KEEPING THE RIGHT PEOPLE
ON THE DNA DATABASE
SCIENCE AND
PUBLIC PROTECTION
MAY 2009
You should also contact the Police Powers and Procedures Team should you require a copy of this
consultation paper in any other format, e.g. Braille, Large Font, or Audio.
Foreword
Foreword
Science and technology
provides a major opportunity
to help detect and convict
criminals. The UK is proud
that it leads the world in the
field of forensic science and
none more so than in DNA.
I measure the success of that
work on how it impacts on real people, how it helps the
level of public protection and how it enhances public
confidence in the criminal justice system.
The use of profiles stored on the National DNA
Database is a prime example where all those objectives
come together. We know that between April 1998 and
September 2008, there were over 390,000 crimes with
DNA matches, providing the police with a lead on the
possible identity of the offender. In 2007-08, 17,614
crimes were detected in which a DNA match was
available. They included 83 homicides, 184 rapes and a
further 15,420 additional detections
My aim is to ensure that those who should be on the
database, are on the database. Public protection lies at the
heart of our criminal justice system. That means having
a structure which ensures the safety and security of all
our citizens. Public protection also means delivering our
commitment to protecting the rights of the individual.
I want consideration of this important issue to focus on
what can and should be done with biometric data to help
detect and convict criminals; and what can be done to
use the message of the success of the DNA database to
stop re-offending and to prevent offending happening in
the first instance.
Jacqui Smith
Home Secretary
These are real crimes affecting real people. The impact
of crime on victims and their families should never be
underestimated. I am conscious of the impact that any
changes in our ability to detect offenders and potential
offenders sends out those who have suffered directly
from the consequences of crime.
That is why in a speech I made last December I
responded directly to the Bowman family whose
daughter Sally Anne was murdered in 2005 when I said
“I have real sympathy for all those with concerns that any
move could undermine a system that helped trap Sally
Anne’s killer. And I want to reassure Sally Anne’s father
that I will not let that happen”.
The proposals in this paper confirm that commitment.
1
2
Contents
Contents
Section 1
Consultation Aims
4
Section 2
Executive Summary
5
Section 3
Public Protection
8
Section 4
DNA in Context
9
Section 5
DNA Samples
12
Section 6
Implementing the Judgement: DNA Profiles
14
Section 7
Taking Samples – Additional Categories
19
Section 8
Fingerprints
20
Section 9
Volunteer Samples and Profiles
21
Section 10 Governance and Accountability
22
Annexes
Annex A:
Responding to this consultation
23
Annex B:
DNA database: Information held on profile records
24
Annex C:
DNA Retention after S and Marper: Jill Dando Institute April 2009
25
Annex D:
Impact Assessment
40
Annex E:
Confidentiality and Disclaimer
95
3
Keeping the right people on the DNA database: Science and public protection
Section 1:
Consultation Aims
1.1 On 4 December 2008, the European Court of
Human Rights delivered its judgment in the case
of S and Marper. Domestic courts had found in
favour of the Government’s position. However,
the Court found that the blanket policy in
England and Wales of retaining indefinitely the
fingerprints and DNA of all people who have
been arrested but not convicted was in breach
of Article 8 of the European Convention on
Human Rights.
1.2 The Court did, however, indicate that it agrees
with the Government that the retention of
fingerprint and DNA data “pursues the legitimate
purpose of the detection, and therefore,
prevention of crime”. This paper
l
sets out our proposals which will remove the
current “blanket” retention policy and replace
it with a retention framework which, in the
words of the judgment, will “discriminate
between different kinds of case and for the
application of strictly defined storage periods
for data”.
ObJECTIVE
1.3 To develop a DNA framework which has the
support and confidence of the public and
achieves a proportionate balance between
the rights of the individual and protection of
the public.
4
Section 2: Executive Summary
Section 2: Executive
Summary
2.1 The UK leads the world in the use of DNA
to solve crimes, to catch criminals and to clear
the innocent. Without this ability, we would be
less safe and criminals would be more likely to
get away with their crimes. It is to safeguard
this crucial ability that we must maintain public
confidence in our DNA database.
l
l
2.4 We consider that the existing threshold in PACE
for taking DNA and fingerprints on arrest
from a person detained at a police station for a
recordable offence is appropriate. This was not
called into question by the ECtHR. Therefore,
the paper sets out the future framework for
retention, destruction and governance of DNA
and fingerprints. The key recommendations are
as follows:
DNA Profiles
—
2.2 Furthermore, we are committed to complying
with the ruling of the European Court of
Human Rights (ECtHR) in the S and Marper
case. This paper sets out options, while also
indicating clearly where we have a preferred
approach to achieving compliance with the
judgement while maximising public protection.
The paper also sets out some areas where
additional powers could help secure improved
public protection.
2.3 It is not the purpose of this consultation paper
to rehearse the arguments in the case. But
the judgment clearly allows a retention policy
provided it is not “blanket and indiscriminate”.
This paper focuses therefore on the details of
retention, recognising the important distinctions
made in the judgement between cellular samples,
which contain an individual’s actual DNA, the
DNA profiles on the database which simply
describe for identification purposes certain
non coding parts of the individual’s DNA, and
finally fingerprints.
Samples – we are proposing the destruction
of all samples taken from suspects on arrest,
whether the individual goes on to be
convicted or not. Samples could only be
retained for as long as necessary to create a
profile suitable to be uploaded onto the
database, and for six months at the maximum.
Samples recovered from crime scenes would,
of course, be retained .
—
—
—
l
l
Adults convicted of a recordable offence
will have their profiles retained indefinitely
Adults arrested for a recordable offence
which is not a serious violent or sexual or
terrorism-related offence, but not
convicted will have their profiles
automatically deleted after six years
Adults arrested for a serious violent or
sexual offence or terrorism-related offence
but not convicted will have their profiles
automatically deleted after twelve years
Profiles from individuals volunteering to
have their DNA taken, for example for
elimination purposes, will not be stored on
the database
Fingerprints – retention of all fingerprints
and deletion after 6 years for those arrested but
not convicted; and after 12 years for those
arrested and not convicted of violent sexual or
terrorist related offences.
Exceptional grounds for earlier
destruction of profiles – the paper sets out
possible grounds for earlier destruction of
profiles. These could be requested by
application to the Chief Constable. Grounds
might include cases of wrongful/unlawful
arrest, mistaken identity, or in cases where it
emerges no crime has been committed. The
critieria which need to inform the Chief ’s
decision could then be codified or set out
in regulations.
5
Keeping the right people on the DNA database: Science and public protection
l
Children – the Home Secretary has stated the
need to be more flexible in the approach to
children
—
—
—
—
—
l
The DNA of all children under 10 – the
age of criminal responsibility – held on the
database has already been removed and will
not be retained in future either
Those under 18 years old who are
convicted of serious violent or sexual or
terrorism-related offences will have their
profiles retained indefinitely, along the lines
of adults
Those under eighteen who are convicted
on only one occasion of a lesser offence
will have the profile removed from the
database when they turn eighteen
Those under eighteen who are arrested but
not convicted of a serious violent or sexual
or terrorism-related offence will have the
profile retained for twelve years, in the
same way as for adults
Those under eighteen who are arrested but
not convicted of a lesser offence on one
occasion will have the profile deleted after
six years or on their eighteenth birthday,
whichever is the sooner. If the individual is
arrested again, the same retention periods
as for adults will apply
Governance – the existing NDNAD Strategy
Board will be rationalised and have a greater
mix of operational and independent members;
and an independent monitoring structure on
implementation of the regulations will be
introduced which will report directly to
Ministers.
2.5 We believe that these measures will provide an
open, transparent and accountable framework
for the taking and retention of biometric
data under the Police and Criminal Evidence
Act (PACE). We have sought to build our
options on principles set out in the ECtHR
judgement, backed up wherever possible with
an evidence base to justify proposed retention
periods. This is not easy, and there is no existing
6
evidence underlying retention regimes in other
jurisdictions. That is because we are at the
cutting edge of forensic development. In several
cases we are relying on new work which has not
yet been fully peer reviewed in the time available.
But we now have a strong evidence base to
support our proposed new retention framework.
2.6 The core judgement is around a six year
retention period for the vast majority of profiles
for persons arrested but not convicted. The
ECtHR ruling stressed that we needed to treat
non convicted individuals differently from
those convicted. At the same time, the Court
recognised we needed to take risk into account,
and praised systems like that in Scotland
where profiles are retained even in the case of
individuals not ultimately convicted in certain
circumstances.
2.7 To ensure public protection, we need to
understand how long it takes after an arrest for
a person to have no higher risk of re-arrest than
a member of the general public. Some work on
this suggests a figure of more than five years,
other work points to between 13-18 years. A
provisional model we have developed suggests a
figure of 4-15 years, which forms the framework
for the retention periods we are recommending.
2.8 We have selected the retention periods of 6
years and 12 years based on the likelihood
of people who have been arrested and not
convicted but who may go on to commit an
offence. Part of that analysis has included data
of those who have been arrested and convicted
based on independent research carried out
by the Jill Dando Institute (JDI) which found
that offending rates of those arrested but not
convicted were not significantly lower than
for those convicted and not given a custodial
sentence. The impact assessment research
show that it takes 15 years before the risk of
offending is at the same level as that for the
general population. The JDI research shows
that 52% of re-offending happens within six
years. We have taken a value judgment on
the associated level of risk that retention for
six years provides and combined within the
Section 2: Executive Summary
ECtHR judgment, concluded that this provides
a proportionate retention period. Two-thirds
of re-offending happens within 12 years. We
believe this a suitable period of retention for
those arrested but not convicted for violent,
sexual or terrorism-related offences in view of
the potential level of harm associated with such
offences and the issues of public confidence.
With our proposal to re-start the clock of 6
years or 12 years after any subsequent arrest, we
believe that a significantly greater proportion of
all offending will remain detectable.
2.9. The impact assessment for this paper sets out
the underlying assumptions in more detail. It
is important to note that any change to the
existing policy is likely to reduce the number of
detections that DNA delivers, and will therefore
have some adverse impact on public protection.
Our policy is designed to minimise this risk while
complying with the ECtHR ruling.
the law. We consider that the proposed timelines
will enable a suitable and realistic operational
response to this important judgment.
2.13 The Home Secretary made clear her intention to
do more to strengthen the dividing line between
guilt and innocence. In her speech to the Intellect
Trade Association on 16 December 2008, the
Home Secretary said that “for those who have
committed a serious offence, our retention
policies need to be as tough as possible.” That is
why the consultation paper also proposes that:
l
l
2.10 The destruction of all existing or legacy
samples would be a significant and lengthy
process and could realistically take up to two
years to complete that work. The destruction
of individual samples taken following the
introduction of new regulations should be done
within a maximum period of six months after
they were taken. In practice, this may be a matter
of weeks following the profile being successfully
loaded onto the National DNA Database.
2.11 In terms of destroying existing or legacy profiles,
we anticipate that a similar period of up to two
years would be required even though there would
be much smaller numbers involved i.e. profiles
relating to those acquitted or not prosecuted
between 1995 and 2003. That is because of the
need to track progress on each case.
those who are convicted of an offence and
whose DNA or fingerprints were not taken
during the criminal justice process would be
subject to a requirement to provide DNA and
fingerprints at any point subsequently. We are
proposing that this be made retrospective, but
limited to violent and sexual or terrorismrelated offenders.
UK citizens and residents who are convicted
overseas of violent and sexual or terrorismrelated offences should be required to
provide DNA and fingerprints on return to
this country.
2.14 Summaries of individual replies and a summary
of responses will be published following the
closing date for responses. Those responses
will be used to help inform the content of draft
regulations which will be subject to statutory
consultation. Draft regulations will indicate the
current preferred options for change. This is
not to prejudge the outcome of the consultation
but instead is intended to inform and assist
understanding of what the suggested approach
might look like.
2.12 We recognise that some will call for faster action
on deleting samples and profiles. We recognise
that concern but we also recognise that the aim
of the ECtHR judgment is not to create chaos
in the criminal justice system nor to divert
operational policing resources away from the
key functions of tackling crime and upholding
7
Keeping the right people on the DNA database: Science and public protection
Section 3:
Public protection
“It is the only deterrent that will stop serious crimes being
committed. I am a mother of four and I have five
grandchildren, I would not worry about any of their details
being held on a computer and everyone in our family feels the
same way. I am sick to death of the people who complain about
this idea. They have no idea what families like mine have been
through.”
Mrs Linda Bowman
(London Evening Standard 1 April 2009)
3.1 The NDNAD is an information database which
provides the opportunity to detect offenders and
to eliminate the innocent from enquiries quickly.
It also holds out the prospect of clearing up
cold cases on the basis of DNA left at the crime
scene. The existence of a profile on the database
does not indicate innocence or guilt of the
individual to whom it relates.
3.2 The Home Secretary’s Introduction and the
Executive Summary make clear that public
protection lies at the heart of the proposed
retention framework. This consultation paper
is about how to preserve public protection as
much as possible while complying with the court
decision in S and Marper.
3.3 There have been calls for a universal database
which would eliminate the suggestion or
perception of guilt. The NDNAD would then be
an even more effective tool as the police would
have instant access to the DNA profile of the
entire population.
3.4 We have never advocated a universal DNA
database. There are significant proportionality as
well as practical and operational issues associated
with such a database. This consultation paper
does not seek views on the arguments for and
against a universal DNA database.
8
3.5 Rather we are focussing on how to construct a
proportionate retention policy which continues
to help detect crime and protect the public
making best use of the database in ensuring
and enhancing public safety and protection
while meeting the terms of the judgment by the
ECtHR and ensuring we respect the rights of
individuals to a private life.
Section 4: DNA in context
Section 4:
DNA in context
4.1 DNA is now part of everyday language.
Advances in the fundamental understanding of
the human genome and the application of DNA
to our everyday life are truly astonishing. The
science of genomics covers all areas of life from
diagnosis and detection of illness and disease,
improving our environment and energy sources,
crops and the food chain, anthropology and
forensic science.
4.2 The National DNA Database was developed as a
means of contributing to the efficiency of crime
detection. The database was set up in 1995 to
store data derived from DNA profiles.
4.3 It operates on the simple premise that
identifying offenders more often and more
quickly should lead to increased detection of
crimes and bring more offenders to justice. As
importantly, it will enable the innocent to be
eliminated from enquiries; indeed the very first
case in which DNA was used enabled the police
to eliminate an individual who had wrongly
confessed to a murder. The DNA database is
also intended to act as a deterrent to offending
and re-offending and, importantly, it should
help raise public confidence that those guilty of
offending can be found and dealt with by the
criminal justice system.
4.4 The use to which DNA taken in the course of
a criminal investigation can be put is strictly
limited under the Police and Criminal Evidence
Act 1984.
4.5 The setting up of the database has revolutionised
the way in which the police work to help protect
the public. The majority of the active criminal
population is now believed to have its DNA
recorded and police forces use DNA profiles to
solve thousands of cases every year.
WhAT hAPPENS TO A DNA SAmPlE?
4.6 Under PACE, samples can be taken from an
individual detained at a custody suite for a
recordable offence – i.e. not a minor offence.
The chain of events on what happens when a
sample is taken from a person and what happens
when a sample is taken from a scene is set out
below (Fig 1) but it is instructive to clarify some
issues first:
WhAT IS A DNA SAmPlE?
4.7 This is a sample taken from an individual, such
as a mouth swab, plucked hair roots or blood
which contains the DNA of the individual for
analysis. A sample may be taken from a person
arrested for a recordable offence and detained at
a police station, or from a volunteer during a mass
screening process for elimination purposes, or
from samples taken at a crime scene. The sample
is handed over by the police to a laboratory for
a profile to be taken. The sample is retained by
the laboratory in secure sterile conditions and
barcoded to enable the sample to be matched to
the profile if necessary.
WhAT IS A DNA PrOFIlE?
4.8 The profile is shown as a numeric code on the
National DNA Database. Accessing records on
the NDNAD is strictly limited. When a police
officer asks for a search to be carried out against
a profile of a crime scene sample, he or she does
not have access to the database.
4.9 Instead they are provided with details only
of those profiles which provide a match. The
profile is the pattern of DNA characteristics
used to distinguish between individuals. The
profile is taken from ‘non coding’ or ‘junk’ parts
of the DNA, and does not contain personal
information other than that listed at Annex B.
4.10 The NDNAD therefore contains profiles which
consist of numeric data. Access to NDNAD
records is restricted to around 30 staff either
working in the NDNAD or the Forensic Science
Service. We are confident that the security
measures in place to monitor abuses or potential
9
Keeping the right people on the DNA database: Science and public protection
abuses of the database are working well. But we
are not complacent. The effectiveness of the
controls is subject to ongoing review
4.11 The physical storage of samples is also subject
to stringent security arrangements. The samples
are held at laboratories on behalf of chief police
officers. The use to which samples can be used
and when are set out in the Police and Criminal
Evidence Act (PACE) 1984.
4.12 The results from the application of DNA
profiling to crime detection have shown the
important contribution made. In 2007-08, there
were 37,376 crimes with a DNA match which
provided the police with an intelligence lead for
further investigative follow-up. These included
serious offences including 363 homicides, 540
rapes, nearly 1,800 violent crimes and more than
8,000 domestic burglary offences.
4.13 We know that from research between May 2001
and 31 December 2005 there were approximately
200,000 DNA profiles on the National DNA
Database which would previously have had
to be removed before legislation was passed
in 2001 because the person was acquitted or
charges dropped.
4.14 Of these 200,000 profiles, approximately 8,500
profiles from some 6,290 individuals have been
linked with crime scene profiles, involving nearly
14,000 offences. These include 114 murders,
55 attempted murders, 116 rapes, 68 sexual
offences, 119 aggravated burglaries and 127
offences of the supply of controlled drugs.
4.15 These results explain why we robustly
defended our retention policy on DNA and
fingerprints through the domestic courts and
onto the European Court of Human Rights.
Having successfully defended the policy up to
and including the House of Lords, we were
disappointed by but accept the judgment
of the Court in the S and Marper case on
4 December 2008.
10
4.16 We need to comply with the judgment in the
S and Marper case, while minimising the risks to
public protection. The proposals outlined here
for consultation seek to do that.
Section 4: DNA in context
Figure 1: Loading a subject profile to The National DNA Database
Police take DNA
samples from a
suspect in police
station, using a
sample kit with a
unique bar code
identifier
Police record details of the sample on DNA
card provided in the sample kit.
Police send
samples and DNA
card to
forensic supplier
Police create a record on the Police National Computer (PNC).
Generates a unique PNC identifier.
Message sent
electronically from
PNC to the
NDNAD which
includes the
bar code and
PNC identifiers
Forensic supplier
analyses half of
the sample to
derive the
DNA profile Forensic supplier
sends DNA profile
electronically
to the NDNAD NDNAD loader
checks that the
bar code and PNC
identifiers on the
DNA profile match
those on the
stub record
DNA Profile
successfully
loaded to the
NDNAD and is
searched against
all other profiles
on the database
NDNAD sends
details of any
resulting DNA
profile matches
to police forces
11
Keeping the right people on the DNA database: Science and public protection
Section 5:
DNA Samples
5.1 The ECtHR judgment highlighted the sensitivity
around retention of samples which contains
cellular information. The Court recognised that
whilst the NDNAD only held a very limited
amount of data for profiling purposes, the nature
and amount of personal information held by the
State was particularly sensitive and constituted
an interference with the private lives of the
individuals concerned.
5.2 Our view is that there is scope for destroying
samples of both those arrested but not
convicted and for those convicted. This goes
substantially further than the requirements
of the S and Marper judgment. The reasons
for doing so are set out below. It is possible
that some unenvisaged circumstances may
occur that would result in a failed prosecution
but this must be weighed against the retention
of potential genetic information for over 4.5
million people. Whilst an outcome may be to
develop a retention policy on samples based
on risk and benefits, we need to be satisfied
that the potential implications for future
public protection are fully identified. We would
welcome comments on the proposal.
5.3 As at 31 March 2009, there were samples relating
to over 4.5 million people held under the PACE
provisions. Some concerns have been expressed
that we are establishing a genetic database with
a suspicion that the samples will be used for
purposes other than crime investigation or
research. The policy to date on retention of all
DNA samples from individuals has its origins
in a belief that we may need to upgrade the
NDNAD as DNA technology develops and
this would require re-analysis of all retained
samples. This is a remote prospect as if we were
to use more discriminating or different DNA
methods then we would run them in parallel
with the current system. Given the size of the
current NDNAD there would have to be a very
12
significant reduction in cost and increase in
effectiveness to justify doing this.
5.4 The destruction of samples raises issues about
the potential legal challenges demanding the
production of the original sample. Given the
purpose of the NDNAD is to identify or
eliminate a possible suspect, they need to be
located and arrested before any criminal justice
action can be taken. This affords an opportunity
to re-sample and confirm a DNA match if
there is any doubt or procedural challenge that
emerges in criminal proceedings.
5.5 There is currently a process in place where the
release of samples requires the request from an
ACPO officer and agreement of the Chair of
the NDNAD. These are relatively infrequent (2-3
per week) and very few situations which are likely
to occur where a criminal prosecution would be
compromised.
5.6 If we were to adopt a destruction policy on
samples, we would need, as part of the quality
control system operated by the forensic
providers, to provide a sufficient period in
which to re-analyse samples. This requirement
could be accommodated by allowing for
destruction after, say, any time up to a maximum
of six months; or for when a satisfactory profile
is loaded onto the database.
5.7 The destruction of samples would be a statutory
requirement under the proposed Regulations.
The likely approach would be for the National
Policing Improvement Agency (NPIA) and the
NDNAD Strategy Board to consider contracting
a suitable organisation to collect samples from
forensic science providers and destroy them as
biological waste. With the samples having an
estimated physical size of 25m3, it is anticipated
that it would take up to 12 months after agreeing
a contract to deal with legacy samples.
Section 5: DNA Samples
Samples: Summary of Recommendations
l
l
l
Samples to be destroyed for all existing and
future cases, both for persons arrested and not
convicted and for those who have been
convicted.
Samples to be retained for up to six months
maximum for possible re-examination purposes
only. Samples to be destroyed when an effective
profile is on the NDNAD.
Legacy samples to be destroyed within
12 months of date of commencement of
regulations.
13
Keeping the right people on the DNA database: Science and public protection
Section 6:
Implementing the
judgment: DNA profiles
6.3 Therefore, in the light of the judgment, research
has been undertaken to establish the latest
evidence to help to consider the options available
to inform a retention period for those arrested
but not convicted or against whom no further
action was taken.
6.1 The ECtHR judgment identified England, Wales
and Northern Ireland to be the only jurisdictions
within the Council of Europe to allow the
indefinite retention of DNA material of any
person of any age suspected of any recordable
offence. Other jurisdictions (e.g. France) have
very long retention periods for some crimes even
in non-conviction cases, sometimes for as long as
25 years.
6.4 First, we considered the simple approach of
destroying all profiles for people arrested but
not convicted. The judgment does not require
that we must delete all fingerprints, samples and
profiles for persons arrested but not convicted.
This option would have resulted in around
850,000 records having to be destroyed. Such a
reduction – going well beyond what is required
by the ECtHR judgment – would have significant
impact on resources and, more importantly, on
future ability to detect offenders.
6.2 The court has accepted there is a justification
in retaining profiles in non-conviction cases.
The question is for how long. We have sought
to examine the evidence base which could
inform this decision. The challenge now is to
devise a framework that is evidence based and
proportionate whilst retaining as far as possible
the benefits to public protection that the existing
scheme offers.
KENSLEY LARRIER – arrested in May 2002 for
the Possession of an Offensive Weapon. His DNA
was taken at this time and loaded to the DNA
database in June 2002. The proceedings were
discontinued in October 2002. Larrier’s DNA was
retained under the provisions of the Criminal Justice
and Police Act 2001.
In July 2004, a rape was committed in the North of
England. DNA from this investigation was speculatively searched against the NDNAD and matched against the acquittal sample. He was arrested, and charged with the offence in November 2004.
Larrier was convicted in June 2005, jailed for 5 years
and was entered on the sex offenders register for life.
14
6.5 The research paper from the Jill Dando Institute
(JDI) is attached at Annex C. This should be
read in conjunction with the Impact Assessment
at Annex D. Importantly, the JDI research
concludes that the seriousness of the initial
offence for which the person was arrested
does not necessarily predict the seriousness of
subsequent offences with which the person may
be associated. As a result, a policy which only
retained profiles where an individual was arrested
for a serious or violent offence (as applies,
for example, in Scotland) would risk missing
numerous detections. For example, the most
common offences for which profiles ultimately
linked to murder cases were originally taken are
drug offences.
6.6 In determining the most suitable retention
period, the key question is one of risk. We have
sought to assess how many years after arrest
an individual’s risk of being rearrested is the
same as the risk of an individual in the general
population.
Section 6: Implementing the judgment: dna profiles
6.7 Some US studies have looked at this from the
point of view of criminal record checking for
potential employees. The first paper (Kurlycheck,
Brame and Bushway) considers the ‘hazard rate’
for a Philadelphia cohort of individuals born
in 1958. The key point from this report is that
after 5 years the difference between hazard rates
for arrested and non arrested individuals is still
significant at over 1%.
6.8 The second paper (Nakamura) considers the
hazard of rearrest for a cohort of people
arrested for the first time in 1980 in New York.
It investigates the effect of being arrested
for different crime types on the time for the
arrestees’ hazard rate to reach the population
hazard rate. The paper concludes that for first
time arrestees of age 18 who were arrested for
robbery, the period is 14 years, for arrestees of
age 16 arrested for burglary, the period is over
13 years.
6.11 This is obviously a controversial assertion, but
it does appear to be borne out by some work
carried out by JDI on a cohort from the mid
1990s (see fig 3). This work suggests that the risk
of subsequent conviction is at least as high in the
group who were subject to no further action to
those who received a caution or a non-custodial
sentence. Nothing here detracts from the legal
principle of the presumption of innocence
of any individual who is not convicted, but
we believe the sort of analysis carried out is
legitimate in assessing underlying risks.
