Forensic genetic databases: ethical and social dimensions
Please note that this is not the final version of this Encyclopedia entry. The final version is available at: http://www.sciencedirect.com/science/article/pii/B9780080970868820620 Please cite as: Prainsack B, Aronson J (2015). Forensic DNA databases: Ethical issues. International Encyclopedia of the Social and Behavioral Sciences 9: 339-345.
Barbara Prainsack Department of Social Science, Health and Medicine King’s College London Strand London WC2R 2LS UK Email:
[email protected] Phone: +44-(0)7912622901 Jay D. Aronson Department of History--240 Baker Hall Carnegie Mellon University 5000 Forbes Ave. Pittsburgh, PA 15213 USA Email:
[email protected] Phone: 1-412-268-2887
Abstract This entry starts with a brief summary of the origins of forensic genetic profiling and databasing. It then provides a typology of forensic genetic databases and discusses key societal and ethical issues related to different configurations of forensic genetic databases across countries. It concludes with an overview of the most current ethical and societal concerns related to DNA databasing in the context of criminal investigation and criminal justice. Keywords crime control, criminal justice, DNA databases, ethnic discrimination, forensics, police, surveillance 1
Background: The beginning of forensic genetic profiling Forensic genetic databases are important tools for criminal investigation and criminal justice in many countries across the globe (Hindmarsh and Prainsack 2010). An answer to the question of when the history of forensic genetic databases began is not a straightforward one. Most official histories of forensic genetic profiling (not databasing as such) start with a discovery that geneticist Sir Alec Jeffreys made in his lab in Leicester, England, in 1984: After finding regions of the mammalian genome that had a particularly high degree of length variability (due to the presense of differing numbers of short repetitive sequences of DNA), Jeffreys and his colleagues decided to find out whether similar variations could also be found in the human genome. Preliminary experiments revealed not only the existence of these polymorphisms (i.e., variations in length) in humans, but also that patterns tended to be inherited in a regular fashion. The more distant the genetic relation between family members, the fewer similarities their DNA profiles displayed on average; the closer the genetic relation, the more similarities in their DNA profile. In cases of monozygotic twins and multiples – who share virtually 100 per cent of their DNA, as well as in cases where two DNA profiles come from the same person, all the markers seen in the DNA profile are identical. Jeffreys called his discovery ‘DNA fingerprinting’ in order to borrow some of the legal and social credibility of its namesake (Lynch et al. 2008). This technique, as well as refinements of the basic idea made by researchers around the world, rendered possible the comparison of people’s genetic make-up to establish whether two (or more) DNA profiles came from the same person. This, in turn, enabled criminal investigators also to link particular people to particular crime scenes (e.g. if, though comparison of DNA profiles, it was found that a person’s DNA profile matched that of a biological trace found at a crime scene). Despite this effort of establishing DNA analysis as a reliable form of evidence, the use of DNA technologies in criminal investigation and in the courtroom was highly contested; experts and practitioners engaged in heated discussions about the scientific robustness and reliability of the technology, as well as various issues concerning the probability that specific DNA variants or even entire profiles might be shared by two different individuals (Aronson 2007; Lynch et al. 2008; Kaye 2010). These controversies are now referred to as the ‘DNA wars’; they were de facto, if not in principle, resolved in the course of the 1990s when experts moved on from discussing whether or not DNA technologies were fit to be used as evidence at all to which tools and methodologies were most robust in the criminal justice context and what regulatory mechanisms needed to be put into place to ensure their validity and reliability. How does forensic genetic profiling work? 2
In contrast to fingerprints, which are recorded in the form of (annotated) visual data, information from modern polymerase chain reaction (PCR)-based DNA profiling techniques is recorded as sets of discrete, numeric data (specifically, the size of a particular DNA fragment). When we look at a predefined set of specific places on the genome, then we find that different people have different numbers of so-called ‘short tandem repeats’ (STR) in these places. Like the variant regions of the genome that Jeffreys focused on, STR are repetions of chains of nucleotides (nucleotides make up DNA). The greater the number of such places (loci) on the genome that are analysed, the smaller the likelihood that two different people coincidentally have the same number of repetitions in all places. The chance that two people do have the same number of repetitions in the same places - the so-called ‘random match probability’ – increases not only when fewer loci are analysed, but also when the number of repeats in a certain locus are relatively frequent in the respective population (in other words, when they are fairly ‘typical’ for this population), or when individuals are biologically related. For example, full siblings share 50 per cent of their DNA on average. This means that on average, half of the genetic information (number of repeats) in the loci that forensic tests