As announced in previous blogs, Nesta is working with the GLA and more than a dozen boroughs in London – and with local authorities, the Digital Catapult and Sunderland Software City in the North East – to pilot data analytics projects that address public service challenges.
This post provides a brief update on the latest developments.
What can you do with data?
For each pilot the first objective has been to identify a public service challenge for which there is: 1) a big problem to solve, 2) good data available, and 3) a strong likelihood of identifying actionable insights that can deliver measurable results within a few months.
To that end, on 21 June, 15 London boroughs came together for a workshop with Andrew Collinge’s GLA data team to explore six challenges suggested by the boroughs themselves. A summary of the challenges and the assessments made of them can be found in a report on the London DataStore blog.
The issue that was thought to have the most potential was identifying unlicensed HMOs – houses of multiple occupancy. (HMOs are properties rented out to at least three people who are not from one ‘household’ – e.g. a family – but who share facilities such as a bathroom and kitchen.) HMO licences place extra responsibilities on landlords to ensure that their properties are safe and suitable for their tenants. According to Local Authority Housing Statistics data returns, there are up to 10,000+ estimated HMOs in some London boroughs. The percentage of those that are licensed varies considerably, but in many boroughs it’s estimated to be less than 10%.
That in itself is a serious issue.
Inspections of unlicensed HMOs frequently reveal conditions with very poor standards of accommodation. Some are genuine health and fire hazards. Vulnerable tenants can be financially exploited. The current hit-rate of finding these properties is low. Over the next few weeks, we’ll be investigating whether data analytics can be brought to bear to improve it.
In the North East, meanwhile, discussions with the seven local authorities have identified a common interest in dealing with issues related to alcohol abuse. Many public sector organisations are affected, from ambulance trusts to the local police, and from social workers to local authorities. A workshop is planned for September to bring together staff from each of those organisations to explore the various aspects of that complex social issue. As in London, the objective will be to pinpoint a specific challenge that meets the three pilot criteria outlined above. A full write-up of the workshop will be available after the event.
Critical to the success of any data analytics project is the quality of the data scientists and the time they have to dedicate to it. Though there are many talented analysts working within the UK public sector, their skills are in such high demand that it simply isn’t feasible to ask them to divert all their time to a data pilot for several weeks.
To resolve this, in London the GLA is working with the ASI – a data science SME – to provide the expertise to support the London Office of Data Analytics pilot. In the North East, meanwhile, Nesta is partnering with the Digital Catapult (which has a North East hub based in Sunderland Software City) to assemble a data science team from local SMEs and universities.
In both areas, there are two key requirements of those teams above and beyond doing the data analysis.
The first is that they share their skills and methods with the data teams within each local authority and public sector body taking part. Building skills and capacity is a vital goal of this project so that it leads into something bigger and more long term.
The second is that the data models they create should be open sourced; freely available for other cities and regions to learn from and reuse. At the end of the pilot, a full toolkit based on all the lessons learned will be published.
Over the summer, the viability of tackling these two issues will be explored further. Each faces a number of obstacles. There are technical barriers (the use of different IT systems), data barriers (the different data standards used and datasets collected), and organisational barriers (the cultural inertia of working in collaboration rather than individually). None is straightforward to resolve.
But as local government and public sector bodies face ever growing pressure to meet rising demand with reduced budgets, it’s never been more important to find ways to do so. To that end, we hope these pilots will be a small step in the right direction.