back to article UK.gov urged to slash DNA retention plan

Government plans designed to bring the National DNA Database in line with human rights legislation have been criticised today by an influential group of MPs as not going far enough. The Commons Home Affairs Select Committee said that DNA profiles from those not convicted of a crime should only be retained for three years. …

COMMENTS

This topic is closed for new posts.
  1. Tom Chiverton 1
    Stop

    Or

    Or, better yet, just stop keeping innocent peoples it at all !

  2. Anonymous Coward
    Anonymous Coward

    Arrested for returning goods???

    "The Home Secretary cited an example of where DNA data should not be retained — that of someone arrested for shoplifting when trying to exchange goods for which she was carrying the receipt,"

    So you can be arrested, and illegally detained in Britain even if you have a store receipt? Your DNA taken, and the Home Secretary, suggests that only then should your DNA not be kept like the criminals you will become? And of course you are likely to be a criminal because you dared return goods for store credit! Unless you keep your nose clean for 5 years!

    What if you are arrested again for demanding to return faulty goods, does the 5 years count start again?

    What's the plan? Treat everyone as criminals, make their lives so miserable and so similar to the life of criminals as to make little difference whether they commit a crime or not?

    What if they DID steal that DVD? You arrest them, check their DNA against EVERY CRIME FOR WHICH DNA WAS EVER RECORDED and they're innocent. What then? You've shown that they have a lower probability of being a criminal than any random person grabbed off the street!

    You could grab an MP from Parliament and they would have a bigger probability of matching the DNA database than the shop lifter.

    Yet you decide to keep their DNA anyway. You've written them off, they are on to make it easier to arrest them in future, even if they're just returning goods for store credit. What difference does innocence or guilt matter? You can never reform any criminal, once an Expenses fiddler, ALWAYS an expenses fiddler.

    Once a liar to Parliament, always a liar to Parliament!

    Vote Labour, tough on crime, even imagined future ones by innocent people!

    Because you know how honest Jacqui Smith was!? Well Alan Johnson is so feeble and weak that he cannot reverse any of her bad decision! She left in shame, yet the choices she made he cannot change.

    Yeh, Vote Labour, I dare you.

    1. John Smith 19 Gold badge
      Happy

      AC@15:59

      "What's the plan? Treat everyone as criminals, make their lives so miserable and so similar to the life of criminals as to make little difference whether they commit a crime or not?"

      Yes that's pretty much the view of various senior civil servants and their Home Office sock puppets in the UK. The same POV that got the EU Data Retention Directive written (while the UK was in charge but passed when the Germans took over, so did not look so bad). The word of the Gulag Archipeligo spring to mind "We never make mistakes, they are *all* guilty of something. It's just a question of what that is and when they commit the crime," so might as well start the paperwork now, eh?

      Why else would they be having this discussion about *innocent* people. Not even *charged* in some cases.

      And let's not forget the permanently unleashed attack dog that is ACPO LTD. an institution with *no* liability, no controls and (apparently) quite a lot of power in setting the agenda.

      Note that 0.3%. That's < 1:330 cases where DNA is *involved*.

      I trust that clears up your confusion.

  3. EvilGav 1

    So . . .

    . . . we have no evidence to suggest that an innocent persons DNA has been used to solve another crime, ever.

    Yet we still want to retain it for several years, for people who haven't been convicted of a crime.

    How about you just fuck off and stop this bastard version of a democracy.

    1. Peter Gathercole Silver badge
      Flame

      Nothing to do with democracy

      We have a representational democracy in the UK. If Parliament vote to accept indefinite retention, there is nothing undemocratic about this.

      Unethical, immoral, non-productive, stupid and possibly illegal according to European law maybe, but undemocratic - never.

      This is even the case if the policy does not match the feeling of the majority of voters. The fact that you voted your MP into office (or didn't if you either did not vote, or supported one of the loosing candidates) means that you delegated to them the authority to vote on your behalf. Only problem is that they rarely do...

      If this really upsets you, lobby your MP, or vote for a party that is more sensible (if there is one!), or even if you feel strongly enough, stand for Parliament yourself!

