A television in Baltimore Circuit Court showed a magnified picture of spent cartridge casings taken through the lens of a microscope. If a juror didn’t look closely, they might confuse two rounds for one.
Half the image showed a casing found on a sidewalk in Southwest Baltimore, where a man and woman were killed early on a frigid morning in November 2019; the other half showed a casing test-fired from a handgun confiscated later that day from one of the defendants.
Tiny, nondescript dents and scratches in the metal, a city firearm examiner testified at the man’s trial in August, were left by the gun that fired those rounds. Look carefully, Christopher Faber told the jury, at “lines, or striations, that are continuously flowing from one cartridge casing to the next.”
What Faber said next goes to the heart of a question before Maryland’s highest court: Should experts like him be allowed to testify that a specific gun fired a specific bullet? Or should they be limited to saying that a gun cannot be ruled out as having fired the projectile in question?
The court is considering a murder case in Prince George’s County that also used firearm “tool mark” analysis as evidence, and the appeal adds to a national groundswell of challenges to the field. There is no timeline for the Maryland high court’s ruling, but its decision could be significant. Such evidence regularly plays a role in shooting cases. Not including suicide, an average of 462 people are killed with firearms every year in Maryland, and many more are shot, according to the nonprofit Everytown for Gun Safety.
Experts say trial courts in several states and federal districts have restricted firearm testimony on a case-by-case basis, and one appellate court has imposed limits on its use.
The forensic science postulates that machines used to make guns leave microscopic imperfections on their components, which, in turn, impart unique marks on the bullets — composed of softer metal — when fired. Firearm examiners — and law enforcement officials who rely on their conclusions to solve and prove cases — maintain the minute scratches on bullets allow them to determine which gun fired them.
Opponents contend the field’s methodology is flawed and hasn’t been subject to testing rigorous enough that the evidence can be relied on when a person’s liberty is at stake.
“The testing that’s been done is inadequate to support the idea that every single gun produces an absolutely unique set of features — that’s the baseline premise — on an item of ammunition,” said Maneka Sinha, an associate professor at the University of Maryland Francis King Carey School of Law who studies forensic sciences.
Even if one assumes the theory is true, Sinha said, “the better question is, ‘Is a firearms examiner capable of discerning that difference?’”
A crime scene technician hunched over to collect cartridge casings and bullet fragments at the scene of the double fatal shooting in Baltimore’s Carrollton Ridge neighborhood about 2:30 a.m. on Nov. 14, 2019.
Hours after the shootings, Baltimore County Police arrested two men and a teenage boy after a brief chase of two stolen cars. One man had a 9 mm Taurus. The three were charged in a series of armed robberies and carjackings.
Weeks later, a law enforcement ballistics database alerted investigators about a potential match between the casings from the city homicide scene and those test-fired from the gun in the county.
Most firearm examiners work in police department crime labs. Faber was assigned at the Baltimore Police Department’s Forensic Laboratory Section to examine the test-fired rounds, or the known sample, and the casings from the scene, the unknown, under a comparison microscope. Such devices allow examiners to look at two pieces of ammunition simultaneously.
A casing encapsulates a bullet. After a trigger is pulled, a firing pin rams the back of the casing, which ignites an explosive substance inside. That propels the bullet down a barrel, which features twisted metal known as “rifling” inside, to spin the projectile for accuracy. A semi-automatic handgun ejects a casing after each shot.
The firing pin is one of as many as 10 or more pieces of a gun that leaves marks on casings, said Michael Haag, a private forensic consultant, firearm examiner and instructor who worked for the crime lab in Albuquerque, New Mexico, for 25 years.
“My protocol was to fire at least three test-fire bullets, three casings, intercompare those, so I’m learning that gun. I’m seeing what marks are or are not reproducing from that particular known specimen,” Haag said. “Then comparing those to unknowns and determining whether or not, through the examination process, I can come to a conclusion, based on observations, of common source: Did this tool create these tool marks?”
The Association of Firearm and Tool Mark Examiners says examiners can declare a match when an examiner sees “sufficient agreement” between two projectiles. The organization, which establishes standards for the field, says agreement between two rounds is significant when a set of markings is of “a quantity and quality that the likelihood another tool could have made the mark is so remote as to be considered a practical impossibility.”
“Their entire theory of identification essentially boils down to, ‘I know it when I see it,’” said Jeffrey Gilleran, chief of the forensics division at the Maryland Office of the Public Defender.
“There’s no objective benchmarks or requirements placed on these examiners,” Gilleran said. “There are no statistics, like in DNA analysis, about how rare or common any set of features are. The standard is instead completely circular and fully subjective.”
In 2008, the National Academy of Sciences’ National Research Council said the underlying theory of tool mark identification had “not yet been fully demonstrated.” More than a decade later, with subsequent critiques from the academy and the President’s Council of Advisors on Science and Technology, the scientific community maintains that view.
