Episode 30: Do Judges Actually Read Search Warrants?

[00:00:00] Speaker A: Imagine a justice system built on rigorous evidence, not gut instincts or educated guesses about what works and what doesn't. More people could access the civil justice they deserve. The criminal justice system could be smaller, more effective, and more humane. The Access to Justice Lab here at Harvard Law School is producing that needed evidence. And this podcast is about the challenge of transforming law into an evidence based field. I'm your host, Jim Griner, and this is Proof Over Precedent. Welcome back to another episode of Proof Over Precedent, the Access to Justice Lab's podcast on efforts to try to transform law into an evidence based field. I have three terrific people with me today. I'm Jim Greiner, I should say, and faculty director of the Access to Justice Lab. And I have three terrific people today to discuss an interesting article that they co authored, recently published in the Harvard Law Review called Unwarranted Warrants, about an empirical analysis of judicial review and search and seizure. And so I'll let each of you introduce yourselves. Can you give us just a brief background on who you are and maybe a fun fact? And so going in the order that you appear on my screen. Miguel, could you lead us off? [00:01:18] Speaker B: Yeah, sure. I'm Miguel de Figueiredo. I am a professor of law at the University of Connecticut. And I, as a fun fact, I guess you wouldn't believe it now, by the way I look, but in high school, I actually went to the national championships for rowing. So that was a fun thing that I did and very much a part of my previous life. But I do work on a lot of empirical work on criminal justice, on judicial behavior, and also on law and development. [00:01:52] Speaker A: Actually, I think one thing common to all three of you is in the adage, you all three know how to count. And so this is extremely welcome, especially welcome here. So terrific. Brett, why don't you go next? [00:02:04] Speaker C: I'm Brett Hashimoto. I'm a professor here in the linguistics department at Brigham Young University. Most of my work revolves around empirical linguistics, focusing on legal linguistics stuff and second language acquisition and second language learning and teaching, that sort of thing. In terms of a fun fact about me, I've been through the desert on a horse with no name. [00:02:31] Speaker A: Really? Terrific. [00:02:32] Speaker C: Yeah. [00:02:33] Speaker A: And does the rest of the song also apply to you or just that one particular line? [00:02:36] Speaker C: Just. I was riding a horse in near Havasu Falls by the Grand Canyon, and I asked our guide, what's the name of the horse? And he said, no name. [00:02:46] Speaker A: No name. Terrific. So absolutely super. Dane, why don't you go next? [00:02:52] Speaker D: Sure. Dane Thorley. This Isn't my fun fact, but I was just looking at the date. I think today is my last day as an associate professor of law at BYU Law School. I think I'm graduating to professor of law tomorrow, which is fun, like Miguel. So I'm at the law school focused on empirical work, do work generally focused on courts and judicial behavior, with a specific emphasis on field experiments in rct. So I've been a big fan of the Access to Justice Lab and what you're doing for a long time. Basically, since. Not basically, since it started before it started. I think my fun fact, this is very stereotypically Utah, but my family, we're all into skiing and I'm particularly into backcountry snowboarding. So I have a set of split boards which split in half for the way up, and I use them as skis. And then I latch it together and ride it as a snowboard on the way down. That's what I like doing during the winters. [00:03:53] Speaker A: Terrific. So we have a rower, a horse rider, and a snowboarder, which sounds like the beginning of a bad joke, but also in common free tastes I never personally acquired. So my congratulations to all three of you for acquiring and pursuing those interests. Let's get straight to your article here. It talks about search warrants. It is one of the first empirical examinations that I have ever seen. And y' all cite to a couple of them, but certainly none of them occurred on the scale that y' all pursued in this article. And quite frankly, none of them were as close to as rigorous or comprehensive. At least none of the ones I've seen were as close to as rigorous and comprehensive as what y' all accomplished here. But I do wanna set the setting for people who may not be familiar with the way that search warrants work. And so, and, and what a warrant is when a. When a police officer might need one or might want one. Because sometimes police officers may have the option to proceed without one, or at least. And you know, why they would. Why they would pursue one. And then what turns on the idea, in particular, of having a judge take a look at a police officer's application or request for a warrant. So let's start with a. With a basic. With a basic thing a police officer is in the course of investigation, and they decide that they want to search a premises, search a car they want, or that they want to try to seize some object, which I believe was also something that could come within the scope of these warrants as well. They think they want to go look for a particular cell Phone or some, you know, particular. So that's what they think they want to do. And so why might they seek a warrant? I guess we better ask one of the legal types for this first. So, Miguel, why don't you tell us, why might someone. Why might a cop seek a warrant here? [00:05:43] Speaker D: Yeah. [00:05:43] Speaker B: So, first of all, thanks for the incredible praise for the paper and thanks for also inviting us to be on your podcast. It's a real, real treat to be here. So what triggers a warrant is basically we have the Fourth Amendment and essentially certain searches require probable cause. Right. And there's a huge sort of, you know, amount of back and forth and doctrine on what constitutes probable cause and what types of searches can be require a warrant. And this has changed over time since our founding. But basically, when there's a sort of. Without barring a number of exceptions, like, for instance, the officer's safety or the destruction of evidence, there's a need for law enforcement to have a warrant for certain, for a number of types of searches. So for instance, you mentioned, you know, search of a house, right? So if a house is considered sort of sacred, right? Like, that's the. That would be a huge intrusion on the privacy of the individual. Right. So in most cases, if the police want to do a search of a house, it requires a warrant. Like, again, barring a whole set of exceptions. But for instance, a really common one is to do a blood draw, right. If I'm going to. If I suspect someone of driving while intoxicated or. And for certain types of motor vehicle searches, there are also warrants that are required. And so as you mentioned, we're. I think we're not the first, but we're one of the first and certainly the first, to the best of my knowledge, in the last two or so decades to do any empirical work on warrants and really focus on search warrants. And so our goal is to kind of understand this process better, understand what the police are doing, what judges are doing. This is really focused on the judges and what the implications are for privacy, for public safety, and for the search powers of the state. [00:07:55] Speaker A: Terrific. And so just to clarify, is something like a request for a wiretap for folks who have heard of these sorts of things? Is that also covered, but that would also be covered by general search warrant doctrine? [00:08:06] Speaker B: Definitely, yeah. So there are different types of wiretaps, but yes, for the most, I'd be familiar with the TV series the Wire. Right. And so in that series, you see, you know, repeatedly the police going to a judge for a warrant. But I also want to say like that's there's a bit of sort of drugs, you know, guns, home searches. There's also something that's happening that I think affects nearly all of us, which is that nearly all of us have rely on third party providers, right? So we have a cell phone carrier, we have a storage provider, right. We