analyse will be the same. This phenomenon is utilised in a technique called ‘familial searching’. This method assumes profiles that match in many but not all loci an indication that the real perpetrator could be a biological relative of the person whose profile produced the partial match (Murphy 2010, Kaye 2013; see also entry on Forensic Genetics). The basic underlying principle of STR-based genetic analysis is thus one of stratification – of assigning people to different groups according to their genetic markers – and not of absolute individualisation (Amorim 2012). This is an important caveat with respect to this technology, and an important difference to fingerprinting, for example, where we assume that fingerprints are indeed unique for every individual. (It should be noted that the reliability of fingerprint analysis is increasingly being challenged, partly as a result of the emergence of DNA profiling as the gold standard of good forensic science. The problem here, however, lies with methods and practices of comparison, and not with the uniqueness – or lack thereof – of individual fingerprints; Cole 2001; Dror & Mnookin 2010; Mnookin et al. 2011.) Despite the probabailistic nature of DNA profiles, they are often treated in criminal justice as if they could be unambiguiously linked to one person. There are two main reasons for this: The first reason is that although the results of a DNA profile comparison is always probabilistic and never certain, random match probabilities (the chance that a profile matches a randomly drawn profile) are often so high that they become forensically irrelevant. For example, in the context of the world population presently being just under 7 billion, if we were faced with a random match probability of a given profile of 1:10 trillion, it would be practically (albeit not statistically) impossible that the DNA profile 3
had not come from the originator of the matching profile (provided that there were no issues with the preservation and analysis of the profile, such as contamination, etc; see below). The second reason for DNA profiles being treated as if they allowed for the establishment of a certain link between a person and a profile is a phenomenon that has been referred to as ‘the CSI effect’. This label captures, broadly speaking, the notion that the wide popularity of technocentric television drama – series where the main heroes are not people, but technologies – has shaped our perception of what forensic technologies can and should do. The ‘CSI effect’ has its name from the widely popular U.S. series Crime Scene Investigation (CSI). It does not only apply to members of the public, but also to actors within the criminal justice system, such as jurors or judges (e.g. Cole and Dioso-Villa 2007; 2009, Kim et al. 2009), and even prisoners (Machado and Prainsack 2012). Forensic genetic databases: the international landscape For much of the first decade after its introduction into criminal investigation in the 1980s, DNA profiling was used strictly on a case-by-case basis.. This means that DNA samples were taken from crime scenes and suspects for specific investigations, but DNA profiles were typically not stored in a central database. The first central forensic genetic database was set up in England and Wales in 1994, with countries such as the Netherlands (1997), Austria (1997), and Germany (1998) following suit. Other countries in Europe (e.g. Italy, Greece, Malta, or Ireland) as well as many countries outside of Europe (e.g. many African countries), at the point of writing this entry early in 2013 [adjust date and status to the time of the last correction], still did not have centralised forensic DNA databases. In the US, a system for the archiving and comparing of DNA profiles was launched as a pilot program between the FBI and a subset of state and local laboratories in 1990 (Combined DNA Index System, CODIS), was enabled at the national level by the DNA Identification Act of 1994, and formally came into existence in October 1998 (Aronson 2010). Unlike the databases in many other countries, the Federal Bureau of Investigations (FBI) is not responsible for physically uploading or maintaining profiles to the database. Rather, it enables connectivity among a collection of databases controlled by state and local law enforcement agencies (around 190 at the time of publication of this chapter). Decisions about what offenses are worthy of inclusion in the database are determined by state legislatures, not the U.S. federal government, and while the FBI issues requirements for the laboratories that participate in CODIS, it does not fully regulate or control them. CODIS allows investigators in one jurisdiction to look for matches in the databses in jurisdictions other than their own. The CODIS database started out as a repository of DNA profiles of convicted sex offenders, it became more expansive over time, and now includes profiles of different types of convicted offenders and suspects, depending on provisions at state level. While some states restrict inclusion criteria to serious felonies, others allows inclusion for misdemeanors or even arrests. Today, CODIS stores profiles of over five million people, that is, about 1.5 per cent of the nation’s total population [check 4