6.12 The Home Office analysis looked at a cohort of
offenders who had been convicted of an offence
in 2001. The reconvictions for this cohort were
investigated in each of the following six years,
with extrapolations for future years. The research
suggests that within 4 years the ‘hazard rate’
converges with that for the peak offending age
group (males aged 16-20). The cohort converges
with the general population around 14-15 years.
6.9 We have sought to replicate this work looking
at data on the PNC. The nature of the material
here means that we are forced to use PNC
conviction and reoffending data.
6.13 On the basis of our own work and the US
evidence, we have concluded that a retention
period of around 6 years for most offences
seems reasonable, with a longer period for
serious sexual and violent offences. The evidence
for reoffending in more serious and violent
cases is unclear, but we believe a longer retention
period is a commonsense approach given the
more serious consequences of reoffending and
therefore the damage that a missed detection
would imply.
6.10 For this to be valid, we would have to believe
that the risk of offending following an arrest
which did not lead to a conviction is similar to
the risk of reoffending following conviction.
Figure 3: % cases arrested again within risk period
Year (Risk Period)
NFA
Caution
Non-Custodial
Sentence
No of cases
1996 sample (30 months)
28
42
23
206
1995 Sample (42 months)
50
47
50
227
1994 Sample (54 months)
48
50
38
99
Combined
40
46
35
532
Source: Jill Dando Institute, April 2009
15
Keeping the right people on the DNA database: Science and public protection
6.14 Some may want a shorter period of retention,
some may want a longer period and others no
retention period at all. We welcome views on the
suggested approach and supporting evidence
which would assist in determining whether an
alternative period is more appropriate.
TErrOrISm
6.15 We also believe that profiles obtained and retained
in relation to terrorism and national security
should be deleted automatically after 12 years
unless the person is convicted of a recordable
offence. This includes profiles obtained under
Schedule 8 to the Terrorism Act 2000 (from
persons arrested as a suspected terrorist or
persons detained under Schedule 7 to the 2000
Act), profiles of those individuals subject to a
control order and profiles retained under section
18 of the Counter-Terrorism Act 2008.
6.16 In the case of samples taken from controlled
individuals, the profiles will be retained for
a period of twelve years from the point at
which the individual is no longer subject to a
control order.
rETENTION PErIOD FOr ChIlDrEN
6.17 The Home Secretary indicated in her speech
on 16 December 2008 that she wants to adopt
a different approach to young people. This
recognised that whilst the typical residual career
length for those who get involved in crime at an
early (teen years) age is 16 years, for many young
people involvement in crime at that age is often
an isolated incident and can be relatively minor.
6.18 We are therefore proposing a policy of deleting
profiles of children who are convicted once only
of minor offences. If a child commits a serious
offence or two minor offences, the profile will be
retained indefinitely, as for adults.
6.19 Similarly for those arrested but not convicted of
minor offences, we are proposing that profiles
be deleted after six years or on the eighteenth
birthday, whichever is the sooner. For serious
violent or sexual or terrorism-related offences,
16
the same 12 year rule will apply to children
and adults.
DESTruCTION OF PrOFIlES ON ExCEPTIONAl
GrOuNDS
6.20 These automatic destruction periods will ensure
that profiles will be removed at predictable
points without requiring any further action.
There may still be cases, however, where
members of the public feel their profiles and
samples should be removed immediately.
Examples might be where there has been a
wrongful arrest, or a case of mistaken identity,
or where it turns out that no crime has been
committed. There is already a right to appeal
to the Chief Officer on exceptional grounds,
though it is rarely exercised.
6.21 The current process under the Exceptional
Case Procedure of making an application to
the chief officer would remain in place. That
is because Chief Officers have the discretion
to authorise the destruction of DNA and
fingerprints. That discretion would remain
subject to judicial review.
6.22 We do, however, propose two significant
changes. First, procedures should be renamed
‘application process for record deletion’. The
change in title is not cosmetic. Applications
for deletion will still be possible but will need
to be made and considered against defined
criteria. The criteria for deletion would be set
out in Regulations. It is not possible to define
comprehensive criteria in legislation for what
will be in practice a based on the individual
circumstances of each case. However, it should,
for example, involve cases where the arrest was
unlawful, where the taking of the sample was
unlawful, or where no offence existed e.g. where
a suspected unlawful killing turns out to be a
death by natural causes. Draft regulations will set
out proposed criteria.
Section 6: Implementing the judgment: dna profiles
PrOFIlES OF CONVICTED PErSONS
6.23 Apart from the proposals above in respect of
juveniles in specific circumstances, we do not
propose to change the existing indefinite period
for retention of profiles for those convicted of a
recordable offence. This would also cover people
given a caution, warning or reprimand.
6.24 We do however recognise the need to co­
ordinate our approach on DNA profiles with
that of retention of other police records.
Central to this is the development of criminality
information policy and implementation of
Sir Ian Magee’s recommendations following
his independent review. These include
outstanding recommendations from the Bichard
Inquiry report.
lEGACy PrOFIlES OF PEOPlE ArrESTED buT
NOT CONVICTED Or ACquITTED
6.25 The Government has given effect to the S and
Marper judgment by destroying the relevant
samples of S and of Marper and by providing
just satisfaction for costs and expenses. The
judgement was made in respect of the case of
the two applicants.
6.26 But we have to consider the position of people
in similar circumstances to the two applicants. In
other words, the profiles already on the database
of people who have not been convicted.
6.27 There are approximately 850, 000 legacy profiles
of which approximately 500, 000 have no linked
PNC Record. This means it is not possible to
tell whether the latter profiles relate to persons
arrested and not convicted or subject to no
further action, or to people who have been
convicted.
6.29 Therefore, there are two issues relating to legacy
cases: first, those whose profile is linked to a
PNC record. In those cases, we are proposing
that the 6-year or the 12-year retention criteria is
applied depending on the offence
6.30 The second issue relates to profiles were there
is no linked PNC record. There are two options.
The least expensive and most efficient process
would be to delete the 500,000 profiles. The
second approach is for the police to match
profiles against records and where a record is
identified, apply the six-year rule, the 12-year rule
or the conviction rule. The Home Secretary has
made clear her intention that the DNA database
should contain profiles of those who should be
on it.
6.31 Deleting such a volume of profiles without
better understanding of the associated risk is
therefore potentially a high risk option. The
Home Secretary has asked the Association
of Chief Police Officers (ACPO) to carry
out further work on this aspect and provide
a detailed impact assessment which can be
published as part of the Summary of Responses
to the consultation exercise on this paper.
6.32 The reference to subsequent arrest refers to
arrest for an offence not related to the original
offence. For example, a person arrested for
attempted murder subsequently reduced to
grievous bodily harm would not be considered
as a new arrest for biometric retention purposes.
Similarly, the re-arrest of a person on fresh
evidence would not constitute a second or
subsequent arrest for biometric retention
purposes.
6.28 It is possible that a proportion of the nonreconciled profiles relate to a conviction. At this
stage it is difficult to estimate what percentage
would be deleted because the information is not
currently held by police forces in a format which
can provide such an assessment. Dealing with
this group represents the biggest challenge and
has the greatest resource implication.
17
Keeping the right people on the DNA database: Science and public protection
Profiles: Summary of Recommendations
l
l
l
l
l
l
l
18
All profiles to be retained for six years for
persons arrested for a recordable offence but not
convicted.
Profiles of persons subject to arrest within that
period to be subject to an automatic retention
period of a further six years.
Profiles of persons arrested but not convicted for
specified violent or sexual terrorism-related
offences or to be retained for 12 years.
Persons over the age of 10 years and under 18
years of age to have profiles deleted at reaching
18 years old whether or not convicted (subject to
the violent or sexual offences criteria) unless
arrested for a subsequent offence before they
reach 18, in which case the rules applicable to
adults apply.
Regulations to set out criteria for making an
application for deletion of profiles.
Six year and 12 year retention periods to start
from date of arrest, except in the case of a
control date where the period of retention will
commence from the date of the order.
Deletion of profiles on the database for persons
arrested but not convicted to be applied from six
years from the date of commencement of the
regulations.
Section 7: Taking samples – additional categories
Section 7: Taking
samples – additional
categories
7.1 The proposed regulations relate only to the
retention and use of DNA and fingerprints. We
are proposing in future primary legislation to be
introduced when Parliamentary time allows to
provide additional provision for the taking of
samples in three specific instances:
l
l
l
Post arrest – where a person has provided a
sample and it has proved to be insufficient for
profiling purposes. We are proposing that the
police should have a power to require a person
to provide a further sample.
Taking Samples: Summary of
Recommendations
l
l
l
Provide the police with a power to take a sample
and fingerprints following arrest if the initial data
is not sufficient for profiling or IDENT1 needs.*
Provide the police with a power to take data post
conviction of persons who were not sampled or
fingerprinted during the investigation or court
process.*
Provide the police with a power to take data from
UK nationals and UK residents convicted of
violent or sexual offences overseas.*
* Proposals for primary legislation.
Post conviction – currently a sample may be
taken if a person is convicted and in prison
custody and a sample was not taken during the
investigative or court process. If a person is
convicted or charged but not subject to a
prison sentence, the police must request within
one month of the conviction or charge, or
within one month of the police being
informed that the sample is not suitable for
analysis, that the person attends the police
station for a sample to be taken. We are
proposing that the police may require a sample
in these circumstances at any time postconviction. The particular operational focus
will be on ensuring that the profiles of those
convicted of the more serious offending will
be on the NDNAD.
It is important to strengthen public protection
by ensuring that the profiles of those UK
residents and nationals convicted of sexual or
violent or terrorism-related offences overseas
are retained in the NDNAD in view of the risk
they may pose here.
19
Keeping the right people on the DNA database: Science and public protection
Section 8: Fingerprints
8.1 The ECtHR judgment recognised that
fingerprints do not contain as much information
as either samples or DNA profiles. As a
result, the Court found that the retention of
fingerprints has less impact on a person’s private
life than the retention of samples or DNA
profiles. The Court considered that the retention
of fingerprints pursued the legitimate purpose
of the detection and therefore prevention
of crime. The Court also recognised that
fingerprints do not contain subject information
and that, accordingly, they do not have the
same impact on private life as cellular samples
and the DNA profile. Nevertheless, the Court
did conclude that the blanket retention of
fingerprints constitutes an interference with the
rights to respect for private life.
8.2 The main focus of the judgment was, however,
clearly on the impact of DNA samples and
profiles. Concern lay in the potential for detailed
personal information to be used outside
the context of the immediate investigation.
Fingerprints do not provide any additional
information other than being able to confirm the
identify of the person.
8.3 The national fingerprint database is known as
INDENT1. Fingerprints are a key identifier of
a person as well as an investigative tool. We are
proposing that we retain fingerprints for those
arrested and not convicted for 6 years and for 12
years for those arrested and not convicted for a
violent, sexual or terrorist related offence.
8.4 In proposing this approach, we are recognising
the benefits that fingerprints bring and have
brought for over a century to enable the
police to confirm who they are dealing with
and in more personal situations, help in the
identification of victims in disasters and other
incidents. It could be argued that this applies
also to DNA. However, as the ECtHR judgment
recognised, cellular samples and DNA profiles
constitute a much greater risk of being used for
20
subjective analysis. For that reason we have also
impact assessed a retention period of 15 years
for fingerprints.
8.5 PACE places specific requirements on the
retention and use of fingerprints. There is no
provision within the 1984 Act to use fingerprints
in connection with the National Identity Scheme.
8.6 There is a current provision within PACE which
allows a person to make a request to witness the
destruction of his or her fingerprints. We are
proposing to remove this entitlement.
Fingerprints: Summary of Recommendations
l
l
l
l
INDENT1 database to retain for 6 years for
persons arrested but not convicted on all offences;
and 12 years for those arrested and not convicted
for a violent, sexual or terrorist related offence.
Audit trail of any copying of fingerprints and their use.
Automatic destruction of copies when no longer
required for investigative purpose.
Removal of individual’s ability to witness destruction of fingerprints.
Section 9: Volunteer samples and profiles
Section 9: Volunteer samples and profiles
9.1 Volunteers consent to provide their samples
in one of two ways: either as part of a mass
screening in a geographical area or on an
individual basis. In both instances, the request
must be related to a specific offence. The
profiles are searched against the relevant crime
scene sample.
Volunteers: Summary of Recommendations
l
l
l
Existing volunteer samples to be removed from
the NDNAD.
Future profiles and samples to be destroyed
when no longer required for investigative
purposes.
Future volunteer samples and profiles to be
subject to distinct processes from speculative
searching on the NDNAD
9.2 In giving their consent to the sample, the
volunteer is also asked whether they wish to give
their consent for their profile to remain on the
NDNAD. If such consent is given, the volunteer
is not then able to subsequently require that the
sample and profile are destroyed.
9.3 We are proposing that a volunteer who gives
their samples for elimination purposes are
not placed on the NDNAD. Whilst consent
will continue to be required for the taking
of the sample, consent will not be sought
for the sample or fingerprints to be retained
on a national database and subject to future
speculative searches.
9.4 Existing ‘volunteer’ samples will be removed
from the database whether or not the person
has consented for its retention. That process is
already under consideration by the NDNAD
Strategy Board and ACPO will be writing out
shortly to all chief officers to inform them that
future volunteer samples and profiles should be
handled through a distinct and separate process
from the NDNAD and that existing samples
should be removed from the NDNAD.
9.5 This will mean that future volunteer profiles
will only be searched against crime scene
samples relating to the specific offence under
investigation.
21
Keeping the right people on the DNA database: Science and public protection
Section 10:
Governance and
accountability
10.1 We have responded promptly to the judgment
by removing S and Marper’s relevant details
from the NDNAD; made payment on costs
and expenses as required by the judgment; and
submitted an initial report to Council of Europe
Committee of Ministers on implementation
progress.
10.2 The Home Secretary’s speech on 16 December
just 12 days after the judgment, reflected the
importance attached to this important area and
the need for change. Since the Home Secretary’s
commitment in that speech to deal with under
10s on the NDNAD, their profiles have all since
been removed.
10.3 This paper sets out proposals to introduce
statutory regulations setting out the criteria for
the making and consideration of applications
to have DNA and fingerprints deleted where
someone considers that their data should not be
retained. Procedures on deletion of such data by
the police will be more transparent.
10.4 We are further proposing that a strategic and
independent advisory panel is tasked with the
function of monitoring the implementation and
operation of the Regulations. This would not be
an appeal mechanism – it would be inappropriate
for an administrative body to rule on the
decisions of a chief police officer – but would
exist to monitor the application of the new
approach and provide advice and guidance to
Ministers through an annual report. Part of their
function would be to comment on application
of the regulations by individual forces. This
would be achieved by examination of statistical
information, currently some of which is already
supplied by forces to the NPIA. It may be that
22
this role can be carried out by an existing body
but we are keen to ensure that such a group
is able to adopt an entirely independent and
constructively critical approach.
10.5 The National DNA Strategy Board has
already commenced a review of their existing
governance structure. A key part of that
consideration is focussed on having more
external and independent membership
represented on the Board.
10.6 The introduction of the Regulations would be
accompanied by a code of practice under section
66 of PACE. Currently Code D deals primarily
with the taking of biometric data as well as
other means to identify the person. There is
considerable scope to expand the Code to reflect
the proposed new Regulations on the retention
and destruction of biometric data. Work has
already started as part of the PACE Review on
re-drafting and re-designing the Notes of Rights
and Entitlements given to detainees.
Summary of Recommendations:
Accountability and Governance
l
l
l
l
Restructuring of the National DNA Strategy
Board to have more external, independent
membership.
The establishment of a strategic and
independent advisory panel to monitor and
scrutinise the retention policy and the processes
of consideration for destruction of profiles.
Annual reporting by the independent advisory
panel to Ministers.
Quarterly/Annual publication of the key
statistics on NDNAD numbers, speculative
searches, deletions and applications for deletions.
Annex A: Responding to this consultation
Annex A: Responding
to this consultation
SCOPE
Topic of this consultation:
Secondary legislation proposed under the Policing and
Crime Bill to amend section 64 of the Police and
Criminal Evidence Act (PACE) 1984 and make more
detailed provisions on the retention, use and destruction
of DNA data and fingerprints.
Scope of this consultation:
The regulations that will need to be put in place to
implement the judgment of the European Court of
Human Rights in the case of S and Marper; and other
measures to raise the effectiveness of the use of
biometric data to ensure we achieve a proper balance
between protecting the public and safeguarding
human rights.
Scope:
England and Wales; Northern Ireland – Consultees
should note that: The policy principles and future
approach to implementation, within the consultation
document, also apply to Northern Ireland; – any
reference to PACE also includes PACE (Northern
Ireland) Order 1989.
Enquiries:
A copy of this consultation paper and further
information on DNA and fingerprints can be found at
http://www.homeoffice.gov.uk/about-us/haveyoursay/
current-consultations/
How to Respond:
Responses should be sent by email to:
[email protected]
By post:
Alan Brown
DNA Consultation
Home Office
Policing Powers and Protection Unit
4th Floor Peel NW, 2 Marsham Street, London
SW1P 4DF
Additional ways to become involved:
The consultation will largely be a written exercise.
After the responses to the public consultation:
A document summarising the consultation will be
published together with a further draft of the regulations
for comment. The responses on the draft regulations will
be published before the draft regulations are finalised
and laid before Parliament for approval by both Houses
of Parliament. Respondents to the consultation will be
identified unless specifically requested otherwise.
Impact Assessment:
An impact assessment is attached at Annex D.
bASIC INFOrmATION
To:
This is a public consultation and anyone may comment.
Individuals and organisations who responded to the
PACE Review consultation exercise have been notified
direct of the consultation as have other groups who may
have a particular interest in the subject.
Duration:
This is a 12-week consultation starting from 7 May 2009
and closing on 7 August 2009.
23
Keeping the right people on the DNA database: Science and public protection
Annex B: DNA
Database: information
held on profile records
17. Date of Birth
All NDNAD records have the same structure consisting
of 36 data fields. However, some of these fields are only
relevant to subject profiles, and some are only relevant to
crime scene profiles. Therefore, no record will have all
data fields completed.
20. Gender
1. DNA Sample Barcode
23. Batch number
2. Arrest Summons Number
24. Batch Year
3. Class code (this denotes whether the sample was
from an arrestee, a volunteer etc)
25. Number in Batch
18. Alias 1
19. Alias 2 (aliases do not originate from the PNC
record and are not often used).
21. Ethnic appearance code
22. Sample Type (i.e. whether from saliva, hair, blood)
26. Track number (denotes position of sample in gel)
4. PNC status (reconciled or non-reconciled; denotes
whether there is a link to PNC)
5. Criminal Record Office number
27. Date sample taken (originates from PNC)
28. Date Sample Loaded (relates to date sample stub
record is loaded)
6. PNC ID Number
29. Gel Number
7. Case Lab code
}
30. Gel Year
8. Case year }
9. Case number
All these relate to crime
scene samples
31. Test method (e.g. SGM or SGM+ which refers to
the DNA profiling system used)
10. Offence code
}
32. Date Batch first added (used in 1990s but no
longer used for loading profiles)
11. Job number
}
33. Date Profile loaded
12. Item number
}
34. Forensic supplier
13. Crime Number
}
35. Unit within forensic supplier
14. Police force
15. Station
16. Name
24
}
36. Whether record searchable
Annex C: DNA Retention after S and Marper: Ken Pease Jill Dando Institute April 2009
Annex C: DNA Retention after S and Marper:
Ken Pease Jill Dando Institute April 2009
DNA Retention after S & Marper
Ken Pease
April 2009
25
Keeping the right people on the DNA database: Science and public protection
SummAry
The S & Marper judgement casts the retention issue as
one of balance between the principles of individual
privacy and public protection. This means that there is a
level of public protection conferred by retention of
DNA from arrestees against whom no further action is
taken (hereinafter NFA arrestees) which would justify
that retention.
Analysis of a sample of such cases from the
Metropolitan Police force area suggests that the level of
subsequent criminality over three years is on a par with
the subsequent criminality of those given non-custodial
sentences or cautioned. This is taken to justify analysis of
this group in terms of the extant criminal career
literature to derive estimates of the value of retention for
varying periods. This suggests that a six year retention
period would cover half the residual number of offences
(RNOs), ie would lose half the maximum downstream
detective value.
Shorter periods would lose more. It is estimated that a
retention period of some 24 years would be necessary to
maximise profile value in detecting future crime. Selective
deletion by offence seriousness or youth of arrestee
would remove most of the detective value of retention
for the arrestee-NFA group.
A wider criminological research programme is advocated
to maximise the prospective detective value of forensic
DNA work, some of whose possible elements are
described.
INTrODuCTION
The judgement in S and Marper1 held that the practice in
England and Wales of retaining on the national database
the DNA samples and profiles on those arrested but
where guilt was not subsequently established or admitted
should be discontinued and presently held samples and
profiles discarded. Verbatim,
“1. In conclusion, the Court finds that the blanket and
indiscriminate nature of the powers of retention of the
fingerprints, cellular samples and DNA profiles of persons
suspected but not convicted of offences, as applied in the case of
the present applicants, fails to strike a fair balance between the
1
26
competing public and private interests and that the respondent
State has overstepped any acceptable margin of appreciation in
this regard. Accordingly, the retention at issue constitutes a
disproportionate interference with the applicants’ right to respect
for private life and cannot be regarded as necessary in a
democratic society.”
This is not the place to analyse the S & Marper
judgement in detail, but it is pertinent to what follows to
note that the judgement quoted above and elsewhere is
couched primarily in terms of proportionality in the
balancing of principles, setting the right of privacy of
those arrested against the protection of the public to be
gained by retention of DNA samples and profiles. Had
the judgement held that there is no level of public
protection so great that it could not be forgone to
maintain the principle of DNA privacy in those not
found or admitting guilt, the debate would end there.
However, this is not the argument.
For example
“104. The interests of the data subjects and the
community as a whole in protecting the personal data,
including fingerprint and DNA information, may be
outweighed by the legitimate interest in the
prevention of crime (see Article 9 of the Data
Protection Convention).”
And
“112….The Court observes that the protection afforded by
Article 8 of the Convention would be unacceptably weakened
if the use of modern scientific techniques in the criminal-justice
system were allowed at any cost and without carefully balancing
the potential benefits of the extensive use of such techniques
against important private-life interests”.
Implicit in this emphasis is the possibility that there is a
level of public protection which could outweigh the right
to privacy. There are four ways in which the balance
might change. These involve the adjustment upwards or
downwards of the two balancing concerns. The balance
could be reset by challenging some of the assumptions
underpinning the S & Marper judgement (for example its
conflation of downstream usage of samples and profiles
European Court of Human Rights, Grand Chamber, S & Marper v United Kingdom. Applications 30562.04 and 30566.04, Judgement of
4th December 2008.
Annex C: DNA Retention after S and Marper: Ken Pease Jill Dando Institute April 2009
to disclose aspects of phenotype and health could be
held to overstate the extent to which DNA profiles
compromise privacy). It could also be reset by fuller
demonstration of the utility of DNA profiles in
deterring or detecting future crime. This would have to
be beyond the numbers attributed to Lord Steyn and
reported in S & Marper as follows:
“2. Lord Steyn also referred to statistical evidence from which it
appeared that almost 6,000 DNA profiles had been linked
with crime-scene stain profiles which would have been destroyed
under the former provisions. The offences involved included
53 murders, 33 attempted murders, 94 rapes, 38 sexual
offences, 63 aggravated burglaries and 56 cases involving the
supply of controlled drugs.”2
That the data cited by Lord Steyn came from only a
sample of police force areas, and will increase over time,
it may be that the larger numbers which would now be
presented may in themselves tip the balance. It is
interesting that Lord Steyn expressed his statistics as a
count rather than a rate. By doing so, the erroneous
impression is given that this represents the sum total of
downstream DNA-detectable crime.
Characterising the retention question as involving a
trade-off between public protection and the right to
privacy is perhaps to draw it too narrowly. In balancing
individual and public goods, John Rawls’ (1971) notion
of a veil of ignorance is helpful. This “veil” is one that
essentially blinds people to all facts about themselves that
might cloud what notion of justice is developed.
“no one knows his place in society, his class position or social
status, nor does anyone know his fortune in the distribution of
natural assets and abilities, his intelligence, strength, and the
like. I shall even assume that the parties do not know their
conceptions of the good or their special psychological propensities.
The principles of justice are chosen behind a veil of ignorance.
(Rawls 1971 p60)”
According to Rawls, ignorance of these details about
oneself will lead to principles that are fair to all. If an
individual does not know how she will end up in his own
conceived society, she is likely not going to privilege any
one class of people, but rather develop a scheme of
justice that treats all fairly. At the risk of moving too
swiftly from philosophical theorising to the all too
practical issue of DNA retention, what are the
considerations which one should bring to bear from
behind a veil of ignorance? Behind such a veil, the
considerations for the arrestee and his or her family and
friends, the offence victim and his or her family and
friends, and the citizen as taxpayer, as potential arrestee
or victim or their intimates. The interests of the actual or
potential arrestee, and of the citizen engaging in lawful
but embarrassing pursuits which may be revealed by the
presence of his or her DNA at a crime scene, have
perhaps been covered adequately elsewhere (see for
example Nuffield Council on Bioethics 2007), with two,
possibly three points under-emphasised. These are:
1. Vulnerability to confirmation bias by investigating officers.
This is the general tendency, manifest in numerous
miscarriages of justice. “…[W]e are more likely to
notice and focus on characteristics that validate
and conform to … a belief or a hope. ….
Confirming data are weighted highly, and …
disconfirming data that we notice are ignored”.
(Dror and Fraser-Mackenzie 2009). Thus
identification of a ‘prime suspect’ carries the
danger of a search for inculpatory evidence to
predominate. Confirmation bias has been deemed
ubiquitous in human reasoning (Nickerson 1998).
The existence of DNA (or indeed any other
forensic evidence) at a crime scene carries the risk
of triggering the confirmation bias.
2. Exculpatory Power of DNA. In the first ever
forensic use of DNA evidence,3 The prime
suspect in the murder of two young women was a
local boy, who confessed to one but denied the
other. The comparison of semen samples from
both murders against a blood sample from the
suspect conclusively proved that both girls were
killed by the same man, who was not the suspect.
