Evolving Coders

This continues our series of student reflections and analysis authored by our research team.

Evolving Coders

Bridget Dickens

tPP is constantly evolving. Paging through the codebook and manual today, readers will see new additions that were not there two months ago. The final pages of the codebook log the various alterations made every few weeks. Given the complex and ever-changing nature of political violence in America, it makes sense that how we research the topic changes along with it.

The alterations made within the Fall term (September-December 2019) mostly center on the variables and values. “Hate Crime” being added as a new variable in September and the tactic value “Blade or blunt weapon” changing to “Other weapons” in late October are just two examples of the new modifications. While altering the project allows increased accuracy and organization, some issues have popped up. Just because text changes in the manual and codebook, it does not change everywhere else. Hundreds of cases need to be examined and fixed as well.

This process takes up time from members who are already concentrated on coding, checking, and adding new cases. It can be monotonous and sometimes things fall through the cracks. Improving our accuracy and efficiency during this process is essential.  Currently, most members perform multiple functions, with those on the steering committee having even more responsibility. Nevertheless, each person has their own preferences. I enjoy updating cases and source files to ensure they follow the correct procedure. However, my coding partner might hate this task; instead, preferring to add new cases. We would be working more efficiently with coders divided by tasks rather than the level of experience.

Rather than training new members in detail about all areas of the project, they would get a brief overview instead. From there, they could pick a role that they enjoy the most. Roles could be split between adding new cases, completing case starters, updating cases when new information is found on the individual or inserted in the codebook. When changes are made in the codebook, there would be a group of people devoted solely to ensuring the changes are reflected in our completed cases. This method is much more efficient compared to having information updated sporadically or by members who are already busy with other tasks.

While the mission of tPP remains constant, how we code cases does not. The tweaks made over time are designed to make tPP’s research more accurate and informative. As our work grows more complex, it is effective to change the roles and responsibilities of coders as well.

‘USEB98’: The Process of Completing a Twenty-Year-Old Case

This continues our series of student reflections and analysis authored by our research team.

‘USEB98’: The Process of Completing a Twenty-Year-Old Case

Sara Godfrey

On August 7th, 1998, trucks armed with explosives simultaneously detonated outside U.S. embassies in Dar Es Salaam, Tanzania and Nairobi, Kenya, bringing public attention for many to al-Qaeda for the first time. The explosions killed 224 people and injured over 4,500. The bombings took place exactly eight years after US troops were deployed in Saudi Arabia, and were widely believed to be motivated by the extradition and alleged torture of four members of the Egyptian Islamic Jihad (EIJ).

So why did this twenty-year-old case remain only partially coded in tPP? I was surprised to find some of al-Qaeda leaders’ names sitting on our team spreadsheet, colored purple, indicating the case had never fully coded. My partner and I quickly claimed the thirteen defendants involved in the ‘USEB98’ group. At this time, we had no idea what we gotten ourselves in to.

With the help of our team’s legal researchers, we secured a 157-page indictment, and 397-page criminal document detaining the case’s history. With the little research I had done on the case, I found that many of the defendants had been charged but not tried due to their fugitive status, so we began by searching the docket for the word “sentencing,” yielding 198 results.

The docket was huge and overwhelming. As we began to click through the “sentencing” results we soon found that there were far more than only thirteen defendants from the USEB98 case. As we began to work our way through the docket we found seven defendants who had not been listed on any of our team spreadsheets. Now we were faced with a twenty-seven person indictment, twenty of whom my partner and I took on to code.

There was plenty of information on some of the defendants such as Usama Bin Laden, Muhammad Atef, Mamdouh Mahmud Salim, and Sulaiman Abu Ghayth, as these are some of the most well-known al-Qaeda members.

Usama Bin Laden and son-in-law Sulaiman Abu Ghayth

However, it was much more difficult to find information on their lesser-known co-conspirators. We struggled to match these Arabic-to-English transliterated names to the actual defendants, as the docket and indictment had inconsistent spellings. On top of that, many of these defendants had dozens of aliases.

