Cold Cases and Serial Killers (Part 1)

SPEAKER 1: Welcome to Justice Today, the official podcast of the Department of Justice's Office of Justice Programs, where we shine a light on cutting-edge research and practices and offer an in-depth look at what we're doing to meet the biggest public safety challenges of our time. Join us as we explore how funded science and technology help us achieve strong communities.

MARK GREENE: Welcome, listeners, to this episode of NIJ's podcast. I'm Mark Greene. I'm a division director of the Technology and Standards Division here at the National Institute of Justice. My background is in engineering from Northwestern. Today, I'm joined by Eric Martin, who is a social science analyst in the Justice System's Research Division at NIJ. Eric has a background in public policy from Purdue University. And I'm also delighted to be joined also by Lucas Zarwell. He is our office director for the Office of Investigative and Forensic Sciences at NIJ. He has a background—a master’s in forensics from George Washington University. 

Today, we are talking about the prevalence of serial killers. Are they on the increase? What individuals are at most risk? And what is known about serial killers and how cases are built against them? Eric, I am excited today to talk about serial killers. I think this is going to be a great discussion. To begin with, many of the listeners may notice that neither of us have backgrounds that you would expect to be on a podcast about serial killers. Lucas, you--you've got a background in forensics, which is really helpful. You can help clarify some things for us. But, Eric, please tell us--tell us what we're doing here.

LUCAS ZARWELL: Yeah. Yeah, Eric.

ERIC MARTIN: Yeah.

LUCAS ZARWELL: Come on. Fill us in.

ERIC MARTIN: I'm so glad Lucas came on to give a street cred. Well--so before anybody gets downloader remorse here, let me set the record straight for you all. So around April of 2018, Joseph DeAngelo, aka the Golden State Killer, was arrested on 13 counts of murder and 13 counts of sexual assault. It turns--yeah, it made the news. It was quite a big deal. And, while it was in the headlines, NIJ realized that they actually played a role in the case closures against Joseph DeAngelo. It turned out NIJ supported both Ventura County and Orange County investigators, mostly through using federal funds to help with the DNA analysis. So as this story was unfolding in the press, it was making big headlines, NIJ leadership, you know, inspired by our connection to this case, commissioned two members of the Office of Forensic Investigative Sciences, Lucas, and as well as me--and that connection is a little obscure, but I think it was because of my former job as a crime analyst with DC Metropolitan Police before I came to NIJ. But the three of us went through and tried to identify all the areas and cases where NIJ had an impact, and it turns out we were able to identify over 2,000 case closures and cold cases. And this may be an undercount because once the project period ends for NIJ support, agencies may or may not continue to keep NIJ apprised of what's going on with those cases. But what's even more exciting or, you know, particularly relevant for our conversation here, within those 2,000 cases, they were able to identify 16 high-profile serial killers. So we went and wrote an article for the NIJ Journal, and that created some buzz as well. Not as much as the national headlines with the Golden State Killer. But we were asked to record a podcast and that's where Mark and I started to team up on that effort. And we're just kind of updating the field with--you know, kind of exploring this issue more. Since we wrote the article and the podcast, there's been a lot more media attention, both in the news and then in TV shows and documentaries, given to serial killers. So we thought it'd be a good time to refresh and just have another conversation.

LUCAS ZARWELL: Eric, that is so cool. I mean, I totally remember reading these articles and how it's become--how NIJ had such an impact on that. But, you know, one of the things I've always asked and like I've always thought this with--you know, coming from a--from a forensic side, we don't really deal with, you know, necessarily this line of work, and so it's really important. Is there a way that you can like remind us, you know, not just me, but maybe Mark as well, and the public, about what is a serial killer?

ERIC MARTIN: Great question, Lucas. And that's important because that definition's changed over time. So if you're reading…

LUCAS ZARWELL: Okay. 

ERIC MARTIN: …literature and you're looking at this question, you may see competing definitions and you might be like, "What's going on?" So as it stands right now, the FBI defines a serial killer as a person who has committed two or more homicides in separate events.

LUCAS ZARWELL: Two or more homicides in separate events?

ERIC MARTIN: So it's not--yeah. It's not like multiple victims within one event. It has to be separated events. It used to be three. So that's where some of the confusion lies. And that was changed recently.

LUCAS ZARWELL: Why do you think that changed?

