Attorney General Jeff Sessions. (Yuri Gripas/Reuters)

When Jeff Sessions was a senator on the Senate Judiciary Committee, he was part of hearings to address the National Academy of Sciences report on the use of forensics in America’s courtrooms. The NAS report had been commissioned by Congress after DNA testing had revealed not only that hundreds of people had been wrongly convicted of serious crimes like murder and rape, but also that about half those people had been convicted due in part to or because of forensic testimony that could only have been wrong.

Sessions wasn’t buying it. “I don’t think we should suggest that those proven scientific principles that we’ve been using for decades are somehow uncertain and leaving prosecutors having to fend off challenges on the most basic issues in a trial,” he said, rebutting the scientists who had come to precisely that conclusion in their report. The “scientific” and “proven” parts were precisely what the report found lacking in too many forensic disciplines. Sessions either didn’t read it — he has a record of criticizing reports without reading them — or simply dismissed it.

When witnesses noted that there was no scientific research to support the field of handwriting analysis, Sessions remarked, “Well, I’ve seen them testify and I’ve seen blow-ups of the handwriting, and it’s pretty impressive.” Who are you going to believe, a team of scientists, or Jeff Sessions’s sense of wonder?

Longtime critics of bad forensics worried that the NAS report was too diplomatic: For example, it didn’t call for the abolition of any dubious disciplines. But for Sessions, even tepid criticism of the tools he and other prosecutors had used for years was a threat. One witness at the hearings was a prosecutor from Wyoming who was testifying in opposition to the report. Sessions tossed him a softball: “Do you believe that the report, perhaps trying to get our attention, used some pretty strong language suggesting the unreliability of what I have always understood to be proven scientific techniques? Is that something that the district attorneys are finding . . . that this is being thrown up to create the impression with a jury that there’s no basis for these kinds of reports?” His main concern was not whether evidence was accurate, but whether the report could make it more difficult to get convictions.

All of which brings us to the big news this week, which, given Sessions’s history, shouldn’t be terribly surprising.

Attorney General Jeff Sessions will end a Justice Department partnership with independent scientists to raise forensic science standards and has suspended an expanded review of FBI testimony across several techniques that have come under question, saying a new strategy will be set by an in-house team of law enforcement advisers.

In a statement Monday, Sessions said he would not renew the National Commission on Forensic Science, a roughly 30-member advisory panel of scientists, judges, crime lab leaders, prosecutors and defense lawyers chartered by the Obama administration in 2013.

A path to meet needs of overburdened crime labs will be set by a yet-to-be-named senior forensic adviser and an internal department crime task force, Sessions’s statement said.

The announcement came as the commission began its last, two-day meeting before its term ends April 23, and as some of its most far-reaching final recommendations remained hanging before the department.

The NCFS was created by the Obama administration in the wake of mounting evidence that forensic evidence was being misstated and misused in courtrooms. The commission was charged with coming up with broad directives on proficiency training, accreditation and certification, and quality control practices.

Sessions’s decision to end the commission is somewhat complicated by a vote the commission members took earlier this year. Under a memorandum of understanding with the Justice Department, the NCFS was initially commissioned to operate for two years. During discussion over what to put in a summary report of the commission’s work, the members voted 16 to 15 not to include language recommending that the commission be renewed.

That could be interpreted to mean that the NCFS members themselves voted to end their own commission — which would make Sessions’s decision to end it seem less controversial. But sources familiar with committee deliberations tell me that this isn’t quite right — a strong majority of the commission does want to continue its work. Some of the 16 who voted against the measure did so because they thought the summary report should include only the commission’s findings; they felt that adding the language about renewal seemed self-serving. Others believe the commission is important and should continue but were somewhat disillusioned by the new administration’s antipathy toward science. Still others wanted the commission to continue, but under a new structure and organization, preferably outside of the Justice Department. According to the sources, there were a few members who may have voted to end the commission entirely, but they were firmly in the minority.

Sessions may not be finished. There’s another important forensics reform initiative started during the Obama administration that Sessions may target next. The National Institute of Standards and Technology (NIST) also oversees forensic science and has long sponsored “working groups” in various fields of forensics. One such working group is currently looking into fingerprint analysis, a field Sessions seems particularly concerned may soon be undermined by scientists. The working groups have been around for decades, but under President Barack Obama, NIST also established the Organization of Scientific Area Committees, or OSACs. These are committees of practitioners, scientists and skeptics who have been asked to delve into the scientific research supporting or refuting more than two dozen fields of forensics.

NIST is under the Commerce Department, not the DOJ, so Sessions couldn’t himself disband the working groups or the OSACs. But the OSACs are partly funded by the DOJ. Sessions could halt that funding. According to recent public comments from the OSAC’s director, NIST has been looking for alternate sources of funding to keep the committees operational.

Once caveat worth noting here: As is often the case with these sorts of committees, the OSACs themselves have been subject to infighting and politicking. As I’ve pointed out here at The Watch, for example, the OSAC looking at bite-mark analysis was loaded with a majority of members who either practice or openly support bite mark matching. Still, the committee also included skeptics, and any recommendations would have to be approved by the NIST Forensic Science Standards Board, a group made up almost entirely of established scientists. Whatever their flaws, the OSACs are at least taking a hard look at problematic fields of forensics and trying to establish uniform standards in others.

Even if Sessions can’t unilaterally end the OSACs or NIST working groups, he wields considerable influence in the Trump administration. Trump himself, like Sessions, spouts law-and-order rhetoric, seems skeptical of criminal-justice reform, and is a man who has little patience for inconvenient science. It isn’t difficult to imagine Sessions persuading Trump to go after NIST as well. There’s a reason the old guard in the forensics community celebrated Trump’s election.

