Verified by 
“There are ways of telling whether she is a witch… What do you do with witches?”
“Burn ’em! Burn them up!”
“And what do you burn, apart from witches?” …
“Wood!”
“Good. Now, why do witches burn?”
“Because, they’re made of wood?”
“Good. So, how do we tell if she is made of wood? … Does wood sink in water?”
“No, no, it floats!” – Monty Python and the Holy Grail
In the wake of Dobbs, I published a three-part series on how the decision would necessarily influence the legal system. Today, I want to write about a case that was unquestionably swayed by premises defended in Dobbs: that the State’s “legitimate interest” in protecting “potential life” is sufficient to limit the rights of women.
We will review yet another no-crime case in which a tragedy involving a child had to ensnare a woman as well. I want to begin by crediting ProPublica with conducting a thorough analysis of this case (on which I relied regularly for this entry).
At 37 years old, after two difficult pregnancies (both ending in cesarean sections), Moira Akers’ pregnancy would have been considered high-risk. When she became unintentionally pregnant, both Akers and her husband decided they would quietly seek an abortion.
After Akers mistakenly believed it was too late to seek an abortion legally, she decided she would secretly carry the baby to term and give it up at a fire station (under safe-haven laws, many firehouses have baby boxes in which infants can be legally and safely abandoned). On November 1, 2018, she gave birth to her unbreathing son: “He did not move. He did not make a sound”.
Akers, at an understandable loss, wrapped her son in a towel, put the towel in a Ziploc bag, and put the bag in the closet. Upon discovering Akers in pools of blood, her husband dialed emergency services. She initially denied being pregnant, but after being transported to the hospital she informed a nurse that she had given birth to a stillborn infant.
Paramedics returned to her home and discovered the body, left where Akers had described. An investigation ensued, and an autopsy was performed. Two Maryland doctors completed an analysis reporting that the neonate had been delivered at 41-42 weeks of pregnancy (an additional risk factor for complications) and appeared well-nourished.
Their analysis also included a hydrostatic (lung) float test (HFT). This test involves removing the lungs (or a portion of the lungs) of a neonate and observing whether they float in water (or liquid formaldehyde). The state pathologists argued that the results of the HFT were appropriate in this case given that “there was no evidence of air being introduced by means other than breathing…”
This test has existed (largely in its present form) for centuries. And, there is broad consensus that the test cannot be used to conclude whether a child was born alive (even by the doctors who performed it). In several studies, the test has produced false positives (i.e., the lungs have floated despite belonging to a stillborn baby).
Babies do not have to take a breath for air to be introduced into their lungs. Air can be introduced in many other ways. For instance, in this case, the baby’s body was placed in a Ziploc bag by a bereaved mother (not a medical professional)—it seems possible that any pressure applied to a sealed bag (with air) could have introduced oxygen into the baby’s body.
Of course, this is an empirical question. But it's one that should have been answered before presuming that, under the circumstances of this case, the HFT was a sufficiently reliable test.
Dr. Ranit Mishori, a professor of family medicine at the Georgetown University School of Medicine, has been clear about issues underlying the HFT (via ProPublica): “Basing something so enormous on a test that should not be used, that has been completely discredited is absolutely wrong… You can send a person who is innocent to prison for many years.” She is also a senior medical advisor for Physicians for Human Rights, who have been studying the test and deemed it, “inaccurate, outdated, and unethical.”
In my own searches, I found one study that deemed the test largely reliable—a 2013 study in which the authors reported 0 false positives across 208 tests. An extraordinary result that warrants extraordinary replications (which have either not been conducted or have not been published). Yet, it continues to be cited in support of the HFT (including in Akers’ case) despite numerous other studies and reports citing serious concerns regarding its validity.
The state pathologists in this case (and many other pathologists) argue that the HFT (despite its issues) can be used in “concordance across a variety of tests.” Of course, many others argue that an unreliable test is an unreliable test, and observing allegedly concordant evidence does not confer reliability on its results. A test that produces seemingly clear conclusions can be problematic when combined with more ambiguous evidence and natural human biases (e.g., forensic confirmation bias).
During the trial, prosecutors displayed an oversized photo of the baby’s body in a bag. They said the word “bag” over 20 times in closing arguments alone. The prosecutor also stated, “These lungs floated. They floated because this child had breathed and was alive after he was delivered at home that day.” These statements belie a confidence that is entirely unjustified by analyses of the HFT.
In April 2022, the jury found Akers guilty of second-degree murder and first-degree child abuse.
After Akers was convicted, the prosecutor stated, “This was truly one of the most difficult cases my office has ever had to prosecute because we had the burden of proving the baby was born alive…” Perhaps if that burden was felt so heavily, it is because it could not be proven definitively. Perhaps, even further, Akers’ baby was not born alive as she consistently attested. Perhaps the tragedy of a nonliving infant being met with suspicion rather than sympathy is a recipe for disaster.
At her sentencing hearing, Moira Akers said, “My children are my entire world, and I fell in love with my son as soon as I saw him”. Akers mourned a son she wouldn’t even have the opportunity to bury. The judge sentenced her to 30 years in prison.
In my next entry, I’ll dig further into issues with forensic confirmation bias, and how it played a role in Moira Akers’ conviction.
The initial version of this post mistakenly described Dr. Mishori as a "he" rather than a "she". I want to thank Dr. Mishori for reaching out with this correction (which is reflected in the current version).
