Tuesday, 24 November 2009

DNA: looking for the evidence

Today's report (pdf) into the national DNA database by the Human Genetics Commission - the indepedent body charged with advising the government in such matters - is a very worthwhile document. I'd recommend it to anyone who wants a thorough grounding in the history of the database and the current debate surrounding it. News coverage this morning focussed on two striking findings - suggestions (not exactly new) that the police were arresting people for no better reason than to give them an excuse to take their DNA, the claim that three quarters of all black men were now on the database. These are important issues, but there's much more.

The authors make a number of recommendations that, from my point of view, are disappointingly lame. They call for more regulation (has any report ever called for less regulation?) and they have little to say about the government's proposals to circumvent the Strasbourg ruling in S and Marper. But they also frame the debate in unusually clear terms, laying out precisely how it came about that Britain (or rather England) now has, per capita, by far the largest such database anywhere in the world.

Nothing to Hide, Nothing to Fear? makes plain just how piecemeal and uninformed by either principle or evidence the growth of the database has been. Set up almost 15 years ago - under pre-existing legislation passed before DNA fingerprinting was even thought of - it has grown by steady accretion without ever having been debated properly in Parliament. One consequence of this is an "ambivalence" about what the main purpose of the database actually is. There's a distinct lack of evidence as to its "forensic utility" - how efficient is it when it comes to solving crimes or eliminating the innocent - or whether the rapid extension of the database has helped, or hindered, its usefulness. There has been "little concerted opposition" to the growth of the DNA database, not necessarily because of widespread enthusiasm, but because there has been surprisingly little discussion of any kind.

The result has been "function creep", and the growth of the database has become an end in itself. And all this, the report repeatedly stresses, has come about with very little in the way of informed public discussion. There are, moreover, "significant concerns that have never been fully addressed".

No-one doubts that having a DNA database containing records of habitual criminals is a valuable tool for fingering the guilty. Even I wouldn't want to dismantle it entirely. Nevertheless, concentration on a few high-profile and thus unusual cases distorts the public's perception of the issues. Not least, it tends to obscure the fact that, by and large, the police used to manage perfectly well before DNA fingerprinting was ever thought of. It's far from obvious, in any event, that a large database, containing the details of everyone who has ever been arrested, is more effective than one narrowly focussed on those who have actually been convicted. It will help to solve some extra crimes, but the cost in storage and administration is far from negligible - and these resources might have been spent on other crime-detection or prevention initiatives. The argument for an arrest-based system only makes sense if you believe that there are a large number of lifestyle criminals who consistently manage to avoid being convicted. There's also an argument that the database is a deterrent, but that case is rarely made in public, and (as the report notes) there is as yet no evidence to support it.

Attachment to a mega-database distorts police practice and fundamentally alters the relationship between the citizen and the state. That this is so emerges clearly from this passage:


A retired senior police officer ... described his own early training, which emphasised the seriousness of depriving a citizen of their liberty by using the common law or statutory powers of arrest, and the preference, in the light of this, for alternative measures, such as reporting a suspect for summons by a magistrate, unless the offence was very serious or the suspect likely to abscond. Then he continued:

“It is obvious … that the system I have described no longer prevails. It is now the norm to arrest offenders for everything if there is a power to do so … It is apparently understood by serving police officers that one of the reasons, if not the reason, for the change in practice is so that the DNA of the offender can be obtained: samples can be obtained after arrest but not if there is a report for summons. It matters not, of course, whether the arrest leads to no action, a caution or a charge, because the DNA is kept on the database anyway.”


It's a remarkable statement, not just because of what it says about the DNA database, but of what it says about how things have changed in the past couple of decades. We are all suspects now.

But there are more subtle forces at work, too. The report suggests that "by shaping the context in which it is used, DNA-based forensic policing produces the conditions for establishing acceptance of its own legitimacy and for increasing the criminal justice system’s dependence upon it." A ratchet effect is introduced, making each subsequent extension more difficult to resist. And it has a distorting effect on other parts of the criminal justice system. "A challenging finding," it argues,

would be not merely that DNA analysis has entered the context of crime management and investigation but that its introduction has been able to shape the legal, operational and political context itself: the way that a decision to create a national DNA database can result in changes to police practice, to the likelihood and procedure of arrest, to decisions about which crimes are investigated, to the way crimes are committed and even to the sorts of crimes that are committed.


Pandora's Box, when opened, often turns out to contain Schrödinger's Cat.

In other words, the feedback between technological development, policing practice and ethico-legal acceptability creates the conditions for further developments that, as we become committed to them in turn, take us progressively further away from the alternative approaches that were equally possible at an earlier stage.

Liberty (the National Council for Civil Liberties) told us that it was "aware of anecdotal evidence that police may drop investigations if DNA evidence is not found at the crime scene."


A potential scandal, if true. A lot of the evidence, both for and against the NDNAD, is anecdotal. Hard evidence is hard to come by.

Here's an interesting quote from Dr Ruth McNally :

People whose profiles are on the database are the ‘pre-suspects’ … the first to be suspected (and eliminated) whenever a new crime scene profile is entered onto the database. In this respect they occupy a different space within the criminal justice system from the rest of the population; they are under greater surveillance and, with the advent of familial searching, this differential status can be extended to their relatives too.


I was quite alarmed by one of the recommendations in the report - that members of the police, and anyone else professionally involved in crime investigations, should have their DNA taken and stored on the database "as a condition of employment". This is presented as the police living up to the "nothing to hide, nothing to fear" soundbite they so often trot out in defence of whatever measure of surveillance or security they are justifying or seeking - and the reluctance of the police to take this step, noted by the authors, is certainly interesting. But the idea strikes me as profoundly dangerous. Firstly, it would represent a further extension of the database at a time when we should be trying to reduce it. Second, once the principle of DNA sampling as a condition of employment, it would surely be extended - initially to all those working with children and "vulnerable adults", or in other sensitive occupations (including, no doubt, private security). It would be the first step towards a universal database.

Worst of all, the move would undermine the idea - already under pressure, and more honoured in the breach - that the national DNA database is a tool of criminal investigation, rather than a vague and blanket monitoring of the population.

In a mildly entertaining passage towards the end, the report's authors try to anatomise the debate surrounding DNA. They identify five broad positions. "Forensic utilitarians" are mainly interested in the database's effectiveness - they are "evidence-based". "Negative libertarians" believe as a matter of principle that the database should be as circumscribed as possible. By contrast, "securitarians" (a great word) believe that the more powers the police have, the better. "Proportionalists" think in terms of a delicate balancing act between privacy and public protection, while "personalists" keep changing their mind. (Putting it more politely, the report notes that "this group recognises, and may themselves demonstrate, how abstract arguments and scientific evidence can be influenced by people’s beliefs and emotions.")

No prizes for guessing which group the Heresiarch belongs to.