"We could all become suspects"

Police begin their work on a crime scene somewhere in Germany. They ask if anyone has seen or heard anything and who could provide decisive leads to help solve the crime. But witnesses can be unreliable. Even when they insist they are telling the truth, their perceptions of events may not accurately reflect what happened. Was it a tall or a medium-sized man? What color was his skin? Did he have light or dark hair? It seems that DNA evidence provides the only reliable account – but which information can be gained from it, or not?

Anna and Veronika Lipphardt are organizing the symposium: Extended DNA Analyses in Forensic Use: Possibilities, Challenges, Risks at the Freiburg Institute for Advanced Studies (FRIAS).
Photo: Klaus Polkowski

Three current, legislative initiatives are aiming to expand massively the possibilities to apply DNA analyses in criminal investigations. The most novel proposal of all would allow expanded analysis known as forensic DNA phenotyping (FDP) that provides information about eye, hair and skin color, and biogeographical ancestry testing (bga), which is believed to be useful for determining biogeographical origin. Yet Prof. Dr. Anna Lipphardt of the Institute for Cultural Anthropology and European Ethnology and Prof. Dr.Veronika Lipphardt of University College Freiburg emphasize that the possibilities opened by using this method are often exaggerated, while the risks, by contrast, are clearly underestimated. Together with an interdisciplinary group, the professors are holding a symposium in June 2017 called Extended DNA Analyses in Forensic Use: Possibilities, Challenges, Risks. Rimma Gerenstein asked both researchers why they find this draft legislation highly problematic and under which conditions the technology could be applied.

At the symposium experts from the fields of bioethics, science studies, the social sciences, genetics, statistics, law and forensics will discuss extended DNA analyses for the first time in Germany. Why is this discourse necessary?

Veronika Lipphardt: The debate about extending DNA analyses is too unilateral in Germany. We want to set it in a broader context. Up to now, it is almost always forensic experts and law enforcement authorities who are providing input for the draft laws – yet they do not represent all the expertise required to address this sensitive and complex subject. If, say, experts on population genetics and human genetics were asked, they would provide very different insights, on biogeographical origin, for example.

Anna Lipphardt: In our own work we've noticed how important interdisciplinary exchange is when dealing with such complex topics. And the more disciplines came into our initiative group, the more sophisticated our analyses became. Again and again, we heard, "Hold on, that doesn't fit from my point of view," "I can only emphasize that from my perspective," or "Careful, the way I see it, this would distort the results". Our group is the first in Germany that has noticed there are serious scientific errors in the draft legislation and the public statements of the politicians who support it. That's alarming in any case.

What errors are there?

Veronika Lipphardt: Let's take probabilities, for example. Currently, it is said that a DNA analysis can, with a probability of 98 percent, correctly discriminate between light and dark skin tones. Yet that isn't quite the case for the application, because you always have to consider the distribution of characteristics like light and dark skin pigmentation in the respective population. Otherwise, you can get the wrong results, "false positives" that can send investigators off in a direction that is completely wrong. The statistical argument, doctors call it "prevalence adjustment," doesn't seem to be known to forensic specialists and investigators in Germany – at least that's simply the impression we've gotten in our conversations. We cannot let politicians make decisions based on such foundations.

What do you fear?

Anna Lipphardt: The statistical weakness of the method comes into play particularly if the test indicates rare characteristics, making such results unreliable. Simultaneously, results like these seem to the investigators to be taking them in the right direction, because it is easier to narrow down a minority than a majority. So there is a fear that the police, led by such test results, would focus on minorities particularly often and unjustifiably. But even the law to align genetic data with fingerprints will have far-reaching societal consequences – and not just for those who perpetrate crimes. We could all become suspects.

How?

Anna Lipphardt: The law makes it possible for investigators, in all types of crime and without the orders of a judge, to gather DNA data and store it. But at a crime scene, it's not only the DNA of the guilty that is found, but also of the innocent, from you and me for example, because we just happened to have been in a specific place. DNA data of many people are also collected during mass screenings. This type of information can – in terms of data protection law – be just as sensitive as data on a mobile phone or surveillance data. What happens to this data? Does it get stored? If yes, for how long? Who administrates it and under what conditions and to whom is it passed on? That's been completely unclear up to now and is also the biggest problem of this debate. Draft laws fail to consider how legal and ethical consequences, including those that impact personality rights rand data protection, could be addressed.

Veronika Lipphardt: In the current state it would be an everything-goes law, which gives investigators very broad powers that are barely monitored by independent authorities. That is why it is so important that experts get involved in the debate who can advocate in the interests of the disproportionately affected groups – mostly minorities, who would unfairly become the focus of investigations.

The case of the Phantom of Heilbronn is a prime example of a failed application of this technology. Anna Lipphardt's has researched the case for a great deal of time. The DNA-based search for the murderer of police officer Michèle Kiesewetter led police down the wrong trail for a long time.

Anna Lipphardt: That's right. The police relied on a laboratory result for "biogeographical origin" that indicated the suspect was a woman from "Eastern Europe." The DNA report was written in Austria, where the practice was already allowed at the time. Combined with other evidence – for example, with the fact that the same DNA showed up at countless other crime scenes, including in Freiburg – it led the Heilbronn police to conclude that the suspect must be a woman from a Roma family. The police then initiated DNA screenings of hundreds of women, the majority of whom were Roma. Even today, we still don't know in detail what happened to this data. Investigators pursued this lead for two years without any audible opposition from the police or judicial authorities. That's alarming. This case demonstrates a very powerful and uncritical belief in technology. The investigators stopped pursuing all the other leads, including, for example, the one that would have led them to the milieu of the extreme right in Heilbronn and the NSU, to the real perpetrators. That's why an extended DNA analysis should always only be used as a last resort, when all the other investigative approaches have been exhausted.

Are there countries in which extended DNA analysis is well-regulated?

Veronika Lipphardt: We use England or the Netherlands as our orientation points, where such investigations are already admissible under law. Nevertheless, the decisive difference is, that there are painstaking and in part very strict regulations for them there. A commission composed of experts from all the relevant areas decides in every single case if such an investigation will be allowed. It's not done thoughtlessly, and extended DNA analyses are rarely used. The German draft laws don't provide for anything like this up to now.

Statement of the Initiative on the three draft laws (in German)

The symposium: Extended DNA Analyses in Forensic Use: Possibilities,Challenges, Risks is taking place on 9 June and 10 June 2017 at the Freiburg Institute for Advanced Studies (FRIAS) at the University of Freiburg. Representatives of the media are cordially invited to attend.

Press release

Website of the initiative to take a critical approach to the use of forensic DNA analysis
(with link to the symposium website)(in German)