“Big public data” as a danger
Professor Markus Schefer
Is “Big public data” an opportunity or a danger to society? A debate on the future of the fundamental right to informational self-determination. An economist and a legal scholar take up positions.
The fundamental right to informational self-determination guarantees each individual the right to decide what personal data she or he wishes to reveal and to whom. This holds especially in relation to the state. The state is permitted to gather, process or pass on people’s personal information only under very restricted preconditions. The constitutional right thereby guarantees an individual the possibility to determine under which identity she or he wishes to be perceived in public dealings. My public persona should be an expression of my autonomy.
In its form up until now constitutional protection has, to a significant degree, been dependent on the kind of personal information in question: The deeper the insights the information in question gives into the personality of the affected person, the higher the hurdles for the state in seeking to gather, process or pass on that information. Against this background the fundamental right to informational self-determination seems primarily to be a right to data protection; thus its description in the Charter of Fundamental Rights of the European Union.
Big data calls into question the very basis of this concept. Primarily, big data changes the connection between the disclosure of personal details by individuals and the knowledge of personal attributes in the possession of third parties, in particular the state. In one concrete case it was demonstrated that anonymized gene sequences placed in research data banks that are accessible to the public could, with the help of just a little more data, be assigned to particular individuals. This makes it possible for third parties to have knowledge of personal details of which the individual may, in certain circumstances, be unaware.
Given this background, state knowledge of a personal characteristic is no longer necessarily based on the fact that the individual has disclosed the information in question. If the state has access to a large amount of different details about people, then big data makes it possible for the state to infer certain characteristics of groups in the population or indeed of particular individuals, and to do so with a certain degree of confidence. For example, with the help of big data police today can localize those areas in which there is an increased risk of burglary. This seems a less problematic issue. But what if the police, without any concrete suspicion, started stopping people in public spaces in the areas described as high risk?
When the state acts towards a group of people merely on the basis of a significantly elevated probability of a particular kind of behavior, then the state is acting on the basis of stereotypes. In the context of big data, these stereotypes are not the result of opinions formed within society but rather the result of algorithms used to make correlations transparent. This does not, however, fundamentally alter the fact that a state’s actions towards individuals, if based on their fitting a stereotype, has the potential to be demeaning.
This is particularly clear, for example, when big data is used as the basis for decisions on releasing individuals from custody. Big data is already used for this purpose by certain parole boards in the United States. The problem is not that the likelihood of successful predictions is too low; in fact it may well be the case that the probability of such a prognosis being correct is greater than a judgment made by a court psychiatrist on the basis of an assessment of the person in question. The problem is, rather, that the affected person is not treated as an individual in his own right but simply as an arithmetical entity.
Against this background, the fundamental right to informational self-determination must be rethought and modified. In these circumstances, the self-determination of the individual in the shaping of his social identity can no longer be protected merely by guaranteeing him the authority to dispose over his personal details. Consent to the use of personal information in this respect is no longer a suitable instrument for protecting privacy.
The focus must be shifted from the disclosure of private details toward the way in which data of any origin is to be handled. Here it is not just a question of highly personal data on a particular person, but also of how the state deals with data that may neither be especially personal nor give any indication of the individual’s identity.
Linking, combining, and evaluating data of any kind has become central to issues on fundamental rights, regardless of whether the data in question is of a more or less personal nature. This is especially the case in relation to data that has been gathered and processed with the consent of the affected person. Here, there is not yet much clarity on the standards set by the fundamental right to informational self-determination. It is the responsibility of legal scholars in particular to drive this new orientation of fundamental rights.
Markus Schefer has been Professor of State and Administrative Law at the University of Basel since 2001. Following a degree in Bern and advanced studies at UC Berkeley and Georgetown University, Washington DC, he completed his doctorate at the University of Bern.