In the USA, The Innocence Project claims 235
post-conviction DNA exculpations to date.
Seventeen of these people had been sentenced to
death before DNA proved their innocence and led
2
The necessary qualification is that these are cases where DNA evidence existed, but this does not mean that a conviction would not have been
achieved in its absence. Against that, it does not include unsolved cases which would have been resolved had DNA evidence been present
4
http://www.forensic.gov.uk/html/media/case-studies/f-18.html, accessed April 14th 2009.
27
Keeping the right people on the DNA database: Science and public protection
to their release. The average sentence served by
those exonerated by DNA has been 12 years. In
almost 40 percent of the cases profiled here, the
perpetrator has been identified by DNA testing.
Exonerations have been won in 33 states and
Washington, D.C.4 The characterization of Britain
as a ‘nation of suspects’5 could equally be
represented as ‘a nation of exculpates’. Of
particular interest is the ethnic subtext of the
debate, with GB concerned that DNA testing
exacerbates ethnic disproportionality, while the
Innocence Project experience clearly sees it is a
weapon against racism in criminal justice.
3. Deterrent Effect of Presence on the DNA database. This
rests on a value judgement that career criminality is
usually not the route to happines. The point is thus
acknowledged as speculative. Awareness of having
one’s DNA on the national database will, if
anything, serve to raise the perceived probability
of detection and hence deter. The writer has
sought published work on the deterrent effect of
the DNA database but can find none. It seems
important to remedy this.
Beyond the sheer numbers of those whose conviction
was attended by DNA evidence, the factors which can
perhaps be put into the balance in favour of retention
include the following.
1. Cost. The cost of a murder is estimated at £1.46
million at 2003/04 prices (Home Office 2005). All
murders by offenders beyond the point at which
they could have been subject to successful DNAfacilitated prosecution thus represents cost and
distress savings. Mutatis mutandis, the argument
applies to all other offences. Murder is a particular
case where time to clearance is bimodal, with the
offence either being cleared very quickly or after
considerable time (Stelfox 2006). Savings in
investigation costs can be achieved by swift
clearance, to say nothing of reductions in putative
further offending and attendant public disquiet.
28
2. Witness avoidance of court process. Guilty pleas induced
by the existence of DNA evidence will avoid the
trauma of court appearance and diminish the
scope for witness intimidation, although the power
of DNA evidence in driving pleas requires further
research. (see Briody 2002, 2004, Briody and
Prenzler 2005).
3. Closure. Convictions afford closure to victims and
families.
ThE ImPACT OF DNA DATAbASES.
The S & Marper judgement contends as follows, in
relation to arguments about the relative size of European
DNA databases:
3. The Court cannot …. disregard the fact that,
notwithstanding the advantages provided by comprehensive
extension of the DNA database, other Contracting States have
chosen to set limits on the retention and use of such data with a
view to achieving a proper balance with the competing interests
of preserving respect for private life. The Court observes that the
protection afforded by Article 8 of the Convention would be
unacceptably weakened if the use of modern scientific
techniques in the criminal-justice system were allowed at any cost
and without carefully balancing the potential benefits of the
extensive use of such techniques against important private-life
interests. In the Court’s view, the strong consensus existing
among the Contracting States in this respect is of considerable
importance and narrows the margin of appreciation left to the
respondent State in the assessment of the permissible limits of
the interference with private life in this sphere. The Court
considers that any State claiming a pioneer role in the
development of new technologies bears special responsibility for
striking the right balance in this regard.
In this context it seems apposite to review briefly some
literature on database effects, with a preamble on
database size. First, the prevalence of convictions in
England and Wales (see Farrington 1981) is such that
lifetime retention of profiles from those convicted would
yield a DNA database numbering some 14000000.6 Thus
the scale of potential increase of the database in a form
4
http://www.innocenceproject.org/know/, accessed April 14th 2009.
5
http://www.independent.co.uk/news/uk/crime/growing-dna-database-turning-britain-into-a-nation-of-suspects-422434.html, accessed April 14th 2009.
6
Assuming stable population size and factors making for criminality.
Annex C: DNA Retention after S and Marper: Ken Pease Jill Dando Institute April 2009
unchallenged by the S & Marper judgement is massive.
The size of the current database is a function of how
widespread convictions are within the population, and
not the overreach of the state in decisions about
retention. Second, and useful only in the provision of
context rather than direct relevance, are three exercises
addressed to the impact of DNA databases generally.
The effect of DNA testing of all prison inmates in New
South Wales mandated under the NSW Crimes (Forensic
Procedures) Act (2000)was assessed (Dunsmuir, Tran
and Weatherburn 2009). The effect was on subsequent
clear-up and charge rates for a series of offence types.
Among other things, this Act gave NSW Police the
power to take DNA samples from offenders serving a
sentence of imprisonment for a serious indictable
offence1 in a correctional centre. From January 2001
onwards, NSW Police began testing inmates serving
sentences for serious indictable offences in NSW prisons.
The Dunsmuir et al. work evaluates the effect of this on
clear-up and conviction rates.
Table1 (simplified from Dunsmuir et al) gives a summary
of the significant DNA terms. Cells in the table coloured
solid red are cases where there is a statistically significant
effect but it is negative, in which case the linear DNA
term corresponds to a worsening of the particular rate
indicated. The cells marked stippled yellow are cases were
the linear DNA term has a statistically significant and
positive impact on rates. Blank cells are cases where the
linear DNA term is not statistically significant. It will be
seen that mandatory DNA testing of inmates was
associated with an increase in clear-up and charge rate of
most offence types. The numbers in the cells represent
the lag at which the association is optimized, eg for sexual
assault, the association is closest when measured sixteen
months after the intervention (mandatory DNA testing).
In a UK study with less detailed information about
methodology, Alaster Smith reported that:
“Overall, the national (UK) detection rate for the police
is 23% of recorded crime. When useable DNA is
recovered and loaded onto the DNA database, this
detection rate rises to 43%….In domestic burglary….the
detection rate rises from 15 to 46%; theft from a vehicle
rises from 7 to 61%; and criminal damage increases from
13 to 52%.”(Smith 2004, p. 14).
The most ambitious evaluation of the effects of DNA
availability, Roman, Reid, Reid, Chalfin, Adams and
Knight (2008) conducted a prospective randomised
experiment of the cost-effectiveness of DNA in
investigating high-volume crimes, including residential
burglary, commercial burglary, and theft from
automobiles. 500 crime scenes in five different
communities were studied and cases randomly assigned
to treatment and control conditions, DNA processing as
well as traditional methods were used to investigate the
case. In the control group, biological evidence was not
initially tested, and case outcomes were due only to
traditional investigation. The study found that property
crime cases where DNA evidence was processed had
more than twice as many suspects identified, twice
as many suspects arrested and more than twice as many
cases accepted for prosecution compared with
traditional investigation.
Table 1: Summary of significant linear DNA effects on all outcome measures.
Assault
Clear-up rate
Charge rate
7
27
Sexual
Assault
Rob.
With
Firearm
Rob.
Without
Firearm
B&E
B&E
Dwell.
Other
18
56
47
29
7
19
54
46
30
10
TOMV
TFMV
0
http://www.dnaresource.com/documents/ChicagoPreventableCrimes-Final.pdf, accessed April 16th 2009.
29
Keeping the right people on the DNA database: Science and public protection
A relevant but inadequately reported study is Chicago’s
‘Study on Preventable Crimes’.7 This claims, reporting a
sample of only eight individuals, that 63 violent crimes
(including 53 murders and rapes) could have been
prevented had DNA been sampled upon arrest rather
than conviction. The judgement in S & Marper does not
prohibit search of the DNA database against swabs
taken upon arrest upon arrest, so any prevention
advantage would be retained after adherence to that
judgement. What is at stake in the Marper judgement is
the prompt downstream detection of cases.
ThE PrESENT rESEArCh.
The work consisted of two parts. Their purposes
respectively were:
1. To establish whether the levels of subsequent
criminality of those arrested with no further action
(hereinafter NFA) taken were roughly comparable
to the subsequent criminality of those cautioned
or proceeding to court where the result was a
non-custodial sanction, ie those cases in which
DNA retention was not challenged under S &
Marper. Those given custodial sanctions were
excluded since their period of incapacitation
would reduce the risk period for subsequent
criminality. In the event that the NFA group
showed much less evidence of subsequent
criminality than the groups otherwise dealt with,
the extensive literature on criminal careers could
not be brought to bear. If they showed much more
evidence of subsequent criminality than the other
groups, then comparisons with the literature would
have to be more guarded, The ideal outcome for
present purposes would be if they showed levels
of subsequent criminality similar to those of the
other groups.
2. The second part of the work, carried out in parallel
to the first, involved re-analysis of data from the
Cambridge longitudinal study of delinquent
development on residual career length (RCL) and
residual number of offences (RNO) to give an
indication of the proportion of offences
remaining in criminal careers after the elapse of n
30
years. This is thought to enable anticipation of
what would be lost in potentially DNA- detectable
crime under any deletion policy.
To address the first purpose, groups of those arrested in
the MPS area from whom DNA samples were taken
were identified. These groups comprised those sampled
on 1st June 2004, 1st June 2005 and 1st June 2006. The
data underwent substantial and lengthy editing to exclude
those with earlier arrests or convictions, and to exclude
pseudo-arrests or convictions (where events recorded
subsequently are in fact linked to the taking of a sample
and should not be classed as subsequent). Some
exclusions were made on the basis of lack of clarity as to
the case outcome.
Table 2 summarises the proportion of those dealt with
re-appearing on the PNC data within three specified
periods (depending on the period available from the
sample date). It will be seen that overall, the subsequent
apparent criminality of the NFA group was somewhat
greater than of those given non-custodial sentences,
with the writer wishing to stress the attendant uncertainty
given the modest sample size and the year to year
variation. These figures include subsequent contacts
where the outcome was also NFA so we need also to
looking at the proportion of those with subsequent
contacts at which guilt was admitted or established. The
NFA group proportion (74%) was slightly higher than
the figure for cautions (61%) and slightly lower than the
group given non-custodial sentences (82%). It may also
be helpful to consider the proportion of downstream
offences involving violence. This was also similar in the
three groups (59%, 63% and 64% respectively). As an
incidental finding, thirteen of the NFA group had a
custodial sentence imposed during the follow-up period.
Given that data came only from the first of a month,
and aware that errors of estimation will be magnified by
multiplying the figures to give a monthly total,
nonetheless the suggestion is that some 390 of those
NFA’d over the course of a whole month would have
had a custodial sentence imposed during the
follow-up period.
Annex C: DNA Retention after S and Marper: Ken Pease Jill Dando Institute April 2009
Table 2: % cases arrested again within risk period8
NFA
Caution
Non-Custodial
Sentence
N of cases
1996 sample (30 month risk period)
28
42
23
206
1995 Sample (42 month risk period)
50
47
50
227
1994 Sample (54 month risk period)
48
50
38
99
Combined
40
46
35
532
As will be recalled, the reason for doing the MPS analysis
was to establish whether the data suggested that the
criminal career literature could be applied to the NFA
group. The conclusion is that the NFA group
subsequently shows itself to be roughly as criminal as the
groups with which it was compared. It is thus defensible
to look at the criminal career literature generally to see
how the group might fare over the longer periods which
are only available for examination in the criminal career
literature. Before moving on to do this, a brief discussion
of the counter-intuitive finding of future criminal
equivalence of the NFA group will be attempted.
There are several factors dating from before a criminal
career is embarked upon which are risk factors for
criminality. These include low IQ, parental criminal
record, measures of impulsivity and troublesomeness at
school (see for example West and Farrington 1973 for an
early treatment of the topic). Programmes aimed at the
pre-delinquent phase of development are collectively
known as secondary crime prevention. An example is the
Home Office’s current On Track programme9 whereby
‘inter-agency co-operation is being developed so that
children at risk of offending are identified early and they
and their families provided with consistent services
through the period of the child’s development.’ Indeed
going further back in the sequence, every time a police
officer turns left rather than right while on patrol on the
basis of views about trouble spots, there emerges
differentiated policing by risk, bringing with it differential
risk of arrest. There is perhaps a debate to be had
concerning whether public protection justifies
intervention before any contact with the justice system
but does not after such involvement by dint of arrest.
Arrest is no less a risk factor than even a non-custodial
sentence, it appears.
It is perhaps best to be candid about the writer’s view of
the composition of the arrest-NFA group. It is believed
that this group comprises the genuinely innocent along
with those where further police action would be fruitless
because of intimidation of witnesses and victims, where
street smarts lead to a lack of cooperation with the police
and Crown Prosecution Service which tip the balance in
favour of discontinuation of the investigative process by
dint of evidential insufficiency. Such a composition
would make sense of the subsequent criminality of the
arrest-NFA group.
Some analyses that were impossible with the data to hand
should be mentioned. Central among these is analysis by
ethnicity. If white arrestees in the NFA group had a
higher rate of subsequent criminality than those of other
ethnicities, a case could be made that injustice on racial
grounds was in play. However if there were no difference
by ethnicity in subsequent contact, this could either
represent colour-blind policing or pervasive racism. If
white arrestees in the NFA group had a lower rate of
subsequent criminality than those of other ethnicities, a
case could be made that inverse racism had occurred.
While the pattern of inference would not be
straightforward, analysis of the data by ethnicity would
exclude some possibilities and would have been valuable.
A larger scale exercise along these lines is recommended.
8
The writer was concerned by the smaller number of cases in 2004. His best guess is that the date fell on the day following a Bank Holiday.
9
http://www.crimereduction.homeoffice.gov.uk/crpinit/crpinit.htm, accessed April 16th 2009.
31
Keeping the right people on the DNA database: Science and public protection
A second analysis which was precluded in the data to
hand was the relationship between the offence
precipitating the initial arrest and the seriousness of
subsequent offending. A time-intensive study of part of
the data is being undertaken but would delay delivery of
the present report. The conclusion anticipated with
confidence is that there will be little or no association
between the seriousness of the events, ie that the
seriousness of the initial offence will not predict the
seriousness of subsequent offences. This is the message
of research on the heterogeneity of criminal careers
generally. Its importance is that a policy of selective deletion is
decided upon, based upon the seriousness of the offence leading to
the initial offence, it will lose most of its potential in downstream
detection.
As noted earlier, the analysis reported above was a test of
whether subsequent criminality of the arrest-NFA group
was such as to justify the plausible application of the
criminal careers literature generally. Concluding that it
could, one brings to bear the notions of Residual Career
Length (RCL) and Residual Number of Offences (RNO)
in criminal careers. Self-evidently, these refer to how
long, and involving how many further offences, will an
offender remain criminally active. These questions bear
precisely on the retention issue. Few studies have
attempted to estimate residual career length or residual
number of offences in criminal careers (Pallone and
Hennessy 1993; Piquero, Farrington, and Blumstein
2003; Von Hirsch 1988, 1998). Prediction studies are
limited by the quality of information accessible to
decision-makers in the criminal justice system
(Blumstein, Cohen, Roth, and Visher 1986a,b). Although
official records are an incomplete source of information
32
on offending, they are often all that is available to
decision-makers. Self-report information could perhaps
allow a more accurate prediction of future offending, but
it is seldom available, and offenders are unlikely to
provide accurate self-reports. It is useful to determine
which variables, available in official records, are most
influential in the assessment of the time remaining in
criminal careers. Piquero et al. (Piquero et al. 2003: p. 479)
outlined the policy implications associated with estimates
of residual career length:
Knowledge on career length and residual career
length is perhaps one of the most critical areas of
research that could best inform criminal justice
policies because it deals directly with sentencing and
incapacitation decisions, which are now so strongly
driven by ideology rather than empirical knowledge
The recent and relevant UK research was carried out by
Lila Kazemian in her doctoral work at Cambridge
(Kazemian 2005; see also Kazemian, Farrington,
LeBlanc and Pease 2007). The writer wishes to thank
Dr. Kazemian, now at John Jay College, New York, for
making her data available in a form which addresses the
retention issue. Figure 1 presents the residual number of
offences after particular periods from first official
process, data from the Cambridge longitudinal study and
reworked from the doctoral dissertation of Lila
Kazemian. This suggests (Fig 1) that deletion after six
years would leave just over half of the cohort’s crime yet
to be committed. Drop lines from any data point will
show the proportion for any of the retention periods on
the y axis.
Annex C: DNA Retention after S and Marper: Ken Pease Jill Dando Institute April 2009
Figure 1: Percentage of subsequent criminality lost by retention period
F
100
90
% subsequent criminality
80
70
60
50
40
30
20
10
0
1
3
5
7
9
11
13
15
17
19
21
23
25
Retention Period
As the data points move along the x axis, they are based
on ever fewer numbers of individuals. The RNO would
be less at any number of years if deletion occurred on a
case-by-case basis after an arrest-free period of n years.
Dr. Kazemian has been asked to rework her data to yield
data which permit the issue to be looked at in this way.
SElECTIVE DElETION by AGE Or OFFENCE
SErIOuSNESS.
The literature is for practical purposes unanimous in its
demonstration that early age of onset is a predictor of
longer criminal careers (Ashford and LeCroy 1988;
Blumstein et al. 1982; Blumstein et al. 1986b; Carney
1967; Farrington et al. 1998; Humphrey et al. 1992;
Piquero et al. 2004; Scarpitti and Stephenson 1971;
Silver et al. 2000). Selective deletion on the basis of
youth would yield a greater reduction in the database’s
potential for prospective detection than (for example)
random deletion.
The second alternative candidate for selective deletion is
seriousness of offence involved at the point of first
10
arrest. This rests upon an assumption of homogeneity
within criminal careers, which recently Roach (2009) has
shown to be overestimated by police officers.
This almost certainly reflects a more general and
misguided assumption of homogeneity in the
population. previous research of the more conventional
kind establishes a high degree of versatility in most
criminal careers (for example Blumstein et al. 1988,
Paternoster et al 1998, Bouhana 2003).10 Although
debate exists about the precise level of specialisation
exhibited by offenders, the degree of their versatility in
both offence and method is substantial. It is difficult to
overstate the implications of this for the targeting of
prolific offenders, by forensic and by other means.
Insofar as offenders are versatile, detection in one
offence offers an opportunity for detection in
subsequent offences of other types. The evidence for
this comes from the detection of notorious offenders It
also comes from research showing the high proportion
of those committing trivial offences who are also
involved in more serious offending. Schneider identifies
A serious indictable offender is defined as a person convicted of an offence carrying a maximum penalty of five or more year’s imprisonment
Haesler 2003).
33
Keeping the right people on the DNA database: Science and public protection
whose arrest for being drunk and disorderly led to a
match with saliva from one of his hoax letters (Roach
2009). Most recently, Stuart Cundy, Senior Investigating
Officer n the murder of Sally Anne Bowman, opined
that had a sample been placed on the national database
after a theft offence committed in 2003, [we] could have
identified Sally Anne’s murderer within 24 hours. Instead
it took nearly nine months before Mark Dixie was
identified and almost two and a half years for justice to
be done.”15
the high rate of self-reported shop theft amongst active
burglars. Wellsmith and Guille11 (2005) show the levels of
active criminality in a sample of those repeatedly subject
to fixed penalty notices. Rose12 found little evidence of
specialisation amongst serious traffic offenders,
compared to mainstream offenders. Further,
Sugg observed a wide range of ancillary offending (preand post-conviction) for attendees of probation
motoring programmes.13
In work highly relevant to the context of this paper,
Frederick et al.14 examined the impact of expanding the
Offender Index (equivalent to UK’s Criminal Justice
samples) of New York State. They found, regardless of
the severity of an individual’s first adult offence a high
degree of versatility for all but a minority of offenders
More systematic is an analysis of the offences leading to
a criminal justice sample being taken which subsequently
led to detection of a serious crime. This is shown as
Table 3 below. It will be noted that the offence which led
to DNA being taken which was later available as evidence
in a murder case was most often a drugs offence (in 29
cases), less often a theft offence (in 10 cases) and least
often in an offence of violence. Selectively deleting
DNA samples taken after drug or theft offences would
thus lose the bulk of its evidentiary potential for the
solution of murders.
The degree of versatility among offenders can be
illustrated in a number of ways. Most dramatically, this
can be illustrated anecdotally, from Dick Turpin (arrested
for theft, identified and hung for murder), including the
Yorkshire Ripper detected through his theft of a
numberplate from a Dewsbury scrapyard, to John
Humble, the hoaxer who derailed the Ripper enquiry,
Table 3: Joint frequency distribution of Precursor and Index offences
Index
Precursor
murder
robbery
Sexual
Offences
Violence
Total
Drugs
29
49
26
11
115
Theft Act
10
59
27
9
105
Other
16
32
31
10
89
7
25
19
5
56
62
165
103
35
365
Violence
Total
34
11
M. Wellsmith and H. Guille. Fixed penalty notices as a means of offender selection. International Journal of Police Science and Management 2005;
7(1): 36-43.
12
G. Rose. 2000. The criminal histories of serious traffic offenders. Home Office Research Study 206. London: Home Office.
13
D. Sugg. 1998. Motor Projects in England and Wales: An Evaluation. Home Office Research and Development Statistics Directorate, Research
Findings 81. London: Home Office.
14
B. Frederick. et al. 2002. Expanding the Offender Index of the New York State DNA Data Bank. Policy Research Report. Albany (NY): Division of
Criminal Justice Services.
15
http://news.bbc.co.uk/1/hi/uk/7259494.stm accessed April 16th 2009.
Annex C: DNA Retention after S and Marper: Ken Pease Jill Dando Institute April 2009
Previous research of the more conventional kind
establishes a high degree of versatility in most criminal
careers . Although debate exists about the precise level of
specialisation exhibited by offenders, the degree of their
versatility in both offence and method is substantial
(Blumstein et al. 1988, Paternoster et al. 1998). It is
difficult to overstate the implications of this for the
targeting of prolific offenders, by forensic and by other
means. Insofar as offenders are versatile, detection in one
offence offers an opportunity for detection in
subsequent offences of other types. A high proportion
of those committing serious offences are also involved in
more trivial offending. Schneider (2005) identifies the
high rate of self-reported shop theft amongst active
burglars. Wellsmith and Guille (2005) show the levels of
active criminality in a sample of those repeatedly subject
to fixed penalty notices. Rose (2000) found little evidence
of specialisation amongst serious traffic offenders,
compared to mainstream offenders. Further, Sugg (1998)
observed a wide range of ancillary offending (pre- and
post-conviction) for attendees of probation motoring
programmes. In work highly relevant to the context of
this paper, Frederick (2002) examined the impact of
expanding the Offender Index of New York State. They
found, regardless of the severity of an individual’s first
adult offence a high degree of versatility for all but a
minority of offenders
STANDING bACk.
The reader’s indulgence is requested for the writer to
make three more general points.
1. Eugenics casts a long shadow. A recent
authoritative book of research on criminal
investigative failures (Rossmo 2009) makes it clear
that there is more cause to worry about cognitive
styles in investigation, and fingerprint
interpretation, than about current DNA usage.
The very different implications of retention of
samples and profiles are conflated in the S &
Marper judgement. If samples were discarded and
profiles retained, legitimate concerns raised in the
judgement about how developments in forensic
genetics could compromise privacy would be
negated, and the judgement in consequence be
much shorter in length and the balance between
the interests of individual privacy and public
protection reconsidered. The crucial distinction
between coding and non-coding DNA sequences
and its implications for safeguards does not feature
in the judgement .
2. While properly absent from the judgement, the
capacity of the police to work round the
restrictions placed upon them by deletion of DNA
samples should not be underestimated. Such
work-arounds are of two kinds. First, speculative
searches of past crime scene samples will still be
permitted under S & Marper. This would probably
lead to less justified arrests of those suspected of a
serious crime (where a crime scene DNA sample is
available) whose profiles have been deleted. There
is a threshold of evidence which triggers arrest.
That threshold will fall if a police officer believes
there to be a chance that an arrest will lead to the
clearance of a serious crime committed earlier.
Second, there will develop forms of words in
police intelligence records which will be used to
reflect the fact that a crime scene and criminal
justice profile had previously matched, although
the criminal justice profile had been deleted.
This will result in de facto retention of DNA
information.
3. There remains an outstanding research
programme which would clarify the issues
surrounding the S & Marper judgement in
particular and the forensic use of DNA profiles
more generally. This would include a larger-scale
replication of the first element of work reported
here, in more than one force, clarifying the
patterns of ethnicity surrounding the taking of
samples in arrest-NFA cases and subsequent
criminality. It would feature a study of public
attitudes generally and of those on the database,
and the offsetting deterrent and stigmatising
effects of presence there. It would include the
overlooked issue of the skill and preferences of
crime scene examiners, since there is reason to
suppose that much forensic evidence is
overlooked. Figure 2 shows unpublished data from
a northern English force indicating that the
number and proportions of DNA and fingerprint
hits varied massively by officer
35
Keeping the right people on the DNA database: Science and public protection
Figure 2: SOCO Performance (6 months)
Figure 2 SOCO Performance (6 months)
70
F/P Idents
60
DNA Hits
Number of Idents
50
40
30
20
10
0
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
R
S
SOCO
Figure 3 shows four contrasting performances of
individual officers. The point being made here is that a
programme of research should look more widely than
the S & Marper issue if it is to realise the potential of
DNA forensic analysis.
Figure 3: SOCO finds by material for 2002
Figure 3 SoCO finds by material for 2002
35
31
30
25
Blood
Cigarettes
Count
20
Saliva
18
15
13
10
8
5
2
2
3
1
1
1
0
SoCO A
36
SoCO B
SoCO C
SoCO D
1
T
U
Annex C: DNA Retention after S and Marper: Ken Pease Jill Dando Institute April 2009
CONCluSIONS
The S & Marper judgement casts the retention issue as
one of balance between the principles of individual
privacy and public protection. This means that there is a
level of public protection conferred by retention of
DNA from arrestees against whom no further action is
taken which would justify that retention. Analysis of a
sample of such cases from the Metropolitan Police force
area suggests that the level of subsequent criminality
over three years is on a par with the subsequent
criminality of those given non-custodial sentences or
cautioned. This is taken to justify analysis of this group
in terms of the extant criminal career literature. This
suggests that a six year retention period would cover half
the residual number of offences (RNOs), ie would lose
half its downstream detective value. Shorter periods
would lose more. It is estimated that a retention period
of some 24 years would be necessary to maximise profile
value in detecting future crime. Selective deletion by
offence seriousness or youth of arrestee would remove
most of the detective value of retention for the arresteeNFA group.
