The Washington PostDemocracy Dies in Darkness

How to reform (and how not to reform) laws governing police raids

Placeholder while article actions load

The legislatures of two relatively conservative states are considering new bills that would put some restrictions on when and how police can break into private homes. First, in Georgia, “Bou Bou’s Law” would require police to show probable cause that suspects could be dangerous to police or could destroy evidence if police were required to knock and announce themselves before forcing their way into a home. It basically raises the standard of evidence for no-knock raids from reasonable suspicion to probable cause. The bill is named for Bounkham “Bou Bou” Phonesavanh, a toddler who was badly burned and nearly killed by a flash grenade during a drug raid last year.

But as Jacob Sullum points out at Reason, it’s far from clear that the new law would have even prevented the incident for which it’s named.

The cops who raided the home where he was staying with his sisters and parents were looking for his cousin Wanis Thonetheva, a small-time meth dealer, and they argued that his history of gun charges suggested they would face violent resistance.
It turned out that Thonetheva no longer lived in the house, where police found no drugs or weapons. He was unarmed when police arrested him later that day at a different location.
The Habersham County grand jury that investigated the raid nevertheless concluded that Thonetheva’s history justified the no-knock warrant. Even under the stricter standard favored by [Sen. Vincent] Fort, a judge might agree.
Assuming police could not meet that test and therefore had to announce themselves, it might not have made any difference, since the raid occurred between 2 and 3 a.m, when everyone in the house was asleep. It is doubtful whether the Phonesavanhs would have heard and understood the police, let alone that they would have had enough time to answer the door before it was knocked down.
Police are supposed to wait a “reasonable time” after announcing themselves, but it’s not clear what that means. According to a 2003 Supreme Court decision, the relevant question is not how long it takes to answer the door but how long it takes to flush drugs down the toilet. In that case, the Court said 15 to 20 seconds was plenty of time. For people awakened in the middle of the night, that is probably not enough.

And this gets to the heart of the problem. The no-knock raid is different from other raids in that it dispenses with the knock-and-announce requirement. This requirement, which dates back centuries to English common law, is part and parcel with the Castle Doctrine, the idea that the home should be a place of peace and sanctuary and that even when police have just cause to search a house, residents should be given every opportunity to come to the door, answer peacefully and let the officers in, thus avoiding damage to their property and violence to their person. (It also gives innocent residents the opportunity to point out to the police that they have made a mistake.) Under the Castle Doctrine, suspects are presumed innocent until proven guilty.

The Supreme Court began chipping away at that presumption in 1963 with Ker v. California. In that case, the majority allowed for a home entry without announcement if police encountered one or more exigent circumstances, specifically if the police believed that announcing themselves would give suspects the opportunity to destroy evidence or to arm themselves. As Justice William Brennan pointed out in his dissent, the majority opinion does “obvious violence to the presumption of innocence.” That is, it begins its evaluation of when police may enter a home without announcing by presuming that the suspects are guilty — not just of the crime of which they’re suspected, but that they’re also capable of committing the additional crimes of destroying evidence and/or assaulting or attempting to kill law enforcement officers.

The case to which Sullum refers, U.S. v. Banksin some ways represents a complete abandonment of the Castle Doctrine. As Sullum points out, the issue in the case is how long the police must wait after knocking and announcing before they can force their way into a home. But the majority goes about calculating that period of time not by estimating how long it would take a resident to come to the door to avoid violence and destruction of property — as it would do if it were starting with a presumption of innocence — but by estimating how long it would take someone to destroy drug evidence, as it would do if it were starting with a presumption of guilt. So while the court’s majority in Banks affirmed the knock-and-announce requirement, it dispensed with the very reason the requirement exists in the first place. It reduced the requirement to a mere formality, a kind of ritualistic but ultimately meaningless tribute to a time when the law took more seriously the presumption of innocence and the sanctity of the home.

