The verification framework: early experiences of implementation
A hard copy of this report summary can be obtained by contacting Paul Noakes [E-Mail: Paul.Noakes@dwp.gsi.gov.uk] or by writing to him at the 'Social Research Division, Department for Work and Pensions, 4th Floor, Adelphi, 1-11 John Adam Street, London WC2N 6HT'.
Research Report No. 137
By Nick Pettigrew, Andrew Thomas, Sara Candy, Lucy Joyce and Tanya Saunders
This research is part of a wider evaluation and describes 14 local authorities’ experiences of the early implementation of the Verification Framework (VF). The main aims of the research were to explore local authorities’ behaviour and experiences in early stages of setting-up and operating the VF.
The research was based on in depth interviews with these 14 local authorities. Local authorities were selected to capture a range of experiences and types of authority. Initial fieldwork to explore the set-up of the VF was carried out in November 1999 – February 2000. Local authorities were then revisited in July – August 2000 to explore how the VF was operating.
The VF is designed to reduce the number of fraudulent claims made for benefit. Minimum standards are defined for collecting evidence when a Housing Benefit (HB)/Council Tax Benefit (CTB) claim is made, as well as for the frequency of subsequent reviews and visits. Operation of the VF is voluntary, with the Department of Social Security (DSS) providing funding for its set-up and on going costs to those Local Authorities (LAs) that wish to implement it.
The key findings include:
- Overall the underlying finding is that LAs remain committed to the VF and although there were early teething problems, many are dealt with by the time the VF has bedded down.
- After implementing the VF, LAs experienced a backlog of claims which were attributed to: the implementation of other initiatives such as Data Image Processing at the same time as VF; underestimating how compliant systems were; software problems; incomplete information from customers to authorise claims or renewals; and an influx of queries from confused customers;
- The first 12 months of operating the VF were generally felt to be the ‘bedding down’ period, although some LAs felt the VF had settled more quickly than others.
- LAs supported the principles and objectives of the VF. Authorities felt that there were a number of indicators to suggest the VF had impacted beneficially on fraud;
- Prior to the VF LAs had procedures to deal with benefit fraud. However, few LAs had written fraud or prosecution policies;
- LAs wanted a long-term guarantee of future funding and for the VF to be made compulsory for all authorities.
Set-up
- LAs struggled when determining costs for VF applications and felt that DSS gave them insufficient guidance on the costs;
- None of the 14 LAs met their initial compliance timetable because they underestimated the time needed to adapt buildings and redesign benefit claim forms, had problems recruiting staff and experienced software problems.
Operation
- During the second phase of research LAs generally felt most problems were behind them and felt positive about the VF although none of them felt they were back to their pre-VF processing times.
- Some LAs felt that DSS was partially responsible for some of the software problems and could have done more to brief software houses about changes that would be needed to support the VF.
- The production of Management Information from software remained one of the most problematic issues. LAs were concerned that this might have a negative impact on their funding and that they could not determine whether the VF was working;
- LAs felt initial implementation of VF had a negative impact on staffing levels, morale and working hours. Again, this seemed to be improving at the second stage of fieldwork;
- LAs’ experiences of communication with the DSS were mixed. Some reported that DSS was always very friendly, helpful and able to generate solutions very swiftly. However, the general opinion was that the DSS had been unhelpful as: they were often difficult to contact and took a long time to respond to written enquiries; they offered contradictory advice, or referred callers to 'the manual': and would not put oral advice in writing.
Summary
Introduction and research methodology
The study, carried out by “BMRB Qualitative,” was conducted in two phases: The first explored the LAs’ behaviours and experiences in setting up the VF and the second focused on their behaviours and experiences when operating it.
Fieldwork for the first phase was conducted during the period November 1999 to February 2000, and the fieldwork for the second phase between July and August 2000.
In total, 14 LAs participated in both phases of the study - ten in England, two in Scotland and two in Wales. LAs were selected to reflect a broad mix in terms of type of authority:
- Type of local authority - district, metropolitan, unitary;
- Catchment area - city, urban, semi-rural;
- Size of HB caseload, determined from statistics compiled by the DSS;
- Whether the processing of HB had been contracted-out;
- The amount of time that the LA had been running VF; and
- Whether the LA stated they were fully compliant.
