On 10th March 2015, a plenary meeting was held to conclude the data sharing open policy process, attended by representatives from privacy groups, civil society organisations and government.
Peter Lawrence (Cabinet Office) introduced the session and gave a recap of the policy process so far. The open policy process has taken place over the past year. It was established due to the recognition that the complexity and potential controversy of data sharing means it could not be developed behind closed doors. The policy process set out to find and build consensus, but where consensus could not be found it would seek to understand the disagreement.
He summarised where each of the strands had got to. This is outlined in the Conclusions policy paper.
Progress of the strands
Simon Burall (Involve) introduced the next session. There are issues still to be resolved across the three strands, but the policy process has reached an area of rest. Participants were split into groups to discuss the strands and given the task of describing what the state of rest looks like. Specifically, each group was asked to discuss what had been banked (i.e. where agreement had been reached); what is still to discuss, and what the different options are.
Tailored public services
- It’s for social policy objectives
- It must demonstrate the benefit to the individual.
- Not punitive intent
- This is to be used if can’t get informed consent
- Make an offer to identified people
- Process transparent
- Social care and health ‘are in’
- Onward sharing of information. Departments – Partner – Onwards?
- Good business case – 1 local authority + 1 department
- How to replicate to be cost effective?
- Explore “fall out” findings further
- Consent/data relating to deceased
- MyData – individuals hold own
- LEPs – where in definition of “bodies”
Research & stats
- Fundamental scope of 3 distinct sub strands (ONS; Trusted Third Party de-identified;HMRC) including how they fit together (coherence)
- Getting HMRC to same starting point as others
- Broad power for all public authorities to disclose to each identified data for research and statistics was a step too far
- Agreed on the need for, and the terminology on, a power for all public authorities to to disclose de-identified data to accredited safe havens using trusted third party indexers for accredited research purposes to accredited researchers
- Bodies delivering health and social care – appropriate they have separate treatment and get no new powers in the de-identified strand
- Any requirement for consent would be unworkable for this strand.
- Trusted Third Party governance and control
- Need for sanctions in de-identified strand?
- Identified strand: are any of the alternatives to Parliamentary scrutiny of Information Sharing Orders for ONS acceptable?
- Impact of European Data Protection Regulation
- Definitions in the legislation still not agreed:
- Public benefit
- Public interest
- Public body
- Flexibility and future proofing.
- Scope? Is there scope in the future for data from health bodies to be included in the powerShou
- Need to make the case continually for the powers
- Public engagement
- Conversations beyond legal barriers
- Next steps – White Paper
- What structure?
- Do all strands belong in the same bill?
- Would any legislation move forward with ONS power, HMRC power and de-identified power?
- Fit with other legislation e.g. new EU Data Protection Regulation
- Governance bodies
- Of data shares
- How much of the policy should be in hard legislation and how much included in statutory or non-statutory guidance
- Most agree error is complex and not suitable / proportionate for data sharing solutions
- Agree that fraud is a significant problem, which needs addressing
- Fraud is a cost everyone
- Pilots provide the opportunity to to better understand tolerance level / cost effectiveness of proposals
- Survey / feedback to understand citizens views of data sharing
- Essential that appropriate privacy safeguards are present in proposals
- Feedback mechanisms built into pilot proposals
- Consensus on PIA / DPA principles
- A number of representatives from the public sector felt there was greater evidence / certainty of value of data sharing than set out in the paper
- Data is a facilitator for decision making, not the solution itself
- Evaluation / what does success look like?
- What are the tolerance levels for fraud?
- Cost effective
- Evaluation is linked to the level, quality of analysis,data cleaning, etc.
- Best approach is to rely on consent if possible
- Only rely on compulsion (i.e. legislate) if absolutely necessary
- Clarified definitional issues
- How will DMI work in practice
- Openness (FoI)?
- Will local authorities be included?
- Is DMI an opportunity to promote best practice and “level up” approach to debtors
- Prioritisation of debts
- Tension between benefit to individual and efforts to “get the money in”
- Sharing with private sector? Individuals have debts with a number of different organisations – private & public
- Balance success between collection and intrusion
- How to get full picture when some debts not in DMI
- How to separate success of DMI as a project/initiative and the data sharing aspects
- Overlap between DMI and statutory debt solutions (e.g. bankruptcy, DRO, etc.)