Judging by his public record, Sessions believes the criminal-justice system’s primary job is to fill up prisons. You need only look at his enthusiasm for the drug war, his efforts to block sentencing reform or, well, just about anything he has said on the topic over the course of his career. And as we’ve seen with his statements on policing, he’s also no fan of federal oversight. It shouldn’t be surprising why he seems so irked by reforms that would undermine both views.

During his time in the Senate, Sessions’s main concern when it comes to forensics has been that the field suffered from a lack of funding, which he worried causes backlogs and over-burdened crime labs — all of which makes it more difficult to put people behind bars. “So, tens of thousands of people, I suggest, are not being promptly tried,” he said at the 2009 hearing on the NAS report. “While they’re out on bail or un-indicted, they’re committing crimes this very moment. A lot of that is because we’ve not invested enough in our forensic sciences so that we can get accurate and prompt reports.”

For Sessions, the federal government’s only real responsibility in the area of forensics is to provide the resources and training necessary to help local cops and prosecutors put people in prison. Any oversight or quality control is meddling. The FBI’s own history shows why he’s wrong. Over the past several years we’ve learned that the agency — whose crime lab is considered one of the most prestigious in the world — faked an entire field of forensic science. Not once, but twice.

But when law enforcement clashes with science, Sessions sides with law enforcement. And this really gets to the  heart of the ongoing problem with forensics. Most fields of forensics were invented and developed by police agencies, not in scientific labs. In fact, for most of the 20th century, the scientific community largely steered clear of the criminal-justice system. Science and law are two entirely different fields. They’re driven by different goals, different processes, and different values. Science is the gradual accumulation of knowledge through trial, error and corroboration. The criminal-justice system tries to get at truth through an adversarial process, after which it prioritizes preserving jury verdicts.

But prosecutors also learned early on that jurors like expertise, and so a demand emerged for expert witnesses. Unfortunately, no one was making sure the expertise on offer was legitimate. And so we got an entire profession of experts who were willing to say things that actual scientists wouldn’t.

Juries crave certainty. They swoon for expert witnesses who can wow them with technology. Real scientists don’t speak in certainties. They talk about margins of error, which means jurors find real science less convincing. Our adversarial system may be the best system available for assessing evidence, but it’s hostile to good science. That’s why it’s important that experts willing to tell juries un-scientific things be barred from the courtroom. It’s why we need judges to consult with real scientists when making such decisions. And it’s why we need prosecutors honest enough to resist the temptation to seal convictions with charlatanism masquerading as expertise.

But the system also makes all of those things difficult, so it wasn’t until the 1990s and DNA testing — technology developed by scientists, not law enforcement — that we began to see just how wrong forensic analysis could be. There were particular problems within the highly subjective disciplines known as “pattern matching” — fields like hair and fiber analysis and bite-mark matching. Even then, it took another couple of decades before the scientific community began to rigorously apply the scientific method to the claims and methods of forensic analysts. Once they did get involved, in report after report — including that 2009 NAS report, reports from the Texas Forensic Science Commission, and the President’s Council of Advisors on Science & Technology (PCAST) report last fall — the “science” in”forensic science” has been found to be somewhere between insufficient and nonexistent.

And yet instead of a sense of humility at the profound implications of these reports, law enforcement officials and prosecutors like Sessions have retreated to the battle lines that defined much of the last century. They’ve attacked the scientists as biased, or made bizarre arguments that forensics should be judged on principles other than scientific principles. (Before we give the Obama administration too much praise, it’s worth pointing out that after PCAST issued its damning report on forensics last year, Obama’s own attorney general Loretta Lynch immediately dismissed it.)

In only the past decade or so, we have finally managed to nudge forensic analysis at least partly out from its dark ages. While the courts haven’t paid much attention yet, these committees and their reports were the first steps toward subjecting forensics to principles like peer review, blind testing, statistical analysis and the more modern concepts grounded in those principles, like sequential unmasking. Sessions is poised to eradicate that progress.

It’s hard to overstate the urgency here. As I’ve written before, DNA testing was a wake-up call. It is not a panacea. In the small pool of cases for which DNA testing is dispositive of guilt, the technology revealed serious problems with our criminal-justice system, and forensic evidence was one of the most significant. But the window of opportunity to correct the mistakes exposed by DNA testing will remain open only as long as DNA exonerations are fresh enough in the minds of the public to sustain support for reform. Soon we’ll have exhausted the pool of cases that are old enough to have been decided before DNA testing would have been done early in the investigation, but recent enough that the DNA sample is still available and hasn’t degraded. If the Jeff Sessionses of the world can put off reforms until those cases are exhausted, it becomes much easier to argue that the problems DNA testing exposed are mere relics of the past — that we needn’t worry about all of this anymore. And if we don’t fix the problems DNA testing has exposed, they’ll continue to plague all the other cases for which DNA testing isn’t useful. There will be few new exonerations to argue otherwise, and in the meantime, there will always be a grisly murder or brutal rape in the news to demagogue about the danger of “tying the hands” of police and prosecutors — a tactic Sessions deploys with ease. Only this time, there might be no new technology to let us know that we’re making mistakes. We’ll continue on, blind to the problems we failed to fix.

The Obama administration finally provided a platform and framework for real scientists to scrutinize the way forensics is used in the courtroom — but then largely ignored their recommendations. The Trump administration now seems intent on eliminating the platform, too.

That window for reform is closing, and Sessions seems content to let it slam shut. He’d prefer to keep forensics in the dark.