A total of 96 LA staff participated in both phases of the research, comprising a mix of:
- HB managers;
- VF managers;
- Processing staff;
- Frontline staff;
- Reception staff;
- Visiting officers;
- Finance officers: and
- Audit officers.
Pre-VF
Apart from HB managers and other senior staff, processing and frontline staff tended to know very little about the levels of fraud that operated in their LA area. Most HB managers felt that their LA area had an average level of benefit fraud. There was some sense that the fraud detected reflected only a small proportion of that committed.
In common with other studies, LA staff identified the following as being the main types of benefit frauds committed:
- People claiming as single parents but living with a partner;
- Working whilst claiming;
- Undeclared capital or income;
- Undeclared changes of circumstances;
- Collusion between claimants and landlords; and
- Continued payment to landlords when the claimant has left the premises.
Organised fraud rings were felt to be prevalent in city areas; landlord fraud was felt to be particularly high in seasonal employment and bed-sit areas.
Prior to VF all the LAs had a set of procedures to deal with benefit fraud, which included acting on 'tip-offs', spot checks on documents, random visits to customers’ homes and acting on 'hunches' about claims that arrived in the office. However, few LAs had written fraud or prosecution policies. Fraud did not seem to be a particularly high priority issue for the LA as a whole, although this may not have been true for individual HB departments. Some LAs did not promote an anti-fraud stance for fear of negative publicity, others felt that the 'customer-orientated' focus of the previous years had meant that LAs had been fairly relaxed in their approach to benefit fraud. This was reinforced by the relative absence of fraud training within LAs prior to the introduction of the VF.
Applying for VF funding
Implementation of the VF is voluntary. LAs could apply for funding both to set-up and operate the VF. In 1998/1999 84 LAs received funding to implement the VF and at the time of report writing, over 250 LAs had received VF funding.
LAs had several reasons for applying for VF funding. First, it provided a new source of revenue for at least two years. Secondly, there was a view that the VF was likely to be compulsory in the future and therefore it was prudent to seek funding early while it was likely to be at its most generous. Thirdly, LAs believed that the implementation of VF would raise the profile of benefit fraud within the LA and add to the LAs overall status. Fourthly, some LAs saw the VF as a way of standardising benefit processing procedures and encouraging best practice.
The process by which applications were put together were often not accessible to the research team, either because this information was no longer available (as staff had moved on) or in the case of contracted-out companies was deemed commercially sensitive.
However, with hindsight it was clear that LAs made two major decisions which seriously hampered the introduction of the VF. First, in putting together their applications LAs often over-estimated how compliant their HB processing system was with the requirements of the VF. Second, LAs often combined the introduction of the VF with other systems such as new benefit processing software, document information processing systems (DIPS) and new ways of working. The impact of these decisions could be considerable, as the implementation of the DIP system impacted upon HB processing times and so further contributed to the backlog of claims.
The process of determining the VF applications was to considerably exercise the LAs, but on reflection, the LAs felt that they essentially 'pulled figures out of the air'. They took into account a wide variety of costs in their applications for VF funding, but realised once they had started the operation process that there were many costs which they had not included, primarily because they had underestimated the amount of work they would have to do to achieve compliance and partly due to the impact of introducing a number of systems simultaneously.
Perhaps the key feature to come out of the application process was that LAs felt that they were given insufficient guidance by the DSS as to the likely costs involved in setting up the VF. This was particularly pertinent as a number of LAs were aware that an early version of the VF had already been piloted yet they felt that there had been no learning from this early work as to the likely costs, or effects, of implementing the VF.
Setting-up and operating the VF
A number of issues were discussed in both phases of the research. These included:
- Training: Staff were trained through a range of external and internal training sessions, plus the provision of training materials, handouts and training summaries. Training was highly welcomed and overall felt to provide a good overview of the VF. However, there was some concern that training was not necessarily provided at the most opportune time and was not always practical. In the second phase of the research, some LAs had provided refresher training but others were yet to do so because of lack of resources, time, and the fact training was a low priority in comparison with clearing the backlogs. However, it was felt that refresher training was an important thing to be provided, as well as gaining feedback from staff.
- Publicity: Some LAs publicised the VF to the general public through newspaper advertisements and editorials; others made their landlords and RSLs aware that the VF was to be implemented. In some cases LAs offered landlords an opportunity to discuss the finer details of the VF, but with extremely little interest being expressed by landlords. In some LAs, where a publicity campaign had been orchestrated, this had given rise to unexpected consequences. However, amongst other LAs the prior publicity of VF had been found to work well. In the second phase, there was little evidence of further publicity and consultation, although some LAs did feel that further publicity was needed to get over or reinforce the requirements of VF.