ERIC MARTIN: I was not consulted in the definition change. I do not know. I think it's just--and there is some literature around it. Not to be too cheeky on the matter, but it--I think it's largely just to make sure they encompass the universe of what's going on. Because you can imagine--we're dealing with very tragic events and these are individuals' lives who are being lost. So you--I think--this is just coming from me, Eric, but that you're trying to study something so you can understand the problem and therefore mitigate it, but you don't want to create like trivial, you know, definitional boundaries, right? So you have to be sure to include all aspects of the problem. 

And that--you know, there's also spree killers as well. The FBI now includes spree killers in their definition of serial killers. Now, this is where it gets even--a little bit more muddled. A spree killer usually kills multiple people over a matter of days instead of over year--months, years, which we tend to think of as the traditional serial killer. Again, I think that definitional change is just to make sure they are the most accurate as possible when trying to figure out the problem. But you can imagine that certain researchers, you know, really build their careers around one type of killer, a spree killer versus a serial killer, and that definitional change does impact research and thought on the subject.

MARK GREENE: I know that serial killers are--seem to be getting more attention in popular cultures like TV shows and--you know, you just mentioned the definitional change from three to two incidents. Are they on the increase? Should we be freaked out? Does that change in definition mean that like we now have more serial killers among us because the criteria is lower?

LUCAS ZARWELL: Good question, Mark. That’s interesting.

ERIC MARTIN: Well--very good question. And, you know, you tend to--I think there's this propensity when you're just confronted with information on a topic repeatedly, you tend to naturally think that this is a phenomenon that is getting worse. I think, largely because these crimes are heinous and violent, they just attract our attention. When you think of--you know, much has been made and written about the serial killer mind, right? And it's just so foreign to how those of us really go about our day and think about things that I think that's where really the intrigue comes about in trying to really get into this genre of film and documentaries. But, to answer your question, all research substantiates that there is actually less serial killers in operation today, down from their peak in the 1980s. 1970s and '80s is where research really points to a peak in serial killer operation writ large in the public, and we have seen a steady decline.

LUCAS ZARWELL: That’s good news, I think, you know?

ERIC MARTIN: Yeah. Yeah.

LUCAS ZARWELL: Very, very different. I did not expect that, you know, especially with all the media attention and all the information that's out there. I mean, I think you're right. There is a perception. It's almost like the CSI effect meets serial killers, right? Interesting.

ERIC MARTIN: Yes. Exactly.

LUCAS ZARWELL: I think one of the things I want to get back to is, you know, since there's less serial killers today than back in the 1980s, like can you help explain that a little more?

ERIC MARTIN: Yeah. And it's hard to know for sure. There's--you know, some of these definitional issues come into play like we just talked about at the onset. Also, how do you know when a serial killer has stopped operating, right, because their offenses could be, you know, drawn out over months, years, even decades in some instances. But best estimates, and there's a whole range of estimates, is I think--it would be safe to say there's around 25 to 50 active serial killers right now at any one time. And that could be as little as a quarter to half of what we saw at its peak in the 1980s. Another way to look at it is--and, again, this is looking at the victims, not necessarily the serial killers, but about 15% of all homicides may be attributable to serial killers. And that's at the high end. I've seen estimates really drop down from there, even as low as 1%. And in that article we originally published, we go through a number of the different estimates. And my method, and I don't know if it's the most accurate, is just to kind of pick the middle, you know, when you have a bunch of conflicting estimates.

MARK GREENE: Right. Right. So I--Eric, I mean, it's such a--it's such a range. Like what--why is it so hard to know for sure?

ERIC MARTIN: Well, much of it depends on deciding what to measure.

MARK GREENE: Ah. Okay.

ERIC MARTIN: And, you know, picking that initial unit of analysis will really dictate what your final estimate may be. And, also, it's--we're talking about crimes, in the reference to the spree killer, that could be very easily observable, right? You know, sporadic, violent events over a period of days. Those could attract a lot of attention. Investigators would make it--yeah, it's--you would think it would be much easier to make those case connections. But with a serial killer who's operating, you know, at a pace of like years, you know, stretching into decades, it--it's hard to make that connection immediately or it may not be apparent, thus, you know, you only could do the best estimates you can to try to capture all of that. And then...

MARK GREENE: I see.

ERIC MARTIN: ...when a serial killer stops operating, unless they've been arrested and brought to justice, it is--you know, it's hard to say. It's hard to count them active with any degree of surety. Are they arrested on another charge and thus, you know, the public is--has a reprieve from risk from that individual? Are they dead? You know, did something happen to them so they're no longer a threat?