We drew upon numerous resources to successfully code these defendants. The most helpful of which may have been a CNN Fast Facts report from 2019 which gave a timeline of relevant events. This gave us information on the three defendants who were successfully charged and sentenced: Mohamed Suleiman Al Nalfi, Sulaiman Abu Ghayth, and Mamdouh Mahmud Salim. From there, we returned to the docket to get the specifics.

This left us with seventeen defendants who did not appear to have been tried- meaning they would qualify for our ‘charged but not tried’ value. This was due to their fugitive status, and with some deep digging, we were able to conclude that the majority of the defendants had been confirmed killed by U.S. airstrikes in Afghanistan and Yemen. Two defendants had died while awaiting trial/extradition, and one had been extradited to Egypt for other crimes and sentenced to death. The remaining defendants Saif Al-Adel and Abdullah Ahmed Abdullah are still on the FBI’s Most Wanted List over twenty years later.



















Large cases dealing with Foreign Terrorist Organizations (FTOs), such as this are challenging for numerous reasons. First of all, transliterated names can be misguiding, and inconsistencies in spellings create challenges when searching for information such as age, veteran status, and additional sentencing information.

Secondly, remembering to always code according to the codebook, and not your own logic can be challenging, particularly regarding the “Veteran Status” variable. As I read about these defendants’ military involvement and roles in al-Qaeda I was tempted to code the “Veteran Status” and “Combat Veteran” variables as “Former/current member of the non-U.S. military” and “Yes” respectively. However, this would be incorrect as our codebook specifies that “Veteran Status” refers to one’s involvement in an official state-sanctioned military (and one recognized by the United States), not one’s involvement in a militant group. Although logically, these men were veterans of al-Qaeda, this does not qualify as formal military service and veteran status. It is worth noting that this key methodological determination was made in close consultation with tPP’s Advisory Board. In September 2019 Advisors were asked to weigh in on this issue, and in the end, they overwhelmingly concluded that non-state actor combat experience with insurgencies and other armed conflicts does not qualify as ‘veteran’ status within the project.

Lastly, cases involving al-Qaeda attacks need to be examined closely, as defendants charged for one attack may be involved in another, and therefore listed twice on our official spreadsheet. This was the case with a handful of the USEB98 defendants, some of which were also indicted for their involvement in the 9/11 attacks and USS Cole bombings in 2000. This needs to be specified on our spreadsheet by placing a (1) and (2) by the multiple listings to indicate the defendant has been indicted on two separate occasions. This can be especially tricky when there are discrepancies in the spelling of names.

Overall, completing the twenty-year-old USEB98 case was a huge challenge. But as a new coder, my partner Courtney and I felt extremely accomplished upon completing such a large case. By the end of the week, Courtney and I had successfully coded twenty defendants, seven of which were completely new case starters. The completion of the USEB98 case was a great way to begin to wrap up my first semester as a tPP coder.

Media Framing in Terrorism Cases

This continues our series of student reflections and analysis authored by our research team.

Media Framing in Terrorism Cases

Selena Mungur

While researching Ismail Hamad’s case, I noticed a very distinct difference in how media sources depicted the 18-year-old immediately after this crime in January, 2019, versus how even the same media source portrayed him earlier this month when he was released from jail to await his May 2020 trial at home instead. An article, MCSO: 18-year-old terror suspect shot by MCSO charged with 2 counts of terrorism from ABC 15, Arizona News on January 16th, 2019: “An 18-year-old man who was shot by a deputy from the Maricopa County Sheriff’s Office in Fountain Hills earlier this month has been charged with two counts of terrorism and aggravated assault with a deadly weapon, according to court records.”

Here, Hamad is called “an 18-year-old man” who is associated with a terroristic act, a little over a week after the crime. Let’s now look at a quote from another ABC 15, Arizona News article, Fountain Hills teen charged with terrorism released from jail prior to trial, on October 4th, the day Hamad was released from jail: “Ismail Hamed, the Fountain Hills teen charged with two counts of terrorism and one count of aggravated assault after threatening an MCSO deputy with a knife, was released Friday as he awaits trial.”

In this opening sentence, we can see Hamad being portrayed as a teenager who was charged with acts of terrorism, instead of a man who was charged with terrorism. Additionally, the images associated with both articles are very telling as well. In the January article, the shown image of Hamad is one where he has a full beard, and he looks much older (see above), versus in the video clip we see of Hamad being released from jail, he has no beard and is being led to the car (see below).

