Home Affairs Committee on the case of the National DNA Database
The inquiry into the National DNA Database by the Home Affairs Committee of the House of Commons, which was published on Monday, comes after much of the ‘inquiring’ - including by Radio 4's More or Less programme in August 2009 - has already been done.
But the Home Affairs Committee provides a most useful and timely summary ahead of the Crime and Security Bill’s return to the Commons for its Report stage. In particular, the Committee was concerned (para 14) that no research had been done on how many crimes have been solved with the help of stored personal profiles of those not previously convicted of a crime; and that a few serious case studies were recycled, not all of which fitted the bill.
On 19 November 2009, the Home Secretary introduced the government’s proposed changes to the system for retaining DNA samples and profiles. The Home Affairs Committee notes:
- that the Home Secretary had not complied with the Committee’s request to provide details of 23 pertinent case studies to which he referred in his speech to Parliament (para 14).
- DNA-detections do not appear to have increased commensurately with the number of subject-profiles held on the National DNA Database - proper analysis is lacking on this, a key issue (para 7 & 8).
- Time to re-arrest is not a suitable proxy outcome for time to conviction when considering the retention time for ‘innocent’ DNA-profiles (para 26 -28; 34)
- Criticisms of Home Office’s extrapolation from an adhoc analysis of time to re-arrest for persons whose ‘innocent’ DNA-profile has been retained (para 35).
- A possible distinction between ‘arrest-never-charged’ and ‘arrested-charged-never-convicted’.
There are even wider limitations (see previous Straight Statistics articles) to the 19 November 2009 analysis than even the Home Affairs Committee has noted.
I had hoped that, sensible of this, the Home Secretary was preparing a third, proper analysis: of time to conviction, with stratification by the sex and age-group at retention of those whose ‘innocent’ DNA-profiles had been retained, for comparison with the conviction rate of the general population belonging to the same age groups.
Finally, since the Home Affairs Committee recommends a maximum retention period of three years – both by analogy with Scotland and to avoid extrapolation beyond actual data – the case for re-analysis is stronger still. Using database linkage, analysts within the Home Office and Ministry of Justice should be able to offer a life-table analysis of time to conviction within three to four years of retention of ‘innocent’ DNA-profiles. Analysts rightly point to waiting times between arrest, charge, and disposal by the courts.
Anonymous (not verified) wrote,
Wed, 10/03/2010 - 22:50
"We are a campaign established by journalists and statisticians to improve the understanding and use of statistics by government,politicians, companies, advertisers and the mass media. By exposing bad practice and rewarding good, we aim to restore public confidence in statistics."
Your article above appears to be a personal commentary on what analysis you 'hope' for, rather than any objective or considered critique [or praise] of what has been provided so far - whether there is any merit as well as any downsides, vased on any objective criteria. If you had asked you might have found out how the much cited Scottish analysis was produced - for all we know it was on the back of an envelope (or more liely a postage stamp) ? ; and whether the analyses you suggest are theoretical ideals or actually practically possible given availlable data sources.
Unfortunately what comes across is not a scientific independent and objective critique of anybody's statistical analysis but instead a high minded egotistical and blinkered academic ivory tower perspective that does you and the idea of Straight Statistics no credit at all :) Please try harder next time.