One of the most eye-catching aspects of David Cameron’s speech on criminal justice yesterday was the suggestion that GPS technology could be used to track offenders. As someone who, during his time as a Civil Servant was actively involved in advising on the use of GPS technologies and has quite a bit of experience in the field I found myself unconvinced by aspects of this proposal. I’m not convinced that it is workable – yet – and I wonder whether No 10 has really sought good technical advice on this.
Politicians and the media in Britain do not, as a whole, exhibit a great degree of scientific or technical literacy. It means they are easily suckered by people selling technical solutions that give the appearance of being plausible; a lot of people in Whitehall spend quite a lot of their time telling Ministers that the latest technical wheeze being offered by manufacturers, however impressive it may look in a small-scale trial, is unsupported by any real evidence that it could be rolled out on a large scale in the messy real world
GPS is a mature technology that has a huge range of real-world applications. Obviously it’s crucial to navigation systems but most smartphones contain a GPS chip, as do all ATMs (for date and time stamping rather than navigation). The signs at bus stops telling you that the next 5A will arrive in 5 minutes use GPS technology. But many commentators show themselves to be clearly clueless about how it works – witness this rather glorious piece of bad science from the Daily Telegraph following yesterday’s announcement. Put at its simplest, satellites do not track and cannot track anyone or anything; the same fallacy dominated the public debate over road pricing launched by the last Labour government. Within Whitehall, Ministers and Special Advisers are notoriously susceptible to this kind of nonsense. A report in the Guardian today tends to confirm the suspicion that the work behind Cameron’s speech is worryingly light on evidence.
GPS works by taking location fixes from satellites. It requires a clear sky view of at least three satellites to do this. In order to track an offender, you would need to generate a series of such fixes, and pass that information to the control centre. Obviously, the more detailed the tracking, the more data you would need to store on the tag and the more frequently you would need to relay that information to the control centre. And of course the tag would need a power source; GPS applications are notoriously battery-draining and the tag would need frequent charging.
Now obviously there is nothing here that couldn’t be done. The most technically problematic part is generating fixes; GPS is notoriously susceptible to difficulties in streets with tall buildings, or where plate glass or water causes reflections and hence the possibility for false signals. Now in the real world you can get round this either by map-matching (i.e. snapping the reads to digital mapping data stored in the tag) or simply discounting the odd outlying fix and using algorithms to approximate the missing data. But in a world in which you are seeking to establish a criminal burden of proof, how far can you do this? Equipment like safety cameras requires rigorous levels of type-approval before its output is admissible in court. What level of locational accuracy would be required to sustain a prosecution of an offender who had broken the terms, say, of his bail? And how much would it cost to produce a type-approved product that could demonstrate that an offender was in a particular place at a given time, beyond reasonable doubt? Would you need corroborating evidence from eyewitnesses or cameras?
GPS does not work underground (obviously). Would London offenders be barred from using the Tube (which could be pretty counter-productive if one object of the scheme is to allow offenders to get back into jobs)?
And GPS can be spoofed, relatively cheaply and easily. How much security would be built into an offender tagging system? At what cost?
And what about data transfer? How do you get data from the tag to the control room? The obvious answer is through mobile telephony, but to claim that you can pinpoint where an offender is at any given time requires real-time updating, and if tagging is to operate on any kind of scale you need to either transfer huge quantities of data in real time – and that can become seriously expensive (especially compared with the option of storing the information on the tag – memory is becoming very cheap – and transferring the data en masse when the networks are quiet).
In theory, there is nothing here that could not be overcome. But it comes with a price-tag and there is the world of difference between running a successful pilot scheme and delivering something that is robust and cost-effective in the real world. And – given that evidenced policy-making has not so far been the coalition’s strong point, I wonder just how far Special Advisers – for whom evidence is often the word of the last industry lobbyist who has bent their ear – are pushing this plan beyond its evidential base, and are relying on sales pitches from the private sector instead. I’d guess there are plenty of private sector providers who are very keen to get their foot in the door to get what could be some very lucrative contracts, and who are confident that they could lay off the risks on to the public sector.
It’s important, because in this case tagging seems to be the quid pro quo for moving towards a justice system which is based on rehabilitation. I would argue for an approach based on rehabilitation and restorative justice; all the evidence I am aware of suggests that it is far more effective and way cheaper than prison, especially for less serious offences. But I am concerned that a glib throwaway remark from a Prime Minister playing that old political trick of appearing to be tough on crime when things are not going well could lead to an expensive disaster, which could end up being used to discredit a move away from custodial sentencing. And I have no confidence that either politicians or the media will give the idea of tagging the scrutiny it needs, or indeed that they are remotely capable of conducting a nuanced and evidenced debate around a really difficult issue.