises it is keen to prosecute.
9 March 2000	Case closed without any sanction having been applied. No further liaison with BFIS.

Source: BFI analysis

2.453 Lambeth commented that the case referred to in Figure 2.70 had been delayed due to staff shortages at the time.

2.454 In our sampling exercise we found evidence of delays following the IUC. For example we found:

claimants who had been interviewed under caution were not informed promptly whether a prosecution would follow
IUCs were suspended and not resumed

a case where 8 months had elapsed from the date of receipt of an HBMS referral, until an IUC was conducted in January 2000. The IUC was suspended, because the claimant denied working for the alleged employer. Despite further and stronger evidence being obtained the day after the interview, no attempt had been made to resume the interview at the time of our inspection

a case where 8 months after the IUC had been conducted no decision had been made on whether to prosecute
a case where the IUC was conducted in July 1999, but the decision on whether to prosecute had not been taken at the time of our inspection in October 2000.

2.455 We consider that such delays:

place unnecessary constraints on the success of Lambeth’s counter fraud activity
prejudice the likelihood of a successful prosecution
allow overpayments to increase
send out the wrong message to fraudsters
are likely to discourage anonymous informants and staff referrals.

2.456 CAFT management acknowledged the existence of these delays and attributed them to under-resourcing in both the fraud investigation teams.

2.457 Lambeth had taken steps to improve its performance in July 2000, shortly before our inspection, and we were informed that fraud investigators had recently been set a target of completing 75% of all investigations within 13 weeks. We found that there had been an improvement in the most recent cases in our sample.

We recommend that, to improve the speed of investigations, Lambeth ensures that:

all cases suitable for investigation are referred to investigators as quickly as possible
fraud managers review progress on open fraud cases at more frequent, regular intervals.

File construction

2.458 We looked at how the investigation files in our sample were constructed and found that they lacked copies of the relevant claim documents and information that should have been readily available to CAFT, such as printouts of the amounts and periods of benefit, council tax registration information and electoral roll entries.

2.459 We also found that file construction was poor in that:

VO non-residency cases lacked an adequate activity log to record visits and outcomes
cheques and giros were not filed in protective wallets to prevent unnecessary handling
documents were not secured within the files
VOs’ control sheets containing sensitive information were also used as statement forms
an anonymous letter was not date stamped received
one case was erroneously created in the name of the informant
activity logs were not routinely completed
there were 5 cases where no file had been raised
notes of telephone calls and messages were unsigned and undated.

2.460 The poor file construction means that:

there is little or no audit trail
there is insufficient information on file to conduct an interview under caution with the claimant
it would be difficult for another investigator to take over the case
there is often insufficient evidence to support the conclusions reached by the investigator, including those for WBS claims.

2.461 We found that fraud investigators used customised investigators’ notebooks produced by Local Authority Investigation Officers’ Group (LAIOG). These notebooks contain serially numbered pages and a margin. We found that these notebooks were well maintained by Fraud Investigation Officers (FIOs). We consider the use of these notebooks good practice and a potential management tool because, if used effectively they can:

improve the quality of time and record keeping
provide valuable management information on the costs, effectiveness and time management of the fraud staff
ensure that a record taken at the time can be produced in court as evidence if there is a dispute or a complaint.

We recommend that, to improve the quality of management information, fraud managers should check investigation officers’ notebooks at regular intervals. This information can be used to inform on the effectiveness of its fraud investigation teams.

Investigative approach

2.462 We found considerable evidence of poor quality investigations in our sample of 70 fraud files. For example, we found one case where the claimant admitted during an IUC that she had a second job, but the statement taken under caution failed to incorporate details of the second job or her earnings.

2.463 Of the 70 cases we examined, 20 cases, relating to the question of non-residency, originated from the VOs. We found that very superficial enquiries had been conducted in these non-residency cases. In almost every case, the investigation consisted solely of 2 ineffective visits and a letter left at the property asking the claimant to contact the CAFT. For example we found very limited evidence of enquiries made with:

landlords

other occupiers

neighbours
BA
other LAs.

2.464 Where a claimant is found to be non-resident or "gone away" the LA is able to claim 75% WBS if the criteria outlined in the Subsidy Order are satisfied. An LA must undertake:

at least 2 unannounced visits to the dwelling in respect of which the relevant benefit was paid or allowed to the claimant and which:
- took place in what were, for the claimant, different benefit weeks; and
- established that the claimant had ceased to occupy the dwelling as his home; or
one announced visit to the dwelling in respect of which the relevant benefit was paid or allowed at which it was established beyond doubt that the claimant had ceased to occupy the dwelling as his home.

2.465 We consider that, to establish whether or not a claimant has ceased to occupy a property as his home, it is necessary to provide corroborating evidence to support the evidence of the ineffective visits.

2.466 The absence of supporting evidence from other potential sources of information meant that, in most cases, non-residency relied upon a failure to respond to the letter left at the property on the second visit. The limitations of this approach are that:

the claimant may in fact still be resident
the date of vacation cannot be established
fraudsters may evade detection for past periods of non residence
the full extent of fraudulent overpayments may not be calculated
WBS claims may be invalid.

2.467 We found that, even where a forwarding address was obtained, there was no evidence to show that enquiries had been pursued at the second address, via other LAs or BA, to try to establish a date of vacation, and potential overpayment.

We recommend that, to ensure the validity of non residency or "gone away" WBS claims and to ensure that fraudulent overpayments are fully identified, Lambeth:

ensures that all investigators are familiar with the no residency or "gone away" guidance contained in Circular HB/CTB F17/98 and the Income-Related Benefits (Subsidy to Authorities) Order 1998
makes full enquiries to confirm that claimants have ceased to occupy properties as their home
seeks to identify the date of change of circumstances wherever possible.

Investigation techniques

2.468 We found poor and inappropriate investigative techniques had been used in our sample of 70 fraud cases. We found that:

enquiries were not always fully pursued
statements were poorly completed
enquiries were limited and generally superficial.

2.469 Out of our sample all cases but one had been investigated. We found examples of poor and inappropriate techniques in 15 (21%) cases. The details are as follows:

2 cases where the claimant should have been interviewed under caution
2 cases where the IUC had been poorly conducted
8 cases where poor quality statements jeopardised the success of future sanctions
3 cases where obvious lines of enquiry and allegations had been completely ignored.

We recommend that, to improve the quality of fraud investigation techniques:

Lambeth provides refresher training on good practices to the fraud investigators
fraud managers set quality standards and conduct regular checks to ensure standards are met.

Outcomes of investigations

2.470 We examined 25 cases where a WBS had been claimed. At the time of our inspection, Lambeth had not submitted its WBS claim to DSS for 1999/2000, so all of these cases were still potentially subject to validation by EA. We tested each of these cases against the criteria set out in the Subsidy Order. Figure 2.71 shows our findings.

Fig. 2.71: Analysis of validity of WBS claims
BFI finding	Number	% of total
Valid	8	32
Invalid	17	68
Total	25	100

Source: BFI analysis

2.471 We found:

7 cases where a WBS had been claimed although fraud had not been established
10 cases where the investigator had not established that the claimant had ceased to occupy the dwelling as his home
100% WBS instead of 75% WBS had been incorrectly claimed in 7 of the 10 invalid "gone away" cases.

2.472 At the time of our inspection, Lambeth was checking all 1999/2000 WBS claims, in response to concerns about the validity of WBS raised by EA during the previous audits.

2.473 Lambeth told us that the errors we found would have been detected as part of the additional management check of WBS claims before submission of the subsidy claim to EA. However, we were unable to test whether this would have been the case, as none of the cases we selected had been subjected to the management check at the time of our inspection.

2.474 We consider that:

there needs to be a more thorough management check of WBS claims
staff need more training to ensure that WBS claims are valid
there are serious doubts as to the validity of the remaining WBS claims made by Lambeth during 1999/2000.

We recommend that Lambeth ensures that WBS claims are valid by:

implementing a thorough management check of all WBS claims
providing staff with appropriate training on the criteria required to claim a WBS.

Duplicate WBS claims

2.475 We examined all the WBS claims recorded on the FACTS database for 1999/2000. We found 16 duplicate claims for WBS, which we alerted Lambeth to. Lambeth found that 8 were legitimate claims as they were for separate investigations but agreed that the remainder were duplicates and agreed to withdraw each claim. The total value of the WBS on these 8 claims was £20,419.84. Lambeth informed us that the duplicate claims arose because different investigators had worked on the same investigation and had subsequently submitted duplicate WBS claims. We discussed this with Lambeth and were told that since April 2000 all WBS claims had been checked by the GFMs to prevent duplication.

