The Trouble with DNA Evidence
DNA matching has come to be regarded as the “gold standard” in forensic profiling. It has even been described as “the hand of God pointing down at the guilty person”, and “the cornerstone of CSI”.
There is no doubt that DNA matching represents an important advance in forensic science. If a suspect’s DNA is found at the scene of a crime, it proves beyond any doubt that they were there.
Or does it?
Actually it doesn’t. It proves only that their DNA was left there. And that’s the elephant in the room of DNA evidence. An elephant whose presence police and forensic investigators have been reluctant to acknowledge.
Unlike fingerprints, DNA can be moved from one place to another, whether by accident or intentionally. If a person’s fingerprints are found at the scene of a crime it is possible to say with certainty that they were left there by that person. But DNA can be transferred from one location to another, or deposited on an object – a murder weapon, for example – with ease. And this is a serious and potentially fatal weakness with genetic evidence.
In a way it is remarkable that DNA evidence has come to be so widely accepted and trusted to the extent that it has. In the beginning, jurors were highly sceptical of genetic evidence. They distrusted it because they didn’t understand the science behind it, and because it dealt with statistics rather than solid facts. They found it hard to believe that a microscopic speck of tissue could be unequivocally identified as having come from a particular person. Defense lawyers took advantage of jurors’ lack of scientific knowledge to cast doubt on the reliability of DNA matches, and many criminals were acquitted despite there being what would now be considered overwhelming DNA evidence of their guilt.
In the OJ Simpson trial in 1995, for instance, jurors rejected the strong DNA evidence implicating Simpson in the murder of his wife, and returned a “not guilty” verdict. Interestingly, when these jurors were later interviewed on TV and asked why they had rejected the genetic evidence, several of them said it was the police, not the DNA itself, that they didn’t trust. In other words they suspected the police of planting OJ Simpson’s blood in order to improve their chances of getting a conviction. As it turned out, their suspicions were well-founded. The detective leading the case, Mark Furhmann, was subsequently accused of falsifying evidence after tape-recordings were played in court in which he could be heard making racist comments and admitting to beating up and intimidating black gang members. And this highlights a major problem with DNA evidence – the fact that the police may be tempted to plant genetic evidence in order to secure the conviction of a suspect they believe to be guilty. I will talk a little bit more about this later.
Simpson was acquitted despite DNA evidence implicating him in the murder of his wife Nicole.
Gradually, people came to accept that every individual’s “DNA fingerprint” was unique to them, and that the technique was a reliable method of identification. Within a matter of just a few years DNA evidence went from being distrusted to being accepted as the strongest type of forensic evidence. After all, the reasoning went, a person could change their name, their address, their identity, and even their facial features; but the one thing they couldn’t change was their DNA. Meanwhile, genetic evidence began to be featured – and glamourised – in movies and on TV shows. DNA became the buzzword in criminal investigation. Juries came to expect genetic evidence, and were disappointed – and less inclined to convict – when none was presented.
The thing that almost everyone overlooked in their enthusiasm for this exciting new crime-solving technique was that, although DNA matching could show that a microscopic quantity of genetic material came from a particular person, it didn’t prove that it had been deposited where it was found by that person. For some reason, the possibility of DNA being transferred was not, and still is not taken seriously, or given the consideration it deserves. The assumption is that if a person’s DNA is found at a particular location, they must have deposited it there themselves. In effect, DNA evidence is treated in the same way that fingerprint evidence is treated. The very term “genetic fingerprint” testifies to this comparison. If anything, genetic evidence carries more weight with jurors than fingerprint evidence, when in reality fingerprint evidence is far more reliable, and far less amenable to errors and falsification.
In fact DNA evidence falls a long way short of the hype that surrounds it. To get an idea of how weak DNA evidence really is, imagine what would happen if the criminal fraternity figured out a way to transfer fingerprints from one place to another, or from one object to another, as easily as a person’s DNA can be transferred. A defense lawyer would only have to point out that the prints offered in evidence might have been deliberately planted for a case to be dismissed, or at least for that particular evidence to be thrown out. And yet this is precisely the current situation with DNA evidence. It is automatically assumed that the sample being offered as evidence wasn’t moved or interfered with in any way. If a defense lawyer tries to raise the possibility that their client’s DNA was planted at the scene of a crime in a deliberate attempt to incriminate them, this is usually seen as grasping at straws.
