Preview Mode Links will not work in preview mode

For a transcript of any episode of this podcast, email cgupodcasts at gmail.com and include the episode title.

Oct 6, 2021

For years, CGU Associate Professor of Economic Sciences Greg DeAngelo has been studying and interpreting data from coast-to-coast. Now, he and CGU’s Computational Justice Lab work closely with agencies in the public sector to address pressing questions of the criminal justice system. We sat down with Professor DeAngelo to discuss the finer details of the Lab’s work, along with its focus on finding causal effects of various justice issues and policies.

For more information about CGU's Computational Justice Lab, visit their website.

TRANSCRIPT

Jeremy Byrum:
Hi, everybody. And thank you for joining us from wherever you may be. It's getting a little chilly out there, so welcome to the campfire. I'm your host, Jeremy Byrum. And today, I'm excited to be joined by our guest, CGU Associate Professor of Economic Sciences and Director of the Computational Justice Lab, Gregory DeAngelo. Dr. DeAngelo works closely with public sector agencies to address pressing questions of criminal justice policy, identifying the causal effects of actions by both legal and extra-legal actors on public safety. Along with the Computational Justice Lab Research, they generate technologies with the potential to counteract any negative externalities of the actors we mentioned. Greg, thank you so much for joining the show.

Gregory DeAngelo:
Thanks so much for having me. I'm excited to be here.

Jeremy Byrum:
Awesome. I'm excited to talk to you. So I want to get into a little bit of your background before we talk about the crux of our discussion here with the lab and what you and the students do in the lab, but how did you, yourself, get into the criminal justice space, in the research space? And then also, how did that bridge into the data analytics space in how that led to you creating the Computational Justice Lab?

Gregory DeAngelo:
Sure, yeah. Happy to explain some of that. As the story go, it's not a terribly interesting story, but ...

Jeremy Byrum:
Well, stories are.

Gregory DeAngelo:
... I went off to school, to college. Really, my parents didn't go to college, nobody really. I was a little weird in my family in a sense. I was early into math and sciency stuff. And my siblings are more on the artsy ... I mean, as it turns out, we've all dabble in the teaching space a little bit. But I was the weird one. And I went off to college and I was interested in technical math type problems. And I was hanging around a bunch of engineers. And they're great people. And I had a lot of fun hanging out with them. But I just was super pulled toward social problems.

Gregory DeAngelo:
And I wanted to find how I could work on social problems. And in particular, at that stage in my life, I was interested in the environmental issues. And I had some great faculty members who had influenced me quite a bit and said, you really ought to think about a PhD. And I remember talking over my parents and they were like, you're getting your degree. You don't need to go get another. I was like, yeah, I think I'm going to do it. Not only am I going to get a degree, I'm going to leave upstate New York where I lived my whole life and move to California.

Gregory DeAngelo:
And that, I think, got a couple of furrowed eyebrows and like, what the heck are you up to here? But anyway, so I got to grad school thinking I was going to do environmental economics. And I was intrigued by environmental economics, but I don't know. It just wasn't the thing that I thought I was going to really dive in on. And I forget which trip home it was. I wish I remembered for a life of me now, but I'd flown back to the East Coast to see my family. And I was on a flight from the East Coast back home, and I sat next to a guy who worked for the Oregon State Police.

Gregory DeAngelo:
And it's a five-and-a-half hour flight. And I remember the guy, he probably wanted to get some shuteye, but he ends up next to me, something like 23-year-old, wet behind the ears, super well caffeinated puppy dog who just wants attention. And so, I'm sitting next to this guy and he starts talking about all these issues going on in Oregon with regards to the policing. And so, that led me to dig in and start asking questions about what the heck's going on there. And long story short, the guy hands me some data. Well, he didn't, another person.

Gregory DeAngelo:
He handed me some data, and that got the ball rolling, me thinking about criminal justice issues. And that's like 16 years or something ago now. And so, I really started becoming intrigued by the criminal justice system and how complicated it is. And the layers, there's so many layers there. And the minute that I finish a paper and I think to myself like, okay, I'm done on that part of the criminal justice system. I'm not going to think about it anymore.