- Staff recruitment: Additional staff were recruited, typically, processing staff, verification staff (in LAs which split the function of benefit processing and verification between different staff), DIPS processing staff and visiting officers. LAs often found it difficult to find staff that met their requirements. In the second phase, further changes had been made in some cases to the team structures and the staffing levels. Although LAs still felt understaffed they felt that the situation had improved since the first phase of research. As benefit administration has a long learning curve, the use of temporary staff proved problematic as some had left for other positions.
- Visiting Officers were more in evidence when the second phase of the fieldwork took place since previously, due to staff shortages, some had redirected to assessment or frontline. Some visiting officers had additional administrative staff to deal with the paperwork. Some visits were now being notified to save time, particularly those to the vulnerable. Visiting Officers did not feel their role was always completely distinct from fraud officers. In both phases of the research Visiting Officer software was problematic, impacting on the number of visits that could be made.
- Changes to LA offices: In many of the LAs a number of physical changes had been made to offices. Reception areas were enlarged and re-arranged to provide greater privacy for customers; and in most participating LAs the post room was made more secure. More secure courier systems were also introduced in order to move valuable original documents from reception areas to the DIPS and verification areas. These were mainly introduced during the first phase of research.
- Computer hardware: With an increasing headcount LAs also purchased additional computer hardware. The setting-up of the computer systems usually occurred without too many mishaps.
- Computer software was usually upgraded and additional VF screens were added to the benefit processing system. LAs reported a succession of teething problems with new software. In one instance a LA was without its computer system for six weeks. Most commonly, the new software was found to be unstable, VF screens were not always fully integrated with the processing software and the electronic visiting officer diaries regularly failed to work. There appeared to be relatively little software training. Although the situation was felt to be improving in the second phase, there was still some way to go before software was operating as originally envisaged. A perception still remained amongst some LAs that the DSS were partly responsible for at least some of the software problems. In these cases it was perceived that the software houses had not been fully briefed in connection with the specifications needed to operate VF.
- Paperwork: LAs had been surprised at the volume of paperwork that the VF had generated, as well as the time it had taken to agree any paperwork that was issued to the public. Typically, new HB application forms were designed, together with supporting documentation. These often took much longer to agree than had been expected. In addition, all documents that were required for the verification process had to be either scanned into a DIP system or photocopied and held on file. The latter produced large quantities of paper. LAs that had separated the processing and verification processes often required two sets of files to hold the relevant documents. During the second phase paperwork was either continually being tweaked or LAs were waiting for a specific amount of time before changing paperwork, with forms usually being made shorter. Some LAs were now acknowledging receipt of forms to prevent customers ringing up to check they had been received.
- DIP system: Eleven of the 14 LAs had a DIP system; in most cases this had been implemented at the same time as the VF. There were many complications with DIP systems. First, LAs had not expected the amount of time it took to scan and enter all supporting documents onto the system, and had the effect of considerably slowing down either the reception staff or the verification teams. Second, it often took some time to work out the best form of indexing the documents and to ensure that staff were consistent in their approach; in some cases LAs could not link all pieces of evidence to a case because of inconsistent indexing. DIP systems were often very unstable in the early months of use. In the second phase, DIPs appeared to be having less impact on the working of VF.
- Management Information (MI): The software was often felt to be unable to produce MI. However, there was also some evidence that MI was available; in some cases the LAs were not convinced it was accurate, in others the volume of material produced was so great that LAs were as yet unable to use it in a useful manner. Of all the issues surrounding VF in the second phase, MI remained one of the most problematic. Some LAs were not producing any of the MI requested by the DSS, whilst others were providing only a part of the requirement, with still software problems and a perception remaining that the DSS kept changing requests for information. Some managers were concerned about not being able to produce MI, as they felt that DSS might take funding away if MI was not supplied, and it led to an inability to prove whether or not VF was effective.
- Timetables: Despite the planning involved none of the LAs in the study met their initial operational timetables (these ranged from six weeks to nine months). Indeed, timetables were often moved a number of times, with the effect that the set-up and operation phases often became one and the same. Timetables lengthened for a variety of reasons: underestimating the time taken to make the adaptations to buildings; difficulties in recruiting staff; the time taken to re-design and approve the HB application form; and considerable teething problems with new software.
