Journal of Law, Information and Science
|
|
Home
| Databases
| WorldLII
| Search
| Feedback
Journal of Law, Information and Science |
|
LAURA THOMAS
Partial DNA matching is a forensic technique that compares the DNA profiles of two individuals to determine whether it is likely that they are close genetic relatives.[1] When DNA is left at a crime scene the technique can be used in combination with a DNA database or a mass screening to ‘track down’ the posited offender,[2] usually when more traditional investigative techniques have failed. In the United Kingdom and the United States the technique has been used in criminal investigations since 2002 and has led to some spectacular successes in solving violent crimes, including in cold cases that would otherwise have remained unsolved. Part One of this article describes the science behind the partial DNA matching technique and ways that it can be used in criminal investigations.
In Australia partial DNA matching has been used to identify unknown deceased persons, including 67 victims of the Bali bombing,[3] but there has not yet been any publicized use of the technique in a criminal investigation. CrimTrac, the Commonwealth government agency that operates Australia’s DNA database, is currently investigating ‘the numerous legislative, ethical, privacy and technical issues’ that would arise if partial DNA matching were used in Australian criminal investigations.[4] Part Two of this article examines some of these issues, including concerns about the privacy of individuals and families, the efficiency of criminal investigations, and disproportionate impacts upon ethnic groups already over-represented in the criminal justice system.
Legislation governing the use of DNA in criminal investigations must balance civil liberties and effective law enforcement. In striking this balance, Australian legislatures have favoured civil liberties more than their counterparts in the United States and the United Kingdom. In keeping with that approach, Part Three of this article proposes a method of judicial oversight of the use of partial DNA matching in criminal investigations. This would allow for a case-by-case balancing of the public’s interest in allowing police to obtain evidence with the public’s interest in upholding privacy rights. It is suggested that the correct balance will depend on the seriousness of the crime being investigated; the use of partial DNA matching will usually only be justified in the investigation of serious violent crimes, where more traditional investigative techniques have failed. Practical ways to minimize potential harms are also suggested in Part Three, including the training of specialist police officers and making genetic counselors available to families under investigation.
2.1 The Science of DNA Profile Matching and the Partial Matching Technique
The DNA profiling technique used in Australia, called Profiler Plus, measures the length of certain portions of the non-coding or ‘junk’ DNA, where short sequences of the genetic code repeat several times. These sequences are known as short tandem repeats (STRs).[5] Every individual has two copies of each STR, one inherited from each parent. For example, at one STR location the sequence G-A-T-A[6] repeats between and 15 times. An individual may have one G-A-T-A STR with 8 repeats, inherited from their mother, and one with 11 repeats, inherited from their father. Profiler Plus measures the length of an individual’s two STRs at nine different locations, and determines the person’s sex, producing a profile consisting of nine pairs of numbers (or markers) and the person’s sex. Because all Australian jurisdictions use Profiler Plus, any two DNA profiles prepared in Australia can be compared.
The use of genetic profiles in the forensic investigation of crime usually involves finding two profiles that match completely – i.e. all 18 markers representing the length of the individual’s STRs are the same in both profiles. The chance of two people who are not identical twins having the same profile is one in many millions or even billions.[7] Because each individual inherits one STR at each location from each parent, close genetic relatives are likely to share many more markers than unrelated individuals. A parent and child will share, on average, about 60% of their markers and siblings share, on average, about 70%. By comparison, unrelated individuals share, on average, about 30% of their markers.[8] In addition, a parent and child will always share at least one marker in each pair – a pattern that is ‘highly characteristic of that specific relationship’.[9] Partial DNA matching involves finding profiles that do not match completely but have an unusually high percentage of shared markers. A complex calculation can be used to determine the probability that the two profiles belong to close genetic relatives.[10]
2.2 Partial DNA Matching in Criminal Investigations
Partial DNA matching was first used in a criminal investigation in the United Kingdom in 2002 and has since been used in other countries. The technique can be used in an ad hoc manner, but is more likely to be used in conjunction with a DNA database or a mass screening of potential suspects.
Partial DNA matching may occasionally be used in an ad hoc manner in police investigations. For example, it has been reported that Kansas police used partial DNA matching in the investigation that led to the arrest and eventual prosecution of Dennis Rader, the BTK (bind, torture, kill) serial killer, who pleaded guilty to murdering ten women.[11] Although the Kansas police have not publicly confirmed how DNA was used in the investigation, several reputable newspapers reported that once Rader became a suspect, a DNA sample from his daughter was used to confirm that the crime scene DNA likely came from her father, thereby providing strong confirmation of police suspicion about Rader without alerting him to that suspicion by requiring him to provide a DNA sample.[12]
The Los Angeles Times reported that the daughter’s DNA sample was obtained from a medical clinic with a warrant; the New York Times reported that she gave the sample voluntarily in order to exonerate her father.[13]
Ad hoc partial DNA matching might occur in a criminal investigation where a technician, asked to compare a suspect’s DNA to a crime scene sample, does not find a perfect match but notices and reports a partial match, casting suspicion on the original suspect’s close family members. Many violent crimes are committed by close relatives of the victim, so in some cases it may be useful to compare the DNA profiles of the victim and the posited offender, to determine whether there is a partial match suggestive of a family relationship.
