For our third installment of city-by-city analysis, we take a look at our nation’s capital: Washington, D.C. With a population of approximately 658,900 spanning over 61.4 square miles, D.C.’s population density is approximately 37 persons per average square city block; or 10,731 persons per square mile. D.C. having one of the highest population densities and a reputation for being one of the “most dangerous” cities in the U.S. we would assume that crimes occur at a highly frequent rate.

Based on the Federal Bureau of Investigation’s Uniform Crime Report (FBI-UCR), D.C. reported 105 murders, 3,231 robberies, and 4,004 aggravated assaults in 2014. Using basic numbers, that assume a uniform distribution over the entire city, this equated to approximately 16 murders, 490 robberies, and 608 assaults per 100,000 persons each. Accounting for all cities with population above 50,000 persons, this would place Washington D.C. at 26th overall in the U.S. for most violence per 100,000 persons. In comparison, Oakland “ranks” at 5th and Milwaukee “ranks” at 9th.

What these rankings, which are almost identical to the old Morgan Quitno rankings, assume is that all cities have the same: square mileage,  population density, and have uniform distribution (probability of a robbery on the street is the same as a robbery in the East Room of the White House).

Washington, District of Columbia

Washington, D.C. is partitioned into a series of eight wards, with a total of 133 listed neighborhoods. With a total square mileage of 61.4 square miles, this breaks down to an average neighborhood being an approximate 0.46 square miles; or a total of about 133 city blocks. The sections of D.C. are aggregated into four primary sections, though not formally defined, indicated as “Southeast D.C.” bordered by Prince George’s County, Maryland and the Anacostia River; “Downtown” bordered by the Anacostia and Potomac Rivers; “Northeast D.C.” bordered by downtown an the northern Prince George’s County border; and “Northwest D.C.” bordered by Montgomery County, Maryland.

Shootings In D.C.

According to SpotCrime data, there were 687 shootings reported in 2010 (most recent data publicly available). Under a uniform distribution, this would indicate that each city block should experience a shooting roughly once every 26 years; a total of 0.0387 shootings per city block per year. However, using the geo-encoded data obtained from SpotCrime, the distribution of shootings in D.C. is not uniform.

Looking at the distribution of shootings in D.C., we see that the shootings primarily exist in Southeast D.C. In contrast, the downtown region of D.C. experiences almost little to no shootings. Zooming in closer, with the sections overlaid on the distribution of shootings, we see that the shootings are in systematic locations within each respective region.

Breaking down the contours of the distributions of shootings, we find that 95% of the reported shootings in D.C. are contained in 40.63 square miles, which is 66.17% of the entire city. Similarly, 75% of the reported shootings in D.C. are contained in 21.34 square miles (34.76% of the city) and 50% of the reported shootings in D.C. are contained in 11.15 square miles (18.17% of shootings). While this is not uniform, Washington D.C. exhibits a wider spread of shootings throughout the city (66.17, 34.76, 18.17) when compared to Oakland (36.54, 17.52, 8.91) and Milwaukee (50.58, 23.51, 11.12).

From the contours, we can then identify the average number of years for a person to expect a random shooting. In the region outside of the 95% contour (red), we see there are approximately a total of 35 shootings over the 20.77 square miles. This equates to 0.005831 shootings per city block. With an average population density of 37 persons per city block, this equates to a person to expect one random shooting every 6,345 years.

Between the 75% and 95% contours (red and blue contours), there were approximately 137 shootings. With an area of 19.29 square miles, this equates to approximately 0.0246 shootings per average city block per year. For the average population density of D.C., this equates to a person to expect one random shooting every 1,506 years.

Between the 50% and 75% contours (blue and green contours), there were approximately 172 shootings. With an area of 10.19 square miles, this equates to approximately 0.0584 shootings per average city block per year. For the average population density of D.C., this equates to a person to expect one random shooting every 634 years.

Within the 50% contour (green contour), there were approximately 344 shootings. With an area of 11.15 square miles, this equates to approximately 0.1068 shootings per average city block per year. For the average population density of D.C., this equates to a person to expect one random shooting every 347 years.

So this identifies that shootings, while occurring typically two per day in D.C., due to the number of person in the city and the spread of the city it is a rare event to experience a shooting. Granted, this is a direct effect to a person (being shot at/shot); which is different to witnessing a shooting (refer to per block amounts for this).

Taking into account only quantities of reporting and population distribution, this indicates that D.C. is on the order if being five times “safer” than Oakland and Milwaukee in terms of number of years to expect shootings relative to percentiles.

Robberies in D.C.

Washington D.C. experienced 3,231 reported robberies during 2014. While this translates to 490 per 100,000, D.C.’s robbery rates are lower than that of Oakland (849) and Milwaukee (586).

Here we see that half of the crimes are contained in approximately 19% of the city, like shootings, but are concentrated in one major area: the Shaw, U Street, and Columbia Heights neighborhoods that straddle the Children’s National Medical Center. In this region, a person can expect to experience a random robbery once every 34.7 years. This means, a person should expect two random robbery attempts in their lifetime in the worst neighborhood for robberies in D.C.

Assaults in D.C.

Washington D.C. experienced 4,004 reported assaults during 2014.Per 100,000 persons, this translates to 607; assaults are contained primarily to the same higher crime regions of D.C. spread amongst a corridor of Northwest D.C., across Southeast D.C. and in few place in Northeast D.C. While there is a small flare-up in downtown, there is again rarely any reported crime in this region. Again Oakland (766 per 100,000) and Milwaukee (809 per 100,000) have higher rates than D.C., identifying Washington D.C. to be randomly “safer” than these two cities.

In summary

We see that while Washington D.C. has a smaller rate of shootings, robberies and assaults when compared to Oakland, California and Milwaukee, Wisconsin, it still has its fair share of “rough” neighborhoods and “safe” neighborhoods. D.C.’s “safest” neighborhoods are indeed some of the safest for major cities across the United States. We do see that particular violent crimes are limited specific areas within the city. We do not have root causes developed for why this is the case, but we at least know where things are happening across the city; and more importantly how rare they occur.