Fujitsu helps Gravesham Borough Council tackle social housing fraud

Council able to repossess six properties thanks to analytics software

A key issue for the UK's local authorities is housing tenancy, benefit and council tax fraud - a problem that costs them in excess of £1.3bn a year, according to the National Fraud Authority.

In order to help tackle these problems Gravesham Borough Council agreed a deal with Fujitsu to pilot its social housing fraud and error technology, a tool that is already used by HM Revenue & Customs (HMRC).

As a result of using Fujitsu's Social Housing Analytics service technology, the council was able to help its investigative team to identify over 75 properties where the council made a range of interventions from repossession of properties to re-housing.

"A couple of years ago, the government put a focus on councils to tackle tenancy fraud in regards to subletting false applications and ‘key selling'," James Flannery, investigations manager at the council told Computing.

"One way of tackling fraud tenancy is by conducting data matching to ensure that the data we hold with respect to our tenants is correct. Fujitsu came in seeking a partner to help pilot some of their data analytical software and we put our name forward to assist and conducted this at the back end of 2011," he explained.

Although Fujitsu could not disclose detailed information of the technology because of the sensitive nature of data and as the project is still ongoing, its lead practitioner for local government Ian Hall explained in brief how the technology works.

"The council provides us with a list of people claiming a particular housing benefit or exemption, we then run this data through a series of analytics models to provide them with a list prioritised by the level of risk they represent - the council then checks this with a view to preventing fraud or errors and recovering any money owed as a result," Hall told Computing.

Flannery added: "With all data matching there is elements of false positives meaning that fraud isn't actually being committed, it's just a data error, either on our site or the data set that Fujitsu were matching against which in this case was credit reference data."

The council previously used credit referencing agencies to match the data, but this produced hundreds of results, of which many were false positives.

"Fujitsu were very easy to work with. They explained what they needed from our tenancy data and then provided instructions on what the data meant when it came back and how we should tackle the data. We had heard of alternative options for data matching and analytical companies, and they all offer different things but data quality was paramount. What interested me in Fujitsu was that the number of matches we were going to get would be very focused," Flannery explained.

Most importantly for the council, the technology allows the team to focus on high-risk cases.

"Using the technology, we were able to identify a specific amount of cases which required some high resource focus that involved people going door knocking and conducting tenancy audits on properties that we identified as discrepancies," said Flannery.

Council able to repossess six properties thanks to analytics software

This allowed the council to track down six cases in which property needed to be repossessed, four properties that were under occupied and 12 illegal tenancy successions.

"The six cases in which property needed to be repossessed yielded commission and had a value of around £18,000 per tenancy, so using those figures we identified about £108,000 worth of tenancy fraud," Flannery said.

But Flannery insisted that the main goal of the pilot was not to take action against people committing crimes but to recover properties for people who actually needed them. He used the example of one family that had a severely overcrowded property - something the council had known but had slipped off the radar - and thanks in part to the technology, the council was then able to accommodate them in a bigger property.

The project ran from December 2011 to August 2012, at no cost to the council. Flannery said that the council would be interested in taking up the service every two years to ensure that their housing data is up to date.

However, if it were to take up the data matching service at another point it would come at a cost.

"The cost of the service would depend on the number of cases reviewed and the type of benefit and service being looked at. As this service is dependent on a range of factors, each project would require a bespoke quotation," said Fujitsu's Hall.

The initial pilot, according to Flannery, amounted to about 120 hours of council employees' time, but he believes this was well spent.

"The costs to the council are insignificant compared to gains we received," he said.

However, Flannery conceded that it was hard to compare Fujitsu's tool to others in terms of the amount of data matches they would have received. He said that although the software streamlined the amount of cases to about 75 compared to credit reference agencies' hundreds - saving staff time – it could be the case that other tools provide similar results.