The Washington, DC region is great >> and it can be greater.

Public Safety

Black homicide rate drops but remains high

Homicide rates in DC have decreased sharply over the past several years. However, even with this overall improvement, the rate of homicide of black DC residents remains significantly higher than that of whites.

Photo by Null Value on Flickr.

In 2009, DC dropped off the list of the cities with the top 10 homicide rates. The absolute number of homicides and the number of homicides per 100,000 residents decreased even further in 2010.

And yet, although there have been considerable improvements in crime statistics within the capital over the past decade, DC's black homicide figures are still much higher than comparable rates at the national level. In fact, on a per resident basis, blacks in the District face over double the homicide rate as blacks in the nation as a whole.

The Metropolitan Police Department publishes data online starting from 1998 indicating each homicide victim's race. Combining the MPD data with population projections derived from official Census figures allows computing the homicide rate per 100,000 residents by race.

Homicides per 100,000 residents, by race. View the raw data (XLS).

An analysis of FBI victim data and population estimates based off of Census Bureau findings reveals that from 1998 to 2003, the number of homicides ranged from 61.6 to 72.9 black victims per 100,000 black residents. Starting in 2004, even as the number of black District residents dropped, the black homicide victim count decreased at an even faster rate.

The rate fluctuated from 49.6 to 58.1 during the five years from 2004 to 2008. Over the past few years rates dropped at an even faster rate, to 43.0 in 2009 and 37.7 in 2010. Although these are notable improvements, for comparison, the national homicide rate was 16.5 black victims per 100,000 black residents in 2009.

Among white residents the homicide rate ranged from 3.3 to 11.3 during the period of 1998 to 2002. From 2003 to 2010, the figure further decreased, ranging from zero to 3.1 per 100,000 residents over the 7-year span. During the five most recent years with data available (2006 to 2010), the rate averaged 1.5 white homicide victims per 100,000 white residents.

In fact, white residents face a lower homicide victim rate in the District than the nation as a whole. Nationally, 2.7 white residents are homicide victims per 100,000 white residents. But, among DC white residents, this figure is lower both when looking at data from either of the most recent two years (zero and 1.3 in 2009 and 2010, respectively) or the latest 5-year average (1.5).

There are many limitations to expressing homicide rates in these terms for any geographic entity smaller than the nation as a whole. Victims at a state or city level may or may not be actual residents of that state or city. Urban areas tend to have a much larger actual number of people present compared to residents because of commuters, tourists and business travelers. However, the data still provides a useful measure of the relative safety of any given location.

It should be noted that a comparable analysis is not available for other races, more than one race, or residents of Hispanic origin. The US Census Bureau tracks Hispanic origin as a separate identifier than race, while MPD uses it as one of the possible race categories. This difference coupled with a consolidated "other" category makes it difficult to extend this analysis to other races and individuals of Hispanic origin.

Of course, race alone does not explain differences in homicide occurrences. Each death is tragic and occurs under different circumstances as Homicide Watch D.C. strives to document.

Nonetheless, analyzing racial trends can provide insight into the perceptions and realities that District residents must negotiate with regard to personal safety. The trends may also prove useful in identifying some of the possible targeted opportunities that relate to public safety, family and community support, education and economic opportunities.

Mitch Wander first arrived in Washington, DC over 25 years ago as a US House of Representatives page while in high school. An avid promoter of DC living, Mitch has lived in wards 1, 2, 3, and 6. He and his wife are proud DC Public School parents. He serves as an officer in the US Army Reserve. 


Add a comment »

I am curious as to what the Latino homicide rate is, although I note that you mention that data on other races isn't available.

That white people are rarely victims of homicide in DC isn't particularly surprising: white residents in DC tend to be overwhelmingly affluent, and homicide--like most other violent criminal behavior--correlates very closely with poverty. To be a white homcide victim in DC, you're likely either the victim of a domestic dispute or a mugging/carjacking gone awry.

by Joel on Jun 21, 2011 2:33 pm • linkreport

what is particularly frustrating is the lack of coverage by newspapers and even internet media when concerned with DC crime- and especially homicide. At one time, you could read the homicide reports in the Metro section of the WaPo and even get the updated statistics- but this is not longer allowed- some kind of political correctness anal retentive policy no doubt is at work here. It is important to know where the crime is going down- and what kind of crime- and the numbers. All of this is grisly- yes- but it is also in a citizen's self interest to know which areas are dangerous and which are safe. The WaPo in their crusdae to paint everyone as the "same" chooses to omit important data that can be of benefit to all. What a shame.