all have these third party providers that, that are really important in our daily lives. We have email providers. And what's happening is if I want to, if I'm law enforcement and I want to do a search of Jim Griner's, let's say your cell phone, I don't even have to go to Jim, right? I can just go to Apple, right? Or I can just go to, you know, your cloud provider, whether it's Microsoft or Google or whomever, or if I want to search your email, I can do that and then that third party provider and then say yes to law enforcement without you even knowing it. And there are hundreds of thousands of these third party searches that are done on a monthly basis of Facebook or Meta, I should say of Apple, of Google. And that's all, you know, that's an invasion or maybe, or maybe it's not an invasion, but something that's a major invasion of our privacy and that's something that affects everybody. And it's not to say that, you know, there are other cases like drug warrants, no knock warrants and so forth that are really important too. But I just want to not, I just want to emphasize that the scope of this is really, really big. And so a lot of the warrants we see are of T Mobile and Verizon and such. And so I just want to emphasize that, not to forget that this has a really wide ranging implications. [00:09:57] Speaker D: Jim, if I can jump on and just sort of re emphasize something that Miguel said, often as legal empiricists, we start our articles by saying something like, oh, we have this novel data center where we're contributing to a lacuna. You know, we, we make these claims. [00:10:13] Speaker A: About filling a gap, right? [00:10:15] Speaker D: We're filling a gap. And as you brought this up and Miguel brought this up. But I want to reemphasize that there really is an incredible gap here. Not just because there's been so little work done, but so little work done in an area that's so important in a practical sense to people out in the world and important in a legal theory and jurisprudential sense. I mean, I think there's few topics that have been written on more in law reviews than search and seizure and the doctrines surrounding the development of the jurisprudence here. And it's just given how much airtime this takes up in legal scholarship and in court cases, it's really incredible that we have as little information on the process as we currently do. I think this speaks to broader themes that you tend to touch on in this podcast, which is do we really know what we're doing? And the answer is no, we don't. And almost ever, and especially in this scenario, we just know so little about what goes on here that any empirical contribution is a big one in this arena because it's important and because so few people have done it before and it was truly low hanging fruit. I mean, it turns out it was not low hanging in the sense that we, there were some struggles, it wasn't easy to get, but once, once we had it, there was so much new that we could add to the discourse with really basic empirical techniques. Our paper is not. There's a lot of data in it, but we don't do anything complicated. I think I haven't talked with Brett about this, but I know I've talked with Miguel about this. I think this is the simplest empirical piece I've ever done, but maybe the most impactful just because of how much of a gap actually existed here. And it made it really fun, sort of true feelings of discovery as we're looking at the results here. Like, wow, nobody knows this and now we know it and we get to tell people. And it was awesome. [00:12:16] Speaker A: Terrific, terrific. And I want to talk some more about that because among the many things that I liked about this paper, it demonstrates that critical and eye opening empirical scholarship, empirical legal scholarship can be done without obsessing on over causation, which is quite frankly my obsession. Right. That's why I do randomized field experiments. And it demonstrates that how much you can learn by doing just really outstanding work that looks and sees what's going on. But let's continue just to set the stage here. So we got this requirement in the fourth Amendment that the police can't take certain action, you know, such as, you know, ask demanding that a, you know, say anything from demanding that a cell phone carrier produce or an Internet service provider produce certain about somebody's activities to the ability to search a residence or a car or some someplace where a person has a reasonable expectation of privacy or. And this is actually where I wanted to head next just to get a clarification, they can't also kick down someone's door, say of an apartment. Legally they can't. They can't just kick down someone's door and say we get to look around in there without announcing that they're even coming. Without, you know, in other words, they can't by kicking down the door, I mean, by breaking down the door without saying, knocking on it and saying, we're the police department. And so this is where I wanted to head was what, you know, what is. And maybe, Dane, you could take this one. What is a no knock warrant? Because these have made. These have been the subject of significant scrutiny in the press. Not so much recently, but certainly a few years back. You know, with the name Breonna Taylor is seared in my consciousness and not just because a bunch of WNBA players put it on the backs of their shirts. Thank goodness they did that, by the way. But what is a no knock warrant, Dane? [00:14:07] Speaker D: Yeah, so the, the, there's, there's the complicated answer, which is it depends on where you are. The, the development of no knock warrant law is, is diverse. At the state level, there's some, there's some national level legal precedent that sort of bounds what states are allowed to do. But there's a good deal of latitude in what states can do in regard to no knock warrant, which is effectively, you know, the purpose of a warrant is to gather evidence of a suspected crime. And if the police officers feel that they're in a situation where the normal course of enacting a search warrant would result in the loss of evidence. So for example, I think the quintessential example is so you knock on the door and you say police open up and you grab your baggie and you scramble to the toilet and try to flush down the bag. There are reasons why police officers might be worried about making their presence known before enacting a warrant. And so there's been various strategies that they've developed. No knock warrants are one of these where. And also police safety is something we really have to consider. You know, luckily I don't think this happens with the vast, vast majority of warrants. But if you're investigating high crimes, violent crimes, there's potential for violence against the police officers. We want to make sure to protect the police. Officers announcing themselves may actually put them in more danger, unnecessary danger in some circumstances. So no knock is effectively where a police officer can initiate a search in a house. It's hard to think of, I suppose we can think of situations where a no knock type activity occurs outside of a home, but they, in some states they have to announce but wait 10 seconds. In other states they don't have to announce at all. They just come right through and enact the beginning of the search, which in the case of a no knock warrant, almost always includes an arrest or at least attainment of an individual. Again, to secure the premises so that the evidence isn't going to be removed or lost, and also to protect the safety of the police officers. Now, this has, even though this doesn't constitute a very large percentage of warrants, this has constituted a very large percentage of the discourse around warrants you mentioned. Breonna Taylor, Amir Locke is another example sort of tragic outcomes, understandably. I mean, if you're unfamiliar with what a no knock warrant is, and I just gave you that definition, I think any person could very quickly think up scenarios where this might go wrong. And it does go wrong. It goes wrong for the police officers. It often goes wrong for the suspects. In the case of Breonna Taylor and Amir Locke, they may have gone to the wrong place. They may have mistaken someone waking up from a nap on the couch as someone pulling