Adverse impact of failure in benefit administration on fraud investigations

2.476 The weaknesses in benefit administration are having a negative impact on CAFT’s work. In particular we found:

compromised investigations because of delays and lack of documentary evidence
a potential delay of up to 28 days when, after the investigation has been completed, a benefit file cannot be located for determination purposes. In theory, after the investigation has been completed, and where evidence affecting benefit entitlement has been obtained, CAFT refers the evidence to the CAFT Liaison Officers at OMH. The Liaison Officers request a search for the benefit file. In the meantime, no action is taken on the evidence obtained, even where ongoing entitlement is affected. If the file has still not been found after 28 days, the Liaison Officers refer the evidence without the file to the determination officers. The determination officers would then have to make a decision based solely upon the evidence provided by CAFT and whatever information was held on Academy
evidence supplied by Lambeth was that 40% of benefit files could not be found for cases that required a redetermination of the claim following a fraud investigation
evidence that Capita would not allow a live claim to be suspended, despite a strong suspicion of fraud. Neither CAFT nor the CAFT liaison officers have the ability to suspend claims. We were told that this was due to past problems where claims had been suspended without sufficient evidence, causing claimants and the HD to complain about non-receipt of benefit. I addition ,of course, it is a matter for Lambeth not Capita to decide
that information contained in benefit files and on the Academy system could not be relied upon to be up to date because of the backlog in processing claims
that CAFT resources and time were frequently wasted by pointless enquiries and investigations because of unactioned correspondence at OMH
a failure to tailor benefit periods in WFTC cases leading to needless fraud investigations
delays in assessing benefit claims meant that the entitlement information held on the Academy system could not be relied upon for accuracy. This made it difficult for management to check WBS cases within a reasonable period of time
delays in dealing with changes of circumstances created opportunities for fraud because claimants could be tempted to retain benefit to which they knew they were not entitled
a lack of control over incoming correspondence providing fraudsters with the excuse that they had notified the LA of changes of circumstances knowing it would be difficult to establish that they had not.

2.477 We consider that Lambeth implementing our recommendations in Getting it right and Keeping it right should address the weaknesses in benefit administration and minimise the adverse impact on the success of fraud investigations.

Visiting Officers

2.478 VOs work within the CAFT. We were informed that each VO has 2 performance targets, to:

conduct 40 visits per day
achieve £30,000 WBS per month.

2.479 An unannounced visit is made to the property. If no contact is made with the claimant, a letter is left informing them that a second visit will be made in the following week. If no contact is made on the second visit, a further letter is left informing the claimant that they have 14 days to contact CAFT, failing which the claim will be cancelled.

2.480 The sole purpose of the visits is to establish residency and identity. VOs are under instruction not to pursue any suspicion of fraudulent activity that may arise during the course of a visit. Any suspicions should be referred to an FIO.

2.481 As we have already discussed in Getting it right we are concerned that there is no liaison with the benefits section to identify cases where residency needs to be checked by a visit.

2.482 The VOs estimated that contact was made with the claimant in only 20% of first (unannounced) visits. However, they estimated that contact was made with the claimant in 80% of second (planned) visits.

2.483 The VOs told us that the target of 40 visits per day did not allow sufficient time to make any enquiries other than as described above.

2.484 We consider that the effectiveness of the current visiting programme is limited because:

XXXX XXXX XXXX XXXX X
the sole purpose of the visit is to establish residency and not to confirm whether there has been any change in the circumstances of the original benefit claim
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XX
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX X
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX
XXXX XXXX XXXX XXXX XXXX XXXX
the monthly WBS target is an unreliable performance indicator
opportunities to question claimants and third parties to confirm information in the benefit claim may be missed
there is no liaison with BA to avoid duplication of visits.

2.485 We found that in our sample of 70 investigations, the VOs had generated 25 cases (36%). Of these, 4 cases had led to a suspicion of undisclosed income and one case had led to a suspicion of living together as husband and wife.

2.486 However, we found that the quality of the evidence obtained during the visit, and the lack of enquiries pursued into non-residency, led to doubts as to the validity of the WBS. We also found that 100% WBS had been claimed instead of the 75% allowed in "gone away" cases. We comment further on the quality of investigations later in this section.

2.487 We consider that the visiting programme produces few results for the level of resources invested. The success of the initiative could be enhanced by:

better targeting of resources on high-risk cases only
conducting more in-depth interviews to confirm claimants’ circumstances and entitlement to benefit
providing training on evidence gathering to VOs.

2.488 At the time of our inspection CAFT was undertaking a proactive risk project based on the area housing benefit reviews and sampling some 400 households in receipt of benefit. We found, however, that this will have a limited application because:

excluding a section of claimants means that the sample will be skewed
there was a failure to question or challenge inconsistencies during the visits
there is no intention to use the information gained to inform benefit administration.

We recommend that, to improve the effectiveness of Lambeth’s fraud visiting programme:

it is extended to include visits to RR cases
it includes a full verification check on all aspects of the claim by the VO
the outcomes of all visits are reported back to Benefits by CAFT
resources are targeted towards high-risk cases
VOs receive training on evidence gathering
performance targets for VOs are linked to completed visits
VOs liaise with BA.

Management checks

2.489 Regular management checks on casework are essential to ensure proper management control over the investigative process. Implementing a programme of checking at regular intervals would help:

focus investigative effort and resources on the most promising cases
minimise delays in investigation
improve consistency in investigative approach and techniques
drive up the standards and quality of fraud work
identify training needs more quickly
better inform the staff appraisal process.

2.490 We tested the frequency and the quality of management checks and found:

only 10 cases had been subject to a management check
the checks had been conducted at the end of the investigation process.

2.491 We found that there were significant differences between the level and frequency of management checks done in the 2 benefit investigation teams:

the North Team GFM reviewed the caseload of his investigators at fortnightly intervals, whereas the South Team GFM reviews were conducted monthly
some FIOs reported that every case was reviewed, whereas others reported that only a small selection of their cases had been reviewed or that they reviewed their own cases.

2.492 We consider it good practice for fraud managers to occasionally accompany fraud staff on interviews and visits, including unannounced visits. This allows them to:

observe practices, give advice and identify training needs of individual fraud officers
conduct informed appraisal interviews.

2.493 We found that there was no routine programme for GFMs to accompany all FIOS and the practice of accompanying FIOs varied between the GFMs.

2.494 We found evidence that the closure of an investigation was not always authorised by a GFM. We consider that all case closures should be authorised by a GFM to ensure:

management control over fraud investigation work
consistency in approach
all appropriate lines of enquiry have been pursued
separation of duties between the investigative action taken by the FIO and closure action authorised by a GFM.

We recommend that, to improve the effectiveness of management checks on fraud work, CAFT:

implements a thorough and regular checking and monitoring regime by GFMs on all fraud investigation cases
ensures consistency in the frequency and extent of those cases selected for check by developing guidance for the GFMs
introduces a programme of regular accompanied visits, including unannounced visits
ensures that all cases under investigation are reviewed by GFMs at regular intervals to ensure that the necessary progress is being made
requires that closure of all cases is authorised by GFMs.

Sanctions

2.495 Lambeth has a comprehensive prosecution policy and a procedures document for administration penalties. The introduction to the policy states that:

…prosecution action will only be taken where it is in the public interest to do so. The purpose of this policy is to ensure that the decision to prosecute can be justified as fair, reasonable and consistent.

2.496 Lambeth has clear criteria as to whether prosecutions should occur. These are shown in Figure 2.72.

Fig. 2.72: Lambeth’s criteria for prosecution
Value	Action
Less than £400	Recovery of the overpayment will be an appropriate sanction rather than prosecution.
£400 to £1,500	Administrative penalties will be used in most cases as an alternative to prosecution (the case must have been suitable for prosecution within the terms of this policy). Prosecution action should be taken where a person declines to accept the administrative penalty.
£1,500 to £5,000	Consider prosecuting under SSAA 1992, Theft Acts or other criminal offences. Such cases are not suitable for administrative penalties.
£5,000 and over	Referral to the police for prosecution under section 111A of the SSAA 1992, or Theft Act or other criminal offences.

Source: Lambeth

2.497 The policy clearly sets out the criteria to be applied when Lambeth consider whether it is in the public interest to prosecute. Lambeth considers:

whether the person was in a position of trust, for example an employee
whether there was a degree of premeditation
the personal circumstances of the offender.

2.498 The policy also states that each case must be looked at on its merits and a prosecution should only takes place with the agreement of the Operations Manager or the Head of CAFT.

2.499 Cases deemed suitable for prosecution are referred to the Borough Solicitor or the police for action. We were told that a prosecution panel was going to be formed to identify those cases that were likely to be suitable for a prosecution at an earlier stage. Figure 2.73 shows that there have been only 9 prosecutions in the past 3 years. For 1999/2000 the 5 shown represent only 0.4% of the referrals CAFT received.

Fig. 2.73: Prosecutions – 1997/98 to 1999/2000
	1997/98	1998/99	1999/2000
Prosecutions initiated	1	3	5
Successful prosecutions	1	2*	2*

Source: Lambeth

* cases still pending.

2.500 In our sample we found 33 cases where a fraud had been established by investigation and 4 where a prosecution could have been considered. These included:

one case where the Lambeth overpayment was approximately £3,000, an IS overpayment was £6,000 and the claimant had admitted claiming and working.
one case where it was established that the claimant was working and there was an overpayment of £4,200 and a BA overpayment of £7,600
one case where a claimant appears to have made 2 claims using false names. The claimant had had an IUC but there had been no action since May 1999.

