Joyce Daubert and her son Jason in front of Joyce's home in Ramona, California.

Where Science Enters the Courtroom, the Daubert Name Looms Large

Joyce Daubert’s baby weighed six pounds, one-and-three-quarter ounces — but weight was not what set Baby Boy Daubert apart. Nor was it the timing of his birth on July 26, 1973, almost two weeks past her due date. What shocked everyone in the delivery room at Palomar Memorial Hospital, in Escondido, California, were the newborn’s tiny limbs. Joyce, who was 26 at the time, first saw what looked like a vestigial big toe, and she remembered saying: “What’s wrong with his foot?” Her son appeared to have only two fingers on his right hand, and his right arm was rigid, angled up towards his face, almost, she thought, like a chicken wing. “I’m trying to figure out what’s happening and I’m just shocked,” she recalled. “And, you think, ‘Oh my God, what did I do wrong? Is he going to live?’”

In the hours and days after the boy’s birth, doctors swarmed over him. Pediatric specialists performed diagnostic tests. They took X-rays. The newborn’s heart and lungs looked fine. Still, Joyce said, the doctors deemed his anomalies “major.” The boy, whom Joyce and her husband Bill had named Jason, was missing a bone in his forearm and three fingers on his right hand. He had a hernia, an undescended testicle, and a deformed big toe. Hospital staff took him into Joyce’s room swaddled up like a cigar. She nursed, but they advised her not to unwrap his blanket. “I felt responsible,” Joyce said. “Because I carried that baby. It was horrible. I just didn’t know if I had what it would take to raise a deformed child.” For a brief time, Joyce said, she contemplated suicide. But ultimately, she committed herself to giving Jason the best parenting anyone could. “I quickly gathered myself and moved to that perspective,” Joyce said, “and operated on that ever since.”

“Because I carried that baby. It was horrible. I just didn’t know if I had what it would take to raise a deformed child.”

A decade later, Joyce — now the mother to both Jason and a healthy girl, Jessica — came home and picked up the paper from the driveway. By that time, Joyce had gone back to teaching, and Jason had undergone multiple surgical procedures. Although genetic tests would later confirm that he had “apparently normal” chromosomes, the exact cause of his birth defects remained a mystery. Experts told the family the most likely explanation was an exposure to a “teratogen” — a medical term that literally means “monster-producing.” It describes any factor that causes malformations to a developing embryo.

Inside the house that day, Joyce made a snack, and plopped down with her two children. As she remembers it, she flipped open the newspaper, and, in the upper left-hand corner, they saw a photograph of a little girl whose arm looked just like Jason’s. Jason ran to get a pair of scissors and Joyce clipped out the photo. The girl’s name was Mary Oxendine.

In 1983, the Oxendines were part of an early wave of plaintiffs who filed suits against Merrell Dow, a pharmaceutical company that was then a subsidiary of Dow Chemical Company. The family went to trial at a local District of Columbia courthouse and sought financial recovery for damages allegedly caused by a drug called Bendectin. A combination of vitamin B6 (pyridoxine hydrochloride), an antispasmodic (dicyclomine hydrochloride), and an antihistamine (doxylamine succinate), Bendectin was first approved by the U.S. Food and Drug Administration in 1956 for the treatment of nausea and vomiting during pregnancy. It was then the only anti-nausea medication available for pregnant women to treat what is commonly referred to as morning sickness. Her own doctors had never pinpointed the cause of Jason’s birth defects, but Joyce remembered that she, too, had been prescribed Bendectin for nausea in her first trimester. She had taken the drug just three times, she recalled, but in that moment, she felt she’d finally identified the teratogen that harmed Jason.

That night, Joyce squirreled away Oxendine’s photo. “I had dinner to make. We had homework to do,” she said. But early the next morning, she called directory assistance and eventually talked on the phone with the girl’s father in Maryland. On a page from Joyce’s calendar book dated June 2, 1983, she wrote the name of a Washington trial attorney who represented the Oxendines: Barry Nace. (By that time, Bendectin had been reformulated as a two-drug combination without the antispasmodic.)

Nace says he had been approached by the girl’s parents, who were impressed by his background in chemistry. “That’s how I took the case, having no idea, no clue what it was going to blow up into,” Nace recalled in a recent interview. “Because in those days, the scientific evidence crap wasn’t like it is today. You tried your case. You put your evidence forward, which is what I did.” During the Oxendines’ trial, Nace called two experts: Shanna Swan, a reproductive epidemiologist who worked with the California Department of Health Services, offered a statistical re-analysis of existing scientific studies. Separately, in three and half days of testimony, Alan K. Done, a former pediatrician from Wayne State University, explained to a jury that, when considered individually, the existing epidemiological data, chemical structure analyses, animal toxicology studies, and animal experiments may be insufficient to show the teratogenicity of Bendectin, but the combination of all four factors proved their case.

