Wed, 06 Jan 2010

Sorry Affairs Committee

On Tuesday, the Home Affairs Committee was taking evidence on the National DNA Database (NDNAD). Unfortunately, this was a sorry show of how politicians work. The committee members showed an poor understanding of the NDNAD and of the Crime and Security bill. This was exemplified even before the start of the proceedings in the one page leaflet present on the chairs in the room where the evidence session was held at Portcullis House:

Today's evidence session–The National DNA Database

This is a one-off evidence session. The purpose of the session is to examine the allegations of over-representation of young black people on the National DNA Database, and in particular to inquire why the police think it necessary to retain samples [sic] from those never charged or subsequently found not guilty of any offence.

The proposed plan is to destroy DNA samples after DNA profiles have been produced and no later than six months after they were taken. The debate is about the retention of DNA profiles of innocents. The DNA samples are the buccal swabs (or sometime hair) taken at arrest containing the full genome of individuals. The DNA profiles are a series of 20 numbers produced from the analysis of DNA patterns (from what is today considered non-coding regions of the genome) of one of the sample. DNA samples are kept in fridges in the forensic labs contracted to do the analysis, while DNA profiles are held in the National DNA Database. For more on DNA profiles see Information retained in the National DNA Database profile records. The prompt destruction of the DNA samples is the one measure that everyone, but the forensic labs (who are paid to maintain the freezers), is happy with.

This confusion between samples and profiles was common, however it is rather inoffensive as one can usually guess what is it that is talked about solely from the context. A much more embarrassing confusion was between the loading of DNA profiles on the NDNAD and the retention of these profiles. A clear understanding of the difference between the two is essential to review the proposed legislation. The government has helped create this confusion by regularly highlighting examples of horrific crimes solved where a criminal was identified when his DNA profile matched, just after it was loaded on the NDNAD, that of an earlier crime scene, and then using that anecdote as a justification for the retention of DNA profiles. When matches happen as the DNA profile is just loaded on the NDNAD, retention is irrelevant.

For instance, Mark Dixie, convicted for the murder of Sally Ann Bowman, was identified when, following his arrest for a pub brawl, his DNA profile was uploaded to the NDNAD and matched that recovered from the murder scene. This case can be used to justify the taking of DNA at arrest, but is irrelevant in a debate about retention of DNA profiles. (Ironically, this case also illustrates that retention of the DNA profile of an innocent, Kevin Reynolds, was not used to eliminate him immediately from the investigation.) During the evidence session, David Winnick MP mentioned the case of Steve Wright several times. What no-one explained, unfortunately, is that this case didn't rely on DNA evidence at all to identify this murderer. Steve Wright was linked to the five Ipswich prostitute murders during the investigation, not by his DNA, but by his car figuring in CCTV footage in proximity to several victims. Liberty has debunked the reporting of some of these cases, including those of Dixie and Wright.

The NDNAD Annual Report 2007-09 explains that:

The average match rates between crime scene and subject profiles when:

• a crime scene profile was loaded to the NDNAD during 2008/09 was 58.7%
• a subject profile was loaded to the NDNAD during 2008/09 was 2.3%.

In most cases, what is useful is the check against unsolved crimes for which DNA has been recovered when the individual's DNA profile is loaded, the further retention of the DNA profile is of little help. In a recent post, Home Office gets DNA database funding priorities wrong, I explored this further.

Leaks or stitch-up? 0.67% or 0.36%?

After a question is answered, Keith Vaz MP, the Committee's chairman turns to the MP whose question is next. Chief Constable Chris Sims, the Association of Chief Police Officers' (ACPO) lead on the NDNAD revealed that he had received the list of questions he was going to be asked, the previous day:

• Chis Sims: Sorry, we're jumping to the last question?
• Keith Vaz: Have you had a list of questions?
• Chis Sims: I have, yes, is that wrong?
• Keith Vaz: It is unusual.
• Chis Sims: Oh, I do apologise.
• Keith Vaz: How did you get them?
• Chis Sims: It's arrived in my email system.
• Keith Vaz: Right, how extraordinary!
• Chis Sims: How extraordinary!
• Keith Vaz: Do you know who sent them to you?
• Chis Sims: I could find that out, I've no doubt.