This difference in how ABC 15, Arizona News depicted Hamad in their January article versus in their October article is an example of how the media is able to frame a story in a different way in order to push an underlying storyline. Aysel Morin explains in her article Framing Terror: The Strategies Newspapers Use to Frame an Act as Terror or Crime, how much of an effect the media can have on how the public views a defendant in a terrorism case with the example of two almost identical shootings where one man was claimed to be an Islamic terrorist and the other was mentally ill. The media coverage of the “Islamic terrorist” focused on the defendant’s identities of being a Muslim American and his ethnicity. Contrastingly, the coverage of the “mentally ill” defendant took the shooter mostly out of the picture, and instead focused on mental health and gun control. This difference in how the media portrayed both of the shooters strongly influenced how each defendant was viewed by the public.

This same idea of using a different media frame can also be applied to Ismail Hamad’s case as well as we see a difference in how a news source is portraying him. Even though with Hamad’s case, the varying news articles are from the same media source and their are describing the same defendant, the difference in their word choice and provided images make a large difference. By depicting Hamad as a teen who was charged with terrorism, the emphasis is on adolescence which is then associated with innocence, instead of the initial portrayals of him being a man, which insinuates that a man is more capable of making their own choices, and thus cannot have the same veil of innocence that a teenager can.

While further research would need to be done on this topic, it would be likely that this difference in media portrayal of Ismail Hamad may affect how a jury views him in his court case. This could then in turn affect his sentencing and overall how he treated within the court system, as we are already beginning to see by the fact that he was released from jail to await his court date at home instead.

Coding: The efficiency of Group Cases

This continues our series of student reflections and analysis authored by our research team.

Coding: The efficiency of Group Cases

Courtney Faber

The Project’s dataset gives a distinction between cases where there are multiple offenders and those cases where the crime is perpetrated by a single person. The variable co-offenders in our dataset is denoted with a yes for those cases where there are multiple defendants charged for the related crimes and a no for cases where the defendant was not conspiring or working with another on the criminal activity. My coding partner Sara and I have encountered and coded both types of cases and we agree that when it comes time to choose the cases we want to code group cases are the way to go.

One of the reasons that group cases are easier to code is because when it comes to locating court documents (for instance an indictment, plea, sentencing, etc.) defendants are often tried together. (a photo example of an indictment with multiple defendants is pictured at the end of this blog post) When defendants are tried together usually all of the information about the crimes and the defendants are on one single page throughout the course of the case. This fact makes the transfer of information into the cells on the coding sheets we work with faster and easier because multiple variables will be coded the same way for each of the defendants.

Furthermore in my opinion group cases are usually more attention grabbing for many reasons – they are often premeditated, more severe, more interesting, etc.- which leads to more news stories and attention going into the case. For the project’s purposes cases with more media focus often yield more data sources and sources that don’t take as much extensive research to find. In sum cases with multiple offenders are less time-consuming to code because:

  1. you don’t have to look for several files for each defendant but rather you often can find files with all the defenders information in one place,
  2. transferring the information onto the coding spreadsheet is easier since the defendants usually have the same information for multiple cells, and
  3. source files are often easier to find for co-offenders because group cases may get more media attention than individual cases defending on the specifics of the case.

To illustrate the efficiency of coding group cases I will discuss a case I worked with last week. It was a group case about 4 defendants associated with the violent rightist organization known as the Rise Above Movement. A quick summary of the case and its sentencing outcomes from the Department of Justice can be found at this link for all those interested: RAM Case.

In this instance there were several defendants associated with perpetrating violence on behalf of their socio political values at several political protests/rallies, but most notably the Charlottesville Unite the Right Protest in 2017. The indictment shows just how easy it can be to gather information on multiple defendants at one time since they are charged and referenced in one document. Looking at the actual data set for the defendants that my coding pair worked on there are several variables that are labeled the same for each of the defendants since the crimes were so related.