by w on Jun 21, 2011 2:40 pm • linkreport

Controls, please. Just looking at race and saying "whites in dc have lower murder rates than in the nation, blacks have higher" isn't very useful. Are the whites and blacks of DC comparable to whites and blacks in the rest of the nation? Are white people in DC wealthier/more educated/better employed/etc than whites in the nation? What about blacks? These are important questions. You should be concerned about artificially biasing these results towards a racial frame without identifying (let alone answering) why these discrepancies exist between races in DC and among races against the nation.

by jared on Jun 21, 2011 2:58 pm • linkreport

Interesting Thought: Why dont we publish crime statistics by where the CRIMINAL lives instead of the crime location? I think you would see some interesting clustering effects.

I'm always curious when these muggings, robberies, etc happen in NW, Dupont, Georgetown, etc --- where do these people come from and instead of trying to find what random place those criminals may target - how about finding the source?

by David on Jun 21, 2011 3:04 pm • linkreport

I wouldn't say the homicide coverage has dropped off universally. It's still quite prominent in the evening news. You might be focusing on one specific dysfunctional news outlet (WaPo) that has a bunch of other problems as well. They've pretty much ceded local coverage to become a national paper.

I'm not sure the city really wants to fix this problem. The fixes go against the "people in glass houses" arguments that the AA political class seems particularly touchy too. There's pervasive violence in poorer communities, but whereas with poor white violence it tends to be physically removed from most people (it's out in the sticks). For poor blacks, it tends to be much more concentrated (urban density) and thus the violence feeds off itself more quickly (You have to join a gang to get protection from a gang, which creates more opportunity for violence to flare).

While I'd agree that from a sociological standpoint it is rooted in the violence of segregation, I'm not really sure that has any practical implications on it's solution. Certainly the older boomer generation doesn't seem to have this problem with internal violence or drugs, but their children and grandchildren certainly seem to. Any proposal that seems to come from white people is labeled as (and occasionally is) racism even when it's from liberal white people. Providing housing an food assistance as a leg up, certainly doesn't seem to solve the problem as DC has demonstrated. Attempts to improve education outcomes have been met with stiff resistance by the AA community as interference. The traditional societal response of using a local police force as a mechanism for enforcing societal norms has been undercut by the hiring of the very same criminals into the force and the deep rooted and politically expedient demonization of police in general.

There's really nothing left for anyone *other than* the AA community to actually take action, which gets me back to my main point. With the past 30+ years of AA control of this city, not much has been done to fix the violence problem other than to rely on gentrification and federal incarceration.

by Old Head on Jun 21, 2011 3:04 pm • linkreport

@Joel - The homicide victim data for Hispanics and other races is available either on the MPD website in the Annual Reports (,A,1246,Q,555551.asp) or consolidated into a single page if you access the "View the raw data" link directly under the graph in this article. The worksheet called "Raw MPDC data" has what you're seeking.

There was no easy way for me to determine a Hispanic homicide victim rate since Census tracks Hispanic as a separate question from race while MPD tracks it in the same question as race.

by Mitch Wander on Jun 21, 2011 3:21 pm • linkreport

There is a difference between the lack of coverage of an issue and the issue not being available at all. The stats are there, it just takes going to MPD's website to find it. As far as the Post, I don't know anyone who considers it relevant.

@Oldhead, I agree w/most of what you said. However, it's factually incorrect to suggest that improvements to the Edusystem here has been met w/stiff resistance from the black community. Aside from innuendo and assumptions, there aren't any facts to back that claim. Maybe the improvements were not what the black community felt they needed to be and since we are talking specifically about a "group," I would argue that the black community's position on this is equally important as those who suggest what the improvement should be.

Also, I'm not steeped in this but if violent acts committed by blacks are at the same level it was 30 years ago, your thoughts on DC's 30 years of "black" leadership would be appropriate. However, there are too many assumptions built into that line of reasoning.

Obama is THE black man in america and violence is just as bad. His mere presence will not solve the issue. It takes action..from the community and our stakeholders.

by HogWash on Jun 21, 2011 3:36 pm • linkreport

Thanks for the shout-out to Homicide Watch.