In May of 1983, a jury awarded the Oxendines $750,000 — the first major Bendectin settlement in the more than 750 lawsuits that would eventually be brought against Merrell Dow. Thirteen days after that, the company pulled the drug off the market — a move that, according to Michael D. Green, author of the 1996 book “Bendectin and Birth Defects,” convinced many people that “there were hidden teratologic skeletons in Merrell’s closet.”

Still, throughout the 1980s, Merrell stood by Bendectin’s safety, and its own expert witnesses maintained the drug did not cause birth defects. Following the Oxendine verdict, the company issued a statement to The Los Angeles Times saying the verdict was “not consistent with the weight of scientific evidence.”

The Dauberts, meanwhile, assembled a team of experts that offered a different interpretation, and the family would eventually fight the drugmaker all the way to the Supreme Court in 1993. In the process, their name became synonymous with a legal standard that now defines how federal judges admit scientific evidence in courtrooms across the U.S. The “Daubert standard” asks judges to perform a kind of quality control, ensuring that jurors hear from experts who apply the same intellectual rigor that is demanded in their professional work, and that their methods are grounded in empirical research.

“That’s how I took the case, having no idea, no clue what it was going to blow up into. Because in those days, the scientific evidence crap wasn’t like it is today.”

Today, “Daubert” is no longer just the name of a family or the civil suit they brought against Merrell Dow. It is among the most cited Supreme Court rulings, and the name has become a verb. To Daubert an expert is to vet them; it is to scrutinize their expertise. Daubert hearings continue to shape headline-making lawsuits involving the herbicide Roundup and Johnson & Johnson’s baby powder, and the Daubert standard features prominently in discussions around criminal justice reform.

The Dauberts might have joined an elite group of plaintiffs whose names have become ubiquitous in the legal system, but in the sweep of history, the lives of the people whose personal tragedy gave rise to it all had largely been scrubbed from the record. Surprisingly, their years-long fight would be marked by more setback and loss than having a landmark case heard by the nation’s highest court might suggest. Today Jason Daubert is 46 and works in technology in La Jolla, California. He’s come to appreciate how Daubert continues to shape cases brought by people who believe they’ve been harmed by a chemical or a medication and are suing corporate defendants. Jason also knows the standard that bears his name influences trials where criminal defendants’ lives were at stake. He said he felt lucky to be a part of it all, although he didn’t exactly crave recognition or praise.

“If anybody deserves kudos, it’s my mom for dealing with all this stuff,” Jason said when I visited him last summer. “She had to deal with a kid who had these horrible birth defects.”

“Well, horrible from her perspective,” he added. “Not so horrible from mine.”


Plaintiffs seeking monetary damages in the type of personal injury lawsuit the Dauberts filed must prove — by a legal burden of proof, or a preponderance of the evidence — that a substance caused injury. In their case against Merrell Dow, that meant proving the drug cocktail in Bendectin caused Jason’s developmental anomalies. While evidence may mean one thing to a scientist, it can mean something else to a jury. One longstanding complaint was that lawyers favored the least impartial experts and another was that juries too often sympathized with injured parties. Then, in 1975, Congress enacted the Federal Rules of Evidence. These rules govern the introduction of evidence into federal trial courts. Rule 702 concerns the opinions of qualified people testifying about “scientific, technical, or other specialized knowledge.” The rule was inclusive: Expert witnesses were allowed into courts when their testimony was deemed relevant.

By the time the Dauberts’ appeal reached to the Supreme Court, in 1992, the high court took an interest because of inconsistencies in the way judges admitted qualified scientific experts. Some of the federal circuit courts turned to the Federal Rules of Evidence for guidance. Others took an anything-goes approach and left it up to a jury to decide. But a few federal judges applied what’s known as the Frye test.

Frye v. United States was a 1923 criminal case involving the use of a prototype version of the polygraph, and the D.C. Circuit Court of Appeals’ ruling in Frye established the standard for expert testimony: Only propositions that had gained “general acceptance” could be admitted. Frye was rarely invoked in civil proceedings. In criminal cases, the standard applied only to new or novel forensic techniques, meaning that defense lawyers could not challenge techniques such as fingerprint or handwriting analysis that have a long history of acceptance in the courts. As a consequence, evidence could be accepted by large groups of people even if a field has no empirical basis.

