#04 (Kim Rossmo)

I'm Gerry Rackliffe with reducingcrime.com, a podcast featuring interviews with influential thinkers in the police service and leading crime and policing researchers working to advance public safety. Kim Rosmo is a former detective from Vancouver known as a pioneer in the area of geographic profiling. However, in this podcast we discuss his latest work trying to better understand investigative failures. Find out more at reducingcrime.com and on Twitter at underscore reducingcrime. Kim Rosmo has a PhD in criminology and is currently a professor in the School of Criminal Justice at Texas State University. However, before entering academic life, he spent 21 years with the Vancouver Police Department, where he was the detective inspector in charge of their geographic profiling section. His research interests include environmental criminology, the geography of crime and criminal investigations, and he is particularly interested in improving police investigations by studying wrongful convictions and how solvable cases go cold. As part of this work, he's a member of the International Association of Chiefs of Police Investigative Operations Committee, and that's quite a mouthful, and he recently completed a National Institute of Justice project on deconstructing criminal investigative failures, which is the topic of this podcast and what we talk about. Kim, thanks for joining me. You've been working recently on investigative failures. How did you get into doing that? A couple of reasons. One was as a doctoral student, I was involved with a wrongful conviction case study in Canada called the David Milgaard case. What was that about? He was convicted of the sexual murder of a nurse in my hometown, actually, because I was interested in environmental criminology. There were a lot of aspects of the environment of the crime, where it occurred, the weather, the timing that were really intriguing and were things you can analyze. Just in case anybody's new to this kind of stuff, we should tell them that we're not really tall and can make out oil spills in Alaska, right? No, we're talking about time and place. What are the characteristics of the location where the crime occurred, the timing it occurred. In this particular murder, there were aspects about where different witnesses were, the timing, the weather that made me interested in it, but that analysis really didn't add up to the David Milgaard being the guilty party. DNA eventually showed that he was innocent. Someone else had committed a murder. Did he spend a lot of time inside? 23 years in prison. That's true. Oh my goodness. How did that happen? This case involves a multiple number of causes. It's very much like an airplane crash. You have a series of occurrences that all just kind of come together at the wrong time and wrong place, and they lead to a failure. Now, you mentioned this because you know I'm a pilot, right? Yes. Yeah, and you've been in a plane with me. So thanks very much for talking about airplane crashes. But you're absolutely right. I mean, I've studied airplane crashes, and they tend not to have one point of failure. It tends to be a kind of cascade of clusters, right? So it's low fuel plus low clouds and poor decision making by the pilot, or a mechanical failure that the pilot fixates on equipment that's just not showing the right information like a needle in the wrong place, and then the pilot fixates on that and then loses situational awareness. Is that the kind of thing you're talking about? Yes. James Reason, who's a well-known safety expert, talks about the Swiss cheese effect. So you're just lining up all the different holes in different pieces of Swiss cheese, and any one of them could block the hole. So that's why airplane accidents are rare. It takes a lot of things to go wrong at the same time for such a failure to occur. Same thing with a wrongful conviction or criminal investigative failure. So in this case, at least didn't find some of the physical evidence. They had dodgy witnesses, and they didn't check the veracity of what they claimed they saw. Milgaard was in the wrong place at the wrong time, and these all coalesced together and ultimately resulted in a wrongful conviction, one that probably could have been avoided. One of the most compelling things in the whole case, probably the single most important piece of evidence got convicted, was one of his friends, who ultimately, after pressure from the police being locked up overnight, said yes, she saw the murder. Now, these were juveniles. These were teenagers. But what she said she saw was an attack on the street, and then Milgaard stabs this woman, pulls her into an alley. The temperature was colder than 30 below. So it was very, very cold. So this was a Canadian summer, right? Yes. And the trouble was is that there were stab wounds in the woman's back and stab wounds in her coat that matched everything, but there were no stab wounds in her nurse's uniform. That's impossible, given an attack on the street. What had to have happened was, coat came off, the dress came down, the coat went back on again, and then she got stabbed. But none of that fits with the witness's description. By the way, we found this failure to assess the reliability of evidence to be one of the biggest causes of wrongful convictions. Okay, so you've looked at many more cases, not just this one case up in Canada. This is kind of the spark for you, kind