2.501 All of the investigators had experience in carrying out IUCs so the low number of prosecutions and other sanctions in this context is disappointing.
Figure 2.74 shows a case study of a successful prosecution.

Fig. 2.74: Case study of a successful prosecution
Date	Action
18/1/1999	Case referred from benefits section, which was concerned that the claimant’s earnings had not increased for 2 years. Claimant has stated that he earns £45 per week as a part-time worker.
13/4/1999	Request is made to employer for pay details.
30/4/1999	Pay details received from employer who states that the claimant is working as a contract engineer for 37.5 hours per week for a salary of £20,000 a year. Letting agent is contacted to establish that rent is being paid and the current situation with the account which was found to be in arrears.
6/5/1999	Statement taken from employer. It appears that the claimant had been manipulating the earnings certificates from the employer. Printout of pay details obtained.
18/5/1999	Claimant is interviewed under caution where he denies the allegations until the employer’s statement is produced.
14/6/1999	Overpayment is calculated which amounts to over £10,000.
27/7/1999	A summary of offences is prepared for the legal section and the case is authorised for prosecution.
4/1/2000	Claimant appears in court and pleads guilty to 3 charges under the Theft Act.
1/2/2000	Claimant is ordered to perform 160 hours of community service together with compensation of £10,000 and £200 costs.

Source: Lambeth

2.502 This case shows that led to a substantial loss of public funds, could have been eliminated if Lambeth had followed basic procedures, verified properly and had not included the pay form with the HB claim form.

Administrative penalties

2.503 Lambeth has used an administrative penalty twice since the powers were granted. We were told that at least another 8 cases had been identified as potential administrative penalty cases and they were being investigated.

2.504 The effectiveness of administrative penalties relies in turn on the efficient recovery of overpayments. At present only 14% of overpayments where benefit has ceased are likely to result in any form of recovery. It is therefore highly unlikely that administrative penalties will be an effective sanction while the present low level of overpayment recovery continues.

Cautions

2.505 As a prosecuting authority Lambeth is able to administer formal cautions in appropriate cases. It does not have any policy or procedures to do so at present and has not carried any out.

We recommend that Lambeth

seeks to identify at an earlier stage those cases that may be suitable for a prosecution or other sanction
develops procedures to administer cautions where appropriate.

Whistleblowing policy

2.506 In line with the Public Interests and Disclosure Act 1998 Lambeth has a whistleblowing policy. Leaflets are available for all employees explaining how to report any concerns. Posters displayed prominently in council property support this. While CAFT is given as one point of contact there is also a dedicated whistleblowing telephone line that is managed by Expolink. Expolink is an independent external company that specialises in running whistleblower services in both the public and private sectors. The service is provided externally at the wish of Members. All members of staff have been issued with this number in case they wish to report any concerns.

We recommend that:

Lambeth develops a new counter fraud strategy making reference to the BFI’s Good Practice Guide, taking into account the changes within CAFT, the changes in reporting lines and the new decision-making structure
the strategy:
- makes particular reference to HB/CTB fraud
- makes reference to the full range of sanctions open to Lambeth
- includes a definition of fraud and corruption.

2.507 The advantages of implementing this recommendation will be that:

Members can renew their public commitment to fight fraud and corruption
a clear message will be sent to those who seek to defraud Lambeth
policy will be in line with the new strategies applied by Lambeth to fight fraud.

Working with BFIS

2.508 Lambeth interfaces with a number of BFIS sectors, Balham, Brixton, Wandsworth, Bankside and Crystal Palace/Peckham. There are SLAs with each of these sectors, but for ease of monitoring Balham takes the lead on behalf of the other sectors. The SLAs are based on the national model and have not been tailored to local needs.

2.509 The SLA between Lambeth and Balham requires that performance is actively monitored and that a certificate of performance is signed off each year. We found that:

SLA monitoring meetings take place on a quarterly basis
day-to-day relations with CAFT are good
BFIS has a nominated liaison officer based in Brixton
both parties have signed off the certificate of performance.

2.510 However, we do not consider that the SLA is being actively monitored.

2.511 Lambeth and BFIS have undertaken some joint working on students and market traders in the past 2 years that were deemed successful. In the case of student awards there were 48 cases that involved HB. From this, 18 prosecutions and 2 administration penalties resulted, all undertaken by BFIS. A further 24 overpayment cases were identified of which 7 had an HB element. The market traders’ drive identified only 2 cases where there was an HB element. XX
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXX

2.512 A member of CAFT has spent time working with BFIS. A member of BFIS worked with CAFT on joint investigations some 2 years ago. CAFT has suggested undertaking joint IUCs, but we were told that BFIS has been reluctant to do this. However, BFIS has provided fraud awareness training for CAFT, but has not received any from Lambeth. BFIS would be quite willing to provide further training if requested.

2.513 BFIS does not formally overload cases to Lambeth but some cases are referred informally if it is felt there is a Lambeth interest. The process for a formal referral is to record the referral on the Fraud Information By Sector (FIBS) system, and post a copy to the LA with an accompanying HBF2 form. BFIS was concerned that to overload formally would create too much paperwork.

We recommend that:

fraud awareness training for each organisation’s staff is carried out regularly
further exchanges of personnel take place
Lambeth and BFIS develop procedures to monitor the SLA
Lambeth and BFIS develop a strategy for closer working
BFIS formally overloads appropriate cases.

Working with other bodies

2.514 CAFT has helped develop a public sector counter fraud charter under the auspices of the Chartered Institute of Public Finance Accountants (CIPFA) and the Institute of Public Finance. The work CAFT has done on this has been in partnership with the Executive Directorate of Counter Fraud Services of the NHS, the AC, National Audit Office and DA. The charter seeks to identify problems, apply a common counter fraud strategy, identify a common business process and to deploy common specialist skills and raise public awareness over the level of fraud in the public sector.

2.515 CAFT is also involved in improving links with the police under the Crime and Disorder Act 1998. It has signed the protocol for partnership agreed between London Authorities including health authorities Borough Police and the Metropolitan Police Fraud Squad (SO6). The aim of this is to improve working links between the various organisations including named liaison officers, guidelines for co-operation at an operational level.

2.516 While we commend CAFT for its liaison work with other bodies we are concerned in case it is distracted from its core activities of combating fraud in Lambeth.

We recommend that Lambeth, as a senior partner in counter fraud initiatives with external bodies, invites BA/BFIS and BASIS to take part, to widen the representation and effectiveness of these bodies.

Overpayments

2.517 The management of overpayments is an essential part of the government’s strategy to combat fraud and error to ensure that:

the number of overpayments is kept to a minimum
overpaid benefit is recovered, where possible
any debts that cannot be recovered are properly accounted for.

2.518 As part of our inspection process we examined a sample of 70 overpayment cases and analysed these with specific emphasis on the:

identification and recoverability of overpayments
quality of the assessment of overpayments
accuracy of the overpayments
success of recovery action
use of management information.

2.519 BFI’s Good Practice Guide highlights the benefits of effectively managing the end-to-end process of overpayment recovery. We reviewed Lambeth’s management of overpayments by:

interviewing staff
examining written guidance
examining management information
sampling 60 overpayment cases.

2.520 We concentrated on the following areas:

avoiding overpayments, either by acting promptly on notified changes of circumstances or publicising the claimant’s and the landlord’s requirement to report changes
targets
management checks and information
training
guidance and procedures for staff
identification of overpayments
determining cases after a change of circumstances
classification of overpayments
recovery of overpayments.

Fig. 2.75: Overpayments raised and recovered – 1998/99 to 1999/2000 (£ million)

Source:Lambeth, monthly statistics Note: Recovery figures for 1999/2000 include "bulk" adjustment recoveries

2.521 The latest figures supplied by Lambeth (September 2000) show that there is £32.3 million in overpaid HB owed to the LA. Some of these overpayments go back up to 10 years.

2.522 Figure 2.75 shows the amount of overpaid HB that Lambeth identified in 1998/99 and 1999/2000, compared to the amounts recovered in each year. Lambeth has a significant problem in that it recovers less than half the overpayments it raises.

2.523 In 1999/2000 Lambeth paid out HB to the value of £3.34 million due to errors made by the LA, which represents approximately 3% of the total spend on HB. This money is deemed non-recoverable unless the claimant or landlord could have known they were being overpaid.

2.524 However, this is based on Lambeth’s own figures. Our inspection has found that the full amount of an overpayment is not always recorded. We also established that Lambeth does not always record the full value of LA error overpayments, so the figure could be higher.

2.525 Between 1998/99 and 1999/2000 there was a drop of 17% in the value of all overpayments recorded. The biggest drop was in the value of fraudulent overpayments, where there was a 34% decrease. However, this must be set against a backdrop of:

an overall decrease in the amount of HB paid to claimants (between 1998/99 and 1999/2000 there was a 10% drop in the amount of HB paid)
Lambeth’s conversion to the Academy system and the downtime on benefit processing. There could have been an under-recording of overpayment because notifications of changes of circumstances are awaiting action in the backlog of work.