Gregory DeAngelo:
I feel like two or three years later, I end up picking up that data, or going back to that problem and reassessing it, thinking about it through a new lens. And it's made for a career now. And so, I was really into this stuff. I really enjoyed working on empirical issues related to the criminal justice system. I was having a lot of fun with it, but I was also recognizing that a lot of the agencies I worked with had issue. They had real issues that they couldn't sort out internally. Sometimes it was evaluating their practices and policies.

Gregory DeAngelo:
Other times, it was more tangible things like we know there's bad guys out there. We don't know how to catch them because a lot of the marketplace that used to be on the street have moved online and we're not as savvy in that space and we need help. And so, and I could talk about that more, so anyway. There's all these ways. I just found that I was interacting with numerous criminal justice agencies and I was enjoying it. Not to sound insensitive toward my colleagues, but it wasn't just writing academic papers anymore. Everything was coming to life.

Gregory DeAngelo:
I was talking to the people who were creating the data that I was using, and I was just loving every bit of it. But I recognized at that point in time that you got to get pretty serious about your toolkit as a researcher, and so dug in pretty deeply on the data science and computer science toolkit that's needed to work with a lot of these agencies and perform the kinds of analysis that we conduct. And I was really enjoying it. And then fast forward a few years from there, I'd been working on this stuff and I had PhD students around me that were really interested and undergrads that are really interested in this work.

Gregory DeAngelo:
And I started saying, there's more agencies and more data and more problems out there than I, individually, or even me with just a couple of PhD students could ever take on. It'd be interesting to scale this and give PhD students the opportunity to gain some of the education that I had obtained from interacting with real people. So much of what we do in the academy is like a good distance away from reality. Not to sound insensitive or inconsiderate toward my colleagues. But a lot of times, we study people, but we don't ever talk to the people that we study.

Gregory DeAngelo:
And I was getting away from that. I was out there in the field. I was talking to these people and learning about every ... Like if there was a line in the Excel spreadsheet that looked weird to me, I was finding, I could call the person up who probably created that line in the spreadsheet and talk to them about it. And that level of interaction was a wholly different type of education that you can't teach in the classroom. And so, I was thinking, how do we expand this and get more students this kind of an opportunity?

Gregory DeAngelo:
And that's where the notion of the lab, this Computational Justice Lab came from. And I've been fortunate to get ... I should say we, because there's a whole group of faculty now, have been fortunate to get support for the lab that's enabled us to really impact the lives of now we're looking upwards of almost 30 PhD students ...

Jeremy Byrum:
Wow

Gregory DeAngelo:
... that have been, meaning they're now alumni of the lab, or are currently involved in the lab and get crazy access to cool data, need opportunities to go hang out with different law enforcement or prosecutors' officers, different types of criminal justice agencies, and are really making an impact on the community, which is perhaps the most important part of all of this.

Jeremy Byrum:
Definitely. Now I mentioned it a little bit in the intro about what the lab does, but can you elaborate a little bit on ... Again, we talked a little bit about your mathematical and data science background, but can you go into a little more detail of exactly what the lab does, maybe not naming agencies, but the types of agencies you guys work with and what your goal is as a research institution, research entity.

Gregory DeAngelo:
Yeah, absolutely. Sure. And so, the quick part that we can answer that, which is we do the same things that happen at any other academic institution. Our students take classes, we get them trained up on whatever social ... It's typically social science degrees that they're getting. So we're going to get them the formal training in the classroom. That's going to happen for sure. The second part, and that's maybe a little bit different than a lot of other social science degrees, is our students are going to take classes.

Gregory DeAngelo:
And then we have lots of small, little informal training sessions in lots of different computer programming languages. And so, this is where the more technologically advanced listeners are going to ... They're going to be trained in the ARS Data, Python, SQL, GIS, all these different programs. The idea is we're not going to just expect them to learn that as they go. We actually train them on these things while they're getting their training on social sciences. So now they're really well equipped. They're like a data scientist, is the new buzzword.

Gregory DeAngelo:
The big difference for a lot of our students is that they're trained in data science, but they're also trained in this thing called causal inference, which is not the age old adage that correlation doesn't apply causation. Our students are going to dive in deep on that causal side, finding that A causes B, just a much higher standard, research standard to put in front of them. So there's that part of it. And then I'd say the third part is, and I'll elaborate a little bit more on this part, is that we work with lots of different agencies and law enforcement agencies, prosecutors agencies.