of got you interested in this area. Yes, we looked at 50 cases, most from the US, but also some from Canada, United Kingdom, and Europe. Is a cluster of failures common? Is that like a thread throughout these? Yes, they evolved on average seven different factors that led to the failure. We grouped those factors into personnel or individual level factors, organizational factors, and situational. Situational being factors outside the control of the criminal justice system. So which was the most common? The most common was personnel, individual level, so within the control of the detective, him or herself. Is that the way that cops are wired, or is it something that can be trained for us to avoid? What are we talking about? In a few cases, they involved misfeasance or noble cause corruption, efforts to get the bad guy no matter what, but by far and away... Kind of a dirty harry effect. Yes, exactly. But by far and away, the most common problem was some form of cognitive bias. Rush to judgment, premature shift to a suspect-based investigation, instead of an evidence-based investigation, tunnel vision, and confirmation bias. So these sort of mental-optical illusions are what usually get a detective into trouble. Okay, so we're not talking about cops with mental health problems here when we're talking about cognitive problems. We're just talking about behavioral learned traits that lead us in a direction too quickly. Is that the sort of thing? Yes, we all suffer from cognitive biases. Think of them as heuristics or shortcuts. So these are things that lead us, really like optical illusions, but tricks of the mind. We all do them because it takes a lot of effort to analyze and think things through. Now, they work most of the time, but not all the time. Aren't they the kind of process that's going on in the back of our mind as we're going through our day-to-day tasks? Yes, psychologists talk about two types of thinking, the analytic and then the intuitive or the below-conscious type of thinking. And the below-conscious type of thinking, because we're not really aware of it, often will get us to a solution quickly, but if it's wrong, we're not aware of it. To some degree, we need these biases. Is that reasonable to say? Well, like I say, they promote our survival, so that's why we have them. But I'm not on the Serengeti anymore. Yes, and the trouble is, well, it might be appropriate for quickly deciding what to order when you're in a rush at a restaurant. They don't really make sense to imply when you're filling out your tax return. Or if you're doing a complicated major crime investigation that has to go to a court, there we want accuracy. We don't need to be in a big hurry for those things. Okay, so when I'm standing in the Starbucks line, and there's somebody in front of me who suddenly gets to the front counter and decides then at that point to start thinking about what they wanted, you kind of wish they had some cognitive biases and just to get their act together, right? Yes. Just get something that's good enough and move along. Yeah, please. All right, so the problem then kicks in when we're talking about looking at investigations. And there are no really quick decisions to be made in investigations on the whole, aren't they? Well, you may need to make some at the scene in the moment decisions, but that's very different than deciding who you're going to charge, how you're going to prosecute the case, how you're going to pursue the lines of inquiry in the investigation before that. You have a lot of time to think those through carefully. We did find one of the situational factors that was problematic was a high-profile crime usually involving a child victim or something horrible, leading to a lot of media attention, leading to what is called production pressures. So you need to get this solved now. This is a little artificial. It's not about really solving the crime. It's dealing with the politics of the investigation. Yeah, it's not so. There's politics and policing. Unfortunately. I'm just a shock to me. I mean, examples of that, I know I've seen you talk about this starting out. The Mandanox case is one where there must have been a huge amount of pressure on the prosecutors and the detectors involved. Yes. And that was international media pressure. And a lot of it was tabloid journalism. Things like Foxy Noxie became much more important than what the evidence showed. Foxy Noxie. That was the nickname the press gave to her. She was an attractive looking woman. She behaved in some ways that the local Italian police thought was inappropriate. She, like many college students, had engaged in sex. And, you know, all of a sudden this became something much more sinister than it really was. And that particular prosecutor has a bit of a reputation for turning these things into grand cases of... Like a show trial. Yeah. Or satanic sexual overtones and other really strange stuff. He was the same one involved in the Monster Florence case where some of the same issues came up in involving other people. I'm sorry. All the satanic sex stuff then. It's really not like a college student. It's more like the faculty, right? You know that when you said that just a few seconds ago about half the people listening to this podcast just went and Googled Foxy Noxie. How do you extract yourself from that? How, I mean, if you think about it from a supervisor's perspective or the police chief, it