It is so tragic to add victim photo after victim photo to the albums and see young black man after young black man (with some exceptions). Sit through court and you see much the same parade. The one thing that becomes so clear is that in homicides, there are so many more victims than just those that are killed. All the families and so many friends, of both victims and defendants, are impacted and affected by the deaths, too.

As for the homicide rates of other races, this year we've had two Latino victims and one white victim in our database of 50 homicides this year.

by Laura Amico on Jun 21, 2011 3:36 pm • linkreport

In complete seriousness, what is the law enforcement value of tracking crime statistics by race? Shouldn't law enforcement be using its time and resources to group crime statistics in ways that provide for targeted enforcement? Tracking by neighborhood and PSA makes sense because then law enforcement can make deployment decisions, but if you find out a particular race is disproportionately affected by a particular crime, what do you do with that information (unless it involves a race-related/hate crime statistic)? Seems like a waste of time, effort and money.

by Alan Page on Jun 21, 2011 3:51 pm • linkreport

@Alan, you bring up a good point and you're way ahead of yourself at the same time.

Looking at race on its own is totally useless, you're right.

This data though, doesn't ACTUALLY tell us what difference (if any) is attributable to race. To do that, you have to control for other factors first. You have to compare between whites and blacks who are almost exactly the same in all aspects besides race. THEN you know the difference between the racial groups. At that point, looking at race can be helpful for identifying underlying causes associated with race that would cause this difference to exist -- the solution then will go way beyond just policing.

I'd be willing to bet that after controlling for other variables, race drops out as a significant predictor on its own -- that variable is probably absorbing a whole mess of other variables that are important here. I'd guess that the real headline should be "Poor Person Homocide rate Drops but Remains High."

Leaving the data as it's presented here serves only to invoke disturbing and inaccurate assumptions about the conditions and behaviors of both white and black people in the city.

by jared on Jun 21, 2011 4:21 pm • linkreport

And I don't mean to imply that we should dismiss the fact that the black community in DC is disproportionately effected by homicide in DC, just that we should interpret data wisely and responsibly to show that homicide rates are a symptom of other social ills -- not a symptom of blackness.

by Jared on Jun 21, 2011 4:26 pm • linkreport

Actually, now that I think about it...the true headline would probably read "Homicide Rate Continues to Drop, as Average Income Rises in DC." I'd guess that the correlation between average income and homicide isn't linear either (a 10k increase probably effects a 20k earner different than a 100k earner), so I'd suggest logging the variables to control for difference between racial groups.

by Jared on Jun 21, 2011 4:34 pm • linkreport

@Jared - DC Fiscal Policy Institute (DCFPI) reported the folowing based on Census data:

"For Black residents (non-Hispanic), median household income fell between 2000 and 2005 and then increased from 2005 to 2008. While the recent increase was significant — a 10 percent increase — it only offset the earlier decline. Median household income for Black DC residents stood at $39,200 in 2008, just two percent higher than the 2000 figure of $38,400. This change is not statistically significant."

by Mitch Wander on Jun 21, 2011 4:48 pm • linkreport

Most of the homicide is still turf war (drug) related, so the cops don't care as much. In the crack wars DC was likened to a self cleaning oven.

by TGEoA on Jun 21, 2011 11:08 pm • linkreport

@Alan - That's good info to start with, and using the median is definitely smart, but you'd need to regress it along with other controlling variables as well to see if there's a relationship hidden in those numbers. It might be a good idea to control for region in DC as well -- assuming that homicide rates aren't evenly dispersed throughout the city, it's also possible that neither is median income. So, while the city-wide median income may have changed in the pattern you have above, the movement for localized median income might be quite different, and might correlate with homicide rates in those regions as well. And you can't forget other theoretically significant control and test variables either! In a regression, they all matter.

by Jared on Jun 22, 2011 10:10 am • linkreport

Sorry, that last one was for @Mitch, not @Alan.

by Jared on Jun 22, 2011 10:10 am • linkreport

@Jared With all due respect, nothing you said explains why tax dollars need to be expended tracking crime statistics by race; to the contrary, your prediction that "after controlling for other variables, race drops out as a significant predictor on its own" seems to indicate that there is no value at *all* in collating crime statistics by race.

I need someone to make a rational argument in favor of this and explain the law enforcement value of knowing the racial breakdown of criminal behavior or victimhood, outside the narrow context of hate crimes (e.g. interracial crime motivated by bias).