Here’s another way to look at it. Imagine if, instead of finding DNA at the scene of a murder, the police instead found a business card. Would they automatically assume that the person whose name was on the card was the killer? Just imagine how that would play out in court: “The defendant must have been at the scene – his business card was found beside the body.”
There is an assumption – a naïve assumption, in my opinion – that it would never occur to a criminal to plant another person’s DNA, and that the police would never dream of planting their primary suspect’s DNA at a crime scene in order to improve their chances of securing a conviction. The reality right now is that if a person’s DNA is found at a crime scene, and there is circumstantial evidence supporting the case against them, they are almost as good as convicted. DNA evidence has been accepted as proof of guilt even in cases where the suspect had a strong alibi.
In 2006 a British psychiatrist, Dr Jan Falkowski, was charged with raping a female friend of one of his patients. The main evidence against him was a pair of the alleged victim’s panties, which were found by forensic testing to be stained with semen identified through DNA analysis as belonging to him. Falkowski would almost certainly have been convicted were it not for the fact that, just two days before his trial, dramatic new evidence came to light proving that his accuser was a proven liar who had made similar accusations against other men in the past, all of which turned out to be false. The charges against him were dropped, and further investigation revealed that his accuser – who had been stalking him for several years – had stolen a used condom from a bin outside his girlfriend’s apartment and smeared his semen over her own underwear before going to the police with her accusation.
Psychiatrist Dr Jan Falkowski, who was falsely accused of rape by Maria Marchese (inset).
Falkowski had a lucky escape. But his case illustrates just how easy it is to incriminate someone in a crime if you have a sample of their DNA. In 2006 that meant obtaining a specimen of a body fluid such as saliva or semen. Today, microscopic quantities of DNA are enough to identify and implicate a person, and these can be collected and transferred merely by touching an object that has their DNA on it, and then transferring it by handling an object linked to the crime, such as the murder weapon. A sample of a person’s DNA can be swabbed from a discarded cigarette stub, a cup they drank from, even a pen they wrote with. In short, the evidential weight given to DNA evidence is grossly excessive and disproportional to the ease with which DNA traces can be deliberately or accidentally transferred.
For the police, the temptation to plant DNA evidence may be hard to resist in cases where they are convinced of the guilt of a suspect, but don’t have enough evidence to obtain an indictment or go to trial. There is nothing new about the police planting evidence on suspects, of course. It happens more often than most people think. However, planting DNA to implicate a suspect in a crime takes this unethical practice to a whole new level of seriousness for two reasons. First of all, it’s a lot easier, and a lot less risky, for the police to plant DNA evidence than it is for them to plant other kinds of evidence. All that’s necessary is to deposit a microscopic quantity of a suspect’s DNA at the scene of a crime, or on a murder weapon. Then all they have to do is go back to the scene “to make sure they haven’t missed anything”, and “find” some item that hadn’t been tested for DNA – and lo and behold, this item is found to contain the DNA of their suspect, thus connecting him to the crime.
Second, genetic evidence carries far more weight and is far more damning than most other types of evidence. Most kinds of incriminating evidence merely strengthen the prosecution’s case, but they are rarely as decisive a factor in influencing the outcome of a trial as genetic evidence usually is. Even in cases where the other evidence is weak or purely circumstantial, the introduction of DNA evidence is usually enough to swing a jury in the direction of a guilty verdict. In some cases – for example, where the whereabouts of a defendant at the time a crime was committed is critical – DNA evidence almost guarantees a conviction.