will be in the United States who has to probably deal with the politics around this. What can they do? There's a trade-off. The faster the resolution, the more likely they'll be in error. And they just have to resist that. They need to protect investigators from the media, from the politics, from sometimes the district attorney's office. It's much better to go a bit slower and get it right than to end up having literally often paid millions of dollars for a wrongful conviction. You have an extensive police background as a sworn police officer in Vancouver yourself. Has that helped your understanding of the processes? I think so because the other side of this coin is detectives need to solve the case. They want to get the right person. We sometimes say the only way to avoid all wrongful convictions is to never arrest anybody. And that's not workable. So we also think that the same mistakes that lead to a wrongful conviction can also result in a solvable crime not being solved. If you do an investigation properly in an unbiased and thorough fashion, you're going to solve more crimes and you're going to solve them correctly. What do you mean by cases that could be solved, not being solved? Let's take an example of the murder of Chandra Levy, the intern in Washington, D.C., who disappeared and was found murdered in Rock Creek Park a few years ago. There was lots of reasons to think that case could have been solved. But unfortunately, the police focused way too much on Gary Condit, the congressman that she was having an affair with. Now that was salacious. It was a huge media event. It got a lot of press. But there are many reasons to discount him fairly soon as a serious suspect. How did that come about? I mean, were they just focusing too quickly on a suspect and then ignoring everything else that came around it? Yep. We talk about an investigation shifting from evidence-based to suspect-based. If that happens too quickly, you're going to have a problem. That's exactly what occurred there. They made up their minds in this tunnel vision process. Again, the media played a role in making sure that was a case with a lot of pressures on it. Washington, D.C., of course, being the capital probably has more politics than most cities. The fact that their minds were made up, made them overlook or miss a very viable suspect. But at some point, we have to do that. We can't go through the whole process of being an investigator or a detective filled with doubt. At some point, we have to turn around and go, "Okay, there is enough here to look at the suspect." At some point, we have to make that shift, don't we? That is, again, the shift from evidence-based to suspect-based. If you're going to go to trial, you need to do that at some point. The trick is knowing when to do it and to make it happen. Really, you need to collect all your evidence, analyze that evidence in terms of any forensics processing, and check the reliability of that evidence. Then you need to think about it. But if you're making up your mind too early, you're going to have a problem. One of the ways we test for confirmation bias is we take a look at the timeline for evidence collection and analysis. We compare that timeline to the point when the decision was made to focus on a suspect. Can you find that timeline in paperwork? Because what's going to be reflected in the paperwork may not be what's actually in an officer's head. Well, I'll give you another good example. There was a case involving a former Indiana State trooper who was convicted of murdering his wife and two children. Police investigators found a sweatshirt at the crime scene, actually under the body of the son. They submitted it to the lab, but they arrested their suspect literally within a day of the murder before any of the analysis had occurred. Eventually, the analysis of the DNA recovered from that sweatshirt pointed to a convicted felon who was responsible for the murders. That's going to be happening in huge amounts, at least in the United States, because the backlog of DNA evidence submitted to labs around the country is huge. I think NIJ or some of the federal government agencies are syncing millions of dollars just trying to clear the backlog, right? Yes, and the quicker we can get that information into the detectives' hands, actually into their minds, the better. However, in this case, there wasn't a huge delay in the processing. The really tragic part of this occurred when the results were never put into the DNA CODIS database. That's ultimately what cracked the soap, but that literally took years. Okay, so that's not really a failure from the individual investigator. That's more of a failure from the organization to some degree? Again, we have multiple failures. The district attorney got involved right at the beginning. There was a rush to judgment, and a lot of things the on-scene investigators could have done, they didn't do, particularly canvassing. They made mistakes regarding... Canvassing? ...talking to neighbors, other people that lived in that area. It was kind of a cul-de-sac area, so if you drove in that area, a lot of people would have seen you, and they did see things, but they had their suspects, so that really didn't become important. So, if I'm a supervisor, or I'm running a unit, and I'm an investigative, a lead investigator, or running a detective team, what can I do about this? Well, first of all, you can make sure