A wider criminological research programme is advocated
to maximise the prospective detective value of forensic
DNA work, some of whose possible elements are
described.
37
Keeping the right people on the DNA database: Science and public protection
rEFErENCES
Ashford, J.B. and C.W. LeCroy (1988) “Predicting
Recidivism: An Evaluation of the Wisconsin Juvenile
Probation and Aftercare Risk Instrument.” Criminal
Justice and Behavior 15:141-151.
Blumstein A. (1988) ‘Specialization and seriousness
during adult criminal careers.’ Journal of Quantitative
Criminology 4: 303-345.
Blumstein, A., J. Cohen, and P. Hsieh (1982) The
Duration of Adult Criminal Careers: Final Report to
National Institute of Justice. Carnegie-Mellon University,
Pittsburgh, PA.
Blumstein A., Cohen J., Roth J.A., and Visher C.A.
(1986a) Criminal Careers and Career Criminals. Vol 1.
Washington DC., National Academy Press.
Blumstein A., Cohen J., Roth J.A., and Visher C.A.
(1986b) “Dimensions of Active Criminal Careers.” in
Criminal Careers and Career Criminals, vol. 1, edited by
A. Blumstein. Washington, DC: National Academy Press.
Bouhana N. (2003) Profiling Arson. Unpublished
doctoral thesis, Cambridge Institute of Criminology.
Carney, F.J. (1967) “Predicting Recidivism in a Medium
Security Correctional Institution.” Journal of Criminal
Law, Criminology and Police Science 58:338-348.
Dror I.E. (2009) ‘Cognitive Biases in Human Perception,
Judgement and Decision Making: Bridging Theory and
the Real World’. In D.K.Rossmo (ed) Criminal
Investigative Failures. New York: Taylor and Francis.
38
Frederick. B. (2002) Expanding the Offender Index of
the New York State DNA Data Bank. Policy Research
Report. Albany (NY): Division of Criminal Justice
Services
Humphrey, C., P. Carter, and K. Pease. 1992. “A
Reconviction Predictor for Probationers.” British Journal
of Social Work 22:33-46.
Kazemian L. (2005) A Comparative Analysis of the
Duration of Criminal Careers and Desistance from
Crime. D.Phil thesis, University of Cambridge.
Kazemian, L., Le Blanc, M., Farrington, D.P., and Pease,
K. (2007) Patterns of residual criminal careers among a
sample of adjudicated French-Canadian males. Canadian
Journal of Criminology and Criminal Justice, 49(3),
307-340
Nickerson R.S. (1998) ‘Confirmation Bias: A Ubiquitous
Phenomenon in Many Guises.’ Review of General
Psychology, 2, 175-220.
Nuffield Council on Bioethics (2007) The forensic use
of bioinformation: ethical issues. London: Nuffield
Council on Bioethics.
Pallone, N.J. and J.J. Hennessy. (1993) “Tinderbox
Criminal Violence: Neurogenic Impulsivity, Risk-Taking,
and the Phenomenology of Rational Choice.” Pp.
127-157 in Routine Activity and Rational Choice, vol. 5,
edited by R. V. Clarke and M. Felson. New Brunswick,
NJ: Transaction.
Farrington D.P. (1981) ‘The Prevalence of Convictions.’
British Journal of Criminology 21, 171-175.
Paternoster R. (1998) ‘The Forward Specialization
Coefficient: Distributional Properties and Subgroup
Differences.’ Journal of Quantitative Criminology 14:
133-154.
Farrington D.P. (1985) ‘Predicting Self-Reported and
Official Delinquency.’ In D.P.Farrington and R.Tarling
(eds) Prediction in Criminology. New York: State
University of New York Press.
Piquero, A., D.P. Farrington, and A. Blumstein (2003)
“The Criminal Career Paradigm.” Pp. 359-506 in Crime
and Justice: A Review of Research, vol. 30, edited by M.
Tonry. Chicago: University of Chicago Press.
Farrington, D.P., S. Lambert, and D.J. West (1998)
“Criminal Careers of Two Generations of Family
Members in the Cambridge Study in Delinquent.”
Studies on Crime and Crime Prevention 7:85-106.
Piquero, A., R. Brame, and D. Lynam (2004) “Studying
Criminal Career Length Through Early Adulthood
Among Serious Offenders.” Crime and Delinquency
50:412-435.
Annex C: DNA Retention after S and Marper: Ken Pease Jill Dando Institute April 2009
Rawls J. (1971) A Theory of Justice. Canbridge Mass:
Belknap.
Roach J. (2009) Beyond the Usual Suspects. PhD thesis, Jill Dando Institute, University College London.
Von Hirsch A.. 1998. “Selective Incapacitation: Some
Doubts.” Pp. 121-127 in Principled Sentencing: Readings
on Theory and Policy, edited by A. Von Hirsch and A.
Ashworth. Oxford: Hart Publishing.
Von Hirsch A. (1988) “Selective Incapacitation
Roman, J.K., Reid, S., Reid, J., Chalfin, A., Adams, W., and Reexamined: The National Academy of Sciences’ Report
Knight, C. (2008). The DNA Field Experiment: Cost-
on Criminal Careers and “Career Criminals”.” Criminal
Effectiveness Analysis of the Use of DNA in the Justice Ethics 7:19-35.
Investigation of High-Volume Crimes, Urban Institute, Justice Policy Centre, Washington. Wellsmith M. and Guille H. (2005) ‘Fixed penalty notices
as a means of offender selection.’ International Journal
Rose G.N.G. (2000) The criminal histories of serious
of Police Science and Management 7(1): 36-43
traffic offenders. Home Office Research Study 206. London: Home Office
West D.J. and Farrington D.P. (1973) Who Becomes
Delinquent? London: Heinemann.
Scarpitti, F.R. and R.M. Stephenson. 1971. “Juvenile
Court Dispositions: Factors in the Decision-Making Process.” Crime and Delinquency 17:142-151.
Schneider J. (2005) ‘The Link Between Shoplifting And Burglary: The Booster Burglar.’ British Journal of
Criminology, 2005: 45(3): 395 401.
Silver, E., W.R. Smith, and S. Banks (2000) “Constructing Actuarial Devices for Predicting Recidivism: A Comparison of Methods.” Criminal Justice and
Behavior 27:733-764.
Smith, A. (2004) Programme Delivery and the Impact on
Combating Crime and practical implementation issues.
In Beyond DNA in the UK: Integration and
Harmonisation, Forensic Science Conference Proceedings, (M. Townsley and G. Laycock eds). 17-19th
May, 2004, Jill Dando Institute of Crime Science.
Stelfox P. (2006) Factors Influencing the Outcome of
Homicide Investigations. Doctoral thesis, Open
University.
Sugg D. (1998) Motor Projects in England and Wales: An Evaluation. Home Office Research and Development Statistics Directorate, Research Findings 81. London: Home Office
39
Keeping the right people on the DNA database: Science and public protection
Annex D: Impact
Assessment
Summary: Intervention & Options
Department /Agency:
Title:
Home Office
Impact Assessment of the implementation of the S & Marper ECtHR
judgement regarding DNA profile and sample retention.
Stage: Partial 1A
Version: 0.1
Date: 30 April 2009
Related Publications: Annual Report National DNA database: http://www.npia.police.uk/en/11403.htm
Available to view or download at:
http://www.ukba.homeoffice.gov.uk/aboutus/consultations/current
Contact for enquiries: [email protected]
Telephone: 020 7035 4848
What is the problem under consideration? Why is government intervention necessary?
On the 4th December 2008 the European Court of Human Rights (“the ECtHR”) ruled that there had been
a violation of Article 8 of the Convention in the retention of fingerprints, samples and DNA profiles of two
applicants. The ruling decided that the applicant’s right to privacy had been breached and required their
information to be removed from the DNA database and all samples destroyed. Government intervention is
needed to take consideration of this judgement whilst ensuring that public protection is maintained.
What are the policy objectives and the intended effects?
• Create
a new model for managing the DNA database which ensures maximum public protection through early detection whilst complying with the Strasbourg ruling.
• Create a regime which is overall proportionate and workable. • Put in place a governance structure which is open, transparent and accountable
The intended effect of this policy is to ensure the maximum number of detections where a crime scene is matched to an individual on the database thus helping reduce crime and protect the public whilst at the
same time ensuring the right to privacy is protected.
40
Annex D: Impact Assessment
What policy options have been considered? Please justify any preferred option.
Profiles: (1) Do nothing (2) Immediate automatic deletion following decision to not prosecute (3) Delete
after six years if not re-arrested (3a1) Automatically delete records (3a2) Manual check identity before delete (3b) Automatic deletion after 12 year period for those non-convictions in connection with sexual or violent offences, all other non convictions after 6 year period. (3c) Deletion upon request from individual
concerned after 6 year period. Sampling: (A) Do nothing (B) Immediate destruction of all samples once the
profile has been successfully developed (C) Retain samples in line with retention of profiles. Fingerprints: (I)
Do nothing (II) In line with DNA profile policy (III) Deletion after 15 years. Preferred option is 3b, B and III –
this is considered the best balance between detection and privacy.
When will the policy be reviewed to establish the actual costs and benefits and the
achievement of the desired effects? There will be a review in light of the consultation before
secondary legislation.
Ministerial Sign-off For Consultaion stage Impact Assessments:
I have read the Impact Assessment and I am satisfied that, given the available evidence, it
represents a reasonable view of the likely costs, benefits and impact of the leading options
Signed by the responsible Minister: Date: XXXXXXX 2009
41
Keeping the right people on the DNA database: Science and public protection
Summary: Analysis & Evidence
Policy Option: 2
Description: Immediate automatic deletion following decision to
discontinue investigation.
ANNUAL COSTS
One-off (Transition)
Yrs
COSTS
£15,000
Description and scale of key monetised costs by ‘main affected
groups’
Public sector: £15,000 (set up costs of automated computer system).
1
Average Annual Cost
(excluding one-off)
£0
Total Cost (PV) £15,000
20
Other key non-monetised costs by ‘main affected groups’
Increased cost for law enforcement agencies.
ANNUAL BENEFITS
One-off
BENEFITS
£0
Yrs
Description and scale of key monetised benefits by ‘main affected
groups’
1
Average Annual Benefit
(excluding one-off)
£0
20
Total Benefit (PV) £0
Other key non-monetised benefits by ‘main affected groups’
This option leads to the fewest samples being retained. The average annual number of detections is estimated to fall by 6,241 compared to the “Do nothing” case. Right to privacy.
Key Assumptions/Sensitivities/Risks
Risk that an automated system may delete the wrong record leading to potential miscarriages of justice
of missed opportunities to detect crime. Assumption that the number of profiles created stays constant
over time.
Price Base
Year
2009
42
Time Period
Years
20
Net Benefit Range
(NPV)
£-15,000
NET BENEFIT
(NPV Best estimate)
£-15,000
Annex D: Impact Assessment
What is the geographic coverage of the policy/option?
England and Wales
On what date will the policy be implemented?
Which organisation(s) will enforce the policy?
NPIA
What is the total annual cost of enforcement for these organisations?
£
Does enforcement comply with Hampton principles?
Yes/No
Will implementation go beyond minimum EU requirements?
Yes/No
What is the value of the proposed offsetting measure per year?
£ N/A
What is the value of changes in greenhouse gas emissions?
£ N/A
Will the proposal have a significant impact on competition?
No
Annual cost (£-£) per organisation
(excluding one-off)
Micro
N/A
Small
N/A
Are any of these organisations exempt?
Yes/No
Yes/No
Impact on Admin Burdens Baseline (2005 Prices)
Increase of £ 0
Key:
Decrease of £ 0
Annual costs and benefits:
Medium
N/A
N/A
Large
N/A
N/A
(Increase – Decrease)
Net Impact £ 0
(Net) Present Value
43
Keeping the right people on the DNA database: Science and public protection
Summary: Analysis & Evidence
Policy Option:
3a1
Description:
Automated deletion after 6 year period, unless individual concerned
has been re-arrested or convicted during this period.
ANNUAL COSTS
One-off (Transition)
Yrs
COSTS
£15,000
Description and scale of key monetised costs by ‘main affected
groups’
Public sector: £15,000 (set up costs of automated computer system).
1
Average Annual Cost
(excluding one-off)
£0
Total Cost (PV) £ 15,000
20
Other key non-monetised costs by ‘main affected groups’
Has impact on right to privacy which may be seen as a cost to individuals. Increased cost for law
enforcement agencies.
ANNUAL BENEFITS
One-off
Yrs
BENEFITS
£0
Description and scale of key monetised benefits by ‘main
affected groups’
1
Average Annual Benefit
(excluding one-off)
£0
20
Total Benefit (PV) £0
Other key non-monetised benefits by ‘main affected groups’
Average annual number of detections is estimated to fall by 4,505 compared to the “Do Nothing” case.
Key Assumptions/Sensitivities/Risks
Risk of an automated system deleting the wrong record leading to potential for miscarriages of justice
of missed opportunities to detect crime. Assumption that the number of profiles created stays constant
over time.
Price Base
Year
2009
44
Time Period
Years
20
Net Benefit Range
(NPV)
£-15,000
NET BENEFIT
(NPV Best estimate)
£-15,000
Annex D: Impact Assessment
What is the geographic coverage of the policy/option?
England and Wales
On what date will the policy be implemented?
Which organisation(s) will enforce the policy?
NPIA
What is the total annual cost of enforcement for these organisations?
£0
Does enforcement comply with Hampton principles?
Yes/No
Will implementation go beyond minimum EU requirements?
Yes/No
What is the value of the proposed offsetting measure per year?
N/A
What is the value of changes in greenhouse gas emissions?
N/A
Will the proposal have a significant impact on competition?
Annual cost (£-£) per organisation
(excluding one-off)
Micro N/A
Small
N/A
Are any of these organisations exempt?
Yes/No
Yes/No
Impact on Admin Burdens Baseline (2005 Prices)
Increase of £0
Key:
Decrease of
Annual costs and benefits:
Medium
N/A
N/A
Large
N/A
N/A
(Increase – Decrease)
Net Impact £0
(Net) Present Value
45
Keeping the right people on the DNA database: Science and public protection
Summary: Analysis & Evidence
Policy Option:
3a2
Description:
Manual deletion after 6 year period, unless individual concerned has
been re-arrested or convicted during this period.
ANNUAL COSTS
One-off (Transition)
Yrs
COSTS
£100,000
1
Description and scale of key monetised costs by ‘main affected
groups’
Public sector: £15,000 (additional programming costs), £85,730,000
(review costs), £345,000 (training for those conducting reviews).
Average Annual Cost
(excluding one-off)
£4,300,000
Total Cost (PV) £ 52,170,000
20
Other key non-monetised costs by ‘main affected groups’
Has impact on right to privacy which may be seen as a cost to individuals. Increased cost for law
enforcement agencies.
ANNUAL BENEFITS
One-off
Yrs
BENEFITS
£0
Description and scale of key monetised benefits by ‘main
affected groups’
1
Average Annual Benefit
(excluding one-off)
£0
20
Total Benefit (PV) £0
Other key non-monetised benefits by ‘main affected groups’
Average annual number of detections is estimated to fall by 4,505 compared to the “Do Nothing”
case.
Key Assumptions/Sensitivities/
May be considered not to be compliant with the S & Marper judgement, it may be difficult to match to PNC records, any error in data deletion could lead to miscarriages of justice. Assumption that the number
of profiles created stays constant over time.
Price Base
Year
2009
46
Time Period
Years
20
Net Benefit Range
(NPV)
£-52,170,000
NET BENEFIT
(NPV Best estimate)
£-52,170,000
Annex D: Impact Assessment
What is the geographic coverage of the policy/option?
England and Wales
On what date will the policy be implemented?
Which organisation(s) will enforce the policy?
NPIA
What is the total annual cost of enforcement for these organisations?
£
Does enforcement comply with Hampton principles?
Yes/No
Will implementation go beyond minimum EU requirements?
Yes/No
What is the value of the proposed offsetting measure per year?
£ N/A
What is the value of changes in greenhouse gas emissions?
£ N/A
Will the proposal have a significant impact on competition?
No
Annual cost (£-£) per organisation
(excluding one-off)
Micro N/A
Small
N/A
Are any of these organisations exempt?
Yes/No
Yes/No
Impact on Admin Burdens Baseline (2005 Prices)
Increase of £0
Key:
Decrease of £0
Annual costs and benefits:
Medium
N/A
N/A
Large
N/A
N/A
(Increase – Decrease)
Net Impact £0
(Net) Present Value
47
Keeping the right people on the DNA database: Science and public protection
Summary: Analysis & Evidence
Policy Option:
3b
Description:
Special Exemptions for violent and sexual offenders and also for
children.
ANNUAL COSTS
One-off (Transition)
Yrs
COSTS
£15,000
Description and scale of key monetised costs by ‘main affected
groups’
Public sector: £15,000 (set up costs of automated computer system).
1
Average Annual Cost
(excluding one-off)
£0
Total Cost (PV) £ 15,000
20
Other key non-monetised costs by ‘main affected groups’
Increased costs to law enforcement agencies.
ANNUAL BENEFITS
One-off
Yrs
BENEFITS
£0
1
Description and scale of key monetised benefits by ‘main
affected groups’
–
Average Annual Benefit
(excluding one-off)
£0
20
Total Benefit (PV) £0
Other key non-monetised benefits by ‘main affected groups’
This option is likely to lead to only a slight increase in detections over the automated case since violent and sexual offenders only make up 7.25% of those on the NDNAD, however these individuals are likely to cause the most harm. Children will receive additional rights to privacy.
Key Assumptions/Sensitivities/Risks
Risk that an automated system may delete the wrong record leading to potential miscarriages of justice of
missed opportunities to detect crime.
Price Base
Year
2009
48
Time Period
Years
20
Net Benefit Range
(NPV)
£-15,000
NET BENEFIT
(NPV Best estimate)
£-15,000
Annex D: Impact Assessment
What is the geographic coverage of the policy/option?
England and Wales
On what date will the policy be implemented?
Which organisation(s) will enforce the policy?
NPIA
What is the total annual cost of enforcement for these organisations?
£
Does enforcement comply with Hampton principles?
Yes/No
Will implementation go beyond minimum EU requirements?
Yes/No
What is the value of the proposed offsetting measure per year?
£ N/A
What is the value of changes in greenhouse gas emissions?
£ N/A
Will the proposal have a significant impact on competition?
No
Annual cost (£-£) per organisation
(excluding one-off)
Micro N/A
Small
N/A
Are any of these organisations exempt?
Yes/No
Yes/No
Impact on Admin Burdens Baseline (2005 Prices)
Increase of £0
Key:
Decrease of £0
Annual costs and benefits:
Medium
N/A
N/A
Large
N/A
N/A
(Increase – Decrease)
Net Impact £0
(Net) Present Value
49
Keeping the right people on the DNA database: Science and public protection
Summary: Analysis & Evidence
Policy Option:
3c
Description:
Deletion upon request from individual concerned after a 6 year period,
unless individual concerned has been re-arrested or convicted during
this period.
ANNUAL COSTS
One-off (Transition)
COSTS
£110,000
Yrs
Description and scale of key monetised costs by ‘main affected
groups’
Public sector: £10,025,000 (advertising), £345,000 (training).
1
Average Annual Cost
(excluding one-off)
£515,000
20
Total Cost (PV) £ 7,385,000
Other key non-monetised costs by ‘main affected groups’
Correspondence costs for individuals and those in the public sector. Increased costs for law
enforcement agencies.
ANNUAL BENEFITS
One-off
BENEFITS
£0
Yrs
Description and scale of key monetised benefits by ‘main
affected groups’
1
Average Annual Benefit
(excluding one-off)
£0
20
Total Benefit (PV) £0
Other key non-monetised benefits by ‘main affected groups’
Average annual number of detections is estimated to fall by between 2,050 and 4,023 compared to the “Do Nothing” case.
Key Assumptions/Sensitivities/Risks
May be considered not to be compliant with the S & Marper judgement, it may be difficult to match to PNC records, any error in data deletion could lead to miscarriages of justice. Assumption that the number
of profiles created stays constant over time.
50
Annex D: Impact Assessment
Price Base
Year
2009
NET BENEFIT
(NPV Best estimate)
£-7,385,000
Net Benefit Range
(NPV)
£-7,385,000
Time Period
Years
20
What is the geographic coverage of the policy/option?
England and Wales
On what date will the policy be implemented?
Which organisation(s) will enforce the policy?
NPIA
What is the total annual cost of enforcement for these organisations?
£0
Does enforcement comply with Hampton principles?
Yes/No
Will implementation go beyond minimum EU requirements?
Yes/No
What is the value of the proposed offsetting measure per year?
£ N/A
What is the value of changes in greenhouse gas emissions?
£ N/A
Will the proposal have a significant impact on competition?
No
Annual cost (£-£) per organisation
(excluding one-off)
Micro N/A
Small
N/A
Are any of these organisations exempt?
Yes/No
Yes/No
Impact on Admin Burdens Baseline (2005 Prices)
Increase of £0
Key:
Decrease of £0
Annual costs and benefits:
Medium
N/A
N/A
Large
N/A
N/A
(Increase – Decrease)
Net Impact £0
(Net) Present Value
51
Keeping the right people on the DNA database: Science and public protection
Summary: Analysis & Evidence
Policy Option:
B
Description:
Immediate destruction of all samples once the profile has been
successfully developed.
ANNUAL COSTS
One-off (Transition)
Yrs
COSTS
£0
1
-£9,500,000 (refrigeration savings) which would be loss of earnings to those companies.
Average Annual Cost
(excluding one-off)
£4,120,000
Description and scale of key monetised costs by ‘main affected
groups’
Public sector: £91,930,000 (destruction of DNA samples) which
would be a gain in earnings to those companies involved,
Total Cost (PV) £ 58,685,000
20
Other key non-monetised costs by ‘main affected groups’
Future testing methods cannot be used on these samples. Second sample can no longer be used during an investigation for a familial test. It may lead to a few cases may be lost domestically and abroad. Cost of destruction of gels and other records held by suppliers.
ANNUAL BENEFITS
BENEFITS
One-off
Yrs
£0
1
Description and scale of key monetised benefits by ‘main
affected groups’
–
Average Annual Benefit
(excluding one-off)
£0
20
Total Benefit (PV) £0
Other key non-monetised benefits by ‘main affected groups’
Right to privacy.
Key Assumptions/Sensitivities/Risks
Assumption that the number of samples taken does not change over time.
Price Base
Year
2009
52
Time Period
Years
20
Net Benefit Range
(NPV)
£-58,685,000
NET BENEFIT
(NPV Best estimate)
£-58,685,000
Annex D: Impact Assessment
What is the geographic coverage of the policy/option?
England and Wales
On what date will the policy be implemented?
Which organisation(s) will enforce the policy?
NPIA
What is the total annual cost of enforcement for these organisations?
£
Does enforcement comply with Hampton principles?
Yes/No
Will implementation go beyond minimum EU requirements?
Yes/No
What is the value of the proposed offsetting measure per year?
£ N/A
What is the value of changes in greenhouse gas emissions?
£ N/A
Will the proposal have a significant impact on competition?
No
Annual cost (£-£) per organisation
(excluding one-off)
Micro N/A
Small
N/A
Are any of these organisations exempt?
Yes/No
Yes/No
Impact on Admin Burdens Baseline (2005 Prices)
Increase of £0
Key:
Decrease of £0
Annual costs and benefits:
Medium
N/A
N/A
Large
N/A
N/A
(Increase – Decrease)
Net Impact £0
(Net) Present Value
53
Keeping the right people on the DNA database: Science and public protection
Summary: Analysis & Evidence
Policy Option:
C
Description:
Retain samples in line with retention of profiles.
ANNUAL COSTS
One-off (Transition)
Yrs
COSTS
£0
1
Average Annual Cost
(excluding one-off)
£715,000-955,000
Description and scale of key monetised costs by ‘main affected
groups’
Public sector: £15,395,000 - £20,530,000 (destruction of DNA samples) which would be a gain in earnings to companies involved,
(-)£1,090,000 – (-)£1,450,000 (refrigeration savings) which would be loss of earnings to those companies involved.
Total Cost (PV) £ 8.63m-11.51m
20
Other key non-monetised costs by ‘main affected groups’
Future testing methods cannot be used on these samples. Second sample can no longer be used
during an investigation for a familial test. It may lead to a few cases being lost domestically and
abroad. Cost of destruction of gels and other records held by suppliers (which may be substantial
under the requested option).
ANNUAL BENEFITS
BENEFITS
One-off
Yrs
£0
Description and scale of key monetised benefits by ‘main
affected groups’
1
Average Annual Benefit
(excluding one-off)
£0
20
Total Benefit (PV) £0
Other key non-monetised benefits by ‘main affected groups’
Right to privacy.
Key Assumptions/Sensitivities/Risks
Assumption that the number of samples taken does not change over time.
Price Base
Year
2009
54
Time Period
Years
20
Net Benefit Range
(NPV)
£-(-)8.63m-(-)11.51m
NET BENEFIT
(NPV Best estimate)
£-10.07m
Annex D: Impact Assessment
What is the geographic coverage of the policy/option?
England and Wales
On what date will the policy be implemented?
Which organisation(s) will enforce the policy?
NPIA
What is the total annual cost of enforcement for these organisations?
£
Does enforcement comply with Hampton principles?
Yes/No
Will implementation go beyond minimum EU requirements?
Yes/No
What is the value of the proposed offsetting measure per year?
£ N/A
What is the value of changes in greenhouse gas emissions?
£ N/A
Will the proposal have a significant impact on competition?
No
Annual cost (£-£) per organisation
(excluding one-off)
Micro N/A
Small
N/A
Are any of these organisations exempt?