and adjust figure closer to publication date]. In terms of the absolute number of profiles stored, it is the largest DNA database in the world. Within the European Union, the so-called Prüm Decision (2008) required EU member countries that had not established centralised forensic databases to set them up by August 2011, in order to allow authorised actors in law enforcement in each country to search for DNA profile matches across DNA databases in the whole of the EU. The Prüm Decision, while supported by countries such Germany, Austria, and The Netherlands, faced considerable resistance from other countries, such as the UK [closer to production date: update whether the UK has opted out of pre-Lisbon acquis]; and some countries did not make the 2011 deadline due to financial or political struggles (see Prainsack and Toom 2012). While the Prüm Decision is binding for EU member countries only, some of the substantive provisions of the Decision are mirrored in bilateral treaties signed between the US and individual European countries (at the time of writing this entry, these bilateral treaties focused mostly on the exchange of fingerprints; yet they were devised in such a way that they can be extended to the exchange of DNA profiles by signing an additional protocol). Infrastructures for the transnational exchange of forensic bioinformation (DNA and fingerprints), however, predate the Prüm regime. Interpol has been providing an infrastructure for forensic DNA data exchange internationally since 2003. Every member country of Interpol hosts a national Interpol contact point where dedicated matching software is installed on a separate computer system. No nominal data (names, addresses, etc of existing persons) are included in the system, nor it is connected to any other forensic information network. Member countries can define which countries they do and not want to compare data with. The rationale for excluding some countries from the range of those that a particular country compares profiles with can be underpinned by concerns regarding how that other country runs their database, or it can be underpinned by purely practical considerations (for example, European countries which have already ratified the Prüm regime compare their data only via the Prüm network and thus ‘exclude’ each other in the list of countries with which they exchange data via the Interpol DNA Gateway). Whether individual member countries respond to queries from other countries by checking whether the profile matches any profiles in their own database is up to them, and response rates vary considerably across countries. Although the Interpol DNA Gateway stores the DNA profiles that member countries upload to it in a central system, it was never meant to replace national DNA databases; its role is to provide a complementary structure dedicated to the exchange of profiles across borders. While the Interpol DNA Gateway shares this latter goal with the Prüm regime, an important difference between the two – besides the international v. European remit, and the degree of automatisation of the comparison – is that the Prüm system is not a physical database at all: it is 5
merely a mechanism allowing mutual access to the database of EU member countries (i.e. member countries technically search certain elements of each other’s databases directly. See Prainsack 2010). A typology of forensic DNA databases There is no universal set of criteria according to which forensic genetic databases in different countries are organized. There are, however, certain dimensions that can help us to understand some core differences and similarities (see Table 1). classification criterion
features
1
profile inclusion criteria
restrictive / expansive
2
profile retention criteria
restrictive / expansive
3
sample retention
yes / no
4
genetic markers tested
dependent on type of chip and software used (e.g. Next Generation Multiplex - NGM, CODIS)
5
tendency towards expansion, ‘function
yes / no (and how exactly)
creep’ 6
responsibility for, and custody of, the
police / judiciary
database 7
Profiles come from
labs run by the government / private labs / mixed systems
Table 1: central dimensions of a typology of forensic genetic databases (source: authors) The first such dimension are the criteria for profile inclusion. Some countries (and American states) include in their forensic genetic databases only DNA profiles of convicted offenders, or also of suspects of serious offences (for the latter group, in some countries profiles are removed from the database by authorities when these suspects are not convicted later; for more details see Hindmarsh and Prainsack 2010). The underlying rationale for this restrictive approach towards profile inclusion is often a combination of two assumptions on the side of the authorities: The first assumption is that it is mostly a particular subset of the population that commits crimes, and that including a wide range of people, or even the entire population, in the DNA database would increase the administrative burden without significantly increasing detection and conviction rates. The second assumption is that most members of the public do not like the idea of having their DNA profiles stored in any centralised database under governmental control, and that therefore, storing profiles only of convicted offenders or suspects of serious crimes represents a good compromise. Other databases, such as the National DNA Database (NDNAD) in England and Wales, allow for the inclusion also of profiles of people 6