Gregory DeAngelo:
We work with different social services. So we work with a lot of these different agencies and the needs of these agencies are varied. Sometimes it's something like we've introduced a new policy and we don't know if it's a good one or a bad one. Can you evaluate it for us and tell us ...

Jeremy Byrum:
We, as in the state of California or as in ...

Gregory DeAngelo:
Yeah, it could be something that's statewide. It could be something that's countywide. Or just a local police department, for example. Lots of policies are getting enacted at every level, local, all the way to federal level or international level, like their policies are getting enacted all over the place. And one of the things that's just doesn't exist, and I would guess more than 99% of all agencies, whether you name the government entity, and it probably doesn't exist is a team that evaluates if what they're doing is good, if what they're doing is achieving what they've set out to achieve.

Gregory DeAngelo:
Most agencies don't have an evaluation group. And so, that's where we come in. And we'll work with the different agencies and say, hey, you've enacted these policies or you're trying out these new strategies maybe to ... If it's a police department, maybe you're trying out a different patrolling strategy to try to enhance community safety. Are you? Are you making the community safer? Is that new policy, is that new practice that you're implementing actually improving the community's wellbeing or not?

Gregory DeAngelo:
And the reason is that you'd want to know that, of course, is if it's expensive to implement new policies or if it's costing you more than the old policy was. Hopefully, you're getting something for that in the form of improved safety. So these are the sorts of things that we dive in and the faculty will leverage our relationships with different agencies. And through those relationships, we'll bring students in. And it's great for them because they're now learning, not just how to perform research, but how to take the knowledge and the skills that they've learned and communicate that to non-technical experts.

Gregory DeAngelo:
And explain to the practitioner, explain to the people who are running these agencies, this is what we're going to do, this is why we're going to do it that way. And just there's all of these important forms of communication that you learn when you're trying to bridge the gap between being a researcher and who sits in a silo, versus being a researcher who's out trying to work with these practitioners to evaluate different parts of their agency. Or in other instances, trying to actually develop tools for them that help them do better in their jobs.

Gregory DeAngelo:
So that's a whole part that ... There's no way to teach that in the classroom. You got to get people sitting in front of and talking to these practitioners. And so that's, I think, one of the best parts of what we've got going on, is I'd say half of our lab is engaged in these types of interactions at any point in time.

Jeremy Byrum:
Right. Now, clearly, I mean, as you've said, the lab has been very successful. You've had over, I think you said, 30 student to either are currently in there or have already graduated from CGU. And you mentioned earlier as well that groups that have this, or like you said, 99% of these agencies don't have an evaluation group to come in and do what you guys do. So besides that, besides you guys filling that gap, what is it that the lab brings to the table that an internal audit or someone else looking at the discrepancies in the criminal justice system, what is it that you and the students in the lab bring other than just the data expertise. Are there other advantages? Is it taking bias out of the equation? What is it that you guys bring to those agencies?

Gregory DeAngelo:
Yeah. So I think there's two answers to this question. Let's imagine that everyone had the same sort of computer based skillset, that technical skillset. The training that we offer, I'm going to come back to this, the causal training, that's really important. The extremely important part of the equation is that when we sit down, we're not just some person who knows a little bit of statistics and can mess around on a computer. We're adopting a framework and we're going to learn every single piece of nuance about this agency, as we call, institutional details.

Gregory DeAngelo:
Learn all the important institutional details about this agency to ensure that we can get at the causal relationship in whatever policy or practice that we're going to be evaluating. So that's the first thing, and I can come back to that later perhaps. The second thing, which I think is equally important, but I think is more to the point is we're external, we're unbiased. For us, we're not going to just produce the result that the agency wants.

Gregory DeAngelo:
And in many instances, I just want to be clear about this, the agencies don't want us to just produce a ... If there are problems within their agency, to me it's been the most surprising part of my interactions with these agencies, is that they'll say, we don't know if there are problems. We don't know where they exist. That's why you're here. That's why we're working with you, is we want you to tell us what we're doing wrong or where the problems exist in our agency. And then we want to work with you to try to come up with ways to fix those problems.

Gregory DeAngelo:
So I think that third party, external, unbiased component of who we are really is attractive to the agencies because it's not someone internal that in a lot of ways, we don't know all the office dynamics, we don't know the political details of who's trying to get a promotion, and all that stuff. We're blind to that. We're just there to sit down and say, okay, we're going to look at this policy and evaluate it. And we ask a million questions. And I think a lot of times, agencies are like, the heck are these ... Why are they asking all these weird questions?