out a gun. And people die from this. Not always, thank goodness, but certainly more often than they should. And even if it doesn't result in a death, it often results in really traumatic outcomes and experiences for the individuals being searched and sometimes even the police officers. So this has really led the discourse on warrants. Do we need warrants as much as we use them? Who is reviewing these warrants? How did these mistakes get made? The scrutiny level went up a lot during the Black Lives Matter movement. Again, Breonna Taylor, Amir Locke, among other examples, sort of front and center discussions about what are warrants and how could these mistakes have been made? Aren't judges reading these warrants, aren't they going to recognize that there's mistakes in these warrants? I think this is a good thing. The discourse moved, or at least began to include judges and not just police officers. Because when it comes to enacting warrants, this is not supposed to be purely an exercise in law enforcement. This is supposed to be an exercise in law enforcement under the supervision of, you know, a detached and neutral magistrate. They're supposed to put a check on the ability for law enforcement to do their job by. By telling them no. This is not the type of circumstances that the probable cost standard hasn't been met here, or there's an inconsistency in your warrant. This doesn't make sense, or that's not a real address, or you say someone lives here, but then you're searching this other premise. You haven't provided a justification for that. That's what we expect the role of the judge to be in this process. And so when these traumatic outcomes occur, I think judges, in addition to police officers, need to be under the microscope. And that's. Again, Miguel mentioned this, but our paper is really focused on the judges. [00:18:41] Speaker A: Let's get to that, actually, just again, to set the setting for people unfamiliar. Let me just see if I can review my first year criminal law and you'll tell me which parts of this I've gotten wrong since it's only been 30 years. So basically the idea here, because you mentioned the judicial review and the independent magistrate, basically the idea here is that the cop wants to go search a place, search a residence. The cop wants to get records from a Internet service provider, whatever it is. And we say that the cop can't just do that on their own. They can't just issue a warrant on their own either. The cop just can't go to the warrant. Furthermore, the cop can't just go to a supervisor within the police department and say, you know, hey, boss, would you please just, you know, take a look at, you know, what I've gathered so far and decide that it's okay for me to go do this. We actually require the cop to go to an independent branch of government, another branch of government, namely the judiciary, and, and say, can you please issue this piece of paper called a warrant, which then is, as you say, as you said, Dana, you alluded to Miguel, the. The judge could be called, a magistrate is supposed to take a look at the stuff. The evidence that the cop has, has put into an application and said, yeah, that's sufficient. That looks like there's probable cause to do the thing that you want to do, to search the premises that you've identified or to pull those records. But it's not going to give you the opportunity, the justification to search everything on that block or to get, you know, 10,000 people's records. Right? If you just. If what you're suggesting is one person. So it has to be. There has to be evidence, and then there has to be a relationship between the evidence and the thing that is sought. And we require an independent review by somebody who works for a different branch of government, namely the judiciary, as if I recall correctly. And of course, you all are going to tell me more about this. But this is not a trial type setting where we have lawyers involved. We don't have any lawyers involved. This is not like there's an adversarial setting where we notify somebody whose records may be searched. Miguel, you've already admitted, you've already alluded to this, that actually the person whose records are being put pulled may not know that they're being pulled. This is just an application from the cop directly to the judiciary saying, is it okay for me to go do this thing? And we need to have, and the whole point is that we, we want to have an independent look from somebody that doesn't work within the, within the investigative arm of the government to say, yes, there is probably some intrusion here on somebody's privacy interests or on their, on their, maybe their dignity interests. We needed to have an independent look before we authorize that intrusion in the interests of law enforcement. So tell me which part of that any of you did I get wrong? Because again, it's only been 30 years. [00:21:20] Speaker B: Yeah, no problem. That's exactly right. I mean, I guess I'll put a couple of footnotes into what you said, but what you said is basically right. There's a couple of subtle possibilities that could deviate from that, what you said. So what you're talking about first is I think, what's called the particularity requirement, right? Which is that there has to be symmetry between the item being searched, right. The evidence for that item, and then how it relates to a particular crime in the individual circumstance. Right. The origin, by the way of the Fourth Amendment is, and where we depart from our cousins across the pond, so to speak, the uk is that they used to have what you were saying, they would just go by neighborhood and just search houses. And so we wanted to have a real safeguard for the privacy. And the judge is supposed to be something that stands, you know, as an independent authority that could evaluate that trade off between sort of the investigatory powers of the state and public safety on the one hand, and the privacy rights of individuals on the other. So it starts with the officer, you're right, will put in some form of an affidavit, a sworn statement. There's one thing that can happen. In some cases they may go to their supervisor, and some, in most cases they do not. It depends on the department. And then there's one step where they can also seek a prosecutor's or DA's review. And they. That's basically in Utah, that's up to the officer, the law enforcement officer. So in some cases they do, some cases they don't. Then it. And then it goes to the judge, and then the judge makes a decision. [00:22:56] Speaker A: Can I pause here one second, Miguel? If they went to the prosecutor just as a point that for everyone that didn't know this that's actually still within the executive branch. So we haven't gone outside the executive. So both the cops and the prosecutors are considered to be part of the executive, part of the enforcement. [00:23:13] Speaker B: Absolutely right. [00:23:14] Speaker A: Yeah. Right, good. [00:23:15] Speaker B: So that's right. Part of the executive, of the enforcement arm of the state. [00:23:18] Speaker A: It's only when they go to a judge or the judiciary that they've gone to another branch of government which is considered to be the structural safeguard. [00:23:26] Speaker B: That's exactly right. And that is the one safeguard, really, that exists. Okay. And. And you're right, it's not an adversarial process. It's basically a process between the law enforcement officer, potentially a prosecutor, and then the judge. Right. There's no criminal defense lawyers involved and there's actually no audit that's done. There's no. All of the appeals are ex post. Right. So the main. If there's an illegal search that's approved, it's almost always done after the search. Right. Because there's something called the exclusionary rule, which means that evidence which is seized could then be excluded from the record. But there has to be a challenge that happens from the defense lawyer or the defense or the individual defendant. And so what's really important about this process is that you got to get it right up front because there's not a lot of safeguards in place because of what you're saying, Jim. [00:24:22] Speaker D: Right. [00:24:22] Speaker B: There's no, it's not contested. There's an adversarial process and there's only. There's no form