Skip to content


In most areas of our lives, we accept that human bias can influence how we interpret the world. We put checks in place to make sure our personal views do not influence important tasks. We make sure that standardized tests that students take are graded blind. We would never allow drug trial administrators to tell people they received the placebos. We make sure the judge and jurors in a criminal trial do not have prior personal relationships with the defendant.

Yet in crime labs, where issues of life and liberty are also at play, we largely rely on a system – from evidence collection to forensic analysis – that allows human influences to roam free, unchecked. The National Academy of Sciences report warned that a “lack of independence” in crime labs damages the objectivity of forensics. Yet very few labs in the United States are independent.  The first crime labs in the United States were created by police agencies and housed in police departments.  That is mostly still the case.

Read more about various examples of this and then return to the Forensics Reform Toolkit to learn more about how you can help change this.

The George Floyd Autopsy

What would have happened if there had been no video of Derek Chauvin killing George Floyd on May 25, 2020, with his knee on Floyd’s neck for over nine minutes?

Videos made by witnesses and security cameras vividly showed how three officers helped Chauvin restrain Floyd and kept bystanders from intervening. However, before those videos surfaced, the Minneapolis Police Department issued a statement…

Police use forensics in all sorts of cases, from driving while intoxicated cases, drug possession cases, to homicide investigations and police shootings.  From the first moments that an investigation has started by a detective or police officer, a crime scene investigator, or CSI, may get involved. The CSI “is in place to assist with the investigation by finding, documenting, collecting and possible analyzing evidence that may place a suspect of victim at a scene.”   Many people think that such crime scene investigators, and then their colleagues in a crime lab, do highly scientific and high-tech work.  Many have discussed “the CSI Effect” of television shows such as CSI: Crime Scene Investigation and Law & Order. Far from expecting that every case involves DNA or science, instead, most laypeople believe that forensic evidence is 100% accurate.  People do not realize how readily forensics can go wrong, including because most of the CSI and lab professionals are actually acting as law enforcement, and not independent scientists.

There is greater understanding today that policing can be biased, including racially biased, in ways that disproportionately burden minorities with criminal legal involvement.  Forensic science plays an under-appreciated role in that system.  In forensic science, testing is rarely blind.  Traditionally, the forensic examiner talks to the police, and hear all sorts of biasing information about the suspect, like the person’s race and criminal record.  Forensic experts are also biased by the side that they work for.  Experts who see themselves as part of the police and prosecution team predictably shade their results to help the prosecution.

This has been a high-profile problem in investigations of police use of deadly force.  There is evidence that large numbers of deaths — more than half of police killings in the United States — have been mislabeled and not counted as police violence-related deaths, due to medical examiner’s conclusions in death certificates.  As the New York Times noted, “forensic pathologists regularly consult with detectives and prosecutors and in some jurisdictions they are directly employed by police agencies.”  For example, the coroner called Ronald Greene’s death in Louisiana as an accidental and the result of a heart attack, before a video was released showing Greene stunned, beaten and dragged by state troopers.

We need independent scientific crime labs and not police crime labs.  The National Academy of Sciences report warned that a “lack of independence” in crime labs can damage the objectivity of forensic science. Yet very few labs in the United States are independent.  The first crime labs in the United States were created by police agencies and housed in police departments.  That is mostly still the case.  A survey of about 300 labs conducted in the 1980s, found that 79% were located within law enforcement or public safety agencies and most would only examine evidence submitted by law enforcement officials.  As one leading scholar has explained, “the police agency controls the formal and informal system of rewards and sanctions for the laboratory examiners.”

Today, there are over four hundred publicly funded crime labs. The budgets of publicly funded labs have certainly grown. In 2014 they totaled $1.7 billion. There are more labs that are independent, but not many more.  In a handful of states, there is some amount of independence to a lab’s budget and organizational structure.  They still largely conduct testing requested by law enforcement, and return results back to law enforcement.  Their budgets are still largely law enforcement-directed.  Indeed, in some states, any criminal defendant is charged a fixed crime lab fee. We should not be funding forensic evidence with an incentive to convict and draw funds from people charged with crimes. 

Biased Forensics

It is not just the budgets and structure, but also the work that forensic examiners do that is linked to police.  Evidence collection is often handled by police.  They decide what to collect and what to send for forensic testing. Even the forms that police use to submit evidence to a crime lab can include all sorts of biasing information, like a suspected person’s race and criminal record.  

Independence and scientific accountability are deeply needed.  Several high-profile research studies also highlight the importance of this problem. Fortunately, in addition to lab independence, a range of reforms can ensure that the work that forensic examiners do is independent.  They need to be de-biased.  The National Academy of Sciences’ 1996 DNA Report observed that “laboratory procedures should be designed with safeguards to detect bias.” Forensic examiners should not receive biasing information from police.  They should only receive information relevant to their scientific task.

Autopsy of a Crime Lab: Interview with Duke Law Professor Brandon Garrett

Bias can also be baked into technologies that police and crime labs use. Willie Allen Lynch, convicted of drug charges, is serving an eight-year sentence largely based on a blurry cellphone photo. Undercover detectives in Jacksonville, Florida, conducted a $50 crack purchase, and while they did not make an arrest at the time, one took a photo of the seller.  The local police in Jacksonville accessed the FBI’s facial recognition database, the Facial Analysis, Comparison, and Evaluation (FACE) service, with access to 641 million photos (almost double the U. S. population). Just eight days before trial, Lynch’s lawyers learned he was identified using such facial recognition. Apparently, four other faces were also identified. His photo was given a “one star” rating. Lynch’s lawyers were not informed what that meant, or what the other four faces looked like. Even at trial, the judge did not allow the defense any access to forensics database or to complete results.

 Lynch’s lawyers argued that Lynch, who is Black, may have been more likely to have been wrongly identified due to his race.  Several facial recognition databases have been shown to make more errors when searching black faces. And this software had never been tested. Plus, there were specific reasons for concern in Lynch’s case. A lab analyst showed a detective who had observed the drug transaction a single photo of Lynch, asking if he was the one. That was very suggestive: they did not use a proper police lineup, with a group of photos. The analyst also told the detectives that she knew Lynch’s criminal history, which included prior convictions for drug sales.

The appeals judges explained that the prosecutors were not required to turn the evidence over to Lynch. In 2019 the Florida Supreme Court declined to hear the case, so Lynch is still serving an eight-year sentence. You may be in that FACE database.  Half of all adults in the United States had their faces included by 2016, and many more have their faces added each year. The Georgetown Law School Center on Privacy and Technology, the first to report this, aptly calls it a “perpetual lineup.” Law enforcement should not be allowed to use these algorithms until we know how accurate they are and whether they are racially biased.


Dept. of Forensic Science – Washington, D.C.

The Department of Forensic Science in Washington, DC was supposed to be a model lab that would change lab work on crime scene evidence for the better.  The DC government passed a “Department of Forensic Sciences Act of 2011,” to make the lab more independent of police.  They paid over $210 million for a large new building.  After years of concern that crime scene evidence was not being professionally tested, it seemed like everything was finally going right…