that detectives are protected from these external pressures. You can make sure that they don't come up with illogical conclusions, because some of these conclusions that are used to justify a rush to judgment are bizarre and don't make any sense. In the one case I just told you about in Canada, the Milgaard case, even after the DNA showed that a serial rapist who caught the bus every morning at the same bus stop as the victim did, it was his DNA. Some detectives tried to explain it away by saying, "Well, no, no, we're sure Milgaard killed her." And then this rapist came along, and mine is 30 degree temperature found her body in an alley and decided to have sex with it. You know, that's not very logical. You could also make sure that there is no bending of the rules. Sorry, I'm kind of fascinated by this. I mean, absolutely. I mean, anybody listened to this thing, that's probably amongst one of the craziest things they've heard. So, how did the detectives, how did these guys get into that mindset, that they could make this massive leap from logic, reason, and reality and justify that approach in their mind? Did you ever get to the bottom of that? Well, I think that's just a basic human failing. Once we make our mind up, it's hard for us to change our minds. And then, you know, some detectives have big egos. There's a lot at stake here. There's a good quote from Richard Hewer, who's a CIA analyst who deals a lot with decision making. And he says, "Once the bell is rung, you can't unring it. Premature judgment can be really hard to push back against. And if there's some costs, i.e., it's now in the newspapers, you've charged the person, your reputation is involved, it becomes really difficult to go back from that. It's a lot easier to try to irrationally attack the decision from the other side." That brings up another response to your question. And it's something that happens all the time in the United Kingdom. If there's a problem with an investigation, if it remains unsolved after a certain time, in England, what they'll do is they'll get a senior investigating officer from a different jurisdiction to come in and look at the case. Fresh set of eyes. And that does two things. First of all, the original detective knows that it's going to happen. So they're a little more careful. And second of all, if necessary, this external review can be very helpful to the case. You mentioned that gulf between British policing and the U.S. policing. And it's not because I'm British. I have a bias in that regard. But it also helps that the homicide rate is a tiny fraction of the U.S. homicide rate. But homicide investigations in the U.K. are this huge construction with a lot of resources, teams of detectives, a lot of uniform resources for canvassing. It's a huge effort. And homicide investigations, especially in some of the urban cities in the United States, is kind of a couple of detectives get assigned for a few weeks and that's it. And they're also picking up and dealing with their previous caseload and everything else that they've been dealing with. The gulf between the investigative effort from these two countries seems huge. It is huge. And that's why there's over a 90% clearance rate for homicides in the United Kingdom versus somewhere in the 60% in some particular cities. Right now, we're in New Orleans and the homicide clearance rate in this city is nowhere near that. It's really bad. And it comes down to, do you have the necessary resources? Or do you have the right people doing the right job? And he has to play a part in that. Yes. Selection and training. And I've had some interesting discussions with people that they said, not everyone can be a good detective. You have to be willing to admit you made a mistake. And if you're so... That's so against human nature. It's so against how cops feel about stuff, too. Yes. But they've suggested maybe we need psychological tests for picking up on people that are resistance to changing their mind even when new facts come along. That's interesting because when we're training police officers, early career police officers sent out to deal with an incident often on their own. And so decisiveness is a factor in that. Yes. So it's a bit contradictory. We like to encourage decisiveness in early career. And then later on, we also want that more reflective practitioner who's prepared to accept fault and take criticism. And that's almost an entirely different skill set. And often you're picked for that job because you were good at the first job. Yes. But they're different jobs, if you think about it, quite different in fact. I get there's different jobs, but this just sounds so much like the policing I know, which is you're really good at the current job. So let's promote you to a entirely different job. You know, the Peter principle comes into play. Which is? Oh, the Peter principle? What happens in any organization is people get promoted to their first level of incompetency for just the reasons you talked about. You're good at this job, we'll promote you here. You're good at that job, we'll promote you at another level. Oh, you're not any good at that job, but we'll just leave you there. I have some sympathies at some point for police commanders who are in that situation. I mean, especially in organizations that don't provide much in the way of training. It's kind of like we promoted you to detective. So come and in plain clothes on Monday