Feed me Alan, make me strong!

| 3 Comments
| More

It had to happen sooner or later - the police have been accused of arresting people solely for the purposes of adding further 'suspects' to the National DNA Database. It's time to revisit why the current approach threatens to undermine the whole worth of DNA as an evidential tool.

The claim, made by the Human Genetics Commission, is that police forces now obtain DNA samples simply to build a suspects database; that the value of DNA itself is ignored, since in many cases there are easier and more obvious ways to identify suspects; and that the stigmatisation of individuals, particularly certain ethnic groups (it is claimed that up to three-quarters of young black men are in the database) is driving a wedge into society.

The National DNA Database is as strange and utterly unique creature. No other country has one like it. In the quest for target-driven, pre-cog style policing, the government has created a monster that can only deliver the results expected of it if it is fed more and more DNA information - and since it will no doubt be expected to achieve 'better' results against these targets every year, that amount of DNA will have to keep rising. Home Secretary Alan Johnson has vigorously defended the government's approach. However, the outcome is going to look more 'Little Shop of Horrors' than Minority Report as the beast keeps demanding meat to feed its growth.

little_shop_horrors_xl_01.jpg

A leaked shot of the DNA Database requesting more flesh from its unwitting creators

There are a huge number of reasons why the current approach towards the National DNA Database is wrong, but here are few:

  1. Stigmatisation and misuse of results: Even if the National DNA Database (NDNAD) is accurate and correctly used by its operators, that level of process will not extend out into society. Individuals will be stigmatised simply by appearing in it. Ironically, it is the Home Office itself that has set an example of how to misuse the results, after a lawyer claimed she was dismissed from her position when background checks for a contract role in the Identity & Passport Service revealed a match in the database. Rather than the Home Office and her employer accept that these arose from a malicious accusation that did not lead to a conviction, she lost her job. Cynics have used cases such as this to argue that if everyone were in the database, then that would prevent this problem. But that's wrong because...
  2. Assumed infallibility of the database: ...databases are built and operated by humans, and that means they contain errors. Maybe not many, but when a database is assumed to be correct, that makes it very hard for an individual to clear their name. Victims of errors in the NDNAD will find themselves being wrongfully arrested, failing to get jobs because they fail security vetting, and being stigmatised by a society obsessed with CRB checks. And assumed infallibility can kill - earlier this year a young South African man took his own life because the authorities refused to believe he was a South African national, and refused him the papers he needed to work. In the UK, a woman claimed her son had killed himself after being stigmatised for inclusion in the database. Even the government's own advisors have publicly dissociated themselves from the way the government has misused research to prop up the perceived infallibility of the system.
  3. Breakdown of procedures: But even if the database were infallible, what about the procedures around it? The likes of CSI and Silent Witness would have us believe that DNA evidence is only ever handled by white-coated scientists in labs, but it seems inevitable that in the near future the actual analysis will happen in police stations, in police cars, or even by police officers on the beat. Mistakes will happen, samples will be mixed up, contaminants introduced - German police hunted the phantom of Heilbronn for 15 years until she was found to be a worker in a DNA swab factory. And this will have three consequences: firstly, police will rely on DNA evidence without bothering to find anything else to support it. If they have DNA at a crime scene, other evidence will be treated as secondary, and the suspect will be pressured into admitting the crime. Secondly, DNA will be used in the most minor of cases, and we will all end up being arrested for being sick on a pavement in our student days, then bullied into accepting a caution and hence a criminal record. And thirdly, sooner or later a high profile (and most likely very serious) criminal case will collapse when it is found that there were errors in the handling of the DNA evidence - and that will be the end for DNA as a trusted forensic tool.
  4. Familial/predictive use: Perhaps the most disturbing potential use of the NDNAD is that of spotting family connections, and this has rumbled around for many years now: if we can get a partial match to a sample, then let's go and question that person until they tell us which family member committed the crime. From there it is only a short step for the government to decide to use the NDNAD to predict which families are most likely to commit crimes, and intervene before it happens. Children will be born stigmatised by the system regardless of circumstance.

There are of course circumstances in which the retention of DNA evidence is perfectly reasonable: few would question the proportionality of storing DNA records of individuals convicted of violent or sexual crimes against the person, for example. There's a strong case to argue in favour of a short period of retention of everyone convicted of any crime, or currently under investigation after arrest, but that data has to be destroyed once its purpose is complete. And as for the million or more unconvicted suspects, victims, witnesses and volunteers in the National DNA Database? Their records should be expunged immediately.