Photos of Jason Daubert when he was a young boy, displayed on the dining room table. “If I could have given Jason my right arm, I would’ve done it in a heartbeat,” Joyce says. Visual: Ariana Drehsler for Undark

The Dauberts’ suit was a toxic tort, a personal or group injury claim that stems from an exposure to a particular chemical. At the time, the Bendectin lawsuits snowballed into what’s known as a mass tort, where multiple plaintiffs bring civil action against a single corporate defender. At the time, the Bendectin suits landed squarely in the debate over “junk science,” a term popularized in the early 1990s by Peter Huber, a senior fellow at the Manhattan Institute, a conservative think tank in New York City. In his 1991 book “Galileo’s Revenge: Junk Science in the Courtroom,” Huber cited numerous cases where entrepreneurial lawyers had an economic incentive to exaggerate the dangers of certain products, which had spawned a litigation explosion and a cottage industry of purported experts willing to claim anything. He highlighted one case in Philadelphia where a jury awarded nearly a million dollars to a psychic who claimed to have lost her psychic powers after going in for a CT scan, although critics contend her case is something of a legal fable. “‘Let-it-all-in’ legal theory creates the opportunity,” Huber wrote. “The incentive is money: the prospect that the Midas-like touch of a credulous jury will now and again transform scientific dust into gold.”

To tort reform advocates, the lawsuits targeting Bendectin were a perfect example of experts spouting off baseless claims. Charles Fried, former solicitor general who argued the case for Merrell Dow before the Supreme Court, said he felt that the plaintiffs were pushing for billions to trade hands on the basis of junk science. “Some of the things that people were willing to give sworn testimony to was just shocking,” Fried said. “They had white coats and Ph.D.s and they hoped that they could get past summary judgment” — a judgment without a jury trial — “and, if they got past summary judgment, they could squeeze a settlement. That was their hope.”

The plaintiffs countered that corporations and other industries with deep pockets ginned up research solely for the purpose of discrediting litigation. Scientists paid by the asbestos industry, for instance, had suppressed results that could damage the product’s reputation. As Nace explained in an interview with The Washington Informer, Merrell Dow had been “financing scientific articles to get their way.” In some ways, here was the fake news of its time: If you lacked any compelling scientific support for your case, one way to undermine the credibility of your opponents was by calling their evidence “junk science.”

D. Michael Risinger, an evidence scholar and retired professor of law at Seton Hall University in New Jersey, said the claims made by both sides were exaggerated. Still, he conceded that the 1970s and 1980s saw an unbridled expansion of self-anointed experts. Conservative activists wanted to limit corporate liability, but the tort-reform advocates, Risinger said, did have a point. “What their big deal was — and they were not entirely wrong — was that causation in toxic torts had been allowed to become extremely lax.”

In some ways, here was the fake news of its time: If you lacked any compelling scientific support for your case, one way to undermine the credibility of your opponents was by calling their evidence “junk science.”

The question at the core of the Dauberts’ case was whether the plaintiffs’ experts met the existing standard. Initially, the courts concerned themselves with a narrow set of questions: Was Bendectin a teratogen? Did the drug cause birth defects? And, more specifically, was there evidence proving it caused this particular individual’s injuries? But the Dauberts’ appeal soon became a vehicle — or, some argue, more of an ice breaker — for something far more consequential and far-reaching: Does empirical data support an expert’s testimony? Who decides what evidence the jury hears, especially when there’s uncertainty or debate about that evidence?

For the Daubert family, there was an additional layer of complexity. Joyce’s oldest sister married a high-level executive at Dow, and Joyce worried that a fight on behalf of her child against Merrell Dow might fracture the family. She called her father who, she says, told her what he always had. “He made it clear it to me that it wasn’t really a choice. I just needed to stand up and make it happen,” Joyce said. “If I didn’t, how could I ask my students to do the right thing? Or my husband to do the right thing? Or my children to do the right thing?”


On July 26, 2019 — Jason’s 46th birthday — I drove to Ramona, a little desert town in San Diego County to meet Joyce and Bill. Inside, Joyce led me to a back office. Three newspaper clippings hung on the wall: Two about her father’s Christmas tree farm and one with a photo of her and Jason that ran in The New York Times. In that photo, the two wore big, matching smiles.

In 1983, just as news of the Oxendine settlement made headlines all across the county, and not long after Joyce first contacted Nace, the family began working with Mary F. Gillick at Luce, Forward, Hamilton & Scripps, a local law firm in San Diego. Joyce remembers going to a conference room downtown, a book-lined room with big windows that looked out onto the Pacific Ocean. “What the attorneys had said to us,” she said, was “‘You’re young and beautiful people, you’re well educated, your husband’s a veteran. You’ve lived the American dream, this perfect life and you’ve made good in the world.’” She seemed like the perfect witness. Joyce remembered details and dates; she’d obtained her medical records; and, her story was heartbreaking. “Everybody was crying,” she said. “The attorneys were crying.”

In her home office, Joyce pulled open a drawer and began leafing through documents. She’d kept a Polaroid of Jason leaping headlong into a bright blue swimming pool, his left arm straight overhead and his right arm, visibly smaller, angled to his side. She had filed away several pages from a calendar book — one with Nace’s phone number. On another sheet of paper she had calculated her date of conception. There was a settlement offer from Merrell Dow. (Joyce on receiving it: “Put that right in the ‘bite me’ folder.”) There were written responses from their suit, which detailed the three times Joyce took Bendectin in 1972: Once in late October or early November, again in December, and later that month on a trip to Tijuana.