Yes/No
Yes/No
Impact on Admin Burdens Baseline (2005 Prices)
Increase of £0
Key:
Decrease of £0
Annual costs and benefits:
Medium
N/A
N/A
Large
N/A
N/A
(Increase – Decrease)
Net Impact £0
(Net) Present Value
55
Keeping the right people on the DNA database: Science and public protection
Summary: Analysis & Evidence
Policy Option:
II
Description:
Deletion of fingerprint records in line with DNA profile policy.
ANNUAL COSTS
COSTS
One-off (Transition)
Yrs
£15,000-100,000
1
Description and scale of key monetised costs by ‘main affected
groups’
Public sector: £15,000 (additional programming costs), £0 -£155 million (review costs), £13.6 million (destruction of paper record).
Average Annual Cost
(excluding one-off)
£680,000-8,520,000
Total Cost (PV) £ 9.5m-100.7m
20
Other key non-monetised costs by ‘main affected groups’
ANNUAL BENEFITS
One-off
Yrs
BENEFITS
£0
Description and scale of key monetised benefits by ‘main
affected groups’
1
Average Annual Benefit
(excluding one-off)
£0
20
Total Benefit (PV) £0
Other key non-monetised benefits by ‘main affected groups’
Right to privacy. Average annual number of detections is estimated to fall by between 12,004 and
26,640 compared to the “Do Nothing” case.
Key Assumptions/Sensitivities/Risks
Assumption that the number of samples taken does not change over time.
Price Base
Year
2009
56
Time Period
Years
20
Net Benefit Range
(NPV)
£(-)9.5m-(-)100.7m
NET BENEFIT
(NPV Best estimate)
£-55.1m
Annex D: Impact Assessment
What is the geographic coverage of the policy/option?
England and Wales
On what date will the policy be implemented?
Which organisation(s) will enforce the policy?
NPIA
What is the total annual cost of enforcement for these organisations?
£
Does enforcement comply with Hampton principles?
Yes/No
Will implementation go beyond minimum EU requirements?
Yes/No
What is the value of the proposed offsetting measure per year?
£ N/A
What is the value of changes in greenhouse gas emissions?
£ N/A
Will the proposal have a significant impact on competition?
No
Annual cost (£-£) per organisation
(excluding one-off)
Micro N/A
Small
N/A
Are any of these organisations exempt?
Yes/No
Yes/No
Impact on Admin Burdens Baseline (2005 Prices)
Increase of £0
Key:
Decrease of £0
Medium
N/A
N/A
Large
N/A
N/A
(Increase – Decrease)
Net Impact £0
Annual costs and benefits: Constant Prices
(Net) Present Value
57
Keeping the right people on the DNA database: Science and public protection
Summary: Analysis & Evidence
Policy Option:
III
Description:
Automatic deletion of fingerprint records after 15 years.
ANNUAL COSTS
One-off (Transition)
Yrs
COSTS
£15,000
1
Description and scale of key monetised costs by ‘main affected
groups’
Public sector: £15,000 (additional programming costs), £13.6 million
(destruction of paper record).
Average Annual Cost
(excluding one-off)
£680,000
Total Cost (PV) £ 9,535,000
20
Other key non-monetised costs by ‘main affected groups’
ANNUAL BENEFITS
One-off
Yrs
BENEFITS
£0
Description and scale of key monetised benefits by ‘main
affected groups’
1
Average Annual Benefit
(excluding one-off)
£0
20
Total Benefit (PV) £0
Other key non-monetised benefits by ‘main affected groups’
Right to privacy. Average annual number of detections is estimated to fall by 5,797 compared to the
“Do Nothing” case.
Key Assumptions/Sensitivities/Risks
Assumption that the number of samples taken does not change over time.
Price Base
Year
2009
58
Time Period
Years
20
Net Benefit Range
(NPV)
£-9,535,000
NET BENEFIT
(NPV Best estimate)
£-9,535,000
Annex D: Impact Assessment
What is the geographic coverage of the policy/option?
England and Wales
On what date will the policy be implemented?
Which organisation(s) will enforce the policy?
NPIA
What is the total annual cost of enforcement for these organisations?
£
Does enforcement comply with Hampton principles?
Yes/No
Will implementation go beyond minimum EU requirements?
Yes/No
What is the value of the proposed offsetting measure per year?
£ N/A
What is the value of changes in greenhouse gas emissions?
£ N/A
Will the proposal have a significant impact on competition?
No
Annual cost (£-£) per organisation
(excluding one-off)
Micro N/A
Small
N/A
Are any of these organisations exempt?
Yes/No
Yes/No
Impact on Admin Burdens Baseline (2005 Prices)
Increase of £0
Key:
Decrease of £0
Medium
N/A
N/A
Large
N/A
N/A
(Increase – Decrease)
Net Impact £0
Annual costs and benefits: Constant Prices
(Net) Present Value
59
Keeping the right people on the DNA database: Science and public protection
Evidence Base (for summary sheets)
Background
1) DNA samples obtained for analysis from the
collection of DNA at crime scenes and from samples
taken from individuals in police custody are currently
sent to a forensic laboratory where they are read
in a process known as profiling. This profile is held
on the national DNA database. This is a key police
intelligence tool.
It helps to:
• quickly identify offenders • make earlier arrests • secure more convictions • provide critical investigative leads for police investigations.
2) The core legislation that underpins the taking
of samples and retention of DNA information is the
Police and Criminal Evidence Act (PACE). PACE was
amended in 2001 by the Criminal Justice and Police
Act, which removed the obligation to destroy the
DNA in the event of there being no prosecution,
or an acquittal, as long as the samples had been
lawfully taken. The Criminal Justice Act 2003 further
amended PACE to allow the police to take DNA and
fingerprints without consent from anyone arrested
for a recordable offence and detained in a police
station. Entry on to the DNA Database does not
signify a criminal record.
3) Taking a DNA sample and fingerprints from
someone who has been arrested for a recordable
offence and detained in a police station is now part
of the normal process when someone is taken into
police custody. The benefits of DNA are not only
in detecting serious crime and preventing further
offending, but also in eliminating the innocent from
inquiries.
60
4) Between April 1998 and September 2008, there
were over 390,000 crimes with DNA matches,
providing the police with a lead on the identity of
the possible offender. More recently in 2007-08, 17,614 crimes were detected in which a DNA match
was available. They included 83 ‘homicides’ (this
includes murder and manslaughter) and 184 rapes.
Sampling people who have been arrested but not
proceeded against has yielded a match with a crime
scene in over 3,000 offences including 37 murders,
16 attempted murders and 90 rapes.
Problem under consideration and Rationale for
Intervention
5 On the 4th December 2008 the European Court
of Human Rights (ECtHR) ruled that there had been
a violation of Article 8 of the Convention (the right
to a private life) through the indiscriminate retention
of fingerprints, samples and DNA profiles of two
applicants (S & Marper) – hereafter referred to as the S & Marper ruling. The ruling was that the applicant’s right to privacy had been breached and required their
information to be removed from the DNA database
and all samples destroyed. The Court found that the
blanket policy of retaining the fingerprints and DNA
of people who had been arrested but not convicted
or against whom no further action was taken was in
breach of Article 8 of the European Convention on
Human Rights (ECHR) for those people.
6) However, the Court did also indicate that it
agrees with the Government that the retention
of fingerprint and DNA information pursues the
legitimate purpose of the detection, and therefore,
prevention of crime. The Court also recognised the
need for “an approach which discriminated between
different kinds of case and for the application of
strictly defined storage periods for data”.
Annex D: Impact Assessment
7) Government intervention is needed to take
consideration of this judgement whilst ensuring
that public protection is maintained. Whilst the
judgement makes it clear that the regulations
for storing information on the database need to
change it is important to ensure that the database
continues to provide the maximum support to law enforcement possible whilst ensuring that
individual’s rights to privacy are respected. Any new
policy needs to be proportionate and balance the
rights of the individual to privacy against the right to
be protected from serious (and often violent) criminal
offences.
8) The ruling considered both samples and profiles.
When an individual is arrested two swabs are taken
at the police station and are sent of for testing.
One sample is used to read the DNA profile and the
second is stored in case there is a problem reading
the first sample1. The electronic profile created is
then stored on a computer. As such the sample is the
physical product whereas the profile is the electronic
record created. Currently both are stored indefinitely
- the S & Marper ruling applies to both samples and profiles (and also fingerprints).
Policy Objectives and Intended effects
9) In order to address the problem under
consideration it will be necessary to develop a model
for handling the database which has the following
objectives:
• Create a model for the handling of the DNA database which ensures maximum public protection through early detection whilst
complying with the ECtHR ruling.
• Create a regime which is, overall, proportionate and workable.
1
• Putting in place a governance structure which is open, transparent and accountable.
10) The key measurable and intended effect will
be to maximise the number of detections achieved through the use of the database whilst remaining
compliant with the S & Marper ruling. A detection, through the use of the database, occurs when a
sample taken from a crime scene proves to have
useful DNA and this DNA is matched to a currently
held sample on the DNA database.
11) This is not the same as the situation in which
a suspect is arrested on suspicion of committing
a specific crime and his/her DNA is sampled and
compared to a crime scene DNA sample. This
evidential matching is considered to be outside the
S & Marper ruling. The key difference here is that in evidential matching a suspect has already been
identified, in a database detection no suspect has
previously been identified.
12) Nor can it be taken that detection and
conviction are the same thing. Not all detections will
lead to convictions for a variety of reasons. A match
on its own would rarely provide sufficient evidence
for a conviction and there will be cases where the
crime scene DNA does not belong to the guilty
party. Equally a failure to detect a crime using the
DNA database does not imply a failure to convict.
There are many other methods available to police
to identify guilty parties and the crime scene DNA
can be used once such a detection has occurred. It is
therefore also not possible to say how many crimes
would be prevented under different options.
13) For these reasons the measurement of the
benefit of the objectives is considered in terms of
the volume of detections made possible and not
convictions secured.
In addition to the second sample there are also a set of gels which were used in the profiling procedure which contain genetic material and could be
used to re-extract a profile. These would need to be dealt with in line with the sample destruction procedure.
61
Keeping the right people on the DNA database: Science and public protection
Sectors and Groups Affected
14) The main groups who will be affected by this
policy are:
a. Individual members of the public: They will
be affected both through the imposition
such a policy will place on their rights to
privacy and also through the impacts of the
policy on law and order and the detection
and prevention of crime.
b. Private Sector Firms: There are a very small
number of private sector firms (only 5 are
regularly involved) which provide DNA
sampling services to law enforcement
agencies. Whilst there would be no
regulatory burden on these companies
there might be some loss of business
especially regarding the storage of samples.
c. Law Enforcement Agencies: Any model
which leads to fewer profiles being stored
will have an impact on agencies. Even
where other methods of detection are
available these are likely to be more costly
or time consuming than DNA or fingerprint
matching.
Policy options
Profiles:
1. Do Nothing: Continue to hold DNA profiles
indefinitely for all individuals. This would not
be compatible with the S & Marper ruling but is included as a baseline for comparison.
2. Immediate automatic deletion: Following
a decision by the police to discontinue the
investigation any individual not convicted would
have their profile immediately destroyed, no
matter what crime had been committed.
2
62
3. Deletion After 6 Years: A suspect’s profile
would be destroyed after 6 years from the end of
the investigation. The period will re-commence if the person is subject to further arrest before
expiry of the fixed period of retention. There are three sub options here.
3a All individuals deleted:Any individual not
convicted would be deleted, no matter what
crime had been committed.
3a1 Automatic deletion: Profiles would
be deleted using an automated computer
system
3a2 Manual deletion: Profiles would be
deleted following a manual review to
ensure the correct individual had been
identified.
3b Certain Individuals treated separately:
Individuals arrested in relation to a violent or
sexual offence would be subject to a 12 year retention policy after which profiles would
be deleted in line with 3a above. Profiles of
children under the age of 18 would be stored
up to the age of 18 if they have only been
arrested or convicted once otherwise would
be stored in line with 3a1. All individuals
apart from those arrested in relation to
violent or sexual offences would be treated as in option 3a1.
3c Deletion upon individual request:
Individuals who had not been arrested or
convicted2 during the 6 year period would have
the right to have their details removed on request.
The period will re-commence if the person is subject to further arrest before expiry of the fixed period of retention.
Conviction includes all outcomes where a person was found guilty. This would include cautions and the like but would not include minor traffic
offences such as speeding or parking fines.
Annex D: Impact Assessment
Samples:
A Do nothing:Continue to hold DNA samples
indefinitely for all individuals. This would not
be compatible with the S & Marper ruling but is included as a baseline for comparison.
B Immediate destruction:All samples will be
immediately destroyed once the profile has been
successfully developed. The maximum period a sample may be stored for is 6 months or on
completion of the successful profile – whichever
comes first.
C Retain samples in line with retention of
profiles: Whatever the preferred option 1-3 is this would be applied to samples – i.e. delete
after 6 years, delete after 6 years with special
circumstances, or requested deletion after 6 years.
Fingerprints:
I
Do Nothing:Continue to hold fingerprint
indefinitely for all individuals. This would not
be compatible with the S & Marper ruling but is included as a baseline for comparison.
II Deleted in line with profile deletion policy.
Whatever the preferred option 1-3 is this would be applied to fingerprints – i.e. delete
after 6 years, delete after 6 years with special
circumstances or requested deletion after 6 years.
III Delete after 15 years. All finger prints would
be deleted after 15 years no matter what offence
had been committed.
Options review
15) The preferred options are 3b, B and III. This
would involve deletion of all samples as soon as
practically possible and all profiles automatically
at the 6 year mark with the exception of the those 3
individuals previously arrested for serious violent
or sexual offences and children under 18. All fingerprints would be stored for 15 years.
16) The 6 year figure has been chosen based on
evidence of the likelihood of people who have been
arrested , but against whom no further action is
taken, going on to commit further offences. The
data available for research contained those who
had been convicted and also if they had been
reconvicted. Reconviction was assumed to be a
reasonable approximation for those who had not been convicted based on independent research
carried out by the Jill Dando Institute3 which found
that the reoffending rates of those where no
further action was taken is not significantly lower
than for those who were convicted. Our research
on convictions showed that if an individual had
not reoffended after 4 years the risk of them being
reconvicted was approximately the same as the risk of any 16-20 year old male being convicted (1620 year old males are the most likely to commit an
offence). It takes 15 years before the risk is the same
as the general population. This suggested that a
range of between 4 – 15 years is appropriate. More detail on the research can be found in Annex D below. 17) At the same time it was decided not to make any
exemptions for individuals who were only arrested for relatively minor offences. Annex D suggests that most criminals tend to be “generalists” in that
they tend to commit a wide variety of offences. This
means that an individual arrested for a relatively
minor offence is almost as likely to go on to commit
a major offence as someone initially arrested for a
major offence (the only exception to this is for violent and sexual offenders who are discussed more in the annex). As such exempting minor offenders would be inconsistent with storing information on people
arrested for more major offences.
Source: Annex B of Consultation Document.
63
Keeping the right people on the DNA database: Science and public protection
18) The preferred option set is recommended as it
ensures the maximum public protection whilst being compliant with the S & Marper ruling and also taking special account of the need to protect children. At
the same time it provides best value for money to
government of the viable options.
19) It is assessed that this option would lead to a
reduction up to 206,0004 detections and would cost
£119m5 over the next 20 years as compared to the do nothing option.
Legacy
20) The policy for legacy deletion will be in line with
future deletion rules – that is to say that all individuals
who have not been arrested in the previous 6
years, and who do not fall into the violent and
sexual category, will have their samples and profiles identified and deleted in the case of DNA and
profiles deleted after 15 years for fingerprints.
21) However legacy will not be considered in the
assessment of the options below as it is essentially
forward looking and does not consider the issue
of identifying those already on the database who
would need to be removed. This is due to a lack of
data as to the exact numbers involved or costs of finding and removing. Considerable further work
will be required to fully answer this question.
22) It is therefore not possible to say how much
legacy will cost or what impact it will have on
detections at this point.
64
Detail
Costs and Benefits model
23) In order to assess which option set provides the
best value for money it was necessary to construct a
model which allowed for a calculation of the number
of detections under each option including the do
nothing. All costs and benefits were assessed using
a 20 year time horizon, a long time horizon so the
impact beyond the point where the profiles would
be deleted could be considered.
24) The number of detections was calculated using
a simplified model which looked at the flow of
people through the criminal justice system allowing
consideration of what was the probability of each
outcome occurring. For instance there were 580,000
people were arrested and their sample placed on
the database in 20086. Since the conviction rate is
approximately 38% of arrests it was assumed that 197,000 of these people obtained criminal records
and therefore the profile was not eligible for deletion
under any option. Probabilities were then associated
to such factors as getting re-arrested, committing an offence and leaving a usable sample during the
offence.
25) In order to construct the benefit model it was
necessary to construct a number of key probabilities
and assumptions. These were:
• The probability of being convicted of the offence originally arrested for7 is 38%8
4
This is likely to be an overestimate of the number of detections lost because criminals may leave both fingerprints and DNA at the same crime scene.
5
All cost figures presented here are in net present value terms as per Green Book guidance.
6
This is the number of new samples which were new to the database. Most individuals (though not always all) are sampled on arrest but only those who are not already on the database – i.e. have not been arrested before – would be new to the database.
7
This included all convictions, cautions etc – i.e any outcome which includes a guilt verdict.
8
Source: Criminal Statistics England and Wales based upon 1,350,000 arrests for recordable offences and 205,000 cautions, 250,900 convictions at
Magistrate Courts and 61,500 convictions at the Crown Courts.
Annex D: Impact Assessment
• The probability of committing an offence following an arrest, but no further action, is
the same as the probability of committing and
offence following conviction. The probability of
this is 40%9
27) In order to construct the cost model it was
necessary to include a number of key assumptions.
The cost models for all options can be found in Annex B below. The general cost assumptions relevant to all
options are:
• The probability of being matched to the DNA database for a crime other than that which they
were arrested for is 1.4%10
• Where a computer programme has to be modified – to flag record for review, or to automatically
delete records- it is assumed that this will be in the • The probability of being arrested subsequent to no region of £15,000.12
further action being taken is 18%. Also that they • Individuals who have been arrested are assumed tend to offend for a further 10 years.11
to all have associated Police National Computer
• If a person whose details are on the database (PNC) records.
leaves a sample at a crime scene then a match is
• Cost savings of 90p per annum per sample made.
destroyed though a reduction in refrigeration
costs.
26) The model allowed for the calculation of the
total number of detections which would occur
under each option as a result of matches to the
DNA database only(hereafter referred to simply as
detections), the model excludes those detections which would have occurred through other means.
The flow model used can be found in Annex A to the rear of this document whereas more detail of the
model and its results can be found at Annex E. These figures should be taken cautiously since the model
is only approximate and therefore may not be highly accurate. Some of the figures also consider people
who have no further action taken against them to
have similar characteristics to people who have been
convicted (possibly more than once). Whilst there is
research (mentioned above) to support this position
it is likely to mean these figures represent the upper
bound of the likely detection losses. As such all
benefits should be considered to be rough order of
magnitude estimates only.
9
• The cost of destroying a sample is estimated to be about £8.50.
• No growth in the number of DNA and fingerprint samples taken and the number of profiles created.
• The number of suppliers remains constant.
Options
1. Do Nothing – Profiles
Description
28) Under this option there would be no change
from the current process by which DNA profiles
are stored indefinitely by default but requests
can be made to destroy profiles under “special
circumstances” guidance provided to Chief Police
Officers.13 Whilst this option is not compatible with
S & Marper it is included to provide a benchmark against which the other options are assessed.
Source: Re-offending of adults 2006 cohort http://www.justice.gov.uk/publications/reoffendingofadults.htm
10
Source: NPIA Figures
11
Source: Initial estimate based on PNC data.
12
Assumed that it would take one individual 3 months to re-programme the system.
13
So far there have been 868 such requests of which 495 were granted.
65
Keeping the right people on the DNA database: Science and public protection
Costs
29) There are no financial costs to this option over
and above running the database. None of the
options presented below would change this running
cost and therefore this running cost has not been
considered further.
30) This option has a high non monetised cost to
the individual in terms of a loss of privacy. The S & Marper ruling was that such a cost was too high and not proportionate to the corresponding benefit of
increased crime prevention and protection.
Benefits
31) This option leads to the greatest number
of samples being retained on the database and
therefore the greatest number of detections. It also
has the greatest deterrent effect since all criminals
will be aware that once they have been arrested their
profile will be indefinitely retained.
2. Immediate automatic deletion following
decision of no further action.
Description
32) Under this option all individuals would be
removed from the database immediately once the
investigation was concluded and a decision was
reached that no further action would be taken.
There would be no need for an individual to contact
the police to request removal of their profile since
this deletion would be triggered automatically and
the profile would be deleted using an automated
system. Individuals would not be informed of the
deletion since the assumption would be that deletion
had occurred.
Costs
33) The number of records deleted under the
option of immediate automatic deletion following
the decision on no further action was assumed
to be 115,000 per annum. This was calculated by
66
averaging the 900,000 records identified for removal
during the period May 2001 to Dec 2008.
34) Monetised costs for this option relate to the setup of an automated computer system for the public
sector. A change in supplier processes may also be
required and computer software written; however
given current information we have not been able to
monetise this.
35) This option would lead to an improvement in
the individual’s right to privacy over the do nothing
option. The only profiles which would be stored
would be those of convicted criminals and it is
accepted by the ECtHR that the public protection
interest in storing these profiles outweighs any loss in
privacy.
Benefits
36) This option leads to the fewest samples being
retained since deletion occurs immediately for all
individuals for whom no further action is taken
meaning the number of detections is lower under
this option than under any of the other options. This
does not mean that there are no detections made.
On arrest a sample would still be taken and matched
against crimes currently on the database which
would lead to a number of detections being made.
Only if an individual went on to commit a further
offence after being released would the detection
be lost.
Cost and benefit
37) We assessed that the additional costs would
be £15,000 whilst there would be 125,000 less
detections over the 20 year time frame.
Risks
38) There is also a higher risk of an automated
system deleting the wrong record leading to the
potential for miscarriages of justice of missed
opportunities to detect crime.
Annex D: Impact Assessment
3a1 Automated deletion after 6 year period
Description
39) Under this option all appropriate individuals
would be removed from the database 6 years after
their last arrest. There would be no need for an
individual to contact the police to request removal
of their profile as this deletion would be triggered
automatically and the profile would be deleted
using an automated system. Individuals would be
informed in writing at the completion of this exercise that their profile had been removed.
Costs
40) The estimated number of records deleted each
year under the automatic system was estimated at
approximately 115,000 per annum (as described above).
41) Associated costs for this option relate to the
creation of an automated computer system for
suppliers and the public sector. We have not been
able to monetise the costs to the suppliers but
further information has been requested which
should further develop the monetisation of this
option. The cost to the public sector is estimated in
the region of £15,000.
44) Whilst there are no non-monetised benefits of this option over the do nothing case, there is a higher
risk of an automated system deleting the wrong
record compared to a manual system leading to a
higher potential for miscarriages of justice of missed
opportunities to detect crime.
Cost and benefits
45) We assess that the additional costs would
be £15,000 whilst there would be 90,000 fewer
detections.
Risk
46) The risk associated with this option are:
a. Compliance: It may not be considered to be
complaint with S & Marper.
b. Matching records. To delete data there needs to be a match to a PNC record of arrest.
This may be difficult especially considering
a recent ruling by the Information
Commissioner that storing such data for
petty crimes may be excessive. c. Miscarriages of justice: Any error in data deletion could lead to a miscarriage of justice.
3a2 Manual deletion after 6 year period
42) This option would lead to an improvement in
the individual’s right to privacy over the do nothing
option There are no records left which could be
considered to be an additional cost to individual
privacy above the do nothing case.
Benefits
43) This option leads to far fewer samples being
retained compared to the do nothing. That said a
considerable number of profiles are still not deleted
at the end of 6 years due to the assumption that
there is a 55% probability that the individual will have reoffended during this period meaning that
their profile has not been removed.
Description
47) Under this option all individuals would be
removed from the database 6 years after their last
arrest. There would be no need for an individual to
contact the police to request removal of their profile
since this deletion would be triggered automatically
and the profile would be deleted after a manual
review to ensure that the correct individual’s details
had been identified. Individuals would be informed
in writing at the completion of this exercise that their profile had been removed.
67
Keeping the right people on the DNA database: Science and public protection
Costs
48) It is assumed that the same number of records
would be deleted as in the automatic option,
however the costs of deleting these records based
on a review is much more costly. The cost of a review
in this assessment has been based on the proxy of the average cost of a criminal records check. This
is equivalent to about £35.50 per review. Over the
20 year time frame this is estimated to amount to
almost £52 million (in present value terms).
Benefits
49) This model leads to fewer samples being
retained compared to the do nothing option since
deletion occurs in all appropriate cases. That said a
considerable number of profiles are still not deleted
at the end of 6 years due to the assumption that
there is a 55% probability that the individual will have reoffended during this period meaning that
their profile has not been removed.
50) A manual review is likely to be more accurate
and reliable than an automatic review (as described
above) especially where is it difficult to clearly
ascertain the identity of a sample or an associated
Police National Computer (PNC) record.
c. Miscarriages of justice: Any error in data deletion could lead to a miscarriage of justice.
3b Exemptions for certain individuals
Description
53) Under this option those individuals arrested
for, but not convicted of, certain violent or sexual offences (see Annex C for list of offences) would be held for 12 years instead of the 6 proposed for other
offences. In all other respects they would be treated
the same as other offences.
54) At the same time there would also be a special
exemption for children under the age of 18. This would be applied on a “one strike” principal. Any
child who had only been arrested or convicted a
single time would have their profile deleted at the
age of 18 irrespective of what age they were when
they committed the offence. If two or more arrested
occurred then the profile would be stored for 6 years
(unless for a violent or sexual offence when it would be stored for 12).
Risk
55) Treating children separately to adults is
considered appropriate however it is also recognised
that many of the most prolific offenders are in this
age group. Such offenders would still be stored
due to having committed more than one offence
meaning this provision will only impact on children
who are not prolific offenders. Those convicted of
violent of sexual offences will also be excluded from this provision.
52) The risk associated with this option are:
Costs
Cost and benefits
51) We assess that the additional costs would be
£52 million whilst there would be 90,000 fewer
detections.