What about volunteering DNA? In his blog, Philip recently stated that he'd like his DNA on file in case his identity credentials were misused. Sorry Philip, but you've been suckered by the propoganda: there is a fundamental conflict between provision of trusted identity and the fight against serious and organised crime. If your credentials are misused, we need a system that has self-healing qualities with the ability to revoke credentials promptly, backed up by a pervasive infrastructure that can only be achieved through federation rather than the current centralised approach to ID favoured by government. I suspect there's nothing the government would like more than to get everyone's DNA on file so that it can build a predictive police service, but the actual outcome will be treating everyone as a suspect, and still ignoring the victim.

The Conservatives have published their commitment to comply with the European High Court ruling in S v Marper to delete the records of innocent individuals, so there's a good chance we could see a return to a more proportionate, scientific and trustworthy use of DNA. In the meantime, maybe we should change the NDNAD acronym to the AUthorised Dna REpositoritY - or AUDREY for short...

3 Comments

  • I'm not sure if I want to even say this, but I'm beginning to wonder if the only consistent position on this is for _everyone_ to be on the DNA database. From a societal perspective, this would be the best outcome (surely) because it would encourage people to modify their behaviour but in an open and well-understood way.

  • Dave, no! Don't say say that sort of thing, your madcap comments have a habit of bubbling up in policy sooner or later. Maybe there's a role for such a system, but not in the hands of the government, not with such a mixed range of uses, and not without any sense of benefits to law-abiding individuals. The world in which government can be trusted with this information is a long way off yet...

  • Leaving aside all the (vital) privacy and civil liberties concerns, adding everyone's DNA to the database would be a stupendous waste of public money, since it wouldn't help catch more offenders.

    How can I be so confident about this? Because the Home Office has been running the necessary field trials for the last 5 years, and has published the results. You can do the analysis yourself. Here's what you need to know in order to run your very own Evidence-based Policy-Making exercise (as opposed to Policy-Based Evidence-Making, which is what the Home Office seemed to be trying to get the Jill Dando Institute to do on its behalf):

    1. Like the good scientist that you are, start by formulating your hypothesis. "Adding the DNA of those never charged with an offence will help clear up more crimes" would do nicely.

    2. Observe that in 2001, the law in England and Wales was changed to allow DNA profiles taken from people who had been charged to be permanently retained even if they were not convicted. In April 2004, a second legal change came into force that allows the police to take DNA samples without consent on arrest, rather than charge. Prior to this, the police could take some DNA samples on arrest, but only if they needed the DNA to investigate the specific offence for which the person had been arrested. Because of the 2001 change to the law, all these DNA profiles are also retained permanently. The result is that since 2004 hundreds of thousands of DNA profiles have been retained that would otherwise have been deleted. It is now estimated that approx 1 million of the 5 million profiles on the database belong to those never convicted of a crime.

    3. Go here to download the Home Office post-2002 crime statistics spreadsheet:

    http://www.homeoffice.gov.uk/rds/pdfs09/recorded-crime-2002-2009rev.xls

    Load the annual "Recorded Crime" numbers into a spreadsheet.

    4. Amble over to the National Police Improvement Agency site and get a copy of the National DNA Database report 2007-9:

    http://www.npia.police.uk/en/14395.htm

    Go to page 35 and load the annual numbers of "Detections of crimes in which a DNA match was available" into your spreadsheet. Pause to note that these aren't necessarily crimes where detection resulted solely from a DNA match - just those where a match was available, possibly amongst other evidence.

    5. Divide the annual detection numbers by the annual crime numbers. Observe that the proportion of recorded crime involving direct DNA detection has stayed within a whisker of 0.37% since at least 2002. Since this covers the time period where hundreds of thousands of unconvicted people have their DNA permanently added to the database, conclude that the hypothesis is not proved. Adding the DNA of those never charged with an offence does not help clear up more crimes. Conclude that we should stop spending public money on doing it, and return to pre-2004 retention rules.

    Congratulations - you've just become an evidence-based policy analyst.

    Acknowledgements: Genewatch's excellent analyses, especially "The DNA Expansion Programme: reporting real achievement?"

    http://www.genewatch.org/uploads/f03c6d66a9b354535738483c1c3d49e4/DNAexpansion_brief_final.pdf

  • Leave a comment

    Disclaimer

    The views expressed in this blog are my own, and do not necessarily reflect those of any client or other organisation.

    Subscribe to blog feed

    Archives

    Categories

    Toby on Twitter

      Recent Comments

      Andrew Watso on Feed me Alan, make me str... : Leaving aside all the (vital) privacy and civil li...
      Toby Stevens on Feed me Alan, make me str... : Dave, no! Don't say say that sort of thing, your m...
      Dave Birch on Feed me Alan, make me str... : I'm not sure if I want to even say this, but I'm b...

       

      -- Advertisement --