MARK GREEN: Right.

ERIC MARTIN: There's no good way to capture this, you know, until the person is identified.

LUCAS ZARWELL: Yeah. That's fascinating. You know--sorry, Mark. You know, one of the things that's come to my mind now is that I imagine that, you know, the unsolved homicides can complicate these estimates, right? Because if you--if you don't know who committed the crime--right? So can you expand on that? Like I don't really get that.

ERIC MARTIN: Yes. Thanks, Lucas. And that was the purpose of our original article, looking at cold cases. And, again, that's where the bulk of NIJ's support in this area has been, just, you know, helping agencies pursue cold cases, bringing--identifying these individuals and bringing them to justice. Interestingly, homicide has the highest clearance rate of any major crime, according to FBI Uniform Crime Reporting statistics. Around 60% of homicides are cleared. And that definition probably could be expanded on a little here. You take the number of cases identified in a year and then the ratio is built by the number of cases closed. So you could have prior year closures count towards the clearance rate.

LUCAS ZARWELL: You're talking about 60% of homicides compared to like--you know, like let's say home invasions, are usually solved, or auto break-ins.

ERIC MARTIN: Much lower. Yes. Yeah. Yes.

LUCAS ZARWELL: But there's a lot of resources put into the--homicides, right? So 60--so...

ERIC MARTIN: They're very heinous crimes. They have a huge impact on the survivors. And thus, rightfully so, police and prosecutors dedicate a tremendous amount of resources to solving these cases. But even at the highest clearance rate of any major crime, 40% of cases in any given year run the risk of going cold. And what is more surprising is while we had--we've identified the 1980s as the height of serial killers operating in the United States, the clearance rate was much higher at that time for homicide. So I think investigators--even though I could say with some degree of surety based on the research, that there are less serial killers operating today, investigators are dealing with that unknown, that population of unsolved cases that are piling up in many jurisdictions that they have to--you know, were helping to pursue, so--especially to limit the impact of any potential serial killers within those cases.

MARK GREENE: So--wait. Eric, what--one question. So in the 1980s, the clearance rate was higher than today?

ERIC MARTIN: Yes.

MARK GREENE: But--I guess, Lucas, you could--you could chime in. The actual forensics tools available today are better than they were in the 1980s. So that's sort of interesting.

LUCAS ZARWELL: It is interesting. Because, you know, like DNA didn't really make its like commercial arrival until the '90s, right? And then...

MARK GREENE: Yeah. That’s like the ‘90s. Right. Right.

LUCAS ZARWELL: Uh-hmm. Which is not that long ago when you really think about technology and the way people have been investigating crimes for so many years. That's a good thing. Hey, Eric, could you--do--can you opine on like what it's like when a case becomes cold? Like how long is it before a case becomes cold? Is that something that we can...

ERIC MARTIN: It's--there's really no, you know, one accepted definition. And I think--you know, there's a lot of diversity in policing in America with so many, uh, law enforcement jurisdictions and so many jurisdictions that investigate major crimes. But I think closure going longer than three years.

LUCAS ZARWELL: I see.

ERIC MARTIN: Does that make sense?

LUCAS ZARWELL: Yeah.

ERIC MARTIN: So if a case is open for three years or more, we would tend to consider that cold. Now--and we'll talk about this more in the next episode of this podcast, but not every agency in this country has the resources for a dedicated cold case unit. And we're going to talk about--just to be a little bit of a teaser for the next--part two of this podcast, we're going to talk about what research says about how to pursue cold cases and how a jurisdiction could really try to maximize its investigative leverage there.

LUCAS ZARWELL: Interesting. And, you know, Eric, that reminds me of another issue because, you know, forensics we’re often dealing with--because we're dealing with DNA and identifications and just the evidence that's found, you know, is it possible that, you know, there's homicides out there that we just don't even know about, right? If I'm a serial killer, I may, you know, not want a person to find somebody, right?

ERIC MARTIN: Yes. Uh-hmm. NIJ has--is really leading the way in this area as well. NIJ hosts the National Missing and Unidentified Persons System, also known as NamUs. And this is a database where information about unidentified persons can be shared. Kind of a linkage database to really help identify people who are missing or bodies who are discovered who don't have identification. Using FBI data, and this is posted right on the NamUs front page on their website, around 600,000 people a year go missing in this country.

