A closer look at the numbers behind D.C.’s gun violence

Last week the Mayor held a press conference to address an increase in violence in the city. While the total number of crimes committed so far this year is about the same as by this time last year, some violent crimes have increased. By our count, the number of crimes committed with a gun this year is 20 percent higher than last year.

So what’s behind this surge in gun violence? The crime data we looked at, which is available to the public through the D.C. Data Catalog, revealed some interesting patterns.

First, as many people have speculated, the increase in gun violence this year has been driven by crime this summer. Nearly half of the increase in crimes committed with guns this year is from an increase in gun crimes this June and July. The number of gun crimes this summer is also higher than in the summers of 2012 and 2013, even though in those years the total number of gun crimes was about the same as this year.

gun crimes-export

We also found that gun violence has not increased uniformly across the city. Out of the 39 neighborhood clusters for which the police department reports crime data, 22 saw an increase in the number of crimes committed with guns this year.  The biggest increases were in a few neighborhood clusters: north-central clusters like Petworth/Brightwood Park/Crestwood and Takoma/Brightwood/Manor Park; eastern neighborhood clusters west of the Anacostia River like NOMA/H Street/Capitol Hill and Capitol Hill/Lincoln Park; clusters near the southern tip of the city, like Congress Heights/Bellevue/Washington Highlands; and other southeast neighborhood clusters east of the Anacostia, like Woodland/Fort Stanton/Garfield Heights/Knox Hill.

Seventeen of the neighborhood clusters saw a decrease or no change in crimes committed with guns this year compared to last year. The biggest decrease in gun crimes occurred in neighborhoods in the eastern corner of the city. Several of these neighborhoods, though, still had gun crime numbers that were among the highest in the city.


It’s still unclear why this surge in violence is happening. Some people believe it’s from synthetic drugs or part of a nationwide trend. We’re curious to see if anyone can find more answers using data.

What exactly is this data?

We downloaded the crime data from the D.C. Data Catalog on August 4, 2015. Because the data is updated periodically, numbers pulled later might differ slightly from the numbers we have in our post.

We looked at all crimes in D.C. reported between January 1, 2012 and July 31, 2015 where the method listed for the crime was “gun”.

4 thoughts on “A closer look at the numbers behind D.C.’s gun violence

  1. Your method and map insinuate that Georgetown and Dupont are less safe than Edgewood or Benning, which is certainly not the case.

    It would probably be more useful/meaningful to compare percent increases/decrease, as well as ranking each neighborhood in 2014 and 2015 by raw number of gun crimes and measure the change in rank.

    Additionally, I would venture to say that the increase in gun violence is due to gentrifying neighborhoods, which increases opportunities for armed robberies/home invasions/muggings. It is obviously due to other factors as well, but without seriously studying the data, it is imprudent to venture any other guesses.


    1. Yep, using percent change or ranking neighborhoods by raw crime numbers would be another way to look at the data. Each method tells a different story.


  2. One thing to note, and this is important, this map does not line up with the police district map, which makes it somewhat suspect. For example you mention Takoma as someplace where there has been a steep increase, but that has not really been the case at all,, Takoma is part of the 401 police service area (with Shepherd Park and Colonial Village), yet you have it grouped into something else, which has seen an increase, but is considerably smaller. My advice is quit relying on these old maps. This map is not lining up to the police service areas, it uses very outdated boundaries. No offense, but this does not really fall in line to how the data itself is collected if you pay attention to how it is actual geographically mapped in terms of crime data. Police Service Areas actually do make far more sense when analyzing crime data, as they are more reflective of neighborhood boundaries and groupings. For example Takoma is mapped with Shepherd Park, and Colonial Village, instead of Manor Park and Brightwood. If you actually are familiar with the neighborhoods, this indeed makes a bit more sense. Likewise, they are also more reflective of how the crime data is collected, which is not based on this map, but on Police Districts, subdivided by Police Service areas. If you dig deeper, you will notice for example, the problems are not so much in Takoma, as they are in Brightwood and Manor Park. My advice is to go back to the drawing board and use the Police Service Area maps, because right now from a data analysis perspective this really is a fail. It would help you better identify where the real issues are by area. But also provide better analysis. As it stands this lacks a real understanding from a quantitative, qualitative, and data grouping perspective.


    1. Yes, you’re right, we used neighborhood clusters not PSAs. We do lose some precision when we look at groups of neighborhoods. For instance, the neighborhood cluster of Petworth/Brightwood Park/Crestwood had one of the biggest crime increases, but my bet is not a lot of that is from Crestwood.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s