When did you start thinking about the importance of institutions and how they work when it comes to making people’s lives better, and about policing in particular?
My original intent going into getting an MBA was to promote entrepreneurship in Latin America. There was a very clear focus in my mind that if we wanted to promote development in Latin America, we needed to promote entrepreneurship. There was just not enough venture investing in Latin America. That was one of the things that was missing for startups.
I got a job in venture capital and as I was pursuing my MBA, I started working with some faculty who were doing work in institutions and entrepreneurship. That had a massive effect in reframing how I was thinking about this. I started thinking that the presence or absence of venture capital is not something that just happens spontaneously; it’s responding to something else that is deeper and that is affecting both entrepreneurship and venture capital, and that is about the institutional landscape.
That’s why I decided to get a PhD. I realized that the depth of the questions that I wanted to answer went beyond what I was working on as an MBA student and what I would have been doing as a venture capital employee.
As I started studying institutions, what became really interesting to me was that most of the focus on institutions has been to explain how they functioned and why they were so stable. That’s indeed why we like institutions. They provide stability. They provide certainty. But that means that by definition, they’re difficult to change. If you’re in a place that has a good institutional setup, that’s wonderful. If you’re in a place, like Latin America, that has crappy institutions, then you have a problem. You’re stuck in a crappy equilibrium.
I wanted to focus on this question of institutional change. We know that institutions do change because we see it. How is it that people go about changing institutions?
My entry into policing was, in some ways, similar to my entry into institutions in general. I was looking at the big problems that are affecting Latin America. For the past 20 years, there has been a real increase in the extent to which violence and especially organized crime violence is affecting the region, especially in Mexico. I started looking at why violence has increased so much. Why are criminal organizations changing their behavior, all of a sudden, and becoming so much more violent?
I didn’t make this connection at the time, but it’s analogous to this entrepreneurship versus venture capital issue. When we were looking at violence, we’d been looking at the most evident component of our problem in Latin America, which is violent behavior by organized crime. It turns out that that is also just a consequence of something else.
There’s actually a lot of organized crime in the U S as well. If there wasn’t organized crime, then there’s no way you could explain why there’s such efficient distribution of illegal drugs in the whole territory. But somehow, even though there’s a lot of organized crime in the U.S., it’s not as violent as organized crime in Mexico or in Honduras or Guatemala. So it cannot be that organized crime itself explains violence. There’s something else that’s happening. If you compare the U.S. and Mexico, they have the same guns, the same money, the same drugs, even the same organizations, sometimes. Yet there are very different levels of violence. That’s because the true problem is not an organized crime problem, but a public safety problem.
If you unpack what’s going on and why organized crime behaves so differently in the U.S. than in Mexico, it’s because the institutional setup for public safety is very different in these two countries. In Mexico we have enormous weaknesses in our institutional setup when it comes to public safety, especially at the local level, and therefore the country’s ability, and the entire region’s ability, to constrain the behavior of criminal actors is limited.
If I’m a criminal organization and I’m trying to grow my business, and I see a weakness in the institutions of public safety, of course I’m going to exploit it. Of course I’m going to try to make them weaker because the weaker they are, the freer I am in how I conduct my business. Organized crime organizations in Latin America amplify and take advantage of the weaknesses in public safety, but they don’t create them.
The first link in the chain of public safety is police organizations. Those are the ones who are in charge of controlling and regulating what happens in a town or in a city. They are the first responders. Of course, there are other linkages in the justice system, including prosecutors and the penal system and the court system, but without a functioning police, you don’t get to all the other steps. That’s why we focused on police organizations.
When you’re looking at projects and deciding what to work on, do you have a set of values that underlie those choices?
I want to study things that are real problems. I don’t care about just publishing something because some other academic’s going to think that’s a neat theory. I always ask, is this an important problem that is affecting somebody else?
The second thing is I want to study issues for which the perspective that I bring to the table is somewhat novel and therefore can actually be useful. There are many, many big problems out there. One of the ways in which I choose whether I’m going to invest my time in a problem is whether I think that the perspective that I bring, which is this overlap between organizational theory and institutional theory, can actually help.