For instance: The date and date descriptor is the same for the co-defendants since they were indicted together on October 10th 2018, the case name was the same for all defendants since they were charged together: United States of America v. Benjamin Drake Daley, Michael Paul Miselis, Thomas Walter Gillen, and Cole Evan White, location of the crime (city, state, and country) were all the same, target (targeting people at a public event) were the same, ideological target and affiliation was the same for both offenders, tactic to complete the crime was the same (unarmed assault), the defendants were charged with the same exact crimes, and there are several others coded the same way for each defendant.

As you can see group cases make coding easier for the coder since much of the information, compiled in one place, is the same for each conspirator to the crime. This all makes sense given that co-offenders often have the same ideological values, work together on how to best achieve their aims, and decide together who is the aim intending to affect and why. Had this case been individual defendants rather than a codefendants the time it takes to find information and the time it takes to transfer that information to the spreadsheet would add up to have been way more time consuming. Taking advantage of coding group cases makes the coders life so much easier.

Johnny Saldana, in his Coding Manual for Qualitative Researchers, discusses some of the main qualities that coders should strive to have when coding. One of these main qualities he identifies as being creative. He discusses that being creative in the planning stages of coding takes “a little bit of data and a lot of right brain” (Saldana, 39) and that coders need to be able to think about all the ways to approach a problem before coding to do so in the smartest way. I believe that taking an interest in group cases as a way to make coding more efficient is a creative way for the coder to save time and energy later when considered before choosing cases to code. It’s a creative way to make coding easier.

Works Cited:

Saldana, Johnny. The Coding Manual for Qualitative Researchers Third Edition. 3rd edition. Los Angeles, CA: SAGE Publications Ltd, 2015.

The United States Department of Justice. Three Members of California-Based White Supremacist Group Sentenced on Riots Charges Related to August 2017 ‘Unite the Right’ Rally in Charlottesville. 19 July 2019.

The Intricacy of Ideology

This continues our series of student reflections and analysis authored by our research team.

The Intricacy of Ideology

Emma Lovejoy

Through the focus which The Prosecution Project (tPP) gives to the particular ideologies of subjects, our team members have the opportunity to observe a wide array of philosophies.  In many cases, personal definitions and identities by the subject are available to us; especially when they have kept written accounts.  These personal records illustrate the diversity of thought and worldview, even within an ideological category, and underscore the need for highly specific coding categories to create a meaningful dataset.

According to Andrew Heywood, “[ideologies] provide individuals and groups with an intellectual map of how their society works and […] with a general view of the world” (Heywood, 2017).

In order to provide comprehensive representation, tPP coders must identify both the subject’s “ideological affiliation,” and their “ideological target.”  This combination of factors helps to clarify the way their ideology impacts their relationship to the wider society.  It goes deeper than just “rightist” or “leftist,” or “racist,” to include specific information about why a person of a given ideology would target the people or places that they did.  In all tPP has 7 coding categories for “ideological target” (as of November 2019): Government, Industry, Religious, Identity, General Public, Multiple Motivations and Unspecified. The first 4 categories (Government, Industry, Religious, Identity)contain more than 20 additional sub codes We also use 13 coding categories, including subsets of both “rightist” and “leftist” movements (tPP Codebook, 2019).

The goal of taking the time to make these distinctions is to prepare a dataset which is useful for specific questions, not just broad ones.  As stated on the tPP About Page, we are striving to provide data relevant not just to “the manner in which a defendant is charged, prosecuted, and sentenced,” but that can give a fuller picture.  We hope to build a dataset which can be used to examine trends in the way our prosecutors respond to certain types of ideologies (i.e. “rightist” perpetrators v. “leftist” ones, indicted for similar crimes), as well as the prosecutorial response to different kinds of targets.  How does the difference between intention of harm v. actual harm manifest in the judicial system? What has more impact on sentencing, a subject’s personal ideology/motivation, or the groups they may be affiliated with? These are (some of) the applications of highly specific ideological categories.

Because ideology plays such a role in the backstory of a crime, from informing a subject’s worldview to guiding them to a single physical target for attack, it’s important that our coding team takes the time to look for any eccentricities the subject may display.  While it isn’t subtle, the case of Heon Jong Yoo is a good example of a subject defying initial expectations.