If you’re asleep in a bedroom, there’s little difference in whether the police burst in without knocking; announce themselves as the door is coming down; or even knock, announce and wait a few seconds before applying the battering ram. The entire point of “dynamic entry” is to take the occupants of a building by surprise — to overwhelm them with force and violence before they have an opportunity to do much of anything, much less come and answer the door. Once you’ve decided to use dynamic entry tactics, you’ve already dispensed with the entire purpose of the knock-and-announce requirement. In fact, when the police do knock and announce, any resident who does manage to open the door before it is blown open is probably subjecting himself or herself to a lot more danger. (You certainly don’t want to be holding anything in your hand.)

This is why the proposed Georgia bill is basically meaningless. Georgia police could still force their way into homes in the middle of the night. They could still conduct dangerous, highly volatile drug raids that put both police and citizens in peril. They’d just have to take 20 seconds or so to pay homage to a legal principle that was long ago killed off by the courts.

That the Georgia bill still allows no-knock raids to prevent the destruction of evidence is also problematic. Allowing this exception may make sense if the police are apprehending, say, a murderer or some other violent criminal who they fear may destroy evidence of his or her crime. But most of the time, we’re talking about drug cases. I’d submit that if your suspect’s drug supply is small enough to be quickly disposed of before police enter the house, perhaps this isn’t a serious enough criminal to merit the sort of violence that comes with a no-knock raid. (Or, again, with any “dynamic entry” raid, whether or not the police first announce themselves.)

On the other hand, the reform bill just introduced in Utah is quite a bit better. The bill’s chief sponsor is state Sen. Steve Urquhart (R) and is the product of the continuing great work of Connor Boyack and the Libertas Institute, a libertarian think tank in Salt Lake City. The bill would eliminate no-knock raids for the preservation of evidence. It would require the police to show that a suspect poses a risk to their safety for any forced-entry raid (not just a no-knock raid), and it raises the standard of evidence to probable cause. It would require raiding cops to wear uniforms with large, conspicuous lettering indicating that they are police, and it would require all raiding cops to wear a body camera. Incredibly, in Utah (and quite a few other places) the warrants for these raids can be issued by magistrates and justices of the peace who have no training in criminal law. The new bill would end that, too. Most interesting, the Utah bill would statutorily override the Supreme Court’s decision in Hudson v. Michigan. In that case, the court found that even when the police clearly violate the knock-and-announce rule, any evidence they find in the subsequent search can still be used against the suspect in court. Under this bill, such evidence would be inadmissible in Utah.

The Utah bill hasn’t yet passed. I suspect it will be vigorously opposed by police organizations and will probably be watered down a bit if it does pass. Still, it’s probably the most substantive bill of its type that I’ve seen. It’s a bill that demonstrates a clear understanding of the issues and liberties at stake. (There’s a really interesting police reform movement underway in Utah, and Boyack is a big reason why.)

So what would an ideal law governing these raids look like? I think it would bar any forcible entry into a private residence unless the police know that a violent crime is in process or that a violent crime will be committed unless the police intervene. These raids can have fatal consequences. For a recent example in Georgia, see the sad story of David Hooks. For a recent example in Utah, see the sad story of Matthew David Stewart and Officer Jared Francom.

But unless we’re talking about an imminent violent crime, citizens should be given ample opportunity to avoid the violence of one of these raids. An officer with the Houston Police Department recently told me that in that city the police no longer used forced entry for drug raids. Instead, they surround a residence and call out to the people inside, either by telephone or over a megaphone. I haven’t yet been able to verify whether what the officer told me is true, or what sort of exceptions there may be to the policy, but if true, this seems like a far more safe and sensible way of serving drug warrants than barreling into homes in the middle of the night. That is, presuming you believe we need to be serving drug warrants at all.

Though the Georgia bill may be flawed, and the Utah bill is better but not quite ideal, it is encouraging that we’re having this discussion at all. It means that at least a handful of lawmakers are beginning to recognize the inherent danger in allowing police to violently raid homes to serve warrants for relatively low-level crimes — and consensual, nonviolent crimes at that. That in itself is progress.