56) It is assumed that this system would be identical
to the automatic delete option described above. The
costs of developing a system which could identify
b. Matching records. To delete data there needs such individuals and delete to a different time frame
to be a match to a PNC record of arrest.
are assessed as insignificantly different to the costs
This may be difficult especially considering
presented in option 3a1 above.
a recent ruling by the Information
Commissioner that storing such data for
petty crimes may be excessive.
a. Compliance: It may not be considered to be
complaint with S & Marper.
68
Annex D: Impact Assessment
Benefits
57) Research suggests individuals who are arrested
for violent or sexual offences tend to pose a higher risk to society than those charged on other, less
serious offences. As such storing these profiles
for a longer period of time, would allow for an
increased rate of detection to be proportional in
term. Currently 7.25% of those individuals sampled would fall under this category.14 This is likely to lead
to a significant increase in the number of detections
for these offences. DNA also tends to be more
useful in detection of such offences than it is in many
others such as drug dealing for instance. The harm
associated with such crimes is also considerably
greater. Research suggests that a murder does
approximately £1.5m of damage to society compared to £3,200 for household theft.15
58) This option would lead to an improvement in
the individual’s right to privacy over the do nothing
option but would be less of an improvement
compared to options 2 to 3a2 above.
Cost and benefits
59) There is not sufficient data to provide an
assessment of the benefit in terms of detections
of this provision either for either of the special
categories. However both are considered to bring
benefits in terms of public and child protection. The
additional cost over the do nothing option would be
almost £51 million over the 20 year time frame.
Risk
60) The risks associated with this option are that:
a. A youth may be deleted from the database
after only a very short while and then may go
on to reoffend undetected.
b. Some violent or sexual offenders might pose a significantly greater risk even after 12 years
– possibly for life.
3c Deletion upon request from individual
concerned after a 6 year period
Description
61) Under this option the special circumstances
guidance provided to the police would be expanded to allow all individuals the right to have their profile
deleted from the records following six years without being arrested. Deletion would be automatic
following confirmation of the individual’s identity
and deletion would only be rejected if the individual
had been convicted of an offence or had been
arrested again during the last 6 years.
Costs
62) In order to fully inform those who had samples
taken and subsequent profiles developed a
substantive amount of media advertising would be
required. The costs of advertising in this assessment
are based upon building a section on direct.gov.
uk that explains the policy and provides a central gateway for people to apply to have their DNA
removed, search engine optimisation, advertising
on relevant and targeting websites and the design,
print and distribution of leaflets and posters for
frontline partners. Over the 20 year time frame this is
estimated to amount to over £10 million.
63) This cost of the right to privacy under this
option would be considerably less than under the
do nothing option. Whilst there would still be some
cost in terms of loss to privacy it is assessed that this
is proportionate to the need to protect the public
from crime. It also imposes a cost on the individual
in terms of the time and effort needed to contact
the appropriate authority to request deletion of
their profile.
14
Source: NPIA initial estimate.
15
Source: The economic costs of crimes against individuals and households 2003/4 Home Office publication.
69
Keeping the right people on the DNA database: Science and public protection
Benefits
64) This model retains fewer samples than under the
do nothing and this has a subsequent impact on the
number of detections which occur. Under this option
it is assumed that not everyone who is eligible for
removal makes a request to do so. It is not possible
to assess how many people are likely to make such a
request but it was assumed that between 25% and 75% would do so (the same assumption was used in estimating the costs of this option).
Cost benefit ratio
65) It is assessed that the additional cost would
be £10m and it would lead to between 41,000 to
80,000 fewer detections over the 20 years.
Risk
66) The risk associated with this option are:
a. Compliance: It may not be considered to be
complaint with S & Marper.
b. Matching records. To delete data there needs to be a match to a PNC record of arrest.
This may be difficult especially considering
a recent ruling by the Information
Commissioner that storing such data for
petty crimes may be excessive.
c. Miscarriages
of justice: Any error in data deletion could lead to a miscarriage of justice.
A Do nothing Samples
Description
67) Under this option samples would be stored
indefinitely following initial collection. This option is
not compatible with the S & Marper ruling however it is included to provide a base line.
16
70
Source: NPIA data.
Costs
68) There are no financial costs to this option over
and above the current costs of storing the samples.
None of the options presented below would change
this running cost and therefore this running cost has
not been considered.
Benefits
69) The main advantage to storing the second
sample beyond the time needed to develop a profile
is that the second sample can be used during an
investigation for a familial test. This is a technique
which tests to see if the suspect is related to the crime
scene sample. Since investigations can often run for
a year or more this might be needed after the original
sample has been tested.
70) There are approximately 3,000 of these tests run every year16 – or approximately 0.4% of all tests run. It is not know how many of these tests lead to a
conviction.
B Immediate destruction of all samples once
the profile has been successfully developed
Description
71) Under this option the second sample would be
destroyed as soon as a usable profile was created and
quality assurance checks had been made. This time
period would be limited to a maximum of 6 months. All samples would then be destroyed irrespective of
whether the profile was being kept or not.
Costs
72) Samples for all individuals (those for whom no
further action had been taken and those for whom a
conviction had been brought) would be removed as
soon as the profile has been successfully developed.
It is estimated that this would involve the removal
Annex D: Impact Assessment
of approximately 515,000 samples per annum (based on the number of subject profiles loaded and
retained in 2007-08).
73) The only non monetised costs associated with
this option, is that the familial testing would not
be available any longer. That said in certain cases it
might be possible to retest the individual to allow
such a test to be carried out.
74) It was not possible to fully cost all elements of
this option due to a lack of data regarding cost of the
destructions of the gels and other records held by the
companies involved in DNA profiling.
Risk
75) The risks associated with this option are:
a. Future techniques developed allowing for
more information to be extracted. This may be mitigated against in part by re-sampling the individual (though this may not be
practical or appropriate in a large number of
cases)
b. Cases may be lost both domestically and
abroad due to the inability to produce the
original sample. This is mitigated against by
the possibility for retesting the individual
where appropriate.
76) The risks to this option are higher than the
option below as there is not a six year period of retention to help mitigate against them.
Costs and benefits
77) It has not been possible to assess the benefits
of this option however the additional costs for this
option are estimated to be £59 million. This assessed
cost is likely to be significantly less than the realised
cost since it does not allow for the destruction of the
gels used in the testing process.
C Retain samples in line with retention of
profiles.
Description
78) Under this option the second sample would be
destroyed in line with the destruction of profiles. This
would be automatic and triggered by the destruction
of the matching profile.
Costs
79) Samples would be removed in line with profiles.
Therefore under option 3b for example the samples for those for whom no further action has been taken
would be removed after 6 years but the samples of
those convicted or those arrested in connection to
sexual and violent offences could remain indefinitely.
80) It was not possible to fully cost all elements of
this option due to a lack of data regarding cost of the
destructions of the gels and other records held by the
companies involved which would be needed under
this option.
Risk
81) The risks associated with this option are:
a. Future techniques developed allowing for
more information to be extracted. This may be mitigated against in part by re-sampling the individual (though this may not be
practical or appropriate in a large number of
cases)
b. Cases may be lost both domestically and
abroad due to the inability to produce the
original sample. This is mitigated against
by holding for 6 years and also by the
possibility for retesting the individual where
appropriate.
71
Keeping the right people on the DNA database: Science and public protection
Cost and benefits
82) It has not been possible to assess the benefits
of this option however the additional costs for this
option are £12 million. This assessed cost is likely
to be significantly less than the realised cost since it
does not allow for the destruction of the gels used
in the testing process. It is not know what this might
cost but could easily be as high as an additional
£100 million.
I Do nothing fingerprints
Description
83) Under this option fingerprints would be stored
indefinitely following initial collection. Whilst S & Marper primarily focused on the use and storage of DNA, fingerprints were also considered to be
personal data against which indefinite storage was
not deemed appropriate. As such this option is again
considered for comparative purposes only.
Costs
84) There are no financial costs to this option over
and above the current costs of storing fingerprints.
None of the options presented below would change
this running cost and therefore this running cost has
not been considered.
87) Fingerprints are also used for a variety of other
purposes other than the detection of crime, such as
identification of arrestees, Criminal record bureau
checks, immigration status checks, identification
of bodies and missing people etc. The storage of
fingerprints subsequent to arrest can therefore have
benefits to the person to whom the fingerprints
belong.
II Deletion of fingerprint records in line with
DNA profile policy
Description
88) Under this option all appropriate individual’s
fingerprints would be removed from the database in
line with whatever policy option is chosen for DNA
samples. The physical record (i.e. the card with the
original ink fingerprints) would be destroyed as soon
as there was a valid and useable record on IDENT1
(the Fingerprint database system).
Costs
89) The estimated number of records deleted
each year was approximately 210,000 per annum (1.6 million records identified for removal for period
May 2001-Dec 2008).
90) Associated costs for this option relate to the
creation of an automated computer system for the
85) The S & Marper ruling was that the cost of doing public sector (estimated in the region of £15,000),
nothing on the right to privacy was too high and
destruction of the paper record (which is assumed
not proportionate to the corresponding benefit of
to take 8 minutes per case) and training and review
increased crime prevention and protection.
costs under the manual case.
Benefits
86) This option leads to the greatest number of
fingerprints being retained on the database and
therefore the greatest number of detections. It
will also have the greatest deterrent effect since
all criminals will be aware that once they have
been arrested their fingerprints will be indefinitely
retained.
72
91) Compared to the do nothing case there are also
reductions in non monetised costs in the form of
the impact this option has on the right to privacy as
enshrined by the ECHR. Whilst these costs are similar
to DNA storage, fingerprints are not considered
to place such a high cost in terms of privacy due to
the additional, and sometimes beneficial, uses to
fingerprint storage.
Annex D: Impact Assessment
Benefits
removal of their fingerprints as this deletion would
be triggered automatically and would be deleted
92) This option will lead to fewer fingerprints
using an automated system. Individuals would be
being retained compared to the do nothing option
regardless of the preferred retention policy. That said informed in writing at the completion of this exercise a considerable number of profiles are still not deleted that their fingerprints have been removed. The
physical record would be destroyed as soon as there
even by the end of 6 years due to the assumption
that there is a 55% probability that the individual will was a valid and useable record on IDENT1.
have reoffended during this period meaning that
Costs
their record will not have been removed.
96) The estimated number of records deleted each
Cost and benefit
year was estimated at approximately 210,000 per annum (as described above).
93) We assess that the additional costs would be
£9.5 million – £100.7 million whilst it would lead to
between 240,071 and 532,799 fewer detections
over the 20 year period- please note however that some of these detections may also be included in the
DNA assessment (i.e. fingerprints and DNA may be
left at the same crime scene meaning two detections
missed but only one crime committed).
Risk
94) The risk associated with this option are:
a. Compliance: It may not be considered to be
complaint with S & Marper.
b. Miscarriages of justice: Any error in data deletion could lead to a miscarriage of justice.
c. Difficulties created with other areas where
fingerprints are used. For instance it might
not be possible to identify a body.
III Deletion after 15 years
Description
95) Under this option all appropriate individuals (all
individuals including those arrested in connection
with and violent and sexual offences and not convicted or rearrested) would be removed from the
database after 15 years. There would be no need
for an individual to contact the police to request
97) Associated costs for this option relate to the
creation of an automated computer system for the
public sector (estimated in the region of £15,000),
and destruction of the paper record (which is
assumed to take 8 minutes per case).
98) This cost of the right to privacy under this
option would be considerably less than under the
do nothing option. Whilst there would still be some
cost in terms of loss to privacy it is assessed that this is
proportionate to the need to protect the public from
crime. Whilst the longer retention period would have
a greater cost than the previous option again this is
not considered to be as great as for the DNA option.
Benefits
99) This option will lead to fewer fingerprints
being retained compared to the do nothing option
however the effect it would have on detections
would not be great. Research shows that by 15
years the risk of someone reoffending is no greater
than that of the population at large At the same
time a considerable number of fingerprints would
still not deleted even by the end of 15 years as many
arrestees would have been re-arrested during that time. This option therefore delivers greater benefit
in terms of increase detections and reduced risk
compared to option II.
73
Keeping the right people on the DNA database: Science and public protection
Cost and benefits
100) It is assessed that the additional costs would
be £9.5 million whilst it would lead to 116,000
fewer detections- please note however that some of these detections may also be included in the DNA
assessment.
Risk
101) The risk associated with this option are:
a. Compliance: It may not be considered to be
complaint with S & Marper.
b. Miscarriages of justice: Any error in data deletion could lead to a miscarriage of justice.
c. Difficulties created with other areas where
fingerprints are used. For instance it might
not be possible to identify a body.
Q: Have you any further evidence which
would inform the cost or benefit figures
given in paragraphs 24 to 103 above?
Q: Are there any additional substantive risks associated with any of these policy options which have not been identified?
Preferred Option Set
102) The option set which is recommended are that
profiles be automatically deleted after 6 years for all
people except those arrested for violent or sexual offences, samples be destroyed immediately and
fingerprints be stored for 15 years. The different uses
of fingerprints and DNA and the differing impact on
the right to privacy means that it is not considered
incompatible to store fingerprints for longer than
DNA.
74
103) The preferred profile option has been chosen
as it is felt to achieve the right balance between
protecting individuals from the harms associated
with crime whilst also protecting their right to
privacy. It is also believed to be compatible with the
S & Marper ruling.
104) The immediate deletion model was rejected
since it was considered to impose too high a cost in
terms of detections lost and therefore in terms of
public protection. This would not be proportional
to the intrusion upon privacy that the storage for
6 years would impose.
105) The manual deletion option was rejected
because it was not assessed to provide sufficient
value for money when compared to the automatic
deletion model. Whilst the risk was somewhat higher
under this option the difference is not considered to
be significant and does not justify the additional cost
of this option.
106) The retention of samples option was rejected
as it also did not represent value for money. Again
there is a slightly higher risk associated with this
option but these are not considered to be significant
risks and do not justify the spending of additional
money to mitigate.
107) The option of storing fingerprints in line with
DNA samples was rejected since it was felt the
imposition on individual’s right to privacy was less
for fingerprints than for DNA and therefore the non
monetised costs were lower and did not justify the
additional loss in detections.
Annex D: Impact Assessment
Specific Impact Tests: Checklist
Use the table below to demonstrate how broadly you have considered the potential impacts of your
policy options.
Ensure that the results of any tests that impact on the cost-benefit analysis are contained within
the main evidence base; other results may be annexed.
Type of testing undertaken
Results in Evidence Base?
Results annexed?
Competition Assessment
No
No
Small Firms Impact Test
No
No
Legal Aid
No
No
Sustainable Development
No
No
Carbon Assessment
No
No
Other Environment
No
No
Health Impact Assessment
No
No
Race Equality
No
No
Disability Equality
No
No
Gender Equality
No
No
Human Rights
No
No
Rural Proofing
Yes/No
Yes/No
75
Keeping the right people on the DNA database: Science and public protection
Annexes
Annex A: Benefits Model Flow
Diagrams
No Change Proposed
Individual X
Arrested
Sampled
Profiled
Compared
No Match
NFA
Commits Offence
within X period
Sample
Recovered
Detected
No Match
Convicted
Does not commit
offence
Sample Not
Recovered
Detected by
other means
Match Detected
and Convicted
Match Detected
but not Convicted
Commits Offence
after X years
Detected by
other means
Profile Options
Not Detected
Not Detected
Sample stored for
Y months
Sample Options
Destroyed
Sample
needed
Individual
retested
Detection
76
Sample not
needed
Individual not
retested
Detected by
traditional
Not detected
Annexes
Year 1
NFA
Offend
Don’t offend
Leave sample
Detect
Don’t leave
Detect by
Not detected No Action
Offend
Leave
Year 2
Don’t offend
Don’t leave
Detect
Detect by
Not detected
No Action
On database
Removed
Offend
Leave
Detect
Year 3
Don’t offend
Don’t leave
Detect by
Not detected
No Action
Model drawn for 2 year retention and three year appraisal for ease of drawing only.
77
78
Public sector
Public sector
All
Programming costs
Re-adding of profiles that have been
deleted
Total costs
-
All
Number of detections
Total benefits
Public sector
Public sector
All
Additional programming costs
Re-adding of profiles that have been
deleted
Total costs
-
All
Number of detections
Total benefits
Public sector
Public sector
Forensic ser­
vice sector
Public sector
Public sector
All
Review costs
Removal of profile
Training
Training
Re-adding of profiles that have been
deleted
Total
Sector affected
-
All
Type of benefit
Number of detections
Total benefits
Benefits
Sector affected
Public sector
Type of cost
Additional programming costs
Costs
Manual removal (reviewed)
Sector affected
Type of benefit
Benefits
Sector affected
Type of cost
Costs
Automated removal
Sector affected
Type of benefit
Benefits
Sector affected
Type of cost
Costs
Immediate destruction
PROFILE OPTIONS
-90,100
-90,100
Total quantified
benefits
86,093,864
0
346,364
0
0
85,732,500
15,000
Total monetised
costs
-90,100
-90,100
Total quantified
benefits
15,000
0
15,000
Total monetised
costs
-124,811
-124,811
Total quantified
benefits
15,000
0
15,000
Total monetised
costs
0
0
yr0
101,591
0
86,591
Negligible
0
0
15,000
yr0
0
0
yr0
15,000
0
15,000
yr0
0
0
yr0
15,000
0
15,000
yr0
0
0
yr1
0
0
0
0
0
0
0
yr1
0
0
yr1
0
0
Negligible
yr1
-1,153
-1,153
yr1
0
0
0
yr1
0
0
yr2
0
0
0
0
0
0
0
yr2
0
0
yr2
0
0
Negligible
yr2
-2,210
-2,210
yr2
0
0
0
yr2
0
0
yr3
0
0
0
0
0
0
0
yr3
0
0
yr3
0
0
Negligible
yr3
-3,180
-3,180
yr3
0
0
0
yr3
0
0
yr4
0
0
0
0
0
0
0
yr4
0
0
yr4
0
0
Negligible
yr4
-4,071
-4,071
yr4
0
0
0
yr4
0
0
yr5
0
0
0
0
0
0
0
yr5
0
0
yr5
0
0
Negligible
yr5
-4,887
-4,887
yr5
0
0
0
yr5
-4,373
-4,373
yr6
4,099,818
Negligible
17,318
Negligible
Negligible
4,082,500
Negligible
yr6
-4,373
-4,373
yr6
0
Negligible
Negligible
yr6
-5,636
-5,636
yr6
0
Negligible
0
yr6
-5,049
-5,049
yr7
4,099,818
Negligible
17,318
Negligible
Negligible
4,082,500
Negligible
yr7
-5,049
-5,049
yr7
0
Negligible
Negligible
yr7
-6,324
-6,324
yr7
0
Negligible
0
yr7
-5,675
-5,675
yr8
4,099,818
Negligible
17,318
Negligible
Negligible
4,082,500
Negligible
yr8
-5,675
-5,675
yr8
0
Negligible
Negligible
yr8
-6,954
-6,954
yr8
0
Negligible
0
yr8
-6,250
-6,250
yr9
4,099,818
Negligible
17,318
Negligible
Negligible
4,082,500
Negligible
yr9
-6,250
-6,250
yr9
0
Negligible
Negligible
yr9
-7,533
-7,533
yr9
0
Negligible
0
yr9
-6,250
-6,250
yr10
4,099,818
Negligible
17,318
Negligible
Negligible
4,082,500
Negligible
yr10
-6,250
-6,250
yr10
0
Negligible
Negligible
yr10
-7,533
-7,533
yr10
0
Negligible
0
yr10
-6,250
-6,250
yr11
4,099,818
Negligible
17,318
Negligible
Negligible
4,082,500
Negligible
yr11
-6,250
-6,250
yr11
0
Negligible
Negligible
yr11
-7,533
-7,533
yr11
0
Negligible
0
yr11
-6,250
-6,250
yr12
4,099,818
Negligible
17,318
Negligible
Negligible
4,082,500
Negligible
yr12
-6,250
-6,250
yr12
0
Negligible
Negligible
yr12
-7,533
-7,533
yr12
0
Negligible
0
yr12
-6,250
-6,250
yr13
4,099,818
Negligible
17,318
Negligible
Negligible
4,082,500
Negligible
yr13
-6,250
-6,250
yr13
0
Negligible
Negligible
yr13
-7,533
-7,533
yr13
0
Negligible
0
yr13
-6,250
-6,250
yr14
4,099,818
Negligible
17,318
Negligible
Negligible
4,082,500
Negligible
yr14
-6,250
-6,250
yr14
0
Negligible
Negligible
yr14
-7,533
-7,533
yr14
0
Negligible
0
yr14
-6,250
-6,250
yr15
4,099,818
Negligible
17,318
Negligible
Negligible
4,082,500
Negligible
yr15
-6,250
-6,250
yr15
0
Negligible
Negligible
yr15
-7,533
-7,533
yr15
0
Negligible
0
yr15
-6,250
-6,250
yr16
4,099,818
Negligible
17,318
Negligible
Negligible
4,082,500
Negligible
yr16
-6,250
-6,250
yr16
0
Negligible
Negligible
yr16
-7,533
-7,533
yr16
0
Negligible
0
yr16
Annex B: Incremental costs and benefits above the “Do nothing” baseline.