The potential use of partial DNA matching in criminal investigations raises specific ethical concerns in the context of mass screenings, colourfully referred to as ‘DNA dragnets’ in the United States. In a mass screening police take DNA samples from as many potential suspects as possible, usually in a defined geographical area. The first mass screening occurred in the United Kingdom in 1987, when samples were taken from 4000 men in Leicestershire in a murder investigation, before the offender was caught when he tried to convince another man to submit a sample for him. Probably the largest screening occurred the following year in Germany, when 16,400 people provided samples in a rape and murder investigation. In 1997 the first mass screening in Australia tested many of Perth’s 3500 taxi drivers in the investigation of the Claremont serial killings, which remain unsolved.[14]
In 2002 New South Wales police investigating the sexual assault of a 91 year-old woman in Wee Waa took DNA samples from 500 men aged between 18 and 45, most of the town’s male population in that age group. Stephen Boney confessed to the crime before his sample was analysed.[15]
Mass screenings are resource intensive and the DNA samples come from volunteers, not suspects or people previously convicted of crimes. As two of the above examples illustrate, mass screenings often succeed because their perceived effectiveness prompts the offender to confess, or because the offender’s behaviour in trying to avoid the screen casts suspicion upon them. Partial DNA matching has the potential to greatly increase the reach of mass screenings by effectively screening not just the pool of volunteers, but also their close relatives, who have not consented to assisting the criminal investigation.
Partial DNA matching is most powerful, and problematic, when used in conjunction with a DNA database. The first DNA database established for use in criminal investigations, the United Kingdom’s National DNA Database (NDNAD), commenced operation in 1995. It now contains more than 3.4 million profiles, representing more than 5% of the United Kingdom’s population, and ‘the majority of the known active offender population’.[16]
More than £300 million has been invested in the NDNAD and it is estimated that it will eventually hold profiles from 25% of the United Kingdom’s adult male population and 7% of the adult female population.[17] In the United States the National DNA Index System (NDIS), contains 5.2 million profiles, representing about 1.7% of the US population.[18]
With about 1 million people convicted of a felony each year in the United States, it has been reported in relation to the growth of the national database that: ‘the rate-limiting factor has been the speed of states in spending the money needed to analyze the DNA samples they receive from convicts’.[19]
Australia’s National Criminal Investigation DNA Database (NCIDD) is much smaller. At 30 June 2007 it contained just over 300,000 profiles, representing 1.4% of the Australian population.[20] In 2005, Portugal announced its intention to create the first DNA database covering a country’s entire population. It will eventually contain samples from all 10 million Portuguese citizens.[21]
Williams and Johnson argue that, if a person’s DNA profile is stored in a DNA database that police can use, they are subject to ‘indefinite genetic surveillance’[22]
to which people whose profiles are not on the database are not subject. If police find the DNA of a person whose profile is on the database at a crime scene, that person can be quickly identified. If the crime scene DNA belongs to a person whose profile is not on the database, it is merely evidence that might be used against that person if police can locate them using other investigative techniques. In this way, a DNA database is similar to a fingerprint database, which makes it easier for the police to identify a person who leaves fingerprints at a crime scene only if they have previously been fingerprinted, usually after an arrest or conviction. The significance of partial DNA matching is that it extends the reach of genetic surveillance beyond the people whose profiles are stored on the DNA database, to also include their close genetic relatives. A workshop organized by the American Society of Law, Medicine and Ethics to examine this issue called searching for partial matches ‘an implicit database expansion that should be open to public debate.’[23 ]
The first use of partial DNA matching using a DNA Database in a criminal investigation occurred in the United Kingdom in 2002. DNA was extracted from evidence preserved from the crimes scenes of the murders of three women in 1973. The resulting profile was used to search the NDNAD, producing a partial match with Paul Kappen’s profile. After obtaining samples from his mother and siblings, DNA analysis was used to determine that the crime scene DNA probably came from Paul’s father, Joseph Kappen, who was deceased. Police obtained permission to exhume Joseph’s body and they confirmed that he was a direct match to the DNA found at the scenes of the three murders.[24]
In 2004 the United Kingdom’s Forensic Science Service reported that approximately 20 searches for partial DNA matches had been conducted using the NDNAD and about a quarter of them yielded ‘useful intelligence information.’[25] Partial matching using DNA databases has been used in several cases in the US,[26]
including one in which a man who had served 18 years in prison for rape and murder was exonerated using DNA evidence and a partial match was then used to identify the offender, who pleaded guilty.[27]
Williams and Johnson draw the following distinction between traditional uses of DNA evidence and searching for partial matches in DNA databases:
this technology is increasingly being applied inceptively rather than reactively. In other words, it shapes an inquiry by identifying potential suspects from the start rather than merely supporting their incrimination or exoneration after they have been nominated for attention by other more traditional – and often protracted – forms of investigative practice.[28]
Searching for partial matches in a DNA database is likely to yield a significant number of false positives, where the partial match is to the DNA profile of a person who is not related to the posited offender.[29]
In cases where the crime scene DNA profile contains rare markers there will likely be significantly fewer false positives, meaning the usefulness of partial matching in a criminal investigation is partly dependent on the genetic makeup of the posited offender.[30]
Put another way: ‘The partial match is only a lead – a relatively weak one for a common genotype though possibly a strong one for a rare genotype’.[31]
In most cases searching for partial matches on a DNA database will only be useful where police have other information – such as the region where the offender probably resides, or the offender’s suspected age or sex – to help them determine which partial matches to investigate.[32]