2.526 Figures for the period September 1999 to August 2000 show an average monthly overpayment of £0.99 million, with a peak of £1.56 million in August 2000, as shown in Figure 2.76.

Fig. 2.76: Overpayments raised – September 1999 to August 2000 (£ million)

Source: Lambeth, client monthly pack September 2000

Note: December 1999 is not included as figures were unavailable due to the conversion to the Academy system

2.527 Despite some fluctuations, based on current trends we expect overpayments to be in the region of £2 million per month by March 2001. We judge that there have been a number of factors that have led to Lambeth’s poor performance with regard to managing overpayments. We discuss these below.

Managing overpayment recovery work

2.528 Lambeth has a number of approaches to managing the debt it is owed:

a Council-Wide Debt Forum (CWDF). The Forum’s terms of reference set out that it is to:

act as a body where performance on debt is monitored and where good practices are shared

the overpayments project team, responsible for analysing some of the older debts held by Lambeth. Where possible these debts will be collected. The team is employed directly by Lambeth reporting to the HCC
overpayments are initially identified and determined as with any other HB or CTB determination. The assessment and determination teams decide on recoverability before passing the debt to an overpayment recovery team, part of Capita. This team is responsible for all other aspects of HB overpayment recovery in the LA.

2.529 We were extremely concerned to note that, as late as October 2000, there was still an ongoing issue between Lambeth and Capita as to who had responsibility for pre and post contract debt. We consider that, other than as a short-term measure, recovery and management of debt should sit with one team. This reduces the potential for duplication and poor customer service.

We recommend that all recovery action should sit with one team to prevent duplication of effort.

Management and recovery of overpayments

Policy and procedures

2.530 BFI’s Good Practice Guide highlights the benefits of having up-to-date policy and procedures available for staff to refer to. We compared the processes in Lambeth with its advice.

2.531 There is no one source of guidance on overpayments to which staff can refer covering:

prevention of overpayments (such as suspending a claim when it seems that there has been a material change in the claimant’s circumstances)
ensuring that the correct periods of the overpayment are input
classification
recoverability
from whom to recover
amounts to be recovered
methods of recovery
write-off.

2.532 Lambeth has the following written guidance available to staff:

a user manual for the Academy system
a working draft of a write-off policy. This is a very comprehensive document covering:
- reasons for writing off the overpayment
- action taken by staff to recover the overpayment
- action taken to trace the debtor
a best practice user guide for overpayments, although this was written for the old benefits system and is now out of date..

2.533 We found that not all staff were aware of the guidance available and those that were did not always refer to it. Staff were unsure who had responsibility for updating procedural guidance.

2.534 Changes are communicated to staff via circulars. However, staff told us that there is insufficient time to ensure that they understand fully the impact of these circulars. Staff told us that this would often mean that determination officers or assessment staff were working to different ideas of what the HB and CTB Regulations mean.

We recommend that Lambeth review its current policy and procedures documentation to ensure that the various sources of documentation are brought together. There should be one up-to-date reference source covering:

all of the legal requirements placed on LAs when assessing and determining overpayments

any internal polices, such as at what level to set recovery, when to recover (which can have both legal and policy dimensions) and the checks to go through when assessing claims to HB and CTB to ensure that outstanding overpayments can be recovered from current benefit entitlement.

Identifying overpayments

2.535 Overpayments are identified when there has been a change in the claimant’s circumstances leading to a reduction or cessation in the claimant's entitlement to HB or CTB. The Academy system is used to calculate any resulting overpayment.

2.536 The current backlog of work is having a serious effect on Lambeth’s ability to process notified changes of circumstances on time. We sampled 60 overpayment cases and found that in 43 (72%) of them part or all of the overpayment could have been avoided.

2.537 In accordance with BV requirements Lambeth monitors the length of time taken to process changes of circumstances. Figure 2.77 shows Lambeth’s reported performance since April 2000.

Fig. 2.77: Length of time taken to process changes of circumstances
Period	Time taken (days)
April–June	86
July–September	89
October–December	74

Source: Lambeth

2.538 However, our sampling of 60 overpayment cases showed the average delay between the notification of the change to determination of the overpayment was 102 days.

Documenting overpayments

2.539 Capita is aware of the need to document its decisions when deciding whether an overpayment is recoverable, especially in cases where non-recoverable LA error overpayment has occurred (page 14 Working Draft, Housing Benefit overpayment write-off procedures for determination section, 5 November 1999). However, we think that Lambeth needs to improve its documentation on overpayments because:

there were 2 cases where there was no determination certificate on file. Staff on the determination team have also identified problems with this as:
- certificates are not always fully completed
- supporting documentation is not always in order, which leads to the determination officer having to sift through the case, which wastes time
due to the absence of all documentation we were unable to tell in 3 cases the date from which the change took effect
there is no place on the determination sheet to ask if the overpayment is recoverable. Therefore by default the overpayment is classed as recoverable unless the claimant appeals.

We recommend that:

Lambeth introduces the procedures and guidance for documenting its decisions and ensures that these are properly communicated to all staff making decisions on HB and CTB overpayments
to ensure consistency the updated policy and processes should contain guidance on when to attempt to recover a recoverable overpayment.

Classification

2.540 We sampled 60 overpayment cases to test whether:

the correct classification had been applied
the periods for each type of classification were correct
there was sufficient documentary evidence to support the classification.

Figure 2.78 illustrates our findings.

Fig. 2.78: Analysis of Lambeth’s classification of overpayments
Area examined	Number found	%
Classification correct	39	65
Classification incorrect	18	30
Number of cases unable to determine if correct	3	5
Total	60	100

Source: BFI analysis

2.541 There was a mixture of classification types usually spilt between Other, Fraud and LA error. We were unable to tell whether the correct classifications had been applied to 3 cases due an absence of documentation on the file relating to the change of circumstances. This reinforces our findings that Lambeth needs to ensure that its files are properly maintained.

2.542 Of the 18 cases where Lambeth had applied all or part of an incorrect classification we found:

14 had been due to Lambeth assigning a longer period of claimant error A longer period of LA error should have been applied
one case where DSS error, rather than LA error, would have been appropriate
3 overpayments classified as fraud. However, based on the evidence on the file there was insufficient proof to show that the claimant had intended to obtain HB or CTB to which they were not entitled. DSS Circular A57/98 provides guidance to LAs on the test to apply in these cases.

2.543 These incorrect classifications were due in part to bad practice and staff’s lack of understanding. This has been allowed to go unchecked due to:

inadequate control and checking of procedures in use by CAFT
the absence of system of management checks to identify weaknesses in staff’s performance
a lack of training for staff.

We recommend that Lambeth thoroughly reviews the circumstances of each overpayment and ensures that the correct classification and periods of the overpayment are applied.

Recoverability and whether to recover

2.544 HB regulations 91(3) and 99 and CTB regulation 84 set out the conditions for deciding whether or not an overpayment or excess benefit is recoverable. Nearly all overpayments are recoverable, even those deemed official error, if the LA can show that the claimant was aware that they were receiving benefit to which they were not entitled. In these cases the burden of proof rests with the LA.

2.545 There is no place on Lambeth’s determination certificate to ask if the overpayment is recoverable and more importantly to ask why it is recoverable. Therefore by default the overpayment is classed as recoverable unless the claimant appeals. This breach of regulations is addressed later in the report.

2.546 While we were on site Lambeth was drafting a form to help staff making determinations on HB and CTB claims decide whether an overpayment is recoverable. We welcome this.

2.547 Even though an overpayment has been deemed recoverable an LA can exercise discretion in deciding whether to actually attempt recovery. However, we found that that Lambeth does not do this and will attempt to recover all recoverable overpayments. Our sampling had 8 cases where we consider Lambeth could have exercised its discretion and not recovered the overpayment.

We recommend that Lambeth:

ensures that its staff understand:
- HB regulations 91(3) and 99 and CTB regulation 84
- that not all recoverable overpayments have to be recovered
provides guidance on how to document and record decisions taken on the recoverability of overpayments.

Deciding who to recover from

2.548 HB regulation 101 and CTB regulation 86 require LAs to decide from whom to recover overpaid benefit.

2.549 At the time of our on-site inspection Lambeth provided only limited guidance to help its staff to decide from whom to recover overpaid HB and excess CTB. However, staff told us that they would generally recover from the person to whom payment was made. We consider that Lambeth should address this and provide staff with guidance to direct them as to the best person to recover from.

2.550 In our sampling Lambeth had determined all 60 cases as recoverable. We found:

5 cases where the claimant had received payment
23 cases where the landlord had received payment
32 cases where payment had been made to the HD rent account. Lambeth treated these as if payment had been made to the claimant.

2.551 The analysis of the sample of persons from whom Lambeth decided to recover the overpayment is shown in Figure 2.79.

Fig. 2.79: Persons from whom overpayment was recovered
Person from whom recovery is sought	Number	%
Claimant	40	66
Landlord	20	34
Total	60	100

Source: BFI analysis

2.552 Our sampling indicates that in general Lambeth decides to recover overpayments from the party to whom benefit was paid.