By September of 1989, a judge set a trial date for the Dauberts in the District Court for the Southern District of California, in San Diego. Merrell Dow’s legal team submitted a motion to dismiss the case. In response, Nace and the Dauberts’ legal team offered testimony from eight experts. Alan K. Done, the pediatrician who had appeared in the Oxendine case, would base his opinion on an analysis of four different types of studies (none of which he conducted) to show with a reasonable degree of certainty that Bendectin was a teratogen. Swan, the reproductive epidemiologist, served as a rebuttal witness. By re-analyzing epidemiological data, Swan suggested the evidence was not strong enough to rule out the possibility Bendectin caused defects. (As Swan summarized for me: “No, these do not establish safety.”) The others had expertise in developmental biology, toxicology, biostatistics, and pharmacology.

Merrell Dow argued that the case should not go to a jury. The thrust of their argument was that the science is overwhelming. In November, the San Diego trial judge dismissed the Dauberts’ case, writing that an epidemiological study on more than 130,000 women, which compared those who took the drug against an unexposed population, found no “reproducible or consistent association of birth defects with Bendectin exposure.” (This study, which had been conducted on behalf of a judge in Ohio, was later published.) The Dauberts’ lawyers appealed. Two years later, they lost again when the 9th Circuit Court of Appeals in San Francisco affirmed the decision. In his ruling, Judge Alex Kozinski cited the 1923 Frye ruling, which held that expert testimony must have a consensus within that particular field. The plaintiffs’ experts had not published their analyses in a scientific journal. Their methods had not been subject to peer review, the standard process of evaluating scientific evidence and scrutinizing research prior to publication. What the Daubert family’s experts planned to say at trial, Kozinski wrote, failed to meet all the requirements imposed by the scientific community. Their evidence was not the “generally accepted” means of establishing causation.

But the Dauberts weren’t done, and appealed to the Supreme Court. Even before Jason and Joyce slipped in a side door of the Supreme Court for the hour-long arguments in Daubert v. Merrell Dow Pharmaceuticals, Inc. in March of 1993, their case had attracted a lot of attention. Twenty-two amicus briefs had been submitted by “friends of the court,” who were interested in but not a party to the case, including trial lawyers and epidemiologists, the Pharmaceutical Manufacturers Association (now the Pharmaceutical Research and Manufacturers of America), and the New England Journal of Medicine, along with several other groups of academic researchers.

The arguments got off to an awkward start when Chief Justice William Rehnquist called the court to order. He pronounced Daubert as if their name rhymed with Flaubert. (doe-BEAR is a frequent mispronunciation; the family says DAW-burt.) Over the course of the next 58 minutes, the opposing sides argued their cases. The Dauberts were represented by Michael Gottesman, who said that courts couldn’t always wait for a conclusive answer. Expert witnesses, he said, had to work with the best available science, more like a physician using the newest evidence to treat a patient and less like a scientist trying to prove a new law of gravity. Fried, who represented Merrell Dow, knew Frye was favorable to his client. But Fried doubted a 1923 ruling would take precedent over the 1975 Federal Rules of Evidence, so he singled out Swan and emphasized that unpublished opinions were not reliable since her conclusions had not been vetted by the scientific community.

A newspaper clipping from The New York Times includes a photograph of Joyce and Jason. It now hangs on Joyce Daubert’s wall. Visual: Ariana Drehsler for Undark

The photo that ran in The New York Times — the one Joyce kept on the wall in her home office — was taken after these arguments before the nation’s highest court. She and Jason look equal parts baffled and ecstatic. Afterward, they headed to a reception with their attorneys. “Everybody was there celebrating,” Joyce said, “because they thought that it had gone very well.” How well, they would not learn until three months later, when the Court handed down a ruling, reversing the lower court’s decision — a definitive win for the Dauberts.

But others took a different view. On June 28, 1993, the afternoon of the Supreme Court ruling, the phone rang at Nace’s office in Washington D.C. As Nace remembered, it was a reporter in Kansas City calling with questions that led Nace to believe that Merrell Dow claimed the ruling was a victory. The Times story the next day reported that the ruling in Daubert invited judges to be aggressive in screening out speculative theories masquerading as science — suggesting that the new Daubert standard raised the bar for admitting scientific evidence. A story in The Washington Post reported that the court had relaxed the standard — a lowering of the bar. Lawyers from both sides, as well as several scientists, expressed satisfaction, a worrying sign for anyone trying to make sense of the ruling.