LUCAS ZARWELL: Yeah. That--yeah. That's an FBI statistic. I do know this one fairly well. And that’s basically who goes missing in a year and not all those… 

ERIC MARTIN: Uh-hmm. Yes.

LUCAS ZARWELL: Some of those people come back. You know, some of--some of them just are runaways. But there are people that become long-term missing cases, right? And...

ERIC MARTIN: And I think the number of long-term missing is much lower, correct, Lucas?

LUCAS ZARWELL: Yeah. I think it is as well, but I don't know if--I don't--the FBI might have better data on that than we do.

ERIC MARTIN: Yeah. I think it's around 10,000. So, you know, the 600,000, thankfully, many of those are--the whereabouts are identified. But within those 10,000 missing persons, there's also 4,000 unidentified bodies recovered on average in a year in this country. And that speaks to your point you made earlier, Lucas, about--or it may have been Mark. I'm sorry. About, you know, if a homicide is not discovered, you know, that'd be one way a serial killer could contain, or retain, some anonymity.

LUCAS ZARWELL: Yeah. No, that's fine, Eric. We get confused sometimes because we both have beards. But we just have to remember, Mark has the better hair. The...

ERIC MARTIN: Yeah.

LUCAS ZARWELL: I wanted--I wanted to--I wanted to add to that just really quickly, Eric, and also say that I think that's an important, you know, thing, right, because we do know that being missing is not necessarily a crime, but if evidence leads to the evidence of a crime, then we've got, you know, the potential to uncover more serial killers, which is, I think, where the linkage is and--you know, I totally agree.

ERIC MARTIN: Uh-hmm.

LUCAS ZARWELL: And--so thanks for making that connection to the NamUs program. I really do.

ERIC MARTIN: Yeah.

MARK GREENE: Eric, I have a question. So you mentioned over 4,000 decedents, you know, recovered every year. I mean, is there some--like what if two or more of those bodies are homicides attributed to the same person. What--how does law enforcement deal with that?

ERIC MARTIN: Yeah. That's the big question, right, is, you know, so much of the definition of when a active serial killer is in an area, and we've seen historically it is not necessarily tied to one city or one geographic area, it is through case linkages. And there are a number--you know, this is probably the area that has gotten the most attention in TV and print literature, you know? I think everybody has kind of a archetype of the--you know, the criminal profiler. But there's actually a number of ways cases can be linked together. We've met--made a lot of--we've dedicated a lot of time in this podcast to DNA and with good reason. It is the most definitive way to make a case linkage today. And we talked a lot about this in the article, in previous podcasts. Lucas has a wealth of knowledge on the issue, how CODIS works, and how those profiles can be analyzed and linked together. But the basic way this is done is DNA specimens are collected and they can be compared from different crime scenes. And you can identify that the same person left that DNA specimen in multiple crime scenes. And there you go. You have a case connection. You don't have an identified suspect yet, if the identity is unknown who left that specimen, but you know those cases are connected now.

LUCAS ZARWELL: Yeah. Yeah. And it’s cool, Mark. It is. It is cool how the system can actually search and match two profiles that are the same but, at the same time, if the is not in CODIS, it doesn’t become--until it--until that happens, right, they can’t make all those connections. So there’s a lot of arguments for ensuring that we get that data into the database, you know, from people who’ve been convicted of crimes, so that we can make those associations and possibly find out about additional victims.

ERIC MARTIN: Uh-hmm. And, you know, the field is moving beyond just the conventional CODIS database. As--we mentioned, the Golden State Killer at the outset of this podcast. That was through forensic genealogy. What happened was a relative of the Golden State Killer participated in a gene--genealogy DNA database. And through familial matching, they were able to identify the Golden State Killer. So it is more than just CODIS. And we've spent some time in the previous article discussing that. Lot of advances, you know? And I think the field keeps going, both in terms of how precise and how they can collect DNA specimens, where, you know, 10, 15, 20 years ago, that probably--that collection wouldn't have been viable. And then also in their analytic techniques as well. 

We mentioned at the outset too that, you know, there were 16 serial killers, at least, identified within the 2,000 cases that NIJ knows that it contributed to their closure. DNA confirmations were present in confirming Albert DeSalvo, also known as the Boston Strangler; John Bittrolff, the Long Island Killer; Gary Ridgway, the Green River Killer; Lonnie Franklin, the Grim Sleeper; and Joseph DeAngelo, the Golden State Killer. So this is where NIJ's made the lion's share of its contribution and this is where the field is, as far as the state-of-the-art of making a accurate case connection. 