We recommend that Lambeth provides guidance for staff to help them decide the best person to recover overpaid HB from.

Training for staff

2.553 There are a number of benefits to ensuring that staff are properly trained, not least of which is minimising the potential losses in money that Lambeth faces. We consider that Lambeth could reduce and recover overpayments better by improving its training arrangements for staff.

2.554 A training needs analysis has been carried out, but we found that:

there is no structured approach to training and development
staff do not have Personal Development Plans (PDPs)
management checking is limited, so staff training needs can not be identified. The results of what little management checking there is are not fed into a wider training plan
there is no one person with responsibility for communicating changes to policy or procedures. This means that Lambeth’s determination officers and Capita are working to different standards
anticipated training needs are not addressed. For example, we found that determination staff were unsure how to access and view data that had been archived after the conversion from the previous computer system to Academy. This was of particular concern as the conversion took place in January 2000.

2.555 Our sampling of overpayment cases highlighted staff’s lack of understanding as to how the overpayment process should be managed effectively. We found that staff had, for example:

applied incorrect periods to overpayment cases
applied incorrect classifications to overpayments. For example, fraudulent classifications had been given to cases where the conditions for recording such an overpayment had not been met. This would have been less likely to happen if staff had received appropriate training
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX X
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XX
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXX

We recommend that as a matter of urgency Lambeth develops a training plan to:

assess gaps in staff’s knowledge of the HB and CTB Regulations relating to overpayments
make sure that staff are aware of the need to check the various computer systems for previous debts and take the necessary action
ensure that staff clearly understand Lambeth’s internal policies on overpayments and the action to take.

Overpayment recovery

2.556 BFI’s Good Practice Guide details the benefits accruing from effectively managing the recovery process.

Legal basis of recovery

2.557 For an LA to recover overpaid HB and CTB there must have been:

a determination in accordance with the HB and CTB Regulations, by an officer of the LA, as to the status of the overpayment (recoverability and so on)
a notice of determination that complies with Schedule 6 to the HB and CTB Regulations.

Legality of determination

2.558 It is a pre-requisite of recovery that determinations of HB and CTB should be made by Lambeth. We highlight throughout the report Lambeth’s failure to make legal determinations on HB and CTB claims.

2.559 We found 2 cases in our sampling where there was no determination certificate on file. This indicates that:

Lambeth is failing to make even a retrospective check on the correctness of the determination and recoverability of HB and CTB overpayments
there are inadequate procedures in place to prevent benefit payments being made without first referring them to the determination team. This raises all kinds of issues, not least of which is the risk of internal and external fraud.

Older notices of determination

2.560 Lambeth’s previous notices of determination did not comply with Schedule 6 XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XX
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX X
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XX
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX X
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX
XXXX XXXX XXXX XXXX XXXX

2.561 Lambeth will have to make a judgement on how to pursue these cases, particularly those where recovery is not already taking place. A way around these constraints would be to issue a valid notice of determination.

We recommend that Lambeth carries out an assessment of the suitability for recovery of all outstanding debts, where no recovery is currently taking place. Where recovery is likely to be possible Lambeth should reissue a notice of determination compliant with Schedule 6 and begin recovery action.

This should form part of a rolling programme that will eventually assess the suitability of all debt for recovery and also guarantee that the legal basis for the recovery of these debts is sound.

Beginning recovery action

2.562 HB regulation 79(2) and CTB regulation 69(2) set out the claimant’s rights of appeal against a determination. Any appeal must be made within 6 weeks, in writing and signed by the claimant.

2.563 XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX X
XXXX XXXX XXXX

2.564 There were 2 cases in our sample where the claimant had appealed against the decision to recover the overpayment. These cases are currently with the appeals team.

Rates of recovery

2.565 In September 2000 DSS made changes to the HB and CTB Regulations affecting rates at which to set recovery. Whilst we were on site, the Academy system did not have a default level at which to set recovery. This has situation has subsequently changed. Lambeth has a recovery rate policy which gives guidance on the rate at which recovery should be set. However, the policy was drafted in 1997 and is due for review.

2.566 The policy directs staff to:

Examine each case on its merits and finding that amounts proposed are reasonable and that no hardship will result the following recovery rates are recommended (sic)

We found that staff were unaware of this policy’s existence and as a result they were not complying with it.

2.567 Following DSS changes to rates for recovery, the rates to set recovery on overpayments raised after 2 October 2000 are prescriptive for claimants in receipt of IS/JSA(IB). However, Lambeth needs to ensure that its policy addresses the circumstances of those claimants not on IS/JSA(IB).

We recommend that Lambeth:

ensures that staff are made aware of the changes to rates of recovery explained in DSS HB/CTB Circular A42/2000
develops a policy for setting rates of recovery
develops management checks to test compliance with the policy.

Recovery team liaison with other sections of the authority

2.568 In its publication Countering HB Fraud – A Management Handbook the AC recommends that overpayment recovery staff should try to co-ordinate their action with colleagues. It found that some authorities had formal debt policies that include criteria for staff to follow when doing this.

2.569 Lambeth told us that there is little formal liaison between the different sections pursuing day-to-day debt recovery. Because of this debts could be pursued in isolation or pursued together, which will mean that claimants could default on one debt whilst paying off another within the authority. The recovery team can access rent accounts to check whether they are in credit. If they are, they can be debited to pay off any existing overpayments of HB.

We recommend that Lambeth formalises liaison arrangements by ensuring that:

the recovery team explores the possibility of formal links with other sections of the LA pursuing debt so that all debts can be pursued as one
suitable arrangements are put in place to ensure that relevant data protection legislation is not breached
there are formalised procedures for the exchange of information.

Overpayment cases with an overpayment history

2.570 Of the 60 cases we sampled we found that 22 (37%) of them had previous overpayments, with a combined value of £27,400. There is currently £14,503 (53%) of this amount outstanding. In each of the cases where there was no recovery action the debtor was no longer in receipt of HB.

Methods of recovery and recovery performance

2.571 Lambeth uses the following methods to recover overpaid HB:

from ongoing RA and RR
from arrears of RA and RR
from RA paid direct to landlords
by invoice and cash instalments if there is little likelihood that that HB will remain in payment
by debiting the rent account when the account is in credit.

2.572 CTB overpayments are added to the Council Tax account or, where a CT account no longer exists, are recovered as any other sundry debt.

2.573 Lambeth told us that it considers Capita’s recovery efforts to be ineffective. However, responsibility for this must rest with Lambeth, for failing to manage the contract effectively.

2.574 Figure 2.80 shows the total amounts of HB that Lambeth has recovered.

Fig. 2.80: Recovery of overpaid HB – 1997/98 to 1999/2000 (£ million)

Source: BFI questionnaire and Lambeth monthly statistics pack September 2000

2.575 Figure 2.80 gives a false impression of Lambeth’s recovery performance since November 1999. Lambeth includes bulk adjustments as recovered. These are recalculations where HB is reassessed and a new lower rate of HB is awarded. This is then offset against the previous award of HB. The old award is shown as fully recovered, when in fact the claimant was entitled to some of the original HB awarded.

2.576 Lambeth’s ability to monitor Capita’s recovery performance is hampered by the change to Academy. On the old system these "bulk adjustments" could be separated out. However, they are now included in the overall amounts of HB recovered, giving a distorted impression of performance.

2.577 Capita’s recovery team has a significant backlog of work. There are currently 36,224 HB debts owed to Lambeth. Figure 2.81 shows the status of each of these.

Fig. 2.81: Status of debts in Lambeth on 25 October 2000
Status of debt	Number
Recovery by cash or other	43
Recovery through ongoing HB	2,239
No recovery	33,942
Total	36,224

Source: Lambeth

2.578 Figure 2.81 shows the very poor performance of Lambeth’s recovery measures. This is compounded by Lambeth not making full use of the different methods of recovery available to it. Our own sampling had only 7 cases (12%) where the debt was recovered in full. All of these had been through ongoing HB.

2.579 The longer the delay between each action, the less likely the debt is to be recovered. We found that there are significant delays at all stages of the recovery process, particularly when HB is no longer in payment. We carried out our sampling in September 2000 covering cases from April 1999 to March 2000. Of the cases where no steps had been taken to recover the debt the average length of time since the previous action was 235 days.

2.580 Lambeth does not have any procedures for refunding cash where the claimant has repaid more than the debt outstanding. This means that the credit stays on the system.

We recommend that Lambeth:

investigates ways of ensuring that "bulk" recoveries are accounted for separately to ensure that managers have a true picture of the recovery status of debts
ensures that debts are regularly monitored and action is taken on cases in a timely manner
implements a refund procedure.

Recovery from ongoing or arrears of benefit

2.581 We found that this is how Lambeth recovers most of its overpaid HB, as shown in Figure 2.82.

Fig. 2.82: Recovery of overpaid HB through ongoing benefit – April 1999 to August 2000 (£ million)

Source: Lambeth, client monthly pack September 2000

Note: December 1999 is not included as figures were unavailable due to the conversion to Academy benefits

2.582 Lambeth seemed to have a steadily improving performance after the introduction of the new benefits system. However, this is explained by the inclusion of "bulk" recoveries or adjustments in the amounts recovered. Using our own sample of 60 cases we tested Lambeth’s performance. We found 22 cases where the debtor was currently in receipt of HB. Figure 2.83 shows our findings.