In the case of police organizations, for example, we’ve known that policing is a tough problem for a long time. It’s a big problem in the U.S. and it’s a big problem in Latin America. When I looked at most of the work that is being done around policing, I noticed that the organizational and institutional perspective is mostly absent. People tend to study policing at the individual level. They tend to look at individual police officers and where you find them and how you equip them and how you deploy them. If there’s a case of police misconduct, we keep having this bad-apples narrative, that it’s just this bad individual.
If you think about it for a little bit, the vast majority of the behavior of a police officer is going to be determined by organizational factors, starting with who their boss is. Tell me who your boss is and I’ll tell you a lot about how you’re going to behave as a police officer. Tell me how you’re evaluated. Tell me how you’re rewarded. Tell me who gets promoted and why. Tell me who gets fired and why. If you tell me all these things—which have nothing to do with individuals, but have everything to do with the structure—I can predict a large percentage of your behavior. I do care about theory as well, so I try to find an overlap between real problems where my perspective can help, but also where the problem can inform how we’re thinking about these issues in general from an academic perspective. If possible, I prefer to look at developing countries just because I think that there’s less research happening in developing countries, and we need people to look at those places.
Has the pandemic and everything else that’s been happening this year changed your focus and priorities?
When we decided to work on policing, we spent a lot of time just raising money so that we could do this well. That allowed me to create a little lab that’s based in Mexico that is doing this work all the time. On the one hand, the thrust of our work remains the same. The focus is how to help police reform and to understand what a police organization should look like. How do we create police organizations that are effective, but that are also resilient and that are trusted by the citizens? That has remained our focus and will remain our focus.
At the same time, the pandemic brought some additional challenges—some interesting questions and some unique chances to make contributions. When we think about the pandemic, we think a lot about nurses and doctors as responders who are deeply affected by this. We don’t tend to think about police officers, but if you think about it for a second, police officers are incredibly exposed. They, in many ways, are more vulnerable than doctors, because doctors and nurses can be very careful in how they protect themselves; police officers, not really. If you’re going to be on the street all day, you can’t be wearing full PPE equipment all day. If somebody is behaving in an aggressive way and you need to arrest them, you don’t get to choose whether you touch them or not.
One of the things that we did is redeploy our resources to get a sense for the vulnerabilities and challenges that police organizations were facing given the coronavirus. We already have this research infrastructure, so we could reach out to all the police organizations in Mexico and ask them, “How have things changed since the pandemic started? What are specific challenges that you’re facing? What are vulnerabilities that your police officers have? Are you receiving enough help? What kind of help do you need?” We were able to collect those data very quickly and that provides an opportunity both to research something interesting and also to provide some useful information to policy makers.
It turns out that some police organizations, during the pandemic, have been able to generate more trust with citizens because they’ve actually reached out and they’ve become more protective and more useful with citizens, whereas for other police forces, as we’ve seen in the U.S., the crisis has actually either exacerbated or created real issues of trust between citizens and the police.
That’s raises a really interesting question. It’s the same pandemic and it’s hitting everybody in similar ways. Yet some police organizations are able to leverage that moment of vulnerability to create a tighter relationship with citizens whereas others are driven further apart from citizens by the pandemic. That tells us a lot about how you go about building trust with citizens, about what it means to be a protector of citizens.
Then there are more practical things. One of the research streams that we’re developing right now is on community policing. We were planning to do a randomized field experiment that required us to do a lot of field work to help different police forces in Mexico enact different practices for community policing. One of the core components of the methodology that we were going to be testing is organizing neighborhood meetings between police and citizens. Obviously, that’s out the window the moment the pandemic hits.
What we’re looking at right now is whether it’s possible to have community meetings online. Who would show up? Does it constrain who shows up or does it open up who shows up? We are now beginning to experiment with this question of whether police and citizens interact virtually and what kind of engagement is possible. In some ways, it’s worse, but in some ways you can imagine that it can be much better because if you do establish a good virtual linkage between citizens and the police, then it’s actually much easier for citizens to provide information about what’s happening in their neighborhood in a way that they feel is less threatening and that is more fluid.
We’re now experimenting with these virtual approaches, just to see if it’s possible. If it is, then we’re going to do a randomized control trial about virtual community meetings.
