Crime and Security Bill

Well, we shall come to the judgment on timing. I was just about to turn to retention, which is central to the debate, and the balance to be struck in terms of the presumption of innocence until proven guilty. I appreciate that the Minister and the Government take the reverse approach of wishing to retain data for as long as possible. That has been their historical approach, and they have made their position clear. We approach the issue from the alternative direction, saying that DNA should be retained for the minimum period that it is safe to do so. That is why we have drawn a distinction: we do not simply say that all DNA profiles of anyone arrested for an offence but not convicted should be removed; we strike that balance—that proportionate balance, as we see it—by having a period during which the DNA of those arrested for an offence but not necessarily charged or convicted should be retained. There is a real difference between us on the retention of profiles, and the Minister alluded to it. The Government argue that, when someone has been acquitted of any wrongdoing, they should treat them as a future or potential suspect in a criminal investigation. Such people are supposed to be innocent in the eyes of the law, but the Government are suggesting a blanket period of six years, regardless of the nature of the offence, the crime that may or may not have been committed or whether anyone was found guilty. The Government have sought to justify their position on two main grounds. Their "hazard rate analysis", as they call it, compares the likelihood of re-arrest within a specified period following an initial arrest with the general propensity for arrest in the general population. The "arrest to arrest hazard curve" is the analysis on which the Government have sought to pin their six-year period. However, by doing so the Government conflate the fundamental difference between arrest and conviction. They use re-arrest as a proxy for having committed an offence, even though a second arrest may be as unsound as the first. They also fail to take account of what has been termed "the confirmation bias by investigating officers"—the fact that, once a person is known to the police, they are more likely to be considered a suspect when future offences are investigated. A bias in the statistics is created as a consequence. The Government acknowledge that the analysis is based on extrapolated, rather than real, data, because they had only a three-year window to consider the data that were available. Therefore, estimated data on the general population was used, the extrapolated risk curve is quite sensitive to errors, and the lines are measured with some degree of uncertainty. It is highly questionable whether this latest Home Office-produced research offers much more than the previous, incomplete research produced by the Jill Dando Institute of Crime Science, which the Home Office published in support of its previous policy of retaining DNA for 12 years. As Gloria Laycock, the director of the Jill Dando Institute, acknowledged, it was a mistake to publish that incomplete research, which was based on data to which it did not even have direct access. She said:""That was probably a mistake with hindsight. We should have just said 'you might as well just stick your finger in the air and think of a number'."" It is also interesting to note that the Information Commissioner has formed a different view from that of the Government on the interpretation of the hazard rate analysis data, stating that he""remains concerned that the way this evidence is being interpreted at present does not provide an appropriate basis for the proposed retention periods."" He suggests that a shorter retention period for those not convicted of any crime should be adopted than that advocated by the Government:""The Commissioner does not consider that the evidence presented supports a general retention period of anything like six years.""