Fig. 2.83: Recovery of overpayments through ongoing HB
Year	Number	%
Ongoing recovery from HB	11	44
Recovered in full	7	28
Claim pending	3	12
Appeal pending	2	8
In receipt of HB but not being recovered	2	8
Total	25	100

Source: BFI analysis

2.583 In 2 cases (8%) where the claimant was on HB, no recovery action had started. We consider that there are a number of reasons for this:

not all staff check for previous debts and begin recovery once the claimant is back on benefit
the recovery team has a backlog and cannot ensure that recovery action is started.

2.584 When a debtor is no longer entitled to HB the benefits system will produce a report to prompt the recovery team to set up an agreement to recover by cash instalments. However, a lack of resources and the current problems being experienced on recovery have made this difficult.

2.585 These problems are further compounded by a lack of training and management checks on the work carried out by staff. We have made recommendations elsewhere in the report that should address these issues.

Recovery from landlords

2.586 When landlords are paid direct they have to sign a declaration that they will notify Lambeth of any changes in their tenants’ circumstances. They also agree to pay any overpayments of HB.

2.587 In our sampling we found that 13 (22%) of the overpayments were due to the claimant having moved house. Of these HB was paid direct to the landlord in 9 instances. Clearly this raises a question over landlord liaison, as landlords are failing to inform Lambeth when their tenant moves on. Lambeth could help to deter landlords from doing this by ensuring that such overpayments in particular are recovered.

2.588 Since 3 November 1997, LAs have had the power under the provisions of section 75(5)(b) of the SSAA 1992, inserted by the Social Security Administration (Fraud) Act 1997, to recover overpayments of HB from benefit payments made to a landlord even when the debt is not in respect of one of their current tenants. However, Lambeth does not make use of this provision. We found 23 cases in our sample that could have been suitable for recovery action in this manner. In 9 of these, invoices had been issued, but no actual recovery of the debt had taken place.

2.589 Lambeth told us that it had considered using this method of recovery, but is under the impression that the Academy system will not allow it do so. We understand that this facility was available in Academy version 17. The use of this additional power of recovery is important, as Lambeth is owed £11.8 million from landlords, which represents over 30% of the total debt outstanding.

We recommend that Lambeth maximises recovery from landlords by:

reviewing existing debts owed by landlords
assessing their suitability for recovery under section 75(5)(b) of the SSAA 1992.

Cash recoveries

2.590 Lambeth shows good practice by allowing 4 weeks before beginning recovery action, giving the claimant time to appeal against the determination or reapply for benefit. A reminder invoice is then issued if the claimant replies. If, after a further 28 days, the claimant has still not contacted the LA then cases should be referred to the legal section. However, this did not happen at the time of our on-site inspection.

2.591 If the claimant contacts Lambeth they can arrange to make repayments either in full or by "cash" instalments. Debtors can make payment at OMH in cash, which are then credited on Lambeth’s debtors system.

2.592 Lambeth’s recovery performance in this area has been inconsistent, as shown in Figure 2.84. This is mainly because Capita made the decision not to issue invoices until 18 August 2000. This accounts for the drop in performance as new debts were not set up for recovery. To ensure that there were no problems with invoices issued after this date Lambeth performs reconciliation between details on the system and the amounts specified on invoices.

Fig. 2.84: Recovery of overpaid HB through cash payments – April 1999 to August 2000 (£ million)

Source: Lambeth, client monthly pack September 2000

Note: December 1999 is not included, as figures were unavailable due to the conversion to Academy benefits

2.593 In the period August 2000 to October 2000 Lambeth issued 420 invoices to debtors, but took no follow-up action. Lambeth should follow up all invoices and set up agreements with the debtor. Payments can be monitored using the Academy system, but we found this facility was not in use at Lambeth.

2.594 The current amount of debt against which there is no recovery, nor an agreement to recover, is £25.6 million. Unless action is taken quickly to recover this money Lambeth will find that the majority of the £25.6 million will be irrecoverable because claimants will be difficult to find. Lambeth may be left with a substantial amount to write off. We carried out our own testing of Lambeth’s performance. We found that Lambeth does not perform well when the claimant is no longer in receipt of HB, as shown in Figure 2.85.

Fig. 2.85: Recovery of overpayments after issuing an invoice
Year	Number	%
Cash recovery	4	12
No recovery action taken	6	17
Invoice issued no recovery	25	71
Total	35	100

Source: Lambeth

2.595 However, the poor levels of recovery through instalments cannot be blamed purely on the conversion to Academy. We found cases up to 18 months old where there had been no follow-up action. We found that:

assessments are not carried out to see if other means can be used to recover debt
controls to prompt staff to maintain debtor accounts are not effectively monitored. This would be identified by a management checking regime
staff do not have targets that would prompt them to monitor and manage their debtor accounts.

We recommend that Lambeth:

urgently analyses its current debt and assesses its suitability for recovery through other means such as:
- recovery from other social security benefits
- recovery from landlords under section 75(5) of the SSAA 1992
- registration of debts at county court
ensures that debtor accounts are effectively monitored and chased where the claimant has failed to repay overpayments.

Using debt collection agencies

2.596 Many LAs we have inspected refer debts to a collection agency when their own efforts have failed to recover debt. Agencies either accept a flat payment for each tracing or keep a proportion of any monies. Lambeth does not make use of debt collection agencies.

We recommend that Lambeth explores the possibility of using debt collection agencies where all other means to trace and recover debt have failed.

Recovery from other DSS benefits

2.597 HB regulation 105 and CTB regulation 91 allow LAs to require BA to recover overpayments in certain prescribed circumstances. Many LAs fail to take advantage of the powers offered under these regulations and Lambeth is no exception. BA told us that it already had procedures in place to set up recovery for LAs in this way. However, we found that Lambeth does not use this facility. BA told us that there had been one or 2 cases in the past where Lambeth had applied to recover debt in this manner. However, there were no cases at the time of our on-site inspection.

We recommend that Lambeth sets up procedures to make use of the powers provided under HB regulation 105 and CTB regulation 91.

Registering debts

2.598 Where recovery is not being made through a prescribed benefit, section 75(7) of the SSAA 1992 permits LAs to register debts at the relevant county court, without the need to bring proceedings. This can enhance an LA’s ability to recover overpayments as it may use the court’s powers of enforcement by use of bailiffs or attachments to earnings.

2.599 Lambeth has failed to make use of these powers and has no procedures in place to allow it do so. We consider that these procedures should go in tandem with pursuing older debts through court. XXXX XXXX XXXX XXXX XXXX XXXX XXXX
XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX XXXX X In cases where all other means have failed we suggest that Lambeth selects some of the larger debts and, where necessary, issues a valid notice of determination.

We recommend that Lambeth puts in place procedures to allow it to make use of section 75(7) of the SSAA 1992.

Pursuing debts through court

2.600 In July 2000 Lambeth’s overpayments project team raised the issue that there were no procedures in place for taking debtors to court. There has been no progress made since then.

2.601 We consider that this failure to address identified problems, such as this, is one of the main reasons why the value of debts outstanding in Lambeth is virtually unmanageable.

We recommend that as a matter of urgency Lambeth develops its procedures for taking debtors to court making full use of all available legislative powers.

Write-off

2.602 Lambeth has a working draft write-off policy which has been in use since November 1999. Members have not approved the policy. This is a comprehensive document containing procedural guidance and sample signatures of those authorised to recommend writing off debts and to write off debts. However, these signatures are not regularly reviewed. This is of particular concern because of Lambeth’s high staff turnover.

2.603 All write-offs have to be authorised by the Director of Finance. There is some scope for Lambeth to streamline these procedures by having levels of debt that can be written off by less senior officers of the LA, although suitable checks and balances would need to be retained

2.604 Lambeth has carried out some aged debt analysis and we were pleased to see that when debts were recommended for write-off each case was considered on it individual merits.

2.605 Regular monthly statistics are produced of the amount of debt waiting to be written off and of the amounts written off. This is broken down by reason and value of the debts being written off. Figure 2.86 shows the amount of HB debt written off by Lambeth.

Fig. 2.86: Overpayments written off – April 1999 to August 2000 (£ million)

Source: Lambeth, client monthly pack September 2000

Note: December 1999 is not included, as figures were unavailable because of the conversion to the Academy system

2.606 The peaks and troughs in Lambeth’s writing off of debt are as a result of:

the overpayments project team concentrating on clearing some of the older debts
Lambeth’s policy of passing batches of write-offs to the Director of Finance who then has to find the time to authorise each case.

We recommend that:

the write-off policy is approved by Members and changed from being a working draft to an agreed policy document
the signatures of authorised signatories are reviewed to take account of changes in staff. This should be carried out on a monthly basis
procedures are streamlined so that smaller write-offs can be authorised by officers other than the Executive Director of Finance and Corporate Services
written off debts should form part of the management checking regime.