      Democracy is basically agreeing to be governed by the majority. The problem is that without a referendum on every subject, you have to trust, rightly or wrongly, someone to act on your behalf.

      BTW. I hate the policy as well. If you are cleared, then my view is that the data should be deleted immediately. Maybe we ought to have a 'not proven' verdict like Scotland, which would allow DNA to be kept for a fixed period, but deleted for cleared suspects.

      1. Graham Marsden

        Re: Nothing to do with democracy

        You got the title right, our current system has *nothing* to do with democracy! As you say: "you delegated to them the authority to vote on your behalf. Only problem is that they rarely do..."

        MPs are supposed to *represent* the views of their constituents in Parliament, unfortunately what most of them tend to do is represent the views of their Party Leaders and Whips to their constituents and basically say "that's how it is, like it or lump it".

        Cameron and his ilk are currently campaigning on an effective slogan of "Vote for us, because we're not them" and telling people that voting for anyone else is a vote for another 5 years of Gordon Brown which sounds like desperation coupled with a lack of anything that makes them actually worth electing.

        IMO the only hope we have of getting any representation back into our electoral process is to tactically vote for a Hung Parliament so one party doesn't have a "majority" (even when they only got a minority of the vote) giving the ability to force through whatever legislation they like.

        1. CD001

          EVERYBODY VOTE

          And have everybody write "no confidence" (or "I have no confidence in the ability of any of these morons to run run a country" - a literal vote of no confidence - *b'dum ching*) across the whole ballot paper - unlike not voting, spoiled votes still have to be counted.

          1. Anonymous Coward
            Anonymous Coward

            So?

            And what do you suggest next?

  4. Anonymous Coward
    Stop

    All elected representatives need to be on this

    It should be part of the job. Anything that is forced on the oiks should be done to those who suggest it.

  5. Jonathan

    the birthday paradox

    don't know if it was another (response to another) article on here or somewhere else, but somebody mentioned the possibly applicability of the "birthday paradox" (in a room of 23 people there is 50% chance that one of them will share my birthday and a 99% chance if there are 57 people) to the DNA database.

    if there are a few tens (maybe even hundreds) of thousands of people in there the chances of there being a false match amongst then are slim, but if there are 60 milllion people in there, it could be almost guaranteed.

    the maths is fine but the knowledge of whats involved in the DNA fingerprinting is beyond me...

    anyone knowledgeable out there?

    1. Anonymous Coward
      Boffin

      Re: The birthday paradox

      The quoted error rate for matching full profile (quite often you get partial match) is 1: 1000000000 which is 10^-9).

      So assuming that we've chosen a totally random person from UK population (which I'm assuming is 60 million; given that database size is so huge and expanding the 'random' part of the assumption will very soon be justified in real life), and using Bayes theorem, the probability that our person is the guy/gal we are looking for is circa 95% if they got a match.

      This means that in 5% of successful matches, they got it wrong.

      Assuming that number of successful matches to date is close to that of the US CODIS database (which has similar size), i.e. 50 000, in 2500 cases the database search has given a match to a wrong guy.

      This is a rough estimate - and given that people are fairly likely to re-offend, it will be lower for people who were convicted.

    2. Anonymous Coward
      Anonymous Coward

      Especially...

      ... as the chances against a mismatch so often touted in in discussions about DNA testing are mathematically-dyslexic rubbish - in forensic terms at least. The chances of a false match are nowhere bear the millions-to-one so often quoted. I don't know if you intended the birthday paradox as humour, but it's pretty near the knuckle, believe me.

      DNA retention isn't only an assault on our civil liberties. In the hands of politicians, bureaucrats and police, to whom an AA battery is indistinguishable from magic, it's simply Bad Science.

    3. John Smith 19 Gold badge
      Unhappy

      @Jonathan

      "don't know if it was another (response to another) article on here or somewhere else, but somebody mentioned the possibly applicability of the "birthday paradox" (in a room of 23 people there is 50% chance that one of them will share my birthday and a 99% chance if there are 57 people) to the DNA database."

      Quite relevant as it happens.