Keith Vaz MP appeared shocked. Not being familiar with the working of these committees, it is unclear whether his apparent surprise at this unusual revelation was genuine or for the records. Without Chris Sims' blunder, I would not have even known the questions were prepared before the session, especially from the apparent lack of preparation of the MPs on this topic. From their answers, it was obvious that Diane Abbott MP and Isabella Sankey, Policy Director, Liberty who were first to give evidence, had not been given sight of the questions.

Some MPs were trying to find out how useful the NDNAD is and whether money spent on it could be more useful for other policing activities. This lead Chris Sims to state that 33,000 of the 4.9 million crimes the police recorded last year were solved solely or largely because of the DNA database. At Keith Vaz MP's request, committee's staff helpfully calculated that this represented 0.67% of the recorded crimes – negligible was the general reaction. Chris Sims, helped by Gary Pugh, Chair, NDNAD Strategy Board, added that DNA matches play a much more important role in solving certain type crimes. That the percentage was around 40% of for burglaries. These figures have been widely reported, without any questioning, by the press.

Based on the official data on recorded crime (Table 2.04 of Crime in England and Wales 2009/09) and on the NDNAD (Tables 2 and 3 of NDNAD Annual Report 2007-09), it is easy to find that these numbers don't add up. The mention of 4.9 million recorded crimes shows that Chris Sims was referring to 2007/08. As data is also available for 2008/09, I've made the calculations for both years! The second number mentioned, 33,000 crimes solved as a result of a DNA match, doesn't seem to be correct [see update below]. The number of matches for that year was slightly higher, however number of matches are not very interesting in this context. Matches 'include cases where the individual had an innocent reason for being at the crime scene and cases where it was not possible to take the investigation forward'. More interesting is the number of detections where a DNA match was available. In these cases, the crimes has been solved. Even then, it does not mean that it is the DNA match that helped solved the crime. As can be seen from the tables below, even when you look at the much larger, but not much relevant, number of matches, it is nowhere near 40%.

Update: An attentive reader pointed out that '[t]he 0.67% is because they have included "indirect" detections (e.g. extra crimes that the suspect confesses to after a DNA detection with the first crime). The main problem with this figure is it is NOT the number of detections made due to their being a database, only a minority of these detections would disappear if there was no database of individuals at all (let alone if you only took innocent people off).' The figure for total DNA-related detection – DNA detections and additional detections – was 33,034 (0.67% of recorded crimes) in 2007/08 and 31,915 (0.68% of recorded crimes) in 2008/09. (The NDNAD Annual report doesn't even bother to break down these numbers by types of crime).

What the committee should have found out by itself when preparing for this evidence session, what Chris Sims should have answered, and what journalists should have written, is that 17,463 crimes (0.37%) of the 4.7 million crimes the police recorded in 2009/09 were solved in which a DNA match was available.

GeneWatch UK had been highlighting this percentage for a long time and found that '[s]ince April 2003 [to 2006], about 1.5 million extra people have been added to the Database, but the chances of detecting a crime using DNA [as a percentage of recorded crimes] has remained constant, at about 0.36%.' As explained in an earlier post in more details, this data shows that retaining DNA profiles of individuals indiscriminately does not help solve crime.

