Weeknotes May 6, 2022

It’s been a short – but really fun! – week for me. Tim has properly joined the team. I’ve been collaborating with the amazing Jess Morley on some work about patient and public participation in health data governance. I managed to finish the first draft of my report comparing food and data regulation. And I was on a panel about the future of UK data regulation.

What a difference an employee makes

The big change this week is that Tim has finally properly started as Research Director. You can read about his week in his weeknotes. The differences this has made already:

• An immediate increase in both the quantity and quality of our work. As well as the longer term stuff he’s working on, Tim grabbed hold of the Data Values Project consultation and whipped up both a Twitter thread and a response which required very little by way of review. And he gave me some great constructive feedback on my work comparing food and data regulation which has definitely improved it.

• On the flipside, having an employee requires quite a bit of administration in terms of registering them, sorting out pensions and insurance, creating organisational rather than personal accounts for things, and so on. Hopefully this is a one time burden though and will be easier when our next employee starts later in the month!

• We’ve had to align around how we want to work together. This has been pretty easy with Tim – he’s happy with async chat in Discord; knows how to use suggestion mode and styles properly in Google Docs; can edit Markdown files and figure out my weird Jekyll configuration sufficiently to create GitHub pull requests to change the website. I doubt this will be the case with many other employees! We’ve started the process of figuring out our way of doing things, like posting stand-ups and stand-downs at the beginning of each day, and getting together for confab every Monday morning. These again will evolve as we work out how to work best together, and as the team grows and changes.

Pondering patient and public participation practices

Now she’s got the Goldacre Review out the way, Jess has started work again on a paper we’re writing together (well, that she’s authoring and I’m throwing a few things into) about patient and public attitudes to the use of health data.

I don’t want to steal the thunder of that work, but there were a few things that struck me about the way in which this research typically gets framed.

First, the questions tend to be framed about what the participants want for themselves. With thanks to all those on Twitter who helped me find the reference, this is different from asking them how they think the system / society should work in general, behind a veil of ignorance. Questions that highlight how they could be on the losing side of choices made by other people – subjects of decisions shaped by others’ data and their data sharing choices – might give different results. Deciding whether data should be shared with companies to help them develop new cancer treatments has a different impact when you’re reminded that you or a loved one might develop cancer one day. It would be interesting to test the degree to which this changes people’s responses.

Second, it seems likely that the accountability and regulatory environment in which data sharing takes place has an impact on the trust people feel in the system. However most public attitudes research seems to focus on the internal aspects of data collection, use or sharing, rather than the external aspects. For example, the conjoint analysis in CDEI’s public attitudes tracker survey looks at things like who is sharing what kind of data, with whom, for what purpose and under what conditions, but not how breaches are detected or conditions are enforced, or what penalties are in place. It would be interesting to incorporate those considerations into future studies.

Finally, practically all the studies seem to frame choices about data governance design as binary – the public has to express a preference between either (a) individual decisions (opt in or opt out consent) or (b) institutional decisions. Independent collective decision making – for example by a Board or a democratic vote – is rarely given as an option. This reduces the question to whether you individually want control or not (with obvious results), rather that enabling consideration of the broader design space of possibilities.

While we’re on public attitudes research, worth also checking out the Ada Lovelace Institute’s review of public attitudes to data and AI regulation, and their write up of the roundtable they hosted (which I also wrote up here). Their top five findings:

1. There is consistent evidence of public support for more and better regulation of data and data-driven technologies. (But more research is needed on what the public expects ‘better’ regulation to look like.)

2. The UK public want data-driven innovation, and they expect it to be ethical, responsible and focused on public benefit. (Determining what constitutes ‘public benefit’ from data requires ongoing engagement with the public.)

3. People want clearer information about data practices and how to enact their data rights. (But what this looks like in practice is not yet fully understood.)

4. Creating a trustworthy data ecosystem is critical to protecting against potential public backlash or resistance. (Emerging research suggests that regulation is a key component in those ecosystems.)

5. Public concerns around data should not be dismissed as lack of awareness or understanding, and simply raising awareness about the benefits of data will not increase public trust. (More research is needed to understand the connection between awareness of and attitudes to data.)

One last thing to note here: public attitudes research is different from participatory data governance. Yes, there is and should be academic interest in what people think, but organisations who are doing things with data should also be creating the governance and accountability mechanisms that mean those thoughts, attitudes and opinions actually affect what organisations do. Again, I think asking people to make real, impactful decisions is likely to lead to rather more nuanced responses than abstract opinion surveys provide.

My obsession with food regulation continues

I’ve finally completed the first draft of the paper / report I’ve been writing comparing food and data regulation. Tim made some very good suggestions about including the role of international regulation, so we’ve added a nod to that.

I’ve circulated the draft to just a few people for a first round of feedback; if you’re not one of them but you’d like an early look then do let me know.

I haz opinions on the future of UK data regulation

Final fun thing of the week was being on a panel about the future of UK data regulation at the New Statesman / TechMonitor Digital Responsibility Summit. I’ve written up my notes from the panel here.

It was a good opportunity to try out some of the arguments that we’re trying to put forward about the importance of community engagement and participation, and collective data governance. It was also nice, from my perspective, to be able to more-or-less focus on that one aspect of data regulation, and not to worry too much about disagreeing with the government line – after all, one of my goals embarking on Connected by Data was to get comfortable with having more of an “outsider” voice.

As mentioned by John Whittingdale MP during the panel, the Data Bill will be announced in the Queen’s Speech on Tuesday. We might get a hint of what it looks like then but I’m not anticipating seeing the draft legislation until a little later in the summer. So far the spin has been largely about getting rid of cookie banners. We’ll be paying particular attention to changes to legitimate interests, if there are any.