who committed very minor offences, or who were only arrested, but never charged, or later acquitted. In the United States, the DNA Fingerprinting Act of 2005 allows the federal government to include the profiles of all persons arrested for federal crimes as well as all non-U.S. detainees. The later catergory is, of course, directly relevant to the American ‘war on terror’. Another major issue in the United States is that state and local authorities often keep a non-CODIS (offline) database with wider inclusion criteria that they use for local investigations. A second, related dimension are the criteria for the retention of DNA profiles. In some jurisdictions, such as, again, England and Wales, once somebody’s DNA profile has been included in the central forensic DNA database, it can stay there indefinitely. Such an expansive approach towards profile retention is typically underpinned by the assumption that people who start out by committing smaller offences often ‘mature’ towards committing more serious crimes, and thus, that having their DNA stored in the database can help increase detection and rates later on. Moreover, legal provisions stipulating that profiles need to removed from the database a certain number of years after inclusion – provided that the profile’s originator has not been found to have committed any offences or crimes in that period, or when the originator has reached a certain age – always imply a duty on the side of the relevant authorities to actually remove the DNA profiles. This is often logistically difficult, and especially also against the backgrop of scarce human and financial resources. A result of the coexistence of restrictive and expansive profile inclusion and retention criteria is that the proportion of a country’s (or state’s) population that has DNA profiles stored in a database for police and forensic purposes varies widely. At the end of 2012, the NDNAD in England and Wales had DNA profiles from the greatest proportion of their total population stored – roughtly ten per cent. (These percentages can change fast when the inclusion and retention criteria in a country change.) Also, while most countries and states include predominantly profiles of convicted offenders, others include profiles also of mere arrestees (who were not charged, or not convicted), as well as profiles of other groups such as victims and volunteers; the latter two groups are included to enable profiles from crime scenes to be excluded from the investigation. For example, if several DNA samples are taken at a crime scene, it can happen that some of the samples taken stem from people who are not suspects, i.e. the victim, family members, friends, or neighbours of the victim, etc. Profiles of these people are often collected – and in some countries, included in the database – to make sure that police does not waste time investigating leads coming from DNA profile matches that are unlikely to have come from the suspect. A third dimension in the classification of forensic genetic databases is sample retention. A DNA sample is the physical biological matter – e.g. a blood or semen stain, a trace of saliva, etc. – that a 7
DNA profile is derived from. DNA samples are never as such included in the dabatase (only profiles are), but practices vary greatly among countries regarding the storage of DNA samples. While some countries, such as Germany, and at least one U.S. state, Wisconsin, destroy samples as soon as the DNA profile has been derived from it, many others, including most US states, store the samples. Especially the retention of so-called subject samples – that is, samples that were not collected at crime scenes, but that were taken by means of a buccal swaps from certain groups of people, such as convicts, suspects, arrestees and, in some countries, also volunteers and even victims – has attracted a great deal of concern. Critics of subject sample retention are worried that stored samples could be used for more wide ranging genetic analysis; for example, the samples could be searched for markers that disclose information about genetic predispositions to diseases or personality traits (which is not possible to infer from the information contained in a traditional DNA profile stored in the DNA database). Many of those who support sample retention, in contrast, hold that this practice poses relatively few additional privacy risks, yet that it has considerable practical advantages, for example, for EU countries in the context of the aforementioned Prüm regime: Here, certain kinds of preliminary DNA ‘matches’ need to be confirmed by an additional DNA analysis. If the sample has not been retained, then police would need to find the originator of the DNA profile and obtain another sample, both of which may not always be feasible. A fourth dimension in our typology pertains to the set of markers included in a forensic DNA profile. Not all countries include the same set of markers in DNA profiles stored in their databases. In Europe, the so-called European Standard Set (ESS) of loci was endorsed by the European Council in 2001. The ESS is a ‘minimal consensus’ of markers that all EU countries are encouraged to include in their DNA profiles to allow comparison across countries, regardless of the particular chip and software they use for sample analysis and profile storage. In the 2nd half of the 2000s, the European DNA Profiling Group (ENDAP), an association of forensic scientists, proposed the addition of five additional, new STR loci to the ESS (Gill et al, 2006) in order to enhance the discriminating power of DNA profile compariosns. The EU Council enforced this in a Council Resolution in 2009. In the US, plans to increase the CODIS markers from 13 to 24 (Hares et al. 2012) were still under discussion at the time of writing this entry. Interpol, because of their role as faciliators of bioinformation exchange for police and forensic purposes internationally, have developed a ‘minimal consensus’ set of seven loci that are included in most databases in the world (the so-called Interpol Standard Set of Loci, ISSOL). The five new ESS markers are now being added to the existing seven ISSOL, so that Interpol effectively uses 12 loci for transnational DNA profile comparisons. A fifth dimension pertains to whether or not the database is showing a tendency towards expansion. Such a tendency can manifest itself in a trend towards becoming more inclusive of the kinds of 8