Gregory DeAngelo:
But through that process, I think they learn, oh, wow, these folks are really digging in to learn every last detail they need to know to be able to say that this policy is causing this outcome, versus this policy is correlated with this outcome. So I think those two reasons, we, as outsiders, are often welcomed into the fold.

Jeremy Byrum:
Right. Now getting into the causality. Now you've mentioned this word several times. Could you go into a little bit more detail on maybe an example of where one might correlate to things, I don't know, such as crime statistics in certain neighborhoods and how many police are called to those neighborhoods, versus are those correlated or are those causal? What kind of examples in the training does that show? Or what shows up in the training in how you discern between the two? Because that sounds really interesting.

Gregory DeAngelo:
Yeah, absolutely. I mean, and to me, I think it's really interesting. And to me, it's the crux of a lot of what we're seeing in terms of the social movements that are out there right now with regards to the criminal justice system. One of the most difficult problems, and I don't know what the listenership is like, so I'm going to try to keep this at a level that is suitable for everyone. But one of the things that's tricky is, and hopefully unsurprising, is that things don't happen just at random in the world.

Gregory DeAngelo:
And police aren't just randomly allocated and prosecution isn't just ... Prosecutors aren't just randomly deciding to press charges against people. There's a lot of decisions that are made. Police agencies are making decisions about where to go patrol. And in a lot of ways, that's reactionary like, well, where is crime happening? Alright. Well, there's going to be more police in those areas. If there's going to be more police in those area, there's likely going to be more arrests in those areas.

Gregory DeAngelo:
If there's more arrests in those areas, there's going to be more prosecution, there's going to be more charges pressed. Those people are more likely to go to jail or to prison. And so, this is the trickiness, is these decisions that are made are nonrandom. If it's the case, that crime rates are higher in one part of the community than in another part community and more police are placed in those parts of the community, then there's going to be more arrests made in those parts of the community.

Gregory DeAngelo:
And if there's more arrests made in those parts of the community, there's going to be more prosecution. Now, what happens is this is an issue of what we're ... Actually, let me say this differently. What we're observing could be that there's actually higher crime in those areas. And that's why there's more police in those areas. It could be that there's some other force, there's some other reason that we're putting more police in those areas.

Gregory DeAngelo:
And we don't observe that as a researcher, but the end result might be that it looks like the police are disproportionately policing in certain communities than they are in other communities. And if you fast forward in the process a little bit further and you were to hone in on prosecutors, for example, you might say, well, then it looks like the prosecutors are being maybe biased toward one group because they're prosecuting more crimes in one part of the community than in another part of the community. And therefore you might start to use words like bias.

Gregory DeAngelo:
You might say, well, it appears that prosecutors are more biased, right? And so, it looks like the prosecutors maybe are going after like one subset or one ethnic group in the community more than another. But that's a problem if you were to just sit and try to run research project, because prosecutors are downstream quite a bit. A lot of decisions are made before an arrest is made and then suggested charges are made. And then those charges end up at the prosecutor's office.

Gregory DeAngelo:
And the prosecutors make decision based on the evidence put front of them, whether or not they decide to proceed forward with the charges. Well, if there's already bias in the system by people who are involved in this arrest upstream of the prosecutors, then you would be falsely attributing the prosecutor of being biased. When, in fact, that's not their fault. They inherited. Maybe they inherited that bias, and you can go further back. Okay, the police. Maybe the police ... You maybe would say, well, the police are biased. And they could be, I don't know.

Gregory DeAngelo:
But it could be that the police got a bunch of information from a dispatcher, and that dispatcher got a bunch of information from a call taker, a 911 call taker. And that call taker, maybe they were biased. And maybe when they were on the call, when they received the call, they asked a whole bunch of inappropriate questions and then conveyed a whole bunch of information to the dispatcher who then sent the information to the police officer that was somewhat racially motivated.

Gregory DeAngelo:
And that primed that police officer to show up to the call with a heightened level of concern and maybe go a step further than they would have otherwise and make an arrest against that individual. So, I say all this. If it sounds complicated, my response is by design. It's that there's these layers of complexity, there's these layers of decisions that are going on in the criminal justice system. And it makes it so that we can't just conduct really simple basic analyses on the system because there's so much bias born into the system.