Even where such filtering criteria can be applied, investigating partial matches is likely to be resource intensive because police need to identify the close relatives of each person whose profile is a partial match before using filtering criteria to decide who to investigate further. An example from the United Kingdom illustrates this. In 2004, somebody threw a brick from a footbridge over a freeway in Surrey, England. The brick smashed through the windscreen of a truck, killing the driver, Michael Little. Blood that did not belong to Little was found on the brick, and its profile matched that of a sample taken from a car that had been broken into on the same evening. The DNA profile did not directly match any profiles on the United Kingdom’s DNA database. After a mass-screening of 350 volunteers in the area and the offer of a £25,000 reward for information yielded no useful results, a search for partial matches on the NDNAD was performed. There were too many partial matches to investigate them all, but the DNA sample belonged to a male and ethnic markers indicated that he was probably white,[33] so a decision was made to restrict investigation to white males aged less than 35 years-old and resident in Surrey and Hampshire, two counties near the crime scene with a combined population of 2.6 million. Police began by interviewing the person with the closest match to the crime scene profile and discovered that he had a 20 year-old brother, Craig Harman, who lived near the crime scene. Harman gave a DNA sample and confessed to dropping the brick when confronted with the evidence of a direct match. He was convicted of manslaughter.[34]
2.3 Regulation of the Use of DNA in Australian Criminal Investigations
In the Commonwealth jurisdiction the collection of DNA in criminal investigations and the NCIDD are governed by Part 1D of the Crimes Act 1914 (Cth). Each State and Territory has equivalent legislation,[35] most of which is similar to the Commonwealth Crimes Act.[36] Matching between profiles supplied to the NCIDD by different jurisdictions did not begin until 2005, and was initially limited to only a few participating jurisdictions. In 2007 a single ministerial agreement was concluded to allow matching between profiles supplied by all Australian jurisdictions, except New South Wales, which currently allows matches with five other jurisdictions subject to bilateral agreements.[37]
Australia has comparatively restrictive sample collection and profile retention rules, which mean that the NCIDD is unlikely to ever cover a portion of the population equivalent to the United Kingdom’s database. Under Commonwealth legislation, DNA profiles may be obtained from people who have committed offences carrying maximum penalties exceeding 2 years imprisonment.[38] Samples can also be taken from people suspected of any indictable offence[39] and from volunteers,[40] but a suspect’s DNA sample must be destroyed and the identifying information removed from the database after a period of 12 months if they have not been convicted[41] and a volunteer’s sample must be destroyed and their identifying information removed from the database if they withdraw consent, or after the lapse of a period of time agreed between the volunteer and the Commissioner of the Australian Federal Police when the sample is given.[42] Similar rules apply in most States and Territories,[43] but there is some variation. For example, in the Northern Territory the Commissioner of Police may retain samples for as long as he or she thinks fit.[44] By comparison, the United Kingdom has ‘the most inclusive and far-reaching legislative framework for authorizing the collection, storage and use of forensic DNA samples in the world.’[45] Samples are taken from all people suspected, charged or convicted of a ‘recordable offence’ which includes all offences that carry a sentence of imprisonment and many other proscribed offences.[46] Samples and profiles can be retained indefinitely and profiles will only be removed from the database in ‘exceptional circumstances’.[47] In the United States some States and the federal jurisdiction retain profiles from people who are merely arrested on suspicion of committing a felony, but not necessarily convicted, while most States retain profiles only from convicted felons.[48] Australia’s comparatively restrictive profile retention rules reflect a choice by Australian parliaments to prioritize civil liberties over efficient criminal investigations, when compared to the United States and the United Kingdom.
The NCIDD has seven indexes of DNA profiles: crime scene, suspects, volunteers (limited purposes), volunteers (unlimited purposes), serious offenders, missing persons and unknown deceased persons. Section 23YDAF of the Crimes Act 1914 (Cth) is headed ‘Permissible matching of DNA profiles’ and provides that matching is allowed between profiles on each of the indexes, except that profiles on the Volunteers (limited purposes) index may only be matched with other indexes if the matching is ‘within purpose’. Allowing volunteers to control how their profiles are used is another example of how the Act emphasizes the protection of civil liberties. It is not entirely clear whether s 23YDAF would allow partial matches between profiles on the database as a kind of ‘matching’. In any case, if partial matching is to be allowed it would be preferable if the Crimes Act 1914 (Cth) and the relevant State and Territory Acts made this explicit.
The informed consent provisions of the Part 1D of the Crimes Act 1914 (Cth) require that the person giving the DNA sample be told, amongst other things: the purpose for which the sample is being collected; that the forensic procedure may produce evidence that could be used against them in court; that their profile will be placed on the NCIDD; and how it may then be used.[49] If partial matching using the NCIDD is eventually allowed, people giving DNA samples must be informed of this. Ideally the Crimes Act 1914 (Cth) and the relevant State legislation should be amended to require specifically that the person be told that their profile may be used to incriminate a family member. The problem of whether to allow partial matching using profiles collected before a change to the informed consent procedures would still remain.
As a Commonwealth government agency CrimTrac is bound by the Privacy Act 1988 (Cth) and must run the NCIDD in accordance with the Act’s Information Privacy Principles.[50] The Principles are not directly enforceable, rather, an individual can complain to the Privacy Commissioner if they believe that a Principle has been breached. The Principles are also useful as a statement of Commonwealth government policy regarding privacy. Nothing in the Principles would prohibit partial DNA matching. Of relevance for the problem of whether to allow searches for partial matches with profiles collected before the introduction of informed consent to partial matching, Principle 10 provides that information collected for a particular purpose shall not be used for another purpose unless ‘the disclosure is reasonably necessary for enforcement of the criminal law’.[51]
The benefits of the use of partial DNA matching in criminal investigations are obvious – the technique can be used to aid police investigations and obtain criminal convictions, particularly in relation to violent crime. In the United Kingdom and the United States the technique seems to have so-far only been used in investigating rapes and murders, and there have been some spectacular successes solving ‘cold cases’. In deciding whether and how partial DNA matching should be used in Australian criminal investigations, these benefits must be weighed against potential negative consequences. The possibility of minimizing negative consequences through best practice and regulation must also be considered.