profiles that are stored and/or retained in the national DNA database (such a trend is currently visible in the Netherlands, see for example Toom 2012, and in New York State’s 2012 decision to include profiles of all convicts,irrespective of whether the crime is mass murder or a minor misdemeanor, in its state database [New York State Division of Criminal Justice Services 2012]); or it could manifest itself in the form of an expansion of the functions for which DNA samples or DNA profiles are being used. This latter phenomenon of expanding the functions and purposes of samples and profiles have been referred to as ‘function creep’ (e.g. Dahl and Rudinow 2009). In some instances, the functions and purposes of materials initially collected and analysed in connection with a forensic genetic database become so widely extended that they are no longer limited to the core functions of the DNA database. The aforementioned method of ‘familial searching’ is an example for this. ‘Familial searching’, aka ‘genetic proximity testing’, can be performed if a DNA profile derived from a crime scene stain results in a partial match with a profile in the database. Police could then apprehend the person identified and investigate whether she has any biological relatives who could have committed crime
(Bieber
et
al.
2006;
Greely
et
al.
2008;
http://www.denverda.org/DNA/Familial_DNA_Database_Searches.htm). Genetic proximity testing can be seen as ‘function creep’ because it adds an additional feature to the existing functions in forensic DNA databases and consequently broadens the scope of individuals who are likely to be targeted by police investigations (see also entry Forensic Genetics). Function creep can also occur when databases become increasingly interlinked. In the European context, the aforementioned Prüm Decision is the most important example. A concern related to the large number of profiles stored in such database-networks is that the chance of false positive matches increases as well (for example, in a network that includes 10 million profiles, a random match probability of 1 in 10 million would mean that statistically, one profile in the database would falsely match the DNA profile in question). A sixth dimension relevant for classifying forensic genetic databases is the question in whose jurisdiction and under whose control a forensic DNA database is. While in some countries, the database is within the authority of the judiciary, whereas in others it is run by police authorities. The seventh and final dimension relates to the kinds of laboratories that carry out the genetic analysis (i.e. who derive the profiles form the DNA samples that are then uploaded to the database). In some countries, law enforcement authorities operate their own labs for this purpose. In other countries, also private and/or university labs who are properly accredited can provide DNA profiles for inclusion in the DNA database. In the United States, public and private labs coexist. In Great Britain, due to the recent closure of the Forensic Science Service (FSS), DNA profiling in has been almost entirely privatised—in keeping with the notion of austerity promoted by that country’s coalition government. 9
Ethical and societal concerns about the forensic genetic databasing DNA databases are not foolproof producers of truth. Besides the fact that, as discussed above, DNA profiling per per se does not produce certain, but only probabilistic links between profiles, there are challenges to be met at every stage of the process, from securing a DNA trace at the crime scene to presenting DNA evidence at court. At the stage of securing traces at the crime scene, insufficient resources and inadequate training of scene of crime officers can cause problems such as overlooking potentially usefully evidence, or causing contamination. Contamination occurs when a DNA sample is accidentally mixed with somebody else’s DNA – such as the scene of crime officer’s, the perpetrator’s, or a victim’s – without there being a record that this has happened (sometimes crime labs analyse traces from crime scenes where it is known that they contain DNA from more than one subject, e.g. from the victim and the perpetrator; such ‘mixed traces’ are not instances of contamination). The risk of contamination can be minimised, but never eliminated entirely, as it can happen also in light of very high standards. Contamination can take place also in other places than the crime scene, e.g. during transportation or storage: Samples could be swapped, misplaced, or deliberately tampered with. Once the sample has arrived at an accredited crime lab, the risk of contamination should in theory be very low, as all accredited labs are obliged to have mechanisms in place to avoid contamination (it should be noted, however, that not all labs in which DNA analyses for police or forensic purposes are carried out have accreditation). Such mechanisms include, depending on the lab, electronic tagging, video taping, or using different facilities for the analysis of samples from victims and those from suspects. Once lab workers have derived the DNA profile, it is uploaded to the central DNA database. We have already discussed issues involved in the calculation of random match probabilities in the course of profile matching. A growing problem is that although forensic scientists and those running DNA databases are usually well aware of what a DNA profile match can tell us about a case and what it cannot, other actors in the criminal justice system are often less aware of such limitations. Because not only many members of the public, but also many judges and jurors regard DNA science as infallible, numerical interpretations of DNA evidence are often taken as facts (related to this is also the aforementioned ‘CSI effect’). Another problem pertaining to using DNA evidence at court is the so-called ‘prosecutor’s fallacy’ in assessing the probabilities of guilt and innocence of a suspect based on the presentation of forensic evidence. For example, the expert witness in a case could state that the chance that a DNA obtained from the crime scene randomly matches somebody other than the defendant’s is one in one billion. In that case, the ‘prosecutor’s fallacy’ would be to misinterpret this 10