2007/08	Recorded crimes	Crimes with a DNA scene-subject match		Detections of crimes in which a DNA match was available
Homicide	784	363	46.30%	83	10.59%
Other Violent Offences	960,391	1,766	0.18%	849	0.09%
Rape	12,639	540	4.27%	184	1.46%
Other Sex Offences	40,838	163	0.40%	64	0.16%
Robbery	84,751	1,432	1.69%	617	0.73%
Domestic Burglary	280,708	8,043	2.87%	3,443	1.23%
Other Burglary	303,006	7,211	2.38%	3,886	1.28%
Theft from Vehicle	432,387	3,544	0.82%	2,201	0.51%
Theft of Vehicle	170,016	4,223	2.48%	1,379	0.81%
Criminal Damage	1,036,236	5,432	0.52%	3,180	0.31%
Drugs Offences	229,903	1,000	0.43%	321	0.14%
All other recorded crime	1,399,845	3,659	0.26%	1,407	0.10%
Total of 12 crime types	4,951,504	37,376	0.75%	17,614	0.36%

2008/09	Recorded crimes	Crimes with a DNA scene-subject match		Detections of crimes in which a DNA match was available
Homicide	648	252	38.89%	70	10.80%
Other Violent Offences	903,345	1,819	0.20%	861	0.10%
Rape	13,133	580	4.42%	168	1.28%
Other Sex Offences	38,355	175	0.46%	106	0.28%
Robbery	80,104	1,462	1.83%	603	0.75%
Domestic Burglary	284,445	8,188	2.88%	3,702	1.30%
Other Burglary	296,952	7,110	2.39%	3,830	1.29%
Theft from Vehicle	396,990	3,484	0.88%	2,036	0.51%
Theft of Vehicle	147,470	3,699	2.51%	1,298	0.88%
Criminal Damage	936,729	5,149	0.55%	2,886	0.31%
Drugs Offences	242,907	1,110	0.46%	397	0.16%
All other recorded crime	1,362,736	3,699	0.27%	1,506	0.11%
Total of 12 crime types	4,703,814	36,727	0.78%	17,463	0.37%

Minister forgets his bill

Could it get any worse? It did when Alan Campbell MP, Parliamentary Under-Secretary of State for Crime Reduction, Home Office got the proposals, in the Crime and Security bill concerning children who have been convicted, wrong. He failed to remember them! None of the committee members came to his rescue so they probably didn't have a copy of the bill, or of the much shorter standard note published by the House of Commons Library (SN/HA/4049 Retention of fingerprints and DNA data). Eventually, several of the minister's minders jumped up from their seats to provide him with this basic and essential information (see picture). If you need to refresh your memory see the preceding post: Home Office still wants your DNA profile, and your PNC record.

Alan Campbell was very clear about his willingness to get rid of the presumption of innocence. He sees three distinct categories of people: those who are guilty, those who are convicted and those who are arrested but not convicted. Having their DNA profile retained for a certain number of years is a price those in this latter group have to pay as 'our research says [...] they are more likely to offend in the next six years'. The minister remains convinced that this bill will be completed before the next general election. He promised that the second reading of the Crime and Security Bill would happen very soon, but wouldn't commit as to whether it'll happen this month or when a date will be set.

Just before the end, there was a rather surreal moment. David TC Davies, Conservative MP (not to be confused with David Davis MP) expressed that his views are 'fairly close to those of the minister on this' (less than a week after his party launched a 'Return my DNA' campaign!). He went on to suggest, 'as a fellow supporter, that the minister would 'allow people to opt-in to get on this database.' He also attempted to get Alan Campbell to agree with him that, 'if there's an issue that young black men are over-represented [on the NDNAD], surely that may be because they're committing more crime?' The minister didn't 'accept this rather simplistic account'. (Home Office research indicates 'people from BME groups are over-represented at each stage of the criminal justice process from initial contact to sentencing. Evidence also suggests that it is not because people from BME groups are more likely to offend.')

Not having much experience of evidence sessions and as the leaflet introducing this session stated this was 'a one-off evidence session', I can only hope that regular sessions are conducted in a more professional manner. You can watch the session now on Parliament TV; uncorrected transcripts will eventually be published on this page. Even though, this was a one-off session, Keith Vaz MP announced that there would be another session next week with Sir Alec Jeffreys and possibly another mystery guest.

First published on 2010-01-06; last updated on 2010-01-07.

---