      Your DNA is *not* stored in total and samples are taken per *crime* (hence some people have multiple samples held against them.

      IIRC The exact location of some highly variable (more-or-less unique) parts of the genome. others have posted that this is like a hash function. While DNA is unique (except to identical twins) the hash function i subject to collision. The higher the number of entries the more likely there *will* be a hit. So when the database throws up (say) 100 matches what will the Police do?

      1) Work through the list of matches eliminating all unreasonable matches and concentrate on ones known to the victim.

      2) Find the first one that matches in the local neighborhood and has no alibi.

      We all *know* which one of those options we would prefer, but which one is more likely?

    4. Anonymous Coward
      Boffin

      re : birthday "paradox"

      "in a room of 23 people there is 50% chance that one of them will share my birthday and a 99% chance if there are 57 people"

      Wrong.

      In a room of 23 people there is an approximate chance of 22/365 that one of them will share YOUR birthday (not allowing for 29th Feb), because you are using a specific date.

      In a room of 23 people there is a roughly 50% chance that two of them will share the same birthday (of ANY date) and a 99% chance if there are 57 people.

  6. Glyn 2
    Boffin

    Titles are for other people

    Have any of the other political parties said they'll delete the DNA as soon as you're proved innocent (note, not found guilty, but proved innocent )????

    1. Intractable Potsherd
      Thumb Up

      I generally agree ...

      ... but the way it used to work in this country (and should definitely be restated and reinforced) is that no-one has to be "proved innocent". A person was (until recently) presumed innocent unless proven *guilty*. The burden of proof lies on the accuser, not the accused. So, to restate your excellent question - has any party said they will reinstate the rule of law, and put the burden of proof back on the prosecution? The answer, as far as I am aware, is no.

    2. Anonymous Coward
      Anonymous Coward

      Eh?

      When you said "proved innocent" I presume you meant "found not guilty". You cannot be proved innocent under our legal system.

      The courts' job is to prove your guilt. In that context prove means "test". It does not mean "demonstrate beyond reasonable doubt".

  7. Anonymous Coward
    Anonymous Coward

    contradict

    1) there is no evidence of innocent people's DNA solving crimes and only 0.3 crimes are solved by this DNA farce anyway.

    YET

    2) The benefit to public safety outweighs the impact on privacy.

    So how does something for which we have no measure outweigh anything?

    1. Anonymous Coward
      Anonymous Coward

      Re: contradict

      If taking something back to a shop for an exchange, whilst in possession of a receipt, is an arrestable offence in this country, then what makes you think that we have any right to question anything our overlords decide to tell us?

      Just put them out of jobs at the election (assuming there is one!), then boycott any company that offers them a directorship or consultancy - especially if it's a media distributor.

    2. Anonymous Coward
      Anonymous Coward

      Because...

      ...you don't count, and the people who seek to control every aspect of your life do...

      thimple...

      And if you think an election is going to solve a damned thing in Banana Republic UK Plc...

  8. Anonymous Coward
    WTF?

    Ah for crying out loud!

    "The Commons Home Affairs Select Committee said that DNA profiles from those not convicted of a crime should only be retained for three years."

    Keeping innocent's DNA will NOT, I repeat, NOT help solve cases for fuck's sake!

    Innocent people's DNA should not be hold 3 years, but never stored in the first place!

    What do these people have floating in their heads where normal clear-thinking people have brains?

    1. Intractable Potsherd
      Thumb Up

      Excellent question!

      "What do these people have floating in their heads where normal clear-thinking people have brains?" I can't be sure, but I suspect it is brown, rather than grey (and probably smells like a cowshed).

  9. Doctor_Wibble

    re: birthday paradox

    Someone mentioned this a long long time ago and I ended up having a go at the calculations but I don't know if these are the ones being referred to...

    But in any case, here's what I came up with, using the Birthday Paradox formulas, and 'bc' to get some absurdly long numbers in there. The biggest assumption is that the loci chosen for the profiles are truly random, and evenly distributed across the entire range of possible values - possibly unwise because nature has a knack of not doing that sort of handy thing.