to mean that the chances that the defendant is innocent are one in one billion. This, of course, is incorrect, because even if it could be established with certainty that the DNA from the crime scene was indeed the defendant’s, that would not prove that the defendant committed the crime: her DNA could have been at the crime scene for legitimate reasons, or it could have been planted, or the sample could have been contaminated at some stage with the defendant’s DNA. Obviously, such ‘misunderstandings’ could have disastrous consequences for the defendant. The term ‘defence attorney’s fallacy’ is used to describe the opposite scenario: In light of a random match probability of, for example, one in one million, some defence attorneys argue that in a city of eight million inhabitants (half of whom can be assumed to be of the same sex as the defendant), there are four other people who could be the originator of the DNA. This is false conclusion, because it ignores all other circumstances of the crime (e.g. not all four million other people had access to the crime scene; some of these four million may be young children; etc.).
Overarching societal and ethical issues In sum, while there is wide agreement that DNA databases are a very helpful tool in criminal investigation and criminal justice more broadly, it is important to be aware of its limitations. Besides the limitations inherent in the (STR-based matching) technology itself, namely that it only ever produces probabilistic results, and that it is impossible to abolish the risk of contamination entirely, there are some overarching societal and ethical issues related to DNA databases for police and forensic purposes. An important one here is the issue of the over-representation of certain population groups. On the basis of known issues with racial discrimination in arresting, charging, and sentencing, it is plausible to assume that this overrepresentation of some minorities is, at least in part, due to such discriminatory practices on the side of law enforcement and state authorities (see Kaye 2013). In the United States, for example, there are significant racial disparities at all levels of the criminal justice system, from the likelihood of being arrested and convicted of a crime to sentencing and to representation in the prison population (Duster 2004). African Americans and Hispanics are the most at risk in these contexts (arrest rates are more than six times higher for blacks and Hispanics than for whites), suggesting that these minority populations will soon be overrepresented in CODIS as well. It should also be kept in mind in this context that behaviours that are more prevalent among underprivileged groups are more prone to be criminalised (Washington 2010). Because forensic genetic databases are seen to exacerbate existing ethnic and socio-economic biases, particularly in the context of familial searching. In this case, the high prevalence of minorties and the poor in the database may increase the likelihood that a poor person or person of color who commits a crime is found through the DNA of their close relatives. In the context of the United States, Levine et al. 11
(2008) worry that CODIS may soon become ‘Jim Crow’s database’ (i.e. represent an instance of de facto racial segregation). Some experts (including Alec Jeffreys) have argued in favour of universal DNA databases that include profiles of every citizen, or even every resident, of a country. No country has so far implemented this solution yet, although some have considered it (such as Portugal, for example). A further issue is that we have no solid data on the effectiveness of DNA databases. DNA evidence is rarely ever used in isoliation, both in criminal investigations and at court; other kinds of evidence, such as witness testimonials, and findings from other forensic analyses (finger- or footprints, toolmarks, etc.), are typically considered in conjunction with DNA evidence. Thus it practically impossible to measure the effect that DNA analysis and DNA evidence alone has had on an investigation or a trial. As a consequence, critics of (the expansion of) forensic DNA databases argue that our assumption that this is an effective tool for criminal justice is based on ideology rather than hard evidence. Critics also contend that the chanelling of ressources into DNA databases contributes to the technologisation of policing, where values such as human experience and judgment are increasingly replaced with machine-generated knowledge that receive excessive amounts of trust. Another overarching issue related to DNA databases is that although legal texts and technical protocols often speak of DNA samples that are given voluntarily, the voluntariness of such an act is dubious in light of the consequences of refusing to ‘volunteer’ as sample in most cases. ‘Volunteers’ fall mostly in one