Today, the ACLU of Massachusetts released a new interactive tool allowing members of the public to visualize and analyze nearly a decade of Boston Police Department SWAT team after-action reports. An ACLU analysis of these reports has identified troubling racial disparities in BPD SWAT raids, particularly when those raids involve drug investigations. 

Review Boston Police SWAT After-Action Reports

About the data 

This analysis draws on after-action reports of BPD SWAT incidents from 2012 to May 2020. After-action reports are created by the BPD SWAT team to document the unit’s deployments. The reports include descriptions of the incident, information about the people involved (the police, targets, and others impacted), the geographic location of the deployment, and details about the level of police force used during the SWAT action, among other details. The BPD released these reports in response to a 2020 subpoena from then-City Councilor Michelle Wu.  

Unfortunately, the documents produced by the BPD were difficult to analyze in bulk; among other problems, they were not machine readable. While the ACLU and our partners made every attempt to ensure the accuracy of the data pulled from these reports, our analysis has some data limitations. For example, the underlying after-action reports occasionally include missing, clearly incorrect, or difficult-to-interpret data. As explained in greater detail in the methodology section at the bottom, some errors may also have arisen in our processing of the records and extraction of information. That said, we engaged in a human review of the documents to ensure any errors in our extraction do not impact the overall analysis of trends.  


The data: Disproportionate SWAT policing of Black and Brown communities and people

There are 262 after-action incident reports representing deployments between 2012 and May 2020. A significant portion of these police actions took place in police precincts roughly corresponding to the neighborhoods of Mattapan, Roxbury, and Dorchester.  

According to data reported by the Boston Planning & Development Agency Research Division, these three neighborhoods had the highest percentage of residents who are non-white (i.e., identifying as a race other than white or as white along with one more other races) of any Boston neighborhood in 2020. These neighborhoods also had a relatively higher share of residents identifying as Black or African American and/or Hispanic than most other Boston neighborhoods.   

After-action reports typically list one or more incident types indicating the reason(s) for the raid. In order of frequency, the most common types were Search Warrant – Other/Unknown, Search Warrant – Drugs, Mental Health Crisis, and Barricaded Suspect. 

Like other SWAT teams, the BPD SWAT unit is deployed in response to active crises and in deliberate, planned SWAT raids. Typically, the latter cases involve the serving of search and arrest warrants. While the BPD has discretion about whether to deploy the SWAT unit in response to calls for service, SWAT deployments pertaining to active emergencies like mental health crises, barricaded suspects, and domestic violence incidents are nonetheless responsive to active emergencies in the community. On the other hand, the deployment of the SWAT unit to serve warrants, particularly search warrants, is more reflective of self-directed police work. For that reason, racial disparities in the use of SWAT units to serve search warrants are a significant cause for concern.  

Data indicate that white and Black people in the United States use and sell drugs at approximately the same rates. But the BPD’s SWAT unit serves drug warrants almost exclusively on people of color.  

The data: Targets of the BPD’s SWAT raids

According to the after-action reports, 784 people were subjected to SWAT raids between 2012 and 2020. 270 of these individuals were specifically marked as the “suspects” or “targets” of the raid.  

The ages of all raid subjects ranged widely from infants to 83-year-olds. Individuals between 20 and 40 years old were the largest group impacted. For those marked specifically as raid suspects or targets there is a smaller range of ages; the youngest person was 16 and the oldest was 67.  

According to the reports, 105 children under the age of 18, including 25 children under the age of 5, lived in homes subjected to BPD SWAT raids between 2012 and early-2020. 

Those impacted by raids were predominantly men, accounting for 88% of those targeted by raids and 62% of all individuals subjected to raids. When considering both race and gender, Black men were the largest group of those both subjected to raids (38%) and marked as raid suspects/targets (50%).

There are significant racial disparities in those targeted and all those impacted by raids 56.3% of those targeted by raids were Black non-Hispanic and 8.5% were Black Hispanic. These figures are disproportionate to the share of Black residents in Boston. As of the 2020 Census, 25.5% of Boston’s population was Black or African American; this figure includes multiracial residents and both Hispanic and non-Hispanic residents. Of this, 22% of Boston’s population was Black non-Hispanic, and 3.4% was Black Hispanic. Hispanic residents are also disproportionately impacted by raids. 28.7% of those targeted by raids were recorded as Hispanic, White Hispanic, or Black Hispanic; however, the 2020 Census found that Hispanic or Latino residents made just 19% of Boston’s population. 

The disproportionate share of Black and Hispanic individuals impacted by raids varies by incident type; the disparity is notable in “officer-initiated” raids* such as those for search warrants. In contrast, the demographics of people subjected to SWAT team involvement in crisis-related incidents, such as mental health crises, more closely reflect the demographics of Boston.  

*To better understand BPD SWAT unit practices, we categorized incident types into “officer-initiated” incidents such as search warrants or arrest warrants and “crisis-related” incident types that reflect an emergency such as a mental health crisis. 

Another way to view this disparity is to look at the share of white non-Hispanic residents impacted by raids. White non-Hispanic people are underrepresented in officer-initiated raids; Boston’s population is 44.6% white non-Hispanic, but the group of people impacted by all officer-initiated raids are only 6.5% white. In contrast, white people were involved in crisis-related raids at a rate more proportionate to the overall population; 33.5% of those impacted by these raids are white.  

The data: SWAT team after-action reports document frightening, militarized surprise raids 

Reports detail the SWAT team battering down doors, deploying aerosol grenades, waking individuals asleep in bed, and detaining relatives of suspects and children. According to the reports, well over half of the people subjected to this level of force were not even the targets of the operation but were just roommates, partners, family members, or children who were present in the home during the raid.  

SWAT raids can even harm people who live nearby a person wanted by police. One of the records details a raid where the entire household was not the target of the operation; the SWAT team failed to go to the correct address on a search warrant and wrongfully invaded another family’s home. In 2018, the ACLU of Massachusetts represented this family in a lawsuit against the BPD. As our lawsuit on behalf of the family alleged, “[the BPD SWAT team] used a battering ram to break down the door, trained guns on and handcuffed the parents … and their 15-year-old child, and ransacked the home with two younger children present.” The family suffered lasting emotional distress from this assault. Ultimately, the case was settled and the City was forced to pay the family $500,000.  

If you are interested in learning more about the practices of the BPD SWAT team, please use the interactive BPD SWAT records review tool. In addition to exploring the data, you can also read the original after-action reports, which include a narrative description of the incident at the end of each document. Please don’t hesitate to contact with any significant findings.

Review Boston Police SWAT After-Action Reports

Data Notes and Methodology

The original incident reports were made public as non-machine-readable PDFs. Therefore, the first task of this project involved determining how to extract information from these files in a usable format. Volunteer data scientists began by OCR-ing and processing the PDF files to extract useful fields from the documents. Due to the nature of the PDFs and occasional image quality issues, the programmatic data processing and extraction of these reports was subject to error. As a final measure to address any errors and ensure overall data quality, selected fields were manually reviewed and updated by volunteers. For additional data documentation please see the “data notes” section of the BPD SWAT tool.  

This data extraction project was possible with the support of: 

  • Tarak Shah, Human Rights Data Analysis Group — programmatically extracted data from the after-action report PDFs 
  • Aaron Boxer, ACLU-MA volunteer — conducted data extraction, processing, manual entry, and analysis 
  • Natasha Ceol, ACLU-MA volunteer — assisted with data quality-checking and redaction of the after-action reports.