Clearly serving the public good
Late last month, I attended the Committee which advises me on the access, use and sharing of public data – the National Statistician’s Data Ethics Advisory Committee (NSDEC).
As I’ve said before, the UK has the potential to be a global leader in data science. But we can only be successful if we maintain the trust and confidence of the citizens and businesses that provide their data. In this context, the role played by NSDEC is crucial.
There were three issues I discussed with members which I thought were particularly interesting.
Firstly, the Committee is taking on a new role, scrutinising projects referred by ONS’s Microdata Release Panel, which governs the Approved Researcher Process. We are expecting a lot more interest in the use of microdata as we work on more administrative and other new data. I was encouraged by the insight Committee members are able to bring to provide us with an additional safeguard designed to ensure that all approved projects meet our ethical standards.
Secondly, a particular priority for us is to make sure we are at the forefront of developing and implementing innovative new methods to improve official statistics. Web-scraping – a tool for extracting information from the underlying HTML code of websites – is one method we have been exploring for some time. As we deepen our understanding of what is possible, new questions are raised and the Committee provides a valuable way of looking at our ideas through an ethical lens.
As members of the Committee discussed, there is a complex legal and ethical landscape for ONS to navigate in its use of web-scraped data. Not only do we need to meet the extensive terms and conditions of websites, we must also be able to demonstrate to the public and businesses that collecting data in this way serves user needs and the public good.
To help staff navigate this complex landscape, members of NSDEC are considering new guidance for ONS staff on using web-scraped data. Members of the Committee provided a range of suggestions about this guidance, and will be considering the issue again once recommendations have been implemented.
And finally, the last item I sat in on was from the Data Science Campus, who updated the committee on their plans to explore the usability of a range of data sources for the production of migration and tourism statistics.
The Committee members had mixed views on the plans. While some highlighted the benefits of using, for example, publicly available geo-located social media data to improve official statistics, others highlighted ethical implications relating to informed consent.
For me, this sort of discussion highlights the value of the Committee. Drawing from the diverse background of Committee members, we are able to improve our understanding of people’s concerns around the sharing of data. We can also begin to think about what reassurance we need to provide to the public and businesses that their data will be used legally, ethically and safely at all times, and only for the purposes of producing aggregate official statistics.
With proposed new uses of data central to our strategy it’s important that we consider the ethical implications of accessing, using and sharing data. I would therefore ask you to consider ethics as part of your own work and make use of our ethics advisory committee in steering your work to ensure it is ethical and for the public good.