What if the stranger, now suspect, is a homeless alcoholic? With a rap sheet longer than a giraffe’s neck? Ask twelve other people: parents, friends, or any potential juror, those same questions.
By: Ann Light
March 27, 2023
This blog addresses the unreliability of DNA evidence, including the stark differences between source probability and random match probability (RMP). Also explained is why U.S. courts should ensure jurors understand these differences and how the United Kingdom has attempted to address some of these issues.
Prosecutors, defense attorneys, judges, and jurors in U.S. courts may not understand the differences, which may or may not be accurately explained by expert witness testimony. The FBI often dumbs down the statistical analysis for a jury by using calculations based on a U.S. population 100 times its true size. Consequently, some juries are left confused, misled, or oblivious by no fault of their own. Even worse, an expert witness could persuade a jury to believe any DNA match is a “gotcha,” “bingo,” or “bullseye.” After all, society’s general belief is that DNA matches are essentially accurate.
Before diving into the statistical analysis, unknown DNA must first be collected. Samples taken from crime scenes are far from perfect. They are often contaminated by other elements. Samples could be old and deteriorated, especially from weather exposure. Moreover, several people’s DNA could be included in one sample, making it difficult to determine what DNA belongs to what person. However difficult the determination may be, a lab analyst’s subjective conclusions are admitted into evidence in a criminal trial. Some labs move forward with the analysis even when a sample contains a small amount of DNA. Regardless of the sample size, any lab or analyst could simply make a mistake. One final imperfection: DNA is everywhere! (Hold onto that idea for now.)
As of November 2022, the FBI’s National DNA Index System (NDIS) contains almost sixteen million convicted offenders, detainees, and legal profiles. A “cold hit” occurs when DNA found at the scene matches DNA saved in the FBI’s database. Viola! The government can then use the match and other basic “evidence,” such as race or age, to narrow the field of suspects down to one, the defendant.
The United Kingdom maintains the world’s largest DNA database. “Cold hits” are certainly found there and in other databases worldwide. Legislation in each European country can limit the size and categories of people contained in a database. For example, Sweden’s databaseincludes only convicted offenders and, depending on the sentence length, remains in the database for a set period. Suspects are removed upon acquittal in Sweden. Latvia, by contrast, keeps convicted offenders in its database for seventy-five years. Even after a suspect is acquitted, the profile remains in Latvia’s database for ten years and is susceptible to a cold hit. On the other hand, Scotland destroys samples if a suspect was not charged or convicted. England, Wales, and Northern Ireland kept even a suspect’s DNA indefinitely until the United Kingdom’s Protection of Freedoms Act 2012 was enforced. Now, acquitted suspects will be dropped from the database after three years if they maintain a clean criminal record and are no longer a suspect in the case.
Despite the differences in database collection and retention, analysis begins once a match is found. However, there are several ways labs slice and dice DNA. Source probability is the probability that the defendant whose DNA matches the DNA found at the crime scene was the trustworthy source of that evidence. Conversely, random match probability (RMP), or profile frequency, is the probability that the DNA of a randomly selected person from a relevant population will match the DNA found at the scene. The denominators used in RMPs could be larger than a quintillion, astounding statistics for any jury’s ear.
One reason the court in United States v. Davis found the DNA conclusions unreliable because RMP matches can occur by mere coincidence. Matching profiles between two unrelated people in Arizona and Maryland were found using RMP. The Prince George’s County database in Maryland contained less than 10,000 profiles. Two unrelated individuals matched seven out of thirteen loci within that small batch. The result: RMP of 1 in 1 trillion. The world’s population is only around 8 billion, far from a trillion.
Still confused?
The U.S. The Supreme Court took a stab at it in McDaniel v. Brown, 558 U.S. 120 (2010) by giving the following explanation:
[I]f a juror is told the probability a member of the general population would share the same DNA is 1 in 10,000 (random match probability), and he takes that to mean there is only a 1 in 10,000 chance that someone other than the defendant is the source of the DNA found at the crime scene (source probability), then he has succumbed to the prosecutor’s fallacy. It is a further error to equate source probability with probability of guilt, unless there is no explanation other than guilt for a person to be the source of crime-scene DNA. This faulty reasoning may result in an erroneous statement that, based on a random match probability of 1 in 10,000, there is a .01% chance the defendant is innocent or a 99.99% chance the defendant is guilty.
If the problems are not glaringly obvious, here is another: the problem of deference that U.S. courts rely on precedent. If previous courts have found RMP reliable, then it is. Moreover, even if courts wanted to investigate older DNA cases, the task seems complicated because it is.
The United Kingdom’s justice system analyzes DNA evidence in any criminal case. Although, it’s through independent educational platforms that have attempted to address DNA issues. The Code for Crown Prosecutors was created to ensure “[…] decisions are fair, transparent, and consistent.” Its guidance provides in-depth explanations of various methods used for DNA profiling, including understanding the match probabilities of each. The Royal Society and the Royal Society of Edinburgh published Forensic DNA analysis: a primer for courts. It aids the judiciary in understanding the science of DNA evidence. Although a step in the right direction, outside guidance and suggestions, are just that. Courtrooms worldwide need to do better.
Remember the homeless alcoholic’s DNA that was found under the fingernails of a suffocated murder victim? He has a name: Lukis Anderson. In 2012, Lukis was charged with the murder of a sixty-six-year-old man. Paramedics on the scene declared Raveesh Kumra dead after responding to the 911 call made immediately by his surviving partner after the attack. Weeks later, Lukis’ DNA was found under the victim’s fingernails. Truthfully, his DNA was there. After spending months in jail, another truth arose. At the time of the murder, Lukis was in the hospital, struggling to stay conscious after consuming the equivalent of twenty-one beers. He could not have committed the murder. Instead, Lukis’ DNA had been transferred through the paramedics. Gone unsolved, Lukis could have been sentenced to death. Read the story here. Then let’s get busy.
