The Fourth Amendment in the Digital Era
The Fourth Amendment protects people from unreasonable searches and seizures, requiring that warrants for these activities be backed up by probable cause. But the proliferation of computers and electronic data has raised new questions. What is an unreasonable search and seizure of computer files?
We recently spoke with Orin Kerr of George Washington University Law School, who argues in a new paper that electronic searches and seizures should be limited by what he calls the "ongoing-seizure approach": Searches and seizures become unreasonable when the government uses data that extends beyond the limits of the warrant. The conversation has been edited for clarity and brevity.
In your paper, you repeatedly mention that Riley v. California was something of a game-changer when it came to electronic seizures. What was the Riley decision? And how did it affect your views on computer searches and the Fourth Amendment?
Riley v. California dealt with how the Fourth Amendment applies to searches incident to arrest. The traditional rule is that when somebody is arrested, the government can search everything on their person for evidence, with no limitations. The question in Riley was whether that rule applies when the item is a cell phone. And the Supreme Court said there's a different rule for cell phones because of the nature of computer searches: Computer and cell-phone searches are so different, so much information is stored there — and such personal information — that if the government wants to search a cell phone incident to arrest, they need a warrant. And the result is a computer-specific rule: one rule for physical searches, another rule for computer searches.
This doesn't really change my view of computer searches, because the Court adopted the approach that I've been saying they should adopt, so I was pleased to see that. It's the first Supreme Court decision on computer search and seizure, and it really points out an important dynamic, which is that computer searches are different in terms of how they're carried out than physical searches. So we need new rules on the traditional limits of the Fourth Amendment in this current environment.
Your paper advocates an "ongoing-seizure approach." Can you tell us about that?
Here's the basic idea: When the government executes a computer search, they usually go into the suspect's home, seize all of their computers, and then take them away for searching later. And they need to do that for practical reasons. It turns out it just takes too long — it can take weeks to search the target's computer — so they usually seize all and search later. And what that means is that the government has access to all of this "nonresponsive" information, information that doesn't relate to the warrant, that they can search at their leisure back in the government's lab.
My argument is that the government is allowed to seize all that nonresponsive information, but they're not allowed to use information that they find that's outside the scope of the warrant when they search through the electronic information.
That means that if the government gets a warrant for fraud records, they can go into a house, seize the computers, and search the computers for fraud records, but they can't use that search for fraud records as an excuse to look for everything else on the computer. They can't turn that into a general search. When they're back in the lab and they're searching the computer for weeks, they might come across information about other crimes or even just information that's embarrassing. I think that when the government tries to use that information in the ongoing seizure of the nonresponsive information, it becomes an unreasonable seizure as the Fourth Amendment prohibits.
Near the end of the paper, you mention that you're not suggesting that the data be destroyed afterwards — you're just saying that it shouldn't be used. What is the difference between having it not be used at all in the future and just destroying it?
I'm skeptical that there's a requirement of destruction, although you could have it. Clearly, if the item is destroyed, it can't be used, but it actually is tricky to figure out what it means to destroy data. Does it mean zeroing out the hard drive? What if there are other copies of the file? I think use is a clearer idea. We could say that disclosure is use, or we could say use as evidence is use.
Use is in some ways simpler a concept to follow, and also it doesn't have a time element. If there's a Fourth Amendment rule that the government has to destroy the nonresponsive record at some time, when do they have to do that? Is it a week? Is it a month? Is it a year?
What if the government needs the original computer to show that there was not exculpatory evidence on the nonresponsive files? If you're a defendant charged in court, you're going to say, "I want to see the full computer because I think all the evidence is showing that I didn't commit the crime." And so there are reasons, for trial integrity purposes, to keep the full computer, at least while the case is pending. After the case is over, it's a different story.
So my approach isn't necessarily rejecting destruction — I just don't think you need it in order to ensure that computers aren't searched in an unlimited way.
Could you connect this to some of the political debates that we've seen over the past couple of months over topics like the NSA?
The NSA debates are mostly over what a search or seizure is, not so much when a search or seizure is reasonable. For example, the Section 215 debates about collecting metadata are about whether non-content records held by the phone company are protected at all. If they're protected by the Fourth Amendment, then the program is very likely unconstitutional. The real debate is what is a search, not what is a reasonable search.
This paper, in contrast, is about what is a reasonable search. Everybody agrees that the contents of your hard drives or the contents of your cell phone are fully protected by the Fourth Amendment, whether in your home or in your pocket or even in the cloud.
There might be a similarity in that the big question is: What do you do with information that's not actually evidence of a crime or not actually incident to a terrorist attack? In all of these cases you've got so much data out there. Some of the data is responsive to the government's concerns, some of it is not. If the government necessarily gets lots of information in the hunt for the important information — they need to get the whole haystack to find the needle — the broad question is similar: What do you do with the data, once the government has found its needle or it turns out there is no needle?
Matthew Disler is a RealClearPolitics intern.