There is little dispute that firearm examiners can look at a casing or bullet and narrow down the range of guns to the type of firearm that may have shot it. That’s because gun makers intentionally manufacture certain components of their firearms in unique ways. Characteristics such as caliber, which refers to the diameter of the inside of a gun barrel, are objective.
“All you need is a microscope and a ruler,” said Sinha, who successfully argued for one of the first limits placed on firearm analysis testimony in the Superior Court of the District of Columbia — a groundbreaking case cited in the Maryland appeal.
A dozen scientists and researchers from institutions ranging from the Johns Hopkins University to the University of Nebraska-Lincoln wrote a brief supporting the defendant appealing the Prince George’s case. They said the theory that each gun is unique is unproven.
“While there are encouraging developments in research design, data from a recent study shows an alarming lack of consistency in decisions when the same examiner was presented with the same evidence twice, and when different examiners were presented with the same evidence,” the scientists wrote. “These new data further undermine the claim of a well-developed, scientifically valid method and cannot go unaddressed.”
Haag said the firearm examiner community was “a little too defensive” in 2008 when the National Academy of Sciences critique came out, rather than “taking it to heart.” He has participated in various studies since and said the attitude has changed over time. The field, he said, also has made improvements, like requiring examiners to document their work with pictures and notes.
Still, Haag said, “we’ve gotten stagnant as far as addressing the idea of how often we can be wrong.”
How often examiners arrive at the wrong conclusion is at issue in the Maryland appellate case, as an established error rate is one of the criteria for permitting expert testimony as evidence in the state. In oral arguments this fall, lawyers on the two sides of the case cited error rates ranging from zero to 50%. The discrepancies in those rates arise from differing opinions about how to handle inconclusive conclusions in studies.
There are five conclusions an examiner can reach: elimination of a bullet as having been fired from a specific gun, identification of a projectile being fired by that gun and three different types of “inconclusive” — one that leans toward excluding a match, another that leans toward a match and one that represents greater uncertainty.
J. Bradford McCullough, the lawyer appealing the case, told the court that the studies were like taking a multiple choice test where four of five answers are correct. If a bullet being studied was a match for a gun, the only wrong answer would be exclusion. So if the tester had any doubt, McCullough said, they’re likely to choose inconclusive — reducing the error rate.
Representing the state, Assistant Attorney General Andrew J. DiMiceli said studies have shown it’s very rare for examiners to mistakenly conclude a bullet was fired from a certain gun, which he argued was the most important thing to a criminal defendant because such a finding would suggest guilt.
McCullough countered with a study that showed examiners correctly eliminated a gun as a firing source only a third of the time.
“That’s what’s really important to a criminal defendant, by the way. If it isn’t really a match, can you determine that? Or do you call it something else other than elimination, including an inconclusive?”
A defense lawyer may argue an inconclusive finding casts doubt on the state’s theory of the case, while the prosecutor maintains it’s evidence of guilt.
For Democratic State’s Attorney Rich Gibson of Howard County, who is president of the Maryland State’s Attorneys’ Association, that’s an example of how our court system provides opportunity for a defendant to challenge testimony.
Prosecutors disclose an examiner’s case file to the defense before trial, so an attorney may question the expert during their testimony about any flaws in that record or hire an expert to review the evidence and present a different opinion.
Also, a defense attorney has “the ability to put the pictures in front of the jury and say, ‘The examiner says there are three lines there, but I see two. How many do you see?’” Gibson said.
Also, Maryland judges tell jurors they can choose to believe all, some or none of an expert’s testimony.
However, a 2020 study, conducted by researchers at Duke University School of Law’s Center for Science and Justice and the University of California Irvine School of Social Ecology, found jurors in a mock trial gave “great weight” to firearm testimony. Cross-examination casting doubt on such testimony did little to affect verdicts.
“Because of the label of ‘science’ and the aura of legitimacy and reliability that comes with it, jurors and judges tend to believe forensic experts, even when it’s incorrect and overstated,” Gilleran said.
In the double murder trial in August in Baltimore, no one testified from the group of people who scattered when gunfire broke out. None of the people robbed, assaulted or carjacked by the men charged with the killings could identify their masked assailants. After the ballistics, grainy surveillance video was the next-best evidence.
When the firearm examiner testified, he pointed to places on the split-screen photograph where scratches on one cartridge casing appeared to continue onto the next. The prosecutor asked what conclusion the examiner reached.
“The fired cartridge cases, the 9 mm Luger cartridge cases, were fired by the Taurus firearm,” Faber testified.
He was one of the last witnesses to take the stand. Three days later, the jury convicted the man caught with the gun of murder.
A judge later sentenced him to two life terms, plus 200 years in prison.