Target setting for overpayment work

2.607 The contract specification has targets for overpayment recovery, as shown in Figure 2.87.

Fig. 2.87: Recovery targets set in the Lambeth benefits contract – 1998 to 2001
Year ending	Target collection threshold by Capita (£ million)
31 March 1998	3.0
31 March 1999	4.5
31 March 2000	6.0
31 March 2001	9.0

Source: Lambeth Benefits Contract Financial documents part VII

2.608 Capita keeps 10% of any recovered HB over these threshold amounts.

2.609 These targets are not being achieved. However, we consider that these targets do not achieve the dual purpose of reducing the number of overpayments created and recovering those overpayments that were unavoidable. We consider that the targets are flawed because:

the thresholds at which the contractor can gain any incentive from recovery increase each year. This makes it increasingly difficult for Capita to remain motivated. We would have expected there to be a higher threshold in the earlier years of the contract when the amount of outstanding debt was high, but that this should reduce
no incentive payments are to be paid for overpayments collected that were created after the contract had started. Some debts will be beyond the control of Capita, such as claimant error or those due to fraud. The contract targets provide no incentive for these debts to be recouped
although the contract makes provision for liquidated damages, default notices were not sent and liquidated damages were not enforced.

2.610 Lambeth should link the recovery of overpayments to Capita’s performance in reducing the numbers of LA error overpayments, with suitable incentives. It should also consider penalties for:

the numbers of LA error overpayments caused by Capita
failure to achieve the target recovery levels.

2.611 Other types of targets that could be set in this area include:

target dates for taking first action and follow-up action on recovery
setting a target for the frequency of trawling old debts
a realistic target to be set for recovery against each type of overpayment raised (BFI and AC recommend a 75% recovery rate). Higher targets should be set for the recovery of fraudulent overpayments and administrative penalties
a maximum number of cases that can be rejected by a designated officer after referral from CAFT
a maximum allowable number of errors on overpayment classifications. This should drop as staff become more experienced
targets to be set against each method of recovery, for example cash recovery might be set at a lower rate than that where HB is still in payment.

We recommend that Lambeth reviews its current target-setting regime to ensure that it incorporates, as a minimum, those targets we have outlined in this report.

Overpayment management information

2.612 Management information is crucial to the effective management of overpayments. It enables managers to assess:

performance against targets
the cost effectiveness of the service
areas of weakness.

2.613 Lambeth uses Microsoft Excel and Academy to provide monthly and weekly management information.

2.614 The monthly information shown is:

the amount and value of each type of overpayment raised
recovery status through ongoing HB and cash payments
debts submitted for write-off
the amount of debt outstanding.

2.615 However, we consider Lambeth should regularly provide:

the number of cases where there is currently no action
the number of cases where an invoice has been issued but no repayment has been made
the number of management checks carried out on overpayment cases and the reasons for discrepancies reported, to help identify training needs of determination staff
the number of cases referred by the investigation section for classification as fraudulent but rejected by Designated Officers, with reasons
the value and number of overpayments generated per week including amount, classification and total numbers per classification type
recovery performance against each classification type and from whom the debt is being recovered
amount of debt outstanding for all client types
debts referred to BA
status reports on all debts passed to the legal section
all debts subject to insolvency
all debts registered as a county court order
debts where recovery action is suspended, including the date when the account was suspended
debts recommended for write-off.

We recommend that Lambeth enhances its current management information to report on all aspects of the overpayment process as detailed in this report.

Overpayment management checks

2.616 BFI’s Good Practice Guide outlines the importance of an effective management checking regime.

2.617 We found that Lambeth does not have an effective management checking regime. We consider that Lambeth is open to a number of risks including internal fraud as it does not have structured or documented management checks to monitor the effectiveness of its overpayment recovery process.

We recommend that Lambeth introduces a series of management checks at a minimum of 10%.

Council-wide debt forum

2.618 Lambeth has delegated much of the responsibility for management and collection of debt to its various department heads. Corporately Lambeth is owed nearly £300 million in outstanding debt (minutes of CWDF 20 August 1999). Lambeth found that it needed to co-ordinate and manage this problem. Lambeth recognises that traditionally:

…the sundry debtor systems and management of income matters across the council had been poorly organised and monitored"(Minutes of the CWDF, 10 November 1999).

2.619 To address this and criticism from EA over its management of debt, Lambeth decided that an interdepartmental overview was necessary to co-ordinate its management and recovery efforts.

2.620 The forum has Member support (the LA’s Financial Secretary sits on the forum) and is attended by all of the senior managers from each Department. Other LAs have adopted this approach to managing debt and have found it to be beneficial (see BFI’s Good Practice Guide case studies on Thanet District Council). We would support this initiative.

2.621 The CWDF’s draft mission statement is to ensure that:

debtor accounts are monitored and managed
all avenues of recovery are employed
written procedures are available and reviewed annually
monitoring and audit trail arrangements are in place
realistic targets are in place for all areas of the recovery process.

2.622 The forum meets monthly and regular reports are received on the types, numbers and amount of debt owed to Lambeth. We fully support the aims and objectives of this forum.

2.623 Our inspection highlights the urgent need for Lambeth to address the issues it has identified. We are concerned by Lambeth’s slow progress against these issues. We found that Lambeth has still to:

introduce an updated policy and procedures manual for staff
provide staff with adequate training to help them maximise the recovery of overpaid HB
adequately monitor debt recovery against realistic and achievable targets
set up procedures to allow debtors to be taken to court.

We recommend that the CWDF observes the progress of the management of HB recovery by monitoring:

monthly reports on progress against action plans
work in progress
the action plan and steps taken to address the issues identified by our inspection and recommendations.

Overpayments project team

2.624 The project team has agreed its terms of reference with the HCC. In summary these are to:

ascertain actual amount, number of cases and status of HB overpayments outstanding by comparing various systems
ascertain the recoverable/non-recoverable split
establish systems to ensure maximum collection of recoverable overpayments
produce write-off procedures for client and contractor for non-recoverable debts
recommend future handling when responsibility is returned to Capita.

2.625 The progress reports produced by the team have been circulated to the Executive Director of Finance and Corporate Services and other interested parties.

2.626 In the report dated February 1999, and in its follow-up report of July 2000, the team identified a number of key issues for the effective management and recovery of outstanding debt:

unknown amount of duplication between the various computer systems holding debt. The team was to establish the extent of debt on these systems and attempt to ensure that the systems reconciled
the setting of targets for the recovery and write-off of outstanding debt
the need to ensure that Capita’s training programme includes measures to minimise the amount of overpaid HB and maximise the recovery of debt. This is one of the areas we identified. As yet no real steps appear to have been taken to address this
the need to pursue recoverable debt, especially for HAs and landlords
the need to address the problems associated with the poor filing of documentation on cases.

2.627 The team has made good progress against some of these objectives:

write-off procedures are now in place
analysis has been carried out on some of the older debts, in particular 250 of the highest value debts. Some attempt has been made to recover these debts. Where this has proved unfeasible the debts have been written off.

2.628 This team’s work must be commended. It has identified a number of important issues that Lambeth must address if it is to deal with the current poor state in which it finds its current handling of overpayments.