of two groups: The first group comprises people who are potential suspects, for the purpose of comparing their DNA with traces from the crime scene. The second group includes those who may have left their DNA at a crime scene for legitimate reasons (friends, employees, etc.). Such individuals are often requested to volunteer a DNA sample to eliminate their DNA profile from the crime scene stains. If an individual in the first group refuses to provide a sample, they typically attract suspicion, the assumption being that if people have nothing to fear, they have nothing to hide, and that in turn, refusing to volunteer a sample is a de facto admission of guilt. This is a very problematic conclusion, of course, because there may be many good reasons for a person wanting to protect her privacy other than the need to hide a criminal deed (similar as for people who draw curtains). In certain jurisdictions and contexts, people refusing to volunteer a sample even render themselves official suspects. If people in the second group refuse to provide a sample, they either attract suspicion as well – for example, a neighbour who was not among the suspects could suddenly become one – or it would be explained to them that providing a sample is in their best interest. In both cases, although an informed consent form would normally be signed, it remains questionable how voluntary such a consent can be considered to be. 12
With the increased ability to extract DNA from even minuscule biological samples, many law enforcement agencies have turned to “abandoned” DNA samples when suspects refuse to comply with requests for samples. In this procedure, police can retrieve DNA from saliva left on cigarette butts, classes, cans and bottles, pizza crusts, or even spit on the sidewalk. American law generally allows such uses of biological material since obtaining them requires neither access to the body nor a search warrant (because the sample is taken from a public, rather than a private, space (Winickoff 2004; Genetics and Public Policy Institute 2008). Finally, with the technological tools advancing fast, the scenario of using the full sequence of a person’s genome, rather than a selected set of genetic markers, for identification purposes may soon be technically and financially feasible. Despite the significant costs that changing DNA databases from STR-profiles to whole-genome information would incur, it is thus unclear for how much longer forensic genetic databases will stick to the present format. Many critics are concerned about governmental authorities holding whole-genome information of indivduals, because of the comparative wealth of information regarding disease predispositions and traits that can be obtained from a full sequence in contrast to a STR-profile. It should be noted, however, that also whole genome sequences do not predict the health or other phenotypic traits of a person entirely, with few exceptions. One of the major lessons from the Human Genome Project so far has been that genetic information alone typically tells us very little. Moreover, it is well possible that once people’s genome sequences are routinely held in the clinical domain, or even on a person’s own computer, it is possible that genomic information will be seen as less sensitive than it is the case today.
Cross References Forensic Genetics; Genotype and Phenotype; Race: Genetic Aspects; Ethical Issues in the New Genetics; Human Genome Project (HGP), history; ELSI initiative; The “New Genetics”; Bioethics and the New Genetics/Genomics; Genetics: Legal Aspects; Genetics and Society; Genetic Admixture; Bioethics in the Post-genomic era; Biobanking: ethical issues; Genetics and the Media; The “New” Genetics and Race; Genetics as a tool for social justice; DNA Phenotyping; DVI, use of genetics; Genetic Ancestry Testing; Genetics and Sociology; Science and Technology Studies, Experts and Expertise; Truth and Credibility in Science; Scientific Controversies; Expert Testimony; Law and Society: Sociolegal Studies; Privacy: Legal and Regulatory Aspects; Science and Law’ Scientific Evidence: Legal Aspects. 13
References Amorim, A. 2010. Opening the DNA black box: demythologizing forensic genetics. New Genetics & Society 31(3): 259-70. Aronson, J.D. 2007. Genetic Witness: Science, Law, and Controversy in the Making of DNA Profiling. New Brunswick, NJ: Rutgers University Press. Aronson, J.D. 2010. On Trial! Governing Forensic DNA Technologies in the United States. In: Hindmarsh, R. & Prainsack, B. (eds). Genetic Suspects: Global Governance of Forensic DNA Profiling and Databasing. Cambridge, UK: Cambridge University Press. 240-261. Bieber, F.R., Brenner, C.H., and Lazer, D. 2006. Human genetics: finding criminals through DNA of their relatives. Science 312:1315-1316. Cole, S.A. 2001. Suspect Identities: A History of Fingerprinting and Criminal Identification. Cambridge, MA: Harvard University Press. Cole, S.A. & Dioso-Villa, R. 2007. CSI and its effects: media, juries, and the burden of proof. New England Law Review 41: 435-470. Cole, S.A. & Dioso-Villa, R. 2009. Investigating the ‘CSI effect’: media and litigation crisis in criminal law. Stanford Law Review 61(6): 1335-1373. Genetics & Public Policy Center (2008). DNA Forensics and the Law: Issue Brief, http://www.dnapolicy.org/images/issuebriefpdfs/DNA,%20Forensics,%20and%20the%20Law%20Iss ue%20Brief.pdf Gill, P., et al. 2006. The evolution of DNA databases - recommendations for new European STR loci. Forensic Science International 156 (2-3): 242-244. Greely, H.T., et al. 2006. Family Ties: The Use of DNA Offender Databases to Catch Offenders’ Kin. The Journal of Law, Medicine & Ethics 34(2): 248-262.