Yoo, a permanent resident in the U.S. originally from South Korea, self-identifies as “the Asian Nazi” in online forums, and attended several college campus demonstrations in Texas in Confederate attire.  Prior to his arrest, he produced web content promoting his white supremacist views. His arrest and conviction were for weapons violations and misrepresenting his citizenship status, but it is noted that people on at least two campuses were actively concerned about his behavior, and the threat he might pose if unchecked.  Because of the uniqueness of Yoo’s positionality, to classify him simply as a “rightist” would fail to describe his actual ideology; that is, without context, such a classification would likely not lead a reader to assume that he is, in fact, a white nationalist, or that he would identify with the Nazi philosophy.  By combining a more specific coded ideology (“rightist: identity focused,) and a description of ideological target (in Yoo’s case, “unspecified,” since his threats were not against one specific group over others,) we are able to produce a more complete and useful database.

Works cited:

Heywood, Andrew.  Political Ideologies: An Introduction. 6th ed. London, UK: Palgrave, 2017.


This continues our series of student reflections and analysis authored by our research team.


Katherine Coate

Fall 2019 was my first semester coding for the Prosecution Project. Throughout that time, something I have been interested in  is coding the ‘Other’ status variable. I had heard the term othering before and knew it had a negative connotation so I was surprised to see it as one of the variables we had to code for. The codebook says this about ‘Other’ Status:




The defendant is marked as othered if they meet or appear to meet any of the following criteria:

  1. Does the defendant have a name not readily understood as European?
  2. Is the defendant Muslim or a Muslim convert?
  3. Is the defendant an immigrant from a  non-Western/European country?
  4. Is the defendant non-white racially as an ‘average person’ would read them (i.e. not passing as white)?

The defendant is marked as Non-othered:

  1. This is used if the defendant is marked as white, non-foreign born, Judeo-Christian and a non-jihadist (i.e. can pass as a white, American-born, Christian).

Within the project, this definition of othering makes sense for the United States where the “norm” is a white christian American.  Some othering cases are obvious such as the case of Naif Abdulaziz M. Alfallaj, a Saudi who trained at an Al- Qaeda training camp and lied to the FBI about it. However, a couple of things surprised me about this definition at first. There is a certain privilege in being able to pass as the norm. I had not thought about. Fitting into this norm is ruined if your name does not sound European though because then people see you as different.

Something else I was surprised about was that gender was not included. At first I believed women would have been included as othered because people see women, especially women who commit political violence, differently. Furthermore, there are many fewer women committing political violence than men. Recently I coded a case about a woman named Patricia Parsons who tried to kidnap someone as part of the Sovereign Citizen movement. While coding the case I assumed women would be a part of the “othered” category but after looking back at the codebook I had to change the value.

However this variable is not about distinguishing smaller groups of people in the data. They are about perception and how the America public views people in general.  Women are viewed differently in American society but not in the same way people of color are viewed differently, especially when it comes to committing political violence. The Guardian says this about othering;

“It is based on the conscious or unconscious assumption that a certain identified group poses a threat to the favoured group” (Powell).

Adding the word threat to the definition makes the coding much easier to understand. Understanding these different views of people who commit political violence is important for understanding the Prosecution Project and why they would choose to include this variable.

Even though othering is a negative and harmful idea it can be very important to understanding prosecution of individuals. If a jury or judge sees a person as an inherent threat without even looking at their crime they are going to punish them worse so they are locked up for longer. Studying these biases and having data to show for them will be an important part of the analysis of this project. I December I finished my first semester of coding and I have learned so much. It has also made me think critically about what surprised me when I first started coding and made me understand why things are done the way they are with the project.

Works Cited

Powell, John A. “Us vs Them: The Sinister Techniques of ‘Othering’ – and How to Avoid Them.” The Guardian, November 8, 2017, sec. Inequality. https://www.theguardian.com/inequality/2017/nov/08/us-vs-them-the-sinister-techniques-of-othering-and-how-to-avoid-them.

The Intricacies of Coding Group Cases and their relationship with Pacer

This continues our series of student reflections and analysis authored by our research team.

The Intricacies of Coding Group Cases and their relationship with PACER

Meekael Hailu

Being a coder for the Prosecution Project, you are exposed to an incredibly expansive spread of cases, as well as a large amount of Group Cases. As I’ve begun my coding journey with this team, I’ve coded a variety of different cases featuring differing inclusion criteria. But an intriguing experience for me was breaking down and coding my first group case.

