At Camden Council, and now at Nesta, lots of people have asked me to explain why Camden created its consolidated residents index - a system that can join together records about individuals from the council’s numerous different IT systems - and what the outcome was.
Many think that the information sharing blockages are too great for them to create their own versions. Yet a number of local authorities have forged ahead, and there are compelling reasons for others to follow suit. Here’s why.
Most data concious local authority leaders will be familiar with using a postcode or Addressbase to cross reference their place data - records about the locations of buildings, parks, bus stops and so on. Some may have a Geographic Information Team to visualise their place data on maps. They may have gained new insights into service delivery by overlaying deprivation levels, pollution, schools, nurseries, parks, anti-social behaviour, lampposts, broadband coverage, etc.
But what about the people who live in those places? Insights about people and the services they consume are almost always absent. It’s not well known that a UK local council will have more than 500 siloed services and more than 200 business applications to manage those services.
More than half the budget of a typical council is spent on supporting people services (such as adults and social care), and this proportion is rising.
There are a handful of councils that have cross referenced their people records between systems, in real time, using modern software tools only recently available. As we shall see, once they have marshalled their own key data across silos, it means councils can not only get a better view of the residents they exist to serve, but also respond more quickly to opportunities to link up their data with other organisations to work collaboratively.
Where residents give different addresses to different services within the council, a residents index can quickly identify areas of fraud, such as when individuals game their address details to get their child into a particular school, or subletting a council property. Repossessing a sublet property saves £168k* per property in the first year. In Camden, 7 of these were identified using the residents index, representing a total saving of £1,176k.
Sharing dates of death across internal council services is another way to keep records up to date and to save money. 150+ concessionary travel passes were cancelled saving £2k per pass per year, a total of £300k. The entitlement for the passes has now been automated using the residents index so that 85 per cent are reissued without any administrative overhead.
£50k per year has been saved annually by using internal data sets to check residency rather than using an agency such as Experian. When councils had to re-build their electoral registers - because of new individual registration rules - Camden was able to find voters and validate their eligibility to vote using internal data. This increased the rate of automatically authenticated voters from 40 per cent to 80 per cent . It meant that first time voters were not disenfranchised.
In short, people data matching tools pay for themselves, and perform functions that cannot be carried out manually.
While some local authorities have done one-off linking exercises using databases or spreadsheets, that approach is not sustainable and cannot deal with dirty or incomplete data.
From the London Office of Data Analytics pilot, a key finding has been that councils with linked data can respond faster to requests for data for regional and city data analytics initiatives. They can also provide anonymised data, such as the number of separate families living at a particular address. This means that more can be done across organisations without having to share any personal data.
Looking outwards, a residents index can mean that more real-time open data can be provided by councils without administrative overheads. It also becomes more obvious where high impact / high value open data sets from central Government would help to save money in local government. For example, regularly updated bulk open data from the General Register Office on dates of death would save £millions across the UK by stopping services promptly so relatives only need to inform government once.
In response to all the requests I have received to provide more details about creating a residents index, I will shortly be working on a 'how to' guide for implementing such a solution quickly, simply and cost effectively. It will cover how to: calculate the business case and return on investment; carefully and securely navigate information sharing and privacy issues; limit the data you hold to the minimum viable set; engage with legal services, information owners and data analysts to release the data.
Have I missed anything you’d find useful in your organisation? Please comment below or find me on Twitter.
*The savings figures are calculated using benchmark figures from the Audit Commission publication ‘Protecting the Public Purse’. The national average of putting a family in temporary accommodation for a year is £18,000. The national average cost of building a one-bed property is £150,000. This totals £168,000 per property in the first year.