The use of ‘ordinary’ direct matching of DNA samples in criminal investigations has become entrenched over the last two decades,[52] yet a significant body of literature has examined its potential and actual problems including:
the threat to the bodily integrity of citizens who are subject to the forced and non-consensual sampling of their genetic material; the intrusion and denigration of privacy rights caused by the storage and use of tissue samples; the potential for the future misuse of such samples held in state and privately owned laboratories; the prospect of long term bio-surveillance occasioned by the storage of genetic information in police databases and biological samples in forensic laboratories; and the possibility for the deceptive use of DNA forensic evidence in police investigations and criminal prosecutions.[53]
This article focuses on problems specific to the use of partial DNA matching, which may overlap with or exacerbate some of these concerns. Information about investigations that have used partial DNA matching comes, whether directly or indirectly, from law enforcement authorities. While spectacular successes are likely to be publicized, negative consequences, including for individuals mistakenly cast into suspicion, are not. Nonetheless a few authors have discussed potential negative consequences of the use of partial DNA matching in criminal investigations.[54]
Their concerns can be divided into three categories; concerns about the efficiency of police investigations, concerns for individuals and families investigated and concerns for ethnic groups already overrepresented in the criminal justice system. This section considers these potential problems and the ways that they might be minimized through best practice and regulation.
3.1 Efficiency of Police Investigations
There are significant and legitimate concerns about the cost-effectiveness of DNA databases and mass screenings and ‘virtually no scientific, comprehensive, independent, peer-reviewed analysis quantifying the overall effectiveness of DNA databases in solving or preventing crimes’.[55]
The cost of performing a search for a partial match on a DNA database is negligible, but the cost of investigating a large number of partial matches, many of which may be false positives, could be great. It might be expected that police could be enthusiastic about the technique upon its introduction but through experience soon learn its comparative efficiency and worth. Although the Memorandum of Understanding governing the use of partial DNA matching in the United Kingdom is operationally sensitive and therefore secret,[56] it appears that searches for partial matches on the NDNAD are only used in the investigation of serious violent crimes (rape and murder)[57] and only then when other investigative techniques have been futile. Given that investigating partial DNA matches is likely to be resource intensive, limiting the use of this technique to serious violent crimes that police have failed to resolve using less resource intensive methods is a sensible approach.
In criminal investigations using a mass DNA screening the mere possibility that profiles could be subject to searches for partial matches could deter people from voluntarily giving samples, undermining the efficiency of the screening. The adage ‘nothing to hide, nothing to fear’ may apply to volunteers in mass screenings who know that they are innocent, but volunteers will not know whether their participation could eventually implicate a close relative. Fortunately, the legislation governing the NCIDD already provides a framework for dealing with this problem. Volunteers may consent to their profile being used for a limited purpose, specifically defined in each case, or they may consent to their profile being used for ‘unlimited purposes’ – being for any criminal investigation or other purpose for which the NCIDD may be used (statistical analysis, identifying disaster victims and unknown deceased persons).[58] Separate database indexes are maintained for ‘limited purposes’ and ‘unlimited purposes’ volunteers. All that would be required to facilitate the use of partial DNA matching in conjunction with mass screenings would be that volunteers are properly informed under the Crimes Act 1914 (Cth) consent procedures that partial matching using their profile could implicate their close genetic relatives. They could then refuse consent to partial matching, or consent to it only in relation to that specific investigation, or they could give informed consent to any partial DNA matching to their profile, which would then be stored on the unlimited purposes index.
3.2 Individuals and Families Involved in Investigations because of Partial DNA Matching
One category of individuals who will be adversely impacted by the use of partial DNA matching in criminal investigations are those who are mistakenly investigated, either because of false positives (suggesting a genetic relationship between two profiles which does not in fact exist), or because a partial match correctly identifies the genetic relationship, but the database profile belongs to a person who has more than one close relative who could be the posited offender. The number of false positives could be dramatically reduced by changing the profiling technique to measure the length of more STRs, producing more markers. Increasing the number of markers to around 30 would virtually eliminate false positives, even if the Australian database was much larger,[59] but this would also raise the possibility of using partial matches to identify second-degree genetic relatives,[60] creating a new problem of false positives if the more powerful technique was used in this way.
The investigation of innocent people is an inevitable consequence of law enforcement, but the impact on those wrongly investigated as a result of partial DNA matching can be minimized if police are properly trained about the meaning of partial matches, particularly regarding the fact that they indicate only a probability that two individuals are related.
All use of forensic DNA in criminal investigations can be seen as impinging on the bodily integrity and privacy of the person from whom a DNA sample is taken, but partial DNA matching raises specific privacy issues in the context of family relationships. It has been argued that partial DNA matching effectively makes the person whose profile is in a DNA database a ‘genetic informant’ on their family members.[61] This may cause significant stress for the individuals concerned and their family relationships. Revelation that a family member’s DNA profile is on the NCIDD may also effectively be a revelation that they have been convicted of a quite serious criminal offence, impinging on that person’s privacy right to keep that information secret from their family members.