In Daubert v. Merrell Dow, the Supreme Court ruled that the Federal Rules of Evidence superseded Frye, the common law’s standard for reliability. The ruling established what is now known as the Daubert standard. Rather than delineating a bright-line rule (for instance, is the expert generally accepted?), the Court instead assigned federal trial judges to act as gatekeepers (for instance, does the expert’s scientific testimony fall under Rule 702 of the Federal Rules of Evidence?). The judge’s task, Justice Harry Blackmun wrote in the opinion, was to ensure “the testimony’s underlying reasoning or methodology is scientifically valid.” To make sure that only expert witnesses with reliable scientific testimony reach the ears of the jury, Blackmun laid out four factors for excluding junk. These included whether a given technique has been tested, if it’s been peer reviewed and published, if it had a known error rate (or whether statistical data supports the certainty of a hypothesis), and whether it’s accepted in the broader scientific community. In essence, the ruling asks scientists who testify in court to follow the scientific method. In a partial dissent, two of the justices wrote that the new regime would force judges to become “amateur scientists.”

For 70 years, Frye had been deferential to the views of experts, not empirical testing. The Daubert standard was different — or, at least it was supposed to be. Legal scholars have called Daubert v. Merrell Dow perhaps the single most important evidence case and a watershed moment, but it wasn’t that the 1993 opinion was particularly cogent or influential on its own. Rather, Daubert was the first of three influential Supreme Court rulings that concerned evidence submitted by technical experts. Daubert is the overarching term that covers the systematic reshaping of case law that took place in the namesake 1993 ruling and two subsequent decisions in General Electric Co. v. Joiner (1997) and Kumho Tire Co. v. Carmichael (1999). This trilogy of cases established the legal framework for any party to challenge an expert’s qualifications and the reliability of their opinions.

The ruling in Kumho is perhaps the most notable in that it clarified that the standard applies to nonscientific experts, and mandated that judges examine expert claims on a case-by-case basis. Simply put, Daubert challenges are not intended to determine whether a given field of science — such as reproductive epidemiology or forensic handwriting examination — are globally reliable in every case. Rather, judges are supposed to determine whether the methodologies in the particular case at hand are reliable.

In 2000, the U.S. Congress amended Rule 702 to codify the Daubert trilogy, so the rule on expert testimony in the Federal Rules of Evidence now reads: “the expert’s scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue; the testimony is based on sufficient facts or data; the testimony is the product of reliable principles and methods; and the expert has reliably applied the principles and methods to the facts of the case.” A majority of states have adopted Daubert in some form. While so much depends on the legal standard, so much more, it seems, depends on who the defendant is. As Ed Cheng, a professor at Vanderbilt Law School and co-author of “Modern Scientific Evidence,” explained it to me, “Whose ox is being gored changes between the criminal and civil side.”


The afternoon I visited Jason Daubert’s parents in Ramona, Bill slipped an old VHS tape into their player, and we sat down in the living room to watch a spirited talk Joyce gave in 1994 to a class of students at California Western School of Law in San Diego about her stake in the Supreme Court’s ruling. At times, I found the video hard to watch: the sound is slightly tinny; she’s recounting familiar points we’ve already discussed; most of all, though, there’s an optimism about her impending path to victory. “This is the hill I’ll die on,” she told the class at one point. “You only go to war when you think you can win.”

By March of 1994, the Supreme Court sent the Dauberts’ case back to the 9th Circuit Court of Appeals, where, as Nace put it, “we saw the handwriting, we knew it was going to happen.” Judge Kozinski issued the appellate court opinion in January of 1995. Just as he had done three years earlier, Kozinski rejected the plaintiffs’ arguments. Only this time he essentially wielded the high court’s ruling — the Daubert standard — against the family.

“This is the hill I’ll die on. You only go to war when you think you can win.”

“The dispute,” Kozinski wrote, “concerns matters at the very cutting edge of scientific research, where fact meets theory and certainty dissolves into probability.” Despite the uncertainties, Kozinski noted that epidemiological evidence suggested that exposure to Bendectin in utero did not double that child’s risk of being born with birth defects. As such, Kozinski found that the plaintiffs lacked the proof to conclude that Bendectin caused Jason’s birth defects. That October, the Supreme Court refused to hear their final appeal. What had seemed like a unanimous victory before the nation’s highest court had not won their experts entry into court. Nor had it given Joyce what she actually wanted: To go to trial and tell their side of the story to a jury.

In the ensuing years, the ruling in Daubert influenced proceedings that involved major scientific controversies. Daubert challenges have played into decisions on admitting evidence about silicone breast implants; electromagnetic waves emitted by cellphones and high-voltage power lines; vaccines; polychlorinated biphenyls (PCBs); sick building syndrome; contaminated groundwater; police shootings; as well as all sorts of faulty consumer products that range from defective ladders to blockbuster pharmaceutical drugs including Baycol, Vioxx, Zofran, and Risperdal.

