The comfortable consensus among court observers goes something like this: the judiciary should stay in its lane. Judges aren't scientists. They shouldn't second-guess expert agencies. Let the EPA do environmental work. Let health officials handle public health. Let the Nuclear Regulatory Commission regulate nuclear plants.
This sounds reasonable. It's the kind of thing that gets nodded at during legal conferences and written into judicial opinions with appropriate gravitas.
But that consensus is doing exactly what it's supposed to do: making us comfortable with an uncomfortable shift already underway. The real question isn't whether courts should defer to experts. It's what gets broken when courts decide which expert conclusions they'll accept and which ones they'll reject.
Consider what's already happening at the margins. A federal appeals court recently upheld an injunction blocking an order to create a federal voter list, citing concerns that weren't primarily about existing law but about what the court believed the policy would do. Nuclear regulators won a case, but only after litigation that essentially forced them to defend their decision not to consider climate impacts as thoroughly as some believed necessary. These aren't edge cases anymore. They're the shape of things to come.
The pattern is this: litigants lose at the agency level, then lose again in administrative appeals, then take their fight to federal court. And increasingly, courts are willing to reweigh the scientific judgments underneath the policy. Not always. Not consistently. But enough to matter.
Here's what makes this dangerous: judges don't get smarter about science just because they put on the robe. They get something else. They get discretion. They get the ability to say that an agency's interpretation of climate data wasn't "reasonable." They get to decide whether a health official's determination was sufficiently supported. And they get to do this while maintaining the appearance of judicial restraint, because they're not directly overruling the agency. They're just saying the agency didn't follow procedure quite right. Or didn't explain itself quite well enough.
This matters because science doesn't work the way courts do. Science is probabilistic. Courts need certainty. Science changes as evidence accumulates. Courts need precedent. Science requires experts to make judgment calls within their authority. Courts want to second-guess those calls if they seem off.
The real break comes when you realize that litigation itself becomes a tool for delaying or derailing scientific consensus. If you can tie up environmental regulations in court for years while arguing about statistical methodology, you've effectively killed the regulation even if you eventually lose. If you can force regulators to defend climate assessments in depositions and cross-examinations, you've transformed scientific work into adversarial theater.
The consensus says judges should defer. Fine. But judges increasingly aren't deferring. They're engaging. And that's not inherently wrong. But it's not neutral either.
What gets broken is the ability of agencies to act with any real finality. What gets broken is the distinction between reasonable disagreement within science and legal disputes. What gets broken is the pace at which governments can respond to problems that don't wait for litigation to settle.
The courts aren't going to become less willing to review agency decisions on their scientific merits. That ship sailed years ago. The question is whether we're honest about what that means: not judicial restraint, but judicial substitution. Not deference to expertise, but a new form of expertise that wears robes.
That's worth discussing. The comfortable consensus isn't.