Gregory DeAngelo:
So what we end up having to do is go out and search for randomness in this very nonrandom world. And so, for example, one of the things that we're getting closer with on some research is, well, when cases show up at the DA's office, it turns out there's one person who ... Or not one person, a group of people who are the very first people that touch those cases. And so, what the clerical staff does is that they'll get a whole bunch of envelopes basically in from different law enforcement agencies who have made arrests, got all the evidence together, sent that evidence over to the district attorney's office in a packet, and have said, we think that you should file these charges for this particular incident.

Gregory DeAngelo:
As it turns out, those envelopes get handed out to these prosecutors that are called review prosecutors. They get handed out a lot like you'd expect to see a Las Vegas card dealer handing out cards for poker. They literally get shuffled. Like you get one, I get one, the next person. And they just ... One for you, one for you, one for you, one for you. And they just randomized this process. Well, that's the kind of thing we're searching for to say, okay, whether Jeremy got that case or Greg got that case or Alice got that case or Mark got that case or Anita got that case was as good as ran random.

Gregory DeAngelo:
Now, let's ask how did those cases that are as good as randomly distributed, how do the prosecutors proceed from that point forward? Now we're at a place where we can start to ask if bias is being introduced into the system by a prosecutor, because it was as good as random that I got the case versus you getting the case. There was nothing systematic. There was no strategy involved. We didn't know what was inside that envelope. They just doled these things out to us.

Gregory DeAngelo:
And based on that allocation decision, we can say, okay, now let's see if Greg is behaving differently than Jeremy. And if so, what impact is that having on justice, basically, on the way that justice is doled out to the community. Is it being handed out asymmetrically? Or does it appear that agencies are handling things in a consistent and what we would probably deem reasonable way?

Jeremy Byrum:
And that sounds to me too like there's an element of just sheer volume sifting through cases prosecutors get, and I'm assuming some of the work you've done, I mean, in quantitative data shows some of that volume perhaps.

Gregory DeAngelo:
Yeah. So one thing I would say, so let's take California, for example. We can go through this history a bit further back if we want. But more recently, there's an act that passed called the Racial Justice Act. And it basically said that if a minority member of the community received a sentence or was treated differently than a non-minority member ... I mean, I'm really, by the way, given the cliffs notes version of this, there's more nuance. It's a policy that has a lot more nuance.

Gregory DeAngelo:
But basically, if there are differences in the way that people are being treated based on their ethnicity, then we want to write that wrong. That's effectively what the act is trying to get at. And I'm working with some agencies in California that are dealing with this, prosecutors agencies who are dealing with, are we doing things wrong? I mean, that's the question they start with. Well, are we engaging with the community in a way that we're being systematically inappropriate toward one racial group? If so, we want to fix that.

Gregory DeAngelo:
But when I start the conversation, they're not saying that's definitely not happening here. That's not at all the way the conversations start, which may surprise a lot of folks. It usually starts with, Jeremy, back to your main point, like we have so many cases coming in the door that we don't have time to look at like, oh, what's this person's race? And, okay, based on the race, I think I'm going to charge differently on this case than on some other. The volume is crazy. I mean, they will actually, at times, get so backlogged in terms of filing cases that they have ... I've heard this from a few different agencies, they have something that's colloquially called a filing party where they will ...

Jeremy Byrum:
Wow.

Gregory DeAngelo:
... literally pull attorneys who are supposed to be doing other things in the office, they'll pull them in and say, look, the stack of cases that we're back here are so backlogged, we need to pull people in to help us out in terms of getting through our backlog. And so, that's the place where a lot of times the conversation starts, is I don't even know how we could be biased because we're so backlogged, that we don't have time to look at those details. We just flip right to the facts of the case. And then we read that over and then we make a decision based on that.

Gregory DeAngelo:
So that's the first thing I'd say. And then also, pushing along the Racial Justice Act, and I think that probably for some of your listeners, this is an area that is a hot button issue at the moment. So it probably speaks to some of what people are curious about in the criminal justice system. I would say that my own experience, and this is work that the lab is continuing to work on at the moment. And some other work that's been produced out of a different lab, that if you want to share that with your reader, I don't know if you can put that in the notes.