and here's a different color badge. Well we're in New Orleans right now, which is a city that's currently under a federal consent decree. Many wrongful convictions have happened in the city and there's minimal, if no training given to a number of the detectives. This is something they're trying to improve upon. But even in places that provide lots of training, most of it focuses on such things as case law, forensics, databases. There's almost no place that I'm aware of that teaches detectives how to think like detectives. So you can screw up the case, but just don't screw up the paperwork. That's, well we're, it's all part of a bureaucracy, isn't it? Right. How much is the push for clearance rates turning into like a perverse incentive? You know, in a similar way that we drive all sorts of problems in terms of the quality of pedestrian stops or field interrogations, investigations, if we start demanding numbers and driving quotas. Is that something else that affects the work of investigators? Undoubtedly. If you ever saw that great television show called Homicide Life on the Streets based in Baltimore, the focus was on- That was the David Simon work. You also did the Y. Correct. They had a big board with all the homicides listed and who was assigned to them. And you didn't want too many cases that were attached to your name that weren't cleared and they actually used different colors to show this. So there is that pressure and sometimes that pressure will cause detectives to take shortcuts and this is where you get noble cause corruption or you've misfeasants coming in. That's where supervisors and managers can play a role, but it looks like in many cases they're kind of nudge, nudge, wink, wink. Yeah, go ahead and do that. We need to solve this crime. They benefit from it. So they can be a little bit complicit, even if they don't mean to be, because it strikes me that's both an organizational pressure, but also a cultural one. As soon as you have a horse in the race in any way, that's going to affect your decision making and introduce biases. Because nobody wants to be thought of as being a crappy detective. And no one wants to have a team of detectives working for them that aren't solving crimes. So they're driving a pressure that might solve the crime, at least initially, in terms of getting a suspect, but when you look at the whole case right the way through to the prosecution, it's succeeding at one end, but driving increased failures at the other. Right. And now with DNA, we're finding out that a lot of those even successful prosecutions were of innocent people. It's true. Are there other kind of supervisory leadership takeaways that you think might be useful for people listening to this? Well, you mentioned the importance of training. Absolutely, I think, is essential for people to do their job. What would that training focus on then, if you could change the current regime? We solve a case through evidence. You have to understand not only just how to get evidence, that's the forensics, that's how to talk to people, witnesses, but also how... That's sort of in current training right now. But what is missing is this whole idea of evaluating the reliability of that evidence. I mean, it's logical. People would say, "Oh, yeah, we do it," but it was one of the major problems that kept coming up over and over again. A jailhouse informant tells me what I need to know to solve this crime. Now, this jailhouse informant's asking for a deal, and guess what? He's done this three times before in his life. How much weight would you give that? Well, objective analysis would say almost none, unless he's telling you something that no one else knew, that only the killer could have told him. OK, so that's a useful clue, which is this guy just spinning you a line, or do they actually have some insight that suggests they actually know something? And this also implies two interrogations. We know now that innocent people will confess to things that they didn't do. We know that certain groups, people that are mentally challenged, juveniles, and people that are naively optimistic and believe in the criminal justice system, think, "Oh, well, I'll just say this to get this over with," but they'll find out. Those people still out there? Where are they? But we need to know that these false confessions can occur. We can evaluate the quality of a confession by the amount of information that's in it, especially information that's not known, and we can make sure that we don't pressure people, especially after hours of interrogation, to say something that they weren't guilty of. Intelligence analysts use a technique called analysis of competing hypothesis, where they look at various different pieces of evidence and then assess that evidence against the different hypotheses of possible outcomes that might be taking place. That sounds to me a little bit like what you're talking about, is to look at the evidence in a more objective fashion and assess its merits. It's very interesting you mention that, because what has happened in a number of these cases is there are reasons to be suspicious of somebody, but the detective sometimes doesn't take the next step and say, "All right, this evidence totally fits with the theory that this guy's the killer," but could it also just as easily fit in with the theory that he's innocent or that other people may have been the killer? Or is there any evidence