Investigations of partial DNA matches may also reveal to investigators that a person who the ‘genetic informant’ believes is their genetic relative is not in fact genetically related to them.[62] This information can be doubly significant because of what it reveals or suggests about family members’ past actions – such as adultery or adoption.[63] Greely et al argue that no harm can come from investigators discovering the non-existence of an assumed genetic relationship, so long as that information is not communicated to the relevant family members. But this can still be seen as an invasion of the family’s privacy right to keep information about their genetic relationships secret from investigators. As Haimes rightly observes: ‘The awareness that “others” (the police, the government) have information about “me” and “us” that I and we did not give them can be threatening, not just because “others” might have knowledge of ‘our genes’ but also because “they” have knowledge of “our family”’.[64] Haimes also argues that even if investigators do not themselves reveal what they discover about genetic links, that information may be revealed if the nature of the criminal investigation prompts family members to start asking unwanted questions of each other.[65]
Whether to reveal unanticipated information about genetic relationships is a question that arises far more frequently in a medical context than in law enforcement, although it may also arise in a family law context where paternity is disputed. Many authors argue that unanticipated information about genetic relationships should always be revealed, a position that been strengthened by the view that ‘individuals need to know their genetic history so that they can know their medical history and, by implication, their future’.[66] Based on studies of adoption and artificial insemination, Haimes argues that families are more robust than is often feared, and that ‘individuals pick and choose which relationships to value, despite the presence, or absence, of genetic connections with other people’.[67] Haimes also states that ‘the style of the revelation, and the context in which it done, appear to matter to family relationships as much as what is actually revealed’.[68] Unfortunately any such revelations arising in the course of a criminal investigation will likely come at a time of stress for the family concerned, but this can be equally true of the revelations in a medical setting and is certainly true of paternity disputes in family law proceedings.
Negative consequences for individuals and families that are investigated because of partial DNA matching can be minimized by ensuring that investigators understand and can properly communicate the meaning of partial DNA matches and that genetic counselors are available, as they would be in a medical setting. To some extent, these harms may be unavoidable, and so infringement of individual and family privacy rights should be balanced against the interests of law enforcement in deciding whether the technique should be used to investigate any particular crime.
3.3 Ethnic Groups already Overrepresented in the Criminal Justice System
Greely et al aptly summarize the fear that: ‘although using familial DNA profiles to identify suspects has no logical connection with the idea that crime runs in either genetic or social families, merely putting the words “crime”, “family” and “DNA” in a sentence might conjure such an association for some readers’.[69] This is a concern best dealt with through appropriate public education, including careful explanation in any publicity of the use of partial DNA matching in police investigations.
Far more concerning is the disproportionate impact that partial DNA matching may have on ethnic groups already over-represented in the criminal justice system. This issue has been raised mostly in the United States, where African-Americans make up 13% of the population but are convicted of 40% of all felonies.[70] This means that African-Americans are disproportionately represented in the State and federal DNA databases. Because of the ability to search for partial matches on those databases, a far greater portion of the African-American population than the white population is subject to genetic surveillance, not because they are convicted felons, but because they are related to one. In Australia Aboriginal people are similarly over-represented in the criminal justice system and concern not to exacerbate this situation is a further reason to restrict the use of partial DNA matching to investigation of only the most serious crimes. Nonetheless Rothstein and Talbott make a good point when they say, discussing the United States:
The overrepresentation of minorities in offender DNA databases reflects the over representation of minorities in the criminal justice and correctional systems, and this overrepresentation is caused by sentencing policies that disparately impact minorities as well as gross disparities in education, employment, housing, health care, and other essential life opportunities. The overrepresentation of minorities among arrestees is caused by ‘racial profiling’ and other dubious law enforcement practices. Having a disproportionate number of minority criminals is a social problem, not a DNA database problem.[71]
One way to eliminate the problems caused specifically by the use of partial DNA matching in criminal investigations would be to establish a universal national database containing every Australian’s profile. If such a database existed ‘family searches would be unnecessary, except for people who had somehow slipped through the cracks of the universal system, and the database would not just be representative of the country’s ethnic and racial diversity, but would fully embody that diversity’.[72] This is an interesting point, but a universal database would be expensive, and probably controversial.[73] Simply limiting the use of partial DNA matching to the investigation of serious violent crime, with a system of judicial oversight, is an approach more in keeping with Australia’s emphasis on civil liberties over the interests of law enforcement in the use of forensic DNA. The problems caused by the use of partial DNA matching in criminal investigations cannot be completely eliminated, but they can be minimized by appropriate amendments to existing Australian legislation, the deployment of genetic counselors where appropriate, and the proper training of law enforcement officers. Because training police officers is itself expensive and time consuming, and investigations using partial DNA matches are likely to remain infrequent, the most practical approach may be to establish a police unit that is specially trained to conduct investigations that use partial DNA matches. Genetic counselors could assist the unit to understand partial matches and be available to counsel individuals and families investigated.
Even where best practice is used to minimize the negative consequences of allowing partial DNA matching in criminal investigations, some negative impacts cannot be eliminated. This must be considered in deciding whether to use partial DNA matching in the investigation of any particular crime. In Australia, judicial oversight is a common mechanism for supervising investigative authorities’ use of techniques that impinge upon privacy. For example, law enforcement authorities must obtain a warrant from a judicial authority before intercepting a telephone call, email or other private telecommunication.[74] A similar form of judicial oversight should be used to limit the circumstances in which partial DNA matching can be used in criminal investigations.