In 2019, Daubert hearings were held in two mass torts, both with multi-million-dollar settlements at stake. Tens of thousands of plaintiffs claim that glyphosate, the world’s best-selling herbicide, which was sold as Roundup by Monsanto and is now owned by Bayer, causes non-Hodgkin’s lymphoma; Johnson & Johnson is also facing thousands of suits from plaintiffs who claim its baby powder contains asbestos that caused ovarian cancer. “Those Roundup and talc cases are just Daubert all over again,” said Joseph Sanders, author of “Bendectin on Trial” and a professor at the University of Houston Law Center. “Not necessarily that they’re losing cases, but there’s another group of plaintiffs who were seriously injured and think that a company produced a product that caused the injury.”

Despite their apparent win at the Supreme Court, the Dauberts had been dismissed without a full trial. In some ways, their fate presaged the outcome for plaintiffs bringing similar types of suits. According to Cheng, the evidence scholar at Vanderbilt, “In general, the courts have upped the ante.” Concrete data is difficult to find, especially in state-level cases, but, in 2015, the Searle Civil Justice Institute at George Mason University reviewed 2,127 filings between 2003 and 2014, and found that Daubert motions limited expert testimony nearly half the time. By the mid-1990s, several published studies failed to establish any link between Bendectin and birth defects, and, Cheng said, the courts were right to dismiss the plaintiffs’ claims. From that point forward, he added, judges tend to prevent plaintiff experts from advancing maverick hypotheses that contradict largescale epidemiological evidence, effectively ending litigation and making it harder to win a lucrative payout.

Despite their apparent win at the Supreme Court, the Dauberts had been dismissed without a full trial. In some ways, their fate presaged the outcome for plaintiffs bringing similar types of suits.

As a form of social policy, some scholars contend that the Bendectin litigation was an illegitimate form of pharmaceutical regulation. Physicians have argued that the suits adversely affected public health — with some adding that the litigation exacerbated the lack of clinical testing on pregnant women. Others said the procedural hurdles that force plaintiffs to win twice — once at a Daubert hearing, and then again at a jury trial — make it more difficult to litigate against corporate defendants. If a defendant Dauberts the plaintiffs’ experts, there’s no trial. “Products slip through for a variety of reasons that are both benign, and not benign,” said Elizabeth Chamblee Burch, a professor at the University of Georgia School of Law and author of “Mass Tort Deals: Backroom Bargaining in Multidistrict Litigation. “And so we have to rely on the litigation system to provide that failsafe to give people recourse against manufacturers.” Prior to approval, though, researchers test products on a relatively small proportion of the total population, and so litigation serves as a stopgap for identifying rare adverse outcomes that could crop up when several hundreds of millions of people are exposed to those products over longer periods of time.

The real difficulty arises, Burch said, when the scientific knowledge lags behind litigation. “Judges just have one chance to try to make the right decision,” she said. “We would love to have perfect knowledge so that we can get these difficult questions right all the time. But it’s about probabilities and guesses. And so Daubert is supposed to provide that screen between what’s too speculative, and what’s a reasonable conclusion.”

The Bendectin litigation exposed these gaps, and it also illustrates how tweaking one safety valve does not mend every break in the system. In 1983, Merrell Dow pulled the drug from the market, citing rising insurance premiums. The company denied wrongdoing; all the cases were eventually dismissed; and studies suggest that the incidence of birth defects remained the same even after the drug was removed. (In 1999, Merrell Dow was acquired by the company now known as Sanofi.) Then, in 2013, Duchesnay USA, an affiliate of Duchesnay Inc., a private Canadian company, released a new and reformulated version of Bendectin called Diclegis, which contains two of the three same ingredients as the original drug. Diclegis is currently approved for the treatment of nausea and vomiting during pregnancy.

“Judges just have one chance to try to make the right decision. We would love to have perfect knowledge so that we can get these difficult questions right all the time. But it’s about probabilities and guesses.”

Joyce is not convinced the drug should be back on the market, and she isn’t alone. Nav Persaud, a family doctor in Toronto who has examined unpublished data and re-analyzed the clinical trials, argues that these drugs’ safety and the relatively low risk of birth defects are irrelevant because, he said, the medications are “definitely not effective.” This matches Joyce’s experience: She says Bendectin never helped her nausea any of the three times she took it. (Duchesnay maintains the safety and efficacy of Diclegis.) Persaud’s analysis suggests another kind of hole in the regulatory fence, but the tort system offers fewer incentives to challenge ineffective drugs that don’t physically hurt anyone, but also don’t really work.

Diclegis, the new and reformulated version of Bendectin, served as a coda that further undermined Joyce’s faith in a just world. The drug had done nothing to alleviate her symptoms, and she believes it had done everything to irrevocably alter her family’s life. Nothing had changed her conviction. Nothing absolved her, either, and Joyce felt she could not help but agonize over what could have been. “If I could have given Jason my right arm, I would’ve done it in a heartbeat,” she said. “I don’t know that I’m talking about feeling guilty, but I’m very responsible. I know that. That’s hard to live with because I can’t — I can’t do anything about it.”

