Why the public sector needs data champions

Opening up the government's data sets will make public service commissioning more effective as well as helping service users, argues Andrew Weston of charity think tank New Philanthropy Capital

The rise of public services being commissioned out to the third sector means that many commissioners are finding themselves wondering what they are buying. 

How can they know that the outcomes or outputs promised by the providers are deliverable? How can they tell if an intervention, for example in a complex area like preventing re-offending, is successful? How best to compare different suppliers? 

Initiatives like the Justice Data Lab, created by New Philanthropy Capital in collaboration with the Ministry of Justice, can help to answer some of these questions.

Dame Julie Mellor: Why departments must take citizens' complaints seriously
Land Registry chief Graham Farrant on breaking down silos, 2017 & why surprises make the best Christmas presents
Autumn Statement and Budget announcements suffer from “evidence gap”, says charity

This portal allows organisations to submit information about reoffending rates following an intervention, and then compare the anonymised reoffending rates against matched comparison groups of offenders using government datasets. 

Once the analysis has been created, the personal data from the organisation is destroyed – ensuring that organisations feel more comfortable about sharing sensitive information. This model allows organisations which lack the resources to do a robust comparative assessment of their own interventions to access a breakdown of their wider impact.

"When data is made freely available for people to access, use, and share, it can stimulate innovation"

At the same time, it provides a common output which means interventions can easily be compared. On average, cases that went through the Justice Data Lab have shown approximately a 2% decrease in reoffending in comparison to the baselines. 

Interestingly, the data also shows cases where those using the services were more likely to reoffend as was the case for 21% of case studies analysed. This, too, allows organisations to change and adapt their services to better support their beneficiaries.

Building on the success of the Justice Data Lab, NPC is now looking at data lab projects in sectors such as health, education and employment — and we are focussing on areas where the data is particularly complex and sensitive and therefore where the model can be of most use. 

Our Freeing Up Health analysis paper recently outlined how a health data lab would help charities, policymakers, commissioners, researchers, and 50 charities have already expressed interest in being involved, with twelve willing to take part in the pilot.

Of course, a data lab-style model is not the only way of effectively sharing data. Another area worth examining open data. Another recent NPC paper, Valuing Data: How To Use It In Your Grant-making, explores effective data sharing for charitable funders, and finds that when data is made freely available for people to access, use, and share, it can stimulate innovation.

Such an approach also opens up the prospect of a data "critical mass", whereby sufficient data exists to fully understand the sector, allowing those who analyse it to accelerate innovation dramatically.

Talk of data always raises questions about confidentiality, as well as concerns about who owns the data. A recent report on data security by the National Data Guardian Dame Fiona Caldicott noted that, in the context of the NHS, the public broadly supports data sharing when the benefits of doing so are clearly explained. 

However, particular issues arise if people feel their data has not been anonymised, or if there is confusion around how it is going to be shared. This suggests a sufficiently robust and transparent open data system could be a practical possibility.

The ideal solution would be a mixture of the two with options for a data lab-style service for the most sensitive data, and to support organisations who lack the resources to analyse data themselves; and a wider open data system to encourage experimentation and innovation.

No matter what the model, none of this can be done without the datasets being unlocked. We need data champions in the public sector who realise the potential of effective data sharing and are willing to step forward and work with the third sector to create the most effective and workable models. 

This will support the service users that charities support. It will also assist commissioners to design and award the most effective contracts. 

Share this page