Right to be Forgotten?
Posted: Tue Apr 26, 2016 4:40 pm
I have lately been pondering the idea of the "right to be forgotten" and its implications for social scientific research. Though most of the discussions surrounding the right to be forgotten have focused on the limits and responsibilities of corporate and media data “controllers” who gather, store, and use vast troves of digital data, I think there are some interesting ethical concerns that should be taken seriously by scholars and that are particularly germane to our discussions of data access and transparency. And these issues impact quantitative and qualitative research alike.
Beyond concerns about anonymity, few have yet questioned the value of retaining and publicly sharing large amounts of data in perpetuity. We generally seem to presume that preserving data is in and of itself desirable. But what might be the unintended consequences of, say, retaining outdated information? Might we wind up creating incentives to use and re-use old, poor, and/or de-contextualized data simply because they are easily accessible in our repositories and the costs of generating and preserving new data so high?
And what of the dignity and privacy of those captured by our data? Perhaps our subjects have fully consented to data use, with or without anonymity guaranteed. Does their signature on a form on April 26, 2016 mean that they can never withdraw--2 years, 5 years, 10 years later? What of those who never actually signed a form--those whose "public" data were collected, for example, from online platforms (where researchers tend to conclude that we can make use of whatever we want as long as the data are not shielded by "privacy settings")? What if any one of these subjects comes to discover that their data are being used in distasteful ways? What if they determine that their data represent episodes or conditions that they find shaming or otherwise harmful? Once data are made public, it becomes exceedingly difficult to adjudicate such matters and effectively impossible to grant one the right to be forgotten.
Of course, none of these concerns are unique to DA-RT. They apply to any shared data. But DA-RT broadens the scale and raises the stakes.
Beyond concerns about anonymity, few have yet questioned the value of retaining and publicly sharing large amounts of data in perpetuity. We generally seem to presume that preserving data is in and of itself desirable. But what might be the unintended consequences of, say, retaining outdated information? Might we wind up creating incentives to use and re-use old, poor, and/or de-contextualized data simply because they are easily accessible in our repositories and the costs of generating and preserving new data so high?
And what of the dignity and privacy of those captured by our data? Perhaps our subjects have fully consented to data use, with or without anonymity guaranteed. Does their signature on a form on April 26, 2016 mean that they can never withdraw--2 years, 5 years, 10 years later? What of those who never actually signed a form--those whose "public" data were collected, for example, from online platforms (where researchers tend to conclude that we can make use of whatever we want as long as the data are not shielded by "privacy settings")? What if any one of these subjects comes to discover that their data are being used in distasteful ways? What if they determine that their data represent episodes or conditions that they find shaming or otherwise harmful? Once data are made public, it becomes exceedingly difficult to adjudicate such matters and effectively impossible to grant one the right to be forgotten.
Of course, none of these concerns are unique to DA-RT. They apply to any shared data. But DA-RT broadens the scale and raises the stakes.