of appeal or anything like that. Right. It just is law enforcement officer, maybe a prosecutor, and then the judge. [00:24:34] Speaker A: So basically, from an access to justice point of view, pure access to justice point of view, we've got a judge just basically acting as the judge with no real adversarial process involved here. Right. We got one party saying, hey, this is what the state of the world is and this is what the evidence is. And we depend on the judge to look and get everything right without an adversarial process. And then, as you say, Miguel, there is no traditional audit that comes in after, because you'll have tens of thousands of warrants in your data set here. You know, some. It's not like somebody's coming in. There's a. The state could do this. It just doesn't. Somebody's coming in after the fact and saying, oh, you know, we took a random sample of, you know, 500 of them and reviewed them or something like that, the way they do, for example, in Medicare, Medicaid claims Right. The government does this regularly, but it doesn't do this for governments don't do this for warrants. And in addition, the, the legality of the judge's decision is as, as you say, Miguel only adjudicated ex post. Right. In other words, when there's a subsequent prosecution and the evidence that was seized in as a result of the warrant, the prosecutor attempts to use that evidence and the defendant moves to exclude the evidence. And then you will litigate the legality of the warrant. But among other things, the warrant either legally or implicitly is entitled to a presumption of correctness. Right. And certainly if cops are acting pursuant to what appears to be a valid warrant, they also that the evidence is also more likely to come in for that reason. So still in my first year criminal law summary is all this COPA said. [00:26:03] Speaker B: That'S all looking good. I'll just add one thing which is interestingly, in Kentucky where the Aprona Taylor case happened, they actually did implement warrant specialty courts and they also implemented, I think, randomized audits of them. So it's exactly what you're saying it is very possible. That's an incredible exception, right? Overwhelmingly. It's very rare that we have that and we tee that up as a possible reform possibility. There also has been a growth of no knock warrant bans in a number of jurisdictions as well. So there is a possibility for reform of this process. We'll get to that probably soon, later. But I want to say that you're absolutely right. Yeah. In terms of your, your summary of. [00:26:42] Speaker D: The process and just one more small detail sort of to not connect the lack of a feedback loop because there is, there's, there's disconnect. But one of the reasons there's a disconnect is because the judges who are reviewing, approving, mostly approving the warrants aren't necessarily and are usually not the judges who are are tasked with overseeing an actual prosecution. And so there's not even informal feedback loops where a judge might say, whoa, I granted this warrant. Now the defense is seeking an exclusion of this evidence. Oops, I made a mistake. There's no real way, even in an informal sense, that a judge might be able to get feedback information. [00:27:23] Speaker A: So from a cognitive point of view, we might actually think that's a feature to the system so that the judge is not judging their own warrant subsequently, because that means the subsequent trial judge who's trying the case might feel more independent about making a decision that says this warrant shouldn't have been granted. On the other hand, as you say, there's no information going back to the, to the judges that make the original decision about whether to grant the warrant. They don't see whether in fact, they granted a warrant that was where the evidence was subsequently excluded because of something they should have picked up. And they don't get to see that and try to correct their decisions going forward. [00:27:56] Speaker D: That's right, yeah. [00:27:57] Speaker A: Super. Okay. So basically then, what we've got here is that a lot depends on this independent judicial look at the submission from the police officer, which is typically, I believe, Miguel, you used the word affidavits. That means it's typically sworn under oath under penalty of perjury. And a lot depends on it. And so the question is, how close of a look are judges conducting and how are they doing this day to day? How fast do they look at them? Or is this a rubber stamp? Is the judiciary basically abdicating its duty to take an independent look? And of course, we can't observe that directly. You know, it's not even possible to do that because you can't climb in someone's head and say, you know, how well are you looking at things? But among other things, what we could do is what y' all did here. And so, Brett, why don't you actually summarize, just big picture. What did you, what did you do here? Right. You, you got a whole bunch of warrants from a computer system in Utah, scraped them a whole bunch of warrants. And you also scrape some metadata about judicial decision making about those warrants. And so just big picture, give us an overview. What sorts of things can we inquire about with those two big pieces of information? [00:29:16] Speaker C: Well, in terms of metadata, one of the key pieces of metadata that features in our data set is that was sort of interesting that this piece was maintained is how long it takes for judges to actually approve the warrants once they have received them and opened them. [00:29:38] Speaker A: Or reject. Approve or adjudicate. [00:29:41] Speaker C: Yeah, make a decision. Yeah, one way or the other. And so in terms of the metadata, we have lots of metadata about who the judges were, who the submitting officers were, those sorts of things. Miguel Dane, feel free to fill in here on the data side. [00:30:01] Speaker D: Well, sure. So just a bit on this timing, metadata. I think that's really what makes this data set so special. So utilize this E warrant system. They're not unique in having E warrants. This is effectively the same sort of warrant system we've always had, except now it's done electronically, where instead of showing up at a judge's doorstep at 2am with a piece of paper explaining why they want the warrant. So their affidavit, they can send it in a system, and the judge who's on call to review gets pinged, and they can open up their computer or their phone and look at them electronically. When Utah started this, there was some metadata on when it was sent and when the decision was made, which is interesting. But over time, they improved the system, and maybe inadvertently, it's not clear to us that this was a design feature. Certainly, I think. Yeah, well, we can get more into sort of judicial responses to this and what they think about it. But now, over time, they increase the number of sort of timestamps to include the moment the judge opens the affidavit for the first time. And so we have this when they first opened it and when they made the decision, which is the maximum amount of time they could have possibly spent reviewing the warrant. Now, I think it could be less. Right. Just because they don't have to have spent all the time between opening it and making a decision, you know, sort of braining on whether the warrant meets the standard. But we assume that that's a pretty accurate measure and sort of conservatively say that's the amount of time they spent reviewing the warrant. This is as close as you can get to getting into a judge's head, I think, in this arena. And I think I'll throw it back to Brett, he can talk about what we do with this. This information vis a vis the word counts for the warrants. But this Miguel was the one that first noticed this. The Salt Lake Tribune had a really great article. They looked at a month's worth of warrants and sort of highlighted the existence of this data. And I can still remember the email. Miguel's like, hey, I want to talk with you about Utah. And I knew he didn't want to talk about the snow part of Utah. It had to have been something data related. But we got on the phone, he's like, there's this data set in Utah, or there's this data in Utah. You could see how long a judge spends reviewing a warrant. I was like, no