The Crimes Act 1914 (Cth) already provides for judicial oversight of the collection of DNA samples in certain circumstances. An order from a magistrate is required to obtain a DNA sample from a suspect who does not consent,[75] to retain a sample from a volunteer who subsequently withdraws consent,[76] or to obtain a sample using an invasive technique from a serious offender who does not consent.[77] Magistrates have a similar role in some States.[78] Under the Crimes Act 1914 (Cth) the matters that a magistrate must take into account before ordering that a sample may be taken from a suspect are quite extensive. Section 23WT of the Act states:
Matters to be considered by magistrate before ordering the forensic procedure(1) The magistrate must be satisfied on the balance of probabilities that:
(a) the person on whom the procedure is proposed to be carried out is a suspect; and
(b) on the evidence before him or her, there are reasonable grounds to believe that the suspect has committed a relevant offence; and
(c) there are reasonable grounds to believe that the forensic procedure is likely to produce evidence tending to confirm or disprove that the suspect committed a relevant offence; and
(d) the carrying out of the forensic procedure is justified in all the circumstances.
(2) In determining whether the carrying out of the forensic procedure is justified in all the circumstances, the magistrate must balance the public interest in obtaining evidence tending to confirm or disprove that the suspect committed the offence concerned against the public interest in upholding the physical integrity of the suspect.
(3) In balancing those interests, the magistrate must have regard to the following matters:
(a) the seriousness of the circumstances surrounding the commission of the relevant offence and the gravity of the relevant offence;
(b) the degree of the suspect’s alleged participation in the commission of the relevant offence
(c) the age, physical health and mental health of the suspect, to the extent that they are known to the magistrate or can be discovered by the magistrate (by asking or otherwise);
(d) if the suspect is a child or an incapable person – the welfare of the suspect;
(e) whether there is a less intrusive but reasonably practicable way of obtaining evidence tending to confirm or disprove that
(f) the suspect committed the relevant offence;
(g) if the suspect gives any reasons for refusing to consent – the reasons;
(h) if the suspect is in custody:
(i) the period for which the suspect has already been detained; and
(i) the reasons for any delay in proposing the carrying out of the forensic procedure;
(4) Without limiting the matters that the magistrate may take into account in considering, for the purposes of paragraph (3)(f), the intrusiveness of the forensic procedure, the magistrate must (where appropriate) take into account the religious beliefs of the suspect.
A similar provision requiring an order of a magistrate before a search for partial DNA matches on the NCIDD could take place should require the magistrate to consider the seriousness of the crime, the likelihood that the perpetrator may re-offend if not caught, the investigative techniques that have already been used and whether the investigation is likely to be successful without recourse to partial DNA matching. The magistrate should then be required to balance the public interest in the obtaining of evidence that might assist in the investigation of the offence against the public interest in preserving the privacy of individuals and families who might be subject to investigation because of the use of the technique. This approach would allow the Australia community to benefit from the use of partial DNA matching to solve serious violent crimes, where other techniques have failed, while minimizing the negative impacts of the technique by limiting its use.
Publicity of the successful use of partial DNA matching overseas will inevitably lead to calls for its use in Australia. This is most likely to occur when the community is faced with particularly horrific crimes or threatened by a serial offender. Despite its encroachment on privacy, the use of the partial DNA matching may be justified in order to solve serious violent crimes, but consideration of how to regulate the technique should occur before a situation where there is pressure to use it arises.
If partial DNA matching using Australia’s NCIDD is to be allowed, this should be made explicit in amendments the Crimes Act 1914 (Cth), and corresponding State legislation. Informed consent procedures should require that people giving samples are told specifically that their DNA sample could potentially incriminate a family member.
There are legitimate concerns about the impact of partial DNA matching on the privacy of individuals and families investigated, the efficiency of police investigations and disproportionate impacts on ethnic minorities already overrepresented in the criminal justice system. While these problems cannot be completely eliminated, the proper training of police officers to ensure that they understand and can communicate the meaning of partial DNA matches and the provision of genetic counselors will help to minimize the impacts on those investigated.
The judicial oversight mechanism proposed here would ensure that partial DNA matching is used only when its negative consequences are outweighed by its potential to assist in the investigation of a serious violent crime, where other investigative techniques have failed. This approach will limit problems that could be caused by the unchecked use of the technique, including unjustified expenditure of police resources and disproportionate impacts on ethnic groups already over-represented in the criminal justice system. Judicial oversight would ensure that the community could benefit from the technique when truly threatened by violent crime, while maintaining the emphasis on civil liberties that is currently found in Australian legislation governing the use of DNA in criminal investigations.
[1] The same technique is also called ‘low-stringency matching’ or ‘familial matching’. This article avoids the term ‘familial matching’ because although familial genetic relationships underlie the usefulness of the technique, the results are probabilistic; a partial match never definitively determines that two profiles come from related individuals. A complex calculation of the probability of two profiles coming from genetic relatives can be made in each case, see David R Paoletti, Travis E Doom, Michael L Raymer & Dan E Krane, ‘Assessing the Implications for Close Relatives in the Event of Similar but Non-matching DNA Profiles’ (2006) 46 Jurimetrics 161.
[2] This article uses the term ‘posited offender’ to describe the person whose DNA is found at a crime scene in circumstances that lead police to believe that it belongs to the offender. There may be other explanations for a person’s DNA being labeled ‘crime scene DNA’ – it may have been left at the crime scene innocently, or there may be a technical mix-up of samples.
[3] CrimTrac, ‘Disaster Victim Identification Assistance’ http://www.crimtrac.gov.au/systems___projects_disaster_victim_ identification_assistance.html (viewed 25 February 2008); CrimTrac, ‘Kinship Matching’ http://www.crimtrac.gov.au/systems___projects_ kinship_matching.html (viewed 25 February 2008).