-6,250
-6,250
yr17
4,099,818
Negligible
17,318
Negligible
Negligible
4,082,500
Negligible
yr17
-6,250
-6,250
yr17
0
Negligible
Negligible
yr17
-7,533
-7,533
yr17
0
Negligible
0
yr17
-6,250
-6,250
yr18
4,099,818
Negligible
17,318
Negligible
Negligible
4,082,500
Negligible
yr18
-6,250
-6,250
yr18
0
Negligible
Negligible
yr18
-7,533
-7,533
yr18
0
Negligible
0
yr18
-6,250
-6,250
yr19
4,099,818
Negligible
17,318
Negligible
Negligible
4,082,500
Negligible
yr19
-6,250
-6,250
yr19
0
Negligible
Negligible
yr19
-7,533
-7,533
yr19
0
Negligible
0
yr19
-6,250
-6,250
yr20
28,594,818
Negligible
17,318
Negligible
Negligible
28,577,500
Negligible
yr20
-6,250
-6,250
yr20
0
Negligible
Negligible
yr20
-7,533
-7,533
yr20
0
Negligible
0
yr20
NA
NA
PV
52,169,318
Negligible
254,532
Negligible
Negligible
51,899,786
15,000
PV
NA
NA
PV
15,000
Negligible
15,000
PV
NA
NA
PV
15,000
Negligible
15,000
PV
Keeping the right people on the DNA database: Science and public protection
Public sector
Public sector
All
Additional programming costs
Re-adding of profiles that have been
deleted
Total costs
-
All
Number of detections
Total benefits
Public sector
Public sector
Public sector
Public sector
Private sector
Public sector
Public sector
All
Programming costs
Correspondence for individuals
Removal and correspondence costs
Advertising campaign
Training companies
Training companies
Training
Total
(-)40,993 (-)80,466
All
Total
Public sector
Public sector
Public sector
Public sector
All
Destruction of DNA samples
Extra cost of resampling
Individual notification of destruction
Refrigeration savings
Total
-
All
ALL BENEFITS
Total
Samples in line with profiles
Sector affected
Type
Benefits
Sector affected
Type (short descriptive)
Costs
Immediate destruction of samples
0
0
Total quantified
benefits
82,425,750
-9,501,750
0
0
91,927,500
Total monetised
costs
(-)40,993 (-)80,466
-
SAMPLE OPTIONS
Total quantified
benefits
Sector affected
ALL BENEFITS
10,371,360
346,360
0
0
10,025,000
0
0
0
Total monetised
costs
-90,100
-90,100
Total quantified
benefits
15,000
0
15,000
Total monetised
costs
Type
Benefits
Sector affected
Type (short descriptive)
Costs
Requested manual removal
Sector affected
Type of benefit
Benefits
Sector affected
Type of cost
Costs
Automated removal with special
circumstances
-
-
yr0
1,957,000
-231,750
Negligible
Negligible
2,188,750
yr0
0
0
yr0
111,590
86,590
Negligible
Negligible
25,000
0
0
Negligible
yr0
0
0
yr0
15,000
0
15,000
yr0
-
-
yr1
3,914,000
-463,500
Negligible
Negligible
4,377,500
yr1
0
0
yr1
500,000
0
Negligible
Negligible
500,000
0
0
0
yr1
0
0
yr1
0
0
Negligible
yr1
-
-
yr2
3,914,000
-463,500
Negligible
Negligible
4,377,500
yr2
0
0
yr2
500,000
0
Negligible
Negligible
500,000
0
0
0
yr2
0
0
yr2
0
0
Negligible
yr2
-
-
yr3
3,914,000
-463,500
Negligible
Negligible
4,377,500
yr3
0
0
yr3
500,000
0
Negligible
Negligible
500,000
0
0
0
yr3
0
0
yr3
0
0
Negligible
yr3
-
-
yr4
3,914,000
-463,500
Negligible
Negligible
4,377,500
yr4
0
0
yr4
500,000
0
Negligible
Negligible
500,000
0
0
0
yr4
0
0
yr4
0
0
Negligible
yr4
-
-
yr5
3,914,000
-463,500
Negligible
Negligible
4,377,500
yr5
0
0
yr5
500,000
0
Negligible
Negligible
500,000
0
0
0
yr5
0
0
yr5
0
0
Negligible
yr5
-
-
yr6
3,914,000
-463,500
Negligible
Negligible
4,377,500
yr6
(-)1,730 (-)3,783
(-)1,730 (-)3,782
-
-
yr7
3,914,000
-463,500
Negligible
Negligible
4,377,500
yr7
(-)2,119 (-)4,430
(-)2,119 (-)4,430
yr7
517,318
517,318
yr6
17,318
Negligible
Negligible
500,000
Negligible
Negligible
0
yr7
-5,049
-5,049
yr7
0
Negligible
Negligible
yr7
17,318
Negligible
Negligible
500,000
Negligible
Negligible
0
yr6
-4,373
-4,373
yr6
0
Negligible
Negligible
yr6
-
-
yr8
3,914,000
-463,500
Negligible
Negligible
4,377,500
yr8
(-)2,506 (-)5,037
(-)2,506 (-)5,037
yr8
517,318
17,318
Negligible
Negligible
500,000
Negligible
Negligible
0
yr8
-5,675
-5,675
yr8
0
Negligible
Negligible
yr8
-
-
yr9
3,914,000
-463,500
Negligible
Negligible
4,377,500
yr9
(-)2,886 (-)5,601
(-)2,886 (-)5,601
yr9
517,318
17,318
Negligible
Negligible
500,000
Negligible
Negligible
0
yr9
-6,250
-6,250
yr9
0
Negligible
Negligible
yr9
-
-
yr10
3,914,000
-463,500
Negligible
Negligible
4,377,500
yr10
(-)2,886 (-)5,601
(-)2,886 (-)5,601
yr10
517,318
17,318
Negligible
Negligible
500,000
Negligible
Negligible
0
yr10
-6,250
-6,250
yr10
0
Negligible
Negligible
yr10
-
-
yr11
3,914,000
-463,500
Negligible
Negligible
4,377,500
yr11
(-)2,886 (-)5,601
(-)2,886 (-)5,601
yr11
517,318
17,318
Negligible
Negligible
500,000
Negligible
Negligible
0
yr11
-6,250
-6,250
yr11
0
Negligible
Negligible
yr11
-
-
yr12
3,914,000
-463,500
Negligible
Negligible
4,377,500
yr12
(-)2,886 (-)5,601
(-)2,886 (-)5,601
yr12
517,318
17,318
Negligible
Negligible
500,000
Negligible
Negligible
0
yr12
-6,250
-6,250
yr12
0
Negligible
Negligible
yr12
-
-
yr13
3,914,000
-463,500
Negligible
Negligible
4,377,500
yr13
(-)2,886 (-)5,601
(-)2,886 (-)5,601
yr13
517,318
17,318
Negligible
Negligible
500,000
Negligible
Negligible
0
yr13
-6,250
-6,250
yr13
0
Negligible
Negligible
yr13
-
-
yr14
3,914,000
-463,500
Negligible
Negligible
4,377,500
yr14
(-)2,886 (-)5,601
(-)2,886 (-)5,601
yr14
517,318
17,318
Negligible
Negligible
500,000
Negligible
Negligible
0
yr14
-6,250
-6,250
yr14
0
Negligible
Negligible
yr14
-
-
yr15
3,914,000
-463,500
Negligible
Negligible
4,377,500
yr15
(-)2,886 (-)5,601
(-)2,886 (-)5,601
yr15
517,318
17,318
Negligible
Negligible
500,000
Negligible
Negligible
0
yr15
-6,250
-6,250
yr15
0
Negligible
Negligible
yr15
-
-
yr16
3,914,000
-463,500
Negligible
Negligible
4,377,500
yr16
(-)2,886 (-)5,601
(-)2,886 (-)5,601
yr16
517,318
17,318
Negligible
Negligible
500,000
Negligible
Negligible
0
yr16
-6,250
-6,250
yr16
0
Negligible
Negligible
yr16
-
-
yr17
3,914,000
-463,500
Negligible
Negligible
4,377,500
yr17
(-)2,886 (-)5,601
(-)2,886 (-)5,601
yr17
517,318
17,318
Negligible
Negligible
500,000
Negligible
Negligible
0
yr17
-6,250
-6,250
yr17
0
Negligible
Negligible
yr17
-
-
yr18
3,914,000
-463,500
Negligible
Negligible
4,377,500
yr18
(-)2,886 (-)5,601
(-)2,886 (-)5,601
yr18
517,318
17,318
Negligible
Negligible
500,000
Negligible
Negligible
0
yr18
-6,250
-6,250
yr18
0
Negligible
Negligible
yr18
-
-
yr19
3,914,000
-463,500
Negligible
Negligible
4,377,500
yr19
(-)2,886 (-)5,601
(-)2,886 (-)5,601
yr19
517,318
17,318
Negligible
Negligible
500,000
Negligible
Negligible
0
yr19
-6,250
-6,250
yr19
0
Negligible
Negligible
yr19
-
-
yr20
6,102,750
-463,500
Negligible
Negligible
6,566,250
yr20
(-)2,886 (-)5,601
(-)2,886 (-)5,601
yr20
517,318
17,318
Negligible
Negligible
500,000
Negligible
Negligible
0
yr20
-6,250
-6,250
yr20
0
Negligible
Negligible
yr20
0
0
PV
58,684,338
-6,819,199
Negligible
Negligible
65,503,537
PV
NA
NA
PV
7,385,730
254,529
Negligible
Negligible
7,131,202
Negligible
Negligible
Negligible
PV
NA
NA
PV
15,000
Negligible
15,000
PV
Annexes
79
80
Public sector
Public sector
Public sector
Public sector
All
Destruction of DNA samples
Extra cost of resampling
Individual notification of destruction
Refrigeration savings
Total
-
All
ALL BENEFITS
Total benefits
Public sector
Public sector
Public sector
Public sector
Public sector
All
Additional programming costs
Additional programming costs
Re-adding of fingerprint profiles
Destruction of paper record (without V,
S or T offences)
Destruction of paper record (V, S or T offences)
Total
-
All
ALL BENEFITS
Total
0
8,853,677
Public sector
Public sector
Public sector
Public sector
Forensic
services
Public sector
Public sector
Public sector
Public sector
Review costs (NON V/S/T)
Review costs after 12 years (V/S/T)
Removal of profile
Training
Training
Re-adding of profiles that have been
deleted
Destruction of paper record (without
V/S/T offences)
Destruction of paper record (V/S/T offences)
4,767,365
346,364
0
0
54,794,250
101,760,750
15,000
Sector affected
Additional programming costs
Total monetised
costs
(-)240,071 (-)532,799
(-)240,071 (-)532,799
Total quantified
benefits
13,636,042
4,767,365
8,853,677
0
0
15,000
Total monetised
costs
0
0
Total quantified
benefits
(-)14,308,875 (-)19,078,500
(-)1,086,750 (-)1,449,000
0
0
15,395,625 20,527,500
Total monetised
costs
Type (short descriptive)
Costs
6 years manual
Sector affected
Type
Benefits
Sector affected
Type of cost
Costs
6 years automated
FINGERPRINT OPTIONS
Sector affected
Type of benefit
Benefits
Sector affected
Type of cost
Costs
0
0
0
86,591
Negligible
0
0
0
15,000
yr0
0
0
yr0
119,184
221,342
0
0
Negligible
0
0
0
0
yr1
0
0
yr1
340,526
119,184
0
15,000
221,342
0
Negligible
0
yr1
-
-
yr1
0
0
0
0
0
yr1
0
0
0
15,000
yr0
-
-
yr0
0
0
0
0
0
yr0
238,368
442,684
0
0
Negligible
0
0
0
0
yr2
0
0
yr2
681,052
238,368
442,684
0
Negligible
0
yr2
-
-
yr2
0
0
0
0
0
yr2
238,368
442,684
0
0
Negligible
0
0
0
0
yr3
0
0
yr3
681,052
238,368
442,684
0
Negligible
0
yr3
-
-
yr3
0
0
0
0
0
yr3
238,368
442,684
0
0
Negligible
0
0
0
0
yr4
0
0
yr4
681,052
238,368
442,684
0
Negligible
0
yr4
-
-
yr4
0
0
0
0
0
yr4
238,368
442,684
0
0
Negligible
0
0
0
0
yr5
0
0
yr5
681,052
238,368
442,684
0
Negligible
0
yr5
-
-
yr5
0
0
0
0
0
yr5
238,368
442,684
Negligible
17,318
Negligible
Negligible
0
4,845,750
Negligible
yr6
(-)10,349 (-)26,294
(-)10,349 (-)26,294
yr6
681,052
238,368
442,684
Negligible
Negligible
0
yr6
-
-
yr6
733,125 977,500
0
Negligible
Negligible
733,125 977,500
yr6
238,368
442,684
Negligible
17,318
Negligible
Negligible
0
4,845,750
Negligible
yr7
(-)12,571 (-)30,165
(-)12,571 (-)30,165
yr7
681,052
238,368
442,684
Negligible
17,318
Negligible
Negligible
0
4,845,750
Negligible
yr8
(-)14,754 (-)33,689
(-)14,754 (-)33,689
yr8
681,052
238,368
442,684
442,684
238,368
Negligible
Negligible
0
yr8
-
-
yr8
655,500 874,011
(-)77,625
(-)103,488
Negligible
Negligible
733,125 977,502
yr8
Negligible
Negligible
0
yr7
-
-
yr7
655,500 874,012
(-)77,625
(-)103,487
Negligible
Negligible
733,125 977,501
yr7
238,368
442,684
Negligible
17,318
Negligible
Negligible
0
4,845,750
Negligible
yr9
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr9
681,052
238,368
442,684
Negligible
Negligible
0
yr9
-
-
yr9
655,500 874,010
(-)77,625
(-)103,489
Negligible
Negligible
733,125 977,503
yr9
238,368
442,684
Negligible
17,318
Negligible
Negligible
0
4,845,750
Negligible
yr10
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr10
681,052
238,368
442,684
Negligible
Negligible
0
yr10
-
-
yr10
655,500 874,009
(-)77,625
(-)103,490
Negligible
Negligible
733,125 977,504
yr10
238,368
442,684
Negligible
17,318
Negligible
Negligible
0
4,845,750
Negligible
yr11
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr11
681,052
238,368
442,684
Negligible
Negligible
0
yr11
-
-
yr11
655,500 874,008
(-)77,625
(-)103,491
Negligible
Negligible
733,125 977,505
yr11
238,368
442,684
442,684
238,368
Negligible
17,318
Negligible
Negligible
2,609,250
4,845,750
Negligible
yr13
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr13
681,052
238,368
442,684
Negligible
Negligible
0
yr13
-
-
yr13
655,500 874,006
(-)77,625
(-)103,493
Negligible
Negligible
733,125 977,507
yr13
Negligible
17,318
Negligible
Negligible
2,609,250
4,845,750
Negligible
yr12
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr12
681,052
238,368
442,684
Negligible
Negligible
0
yr12
-
-
yr12
655,500 874,007
(-)77,625
(-)103,492
Negligible
Negligible
733,125 977,506
yr12
238,368
442,684
Negligible
17,318
Negligible
Negligible
2,609,250
4,845,750
Negligible
yr14
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr14
681,052
238,368
442,684
Negligible
Negligible
0
yr14
-
-
yr14
655,500 874,005
(-)77,625
(-)103,494
Negligible
Negligible
733,125 977,508
yr14
238,368
442,684
Negligible
17,318
Negligible
Negligible
2,609,250
4,845,750
Negligible
yr15
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr15
681,052
238,368
442,684
Negligible
Negligible
0
yr15
-
-
yr15
655,500 874,004
(-)77,625
(-)103,495
Negligible
Negligible
733,125 977,509
yr15
238,368
442,684
Negligible
17,318
Negligible
Negligible
2,609,250
4,845,750
Negligible
yr16
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr16
681,052
238,368
442,684
Negligible
Negligible
0
yr16
-
-
yr16
655,500 874,003
(-)77,625
(-)103,496
Negligible
Negligible
733,125 977,510
yr16
238,368
442,684
Negligible
17,318
Negligible
Negligible
2,609,250
4,845,750
Negligible
238,368
442,684
Negligible
17,318
Negligible
Negligible
2,609,250
4,845,750
Negligible
yr18
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr17
(-)16,866 (-)36,888
yr18
681,052
238,368
442,684
Negligible
Negligible
0
yr18
-
-
yr18
655,500 874,001
(-)77,625
(-)103,498
Negligible
Negligible
733,125 977,512
yr18
(-)16,866 (-)36,888
yr17
681,052
238,368
442,684
Negligible
Negligible
0
yr17
-
-
yr17
655,500 874,002
(-)77,625
(-)103,497
Negligible
Negligible
733,125 977,511
yr17
238,368
442,684
Negligible
17,318
Negligible
Negligible
2,609,250
4,845,750
Negligible
yr19
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr19
681,052
238,368
442,684
Negligible
Negligible
0
yr19
-
-
yr19
655,500 874,000
(-)77,625
(-)103,499
Negligible
Negligible
733,125 977,513
yr19
357,552
664,026
Negligible
17,318
Negligible
Negligible
33,920,250
33,920,250
Negligible
yr20
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr20
1,021,578
357,552
664,026
Negligible
Negligible
0
yr20
-
-
yr20
5,054,250 6,739,000
(-)77,625
(-)103,500
Negligible
Negligible
5,131,875 6,842,500
yr20
3,332,530
6,188,984
Negligible
254,532
Negligible
Negligible
29,332,257
61,602,790
15,000
PV
NA
NA
PV
9,536,513
3,332,530
6,188,984
Negligible
Negligible
15,000
PV
0
-
PV
8,630,423 11,507,231
(-)689,609 (-)919,479
Negligible
Negligible
9,320,032 12,426,709
PV
Keeping the right people on the DNA database: Science and public protection
-
All
Total
Public sector
Public sector
Public sector
Public sector
All
Additional programming costs
Additional programming costs
Re-adding of profiles that have been
deleted
Destruction of paper record
Total
-
All
ALL BENEFITS
Total
Public sector
Public sector
Public sector
Forensic
services
Public sector
Public sector
Public sector
All
Additional programming costs
Review costs (NON V/S/T)
Removal of profile
Training
Training
Re-adding of profiles that have been
deleted
Destruction of paper record
Total
-
All
ALL BENEFITS
Total
(-)36,289 (-)115,944
(-)36,289 (-)115,944
Total quantified
benefits
170,537,406
13,621,042
0
346,364
0
0
156,555,000
15,000
Total monetised
costs
(-)36,289 (-)115,944
(-)36,289 (-)115,944
Total quantified
benefits
13,636,042
13,621,042
0
0
15,000
Total monetised
costs
(-)240,071 (-)532,799
(-)240,071 (-)532,799
Total quantified
benefits
170,537,406
0
0
yr0
101,591
0
0
86,591
Negligible
0
0
15,000
yr0
0
0
yr0
15,000
0
0
0
15,000
yr0
0
0
yr0
101,591
0
0
yr1
340,526
340,526
Negligible
0
Negligible
0
0
Negligible
yr1
0
0
yr1
340,526
340,526
0
Negligible
0
yr1
0
0
yr1
340,526
0
0
yr2
681,052
681,052
Negligible
0
Negligible
0
0
Negligible
yr2
0
0
yr2
681,052
681,052
0
Negligible
0
yr2
0
0
yr2
681,052
0
0
yr3
681,052
681,052
Negligible
0
Negligible
0
0
Negligible
yr3
0
0
yr3
681,052
681,052
0
Negligible
0
yr3
0
0
yr3
681,052
0
0
yr4
681,052
681,052
Negligible
0
Negligible
0
0
Negligible
yr4
0
0
yr4
681,052
681,052
0
Negligible
0
yr4
0
0
yr4
681,052
0
0
yr5
681,052
681,052
Negligible
0
Negligible
0
0
Negligible
yr5
0
0
yr5
681,052
681,052
0
Negligible
0
yr5
0
0
yr5
681,052
0
0
yr6
698,370
681,052
Negligible
17,318
Negligible
0
0
Negligible
yr6
0
0
yr6
681,052
681,052
0
Negligible
0
yr6
(-)10,349 (-)26,294
(-)10,349 (-)26,294
yr6
5,544,120
0
0
yr7
698,370
681,052
Negligible
17,318
Negligible
0
0
Negligible
yr7
0
0
yr7
681,052
681,052
0
Negligible
0
yr7
(-)12,571 (-)30,165
(-)12,571 (-)30,165
yr7
5,544,120
0
0
yr8
698,370
681,052
Negligible
17,318
Negligible
0
0
Negligible
yr8
0
0
yr8
681,052
681,052
0
Negligible
0
yr8
(-)14,754 (-)33,689
(-)14,754 (-)33,689
yr8
5,544,120
0
0
yr9
698,370
681,052
Negligible
17,318
Negligible
0
0
Negligible
yr9
0
0
0
0
yr10
698,370
681,052
Negligible
17,318
Negligible
0
0
Negligible
yr10
0
0
yr10
681,052
681,052
yr9
681,052
0
Negligible
0
yr10
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr10
5,544,120
681,052
0
Negligible
0
yr9
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr9
5,544,120
0
0
yr11
698,370
681,052
Negligible
17,318
Negligible
0
0
Negligible
yr11
0
0
yr11
681,052
681,052
0
Negligible
0
yr11
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr11
5,544,120
0
0
yr12
698,370
681,052
Negligible
17,318
Negligible
0
0
Negligible
yr12
0
0
yr12
681,052
681,052
0
Negligible
0
yr12
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr12
8,153,370
0
0
yr13
698,370
681,052
Negligible
17,318
Negligible
0
0
Negligible
yr13
0
0
yr13
681,052
681,052
0
Negligible
0
yr13
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr13
8,153,370
0
0
yr14
698,370
681,052
Negligible
17,318
Negligible
0
0
Negligible
yr14
0
0
yr14
681,052
681,052
0
Negligible
0
yr14
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr14
8,153,370
(-)6,048 (-)19,324
(-)6,048 (-)19,324
yr15
8,153,370
681,052
Negligible
17,318
Negligible
Negligible
7,455,000
Negligible
yr15
(-)6,048 (-)19,324
(-)6,048 (-)19,324
yr15
681,052
681,052
Negligible
Negligible
0
yr15
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr15
8,153,370
(-)6,048 (-)19,324
(-)6,048 (-)19,324
yr16
8,153,370
681,052
Negligible
17,318
Negligible
Negligible
7,455,000
Negligible
yr16
(-)6,048 (-)19,324
(-)6,048 (-)19,324
yr16
681,052
681,052
Negligible
Negligible
0
yr16
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr16
8,153,370
(-)6,048 (-)19,324
(-)6,048 (-)19,324
yr17
8,153,370
681,052
Negligible
17,318
Negligible
Negligible
7,455,000
Negligible
yr17
(-)6,048 (-)19,324
(-)6,048 (-)19,324
yr17
681,052
681,052
Negligible
Negligible
0
yr17
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr17
8,153,370
(-)6,048 (-)19,324
(-)6,048 (-)19,324
yr18
8,153,370
681,052
Negligible
17,318
Negligible
Negligible
7,455,000
Negligible
yr18
(-)6,048 (-)19,324
(-)6,048 (-)19,324
yr18
681,052
681,052
Negligible
Negligible
0
yr18
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr18
8,153,370
(-)6,048 (-)19,324
(-)6,048 (-)19,324
yr19
8,153,370
681,052
Negligible
17,318
Negligible
Negligible
7,455,000
Negligible
yr19
(-)6,048 (-)19,324
(-)6,048 (-)19,324
yr19
681,052
681,052
Negligible
Negligible
0
yr19
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr19
8,153,370
(-)6,048 (-)19,324
(-)6,048 (-)19,324
yr20
120,318,896
1,021,578
Negligible
17,318
Negligible
Negligible
119,280,000
Negligible
yr20
(-)6,048 (-)19,324
(-)6,048 (-)19,324
yr20
1,021,578
1,021,578
Negligible
Negligible
0
yr20
(-)16,866 (-)36,888
(-)16,866 (-)36,888
yr20
68,879,396
NA
NA
PV
90,531,463
9,521,513
Negligible
254,532
Negligible
Negligible
80,740,418
15,000
PV
NA
NA
PV
9,536,513
9,521,513
Negligible
Negligible
15,000
PV
NA
NA
PV
100,726,092
q: have you any evidence that validates or would better inform the assumptions in the evidence base and estimates set out in the table
above?
Sector affected
Type
Benefits
Sector affected
Type (short descriptive)
Costs
15 Years Manual
Sector affected
Type
Benefits
Sector affected
Type of cost
Costs
15 Years Automated
Public sector
ALL BENEFITS
All
Type
Benefits
Total
Annexes
81
Keeping the right people on the DNA database: Science and public protection
Annex C: Offences which would
fall under Violent and Sexual
Offences Provision
SPECIFIED VIOlENT OFFENCES
1 Manslaughter.
2 Kidnapping.
3 False imprisonment.
4 An offence under section 4 of the Offences against
the Person Act 1861 (c. 100) (soliciting murder).
5 An offence under section 16 of that Act (threats
to kill).
6 An offence under section 18 of that Act (wounding
with intent to cause grievous bodily harm).
7 An offence under section 20 of that Act (malicious
wounding).
8 An offence under section 21 of that Act (attempting
to choke, suffocate or strangle in order to commit or
assist in committing an indictable offence).
9 An offence under section 22 of that Act (using
chloroform etc. to commit or assist in the
committing of any indictable offence).
10 An offence under section 23 of that Act
(maliciously administering poison etc. so as to
endanger life or inflict grievous bodily harm).
11 An offence under section 27 of that Act
(abandoning children).
12 An offence under section 28 of that Act (causing
bodily injury by explosives).
13 An offence under section 29 of that Act (using
explosives etc. with intent to do grievous bodily harm).
14 An offence under section 30 of that Act (placing
explosives with intent to do bodily injury).
15 An offence under section 31 of that Act (setting
spring guns etc. with intent to do grievous bodily
harm).
16 An offence under section 32 of that Act
(endangering the safety of railway passengers).
17 An offence under section 35 of that Act (injuring
persons by furious driving).
82
18 An offence under section 37 of that Act (assaulting
officer preserving wreck).
19 An offence under section 38 of that Act (assault
with intent to resist arrest).
20 An offence under section 47 of that Act (assault
occasioning actual bodily harm).
21 An offence under section 2 of the Explosive
Substances Act 1883 (c. 3) (causing explosion likely
to endanger life or property).
22 An offence under section 3 of that Act (attempt
to cause explosion, or making or keeping explosive
with intent to endanger life or property).
23 An offence under section 1 of the Infant Life
(Preservation) Act 1929 (c. 34) (child destruction).
24 An offence under section 1 of the Children and
Young Persons Act 1933 (c. 12) (cruelty to children).
25 An offence under section 1 of the Infanticide Act
1938 (c. 36) (infanticide).
26 An offence under section 16 of the Firearms Act
1968 (c. 27) (possession of firearm with intent to
endanger life).
27 An offence under section 16A of that Act
(possession of firearm with intent to cause fear of
violence).
28 An offence under section 17(1) of that Act (use of
firearm to resist arrest).
29 An offence under section 17(2) of that Act
(possession of firearm at time of committing or
being arrested for offence specified in Schedule 1 to
that Act).
30 An offence under section 18 of that Act (carrying a
firearm with criminal intent).
31 An offence under section 8 of the Theft Act 1968
(c. 60) (robbery or assault with intent to rob).
32 An offence under section 9 of that Act of burglary
with intent to—
(a) inflict grievous bodily harm on a person, or
(b) do unlawful damage to a building or anything
in it.
33 An offence under section 10 of that Act (aggravated
burglary).
Annexes
34 An offence under section 12A of that Act
(aggravated vehicle-taking) involving an accident
which caused the death of any person.
35 An offence of arson under section 1 of the
Criminal Damage Act 1971 (c. 48).
36 An offence under section 1(2) of that Act
(destroying or damaging property) other than an
offence of arson.
51 An offence under section 9 of that Act (hijacking
of ships).
52 An offence under section 10 of that Act (seizing or
exercising control of fixed platforms).
53 An offence under section 11 of that Act (destroying
fixed platforms or endangering their safety).
54 An offence under section 12 of that Act (other acts
endangering or likely to endanger safe navigation).
37 An offence under section 1 of the Taking of
Hostages Act 1982 (c. 28) (hostage-taking).
55 An offence under section 13 of that Act (offences
involving threats).
38 An offence under section 1 of the Aviation Security
Act 1982 (c. 36) (hijacking).
56 An offence under Part II of the Channel Tunnel
(Security) Order 1994 (S.I. 1994/570) (offences
relating to Channel Tunnel trains and the tunnel
system).
39 An offence under section 2 of that Act (destroying,
damaging or endangering safety of aircraft).
40 An offence under section 3 of that Act (other acts
endangering or likely to endanger safety of aircraft).
41 An offence under section 4 of that Act (offences in
relation to certain dangerous articles).
42 An offence under section 127 of the Mental Health
Act 1983 (c. 20) (ill-treatment of patients).
43 An offence under section 1 of the Prohibition of
Female Circumcision Act 1985 (c. 38) (prohibition
of female circumcision).
44 An offence under section 1 of the Public Order Act
1986 (c. 64) (riot).
45 An offence under section 2 of that Act (violent
disorder).
46 An offence under section 3 of that Act (affray).
57 An offence under section 4 of the Protection from
Harassment Act 1997 (c. 40) (putting people in fear
of violence).
58 An offence under section 29 of the Crime and
Disorder Act 1998 (c. 37) (racially or religiously
aggravated assaults).
59 An offence falling within section 31(1)(a) or (b) of
that Act (racially or religiously aggravated offences
under section 4 or 4A of the Public Order Act 1986
(c. 64)).
60 An offence under section 51 or 52 of the
International Criminal Court Act 2001 (c. 17)
(genocide, crimes against humanity, war crimes and
related offences), other than one involving murder.
47 An offence under section 134 of the Criminal
Justice Act 1988 (c. 33) (torture).
61 An offence under section 1 of the Female Genital
Mutilation Act 2003 (c. 31) (female genital
mutilation).
48 An offence under section 1 of the Road Traffic Act
1988 (c. 52) (causing death by dangerous driving).
62 An offence under section 2 of that Act (assisting a
girl to mutilate her own genitalia).
49 An offence under section 3A of that Act (causing
death by careless driving when under influence of
drink or drugs).
63 An offence under section 3 of that Act (assisting
a non-UK person to mutilate overseas a girl’s
genitalia).
50 An offence under section 1 of the Aviation and
Maritime Security Act 1990 (c. 31) (endangering
safety at aerodromes).
64 An offence of—
(a) aiding, abetting, counselling, procuring or
inciting the commission of an offence specified
in this Part of this Schedule,
83
Keeping the right people on the DNA database: Science and public protection
(b) conspiring to commit an offence so specified, or
(c) attempting to commit an offence so specified.