that disproves this suspect? Yes, you need to consider both the pros and the cons. This is where confirmation bias gets in. Oh, you have an alibi. Well, yeah, but we're not going to believe any of those people that provide the alibi for you, because we don't want to. Not because there's no any objective reason to discount that. And sometimes those people could be lying, but you have to assess the merits of it, not just discard it completely. Exactly. You want to do that in an objective fashion. That's where problems occurs when that objectivity gets lost. Because you know at some point those witnesses are going to appear at trial. Yes. Or in the civil case down the road, if it ends up being a wrongful conviction. Right. And then it gets really expensive. Yeah. The Washington Sniper case. How do we end up scouring half of Northern Virginia for a white van? Initially, the white van was reported as a white box truck that was seen in the area. Now, that makes it a suspect, but not the vehicle driven by the murderers. And then it became a white van. Now, no one seemed to focus on the disconnect between a van and a white box truck. Then the white van began to be seen at other crime sites, even after the shooting. Like the killers would engage in this act of murder and then drive by again just to take a look at the police response, which of course makes no sense. If you stand at any busy intersection in the DC area, you're going to see a white van drive by within five minutes. Every delivery vehicle in the area is a white van, right? It's a really good example of what we were just talking about, Jerry. Yeah, that might be an indicator of something that is interesting, but it could also just be totally spurious because it's a common factor. You've been doing all this work looking at all these cases. Have you been talking to the detectives and investigators involved? And for the research project, no. But I've given a number of presentations over the years. I wrote a book on criminal investigator failures in 2009, and I've received a lot of requests for presentations. And quite surprisingly, most homicide detectives and other investigators want to know about this stuff. Now, they might be resistant or defensive, as we would all be if you're talking about their case, but generally they want to follow the best practices, and they have a lot of interest in these issues. So I take that as a sign of encouragement, but on the other hand, we know that these biases are so much part of how everyone thinks. That's a harder thing to break, and we need to provide ways of teaching people to avoid falling into those traps. A lot of that is associated with a blame mentality that I think we have in society generally, especially towards people working in the public service. Yes, and towards policing. Now, there's this whole initiative that the National Institute of Justice has developed, kind of spearheaded by James Doyle, a defense attorney out of Boston called the Sentinel Events Initiative. So if there's an airplane accident or a train accident or an accident in the operating room, a Sentinel Events approach tries to figure out how and why did things go wrong. They're not really interested in the blame part. They just want to make sure they understand how that failure accident occurred so they could prevent it in the future. As a consequence, air travel is tremendously safe and safer than it was 30 years ago. I'm glad you said that because you've been up in a plane with me. And that's the fact the aircraft you've been up in. I'm on a list server for pilots who fly that aircraft. And what I've been amazed by and impressed by is how open people are with other pilots when they've made the mistakes. Because it's a sea plane, people who have landed on water with the wheels down and flip the plane and sunk it or half sunk it. They've explained what systems broke down in terms of their personal checklists or other problems that they were distracted that made that happen. And I've really been impressed with the honesty and the willingness to share and be open about mistakes. And that's just a cultural thing I do not see in policing at all. That's the problem with the blame game. I mean, to be fair, it can be career limiting in policing. Very much so. However, if we don't understand how these things happen, it's so hard to fix them. It's safe to say that very, very few detectives want to arrest the wrong person. So they honestly believe in their incorrect conclusions. So that's a source of hope. It's also a source of some frustration. When pilots do something wrong, such as accidentally fly into restricted airspace, they can go onto a website that was being set up by NASA and actually sort of confess their sins and say, "This is me and this is what I did and this is the mistake I made." In the process of doing that, a lot of the time, not totally, but most of the time they can avoid any kind of penalty because most of the time it's genuine mistakes. The value of that is that becomes a huge learning tool for the aviation industry to understand why and under what circumstances pilot or bust an airspace. Would you ever see a chance when we can develop something similar to that in policing or in such a state nowadays where there seems to be no space for honest mistakes? I would like to think we could and actually that's one of the objectives of the whole Sentinel Events Initiative within criminal justice. We