It is an encouraging sign that the Department of Justice is finally starting to grasp the importance of sound science in the courtroom. On June 3rd, the department released its first set of rules regarding courtroom testimony that pertains to forensic science – particularly its reliability and the conclusions that can be drawn from it. It may come as a surprise to many that such rules, which have yet to be finalized, do not already exist. Some may feel that such regulations are unnecessary: our impression of forensic scientists is that they are always honest and upfront about what the data tells them.
Unfortunately, not all forensic experts have the personality and infallibility of CSI’s Gary Sinise. A recent analysis by the FBI which examined 268 criminal cases from 1985 to 1999 found that over 95 percent of those cases involved flawed testimony relating to forensic science and its implications of guilt or innocence. This is a shockingly high fraction, and in at least a dozen cases, the convicted defendants were either executed or have died in prison.
Part of the problem is that scientists who testify for the prosecution often overstate the accuracy of their analyses, for example, claims that two fingerprints are a conclusive match when in fact a much high margin of error exists. Understanding that forensic science is often inexact is critical because in many cases only a single piece of biological material can link a defendant to a crime scene or victim. Far too often, wrongful convictions result from slipshod science advertised as unerringly precise by expert witnesses testifying in front of jurors who do not understand the significance of error rates in scientific procedures such as follicle comparison or serology. And, that is to say, nothing of forensic investigators who become biased in their subsequent interpretation of evidence after being told who the prime suspect is a priori. For example, a CSI technician might be more likely to report a match between a hair sample from the crime scene and another sample marked ‘suspect’.
In effect, the new rules are disclosure rules; testimony pertaining to forensic procedures or findings must include a disclosure detailing the probability of a conclusion (e.g. a DNA match) within an “acceptable range of opinion”. The rules also detail what type of statements forensic experts may or may not use while testifying in court. This is typically sound practice – just like reporting the margin of error with the results of a survey or poll.
Though the disclosure rules are a necessary first step toward establishing proper scientific standards in criminal prosecutions, it is hardly the only step that must be taken.
To understand what is actually needed in order to make investigations and prosecutions more evidence-based, consider what the role of a scientific expert ought to be. Is the role of an expert to stay true to the facts and the appropriate method of deductive analysis? Or is it to assist lawyers to manipulate juries to reach a desired verdict? Science, by definition, means engaging in rigorous analytical procedures and following the logical conclusions of that analysis no matter where they lead. If this ideal is what forensic scientists adhere to, then it’s not clear how or why there are often two sets of expert witnesses at odds with each other during a trial, each testifying for different sides. After all, how can two scientists who engage in suitable analysis reach opposite conclusions from the same set of facts? The answer is obvious: forensic scientists, whether they testify for the prosecution or the defense, are closer to paid spokespeople advocating a particular Lego theory rather than objective investigators. It’s sort of like a pharmaceutical conglomerate having an in-house “expert” testify about the safety of a new drug it has developed.
If the forensic sciences were systematic and rigorous as other established sciences, there would exist a set of agreed-upon principles and standards that would allow pieces of evidence and facts to converge toward a singular conclusion. Either the blood spatter is consistent with the action of stabbing – or it isn’t. Either there is a 50% chance the suspect’s hair matches the hair follicle found at the crime scene – or there isn’t. Science exists for the sole purpose of eliminating personal opinions and hunches from the process of truth finding, yet in the criminal justice system, forensic science is often misused for the purpose of idea-mongering.
All this suggests that if forensic science is to be appropriately utilized for the task of establishing truths in the courtroom, any and all evidence collection, analysis, and testimony should be conducted by an independent third party group of specialists who have no personal or professional relationship with the police, prosecution, or defendant. This third party will adhere to peer-reviewed and accepted forensic scientific protocols will be accountable to an independent office not subsumed by any law enforcement body, and will have their labs in a completely separate location inaccessible by the police and prosecution. This would prevent test tampering or unexplained destruction/loss of evidentiary material. It would have the added benefit of preventing the prosecution from hiding evidence from the defense as it would be under the control of a group of scientists who have no incentive to support either side in a criminal proceeding. Experts from this third party would testify in court between the prosecution’s and defense’s case-in-chief and both sides would be permitted to cross-examine them.
Peter Neufeld, a founder of the Innocence Project, stated that the DOJ ought to have the new disclosure rules evaluated by outside scientists instead of relying on in-house specialists. He noted that outside scientists would have “significant and substantial concerns” after reviewing the proposed framework.
Still, there are signs that the federal government is starting to take the precision (or lack thereof) of forensic science seriously. The Justice Department has offered new hair sample testing to hundreds of convicted defendants whose trials were adversely affected by faulty analysis and testimony and a few of them have already been exonerated. The department is also checking to see if errors have been made with regard to other types of evidence. These efforts follow a 2009 Research council study that found problems in the FBI’s courtroom testimony involving various types of forensic evidence and noted specifically that there was “no scientific support” for positively identifying a suspect from a hair sample.
The DOJ plans to release a second set of proposed disclosure rules involving DNA, explosive devices, hair analysis, and handwriting. By emphasizing proper forensic science, the department has the potential to completely change how criminal law is practiced in this country- and all for the better.