65 An attempt to commit murder or a conspiracy to
commit murder.
SPECIFIED SExuAl OFFENCES
66 An offence under section 1 of the Sexual Offences
Act 1956 (c. 69) (rape).
83 An offence under section 22 of that Act (causing
prostitution of women).
84 An offence under section 23 of that Act
(procuration of girl under twenty-one).
85 An offence under section 24 of that Act (detention
of woman in brothel).
67 An offence under section 2 of that Act
(procurement of woman by threats).
86 An offence under section 25 of that Act (permitting
girl under thirteen to use premises for intercourse).
68 An offence under section 3 of that Act
(procurement of woman by false pretences).
87 An offence under section 26 of that Act (permitting
girl under sixteen to use premises for intercourse).
69 An offence under section 4 of that Act
(administering drugs to obtain or facilitate
intercourse).
88 An offence under section 27 of that Act (permitting
defective to use premises for intercourse).
70 An offence under section 5 of that Act (intercourse
with girl under thirteen).
89 An offence under section 28 of that Act (causing or
encouraging the prostitution of, intercourse with or
indecent assault on girl under sixteen).
71 An offence under section 6 of that Act (intercourse
with girl under 16).
90 An offence under section 29 of that Act (causing or
encouraging prostitution of defective).
72 An offence under section 7 of that Act (intercourse
with a defective).
91 An offence under section 32 of that Act (soliciting
by men).
73 An offence under section 9 of that Act
(procurement of a defective).
92 An offence under section 33 of that Act (keeping a
brothel).
74 An offence under section 10 of that Act (incest by
a man).
93 An offence under section 128 of the Mental Health
Act 1959 (c. 72) (sexual intercourse with patients).
75 An offence under section 11 of that Act (incest by a
woman).
94 An offence under section 1 of the Indecency with
Children Act 1960 (c. 33) (indecent conduct towards
young child).
76 An offence under section 14 of that Act (indecent
assault on a woman).
77 An offence under section 15 of that Act (indecent
assault on a man).
78 An offence under section 16 of that Act (assault
with intent to commit buggery).
79 An offence under section 17 of that Act (abduction
of woman by force or for the sake of her property).
80 An offence under section 19 of that Act (abduction
of unmarried girl under eighteen from parent or
guardian).
81 An offence under section 20 of that Act (abduction
of unmarried girl under sixteen from parent or
guardian).
84
82 An offence under section 21 of that Act (abduction
of defective from parent or guardian).
95 An offence under section 4 of the Sexual Offences
Act 1967 (c. 60) (procuring others to commit
homosexual acts).
96 An offence under section 5 of that Act (living on
earnings of male prostitution).
97 An offence under section 9 of the Theft Act 1968
(c. 60) of burglary with intent to commit rape.
98 An offence under section 54 of the Criminal Law
Act 1977 (c. 45) (inciting girl under sixteen to have
incestuous sexual intercourse).
99 An offence under section 1 of the Protection of
Children Act 1978 (c. 37) (indecent photographs of
children).
Annexes
100 An offence under section 170 of the Customs
and Excise Management Act 1979 (c. 2) (penalty
for fraudulent evasion of duty etc.) in relation to
goods prohibited to be imported under section
42 of the Customs Consolidation Act 1876 (c. 36)
(indecent or obscene articles).
101 An offence under section 160 of the Criminal
Justice Act 1988 (c. 33) (possession of indecent
photograph of a child).
117 An offence under section 16 of that Act (abuse of
position of trust: sexual activity with a child).
118 An offence under section 17 of that Act (abuse
of position of trust: causing or inciting a child to
engage in sexual activity).
119 An offence under section 18 of that Act (abuse of
position of trust: sexual activity in the presence of
a child).
102 An offence under section 1 of the Sexual Offences
Act 2003 (c. 42) (rape).
120 An offence under section 19 of that Act (abuse of
position of trust: causing a child to watch a sexual
act).
103 An offence under section 2 of that Act (assault by
penetration).
121 An offence under section 25 of that Act (sexual
activity with a child family member).
104 An offence under section 3 of that Act (sexual
assault).
122 An offence under section 26 of that Act (inciting a
child family member to engage in sexual activity).
105 An offence under section 4 of that Act (causing
a person to engage in sexual activity without
consent).
123 An offence under section 30 of that Act (sexual
activity with a person with a mental disorder
impeding choice).
106 An offence under section 5 of that Act (rape of a
child under 13).
124 An offence under section 31 of that Act (causing
or inciting a person with a mental disorder
impeding choice to engage in sexual activity).
107 An offence under section 6 of that Act (assault of
a child under 13 by penetration).
108 An offence under section 7 of that Act (sexual
assault of a child under 13).
125 An offence under section 32 of that Act (engaging
in sexual activity in the presence of a person with a
mental disorder impeding choice).
109 An offence under section 8 of that Act (causing
or inciting a child under 13 to engage in sexual
activity).
126 An offence under section 33 of that Act (causing a
person with a mental disorder impeding choice to
watch a sexual act).
110 An offence under section 9 of that Act (sexual
activity with a child).
127 An offence under section 34 of that Act
(inducement, threat or deception to procure sexual
activity with a person with a mental disorder).
111 An offence under section 10 of that Act (causing
or inciting a child to engage in sexual activity).
112 An offence under section 11 of that Act (engaging
in sexual activity in the presence of a child).
113 An offence under section 12 of that Act (causing a
child to watch a sexual act).
114 An offence under section 13 of that Act (child sex
offences committed by children or young persons).
115 An offence under section 14 of that Act (arranging
or facilitating commission of a child sex offence).
116 An offence under section 15 of that Act (meeting
a child following sexual grooming etc.).
128 An offence under section 35 of that Act (causing
a person with a mental disorder to engage in or
agree to engage in sexual activity by inducement,
threat or deception).
129 An offence under section 36 of that Act (engaging
in sexual activity in the presence, procured by
inducement, threat or deception, of a person with
a mental disorder).
130 An offence under section 37 of that Act (causing a
person with a mental disorder to watch a sexual act
by inducement, threat or deception).
85
Keeping the right people on the DNA database: Science and public protection
131 An offence under section 38 of that Act (care
workers: sexual activity with a person with a mental
disorder).
132 An offence under section 39 of that Act (care
workers: causing or inciting sexual activity).
133 An offence under section 40 of that Act (care
workers: sexual activity in the presence of a person
with a mental disorder).
134 An offence under section 41 of that Act (care
workers: causing a person with a mental disorder
to watch a sexual act).
135 An offence under section 47 of that Act (paying
for sexual services of a child).
136 An offence under section 48 of that Act (causing
or inciting child prostitution or pornography).
137 An offence under section 49 of that Act
(controlling a child prostitute or a child involved in
pornography).
138 An offence under section 50 of that Act (arranging
or facilitating child prostitution or pornography).
139 An offence under section 52 of that Act (causing
or inciting prostitution for gain).
140 An offence under section 53 of that Act
(controlling prostitution for gain).
141 An offence under section 57 of that Act
(trafficking into the UK for sexual exploitation).
142 An offence under section 58 of that Act
(trafficking within the UK for sexual exploitation).
143 An offence under section 59 of that Act
(trafficking out of the UK for sexual exploitation).
144 An offence under section 61 of that Act
(administering a substance with intent).
145 An offence under section 62 of that Act
(committing an offence with intent to commit a
sexual offence).
146 An offence under section 63 of that Act (trespass
with intent to commit a sexual offence).
147 An offence under section 64 of that Act (sex with
an adult relative: penetration).
148 An offence under section 65 of that Act (sex with
an adult relative: consenting to penetration).
86
149 An offence under section 66 of that Act
(exposure).
150 An offence under section 67 of that Act
(voyeurism).
151 An offence under section 69 of that Act
(intercourse with an animal).
152 An offence under section 70 of that Act (sexual
penetration of a corpse).
153 An offence of—
(a) aiding, abetting, counselling, procuring or
inciting the commission of an offence specified
in this Part of this Schedule,
(b) conspiring to commit an offence so specified, or
(c) attempting to commit an offence so specified.
Annexes
Annex D: Research Base for Retention Period
1 bACkGrOuND
1.1 The length of retention of DNA profiles17 should,
overall, be proportionate to the risk of offending by
those people from whom a DNA sample is taken.
We are concerned with the risk of offending of
those arrested but not charged or cautioned, or if
charged, not convicted. The DNA of those
convicted is not affected by the ruling and will
remain on the database indefinitely.
1.2 The risk of an arrestee being re-arrested will fall as
the number of years from the arrest increases. We
refer to this as a declining “hazard-rate”. We want to
estimate how long it will take for the risk of re­
arrest for someone who is arrested but not
convicted or cautioned to return to the same level as
the risk of arrest for the general population. If the
risk of arrest remains higher than that for the
general population, there is more justification for
retaining the DNA profile on the database.
1.3 The available data mean that these arrest rates
cannot be calculated and compared directly. In
order to obtain traction on the problem we make
two key assumptions:
(a) we assume that the “hazard rate” of arrest for
someone previously arrested can be estimated
by the conviction rate of people previously
convicted; and
(b) we assume that our estimate of the point at
which the hazard rate for the arrestee
population and the general population are
equalised is the same whether we examine the
rates for all those convicted, or for those
convicted for the first time.
If assumption (a) is accepted, we are then able to
use data from the Police National Computer on
convictions to investigate this question. This is a
17
strong assumption, but is partially supported by
analysis from the Jill Dando Institute (JDI). The JDI
research compares the behaviour of offenders who
were arrested and either acquitted or had no further
action taken against them, with the behaviour of
those who were cautioned or proceeded to court
but received a non-custodial sentence. Over a
three-year period, the research suggests that the
subsequent criminality of the group against whom
no further action was taken is similar or even higher
than the criminality of the convicted group. The
effect of assumption (b) is discussed in section 2.10
2 CONVICTION bEhAVIOur ANAlySIS
2.1 We examined a cohort of offenders who had been
convicted of an offence in 2001. The members of
the cohort had been first arrested and subsequently
found guilty of an indictable offence, or of one of a
number of summary offences. This cohort contains
both first time and repeat offenders since it was not
possible to separate out first time offenders from
the rest, as mentioned above. The potential effect
of this is discussed in section 2.10
2.2 The reconvictions for this cohort were then
investigated in each of the 6 years after 2001. Figure
1 shows (i) the hazard rate for the PNC cohort; (ii)
the conviction rate for men (ages 16-20); and (iii)
the conviction rate for all men and women. The
hazard rate is calculated as.
Hazard rate =
number of people who are reconvicted in the year
total number of people who had not been
reconvicted up to that year
2.3 The hazard rate shows how the risk of conviction
for our cohort of people convicted in 2001 changes
over time. Over the six years it falls from just short
of 30% to around 5%. The other lines show (i) the
risk of conviction of any member of the general
population whether or not they were previously
convicted and (ii) the similar risk for 16-20 year old
males. As the graph shows, the hazard line for the
PNC-derived cohort and the 16-20 male line cross
A profile is an electronic representation of a physical DNA sample.
87
Keeping the right people on the DNA database: Science and public protection
after 4 years. This implies that after 4 years the probability of any person re-offending and being convicted following an earlier conviction is the same as the probability that any 16-20 year old male offends and is convicted whether or not they have
been previously convicted. Since 16-20 year old males are the most likely to be convicted this gives
an estimate of the minimum advisable retention
length of a DNA sample.
Figure 1: Hazard rates with initial conviction in 2001 with population conviction rates
Figure 1: Hazard rates (reconviction) compared with national conviction rates
30%
Hazard rate (Reconvictions)
Male Peak: 16-20 years old
Likelihood of Conviction or Caution
25%
All people all ages
20%
15%
10%
5%
0%
1
2
3
4
5
6
Years after a 2001 conviction
2.4 The hazard curve is derived only for a six year
period. But by fitting the hazard curve with a power
function and extrapolating for future years we find
that the PNC cohort hazard curve crosses the “all
population” line at a point between 14 and 15 years,
indicating that it takes around 15 years for the PNC
conviction cohort to return to the overall
population probability of conviction. This gives an estimate of the maximum length of time over
which the probability of re-conviction falls to that
no greater than that of the general population. It may therefore indicate the maximum length of time
a DNA sample might be retained.
88
2.5 We can further extend this analysis by looking at
whether there is any effect of the initial conviction
type on the hazard rate. Figure 2 shows the hazard
rates for serious violence (ABH, GBH and more
serious types of violence) and also sexual offences
compared to the hazard rate of reconviction for
all crime.
Annexes
Figure 2: The hazard rate for sexual offences, serious violence and all offence types
Figure 2: The hazard rate for sexual offences, serious violence and all offence types
30.0%
Sexual Offences
25.0%
Violence
All
20.0%
15.0%
10.0%
5.0%
0.0%
1
2
3
2.6 For sexual offenders, the low numbers convicted
combined with the long sentence length means that
the curve becomes more unstable at the 3 year
point. This meant it was not possible to estimate the
long term implications of re-offending for such
offenders.
4
5
6
2.7 The graph of the hazard rate for violence has a
shallower curve in comparison to the hazard line for
all crimes. When fitting a curve as shown in Figure
3, below, we can see that it takes in excess of 20
years for the hazard rate to return to the overall
population rate. This may be in part due to the
longer sentence lengths received by such offenders
leading to a later release and thereby reducing the
numbers able to re-offend in early years. Lack of
data means it is not currently possible to
compensate for this.
89
Keeping the right people on the DNA database: Science and public protection
Figure 3: Hazard rate for serious violence with the fitted curve projected forwards to 20 years.
Figure 3: Hazard rate for serious violence with the fitted curve projected forwards to 20 years
30.0%
25.0%
Fitted Curve
20.0%
Actual Figures
All People
15.0%
10.0%
5.0%
0.0%
1
2
3
4
5
6
7
8
9
2.8 Different hazard rates for different offence types
suggest that there are potentially different retention
schedules which could be used for people who have
committed different offence types, and could give
different retention periods especially for the minor
crime types. However, the evidence does not
support giving minor crimes a different retention
period. Figure 4, below, shows the probability of
being convicted in 2002 for the given offence types,
given that an offender was originally convicted for
an offence type and was reconvicted (those who did
not re-offend are therefore excluded). It shows that
for most crime types an offender is considerably
more likely to be convicted for a different crime
type the next time than the same crime type. In
essence it suggests that criminals are generalists and
that being initially convicted of a minor crime does
not imply a significantly lower risk of going on to
commit a more serious one.
90
10
11
12
13
14
15
16
17
18
19
20
Annexes
Figure 4: Reconviction
crimes probabilities
Figure 5: Reconviction
criimes probabilities
Probability of those convicted of a crime type in 2001, to re-offend in a given crime type, given they are reconvicted
Probability of reconviction in the given crime type for
those who re-offended
1.00
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
He
ce
Vi
ol
en
t
ef
nc
Se
al
xu
th
al
Of
fe
Th
es
s
le
ic
Sa
d
an
an
ry
ge
Fo
r
Ve
h
ty
fe
ud
d
Ga
d
an
s
ug
Dr
As
sa
Fr
a
ng
s
er
rd
so
Di
ul
Al
ta
co
nd
ho
Da
lM
m
is
ag
us
e
e
0.00
Offence type convicted of in first year
Violence
Theft
Forgery and Fraud
Sexual Offences
Drugs and Gangs
Vehicles
Disorder
Health and Safety
Assault and Damage
Alcohol Misuse
2.9 Figure 5, over, takes the above analysis and
concentrates on the most serious crimes, sexual
offences and serious violence. Although as we have
seen, offenders tend to be generalists there is a
higher risk over a year for those offenders convicted
of sexual crimes to re-offend in a sexual crime than
those offenders who were convicted of any other
crime. A similar effect occurs for serious violent
offenders however the increase in risk is much
smaller. This suggests that although such offenders
are generalists, there is an increased risk associated
with the more serious crimes. A lack of data at
present means it is not possible to say if this effect
persists over multiple years.
91
Keeping the right people on the DNA database: Science and public protection
Figure 5: Probability of having a conviction of offence type in the second year, given the
Figure 5:
criimes
probabilities
conviction
ofReconviction
a given offence
type
in the first year
Probability of those convicted of a crime type in 2001, to re-offend in a given crime type, given they are reconvicted
Probability of reconviction in the given crime type for
those who re-offended
1.00
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
ce
Vi
ol
en
t
ef
nc
fe
Of
xu
al
Th
es
s
le
ic
Sa
d
an
th
Ve
h
ty
fe
ud
an
d
d
Ga
Fr
a
ng
s
er
rd
so
Di
an
ry
co
nd
ho
Da
lM
m
is
ag
us
e
e
0.00
Se
al
He
ge
Fo
r
s
ug
Dr
As
sa
ul
Al
ta
2.10 The two bounds of 4 years and 15 years give us a
3.2 The first paper, (Kurlycheck, Brame and
lower and upper bound for retention periods, based
Bushway18), considers the hazard rate for re-arrests
Offence
of inafirst
year
on the conviction probabilities of the highest
risktype convictedfor
Philadelphia
cohort of individuals born in
segment of the
population
and
the
overall
1958.
looks
along similar lines to our analysis
Sexual Offences
Theft
Vehicles
Violence
HealthItand
Safety
population respectively.
As
previously
mentioned,
although
it
concentrates
on the
behaviour of people
Alcohol
Misuse
Assault and Damage
Drugs and Gangs
Forgery and Fraud
Disorder
the analysis of the PNC cohort examines the hazard
following arrests. The main conclusion from this
rate for all offenders, not just first time offenders.
analysis is that after 5 years the difference between
Looking only at first time offenders might reduce
hazard rates for arrested and non-arrested
the hazard rate, shifting the line downwards overall.
individuals is still statistically significant at over 1%.
But the 16-20 male conviction line also includes
This analysis examines only a short set of years due
both first and repeat offenders. Removing repeat
to the limitations on the dataset used.
offenders is likely to reduce this rate too. Therefore
we do not expect that removing repeat offenders
3.3 The second paper, (Nakamura19), considers the
from the analysis entirely would have a significant
hazard of re-arrest for a cohort of people arrested
impact on the conclusions reached, especially on the
for the first time in 1980 in New York. It
calculated lower retention bound.
investigates the effect of being arrested for different
crime types on the time (T*) for arrestees’ hazard
3 INTErNATIONAl rESEArCh
rate to reach the population hazard rate. It
3.1 Two US papers examine the same problem from the
concludes that for first time arrestees of age 18 who
point of view of criminal record checking for
were arrested for Robbery, T* is 14.5 years. Similarly
potential employees.
for arrestees of age 16 arrested for Burglary, T* is
13.5 years.
92
18
“Scarlet Letters and Recidivism: Does an old criminal record predict future offending?” – Kurlycheck, Brame and Bushway 2006, Criminology &
Public Policy, 5(3): 483-504
19
“Redemption in the Presence of Widespread Criminal Background Check” – Kiminori Nakamura, 21 April 2008, Heinz School second paper,
http://www.heinz.cmu.edu/research/233full.pdf.
Annexes
3.4 The length of time to reach the population hazard
level as indicated in the second paper is lower that
those seen in our analysis, the likely reason for this
difference is the focus in the Nakamura paper on
different offence types while as mentioned earlier
our analysis considers all types of indictable and
some summary offences.
93
Keeping the right people on the DNA database: Science and public protection
Annex E – The Detection Model
1.3 This gave the probability that the DNA database
would allow a successful detection given that the
individual had committed an offence. As such it
excludes those who left their DNA at the crime
scene accidently thus avoiding the potential for over
counting. This probability was then multiplied by
the number of individuals arrested and sampled
each year to give the number of detections made
possible by the database.
1.1 The DNA database will only be useful for those
crimes where DNA is left at the crime scene. The
evidence presented in the previous annex does not
consider whether the DNA database would be
useful in preventing the crimes which they were
convicted for.
1.2 This research was therefore expanded into a
1.4 Figure 6. shows the results of this analysis. Even
separate model based on the flow diagrams which
with a zero retention period it assumes that a sample
are provided in Annex A above, especially the
is taken on arrest and compared to previous crime
second one. The probabilities which were associated
scene samples which are stored on the database, this
with the arms of the flow diagram are given in the
should match to around 8500 crimes. If the DNA is
main Impact Assessment and using these it was
then immediately destroyed this means some future
possible to calculate the probability for an individual
crimes are undetected, leading to an average 6200
who was released without charge of going on to
potential detections being lost per annum. As the
commit an offence and leaving a sample of DNA
retention length on the database increases the
(or a fingerprint for that matter) at the crime scene.
hazard rate falls meaning the likelihood of
The model also included the probability that the
detections being lost also falls, after 4 years it falls to
individual was still on the database by considering
4800 and after 15 years 450. The 6 year
the likelihood both of the individual not
recommended period then leads to 3800 fewer
committing further offences and also the likelihood
detections each year than the do nothing option.
of them committing a further offence but not being arrested for it.
Figure 6: Detections Lost per annum
Number of Detections Lost by Retention Length
7,000
0, 6241
Current Proposal
Detections lost (pa average)
6,000
5,000
4, 4844
4,000
6, 3840
3,000
2,000
15, 452
1,000
0
0
1
2
3
4
5
6
7
8
9
10
11
12
Retention Length
94
13
14
15
16
17
18
19
20
Annexes
COmmENTS by ChIEF ECONOmIST
I have examined the impact assessment relating to the
implementation of the “S & Marper” judgement in the
European Court of Human Rights, and I have been
advised by the economists who have constructed it.
Overall, I am satisfied that the impact assessment
provides a reasonable assessment of the likely costs and
benefits of the proposal and that the key risks to the
project have been identified.
v) The need to complete this work to a very short
time table means that the modelling has not
captured all costs and benefits as completely as I
would ideally like to have seen. There may be a
need to revisit this assessment before a final
decision is made.
John Elliott
April 2009
I draw attention to the following five points:
i) The benefit of not deleting all DNA profiles
immediately is clear – there is an increased
probability of detection of offenders who will
be engaged in criminal activity in the future. In
the preferred option, the estimated number of
additional detections is greater than in the case
when all eligible profiles are deleted after 6 years
– but probably not much greater since sexual
and violent crimes are relatively few in number.
That said they cause considerable additional
harms to society which the Department
estimates and publishes. Therefore the
significantly longer retention period for these
offences should be considered in this context.
ii) Fingerprints have been treated differently to
DNA profiles. Whilst it would be possible to
treat them identically the significant additional
benefits to storing fingerprints and the lower
impact on privacy suggests this is not necessary.
iii) The assessment is forward-looking. It does not
consider the costs of dealing with those DNA
profiles already on the database – the legacy
cases - which would also need to be deleted. The
associated costs are likely to be significant, but
there has not been time to quantify them.
iv) If deletion is not to be immediate, a retention
period must be set. The decision to opt for 6
years is likely to receive considerable scrutiny
but is based on only limited evidence. Ideally a
fuller consideration of different retention
periods would have been helpful but I accept
this was not possible in the time available.
95
Keeping the right people on the DNA database: Science and public protection
Annex E: Confidentiality and
Disclaimer
The information you send us may be passed to
colleagues within the Home Office, the Government or
related agencies.
Information provided in response to this consultation,
including personal information, may be subject to
publication or disclosure in accordance with the access to
information regimes (these are primarily the Freedom of
Information Act 2000 [FOIA], the Data Protection Act
1998 [DPA] and the Environmental Information
Regulations 2004).
If you want other information that you provide to be
treated as confidential, please be aware that, under the
FOIA, there is a statutory Code of Practice with which
public authorities must comply and which deals, amongst
other things, with obligations of confidence.
In view of this it would be helpful if you could explain to
us why you regard the information you have provided as
confidential. If we receive a request for disclosure of the
information we will take full account of your
explanation, but we cannot give an assurance that
confidentiality can be maintained in all circumstances.
An automatic confidentiality disclaimer generated by
your IT system will not, of itself, be regarded as binding
on the Department.
The Department will process your personal data in
accordance with the DPA and in the majority of
circumstances this will mean that your personal data will
not be disclosed to third parties.
CONSulTATION CO-OrDINATOr
If you have a complaint or comment about the Home
Office’s approach to consultation, you should contact the
Home Office Consultation Co-ordinator, Nigel
Lawrence. Please DO NOT send your response to this
consultation to Nigel Lawrence. The Co-ordinator works
to promote best practice standards set by the
Government’s Code of Practice, advises policy teams on
how to conduct consultations and investigates
complaints made against the Home Office. He does not
process your response to this consultation.
96
The Co-ordinator can be emailed at: Nigel.Lawrence@
homeoffice.gsi.gov.uk or alternatively write to him at:
Nigel Lawrence, Consultation Co-ordinator
Home Office
Performance and Delivery Unit
Better Regulation Team
3rd Floor Seacole
2 Marsham Street
London
SW1P 4DF
The Consultation follows the Government’s Code of
Practice on Consultation – the criteria for which are set
out below:
Criterion 1 – When to consult – Formal
consultation should take place at a stage when there
is scope to influence the policy outcome.
Criterion 2 – Duration of consultation exercises
– Consultations should normally last for at least 12
weeks with consideration given to longer timescales
where feasible and sensible.
Criterion 3 – Clarity of scope and impact –
Consultation documents should be clear about the
consultation process, what is being proposed, the
scope to influence and the expected costs and
benefits of the proposals.
Criterion 4 – Accessibility of consultation exercises
– Consultation exercises should be designed to be
accessible to, and clearly targeted at, those people
the exercise is intended to reach.
Criterion 5 – The burden of consultation – Keeping
the burden of consultation to a minimum is
essential if consultations are to be effective and if
consultees’ buy-in to the process is to be obtained.
Criterion 6 – Responsiveness of consultation
exercises – Consultation responses should be
analysed carefully and clear feedback should be
provided to participants following the consultation.
Annexes
Criterion 7 – Capacity to consult – Officials running
consultations should seek guidance in how to run
an effective consultation exercise and share what
they have learned from the experience.
The full Code of Practice on Consultation is available at:
http://www.berr.gov.uk/whatwedo/bre/consultation­
guidance/page44420.html
97
© Crown Copyright 2009
Produced by the Home Office. May 2009.
Ref: 295723. ISBN: 978-1-84726-911-9