14
Hares, D. R. (2012). Expanding the CODIS core loci in the united states. Forensic Science International: Genetics 6(1): e52-4. Kaye, D. 2013. The Genealogy Detectives: A Constitutional Analysis of “Familial Searching”. American Criminal Law Review 51(1) [add page numbers closer to publication date]. Kim, Y.S., et al. 2009. Examining the “CSI-effect” in the cases of circumstantial evidence and eyewitness testimony: multivariate path analyses. Jouranl of Criminal Justice 37: 452-460. Dahl, J.Y. & Sætnan, A.R. 2009. ‘It all happened so slowly’: on controlling function creep in forensic DNA databases. International Journal of Law, Crime and Justice 37(3): 83–103. Dror, I.E. & Mnookin, J.L. 2010. The use of technology in human expert domains: challenges and risks arising from the use of automated fingerprint identification systems in forensic science. Law, Probability and Risk 9(1): 47–67. Hindmarsh, R. & Prainsack, B. (eds.) 2010. Genetic Suspects: Global Governance of DNA Profiling and Databasing. Cambridge, UK: Cambridge University Press. Keating, B. et al. 2012. First all-in-one diagnostic tool for DNA intelligence: Genome-wide inference of biogeographic ancestry, appearance, relatedness, and sex with the identitas v1 forensic chip. International Journal of Legal Medicine [online first; update closer to publication date]. Krimsky, S. & Simoncelli, T. 2011. Genetic Justice: DNA Data Banks, Criminal Investigations, and Civil Liberties. New York: Columbia University Press. Lynch, M., et al. 2008. Truth Machine: The Contentious History of DNA Fingerprinting. Chicago: University of Chicago Press. Machado,H. & Prainsack, B. 2012. Tracing Technologies: Prisoners’ Views in the Era of CSI. Farnham: Ashgate. Mnookin, J.L. et al. 2011. The need for a research culture in the forensic sciences. UCLA Law Review 58: 725–79.
15
Murphy, E. 2010. Relative doubt: familial searches of DNA databases. Michigan Law Review 109(3): 291-348. New York State Division of Criminal Justice Services. 2012. The NYS DNA Databank and CODIS. Available at: http://www.criminaljustice.ny.gov/forensic/dnabrochure.htm Prainsack, B. 2010. Key issues in DNA profiling and databasing: Implications for governance. In: Hindmarsh, R. & Prainsack, B. (eds). Genetic Suspects: Global Governance of Forensic DNA Profiling and Databasing. Cambridge, UK: Cambridge University Press. 15-39. Prainsack, B. & Toom, V. 2013. Performing the Union: The Prüm Decision and the European Dream. Studies in the History and Philosophy of Science (in press) [update closer to publication date]. Toom, V. 2012. Bodies of science and law: forensic DNA profiling, biological bodies and biopower. Journal of Law and Society 39 (1): 150-166. Washington, H. 2010. Base assumptions? Racial aspects of US DNA forensics. In: Hindmarsh, R. & Prainsack, B. (eds). Genetic Suspects: Global Governance of Forensic DNA Profiling and Databasing. Cambridge, UK: Cambridge University Press. 63-84. Winickoff, D. (2004). The Constitutionality of Forensic DNA Databanks: 4th Amendment Issues, updated 2005. http://www.aslme.org/dna_04/reports/winickoff_update.pdf (Accessed 16 January 2013).
16
