Alex Ciomek

Facebook Twitter Google Print Friendly and PDF

Alexandra Ciomek is a Ph.D candidate in the Sociology Department at Harvard University broadly interested in urban neighborhoods, public safety, and crime prevention. At BARI, she works to better understand the city and the use of its services through work on various data sets from the Boston Neighborhood Survey to 911 calls. We talked to her recently about work she’s doing for her dissertation about crime and gangs in Boston.

BARI: Thanks for taking the time to talk about your work with us! Can you start by telling us a little bit about your background before you started your PhD at Harvard?

Alex Ciomek: Of course! I went to the University of Pennsylvania for my Bachelor’s and Master’s in Criminology. I actually majored in both Math and Criminology. Penn has a great program where they allow you to complete your Masters during the last two years of your undergrad, so I completed my masters in criminology during my time there.

BARI: Was there a specific thing that got you interested in criminology, or was it just a generally interesting topic to you?

AC: A bit of both. I was always interested in crime, and trying to better understand policing and how we can improve policing to better prevent crime. Because of the Criminology program at the University of Pennsylvania, I decided to take a few classes and I really enjoyed them. Then, I wanted to learn more and do more research, so I continued on that path first with the masters and then following through to a PhD in Sociology. I’ve been able to follow through on my research interest in policing and crime prevention, as well as expand to gangs, social networks, and urban sociology more broadly.

BARI: Can you can you walk me through the research you’ve been doing for your thesis, and where that work stands at the moment?

AC: My primary areas of interest are urban sociology, the criminal justice system, network science, and quantitative methodology. My dissertation explores the co-offending network of individuals with contact with the criminal justice system in Boston, especially looking at street gang members and the relationships between them. In that work, I use social network analysis on the data on police contacts to better understand the overall gang member experience and its relationship to crime in an urban context. In particular, I study the relationships between members of different gangs, the meaning of the gang member label, and how exposure to a particularly high risk subsection of an urban network can influence life course outcomes, specifically gunshot victimization. My dissertation is co-chaired by Robert Sampson and Bruce Western.

BARI: What are some of the takeaways you’re finding? How do those social networks affect people’s life outcomes? How does police contact affect people’s life outcomes?

AC: I find evidence that [in Boston] there is actually a strong similarity, though at different magnitudes, between co-offending within gangs vs. between gangs, which shows that Boston gangs might have less cohesion than we typically associate with gangs, often in L.A. and Chicago, where they are more national, generally more organized, and generally larger than gangs in more typical American cities like Boston. I also find that gang members commit more crime overall than other co-offending groups that I find in the social network data. But the differences in co-offending are less so. This shows that current gang classifications may not capture all criminal groups, at least in the context of co-offending, which suggests that others are at risk of both engaging in criminal activity as well as becoming victims. In fact, my third chapter (which is work with Anthony Braga and Andrew Papachristos) does show that individuals socially closer to firearms and also socially closer to gang members are more likely to become victims, and that being close to a firearm with characteristics of illegal trafficking is particularly dangerous.

BARI: Interesting. Do you find significant differences between cities that have multiple different gangs versus ones that they may have fewer organizations within the city? I would imagine that a city like L.A. or Chicago might have more different gang organizations than a smaller city like Boston.

AC: At the moment, I have yet to study cross-city comparisons, but I would say based on the literature that Boston gangs are much more geographically concentrated, often by neighborhood, and generally smaller, so they are more similar to each other in the sense that they are organized in a similar way and potentially have a similar specialization. However, my current focus is not the differences between gangs in Boston. I’m more interested in looking at the relationships between gang members at the individual or dyad level to better understand what people in gangs are doing, to better inform whether and how they may act on behalf of the whole gang, and how this relates to the relationships between gangs more broadly.

BARI: How are you using network science to study this?