She had tried. In denying the Dauberts, the courts denied their beliefs. Joyce felt powerless to make things right and, to her mind, the entire legal system remains profoundly disillusioning. “And pretty disheartening, and, very personally, it makes me feel stupid that I was so naïve to think that, after we’d invested so much in the case, that we would get justice,” she said. “We didn’t get into court. We didn’t get our case heard.” When she’s called for jury duty, she introduces herself: “I’m Daubert of Daubert versus Merrell Dow,” she says. “I don’t want to sit on this jury and pretend that I can pass judgment on somebody when there is no justice. Please allow me to be excused.”

Jason and Joyce Daubert look through old family albums. Visual: Ariana Drehsler for Undark

The Daubert standard concerns more than just corporate defendants: It opened the possibility of challenging any scientific expert. Fried, who argued on behalf of Merrell Dow and is now a professor at Harvard Law School, said the rules of evidence apply to all federal cases, not just civil litigation. “We’ve moved towards a more rational system of justice,” he said, “a system that’s based on reason rather than on how many paid liars you can line up.”

But when the first significant Daubert challenge in criminal courts came up in 1995, it offered a revealing window into how the standard would be unevenly applied. At the time, federal prosecutors in New York accused Roberta and Eileen Starzecpyzel of stealing artwork, selling it to Sotheby’s and Christie’s, and forging the owner’s signature — Ethel Brownstone, Roberta’s aunt — on two documents. The government presented a forensic document examiner who claimed the signature was forged. The Starzecpyzels’ attorney mounted a Daubert challenge. Risinger, the retired Seton Hall law professor, and several colleagues, maintained that there were no credible studies to support the examiner’s methods. As the trial judge explained, “forensic document examination, which clothes itself with the trappings of science, does not rest on carefully articulated postulates, does not employ rigorous methodology, and has not convincingly documented the accuracy of its determinations.” Risinger watched the testimony, and he did not see how the Starzecpyzels could lose. Yet, they lost. Precisely because handwriting identification was not considered “scientific,” the judge ruled that Daubert did not apply. Both women were convicted; Eileen was sentenced to more than two years in prison and Roberta to three-and-a-half.

In 1999, defense attorneys challenged fingerprint identification — another forensic science that had not evolved out of rigorous testing of experimental hypothesis and did not appear to meet the criteria set out in Daubert. Again, the challenge failed. But these types of rulings helped expose a lack of objective standards among forensic experts who matched microscopic patterns from bullets and spent cartridges, hair samples, dental impressions, footprints, tire tracks, and blood spatter. Many of these fields emerged out of law enforcement. Practitioners had no scientific basis for their claims, and lacked validation, error rates, and reliability testing. In a 2008 Hastings Law Journal paper, one legal scholar compared these fields to modern-day phrenology, referring to the discredited practice of diagnosing patients based on the bumps found on the surface of their skulls: “The science of the day was more than adequate to demonstrate that phrenology had no basis; but that science was widely ignored.”

Wrongful convictions only served to underscore these problems. Experts compared DNA from a suspect and a crime scene, and this irrefutable evidence began to exonerate criminal defendants. In one review published in the journal Science, DNA evidence revealed that flawed forensic techniques played into 63 percent of cases where innocent people had erroneously been found guilty. Daubert was supposed to offer defendants a means to challenge junk science. But judges rarely sided with the defense. In 2000, Risinger published an article that reviewed hundreds of cases, and found that judges overwhelmingly favored the prosecution. Risinger suspects the outcome was consistent with the view that the Supreme Court’s 1993 ruling in Daubert was first and foremost an effort to cut down on corporate liability and rein in the abuses of plaintiff’s lawyers in civil cases. “They couldn’t explicitly say, ‘This doesn’t apply to prosecution-proffered expertise in criminal cases,’” he said. “They couldn’t say that with a straight face, but functionally, that is the way it was taken by courts thereafter. Nobody ever said it explicitly. It’s just the way it worked out on the ground.”

Judge Kozinski, who ruled against the Dauberts in 1991 and 1995, came to recognize these asymmetrical applications of the Daubert standard. Kozinski was aware that judges held expert testimony in criminal cases to a lesser standard than in civil litigation. Some of this may be attributable to the pro-prosecution biases of judges, especially in states where judges hold an elected position and want to appear tough on crime to voters. It also comes down to resources. Criminal defense lawyers are not ordinarily paid well enough to have the time or resources to find competent researchers to reach higher quality decisions. Over the years, Kozinski voiced concerns about the incentives of expert witnesses who were paid to develop their opinions expressly for purposes of testifying. “I’m a capitalist,” he said at a 2017 symposium. “I believe in the free market.” But Kozinski added that he harbored doubts about forensic labs staffed by government employees whose careers depend on pleasing prosecutors. “Prosecutors want to get convictions, and if they have crime labs that are part of their organization, the implicit and explicit incentives will be to reward evidence that supports the prosecution.”