    On the basis that we are told that the chance of a coincidental match is 'one in a billion', a set of 117540 random samples would give us a probability of 99.9pct that two samples are (apparently) the same (as opposed to being the same sample twice). For a 99.999 percent chance of two samples being the same you need a set of at least 151743. Here 'being the same' means indistinguishable given the criteria of 'one in a billion'.

    I do wonder how this reflects on the actual database and how many 'replicates' (allegedly same person more than once, different names etc) are actually coincidental matches.

    The other 'wrong match probability' figures I did were for a full-population database of 50 million, and came out with a result that with a discrimination of '1 in a billion', you are looking at a wrong-match probability of 5percent on *every* search you make. This for the case where you are looking for a matches to samples you already have.

    E&OE. If anyone can refute the numbers, please be gentle and show your working...

    I dare say I could probably show mine (matron) at some point.

    Oh and there's also the article somewhere about an FBI(?) database showing a lot more coincidental matches than should have been possible...(?)

  10. Anonymous Coward
    Anonymous Coward

    Fallacious reasoning

    "The Committee rejected this option, however, judging that the benefit to public safety of retention for three years outweighed the impact on individual privacy"

    I'm sorry, but this I think that this is almost never true.

  11. I Like Heckling Silver badge

    It's Not Accurate Anyway.

    Fingerprints are not 100% accurate... never have been, but it's only in the last 20yrs that this has become more apparent as the database grows, so does the false positives being recorded.

    The exact same thing applies to the DNA database, the larger it becomes the more likely a false positive becomes. They could end up shooting themselves in the foot when DNA evidence is challenged and found to be false... could render DNA evidence useless... at least as useless as the coppers who only target easy arrests for non crimes.

    But try getting them to deal with the scrote next door who's been driving a car with no tax or mot for 12 months and suddenly they are no where to be seen... But god forbid I forget to put my seatbelt on when returning from the garage to my home which is all of 500 yards.

    1. John Smith 19 Gold badge
      Happy

      @I Like Heckling

      "They could end up shooting themselves in the foot when DNA evidence is challenged and found to be false"

      I think it's your head you should be more concerned about being shot in.

  12. JohnG

    Halving the duration turns wrong into right

    Recommendation: We'll only breach our citizens' human rights for three years instead of six years, in contravention of a treaty to which we are bound.

    So, by the same token, would it be OK if I evaded income tax for only three years instead of six years? The reduced duration of law breaking would make everything OK would it?

  13. David Ramsay

    How about ...

    Refusing to allow DNA to be taken.

    When forced ensure that you have a note of all officers numbers involved.

    Employ a very expensive barrister on Legal Aid and sue them individually for assault in the High Court through to the European Court.

    Ensure that you refuse to pay any costs/fines

    Go to jail rather than pay up if you are asked to contribute/fined

    If everyone was to do this then the whole thing would fail!

    Remember Laws are (or should be) in place with the agreement of the populace and not the other way round.

    1. Anonymous Coward
      Anonymous Coward

      Policing by consent is it?

      It's not about laws being with the agreement of the populace. Were that the case there would probably be no speeding laws, no smoking ban, indeed there are all sorts of laws that are very unpopular, but necessary.

      What is important is policing by consent, unfortunately this government have done more than any previous government (yes, even Thatcher's government) to undermine the public trust in the Police. Nulabour may believe that what they are doing (effectively creating a police state) will reduce crime, but it will not. Creating a such a system will make criminals, albeit reluctant criminals, of most of the population.

  14. Anonymous Coward
    Anonymous Coward

    Here's the thing...

    Strictly speaking DNA testing can only be used to eliminate a person. This was common knowledge only a few years ago, but this government are pushing the database as a means to match people with crimes. That isn't really the way DNA testing works.

    The way crime detection should work is that having collected all their evidence and identified their suspect(s) the police should use DNA testing as a final clincher. The way the government want it to work is that if DNA evidence is available they check it against the database and then go out and collect further evidence to prove the vict^H^H^H^H suspect guilty. That way they'll avoid finding any evidence that casts doubt on a subject's guilt. Oh, and they'll save a fortune on on old fashioned police legwork.