way. There's no way that data exists, and there's no way that it's publicly available. And it turns out it does exist, and it was publicly available. It doesn't mean it was easy to get all the time, but just a really incredible piece of data. And we could do some really cool things in combination with the text of the warrants. And I'll throw it back to Brett for that. [00:32:41] Speaker A: And Actually, Brett, just before you do that, could you get, or maybe this is actually the same question. Could you give us an idea? So these warrants, they probably are like most government forms in the sense that there's probably some stuff there that's, you know, kind of obvious, kind of repetitive, and the judges don't actually need to look at that much. But then there's the operative part where the judge, where the cop describes the evidence that has led them to or what they've seen, the evidence that the, of their observations that leads them to the, to the warrant request and also describes what it is that they want to either search or seize. So can you give us an idea? Because that's probably the part that judges need to focus on. What does that part, that sort of guts of the warrant look like? [00:33:20] Speaker C: So, yeah, as part of our data set we have for the process of a warrant there are sort of three important documents. And those documents have some parts to them. But the three documents are the affidavit, sort of the application for, for the warrant and then the warrant itself. And then the third piece is the return which is the officer reports what was actually found and seized as part of their search. And you're right, these are all generated through the officer will fill out some form and digitally. And then all of that form language actually makes it into the affidavit, it makes it into all of the documents. And so there's quite a bit of just boilerplate language that exists in these documents that sort of frame what the different parts or the different questions that the officers are responding to when they're applying for a warrant. The something really critical. I think data that we keyed in on as part of most of our analyses were there's this section of the affidavits where officers are tasked with describing what the probable cause is and what their qualifications are as an officer is to make the judgment about the quality of that probable cause. And that's really what we keyed into as what the judges would probably be evaluating. We had to go through several types of cleaning up processes in order to boil down some of that, including utilizing some large language models to help us tidy up some of the boilerplate stuff. We did some manual cleaning up and then we tested the large language models and trained it on that manual effort and then discovered that the large language models were pretty good at tidying up the data set for us. So that's what we ended up working off of, was off this sort of cleaned up and subdivided data set based on the classifications and tidying up done by the large language model. [00:35:42] Speaker A: Super. So you've identified two particular pieces of the warrant application that a judge might want to focus on one of them. I did not know this phrase before and I think I'll try to work it into my day to day lexicon. The hero statement, which is basically where the cop brags about all the experience that the cop has and tries to persuade the judge that they have excellent judgment based on their experience, et cetera. I'm a hero type thing. I'm just going to need to go let my kids know that I actually am here. And then the actual description of the specific facts. So on such and such a date I spoke to X person. That person told me that this other, that the, the suspect, you know, entered their own residence with, you know, 10 pounds of cocaine. And we want to, we're asking for a warrant to search this particular address which is the residence of that person as identified by my, by my witness or by my informant, you know, that in order to try to seize, you know, a bag of 10 pound cocaine or something like that and maybe search for some other drug, paraffin A or something like that. Or in the alternative, you know, on such and such a date, I was talking via electronic means to a particular suspect who was. And that person told me how they had downloaded all of this illegal content, you know, child pornography, whatever it might be, from a particular website. And so on the basis of that I now want, please judge, authorize me to go to the Internet service provider and ask and demand the download activity from this individual for this particular date range or something like that. That. Right, that's the sort of guts, you know, along with the hero statement. How long are these things typically? You know, there's going to be a very, very wide range. Brett, back to you, how long are they? You know, but again to sort of give an idea here. Are these things typically, you know, 5,000 words? Are they typically 800 words? What's the sort of. What's the sort of, you know, give us an idea. [00:37:35] Speaker C: That's a good question. I don't know that the average. Do either of you have the hand what the average length is? [00:37:40] Speaker D: I can save you here, Brett. I've got the, I've got the paper up. So it depends on what type of warrant it is. I'll also note and this is. We may get into the key limitations of our study in a bitch and but one that bears mentioning right off the bat is we have a lot of data on all of the warrants. Submitted all of the warrant affidavits submitted in Utah, but we only have the PDF texts of the warrants that were approved. Now, as we'll probably get into, almost all the warrants are approved, so that means we have almost all the data. But I would say that the warrants that were rejected are particularly interesting. This was frustrating for us as we worked on trying to get the data. Understandably, there's no reason why law enforcement would keep, or the courts would keep a warrant affidavit that wasn't approved. There's going to be no legal relevance. [00:38:29] Speaker B: To it at all. [00:38:30] Speaker C: Do you want to just speak to what the percentage of approval was? [00:38:32] Speaker A: Yeah, let's actually get that out on the table. So how, what fraction were approved? That's about 98. [00:38:41] Speaker B: 90, 98 and 93% are approved on the first submission. Right. [00:38:49] Speaker A: And then some were withdrawn. [00:38:50] Speaker B: Right. [00:38:51] Speaker A: So it's not like the judge is rejecting those and then they're later resubmitted and they become. They go, then they're. Then they're later approved. [00:38:57] Speaker D: Right. [00:38:57] Speaker A: So they are a small fraction of these of these are rejected. [00:39:02] Speaker D: Right. [00:39:02] Speaker B: Right. [00:39:04] Speaker A: PDF text are the ones that were approved, but that's almost all of them. And that's actually sort of alarm bell number one is that the. Is the approval rate itself is sort of astonishing, astonishingly high, especially the first submission approval rate, which again, I guess, Miguel, is the point you're making that sort of 93% between 93 and 94 are approved on the first submission. [00:39:26] Speaker D: Yeah. [00:39:26] Speaker A: Right. [00:39:26] Speaker D: And we can maybe get into this later because I think there's some genuinely good questions as to what. Well, what percent should be approved? What percent do we expect to be approved? I think we can get into that once we have a little more context from time. So I think our main results combine the amount of time spent and the overall approval rate. But to answer the question you asked a few minutes ago about the length, it varies depending on what type of warrant we're looking at and it also, there's a good deal of variance within just your standard warrant, but the median length is about 1000 words is sort of your standard warrant. There's some very, very long, you know, upwards of 20,000 word warrants. There's some affidavit. [00:40:07] Speaker B: Right. Not warrant. [00:40:08] Speaker D: Oh, yes, thank you, Miguel. And there's some exceedingly short warrants. We don't get down to zero. You know, you can't have an affidavit at zero, but, you know, one page, one page of text, three