The London borough has reported advances in fighting subletting fraud and managing the electoral roll through use of an IBM data system
The London Borough of Camden has made significant savings over the past two years by using an IBM data system to identify cases of subletting fraud and improve its management of the electoral roll.
Hilary Simpson, Camden’s IT business partner, said these have so far been the stand-out elements of a programme launched two years ago to bring together its data within a single source in the form of its Camden Residents Index (CRI).
It has used the IBM Big Data and Analytics solution to establish the database to facilitate the matching of data about its residents. It can now obtain a single view of its customers and collect relevant data once to use many times.
Recent council figures show that its use of the CRI had identified 752 of its properties that could have been illegally sublet, and the fraud manager in the housing department has been confident that this will result in a quarter of the properties being reclaimed. This is already providing savings, estimated at an average of £18,000 for each property, which are expected to total £3.4 million.
The electoral roll was streamlined using the CRI during 2014. Traditionally the council has matched the roll with data from the Department for Work and Pensions, usually providing a match rate in which nearly half the residents had to be contacted to confirm they should still be on the roll.
This year Camden achieved a match rate of 80% using the CRI, and was able to identify new residents who had not registered to vote. This provided savings of at least £25,000 compared with manual checking.
The database has also been used to automate the checking of personal information held within the accessible transport system, replacing the biannual manual checking system. This has reduced the need for external credit reference checks for the issuing of Freedom Pass travel cards, making it possible to check people’s eligibility automatically as they apply for the passes.
The council was able to immediately verify the eligibility of nearly 12,000 of the 23,000 service users, and learned that 166 had died, preventing it from sending a renewal letter.
The system has also been used to identify cases of school admission fraud, through which parents pretend to live in the borough to get their children into a specific school. Camden has already withdrawn five places from fraudulent applicants.
In addition, it expects to save money that it has previously spent with credit reference agencies in its efforts to reduce fraud. It said its experience so far suggests it achieves higher matching rates with the CRI.
Speaking at a round table on the use of the technology, Simpson said the council is obtaining better quality information from the database than it has obtained from credit reference agencies.
She added that is also prioritising the use of the system to reduce fraud around single person discounts in council tax payments.
“We have a cross-cutting agenda of matching data from different areas, and needed a single view of the customer across Camden,” she said.
“There was no single reference number, hundreds of lines of business and very disparate business areas. The issue is about matching these across the council’s activities.”
Camden set up the system with IBM business partners Scisys and Entity, completing the implementation in just three months. It is based on the IBM InfoSphere Platform, which provides building blocks for data integration, warehousing, master data management, big data and information governance.
Simpson said the council set up a team of data stewards to deal with the information governance issues and provide the human intervention for validation when necessary. They ascribed quality rankings to data from different sources – for example, ranking that from the electoral register highly but giving less weight to that from library records – in order to resolve any mismatches, and used it to create a “golden record”.
This has replaced the data from the council’s customer relationship management system, and Simpson said the updating is a “constant work in progress”.
“Actions had to be taken system by system with a lot of small steps,” Simpson said. “We set up a data quality policy on capturing information – for example, saying that we need a full first name as an initial is not good enough – and this has enabled us to do the matching.
“We worked on this mainly with the customer relationship management teams as they are usually the first line of enquiry.”
While most of the early uses will be directed at reducing fraud, Simpson said that Camden is planning to use the CRI for more predictive analytics in areas such as social care.
One possibility, as part of the government’s Troubled Families Programme, is to identify families with inter-generational problems around issues such as substance abuse and domestic violence. Simpson also referred to the possibility of measuring the link between overcrowding and social attainment.
“At the moment we’ve had it two years but it feels like a 10-year programme,” she said. “We’re finding more and more uses for the system and more areas of savings.”
Camden councillor Theo Blackwell said: “The Residents Index has proven its worth, helping the council to become more efficient, and now contains more than 1 million relevant records.”