Gregory DeAngelo:
But one of the things that we've seen is that removing the race from the incoming information, meaning the case that's being brought forth to the prosecutor's office, like removing race altogether so that you get to this so-called blind justice or these blind charging decisions doesn't seem to change charging practices whatsoever. And I've found that in my own work and other people are finding that in their work as well. And I think, getting back to your point, this has everything to do with just the volume of cases coming through the door to be biased.

Gregory DeAngelo:
Even if it's implicitly biased, that could play out in the data. And we're still sorting that out. But to be explicitly biased, I think, would be ... You'd get behind compared to your counterparts if you were to engage in that behavior, because the volume is just so overwhelming for these individuals. That's been my experience. That's what's falling out of the agent out of the work that we're doing with various agencies thus far. There's a lots of different ways to get at these things.

Gregory DeAngelo:
Some of the results we continue to find, which I would say has been one of the bigger surprises of my research career because I fully anticipated that I'd be having conversations like, yes, your charging practices don't look so good and we need to make some adjustments. And in fact, we're not finding so much of that evidence.

Jeremy Byrum:
Right. Now, when it comes to using this data, as you've mentioned, some of these agencies either don't know if they're doing something wrong or they claim that they don't and you go in to actually see if there is something that they're either doing incorrectly or doing something due to bias or something like that. When you publish a research study, when you publish a research paper from the lab, what is, I guess, the goal of that? Are you just setting out to just publish the data and have people interpret it as they see it?

Jeremy Byrum:
Or are you trying to have other agencies use that data so they too can see what are the issues amongst either their own practices or the criminal justice system as a whole? Of course, that's not something you can "fix" overnight. Rome wasn't built in one day. The criminal justice system isn't something that, unfortunately, can't be fixed in one day. It can't be made perfect. But what is the goal? I guess, when you publish a research paper, when you go into work with the agency, how do you use that data to, I guess, improve other areas or other agencies?

Gregory DeAngelo:
Yeah. Great question. I want to back up one step and say, these agencies aren't perfect. And if I've coming across that way, that would be wrong on my part. There are issues. And when we raise these issues, the question then becomes, well, how do we fix them? What are potential ways we could fix them? That sometimes becomes an internal conversation for them or an HR problem or something along those lines. And that's fine. They can sort that part out.

Gregory DeAngelo:
But I think the first thing that's important as far as why we engage with these agencies is that when we approach them, the whole idea is to enhance community safety. And my fear is that a lot of these agencies are operating somewhat blindly in terms of understanding. If we introduce this policy, if we change this policy, or we change this practice within our office, a lot of times, they don't know if, well, is the current practice that we're engaged in a good one, meaning it's not introducing bias. It actually seems to be pretty good at enhancing the safety of the community.

Gregory DeAngelo:
And are we getting ready to change our policies or practices toward a less desirable outcome? And they don't know the answer to that question because they're operating in the dark. Because again, as we talked about earlier, they don't have an evaluation group within their agency. That's the first thing. That when we approach them, it's about community safety. It's how do we make communities better by helping these criminal justice agencies patrol, if their police patrol in a smarter way, make more consistent decisions within the prosecutor's office.

Gregory DeAngelo:
If it's the jail and they have to let some people go, let some people go early, who are the best people to let go early, who tend not to go out and commit more crime, and for that matter, more violent crime? These are the decisions that a lot of these agencies are saying, "We've just been working off intuition or gut instinct or experience." That's another one I get a lot. Well, based on our experience, I'm like, "Well, have you ever evaluated your data?" And the answer is like, "Well, no. We don't even know how to do that kind of thing." That's the first thing.

Gregory DeAngelo:
That's the reason we even start the research. In many ways, the publication is just all these efforts coming to a head and saying like, here's what we did and here's what we learned. The hope though of the publications, besides letting the world know that we've engaged in this work and hopefully other people have read it and they think it's good work and that they think that it's helping us to understand the criminal justice system and the way that it interacts with our communities.

Gregory DeAngelo:
Beyond that though, there is another part to it, which is advertising to other criminal justice agencies that are out there that maybe aren't engaged with the research community yet about the value of having these practitioner research partnerships. And so, that's, I think, for me, one of the main things I'm trying to get out there, is we in the research community, there are some folks that I think have a very clear agenda and they maybe just want to produce a result because that's what feels right to them.