[4] CrimTrac, ‘Other Projects – Familial Matching’ http://www.crimtrac.gov.au/systems___projects_familial_matching. html (viewed 25 February 2008).
[5] STRs are also known as ‘microsatellites’.
[6] The four letters in the genetic code, G, A, T and C, represent the four DNA ‘bases’, Guanine, Adenine, Cytosine and Thymine, molecules that join in two long chains to form the larger DNA molecule.
[7] The exact chance of two people who are not identical twins having the same profile varies for each specific profile, because some STR lengths occur with greater frequency in the population than others: Jeremy Gans and Gregor Urbas, ‘DNA Identification in the Criminal Justice System’ (2002) Trends and Issues in Crime and Criminal Justice No 226; CrimTrac ‘DNA Profiling Techniques’ http://www.crimtrac.gov.au/ systems___projects_dna_profiling_techniques.html (viewed 25 February 2008); Australian Law Reform Commission, Essentially Yours: The Protection of Human Genetic Information in Australia, Report 96, March 2003 at 974.
[8] These figures are approximate and are based on figures stated in Henry T Greely, Daniel P Riordan, Nanibaa’ A Garrison & Joanna L Mountain, ‘Family Ties: The Use of DNA Offender Databases to Catch Offenders’ Kin’ 34 Journal of Law and Medical Ethics 248, 252 where the authors performed their own calculations based on data about the United States population and DNA profiling method. The Australian figures could be different because Profiler Plus uses different, and fewer, markers, and because of differences between the US and Australian populations.
[9] Greely et al, above note 8, 252.
[10] Paoletti et al, above note 1.
[11] State of Kansas v Dennis Rader, 18th Judicial District Court of Kansas, Sedgwick County, Case No 05 CR 498.
[12] Greely et al, above note 8, 249.
[13] S Simon and P J Huffstutter, ‘Clues Were Clear but Slow to Add Up’, Los Angeles Times, 6 March 2005, A1; M Davey, ‘Computer Disk Led to Arrest in Killings, Pastor Says’, New York Times, 2 March 2005, A12; both cited in Greeley et al, ibid.
[14] Australian Law Reform Commission, above note 7, 1022.
[15] Ibid; Gans and Urbas, above note 7, 3; CrimTrac, ‘Key dates in the History of DNA Profiling’, http://www.crimtrac.gov.au/systems___ projects_key_dates_in_the_history_of_forensic_dna_profiling.html (viewed 25 February 2008).
[16] United Kingdom Home Office, ‘The National DNA Database’ http://www.homeoffice.gov.uk/science-research/using-science/dna-database/ (viewed 25 February 2008). The figure of 3.4 million is correct as of the end of 2005. Based on the rate of additions to the database in the four years to 2005 (approximately half a million per year), the correct figure is likely to now be between 4 and 5 million.
[17] Robin Williams and Paul Johnson ‘Inclusiveness, Effectiveness and Intrusiveness: Issues in the Developing Uses of DNA Profiling in Support of Criminal Investigations’ (2006) 34 Journal of Law, Medicine and Ethics 234, 246.
[18] Combined DNA Indexing System, ‘NDIS Statistics’ http://www.fbi.gov/hq/lab/codis/clickmap.htm (viewed 25 February 2008).
[19] Ibid, 251; Mark A Rothstein and Meghan K Talbott, ‘The Expanding Use of DNA in Law Enforcement: What Role for Privacy?’ (2006) 34 Journal of Law and Medical Ethics 153, 154.
[20] CrimTrac, Annual Report 2006 – 2007, 21 available at http://www. crimtrac.gov.au/corporate_information_annual_reports.html (viewed 25 February 2008).
[21] Rothstein and Talbott, above note 19, 154.
[22] Williams and Johnson, above note 17, 236.
[23] Greely et al, above note 8, 255.
[24] Williams and Johnson, above note 17, 243.
[25] Ibid. Three of these cases are well documented and are discussed in Williams and Johnson’s article.
[26] Greely, et al, above note 8, 249; Tovia Smith, DNA and Familial Testing, Day to Day, National Public Radio (US), 28 February 2007, http://www.npr. org/templates/story/story.php?storyId=7641971 (viewed 25 February 2008).
[27] American Prosecutors Research Institute, ‘Catching Criminals by Investigating Profiles with Allelic Similarities’ (2006) 10 (2) Silent Witness 1, http://www.ndaa.org/publications/newsletters/silent_ witness_volume_10_number_2_2006.html (viewed 25 February 2008).
[28] Williams and Johnson, above note 17 at 234.
[29] The chance of any one person, unrelated to the posited offender, having a DNA profile that is a partial match to the crime scene DNA profile is very low, but the chance of finding one or more such people may not be low if the whole population represented on the database is considered. Paoletti et al, above note 1; Williams and Johnson, above note 17, 243.
[30] Greely et al, above note 8, 252-253; Williams and Johnson, above note 45, 243.
[31] Greely et al, above note 8, 253. A person’s ‘genotype’ refers to their genetic makeup, usually to which genes they have, although in this quote it refers to the STRs which they have. A person’s genotype is contrasted with their ‘phenotype’ describing their expressed physical characteristics.
[32] Ibid, 252-254; Williams and Johnson, above note 17, 243.
[33] While determining the sex of the DNA sample left by the posited offender seems to be uncontroversial, the use of DNA analysis to determine other characteristics, particularly likely race or ethnicity, is highly controversial; see Rothstein and Talbott, above note 19, 158.