pages of text, translate Pages. [00:40:23] Speaker A: When you mention pages to number of words here. So you mentioned that. What was the average again that you mentioned on? Or median. Excuse me? [00:40:30] Speaker D: Median is just under a thousand words. And that would translate. That's right. [00:40:36] Speaker B: Or double spaced pages in word. [00:40:38] Speaker C: Yeah, double spaced. [00:40:39] Speaker B: It's about 250 to 300 words. [00:40:41] Speaker D: Yeah, that's right. And like Brett mentioned, portions of these, and I think we can get into this in a minute. Portions of the affidavit are arguably unnecessary. Maybe you don't even need to read them. We can again get into that in a little bit, but we won't be able to show this to the listeners. But if they really want to see what we're talking about, they can track down the paper. We have these figures that plot the length of the warrant. So again, these are just for approved warrants because we don't have PDF texts of the warrants that were the 2 to 3% of warrants that were withdrawn or rejected. Plots the length of the warrant in words against the amount of time spent on that warrant. And then draw some curves in different colors under different assumptions of reading speeds. And effectively what these plots show is any dots that fall below a given line. So let's say we have a very. I'm not sure if it's conservative or liberal, but let's. We assume judges are reading at 650 words per minute. That would be very, very fast. Extremely fast. Any dots that appear below that line are warrants that we know couldn't have been read in full at that reading speed. And at all variations of reading speeds that we plot on these, on these figures, there's a lot of dots that are falling below the lines. [00:42:02] Speaker A: So the average. The average. The median length. Excuse me, I keep saying the average. The median length. The problem with the average in this setting is that there are some that are these extreme outliers. [00:42:09] Speaker D: Right. [00:42:09] Speaker A: Drag the time that also. [00:42:11] Speaker D: Time. Time especially. [00:42:12] Speaker A: Yeah, exactly. And timing is always a problem. Is always the problem in, in any kind of legal data set involving case activity involving the judiciary is that you have these extreme outliers. So we speak, you know, talking about medians. You know, it's basically, you know, around sort of 3ish to 4ish pages of double space text, around, you know, sort of 800 to 1,000 words is sort of the median thing that we're talking about here. And then let's get the median figure out on the table here. What's the median time figure for judges? And this again is, to be specific, this is the time from when the judge, magistrate opened the affidavit or open to the application and then when they submitted the decision. Right. And is that correct? That's right. [00:43:02] Speaker B: And just to be precise, they open it, they have to read it, digest it. They have to also, if they want to make any changes, they have to do that and they have to then generate the appropriate text. Now most of this stuff can be auto generated, right. But again that all of that is happening from when they open it till the timestamp of when a decision is made. [00:43:23] Speaker A: And so again, this is one of the things that struck me. Is it actually they couldn't be reading that entire time. That's actually possible. So this is in some maximum time. [00:43:33] Speaker B: That it's very conservative estimate. [00:43:35] Speaker A: There's no way they could have actually been reading and thinking that entire time because they've got some other stuff they have to do. It may not take a big, it's not a lot of time, but they have to do some pointing and clicking in order to submit the decision in this case because again, it's the limit that you mentioned earlier, Dane, about only having access to the approved applications. They have to submit the approval. Right. And so there is some pointing in clicking. This is the maximum amount of time. So again, let's just get the basics out on the table. You know, what is the median amount of time here for consideration of these? You know, on median around three to four pages of text. [00:44:13] Speaker D: So it's. Yeah, the if because we have the metadata for all the warrants we can. Again, look, we don't need just for time, we don't need the PDFs. It's three minutes is the median review time for all warrants in 2 minutes and 50 seconds. Seconds is the median review time for approved warrants, which is quite fast. Again, these warrants, they're not super long. Like these aren't. These aren't case, you know, these aren't case filings that. [00:44:42] Speaker A: These are not appellate opinions that you read law school. Right, yeah, got it. Go on. [00:44:47] Speaker D: Yeah. And they tend not to be exceedingly complicated either. But 3 minutes, 2 minutes, 50 seconds is very fast as a median time. Even accounting for the fact that the median warrant is only three or four pages. [00:45:01] Speaker A: And then just. I'm sorry, go ahead, Miguel. [00:45:03] Speaker B: Yeah, and when we spoke to judges, we also, you know, kind of spoke to judges about this. They were sort of divided as to whether three minutes was enough time or not for the median warrant, where there was nearly universal agreement. And universal shock, I will say, is that for 10% of the distribution, which is a pretty weighty sum, maybe some of the distribution. They're deciding them in one minute or less. [00:45:28] Speaker A: That's actually where it was headed. Right. Because. Let's see, let's get an idea here. The total data set, roughly. Brett, tell us this if you can. Roughly doesn't need to be exact, but are we talking a data set of about, you know, 2,000 cases or closer to something like 40,000 cases? [00:45:42] Speaker C: Yeah. The data set in total was 33,000 warrants, more than the. The cases that were approved were about 30,000 of those. [00:45:55] Speaker A: 30,000 of those. Right. And so, Miguel, now back to you. If you're talking about 10, 10% of those, that means that in about 3,000 of the cases in which the warrant was approved, the judge spent. And I believe the. The median there, or the. Or the average, or. No, it was the 10%. So the 10th, the 0.1 decile. Right. It was. In other words, 10% of the cases. Right. [00:46:17] Speaker B: Was one out of every 10 was decided in one minute or less. [00:46:21] Speaker D: Yeah. [00:46:21] Speaker A: So 3,000 of these warrants are decided in less than or fully. And so again, not all this time spent reading fully decided in less than a minute. [00:46:30] Speaker D: Exactly. [00:46:31] Speaker A: Yeah. [00:46:32] Speaker B: And when we spoke to judges, there was near universal concern about that. And just to. If you don't mind, I'll just tell you a little bit about robustness. So we try to address and anticipate all of the potential arguments against this. And so we first, as Jane said, benchmarked against word counts. Most people sort of in our, let's say, cohort of sorts, like most lawyers, judges, they're reading somewhere between 200 and 300 words a minute. Right. That's about a page a minute. That's about roughly what most experts, when they do expert witness testimony, this is anecdotal, but what they bill out at. And so we get, we make very conservative assumptions. 