Reflections from the Minister
Having heard summaries of the group discussion, the Minister for the Cabinet Office reflected on the process. There is a potentially large public benefit to be gained from data sharing, but it is a contentious area with potential risks, and needs to be done in a measured way. There would be no better way to raise suspicion than conducting the policy process in a closed way. Participants had taken part in something groundbreaking – a genuinely open policy making process. He recognised the time and energy that participants had given.
Reflecting on the Tailored Public Services strand, he commented that one of biggest challenges of supporting troubled families is data sharing. Actual and perceived barriers hamper government’s ability to help people. There are two key issues – joining up data and joining up money. Both need to be joined up as close to the individual family as possible, so it’s not about creating big databases. He highlighted the importance of working out the settling point [between opportunities and concerns].
On Research and Statistics, the Minister commented that there are a huge range of issues that we don’t understand, but we have the data that might help us understand. Again he concluded it was about finding the settling point.
Regarding fraud and debt, the Minister stated that there are lots of misperceptions about where data can be shared already. Much of what is needed is about clarity – finding a way through the current patchwork of gateways. On debt he commented that government is not very good at collecting money from those who won’t pay, and not very sensitive to those who can’t pay. Data sharing could help distinguish the two, and support a more humane interaction with those who need it.
To conclude, he highlighted that data sharing is an issue that most new governments don’t want to touch. They get frustrated and see the need for it over time, but then it’s too close to an election. There is no party politics in this and a real opportunity to take forward in early stages of next government.
Q – Why has the role of individuals owning and controlling their data been out of scope for this exercise?
A – Partly because you have to limit things because it could be a conversation you have for ever. Gov.uk’s work on this is now in live beta testing.
Q – Is there scope for civil society organisations to engage on the Debt Market Integrator?
A – Absolutely yes.
Q – What cross party support is there?
A – A lot. All main parties have had recent experience of government and understand the issues.
Q – Potential White Paper – is it overly complicated with strands structure?
A – Good question with no perfect answer. Issue is the simpler you make it, the more all encompassing you make it. It needs to be simplified, but not at cost of benefits we can take from doing in a more complicated way.
Q – It’s been helpful to have a strong but open leadership from Cabinet Office. Would like to build on that post election.
A – Credit to the team. It’s not a comfortable way of doing things. Credit to team for embracing. We now understand what the prize is.
The Minister departed and the conversation moved onto the next steps. Groups were formed around four issues for consideration.
Policy & white paper
- Creating White Paper that’s accessible and engaging. Rational.
- Difficult to understand across the strands. How do we describe strands. Alternative:
- Tagline: Reprocessing data for public good.
- Two themes – administrative (things that have positive impact on individual) – statistical (no impact on the individual)
- Using examples and case studies.
- Identifying cultural as well as legal issue.
- Scrutiny processes that work in a cohesive way.
Citizen engagement and consultation
- Who are we trying to engage with? How to get representative groups?
- Meaningful engagement.
- Need to ensure not hijacked by particular groups.
- Clarity about intention and outcomes.
- Need to understand public attitudes and what individuals think is important.
- Different strands using for operational purposes and others not.
- Sharing consensus points.
- Simple messages that resonate with public
- Need to engage public through case studies and examples – can’t engage in abstract.
- E-petitions will gather lots of feedback. Need to think of digital forms of engagement.
- Red lines: Don’t renegotiate – consult – on banked ideas. But will be a question for the next government. But…
- Can’t guarantee that will be sacrosanct, but will need to be substantial case for change. New evidence, etc.
- Complex message – and different messages on different strands. Essential to use plain and jargon-free language.
- Need a communications strategy. Stakeholder analysis and segmentation to tailor messages.
- Need to have good examples that resonate with people.
- Important to get champions from outside of government.
- Being clear about the public benefits and risks – so can have mature conversation.
- Using the right channels at the right time.
- Can expect negative media attention – need to put consistent and clear message out.
Devolved nations and regions
- Lots of discussion going on about devolution – need to make sure that we’re engaging those areas with discussions
- Control over aims, rather than data. Some data can’t be owned by devolved areas.
- Joining together contacts in devolved regions. Using networks and assets.
- Future proofing. Establishing trust.
Simon Burall (Involve) recognised the time people had given up to participate and thanked the Cabinet Office team for their role.
Peter Lawrence reported that the Cabinet Office team would take on the points they’d heard in the meeting. Over the next few weeks the will start to pull together into a draft White Paper, which a new government can make a decision to take forward if they wish. The Cabinet Office team will have to go quiet during the pre-election purdah period, but will talk to people again when they can .
When there’s a new government, there will be something robust to give to them, with options for next steps. One of the key questions is how to engage the public in this conversation.