The police are often privy to information that, for one reason or another, cannot be presented in court. In many cases they are 100% sure of their suspect’s guilt, but don’t have enough evidence to guarantee a conviction. The suspect may even have made a full confession which, for legal reasons, cannot be used against him at his trial. Or the police may be aware of previous crimes of a similar nature carried out by him, but be debarred from mentioning these in court. Time and again they see dangerous criminals being set free on some legal technicality, or for lack of evidence, after they’ve spent weeks or months tracking them down, capturing them and building what they hope will be a “bullet-proof” case against them. Then, ten minutes into the trial, the case is dismissed on some technicality, and their chief suspect walks out the front door of the courtroom as free as a bird and grinning from ear to ear. This has to be galling and frustrating for even the most conscientious law enforcement officers, and it is not difficult to understand how, after this has happened a number of times, they might be tempted to “reinforce” an important case with “supplemental evidence” in order to improve their chances of getting a successful conviction. After all, they know – or believe they know – that their suspect is guilty. The evidence they have may be inadmissable in court, for whatever technical reason, but does that mean they should allow a violent rapist, a predatory pedophile or a serial killer walk free? In Britain, falsifying evidence to convict the guilty is known as ‘Noble Cause Corruption’, in recognition of the fact that it is often done for the “right reasons”.
How commonplace is this practice? Reliable figures are hard to come by for obvious reasons, but most police officers will admit privately that “flaking”, as they call it, is quite widespread. In 2011, Stephen Anderson, a former New York Police Department (NYPD) narcotics detective who got caught up in a corruption scandal testified that it was common practice to fabricate drug charges against people in order to meet arrest quotas. Anderson stated under oath that he regularly saw police officers plant drugs on innocent people. “The corruption I observed was something I was seeing a lot of, whether it was from supervisors or undercovers and even investigators,” he said.
Former Detective Stephen Anderson (centre) testified under a cooperation agreement with prosecutors.
Jason Williamson, staff attorney with the ACLU which sued Anderson and four other police officers, commented: “We would be naïve to believe that this is not happening on a pretty widespread basis around the country.”
Anderson and the four officers charged with him were responsible for sending 185 individuals to prison by planting evidence on them. Multiply that by every city and district in the country and you have an awful lot of people locked up on the basis of false evidence.
The police in this case falsified evidence to meet arrest quotas. Flaking is also sometimes done to provide the police with a pretext for detaining a legitimate suspect while his possible involvement in some more serious crime is being investigated. For example, a piece of hashish may be slipped into a suspect’s pocket by the police as they are searching him, which they then “find” and charge him with possession of an illegal substance. This allows them to hold him for longer than they would otherwise be able to do.
If, as appears to be the case, falsifying evidence is widespread among the police, is there any reason to believe that they would not plant their primary suspect’s DNA at the scene of the crime with which he has been charged?
Then there is the problem of cross contamination – the accidental transfer of DNA from one place to another, or from one object to another. This was highlighted in the trial of Amanda Knox, the American student convicted of murdering her roommate, Meredith Kercher, in the cottage they shared in Perugia, Italy in 2007. When examined, a clasp from the victim’s bra strap was found to contain the DNA of Knox’s co-accused, Raffaele Sollecito. However, experts hired by Knox’s defense lawyers blamed this on contamination, arguing that the clasp had lain on the floor for several weeks and that a careless police technician may have stepped on Sollecito’s DNA elsewhere in the flat and then entered the room where the murder took place and stepped on the bra clasp, transferring Sollecito’s DNA to it.
Amanda Knox successfully challenged DNA evidence implicating her in the murder of Meredith Kercher.
Contamination was also offered as the explanation of how the victim’s DNA was found on a knife taken from Sollecito’s apartment, in which she had never set foot. In this case, the suggestion was that the knife had become contaminated in the laboratory by technicians who had been handling other items containing traces of Sollecito’s DNA, or at the crime scene, where investigators had collected key items of evidence without changing gloves. To demonstrate just how easily DNA could be transferred from one item to another, the defense’s scientific team picked up used drink cans wearing clean gloves and then placed a new knife into an evidence bag without changing gloves. The knife was subsequently found to have traceable DNA which had been transferred from the can.
The murder convictions of Knox and Sollecito were ultimately annulled in 2015. Much of the DNA evidence had been discredited, and the case served to illustrate how DNA evidence that had initially appeared to be “strong” and “compelling”, could, upon closer examination, turn out to be weak and worthless. One has to wonder how many convictions obtained on the basis of DNA evidence would never have succeeded if the defendants in those cases had been able to afford high-priced genetic experts to challenge the evidence on their behalf. In the normal course of events, expert challenges of this kind rarely happen.