The ability to prove something false is a hallmark of science. Daubert handed judges those tools. As Kozinski realized, failing to hold a Daubert hearing, “was an error of law.” (While he saw himself as a criminal-justice reformer, others saw a perpetrator. In 2017, Kozinski retired after numerous accusations of sexual misconduct. He has not been charged.)

It’s impossible to say how often errors occur since no national agency monitors the implementation of forensic science. Few trial judges granted criminal defendants’ requests for Daubert hearings. As one 2018 analysis of state cases says, “It is incredibly rare to find any discussion of reliability, except in one context: when courts exclude defense experts.” Despite broader calls for reform and despite reports drawing attention to systematic problems — most notably a landmark 2009 report by the National Academy of Sciences — Daubert demonstrated that judicial rulings alone would not revolutionize the criminal justice system. Judges acted not so much as gatekeepers, but more like scientifically naïve ushers.

Daubert offers the possibility of challenging authorities as the source of knowledge. Judges, meanwhile, continue to admit forensic disciplines that have been accepted in the courts for more than a hundred years. Carrie Leonetti, editor of New Criminal Law Review, said when it comes to allowing evidence into a case, judges sometimes grandfather in pattern-matching forensics, which, Leonetti argues is an egregious misapplication of a rule intended for facts that are so obvious, and so well known, that they cannot reasonably be doubted. “I think what’s happened with a lot of the forensic sciences is just nobody wants to start,” she said. “If you start pulling that string, you unravel the whole sweater, right?”


When I met Jason, at his apartment in La Jolla, I began to see how the Daubert family pulled at one of those loose threads, expecting the truth to unravel, only to find themselves with a bunch of tangled knots. And knots, I learn in not so few words, have always presented him with problems. Jason, who is tall and balding with a ring of graying hair, works in information technology. He referred to himself as a geek and also “unconventional and lucky as heck.” We talked for a while in his kitchen, which was ringed with Lego sets, and then headed downstairs to his garage, which was ringed with giant totes filled with Lego parts. Then we hopped in his Tesla. (Jason to me: “Basically, a spyware car.”) Like most of his day-to-day activities, Jason drives without any modifications, or assistive technology. “I don’t act disabled,” he said, “for lack of a better way of saying it.”

Jason said he never really dwelled on the emotional anguish of losing abilities he never had. “For me, it’s like the background,” he says.

Visual: Ariana Drehsler for Undark

There are specific details that Joyce recalls — the day she watched him learn to tie his shoes using his mouth. Jason can’t remember these details, or maybe he doesn’t want to. He shares some impressions of his childhood about being scared of thunderstorms, about being bullied, and about having trouble dating. Jason missed out on swim class because he couldn’t undo a knot in his bathing suit. “Not enough fingers,” he said. But the expectations with which his parents raised him seemed to have a determining effect on who he’d become: Jason blossomed into someone who was unapologetically different. For him, there was never a before, so Jason said he never really dwelled on the emotional anguish of losing abilities he never had. “For me, it’s like the background,” he said. “Like it’s Tuesday. Like, ‘Yeah, I got a deformed right arm just like every other day.”

In downtown La Jolla, no one appeared to give us a second glance. We were just two guys eating fish tacos. In legal circles, though, Jason is a minor celebrity. The citations to Daubert could fill an entire book, but the family’s side of the story was rarely told. The 1993 Supreme Court ruling had been a turning point for the admissibility of scientific evidence. It was also a time when Jason got excited to go to Washington, D.C. to meet friends he knew from playing online games similar to Dungeons & Dragons. When I pulled up the photo of him at court, he remembered only his tie — something he’s rarely ever worn. The case had not ruined his life. Nor, he told me, was its outcome so unequivocally harmful that he harbored any guilt.

Jason said that when he and his then-wife decided to have a child, he harbored all sorts of irrational fears. Even after seeing the sonogram, he’d asked to have a diagnostic test done to double check for chromosomal abnormalities. Jason knew that his own birth defects were likely not genetic, but still felt a profound wave of relief when their daughter, Serena, was born with ten fingers and ten toes. He briefly mulled reopening his case when Diclegis, the updated version of Bendectin, was re-approved.

But he thought better of it when he remembered how hard his mother had already fought only to get smacked down. “Like that’s kind of the universe telling you this isn’t the right way to go.” I asked if the eventual outcome felt like a righteous victory followed by stinging defeat. Jason said no. “It feels like we pulled victory out of the jaws of defeat, but actually it was defeat out of the jaws of victory.” As for as the legal standard, Jason recognized what Daubert stood for: “Good science should be all that gets into the courts.”


Peter Andrey Smith is a freelance reporter. His stories have been featured in Outside magazine, The New York Times, and WNYC Radiolab.

Peter Andrey Smith is a senior contributor at Undark. His stories have also been featured in Science, STAT, The New York Times, and WNYC Radiolab.