For those of you hailing the fact-checking of Trump’s tweets, don’t forget that fact checks are only as good as the fact checkers.

Here’s an AP fact check from the July 2019 Democratic debates that misunderstands the premise so thoroughly that I don’t even know where to begin.

For folks who don’t practice civil rights law regularly: Qualified immunity is a doctrine in civil suits against the government that dramatically raises the standard for when you can recover money for constitutional violations. The constitutional violation must be “clearly established” as a violation of the Constitution at the time it occurs. Believe it or not, this actually matters in a lot of cases, even when you think it shouldn’t.

Take this case, Mullenix v. Luna. It’s not all that out of the ordinary for qualified immunity cases. Police shot and killed the driver of a car in a chase. The driver’s estate sued the police department, claiming excessive force.

The trial court found the police officer liable, and the appeals court agreed. The Supreme Court reversed, because shooting at a moving car to stop the driver was not “clearly established” as a violation of the Fourth Amendment.

In doing so, the Court relied not on a finding that shooting a moving car was constitutional, but that in the past, shooting at a moving car was found to be “not clearly established” as a constitutional violation.

Far from clarifying the issue, excessive force cases involving car chases reveal the hazy legal backdrop against which Mullenix acted. In Brosseau itself, the Court held that an officer did not violate clearly established law when she shot a fleeing suspect out of fear that he endangered “other officers on foot who [she] believed were in the immediate area,” “the occupied vehicles in [his] path,” and “any other citizens who might be in the area.”

Does this sound like some circular logic? I hope so.

As you might imagine, there was an impassioned dissent in Mullenix. But it was a solo dissent by Justice Sotomayor. Ginsburg, Kagan, and Breyer all signed onto the majority opinion (which was released per curiam, so we have no idea who wrote it). There are plenty more horrors embedded in the qualified immunity doctrine that are worth an in-depth look in a longer article.

Qualified immunity isn’t from a statute. It’s a completely judge-made doctrine. Now “judicial activism” is usually just code for “thing I don’t like from a court”, but qualified immunity fits almost any definition. It was never passed by a legislature, and it prohibits people from enforcing their rights, undercutting the guarantees of the Constitution.

Julian Castro’s debate answer back in July was about policing reforms. Getting rid of qualified immunity is and should be a core part of any comprehensive strategy on this front, particularly at the federal level where control over police forces is indirect. As a civil litigation reform, it should and would work in parallel with criminal reforms that would ensure that police officers are held accountable for murder when it occurs (such as sending every police killing to an independent special prosecutor). 

The Associated Press completely misunderstood his answer (which I unfortunately can’t find footage of), even though I readily understood it. This may be because I work in this field, but any lawyer who has experience with federal litigation could have told them this fact check is absurd. Indeed, Slate published this piece in the wake of Castro’s debate answer, perhaps a better series of illustrations of the evils of qualified immunity than the one I’ve just given.

But back to Twitter. We have no idea how this fact-check feature will be used. Will it primarily be used for elected officials, or will other commentators also be subject to it? Will it be used against progressives who agitate for big reforms the way the Castro fact-check was? Is there a team of people who identifies which tweets deserve scrutiny for fact-checking? Are there standards that govern this rule? Even if the rule is of neutral application, it might be implemented in a biased manner.

Or perhaps there is an algorithm Twitter will use in selecting tweets to fact-check. Will the algorithm address threads, or will each individual tweet be scrutinized without context? And algorithms can still be biased if their inputs are biased.

I would like to know plenty more before I believe that this will be a net good for the platform. As of right now, I have concerns.

In any case, none of this might matter all that much. We are dealing with an administration that doesn’t care to traffic in truth, and a large segment of the country that is willing to follow. Even if this fact-checking works perfectly and doesn’t serve to muffle voices for change, it will not save us.

We are the ones we’re waiting for. That’s going to be true regardless of anything Twitter does. Based on recent news, and since most people who read this are white or white-passing, I’d start with some introspection about our role in unequal systems.

We can do this. But it is we who must.