We have seen that DNA can be deliberately planted by unscrupulous police officers or by criminals, and that it can be accidentally transferred from one location to another, or from one object to another, either when it is being collected, or when it is being tested in the laboratory. And the smaller the quantities measured, the greater the risk of accidental contamination. In the Knox-Sillecito case, the quantity of the victim’s DNA found on the suspected murder weapon was so minute that it could not be divided for two separate tests (a second test is normally carried out to replicate and confirm the results of the first test).
But there is an even bigger problem with DNA evidence. Not only can an individual’s DNA be deliberately or accidentally transferred, but scientists have also demonstrated that it is relatively easy to fabricate, or, rather, replicate, a specific person’s DNA. Anyone with an intermediate knowledge of biology and access to a laboratory can manufacture DNA samples to any specification. In a research paper published in the journal Forensic Science International: Genetics, biological chemist Dr. Dan Frumkin (left) demonstrated how DNA samples from blood and saliva can be fabricated from someone else’s body fluids. In other words, cells taken from one person can be used to create DNA that perfectly matches the DNA of another person. Frumkin also showed that if he had access to a DNA profile in a database, he could manufacture a sample of DNA to match that profile without obtaining any tissue or fluids from that person. “You can just engineer a crime scene,” Frumkin told the New York Times. “Any biology undergraduate could perform this.”
The implications of Frumkin’s work are enormous. The fact that an individual’s DNA can be manufactured in a laboratory completely undermines the whole notion of certainty in criminal cases in which DNA is submitted as evidence. The strength of DNA evidence lies in the astronomical odds against two people having identical DNA. But these odds become meaningless if even one in a thousand DNA samples has been manufactured. Or, for that matter, if one in a hundred samples has been planted by the police.
Again, there is an assumption – a wrong assumption in my view – that criminals wouldn’t be imaginative or resourceful enough to take advantage of these shortcomings in the process of DNA collection and analysis. When Dr. Frumkin’s paper was published, British experts immediately rejected the idea that criminals would go to such lengths. Dr Gill Tully (right) of the Government-funded Forensic Science Service said in a statement to the press: ‘You would need a full molecular biology lab, thousands of pounds worth of equipment and a fully competent molecular biology scientist or technician. The vast majority of people who may be involved in criminality would not have access to these materials.”
Another eminent forensic scientist, Dr John Manlove, an expert witness in court cases, echoed this view: ‘Yes, it is scientifically possible,” he told the BBC, “but it is somebody going to an extreme.”
The naïvety of this view is astonishing to me, and betrays a profoundly misguided notion of the criminal mind and criminal behavior. These experts appear to be unaware of the fact that professional criminals can and do pay other people to do things that they are unable to do themselves. The reality is that there are no materials to which criminals cannot obtain access. For example, criminals involved in ID theft, counterfeiting money, manufacturing illegal drugs and so on have no difficulty getting access to valuable equipment and materials, or to professional expertise. Criminals are nothing if not resourceful. There are no lengths to which they will not go to to commit crimes if the rewards are tempting enough. And the suggestion that criminals would not go to extraordinary lengths to avoid being convicted of a serious crime such as rape or murder is nonsensical.
Consider, for example, the case of John Schneeberger, the South African-born physician who drugged and raped one of his female patients, and who for years was able to evade arrest by planting a fake DNA sample inside his own body, successfully foiling tests to identify him as the rapist.
In October 1992, Schneeberger sedated his 23-year-old patient, Candice Foley, and raped her while she was under the influence of the drug. Despite the fact that the sedative she had been given had a strong amnesiac effect, Foley was still able to remember the rape, and she reported it to the police. Schneeberger flatly denied the charge, and readily agreed to give a blood sample for comparison with the samples of the alleged rapist’s semen that had been taken from the victim’s underwear and pants. The test was negative, and Schneeberger was cleared of suspicion. Foley was ostracized by many people for making a false accusation against one of the small-town community’s most respected citizens.
In 1993, at the victim’s insistence, the test was repeated, but the result was again negative. Foley, still determined to prove that her story was true, hired a private detective to investigate the case. He broke into Schneeberger’s car and obtained another DNA sample, which, this time, matched the semen of the alleged rapist. A third official test was arranged, but the sample taken was deemed to be of too poor quality to be useful for analysis.