My first group case was about an 18 person racketeering operation through the Universal Aryan Brotherhood. From the outside, coding a group case might seem like it is an easy task since many of the individuals share the same codes. Although when coding individual or group cases, it’s integral to remain diligent in searching for any deviance in data points to code. I found it crucial during my coding of this large case to try and find as many demographical details regarding each member of the group that I will be coding to maintain the level of detail that I would have for an individual case.

To find the details that are different for each individual, one can hope that the federal indictment would feature said details. One example is the ages of all the individuals. When it came to coding the ages for the members of the group racketeering case for the Universal Aryan Brotherhood, all of the ages were on the federal indictment. But when I dug deeper for more details on that specific case from various other sources on the web, I found additional supplemental members of the racketeering operation that all were not listed on the Federal Indictment.

Another example of the imperative precautions that need to be completed when coding a group case is making sure that the group identifier that is being used is exclusive and unique to your case. This sounds like a minute detail, but if not followed can cause confusion and lost time for various members of our team. These are just some of the handful of precautions a coder must be somewhat familiar with when tackling a group case. This is a great reason why when coding cases, group or individual, having the codebook out in some capacity while the coding is taking place can save the coder future worry and stress. The codebook that is being employed for the Prosecution Project is exclusive to this project, and is constantly being tweaked and edited to make sure that the process from beginning to end of coding a case is becoming more streamlined and easy to accomplish, as well as serving as a common and shared reference point for all the coders when issues do arise. Being a part of an organized team that functions well like the Prosecution Project is a wonderful and extremely foundational aspect of coding group cases.

Some of our materials are sourced using PACER, with the help of a court order which allows some cost savings.  As explained in “The Cost of Electronic Access to US Court Filings Is Facing a Major Legal Test of Its Own”:

“The 10 cents a page that most people are charged to view documents in Pacer, an online database of papers led by litigants in the US federal courts, doesn’t sound like much. But critics of the setup say this cost is just the beginning.” (Browdie, 2016).”

When coding a group case or even just an individual case, having to navigate PACER’s archaic website while getting charged for each case you look up can definitely add up. At times, within the team we realize how strange of a system it is to have to pay to research Federal Court Cases, and various users outside the scope of PACER tend to agree. Group Cases require digging at times, and when you have to repeatedly think back during the process of coding to how much you are getting charged, even if it is less than one dollar, it still to some extent hinders the flow and productivity rate that is crucial when working for projects such as the Prosecution Project.


Browdie, B. (2018, August 10). The cost of electronic access to US court filings is facing a major legal test of its own. Retrieved from https://qz.com/800076/the-cost-of-electronic-access-to-us-court-filings-is-facing-a-major-legal-test-of-its-own/.

Miller, K. (2019, February 21). 18 members of white supremacist prison gang indicted. Retrieved from https://apnews.com/b1b7cafc34f54aea9d01b7bba675fd24.

Coding Hate Crimes

This continues our series of student reflections and analysis authored by our research team.

Coding Hate Crimes

Katherine Coate

In October of 2019 the variable “Hate Crimes” was added to the process of coding cases and the codebook. The values for this variable are simply “yes” or “no”, but understanding exactly what a hate crime is and how it is defined can be confusing. This is my first semester working as a coder for the Prosecution Project and understanding how to code hate crimes was one of the variables I had to study to truly understand. The best place to start for understanding is looking into how the government defines hate crimes. The first federal hate crime law was passed in 1968 and was signed by President Lyndon Johnson.

“The 1968 statute made it a crime to use, or threaten to use, force to willfully interfere with any person because of race, color, religion, or national origin and because the person is participating  in a federally protected activity, such as public education, employment, jury service, travel, or the enjoyment of public accommodations, or helping another person to do so. In 1968, Congress also made it a crime to use, or threaten to use, force to interfere with housing rights because of the victim’s race, color, religion, sex, or national origin; in 1988, protections on the basis of familial status and disability were added.”