- Risk groups were generally felt to work well, although a number of changes were suggested in particular that the self employed and lone parents be put in a higher category. Some LAs also wanted the flexibility to alter their risk categories due to variations in local conditions.
- Postal checks: LAs were pleased that postal checks had been made voluntary since they were felt to be ‘a waste of time’ duplicating the role of visits.
- Evidence of Capital was viewed as the major feature of VF which identifies fraud through customers having to provide two months worth of bank statements. Some staff were in favour of increasing this although they recognised this could create further difficulties. Although collecting ID was seen to slow down the processing of a claim and be frustrating for clients, it was viewed important to collect as evidence. Obtaining evidence of benefits and residency and rent was generally felt to be unproblematic.
The impact of the VF
The initial reaction of LAs was that the VF had been difficult to implement. This reaction was primarily as a result of the backlogs of HB claims that built up after operating the VF. However, it is very clear that while the VF may have contributed to the backlogs, the key reasons for the difficulties that the LAs faced were due to a number of problems. These were:
- Underestimating how compliant their HB processing system was;
- Customers misunderstanding the advance publicity that some LAs published, which resulted in LAs being swamped on the day of VF implementation with unnecessary claims. Customers who did not go to the LA offices, jammed the switchboard with their enquiries;
- The implementation of DIPS, staff's lack of familiarity with DIPS and some initial teething problems had a major impact on processing times;
- Many HB applications received by post were incomplete and had to be responded to individually. Very large 'pending' files were soon in evidence; and
- Upgraded software was either troublesome, failed to work correctly, or failed to work at all.
As a result of these factors plus a lack of familiarity with the VF, large backlogs of unprocessed HB claims built up. Customers complained, either by letter, telephone or in person. The increase in enquiries resulted in a greater workload for the HB staff.
By the time of the second phase of fieldwork, it was clear the outlook was much more positive. Although none of the LAs felt they were back at a pre-VF situation there was general feeling that problems had been dealt with and the VF was being seen in a positive light. The first 12 months of operating VF were generally felt to be the ‘bedding down’ period, although some LAs felt they had bedded VF in more quickly than others. The key determinants of change were:
- Customers becoming familiar with what was required of them, and thus correspondence between customers and LAs was reducing;
- Telephone calls were no longer handled by processing staff in some LAs;
- Processing staff were becoming more familiar with their role;
- Backlogs were clearing, due to an increase in staff, familiarity, and overtime.
The impact on customers
Initially, existing customers ere shocked at the introduction of the VF. Elderly customers could be rather bemused that they were being asked for identification; some other customers were argumentative or abusive. The most accepting customers were 'first-time' claimants. In the second phase there was a general feeling that customers had a much better understanding of what was required of them, although there were some exceptions. In addition the ‘stock’ of claimants had been through the claim process at least once. Queues were reducing and customers were being dealt with more swiftly.
The impact on Staff
The implementation of VF - and all the subsidiary systems - had a major effect on HB staff. Staff resources were stretched, many staff worked very long hours and in some cases a seven-day week to clear backlogs of unprocessed and partially processed HB claims. In addition they had to deal with angry customers and complaints. Despite HB managers and VF team leaders being appreciative of the commitment from staff, morale nevertheless suffered. Levels of sickness increased and there were some resignations that were attributed to the 'VF experience'. In the second phase staff exhibited much lower levels of stress and generally higher morale, since:
- Backlogs were reducing;
- Staff were experiencing less anger and ‘hassle’ from customers;
- Reductions in complaints from customers and members;
- Easements had made the VF easier to use; and
- More familiarity with the VF itself.
However, some stresses and strains had had lasting consequences. For example in one LA teams maintained only the coolest of relationships with each other.
Impact on speed and accuracy
There was a general perception that processing backlogs would not return to pre-VF levels since claims require more processing time and because VF introduced an additional set of opportunities for customers not to comply. Once the VF had been in place for some months there was a sense however that accuracy had increase because of greater attention to detail and increased standardisation.
Impact on Fraud
In the first phase, participating LAs felt it was too soon to assess the overall impact of the VF on fraud. However in the second there were a number of indications that suggested that the VF would have an impact on the levels of fraud primarily in:
- Caseload level. In the first and second phase caseloads were dropping. Some were certain this was a VF effect and others were less sure and wanted another year before they made up their minds. However it was felt that VF had reduced some potentially fraudulent HB claim renewals.