Putting it right – recommendations
We recommend that Lambeth reviews the requirements of the fraud case management system. (Paragraph 2.406)
We recommend that Lambeth reinstates frequent and regular fraud awareness sessions for all of Capita’s benefits staff. (Paragraph 2.411)
We recommend that Lambeth improves the effectiveness of fraud publicity and the hotline by: ensuring that hotline posters display the correct telephone number use of a Freefone number displaying the hotline posters at OMH revising the format of the anonymous telephone call stencil to collect all relevant information providing a stock of report forms and fraud awareness training to the staff at the Coventry call centre monitoring the outcome of hotline referrals. (Paragraph 2.421)
We recommend that Lambeth obtains better quality referrals from the benefits section by: increasing the number of fraud awareness seminars for assessment and counter staff simplifying the procedures for benefits staff to make a referral requiring less information considering permanently locating CAFT staff in the benefits section. (Paragraph 2.439)
We recommend that: all referrals are systematically sifted and prioritised by using risk analysis techniques formal feedback of outcomes is given to individual referrers in the benefits section all referrals are logged on to the FACTS database fraud files are created for all accepted referrals all referrals are risk assessed to aid prioritisation rejected referrals are analysed to inform future fraud awareness training for benefits staff. (Paragraph 2.439)
We recommend that Lambeth: make better use of the Do Not Redirect mail scheme, by extending its coverage to include all outgoing benefits correspondence refer all cases where mail has been returned to CAFT for investigation monitor the number and value of all mail returned. (Paragraph 2.449)
We recommend that, to improve the speed of investigations, Lambeth ensures that: all cases suitable for investigation are referred to investigators as quickly as possible fraud managers review progress on open fraud cases at more frequent, regular intervals. (Paragraph 2.457)
We recommend that, to improve the quality of management information, fraud managers should check investigation officers’ notebooks at regular intervals. This information can be used to inform on the effectiveness of its fraud investigation teams. (Paragraph 2.461)
We recommend that, to ensure the validity of non residency or "gone away" WBS claims and to ensure that fraudulent overpayments are fully identified, Lambeth: ensures that all investigators are familiar with the no residency or "gone away" guidance contained in Circular HB/CTB F17/98 and the Income-Related Benefits (Subsidy to Authorities) Orders 1998 makes full enquiries to confirm that claimants have ceased to occupy properties as their home seeks to identify the date of change of circumstances wherever possible. (Paragraph 2.467)
We recommend that, to improve the quality of fraud investigation techniques: Lambeth provides refresher training on good practices to the fraud investigators fraud managers set quality standards and conduct regular checks to ensure standards are met. (Paragraph 2.469)
We recommend that Lambeth ensures that WBS claims are valid by: implementing a thorough management check of all WBS claims providing staff with appropriate training on the criteria required to claim a WBS. (Paragraph 2.474)
We recommend that, to improve the effectiveness of Lambeth’s fraud visiting programme: it is extended to include visits to RR cases it includes a full verification check on all aspects of the claim by the VO the outcomes of all visits are reported back to Benefits by CAFT resources are targeted towards high-risk cases VOs receive training on evidence gathering performance targets for VOs are linked to completed visits VOs liaise with BA. (Paragraph 2.488)
We recommend that, to improve the effectiveness of management checks on fraud work, CAFT: implements a thorough and regular checking and monitoring regime by GFMs on all fraud investigation cases ensures consistency in the frequency and extent of those cases selected for check by developing guidance for the GFMs introduces a programme of regular accompanied visits, including unannounced visits ensures that all cases under investigation are reviewed by GFMs at regular intervals to ensure that the necessary progress is being made requires that closure of all cases is authorised by GFMs. (Paragraph 2.494)
We recommend that Lambeth seeks to identify at an earlier stage those cases that may be suitable for a prosecution or other sanction develops procedures to administer cautions where appropriate. (Paragraph 2.505)
We recommend that: Lambeth develops a new counter fraud strategy making reference to the BFI’s Good Practice Guide, taking into account the changes within CAFT, the changes in reporting lines and the new decision-making structure the strategy: makes particular reference to HB/CTB fraud makes reference to the full range of sanctions open to Lambeth includes a definition of fraud and corruption. (Paragraph 2.506)
We recommend that: fraud awareness training for each organisation’s staff is carried out regularly further exchanges of personnel take place Lambeth and BFIS develop procedures to monitor the SLA Lambeth and BFIS develop a strategy for closer working BFIS formally overloads appropriate cases. (Paragraph 2.513)
We recommend that Lambeth, as a senior partner in counter fraud initiatives with external bodies, invites BA/BFIS and BASIS to take part, to widen the representation and effectiveness of these bodies. (Paragraph 2.516)
We recommend that all recovery action should sit with one team to prevent duplication of effort. (Paragraph 2.529)
We recommend that Lambeth reviews its current policy and procedures documentation to ensure that the various sources of documentation are brought together. There should be one up-to-date reference source covering: all of the legal requirements placed on LAs when assessing and determining overpayments any internal polices, such as at what level to set recovery, when to recover (which can have both legal and policy dimensions) and the checks to go through when assessing claims to HB and CTB to ensure that outstanding overpayments can be recovered from current benefit entitlement. (Paragraph 2.534)
We recommend that: Lambeth introduces the procedures and guidance for documenting its decisions and ensures that these are properly communicated to all staff making decisions on HB and CTB overpayments to ensure consistency the updated policy and processes should contain guidance on when to attempt to recover a recoverable overpayment. (Paragraph 2.539)
We recommend that Lambeth thoroughly reviews the circumstances of each overpayment and ensures that the correct classification and periods of the overpayment are applied. (Paragraph 2.543)
We recommend that Lambeth: ensures that its staff understand: HB regulations 91(3) and 99 and CTB regulation 84 that not all recoverable overpayments have to be recovered provides guidance on how to document and record decisions taken on the recoverability of overpayments. (Paragraph 2.547)
We recommend that Lambeth provides guidance for staff to help them decide the best person to recover overpaid HB from. (Paragraph 2.552)
We recommend that as a matter of urgency Lambeth develops a training plan to: assess gaps in staff’s knowledge of the HB and CTB Regulations relating to overpayments make sure that staff are aware of the need to check the various computer systems for previous debts and take the necessary action ensure that staff clearly understand Lambeth’s internal policies on overpayments and the action to take. (Paragraph 2.555)
We recommend that Lambeth carries out an assessment of the suitability for recovery of all outstanding debts, where no recovery is currently taking place. Where recovery is likely to be possible Lambeth should reissue a notice of determination compliant with Schedule 6 and begin recovery action. This should form part of a rolling programme that will eventually assess the suitability of all debt for recovery and also guarantee that the legal basis for the recovery of these debts is sound. (Paragraph 2.561)
We recommend that Lambeth: ensures that staff are made aware of the changes to rates of recovery explained in DSS HB/CTB Circular A42/2000 develops a policy for setting rates of recovery develops management checks to test compliance with the policy. (Paragraph 2.567)
We recommend that Lambeth formalises liaison arrangements by ensuring that: the recovery team explores the possibility of formal links with other sections of the LA pursuing debt so that all debts can be pursued as one suitable arrangements are put in place to ensure that relevant data protection legislation is not breached there are formalised procedures for the exchange of information. (Paragraph 2.569)
We recommend that Lambeth: investigates ways of ensuring that "bulk" recoveries are accounted for separately to ensure that managers have a true picture of the recovery status of debts ensures that debts are regularly monitored and action is taken on cases in a timely manner implements a refund procedure. (Paragraph 2.580)
We recommend that Lambeth maximises recovery from landlords by: reviewing existing debts owed by landlords assessing their suitability for recovery under section 75(5)(b) of the SSAA 1992. (Paragraph 2.589)
We recommend that Lambeth: urgently analyses its current debt and assesses its suitability for recovery through other means such as: recovery from other social security benefits recovery from landlords under section 75(5) of the SSAA 1992 registration of debts at county court ensures that debtor accounts are effectively monitored and chased where the claimant has failed to repay overpayments. (Paragraph 2.595)
We recommend that Lambeth explores the possibility of using debt collection agencies where all other means to trace and recover debt have failed. (Paragraph 2.596)
We recommend that Lambeth sets up procedures to make use of the powers provided under HB regulation 105 and CTB regulation 91. (Paragraph 2.597)
We recommend that Lambeth puts in place procedures to allow it to make use of section (75)(7) of the SSAA 1992. (Paragraph 2.599)
We recommend that as a matter of urgency Lambeth develops its procedures for taking debtors to court making full use of all available legislative powers. (Paragraph 2.601)
We recommend that: the write-off policy is approved by Members and changed from being a working draft to an agreed policy document the signatures of authorised signatories are reviewed to take account of changes in staff. This should be carried out on a monthly basis procedures are streamlined so that smaller write-offs can be authorised by officers other than the Executive Director of Finance and Corporate Services written off debts should form part of the management checking regime. (Paragraph 2.606)
We recommend that Lambeth reviews its current target-setting regime to ensure that it incorporates, as a minimum, those targets we have outlined in this report. (Paragraph 2.612)
We recommend that Lambeth enhances its current management information to report on all aspects of the overpayment process as detailed in this report. (Paragraph 2.611)
We recommend that Lambeth introduces a series of management checks at a minimum of 10%. (Paragraph 2.617)
We recommend that the CWDF observes the progress of the management of HB recovery by monitoring: monthly reports on progress against action plans work in progress the action plan and steps taken to address the issues identified by our inspection and recommendations. (Paragraph 2.623)

Previous
Next

Report

Findings (continued)

Putting it right

Definition

CAFT organisation

CAFT procedures

Case management

Fraud awareness

Fraud publicity and hotline

Fraud investigation training

Fraud referrals

Data matching referrals

Do Not Redirect

Quality of fraud cases

Delays in investigation

File construction

Investigative approach

Investigation techniques

Outcomes of investigations

Duplicate WBS claims

Adverse impact of failure in benefit administration on fraud investigations

Visiting Officers

Management checks

Sanctions

Administrative penalties

Cautions

We recommend that Lambeth

Whistleblowing policy

Working with BFIS

Working with other bodies

Overpayments

Managing overpayment recovery work

Management and recovery of overpayments

Policy and procedures

Identifying overpayments

Documenting overpayments

Classification

Recoverability and whether to recover

Deciding who to recover from

Training for staff

Overpayment recovery

Legal basis of recovery

Legality of determination

Older notices of determination

Beginning recovery action

Rates of recovery

Recovery team liaison with other sections of the authority

Overpayment cases with an overpayment history

Recovery from ongoing or arrears of benefit

Recovery from landlords

Cash recoveries

Using debt collection agencies

Recovery from other DSS benefits

Registering debts

Pursuing debts through court

Write-off

Target setting for overpayment work

Overpayment management information

Overpayment management checks

Council-wide debt forum

Overpayments project team