Hate crimes can simply be added as enhancements on federal cases and those are easy to code because the charges include the words hate crime. For example, Holden Matthews set fire to three historically black churches and was charged with three counts of a hate crime. However this is not what every hate crime case looks like.

There are specific charges that count as hate crimes but do not specifically say hate crime in the charges. These are Criminal Interference with Right to Fair Housing, Violent Interference with Federally Protected Rights, Conspiracy Against Rights and Damage to Religious Property as part of the Church Arson Prevention Act.

For example Graham Williamson was charged with one count of interference with housing rights which would be a hate crime. Williamson, along with a co-conspirator, built and burned a wooden cross near the home of a juvenile victim, M.H., who lived in a predominantly African-American residential area of Seminary. He burned the cross to threaten, frighten, and intimidate M.H. and other African-American residents because of their race and color, and because they lived in and occupied residences in that area of Seminary. These cases are a little less straightforward because you have to know what charges count as hate crimes but as long as you have that list available and remember to check it they are easy to code for.

When I was researching hate crimes for this I learned about how much hate crimes vary from state to state and the federal hate crimes laws do not standardize these. There are 5 states currently that do not have hate crime laws and only 28 that include gender bias in their definition (Shanmugasundaram). Federal hate crime laws have expanded pretty consistently and state laws have not. It is frustrating to code cases that seem like they should be hate crimes but can’t be because the states law are behind national laws. It is also hard to prosecute federal cases because there are requirements that the state has to follow. According to the Southern Poverty Law Center:

“The Shepard-Byrd Act allows the federal government to prosecute hate crimes whenever local or state prosecutors choose not to. However, before the DOJ may prosecute a hate crime, several criteria must be met. The U.S. attorney general must assert that either the state does not have jurisdiction, that the state asks the federal government to assume jurisdiction, that a verdict or sentence obtained under state charges insufficiently eradicates bias-motivated violence, or that it is in the public interest of the United States to prosecute.”

Cases such as Joshua John Leff are not hate crimes, even though he threatened to kill minorities with a weapon and tried to state a race war, because it is a state case and the state laws are not as broad. This is surprising to me because normally state legislation is ahead of national legislation. Hate crime laws are incredibly varied and can be frustrating to code because the legislation is not straightforward. However, progress is being made and hopefully it will continue to increase so that hate crimes are nationally defined in a consistent manner.

Works Cited:

The United States Department of Justice. “Hate Crime Laws,” March 7, 2019. https://www.justice.gov/crt/hate-crime-laws.

Shanmugasundaram, Swathi. “Hate Crimes, Explained.” Southern Poverty Law Center, April 15, 2018. https://www.splcenter.org/20180415/hate-crimes-explained.

The Intersections of Members within Foreign Terrorism Organizations and Gender found While Coding

This continues our series of student reflections and analysis authored by our research team.

The Intersections of Members within Foreign Terrorism Organizations and Gender found While Coding

Meekael Hailu

When coding for the Prosecution Project, I have been exposed to a wide variety of various Foreign Terrorist Organizations that are attached to different cases. While coding said cases throughout my beginning with the project this semester, I have had to perform research on multiple various Foreign Terrorist Organizations to make sure that my insufficient background on the organization was met with knowledge so that I coded effectively. When coding these various prominent Foreign Terrorist Organizations, I always tend to notice the gender of the individuals when I am coding.

In my experience working as a coder of this project so far I have only seen men associated with some of the prominent organizations I have coded so far. After consciously realizing this, I wondered to myself, “why is this?”. After reading an eye-opening article on the relationship of gender and justice titled “Treatment of Terrorists: How Does Gender Affect Justice” written by Audrey Alexander and Rebecca Turkington sheds some light on the disparity and statistics of various both on a domestic front within the United States, as well as globally. They emphasize the fact that there are a strong theme and belief within the United States judicial system as well as different judicial systems all over the world that deal with foreign terrorists that women that are involved in these organizations are often “duped” or “fooled” into joining. This fascinated me because you rarely find this being an explanation for most men that choose to join a Foreign Terrorist Organization. They also state in the article that when studies are conducted, many reasons for both men and women joining these organizations stem from the same motivation and ideologies. (Alexander, 2018).