    Here's a simple scenario where DNA evidence would be completely wrong. Imagine your house has been burgled and a baseball cap has been found in your bedroom. The cap has some hair attached, complete with roots. DNA tests on the hair match a youth who has a record for burglary. An easy collar, right? Wrong. The youth left the hat in a pub one night and the real burglar picked it up, complete with hair, and wore it on the day he comitted the burglary.

    Does anybody think that in the scenario above the police would make much effort to look beyong the DNA evidence? Like fuck they would.

    The whole concept of the DNA database is fatally flawed, even if only the DNA of convicted criminals is retained.

  15. John Smith 19 Gold badge
    Joke

    AC@22:50

    "Nulabour may believe that what they are doing (effectively creating a police state) will reduce crime, but it will not. Creating a such a system will make criminals, albeit reluctant criminals, of most of the population."

    Your reluctance to cooperate in becoming a criminal, and thereby demonstrating the efficiency of the Party's detection policy, has been noted.

  16. David Pollard

    Where's the benefit?

    It might be appropriate to claim a benefit from the 0.3 per cent of cases matched, or the total of 23 serious crimes solved per year, if this were the end of the story.

    But, first of all, there seems to have been no estimate of how many of these crimes would have been solved in more traditional ways had they not showed up on the NDNAD.

    It's clear also that despite a significant increase in the number of profiles recorded, both guilty and innocent, the number of matches isn't increasing. The number of cold cases must be declining as time passes and some criminals at least are presumably now becoming adept at not leaving or disguising DNA traces.

    Crucially, though, the statistics entirely fail to take into account the disbenefits of the DNA database. Without doubt the heavy-handedness of its introduction and use will have reduced public co-operation; arguably the single greatest factor in tackling crime. How many criminals were not caught because of this it is difficult to know. But the number not detected because of public distrust could easily exceed the score that were detected without needing any additional effort because they showed a match on the NDNAD.

  17. John Smith 19 Gold badge
    Happy

    Headline should read

    Power crazed maniacs who thought this up are encouraged to slash their own wrists and donate a litre sized sample to the database*

    *This is not glorifying terror, merely encouraging them to demonstrate that if they have nothing to hide, they have nothing to fear. Although I suspect quite a few people would be more than happy to assist them if they find themselves getting a bit squemish. I'm sure it's nothing a little band aid would not stop although IANAD.

  18. John H Woods Silver badge
    IT Angle

    DNA partial hash

    Warning: hurried and possibly poorly thought out argument follows.

    Why not just store a partial hash of EVERYONE's DNA? Hashy enough to ensure about 50-100 duplicates for any given sample.

    No stigma about being on the database ... because everybody is. But the data is hashed enough, with enough duplicates, to prevent efficent use of the database for tracking or other privacy violations.

    When a sample from a crime scene does get matches you can use REAL POLICE WORK to find the culprit, as you only have a few score of suspects. But you have too many suspects to just nail one of them on the grounds that 'this person is a match' so you need real evidence. Also you might have a quick way of eliminating people from suspicion.

  19. John Smith 19 Gold badge
    Happy

    @John H Woods

    "Why not just store a partial hash of EVERYONE's DNA? Hashy enough to ensure about 50-100 duplicates for any given sample."

    They *do* store a partial hash of DNA at present.

    "When a sample from a crime scene does get matches you can use REAL POLICE WORK to find the culprit, as you only have a few score of suspects. "

    Which I suspect is an order of magnitude above what they have to investigate now.

    Despite numerous cop shows in *reality* they will look for the 1st candidate who is a) on the database b) in the area and c) has no or little alibi. And remember if they don't get a conviction ("But I'm 110% convinced he did it"*) they will re-try the case as the UK scrapped double jeopardy following the Stephen Lawrence retrial (with a little help from the best DS that money could buy).

    They would not *have* to eliminate you from their investigation if you were *not* on the database to begin with.

    Thanks for the warning. You were quite correct.

  20. Jamie Kitson

    Yeah But

    Doesn't this sort of thing make you stop and think?

    http://news.bbc.co.uk/1/hi/england/hampshire/8574507.stm

This topic is closed for new posts.