That’s where matters rested until 1997, when Schneeberger’s wife found out that her husband had repeatedly drugged and raped her 15-year-old daughter from her first marriage. The police ordered a fourth DNA test, and this time multiple samples were taken: blood, mouth swab, and hair follicle. All three matched the rapist’s semen.
So why hadn’t the previous three tests identified Schneeberger’s DNA as that of the rapist? The bizarre solution to the mystery came to light during his trial, when Schneeberger explained that he had implanted a 15 cm Penrose drain filled with another man’s blood and anticoagulants in his arm. When “his” blood was being taken for the DNA test, he tricked the laboratory technician into taking the sample from the part of his arm in which the tube was located. So the blood that was extracted was actually someone else’s.
The point to note is that this outlandish and daring ruse – which, if you read it in a crime novel you’d probably dismiss as being too far-fetched – actually worked. Not just once, but three times. It was only when Schneeberger’s attacks on his step-daughter came to light five years after he raped Candice Foley, and a fourth, multiple sample test was carried out, that he was finally cornered.
To dismiss any criminal endeavour on the grounds that it would involve “going to an extreme” is nonsensical. What criminal wouldn’t go to extremes to acquire a million dollars, or to avoid spending the rest of his life in prison?
Too much credence is given to DNA evidence, and not enough consideration is given to the possibility that it may have been deliberately planted, either by the police or by a clever criminal trying to put the blame on someone else. Or that the sample may not have been collected properly by crime scene investigators, and may have been contaminated by them. Or even that cross-contamination may have occurred in the forensic laboratory.
And there’s another problem. DNA by itself isn’t evidence of anything. It only becomes evidence when the person it came from denies ever being at the place it was found, or coming into contact with the object on which it was found. In most cases – and especially in the case of historic sex crimes – the value of DNA evidence relies on the police being able to trick their suspect into making a denial before letting him know that there is DNA linking him to the crime. Without that prior denial, most DNA has zero evidential value. It has been estimated that over 90% of cases in which people are convicted on the basis of DNA evidence would never have gone to trial, let alone resulted in a conviction, if the accused person had not made a denial that could be disproved using their own DNA.
For example, a man suspected of rape-murder may be asked whether he had sex with the victim. If he says no, and his DNA has been found in the victim’s body, then he has implicated himself by his denial. If, on the other hand, he admits to having had sex with the victim on the day she died, but claims that it was consensual, and that she was alive and well the last time he saw her, then his DNA is worthless as evidence against him. No matter how unlikely his story that the victim agreed to have sex with him might appear, it can’t be completely ruled out. Since the presence of his DNA in the victim’s body doesn’t contradict his version of events, it does not directly implicate him in her murder. He will still be a strong suspect, of course; but there will be a reasonable doubt as to his guilt, and that’s all he needs to secure an acquittal.
Here’s another example. A woman is found dead in her car in a remote area. Robbery appears to have been the motive. A man was seen getting into her car a few hours before her body was discovered. Eye witnesses are able to give a detailed description of the man, and he is quickly identified and picked up by the police. The car is examined by forensic investigators, and DNA is found which turns out to match that of the man in custody. The suspect is asked whether he knew the dead woman, and whether he was ever in her car. If he denies having ever met the victim, or being in her car, he implicates himself in her murder. If, however, he tells the police that he knew her, and that she had given him a ride in her car on the day she was killed, the evidential value of his DNA drops to almost zero. He becomes the last person to see the victim alive, which would obviously make him a prime suspect, but suspicion isn’t proof, and proof is the only thing that matters in a murder trial. In fact the presence of his DNA in the car substantiates his version of events.
So the evidential value of DNA relies, in many if not most cases, on a rather crude and simple trick. That trick worked in the past because DNA matching was new, and for a long time was perceived by criminals to be only used in high profile cases involving celebrities. These days, however, DNA matching has become more commonplace and routine, and criminals are beginning to realize that the difference between making a rash denial and saying nothing can be the difference between a guilty verdict and an acquittal. Or, more likely, no case to answer at all; because without that denial the prosecution’s case, if it relies heavily on DNA evidence, would probably fail. So the question is: for how much longer will criminals continue to fall for this trick? If a suspect is savvy enough to keep his mouth shut, his DNA is unlikely to convict him, regardless of where it was found. He will have ample time, after he has been charged, to come up with a plausible explanation as to how it got there.