I could not definitively explain the statistical reasoning behind why I have not coded any women that have joined the multiple prominent Foreign Terrorist Organizations, but after reading the extremely well put together article written by Audrey Alexander and Rebecca Turkington, I can see the gender prejudice making its way into yet another field. As I continue to code more cases as a member of the Prosecution Project, I hope to get more of a diversified portfolio of individuals who join Foreign Terrorist Organizations. One aspect of this is coding more cases with females, to learn more through the source files that I will be searching through about these women and to see probably how similar their ideologies and motivations are to the men that choose to join the same organizations, even though individuals still choose to think differently just as Alexander and Turkington stated.

Additionally, I desire to continue to gain a more diversified portfolio on the various organizations themselves. I enjoy being able to recognize an organization’s name on BBC World News or hear it on the National Public Radio station and connect it back to a case that I coded for the project, it makes the contextual knowledge seem even more useful to know and understand as an educated scholar.

Overall, I truly do think that every single variable that goes into coding a case is important even when you might think it is not, especially gender. After analyzing the excellent paper Treatment of Terrorists: How Does Gender Affect Justice, I am looking forward to discovering more cases that help me understand the reasoning behind individuals’ decisions to join these Foreign Terrorist Organizations. Additionally, for the remainder of my time with the project, I want to make sure I take more of a deeper look while coding demographic information to understand on a deeper and more holistic basis all of the factors that play a role in what makes up the individual. I believe that doing this will consequently help you perpetually find more cases to code, and help you code them at a higher quality.


Alexander, A., & Turkington, R. (2018). Treatment of Terrorists: How Does Gender Affect Justice?. CTC Sentinel, 11.

The Siege at Ruby Ridge

This continues our series of student reflections and analysis authored by our research team.

The Siege at Ruby Ridge
Liv Sellergren
The purpose of the Prosecution Project is to gather data from specific cases involving felony crimes with an explicit political motivation in order to correlate and predict juridical strategy for the purpose of scholars and policy makers. When coding cases, we come across a variety of crimes. When I came across the Ruby Ridge case I was very hesitant on how it should be coded. For those unfamiliar with Ruby Ridge, it was the case that helped start a US militia movement due to an 11 day ATF shootout between U.S. Marshalls and Randall Weaver.
Prior to the shootout, U.S Secret Service and FBI had interviewed Weaver and his wife, Vicky, because they believed he was involved with the Aryan Nations. Weaver denied having any involvement but the ATF still had him set up with an undercover where Randell had illegally purchased two firearms. When he was arrested, he made bail but then failed to appear in court. During this time he had become skeptical and suspicious of the U.S. Government and vowed to fight rather than surrender. This lead to the ATF and U.S Marshals furthering an investigation and surveillance of his house. When the agents were noticed, Randalls son Sam had gone out with a gun and had shot a U.S. Marshall which prompted the shootout and siege of Ruby Ridge.
After 11 days of isolation and failure to comply, Randall Weaver surrendered.
In the court proceedings, Randall Weaver was indicted on 10 counts but was later acquitted of all charges except failure to appear in court. He was sentenced to 18 months in prison and after his release, was awarded $100,000 for the loss of his son and wife. His three daughters had each received one million dollars for the loss of their brother and mother.
You may think, this case would be easy to code based on his charges but it is not. There is a great amount of detail and thought that go into coding a case for this project. Furthermore, what part of the case am I supposed to code? Was there a “physical or ideological target” since the federal agents came to his property? Was Weaver actually affiliated with the Arian Nations? If he was acquitted on all charges regarding to Ruby Ridge should I be including this case? Was his target people or property? Neither? Both? These are some of the questions that were raised when coding this particular case.
My coding partner and I have decided to include and code this case based on all of the charges he was indicted on that relate to the project. So to answer these questions in particular, I am coding the case:
      • United States of America vs Randell Claude Weaver
      • on the charges of First Degree Murder, Assaulting and Resisting Federal Officers, Conspiracy to Provoke a Violent Confrontation, Making Illegal Firearms, Failure to Appear in Court, Committing Crimes While on Pretrial Release, Using A Firearm to Commit a Violent Crime, etc.

I believe this is the best way to code this individual case because those charges are why The Ruby Ridge case is included in the Prosecution Project.