Explanatory Note - Stop Data
Background: In July 2019, the Metropolitan Police Department (MPD) implemented new data collection
methods that enabled officers to collect more comprehensive information about each police stop in an
aggregated manner. More specifically, these changes have allowed for more detailed data collection on
stops, protective pat down (PPDs), searches, and arrests. (For a complete list of terms, see the glossary
on page 2.) These changes support data collection requirements in the Neighborhood Engagement
Achieves Results Amendment Act of 2016 (NEAR Act).
Data Notes: The accompanying data cover all MPD stops including vehicle, pedestrian, bicycle, and
harbor stops for the period from July 1 – December 31, 2021. A stop may involve a ticket (actual or
warning), investigatory stop, protective pat down, search, or arrest.
If the final outcome of a stop results in an actual or warning ticket, the ticket serves as the official
documentation for the stop. The information provided in the ticket include the subject’s name, race,
gender, reason for the stop, and duration. All stops resulting in additional law enforcement actions (e.g.,
pat down, search, or arrest) are documented in MPD’s Record Management System (RMS). This dataset
includes records pulled from both the ticket (District of Columbia Department of Motor Vehicles [DMV])
and RMS sources. Data variables not applicable to a particular stop are indicated as “NULL.” For
example, if the stop type (“stop_type” field) is a “ticket stop,” then the fields: “stop_reason_nonticket”
and “stop_reason_harbor” will be “NULL.”
Each row in the data represents an individual stop of a single person, and that row reveals any and all
recorded outcomes of that stop (including information about any actual or warning tickets issued,
searches conducted, arrests made, etc.). A single traffic stop may generate multiple tickets, including
actual, warning, and/or voided tickets. Additionally, an individual who is stopped and receives a traffic
ticket may also be stopped for investigatory purposes, patted down, searched, and/or arrested. If any of
these situations occur, the “stop_type” field would be labeled “Ticket and Non-Ticket Stop.” If an
individual is searched, MPD differentiates between person and property searches. Please note that the
term property in this context refers to a person’s belongings and not a physical building. The
“stop_location_block” field represents the block-level location of the stop and/or a street name.
The age of the person being stopped is calculated based on the time between the person’s date of birth
and the date of the stop.
There are certain locations that have a high prevalence of non-ticket stops. These can be attributed to
some centralized processing locations. Additionally, there is a time lag for data on some ticket stops as
roughly 20 percent of tickets are handwritten. In these instances, the handwritten traffic tickets are
delivered by MPD to the DMV, and then entered into data systems by DMV contractors.
On August 1, 2021, MPD transitioned to a new version of its current records management system,
Mark43 RMS.
Due to this transition, the data collection and structures for the period between August 1, 2021 –
December 31, 2021 were changed. The list below provides explanatory notes to consider when using
this dataset.
• New fields for data collection resulted in an increase of outliers in stop duration (affecting 0.98%
of stops). In order to mitigate the disruption of outliers on any analysis, these values have been
set to null as consistent with past practices.
• Due to changes to the data structure that occurred after August 1, 2021, six attributes pertaining