300, 600 words a minute. So that's the first thing we do. And you still see tons of warrants that are going on red. Secondly, we drop going on red. [00:47:22] Speaker A: Meaning the assumption is that if you assumed that that was the rate, there's not enough time. Right. That when you say they're going on red, that's what it means. Is that what you meant earlier, Dane, about being below the ormagala? Forget what being below a line. You basically draw a line on your graph and you say, okay, if we assume that the reading rate is 650 words a minute, which strikes me as bonkers. Right. That just Strikes me as way too fast. Right. Or something like 300 words. These are in the paper, these various lines and dots. Dots represent each warrant. There are always a substantial number of dots on kind of the wrong side of the line. [00:47:56] Speaker B: Right, Exactly. So if you decide on a 500 word warrant and in one minute, right. Then. And you're reading at, you know, 250 words a minute. [00:48:10] Speaker D: Right. [00:48:10] Speaker B: Half of that warrant has gone unread. Right. [00:48:14] Speaker A: One second. Let's get, let's hear from Brett real quick. Go ahead, Brett. [00:48:16] Speaker C: Can I just make a note? I don't know if the typical person knows what a normal reading speed is. I mean, but I mean, there's been lots and lots of study about, about what typical reading speed is, what fast reading speed looks like, what. What skimming looks like. And, and 250 words is a typical reading speed for fiction. For an adult reader, 650 words per minute would be reading. Flying through a document, skimming a document might be in the neighborhood of 450 words a minute, something like that. So 650 words would be beyond what we would call scanning. I'm looking for very specific information in the document and I know it's in there somewhere and I just have to find it. That would be beyond. 650 words would be faster than that. [00:49:05] Speaker A: But even if you're scanning, eventually you have to get to a part where you read, right? [00:49:09] Speaker C: Yeah. [00:49:11] Speaker A: You're scanning. That means you're looking. [00:49:13] Speaker C: So skimming. Yes, skimming and scanning. We're talking like 500 words per minute would be the upper limit of what we suspect would be typically possible for somebody. [00:49:23] Speaker A: So I mean, if I'm a judge, I can imagine scanning to get to a part where I really need to focus and then I've got to read. Right. Then I've got. And the point is, because there are always all these dots, meaning these warrant applications and decisions that are on the wrong side of the line, even with the assumption that you're only scanning the entire time and you're never actually stopping and reading. It's just too damn fast. [00:49:44] Speaker B: Yeah. And then we even do further robustness checks, which is we drop DUI warrants, which tend to be the most formulaic and simple. And then we still see tons going unread. We drop all of the boilerplate language, still see tons under that are going unread in their entirety. We then only look at the heterogeneity across crime categories. You would expect certain types of crimes are more complex than others, and we don't See, you know, huge differences there and lots are going on red. And then finally we say, okay, if they just look at the facts. Right, which is just the bare minimum. [00:50:19] Speaker D: Right. [00:50:19] Speaker B: And we use that, the fact word count. We still see huge portions that are going on read. Assuming these very conservative reading speeds. So again, this, I, Dane, and I think this is the first paper that we have ever done that is not causal in the sense of being experimental or quasi experimental. But what we're doing here is we're getting a ton of descriptive pieces of evidence and putting it together to kind of create a picture of what we think is going on. And we think it's pretty convincing that there's, you know, a large section, significant section of these warrants. Like if you think. I think 10% is pretty large that are going unread. And. And then what we do is we add to that and we have. We also, by the way, drop, you know, we only look at the first instance, right. [00:51:06] Speaker D: We. [00:51:06] Speaker B: And then we drop ones also that are very similar to other warrants if there's a first. [00:51:11] Speaker A: First instance of what? [00:51:13] Speaker B: First instance of a warrant given. And then what we do, I mean, really, Brett, deserves credit to heroic work. We do pairwise comparisons of every permutation of the warrants. And so you can imagine, for instance, in some sort of conspiracy or some sort of accomplice type crime, that they're just changing the name. [00:51:32] Speaker D: Right. [00:51:33] Speaker B: And so we then say, okay, if there's very little word similarity, we drop those and then see if there's. If there's still. If the results hold up and they do. Okay. And so we were really generous with. [00:51:46] Speaker C: The way that we conducted that analysis too. There's lots of ways to do document similarity. And we were really generous and tried to be really generous in the way that we did that as well towards you. [00:51:55] Speaker A: You declared documents similar in a permissive way. [00:52:00] Speaker C: Exactly. [00:52:00] Speaker A: Easy for documents to get classified as similar as opposed to hard for documents. [00:52:04] Speaker C: That's correct. [00:52:05] Speaker B: Yeah. [00:52:05] Speaker D: Yeah. [00:52:06] Speaker B: And then we do a qualitative analysis where you would really take a deep dive on a random sample and we find a range of errors. And I don't want, you know, there's stuff like grammar, which you can for now. [00:52:19] Speaker A: Right. You know, forget the grammar. Right. I mean, you know, I get it, I get it, I get it. Right. [00:52:24] Speaker B: But let's, let's push that for just a second on the grammar, which is that remember that this is a communication that's coming from the state to an individual, and we want that individual to have an understanding of what's going on, why they're being searched, what's being searched, and it's, it's okay if there's a typo. I, you know, maybe we don't care about that. But, but what I am saying is, what I care about is when it's unintelligible, that's the problem. [00:52:48] Speaker A: Right, right. [00:52:49] Speaker B: And that's what we're. And we see some of that, like, completely unintelligible stuff that I'm, we're, you know, Dane, and we're doing a really deep dive on this and we can't understand what's going on. And we're lawyers. Right. We're law professors, and that's problematic. Right. And then, and then again, there's the more egregious stuff, which is sort of, we think there are also illegal warrants and violations of probable cause. And so there's this range. But I also want to say that like, you know, the Angenette Young case or some of these cases where there's addresses that are, you know, that are wrong and then people's, you know, doors are being battle rammed down. There's a sloppiness that I think is important that we want to be wary of. Okay. And that, and that is something that I'm pushing back on a little bit. I'm not saying, you know, a typo here and there is a problem, but I'm saying when it reaches the point of unintelligibility, that's, that's problematic because this is a communication that's coming from the state and it invokes your privacy in really important ways. Yeah, sorry, I didn't mean to. [00:53:45] Speaker A: No, no, no, no, no, I get it. I get it. But when people think, oh, I said is as opposed to was. Right, that's not what you're talking about. What you're talking about is, is that the problems approach incoherence around unintelligibility. Right. Which is, which is a problem both in terms of judicial process right after the fact, when you're litigating exclusionary role motion, an exclusion motion or something like that. It's also probably something we care about here at the Access to Justice Lab in terms of treating the populace of your country with dignity and respect and the communication of that and persuading people to follow the rule of law. You know, the, the, the judiciary has neither, you know, the, the sword nor the, nor the purse, but only persuasion. Right. If you're not, if this is, if this is your persuasive, this is your attempt at Persuasion. It's not. It's not all that impressive, you know, on some. On some of these areas, we're going to run out of time. But let me just say, because this. I think we've gotten to the. To the guts of the paper, and I think folks that want to. Want to get some of the details, which are, again, a commend, y' all for the detailed way in which you've looked at things in the various alternative ways, some of which you just alluded to, Miguel, that y' all have looked at the data set. Let me just ask each of you to go around and say, because there's so much in this paper, it's so rich in terms of what you did. What's one other thing that we haven't kind of touched on or that we've only touched on, but just. But just haven't given enough sort of airtime to. What's one other thing? I'm getting this from each of you, you know, in a minute