have a long ways to go. You just have to take a look at a police shooting that was at least a bit controversial and you already have a civil lawyer talking about suing the police and no one really knows what happened in the case. And everybody battens down the hatches and at that point everybody's just keeping quiet to limit the financial cost. As we've been talking, I've been thinking about, I think you're aware of the military approach of having a red team, almost a deliberately set up opposition. Is that something that you think might work in investigations? Yes. In fact, there are some jurisdictions like Vancouver, my former police agency, when they were recovering from some of the mistakes they made in a major failed investigation, decided to implement devil's advocates. So the role of a devil is advocates to punch holes in the dominant investigative theory, to just try to see what are the weaknesses and can those holes be patched up? Can they be defended? People don't want to criticize their colleagues' work, but if this is a standard practice, it's part of the routine, the culture, then people are going to be much more accepting of it. So rather than we're just picking on you because we don't like your investigation and we're thinking out of problems in it, this becomes a sort of supervisory role for every investigation. Yes. And then we're going to be more willing to listen. Our backs won't get up, we won't be defiant. And if you think about it, this is what we do in academia, when we have peer reviews taking a look at our research and try to find the holes. And then we come back and go, yeah, okay, I need to change what I said there, or I need to fix that, or I'm going to argue my point. It's a healthy process that is better at getting to the truth. Well, it's not that healthy. You've been peer reviewer on number two, probably on too many of my papers, right? The ones that never got published. Number two is always a jerk. I hate reviewer number two. What I really like about the research that you're doing is it just feels so different than so much work that's going on in policing at the moment. I think I remember a study that was saying that even though there have been massive changes in policing over the last 20, 30 years in terms of DNA and technology and forensics and research into hotspots policing, the world of the investigators remained largely untouched with the exception of the sort of DNA and the forensic side of things. What's really fascinating about what you're doing is tapping into that and starting to go back and look at that process again. Strangely, even though the investigative function is a very large and important part of policing, it doesn't get much attention by researchers or others. In fact, someone did a survey of papers given at the Academy of Criminal Justice Sciences, meaning in Los Angeles a few years ago. It's heavily practitioner focused, a lot of policing research. Only 2% of the papers were about criminal investigations. Practition of focus, you mean street policing, patrol, uniform, that side of things. Correct. The investigative work has largely gone untouched and unnoticed. That's right. Even given the practical focus of that particular organization and that conference, only 2% of the papers were about investigations. Now, the last I remember was, to some degree, John X's work in the 1980s that would have had a significant feel on the field, especially around the area of triaging investigations. But since then, it's kind of really been nothing. Why do you think that is? I really don't know. I mean, there is discussions about databases and information and forensics, but not really about thinking like detectives and how can we get to the reality, the truth of what really occurred in a crime. Maybe it's just the particular interest of people. Lawyers will take a look at the legal aspects, and scientists will take a look at the forensic. And there's been some work on interviewing. But it sounds to me that the mistakes can often be made prior to that. So by the time we've reached the interview, we've already focused possibly prematurely from what you're saying on a suspect. We've dragged that suspect in for interview, and it's about the interview process and the tactics. Well, that kind of leads to this whole notion of the focus on the why instead of how. If you take a look at research done by the Innocence Project, they'll identify eyewitness errors, false confessions, faulty forensics as the major causes. That's more of the technical side of things. Exactly. That's the case management, not the case philosophy. So even if you have the best tactics, if you're not thinking about things properly, if you're not given evidence as proper due or weight or giving it too much weight when you shouldn't, you're going to make mistakes. So yes, you want to have the best tactics, but you also need to be thinking of what all those pieces mean and how they fit together. That's why we are so interested in this sort of systemic causal approach to understanding why these failures occurred. Kim, thanks very much for your time. You're welcome, Jerry. That interview with Kim Rosimo was episode four of Reducing Crime, recorded in early 2018. You can find more podcasts like this at reducingcrime.com or wherever you found this. New podcasts are announced on Twitter at underscore reducingcrime. Don't forget the underscore. Be safe and best of luck.