Public dialogue on the use of data by the public sector in Scotland

Conclusion

This public dialogue set out to:

• Explore the ethical implications arising from the use of data by the public sector.
• Develop a set of principles to inform future data-driven research projects and policies.
• Explore the possibility of a longer-term approach to public engagement as a means of providing external scrutiny of the public sector.

This report has highlighted the thoughtful ways in which participants engaged with these big issues and the learning journey they went on to develop their final ethical guidelines.

In this concluding chapter we outline the key findings from this public dialogue that add to the existing body of knowledge on public attitudes to data. It also provides reflections on public dialogue format and the potential for future public engagement on data use.

The panel were generally trusting of the public sector’s use of data, but wanted reassurance that ethical principles would be followed alongside existing legal frameworks.

From the start of the process, participants had fairly high levels of trust in the Scottish Government and wider public sector. This trust was linked to an understanding that use of data by the public sector can have benefits for society, particularly in relation to health emergencies such as COVID-19. Participants were more inclined to trust the public sector with their data than the private sector, as they felt the was more likely to be motivated by public good rather than by commercial gain.

Views on the role of the private sector did develop over time. Although in earlier sessions there was scepticism over the trustworthiness of the private sector, by the final session there were more nuanced perspectives. One perception was that some private sector companies can provide benefit by using data and that it could be acceptable as long as their use of data was not solely driven by profit, or to only benefit an organisation and not the wider public. However, there were differing views on this, with some feeling that involvement of the private sector in any public sector data projects was problematic. As a result, the final guidelines relating to private sector use of data about citizens were not endorsed by all. Despite not being a focus of the panel, participants said they would have welcomed more space and time to explore the role of the private sector in detail. This issue was explored in additional workshops (the findings of this can be found in a separate report on the Scottish Government website).

This sense of trust was based on an expectation that the public sector would follow rules and regulations around the use of data and be held accountable for any misuse. Having learned more about existing data protection legislation, the panel generally felt reassured that systems were in place to govern the use of their data. However, the existence of data protection legislation did not remove the need for ethical principles to be followed and it was clear that both legal and ethical considerations were important for building public trust. Though some of the panel’s ethical guidelines reflected the content of existing legislation such as GDPR (for example principles of purpose, transparency, and accountability), participants nonetheless viewed these as fundamental ethical considerations for the public sector to follow.

Views were influenced by the context in which data was being used, specifically whether it was an emergency situation or not. The panel therefore identified a need for some flexibility to be allowed for, while adhering to basic principles.

A unique feature of this public dialogue was that participants were given the opportunity to scrutinise and share feedback on real public sector data projects. As some of those projects had been delivered under the unprecedented circumstances of the COVID-19 pandemic, this raised specific issues for the panel around how the public sector should approach data in an emergency.

The panel understood that data had to be used quickly to support decisions related to COVID-19 and they recognised the benefits of doing so. To take the shielding list as an example, it was acknowledged that the Scottish Government and partners were dealing with exceptional circumstances and that use of data had public health implications. The panel’s ethical guidelines therefore reflected the need for agility, allowing for quick action to be taken in the event of another pandemic or similar emergency. As a minimum, however, there was an expectation of transparency around what data were being used and for what purpose.

The panel emphasised the importance of data quality as an ethical consideration when using data. They wanted to see a minimum quality standard put in place which future data sharing projects would be required to meet.

In this public dialogue, the panel discussed data quality in some detail and this became an important factor as they developed their final ethical guidelines.

Data quality was seen as a measure of accuracy, and the panel stressed the importance of data-driven decisions being made with the most up-to-date and accurate information. Underlying this view was a recognition of the impacts of some of the decisions that the public sector had, and would have, to make through their use of data. The obvious example, again, was the shielding list and the importance of this being based on accurate information.

Data quality was also linked to fairness, with the panel feeling that gaps in data could lead to individuals being excluded or not benefitting from certain initiatives. The panel also recognised that the quality of data may differ between different groups in society, and they questioned whether this might impact on policy decisions.

The importance of data quality was reflected in the panel’s suggestion of a minimum quality standard for each data project to meet. This recommendation applied to all projects, but where there was private sector involvement the panel specifically wanted to see checks on data quality carried out by the public sector. This, again, reflected the higher levels of trust the panel placed in public sector organisations in comparison with the private sector.

The panel was sensitive to the impacts of data use, both positive and negative, on marginalised groups.

The potential use of special category, or sensitive, data raised specific ethical considerations for the panel. Protection of individual privacy was seen as particularly important when dealing with sensitive data, and the panel had concerns that misuse of this type of data could lead to individuals being discriminated against. For example, when reviewing the Policing the pandemic project there was concern that the use of data on health-related vulnerabilities might increase the likelihood of individuals being subject to police enforcement for non-compliance. Participants raised concerns about the risks of vulnerable groups being stigmatised on account of characteristics such as their underlying health conditions.

The panel’s sensitivity to the impact on marginalised groups was also reflected in their discussions about public benefit. As they grappled with their own interpretations of public benefit, they arrived at the view that this can mean benefits to society as a whole but in some cases can also be restricted to a specific, smaller group. In the latter scenario, they felt there can be public benefit as long as this was not at the detriment or harm of other people.

The process highlighted the value of informed dialogue, and that the opportunity to learn about and deliberate on topic can lead to more informed decision making.

This public dialogue has demonstrated the ability of the public to engage with a complex, and in some cases unfamiliar, topic and develop thoughtful principles for the future. The opportunity to have dialogue with experts, and with each other, helped participants to formulate their views on the most important ethical considerations.

However, as noted in the previous chapter, the panel’s response to certain topics may have been influenced by presentation style and delivery. Over the course of the public dialogue, participants heard from twelve different experts and it was inevitable that these would resonate differently with participants. This emphasises the importance of introducing content gradually and in different formats, giving participants the space to ask questions and seek clarification on key points, and being open to different ways of presenting the information.

There was a great deal of support for future public engagement on the use of data by the public sector, and a public panel was seen as a good approach.

The panel strongly believed that the public had a role to play in decisions about the use of data by the public sector. A unique aspect of this public dialogue was that participants acted as a panel, meaning they had the opportunity to review and appraise past, current and future data-led projects. The panel-style approach placed participants in the role of evaluator, giving feedback (sometimes directly to those involved) on data-led projects in a way that could influence decisions around their future delivery. The panel-style approach (with the group meeting over a four month period, with gaps of up to two weeks between workshops) also meant that they had a fairly long period of time to immerse themselves in the topic, reflect in between sessions and gradually develop their ethical guidelines in response to what they learned.

Reflecting their positive views on their own experience, participants felt that a public panel was a good model to replicate. They specifically felt that the opportunity to hear from and engage with experts, and to provide feedback on specific projects, were valuable. There was also clear appetite for future public engagement to feedback on specific data-led projects.