and a half. What's one other thing that you think is worth mentioning? [00:55:13] Speaker D: Yeah, so I'm gonna. I'm gonna throw up two. And if Miguel doesn't take the other one, then Brett can take. So the first one and the one I'll skip over quickly is differences in how the judges approach this. [00:55:24] Speaker A: About the judges. Right. Y' all gathered some data on judges. Yeah. [00:55:27] Speaker D: Yeah. So I'll. I'll throw that one out there and either Miguel or. Or Brett can grab it if they want the other one. And I'll be really quick on this one because I. This is that we don't make any causal claims. The most complex thing we do, outside of using the LLMs to analyze the data, is run some regressions. And this was simply because we wanted a bit more explanation for, like, what sort of warrants seem to be going fast. Like, what sort of circumstances seem to be the circumstances that have fast warrant review. Again, we don't know if this is what's causing fast warrant review, but a little bit information would be interesting. Something that we, at least I came in with the presupposition that judges who were previously in law enforcement as prosecutors, would approve warrants at higher rates and at higher speeds. And I, conversely, assume that judges who were previously public defenders would approve warrants at lower rates and take longer to review them. And again, it's just a regression. You know, we can't read too much of it. It's certainly not causally, but even correlationally, we may not have our model quite right here. That's not how it panned out. We found that the prosecutors took more time, defense attorneys took less time, and prosecutors were a bit more strict in the gatekeeping. We could talk about that all day. So I'll just throw that out as an interesting mystery that is unsolved in our data, and then we can go to Miguel and Brett. [00:56:51] Speaker B: Okay, so I'm going to cheat and put two really quickly out there. One was what Jane pointed to was the judge level heterogeneity. Although the approval rates are really high, when you break it down by judge, there's still huge variation. So there was an outlier around 70%, but really 80% is where we started seeing kind of most of the lower end, the approval rates for judges all the way up to 100%. So the fact that there's this massive. What we think is a big variation in their approval rates across judges. Right. Shows I think, again, we're looking at the extent to which their caseloads are similar, whether they're sort of what we call quad randomly assigned. But we didn't expect to see that much variation. And then the second thing is a broader point, which is just that the Supreme Court and others will often say, hey, okay, we're gonna. So, for instance, there was a Supreme Court case about requiring a blood draw to have a warrant. Okay. And they basically said, this is a pretty major invasion of privacy. [00:57:50] Speaker A: This is in a DUI setting. For example, you know, cop says, I think somebody's driving drop, and we want to get a blood pool to test blood alcohol level more accurately than breathalyzer can, for example. [00:57:59] Speaker B: Exactly. And when you extract someone's blood, you cannot just learn about their. Their blood alcohol content. But a bunch of other things. It's a big invasion of privacy. You're piercing the skin and so forth. So they said, okay, so what we should do is enact a warrant requirement. And the assumption is that that's going to increase the safeguards for one's privacy. But when we look at this, we're like, well, does it. Right. Yeah, it increases the cost a little bit, but how much of a safeguard really is it? And so I think that's an assumption that's made by the Supreme Court. And when we take this to the data, we see that, hey, maybe it's not there. Right. If to some extent, there's a rubber stamp or the cost of getting a warrant are almost nothing. So those are the two things I would point to. And then if Brett wants to comment. [00:58:39] Speaker A: I'll actually, Brett, we'll go to you that's also just a microcosm for the whole, for the, for the big picture writ large again, which is the structure of this doctrine, the structure of the search and seizure doctrine and the Fourth Amendment depends on the idea of this independent judicial review right that we have. Somebody who doesn't work in enforcement, doesn't work for the executive is going to take a look and guard against the invasions of privacy, etc. So Brett, last. We'll give you the last word, last substantive word. [00:59:05] Speaker C: I mean for me as a, as a linguist, somebody who's, who's not a legal scholar per se. I mean I'm always surprised at how weird and complex legal language is. And so in just, you know, reading through these documents and examining the nature of the language there, I think it, I suspect that it's, there's a lot of nuance in the sort of information provided here and I'm just baffled by how a judge could make a decision on some of these things, how they could approve them without really considering the nature of the way that the officers are talking and how complex some of these decisions must be in order to make a determination. You know, everything that Dane and Miguel have been teaching me about, about the requirements that a person needs to meet in order to, from the theoretical perspective, in order to. Yeah, say this is, this has met the threshold for probable cause. Just how complex the language is and how fast they're reading is just incredible. I know this is something that you're interested in somewhat at the Access to Justice Lab, but just the nature of legal language and how peculia it is, how complex it is, this was not an exception to that circumstance. It is totally, totally different than the rest. How almost all of the rest of natural language works. [01:00:36] Speaker A: Yeah. And this is, it's a great, great comment to close on. It is something that's, that's extremely near and dear to our hearts at the Access Justice Lab. It also, you know, among the many things that you're, the implications that you of what you just said is the, that for the day to day person it renders the entire system inaccessible. Right. That you can't read this stuff. And I've actually had the experience the day to day citizenry can't read this stuff. And actually I've had the experience of trying to write self help materials for unrepresented litigants. And what I've concluded is that I actually can't write them. The only people that can write them. Well, the best self help materials are produced by first and second year law students in combination with a college student because they have the enough of the training to understand the legal issue and be able to read something like an affidavit, and then they aren't so hopelessly inured in the matrix that they can't get themselves out. And meanwhile, I'm plugged in and I there's no way to unplug me, right? I'm signed in the matrix. I'm completely stuck. And I it doesn't matter if you gave me an exit, I can't use it. [01:01:41] Speaker B: Right. [01:01:41] Speaker A: I cannot get my brain to talk that way. And so this is a, this is a problem, right? This is a problem. Yeah. Let me say thank you so much to all three of you. We're going to have to end it there, but Brett, Dane Thorley, Miguel de Figueredo, thank you so much for taking the time to talk to Proof Over Precedent and also to produce such terrific scholarship. [01:02:00] Speaker D: Thanks so much, Jim. [01:02:01] Speaker B: Thank you for having us. [01:02:03] Speaker A: Proof Over Precedent is a production of the Access to Justice Lab at Harvard Law School. Views expressed in student podcasts are not necessarily Those of the ADA JLab. Thanks for listening. If we piqued your interest, please subscribe wherever you get your podcasts. Even better, leave us a rating or share an episode with a friend or on social media. Here's a sneak preview of what we'll bring you next week. It's really important to me that we talk about this topic because it just goes under the radar because most people are focused on getting life without the possibility of parole for people who are on death row, and it's seen as an alternative sentence. But I wanted to argue today that it probably shouldn't be for a number of reasons.