AC: My data comes from the Boston Police Department, specifically arrest data as well as FIOs (Field Interrogation Observation reports) which are a form of intelligence gathering or record of a contact based on reasonable suspicion. FIOs happen when an officer thinks there’s reasonable suspicion of a crime happening. They approach two or more individuals individuals and record that were together. It also happens if officers would like to gather intelligence on who’s hanging out with whom. Both of these sets of data link individuals. In the FIO sense, they’re in the same location for the same reasonable suspicion event or intelligence gathering event. And for arrest, if two individuals are arrested for the same crime, which requires probable cause, then they’re considered to have engaged in the crime together. All of these events link individuals to one another. When two individuals are involved in the same event, there is a tie between them, and these ties can be built out into a full network that is essentially a spider web of individuals connected to one another through shared police contact events. I then use that network to better understand who is in which gang, and when and how ties span gangs and cross gang boundaries. So, if two individuals are the same gang then their tie is considered a within-gang tie. If two individuals are not in the same gang, but still both gang members, that is evidence for a tie between members of different gangs.

BARI: Are there policy implications or policy recommendations that you’ve come up with out of this work? I know a lot of your work is sort of understanding how this stuff works, but do you have something to feed back into policy?

AC: Identifying patterns of gang involvement in crime provides more information for the formulation of appropriate crime prevention strategies. And these strategies can both reduce crime and also prevent both gang members and other offenders from suffering the detrimental effects of criminal justice system involvement. Given that involvement in the criminal justice system can negatively affect offenders in multiple domains, including unemployment, health, and social outcomes, my studies are a step toward shaping urban policy to improve these outcomes. Therefore, understanding how gang members relate to each other and through what activities, including FIO activities but also specific crimes in the arrest data, which I also look at, we can better understand the crimes are more likely to be between individuals of different gangs (compared to between individuals in the same gang) as well as the crimes that are more likely to involve more than one individual in general. We can better formulate strategies to prevent those crimes specifically, which could otherwise lead to escalation over time. Previous research has found that the more individuals involved in the crime, the more likely it is to become a violent event. Maybe not all the relationships involved are positive or non-contentious, and therefore there’s a likelihood of not only violence between the people involved in the events, but also for people to join together and engage in violence against others.

BARI: So once you understand the structure of the relationships, and the structure of who is where, with whom, committing the crime, it can have a predictive power of what kind of scenario is most likely to occur?

AC: Yes. And also looking at the network, you can understand if there are a lot of connections between members of different gangs, maybe we don’t need to focus on one gang, but we need to focus on the people that are between gangs or bridging gangs. The same goes for looking at the whole network (beyond just gang members), and the links between sets of people. The links are called brokers in the network literature. If you can concentrate on them or concentrate on the types of activities brokers do, then you can try to sever their ties with others, which may help to decrease crime and prevent future issues.

BARI: That makes sense. I know you’re working with the Boston Police Department to get the data that you’re basing all of this research on. Have you worked with them on any implementation? Have you worked with them to feed any of this research back into any of their on-the-ground work?

AC: I’ve presented work over the years to them, both my own work and the work with Anthony Braga and Andrew Papachristos on firearms, their role in the network, and their influence on gun traffic conversations. Especially the Boston Regional Intelligence Center (the BRIC) is interested in hearing more about social network analysis and trying to use it. The analysts there are using social network analysis on different scales to better understand relationships between people to inform the formulation strategies.

BARI: I know never to ask a PhD student when they’ll be done, but I know you’re writing a lot of the meat of your dissertation right now. What’s next for you? Do you intend to go into academia? What’s what’s your career plan after this?

AC:  I’d like to continue on in academia. I’m mostly looking for postdoc opportunities at the moment, so that I can have more time to move along with this data into the next project, where I will place this network in space and time and examine how this contextualized network influences where crime occurs over time. To do that, I think a postdoc is a great opportunity on the path toward future career opportunities in academia.

BARI: Is there any other work, either being done by your colleagues or other work you see on the horizon, that is as exciting or interesting to you? Beyond your own work, what’s the cutting edge of this stuff?

AC: I’m very excited by the work being done by Andrew Papachristos and many of his colleagues, where he looks at connecting neighborhoods through typically co-offending, especially co-arrest. This work aims to understand how one neighborhood is linked to another through individuals who commit crime together, hypothesizing that the connections may offer pathways across neighborhoods through which information and support, in either a criminal or non-criminal sense, can flow.

BARI: Thank you for talking to us! How can people learn more about your work?

AC: My website is and my e-mail is

Published On: March 12, 2020