- Fraud and error detection. The most commonly mentioned HB claim that seemed to be prevented was that of pensioners who had not declared capital or second pensions. However, given these had been detected in the first year it was felt that detection rates would decline in the future.
- Fraud prevention. It was felt that the future role of VF would primarily be preventative, stopping newcomers ‘trying it on’. However, it was felt that VF would do little to deter the hardened fraudster, and that Visiting Officers would be unable to prove sufficient cases of working while claiming or cohabiting.
- Fraud awareness. There was some evidence to suggest that VF had made staff more fraud aware.
Communicating with the DSS
Experiences of discussing the set-up and operation of the VF with the DSS were very mixed. Some LAs reported that the DSS were always very friendly, helpful and able to generate solutions very swiftly. However, the general opinion was that the DSS had been unhelpful. As evidence LAs indicated that the DSS:
- was often very difficult to contact;
- took a long time to respond to written enquiries;
- offered contradictory advice, or referred callers to 'the manual'; and
- would not put oral advice in writing.
The pre-Christmas 1999 seminars were welcomed but generally felt to have been offered too late to be of any help.
In the second phase of the research, views about communication continued to be mixed, with some thinking that the DSS had been proactive and others believing that the DSS had not listened to their opinion.
Conclusions
- The HB department anti-fraud culture had been reinforced;
Overall, the LAs in the study were fully supportive of the principle of the VF and, despite their implementation difficulties, remained committed to the VF.
The introduction of VF was felt to have had the following beneficial effects:
- Fraud training was now in greater evidence;
- Fraud was now higher profile within the LA generally;
- LAs were of the opinion that the implementation of the VF had slightly reduced the number of HB claims, however, it was difficult for LAs to determine if this was as a result of the implementation of VF or due to local economic reasons;
- Income evidence requirements had produced many instances of undeclared capital and undeclared income;
- Customers were beginning to better understand their responsibilities in terms of providing evidence in support of their claim.
While HB processing times had increased - significantly in some cases - there was evidence that these were now beginning to decrease. However, none of the LAs considered they were likely to return to the processing speeds they had attained prior to operating the VF.
Overall LAs varied in the degrees of success in setting up and operating VF and dealing with the problems. The following factors seemed to be critical in its success:
- Skilled management;
- Taking a realistic view of compliance;
- Adopting realistic timescales;
- Implementing new systems individually;
- Goodwill of managers and staff.
While easements were generally welcomed there was some concern that they had been introduced rather hastily.
Implementing the VF – an ‘ideal’ approach
The LAs that took part in this study suggested some issues for other authorities to consider when implementing the VF:
- Visit other sites to learn from their experiences of implementing the VF;
- Analyse, in detail, the nature of the existing system in place and the new approach to be introduced;
- Consider a phased approach to the introduction of the new system or initiative;
- Ensure that all new software systems are stable, reliable and tested “before” being installed;
- Only implement one new system at a time;
- Provide initial and refresher training;
- Brief their staff about the likely impact of introducing new systems;
- Use local advertising (carefully);
- Clear any initial backlogs;
- Provide a means of dealing with customer enquiries and problems that is separate from that of processing and verification.
Issues for Consideration
LAs participating in this study outlined some issues for the DSS to consider when implementing the VF. LAs suggested that a national advertising campaign run by the DSS would be useful for customers. They also felt that the Department could do a number of things to help the implementation of the VF run more smoothly, such as: an increased training budget and some centrally arranged training; DSS and software houses working in partnership; and DSS providing advice through site-visits, telephone helplines and written communication.
Although LAs were fully committed to the principle of VF they wanted a long-term guarantee of funding for the continued operation of VF. LAs also wanted VF to be made mandatory to provide a uniform approach to dealing with claims and provide more power of enforcement in the face of fraudulent claims. There was some feeling that after the first year of operation fraud detection rates would fall and the emphasis would move to fraud prevention. LAs felt that fraud detection rates could be maintained with the adoption of other approaches, chiefly the developments of links with other agencies, in addition to the VF.
Publication details
Nick Pettigrew, Andrew Thomas, Sara Candy, Lucy Joyce and Tanya Saunders (2001) “The Verification Framework: Early Experiences of Implementation ”(DSS Research Report No. 137) Leeds: CDS (£27.00)