There’s a set of work that hasn’t changed at all. Before the pandemic, we did a randomized control experiment on training police officers by giving them a toolkit for how to have more productive interactions with citizens. If you’re out on the street and you’re going to interact with a citizen, what should you, as a police officer, bring to that interaction and then what toolkit can you use to make sure that that’s a productive interaction? Regardless of whether it’s a traffic stop, you’re helping somebody cross the street, or you’re arresting someone, there’s a professional way in which you should enact what you do that should always build trust with citizens and everybody who’s watching that interaction. That’s what we train police officers in.
We first did a randomized controlled trial with just a thousand police officers, to see whether the methodology works. We were able to document big effects on a number of things. Based on that, the Mexico City police asked us to help them escalate this training to all of their 90,000 police officers. They have to keep on training police officers, regardless of the pandemic, because they can’t stop their work. That means that we have to help them. That work of scaling up the procedural justice training to the whole Mexico City police force has kept going. In a similar way, the federal government of Mexico asked us to weave our insights from all our research into a national policy that they’re enacting right now to set up a set of national standards for policing, and we’ve, obviously, continued working with them. Everything that doesn’t entail actual field work of my RAs on the street, we’ve tried to continue as best as possible.
Creating effective and trusted police organizations has become even more pressing in the United States. If you were speaking to the Association of North American Police Chiefs and you had just a couple of minutes to tell them what they should be thinking about, what would you say?
The first one is we need to think of police organizations as integral organizations. You can’t just change one little thing about how your police organization operates and expect that to work. Police organizations are complex organizations, and you need to visualize them as complex organizations. If you want to change one practice of police officers here, you need to think about how that’s connected to everything else that you do.
For example, if you want to change the way police officers interact with the public, you can’t just train police officers on how to interact with the public. You also have to think about how you’re evaluating that police officer, how you’re evaluating the boss of that police officer, what metrics you’re collecting on whether these police officers are behaving properly in the field. What does this police officer expect is going to get him or her promoted versus fired? What kind of support systems do the police need in order to have a different type of relationship with citizens?
The second big one that comes out of our research very clearly is, what is your objective as a police organization and what do police officers think their main goal is? In general, what we have, in the U.S. and everywhere else where we’ve interacted with police organizations, is a very crime-focused mentality. Police officers tend to think that their job is to catch bad criminals. If that’s your concept of your job, then at the core of your work is a very small number of people who are responsible for the vast majority of crime, and your focus is going to be always on looking for bad folks and trying to arrest them.
In effective police reform, we’ve observed a reframe from focusing on deterring crime to focusing on helping people feel safe. You move from having the criminal at the center of the police identity to having the citizen at the center of the police identity. If the citizen is at the center of my identity, then my job is to establish a good relationship with that citizen so that that citizen feels protected by me and feels like they can trust me.
A component of that is going to be preventing crime and arresting criminals, but the core of it is having a good relationship with neighbors. It sounds like a small reframe, but it’s actually a big, big change. The way most police organizations evaluate their work today is on crime rates and arrests. That’s how police officers are evaluated. How much crime is there in your neighborhood and how many people are you arresting? Think about the incentives that you’re creating for people when you do that. I have an incentive to inhibit the reporting of crime because at the end of the day, I am evaluated on reported crimes. You’re creating an incentive for me to arrest people. Because we’re training them to focus on criminals, we’re also training them to think about all the ways in which criminals can be dangerous. They bring into their daily jobs this mentality of “I am at risk and I need to arrest people.” They’re not thinking about the vast majority of citizens that are just going about their lives. When you combine those things, that’s an explosive combination.
When we look at police organizations that have achieved good police reform, one of the things that you notice is that they collect a very different set of data. They collect data on how do citizens feel. Do they trust the police? Why are they calling the police? What are they reporting when they call the police? If citizens are only calling the police when there is a murder, then there’s a problem because that means that they don’t trust us to call when something else is happening that they could use our help with.
You see that police forces that have become more citizen focused change everything about the data they’re collecting, how they are evaluating their police officers, how they’re training their police officers, and that really changes the dynamic between police officers and neighbors.
Interview conducted and edited by Ben Mattison.