But as I said early on, this is not the only way to make a case connection. There's behavioral profiling. And that's got--I think probably has the--you know, captured the hearts and minds of most Americans when we think of case connections. That's where a lot of--you know, popular culture has really explored this area. This came about largely through the work of the FBI in the 1980s, but interestingly enough, NIJ had a role in it as well. We helped stand up the National Center for the Analysis of Violent Crime, which then produced VICAP, the Violent Criminal Apprehension Program. And it was through the work with FBI agents at this center, and they started developing these profiling tools. Mostly, you know, they--it's evolved over time, but a lot of attention has been given to their typology of the organized versus the disorganized serial killer. Later iterations of this have really tried to operationalize this profiling more and place it within the investigative process. A lot of attention's been given to that. And then, also, separate from behavioral profiling, there's geographic profiling. So we have DNA analysis. We have behavioral profiling. These are the type of crime scene indicators that the killer may leave at the crime scene, that they try to understanding kind of going--what's going on in the killer's head. And then geographic profiling is all about trying to identify the location of that home address.

LUCAS ZARWELL: Wow.

ERIC MARTIN: You know, there's a number of iterations of geographic profiling. Kim Rossmo pioneered one of these, he did a lot of work in this area. And basically--you know, it's very technical, but just to try to give a rough definition of what's going, they can basically try to identify the suspect's home location based on finding a center of the cluster of the known offenses that could be linked to that one serial killer. And it doesn't have to be in homicide.

LUCAS ZARWELL: So is this like proverbial where they put the strings on the map and then they like are cross-indexing it or is it a little bit more...

ERIC MARTIN: Exactly.

LUCAS ZARWELL: Is it more complex than that, Eric?

ERIC MARTIN: We now--you know, we have GIS, you know, analysis and crime mapping, and a lot of geographic algebra that could sit behind. But, yeah, that’s--Lucas, you’re exactly right. This is, you know--or often precedes science, and this is exactly the same technique but, you know, with a little bit more mathematical bent.

MARK GREENE: Yeah. It's like center--it's like a center of mass calculation kind of thing. Yeah. It’s kind of interesting, yeah.

ERIC MARTIN: Yeah. Exactly.

LUCAS ZARWELL: Well, I'm sure--I mean, I'm sure it's more complicated than strings on a map in a pin board now, so--but, I mean, I think that's really cool because what I think people think about geographical, you know, profiling, that's really kind of--that's the flash version of it, right? But the real version involves a much more in-depth analysis of, you know, the region and the location. I could see that happening. Yeah. Very cool stuff. 

MARK GREENE: But that's like--but that's like the movie version, you know? It's like the person's sitting there after like spending like 48 hours awake, like pouring coffee, and they've got like a million like pictures and files everywhere, and it's like, "Look,"--you know, "Oh, there they are," you know? Like...

ERIC MARTIN: And, you know, it's interesting, and not too divert too much, but all of these methods, you know, fill the need investigators had. Investigators were the ones who knew, "I--if I had this information, I could probably figure out who this is," right? And in all these cases--you know, DNA analysis, you know, started out in paternity testing and other more, you know, medical applications, and then was adopted to criminal investigations because, you know, it was filling that need. And the same way that a lot of psychological profiles and also the geographic profiling.

LUCAS ZARWELL: I love it. And I--and I--and I think it's--I like the way that you brought home the interest in like, you know, investigative genealogy and how it's being--or what they call FGG and how they're using it as a tool, right? But, really, it's that DNA CODIS link where they can get the DNA from the individual, make the match to the crime, and then that's what's prevent--presented as evidence of saying like, "Hey, this person's DNA was now found definitively."

ERIC MARTIN: Yeah. That’s a good point.

LUCAS ZARWELL: Yeah. And I think that's really important to remember too, right, because CODIS is such a powerful tool to this day and it's really great to hear how everything is interconnected.

ERIC MARTIN: Uh-hmm.

MARK GREENE: Hey, Eric, I have learned a ton in this podcast. I'm going to let us take a break right now and our audience take a break. We're going to continue this discussion in part two of the podcast. So, listeners, please stay tuned. Join us for part two of the podcast on serial killers.

SPEAKER 2: To learn more about today's topic or about NIJ, visit the links in the episode description and join us for new episodes every month.