Gregory DeAngelo:
But there's many, many, many really good researchers out there that are saying, we don't know the answers to these questions, and we'd like to work with you to help you learn the answers to these questions. And so that we can start to understand, does what we've learned working with an agency that's in a very urban area apply for more rural areas? Or do we need to go engage in research in those rural areas and find out whether or not the answers differ?

Gregory DeAngelo:
And so, to me, we're all contributing to this broader knowledge base that we're slowly building up about the impact that the criminal justice system is having on the community. And I think if we do that in a really fair and honest way, it's good for everyone. It's good for the agencies. It's good for the community. And this is where I think one of the few areas where I think the academic community could be a really trusted entity to act on behalf of the community and the communities that are feeling like they've been targeted and impacted.

Gregory DeAngelo:
In many ways, we are acting on their behalf to go into these criminal justice agencies and say, hey, look, is there bias? Are there issues here? If so, let's identify them. And let's all work together to come up with solutions so that everybody ... Because again, even the criminal justice agencies ... I shouldn't say even, I should say also, the criminal justice agencies are worried about negatively impacting any one community in a bigger way than any other community. That's overwhelmingly been my experience. And it's not just lip service.

Gregory DeAngelo:
They hand the data over that could show them, that could very well show that they're engaged in practices that are biased. And I think they're doing that because they're worried that that could be the case. And if it is, they want to improve on it. But what they don't want this to be, as best I can tell, is a foregone conclusion that they're screwing up, they're doing bad. And so, therefore change absolutely must happen right away. I think they want to be intelligent about how they introduce change so that don't try to fix something that's not broken.

Jeremy Byrum:
Right. And going back to, I guess, ensuring community safety, and perhaps you've already answered this question in one way or another. But why is that research integrity aspect? Like you said, even when you're looking at just quantitative data, of course, you can find data or you can try to find data or create data that supports your argument or supports what you want to do. But when you're trying to remove that, when you're trying to remove that bias, when you're trying to remove that non ... Or I guess not being as genuine as you could be, why does that research integrity become so important as an academic?

Gregory DeAngelo:
Well, the academic community is big, number one. And so, if you're ... Well, let me start even before what I was going to say. The first thing I would say is the academic community, they're pretty good sniffing out if you've done something wrong. And so, this is the peer review process. It's brutal. It can be absolutely brutal. If you haven't, going back to the example I used before, found a situation where it was as good as random, like a coin flip chance, that Jeremy handled this case and Greg didn't.

Gregory DeAngelo:
Or that Greg handled the next case and Jeremy ... You need that randomness to basically say, okay, it was just an act from somewhere else. Lightning struck and you got the case and I didn't. And because you got the case and perhaps you're a little more lenient of a prosecutor than I am, maybe you didn't press as many charges or you didn't press charges at all. You decided not to proceed forward, whereas I would have. And then we can ask a question about, what's the effect of being more stringent or less stringent on the outcome of the case?

Gregory DeAngelo:
Or maybe you're less inclined to proceed forward on charges against the minority community than I am. And then we can ask, well, what's the effect of being disproportionately tougher on the minority community or pressing more charges against the minority community on their outcomes? We can start to get at these sorts of things, but it's that randomness that we require. And we, as an academic community, require to be able to say that anything is even close to causal. And if we don't have that, then their results simply aren't believable.

Gregory DeAngelo:
The academic community will not publish them. Or maybe they would publish them, but they wouldn't get very much attention. And if the goal is to produce research that has impact, then you want to get some attention from your publications. That's the first thing on the academic side. If I don't reach that threshold, that'd be bad news. And I likely wouldn't be able to produce an academic article where I could disseminate the findings. That's the first thing I would say.

Gregory DeAngelo:
The second thing I would say with regards to really keeping a super high level of quality and ensuring that we get a causal story told in the end, is that's the only way that I even feel remotely close to comfortable in advising an agency on how to proceed forward. Given that lives could be impacted, people go to jail or prison and their lives are totally upended. The only way that we're going to proceed forward on making any sort of recommendation is if we achieve that very high standard of, we know, without any doubt, that this is the causal relationship between specific charging practices and community safety outcomes.

Gregory DeAngelo:
And if we don't have that, we're not proceeding forward. Why this matters on a larger scale though? I would say we have lots of policies that are being enacted right now. The pendulum I would say, especially in California, but I think even more generally in the US, has swung in the direction of decarceration, decriminalization. The intent is to not hurt people by getting them involved in the criminal justice system. And by that, I mean not getting them arrested, not getting them prosecuted, not putting them in jail. And from where I sit, I think we're introducing a lot of policies.