[34] Forensic Science Service, Annual Report and Accounts 2004-05, http://www. forensic.gov.uk/forensic_t/inside/about/docs/04_05.pdf (viewed 25 February 2008).
[35] Police Powers and Responsibilities Act 2000 (Qld); Criminal Investigations (Identifying People) Act 2002 (WA); Crimes (Forensic Procedures) Act 2000 (NSW); Forensic Procedures Act 2000 (Tas); Crimes Act 1958 (Vic); Criminal Law (Forensic Procedures) Act 2007 (SA); Crimes (Forensic Procedures) Act 2000 (ACT); Police Administration Act 1978 (NT).
[36] Particularly Police Powers and Responsibilities Act 2000 (Qld); Criminal Investigations (Identifying People) Act 2002 (WA); Crimes (Forensic Procedures) Act 2000 (NSW); Forensic Procedures Act 2000 (Tas); Crimes (Forensic Procedures) Act 2000 (ACT); one significant difference in the Police Administration Act 1978 (NT) has already been noted.
[38] Crimes Act 1914 (Cth), Part 1D, Division 6A.
[39] Ibid, Division 3.
[40] Ibid, Division 6B.
[41] Ibid, Division 8 and s 23YDAG.
[42] Ibid, ss 23XWT(2)(b) and 23YDAG.
[43] For example see Police Powers and Responsibilities Act 2000 (Qld); Criminal Investigation (Identifying People) Act 2002 (WA); Crimes (Forensic Procedures) Act 2000 (NSW).
[44] Police Administration Act 1978 (NT), s147C.
[45] Williams and Johnson 17, 236.
[46] Ibid, footnote 8.
[47] United Kingdom Home Office, ‘National DNA Database documents and other information’: http://www.homeoffice.gov.uk/about-us/organisa tion/directorate-search/crcsg/ppod/fspu/national-dna-database-documents (viewed 25 February 2008).
[48] Greely et al, above note 8, 250.
[49] Crimes Act 1914 (Cth) s 23WJ (suspects); s 23XWJ (serious and prescribed offenders); s 23XWR (volunteers).
[50] Privacy Act 1988 (Cth), s 14.
[51] Ibid, Principle 10(d).
[53] Williams and Johnson, above note 17, 235.
[54] Erica Haimes, ‘Social and ethical issues in the use of familial searching in forensic investigations: insights from family and kinship studies’ (2006) 34 Journal of Law and Medical Ethics 263; Williams and Johnson, above note 17; Greely et al, above note 8; Rothstein & Talbott, above note 19.
[55] Rothstein and Talbott, above note 19, 154; Williams and Johnson, above note 8, 241.
[56] Williams and Johnson, above note 8, 244.
[57] Haimes, above note 54, 272 citing B. Bramley, ‘Summary of the presentation from Dr Bob Bramley, Custodian, National DNA Database’, Minutes of the Human Genetics Commission Plenary Meeting, 11 February 2004.
[58] Crimes Act 1914 (Cth), s 23XWR.
[59] Greely et al, above note 8, 254.
[60] Half-siblings, cousins, aunts, uncles, grandparents and grandchildren.
[61] Haimes, above note 54, 269.
[62] Ibid. Comparison of two individuals’ profiles can conclusively determine that they do not have a genetic parent/child relationship, but unless the DNA of a third first-degree relative is also analysed, cannot be conclusive in the case of siblings, because there is always a small chance that genetic siblings will not share any profile markers. Despite contrary assertions of some writers (Williams and Johnson, above note 8, 244; Haimes, above note 54, 269-70), Greely et al are correct in their analysis that it is difficult to envision how partial DNA matching could reveal to the ‘genetic informant’ a genetic relationship which they did not previously know existed, because the technique relies on them nominating their close relatives (Greely et al, above note 8, 255). This confusion in the literature demonstrates the need to think carefully about what consequences will and will not flow from the use of the technique. Greely et al also observe that: ‘If an offender names someone as a relative who does not suspect that relationship, DNA analysis could confirm that the relationship was possible or even likely, but, except for identical twins, it would not provide definite proof.’(Ibid).
[63] Haimes, above note 54, 269.
[64] Haimes, above note 54, 271.
[65] Ibid, 272.
[66] Ibid, 266-7.
[67] Ibid, 268.
[68] Ibid, 267.
[69] Greely et al, above note 8, 255.
[70] Ibid, 258.
[71] Rothstein and Talbott, above note 19, 155.
[72] Greely et al, above note 8, 255.
[73] If a universal DNA database for use in law enforcement seems to evoke ‘Big Brother’ in an discomforting way, we should consider whether, from an individual’s perspective, being required to submit a DNA profile to a universal database is really very different to having a close family member’s profile stored on a database on which searches for partial matches may be performed.
[74] Telecommunications (Interception and Access) Act 1979 (Cth), Part 2-5.
[75] Crimes Act 1914 (Cth), s 23WQ-R.
[77] Ibid, s 23XWO. An order of a judge or magistrate is required if the serious offender does not consent, and the sample will be taken by an ‘intimate forensic procedure’, which includes the taking of blood, or a buccal swab of the inside of the cheek, see s 23WA.
[78] See for example Police Powers and Responsibilities Act 2000 (Qld), s 458; Crimes (Forensic Procedures) Act 2000 (NSW), Part 5.
AustLII:
Copyright Policy
|
Disclaimers
|
Privacy Policy
|
Feedback
URL: http://www.austlii.edu.au/au/journals/JlLawInfoSci/2006/5.html