Statistically, violent crimes like rape and murder account for just 1% of all crimes reported, and DNA evidence is presented in only 2% of trials arising from these cases. Furthermore, in most of these cases the DNA evidence isn’t crucial to the outcome of the case – in other words there is enough non-DNA evidence to secure a conviction. In fact the issue in most murder cases is about the degree of intent rather than the identity of the killer – that is, whether the killing was premeditated or spontaneous. It is only in a tiny percentage of cases that the identity of the person responsible is disputed. In reality, therefore, DNA is only instrumental in solving perhaps one in every 1,000 or so serious crimes.
So why all the hype about DNA evidence? Why is DNA matching seen as an effective weapon in the fight against crime when it has all these shortcomings and is only useful in a tiny percentage of the cases in which it is introduced as evidence?
Mainly because it makes for good TV. The types of cases that are solved using DNA matching usually involve sex, violence, murder and an unidentified assailant. And whereas most other kinds of evidence can be interpreted in different ways and argued about endlessly, DNA profiling delivers apparently conclusive and unambiguous results, pointing the finger of blame at a particular person and linking them to the crime, even when several decades have passed since it was committed. What could be more dramatic and televisual than that? The amount of media coverage given to cases in which DNA profiling is used successfully gives the general public the impression that a large number of cases are being solved using this technique, when in fact nothing could be further from the truth.
It has to be acknowledged that DNA matching does indeed represent a major technological breakthrough. And the technique can be spectacularly effective in helping to solve cases, and secure convictions, in crimes such as rape and rape-murder, especially in “historic” crimes of this kind.
The problem isn’t that DNA profiling isn’t effective; the problem is that it is only effective in a tiny percentage of cases. Obviously it’s a good thing that these cases are solved and that rapists and murderers are brought to justice. But DNA matching is a lengthy and expensive process, and it is reasonable to ask whether the money and resources that are put into forensic DNA analysis would not be better used in improving traditional investigative methods, or in crime prevention. There is no doubt that the resources devoted to this field are grossly disproportionate to its overall effectiveness in solving crime.
Despite the fact that genetic technology is improving all the time, the number of cases solved by DNA evidence may actually decrease rather than increase in the future. This is because the very sensitivity of DNA tests is making results more questionable and open to challenge in court. The smaller samples become – and these days a sample can be just two or three cells – the greater the risk of cross contamination, and even atmospheric contamination (ie, airborn particles floating from one place to another). As testing has become more sensitive, the methodology of forensic investigators has come under closer scrutiny, and defense lawyers now routinely scrutinize video footage of the collection, documentation and bagging of evidence to make sure that all the correct protocols were followed. If, for example, investigators fail to change their gloves in between handling items – as happened in the Knox-Sillecito case – any DNA taken from those items will be contested and is liable to be deemed inadmissable as evidence.
Forensic DNA analysis has been most successfully employed in rape and rape-murder crimes that were committed before it became possible to positively identify a person from a tiny sample of their DNA, and these are the cases that have received the most media coverage. It is important to note, however, that the criminals who are now being convicted of historic crimes on the basis of DNA evidence are not being caught because they were stupid or careless. They simply didn’t realize, at the time they offended, that twenty or thirty years down the road it would be possible to identify them from a speck of blood or a drop of perspiration. If they had known, they would probably have been a lot more careful about cleaning up their crime scene.
To summarize, then, DNA can be crucial evidence implicating a specific person in a crime, but only in a tiny percentage of cases, and usually only if the suspect can be tricked into making a false denial. DNA can also be deliberately planted, either by the police or by someone else, and it can be transferred accidentally, either at the scene of a crime or in the laboratory. Finally, “personalised” DNA can be manufactured by anyone with a bachelor’s degree in chemistry.
These shortcomings do not invalidate forensic DNA analysis, but they do mean that its reliability and its value as a crime fighting tool are greatly overrated by the general public and overstated by the media.