Gregory DeAngelo:
And I think the policies are far outpacing our knowledge of the impact of these policies. And that's where I get scared. I mean, right now, where we're at, I mean, in California, between AB 109, which was realignment, they moved a lot of prisoners out of prisons and into local jails, they crowded the local jails. As the local jails became crowded, they started releasing individuals from jail into the community. And there's only a certain amount of community supervision that one can engage in.

Gregory DeAngelo:
And then you fast forward a little bit further to Prop 47 and Prop 57, you've got additional propositions that are reducing sentences or decriminalizing certain behavior. And my experience talking with lots of criminal justice agencies is this is really creating a difficult situation for us in terms of the safety of our communities. And my answer to them is, I don't know if that's a true statement. We would need to test that. And so, I'm really simply pathetic to what they're saying.

Gregory DeAngelo:
Because to use a non-criminal justice example, if you have a child and they're engaging and the kid comes home late at past their curfew and you catch them and you say, I caught you coming home late. I'm not going to punish you though. I just want to tell you that I caught you. And I know that you came home late, but I'm not going to punish you. Would you expect them to stop coming home late? And the answer is like, I was a kid once and I think if my parents didn't punish me for coming home late, I would have stayed out later, longer, and longer, and longer.

Gregory DeAngelo:
Because there were no ramifications of that. And that's the concern that I keep hearing a lot of lately, is these decarceration or decriminalization policies seem to be making it ... That have less of a deterrent effect on people engaging in criminal conduct. Isn't that making our communities less safe? And my answer at this moment, although the lab is deeply engaged in answering these types of questions right now, is we're working on it. We don't quite know all the answers yet, but what you hypothesize seems reasonable. Let's look into your data and check this out.

Gregory DeAngelo:
And so, that's where we've been engaged for the last two, three years since the lab has gotten going, which has been a lot of fun. But at the same time, it's a slow process. Research is slow. And people would like answers yesterday. So, takes a little time, but we're slowly unraveling some answers. And I think we've given a lot of great opportunities to interact with these agencies to be able to provide answers. So I'm really hopeful for what the future will have for us in terms of providing important insights to the various agencies that we're working with.

Jeremy Byrum:
Awesome. Now, we're just at about our time for our show today. Where can they find or where can our audience members find you and the lab's work? Or do you guys have a website? Is there a place we can read your research papers?

Gregory DeAngelo:
Absolutely. The first thing I'd say is we have a lab website, which is the www.computationaljusticelab.org. You can find us there. We've got some Twitter presence, which we can share with you guys and some Facebook presence. The other thing I would say is we have biweekly lab group meetings that are open to the public and the community to sit in on and hear the research ideas we're working on. So if one is interested in participating in that, they can just reach out to me. My email is gregory.deangelo@cgu.edu.

Gregory DeAngelo:
And you can also read up on my research at my webpage, which is gregoryjdeangelo.com. And also, I encourage you guys to check out the faculty and students and interns and alumni that are all part of the lab, because there's just so many great people that are involved in the lab, and so much awesome research going on. We have a couple of junior faculty members in the lab who are engaged in some really cool research. CarlyWill Sloan and Matt Ross are just engaged in some really neat research, which I strongly encourage people to check out. And reach out to us.

Gregory DeAngelo:
You don't have to be an academic to write us and ask us questions. I sometimes get questions and calls from people who are just regular old people, who's hanging out. They've listened to this podcast or seen something that I've written that end up being featured in the newspaper, whatever. And those are some of my favorite exchanges, is when people just reach out to me by email or phone and engage with us. Because I learn so much from the community, especially people who have maybe unique insights into the criminal justice system. We're especially interested in talking with you. If you're thinking about writing, please do. I would look forward to that opportunity.

Jeremy Byrum:
Great. Well, on that note, we'll let you get back to work.

Gregory DeAngelo:
Alright.

Jeremy Byrum:
Sounds like you got a lot of students and are very busy, and we wish you continued success.

Gregory DeAngelo:
Thanks so much for having me.

Jeremy Byrum:
Alright. Thanks, Greg, for coming on. From Studio B3 at Claremont Graduate University, you've been listening to the Campfire. We'll see you next time.