I wrote all my at-large council members on this as well as Wells. They forwarded the matter to the top brass. No answer yet, probably because there is no answer. Another case of race trumping reason, it appears, unless someone can present a rational argument in favor. I don't want my tax dollars going down a rabbit hole like this...

by Alan Page on Jun 22, 2011 4:22 pm • linkreport


I think Mitch actually does address your question quite well in the last paragraph of this piece. I won't quote the whole thing, but basically it "helps target opportunities to improve situations." I think it would be more troublesome if we didn't pay attention when a distinct group of people are disproportionately suffering from any type of crime.

My problem with is simply that looking at race alone without all these other variables isn't very useful and doesn't fully explain the phenomenon. If the problem is "black people get murdered," what's the solution? If the problem is "Low income people who live in areas of certain density, a particular distance from police stations with a certain frequency of police patrols, have a certain level of education, etc... get murdered," then we can start to work on solutions.

I think that in all likelihood, race is collected because it's quite easy to do, and probably doesn't expend that many tax dollars. It would take a lot more resources to collect all the other variables I've suggested (it's the kind of question that academics are more likely to investigate than police!) In the face of limited resources, I can understand why police would collect racial data as a proxy variable for other elements which they have no data on (or for things which can't easily be quantified, like "culture").

I'm not letting Mitch off the hook though -- if he's going to analyze data, the only responsible thing to do is to analyze it fully, or at least make note of what he couldn't/didn't include.

by Jared on Jun 22, 2011 5:01 pm • linkreport

Very late to this party, I know, but the discourse is an important one.

Jared remarks: "My problem with [this] is simply that looking at race alone without all these other variables isn't very useful and doesn't fully explain the phenomenon. If the problem is "black people get murdered," what's the solution?"

Would this same logic apply to education statistics? "Black people get educated poorly (relative to whites)" but unless one has "the solution," please don't make that observation? Many well-intentioned observers who lay no claim whatsoever to solutions believe it is not only appropriate but imperative to make the observation, however they might opt to contextualize it.

Normally, of course, what is considered *really* off-limits in polite sociological conversation is the comparably disproportionate over-representation of blacks within the population not of victims but offenders. Both MPD and Mitch dodge this bullet (to use an idiom that most everyone will surely consider either wholly appropriate or wholly not) by confining the available data and the analysis, respectively, to the race of victims. But even so, this is problematic, because victims--obviously and by definition--require a victimization; and every victimization--obviously and by definition--requires a victimizer or, returning to the parlance of criminal justice, an offender. And, finally, with homicide, far more often than not the race of the victim is the same as the offender.

In any event, Laura at Homicide Watch and Mitch are both quite right in showing the faces of DC murder victims (albeit one through photos, the other statistics) because the risk of such an ongoing social tragedy being ignored or downplayed by elites and by the larger public is, sadly, just too great without such reminders.

Thus, Jared is half-right, perhaps, in suggesting that if we are going we are going to consider race quite so openly, then we should consider the other variables that we imagine to be theoretically important and sociologically operative in trying to understand local homicide.

Unfortunately, other problems arise. First, despite Jared's expansive faith in regression models, there is the galloping problem of multicollinearity. One cannot control statistically for income, education, place, or even prior criminal history in the District, when those independent variables are so often so very related. In other words, not independent.

Secondly, and perhaps a more profound way of saying much the same thing: some scholars have argued quite persuasively that elaborate statistical models which purport to show the *in*significance of "race" in crime by controlling so thoroughly for income, education, urbanization, exposure to poverty, drug and alcohol abuse, prior experiences with police or incarceration and sometimes attitudes, too, for example, religiosity, etc., etc., etc.---some scholars argue that models like this have the effect of statistically removing the central features of race as constructed in American society and modern urban life. In short, they they argue that models which pretend race can be meaningfully isolated from experience and social circumstance are folly.

So perhaps the centrality of race in Mitch's admittedly limited analysis is something much more than insufficiently-sensitive, over-general and pointless? *Perhaps* by underlining that the proportional rate of homicide for black victims in this city is 29X that of white ones, the author has appropriately reminded us--politics notwithstanding--that the immortally clever book title by Cornell West was right: "Race Matters."

by Joe on Mar 3, 2012 5:22 am • linkreport

Add a Comment

Name: (will be displayed on the comments page)

Email: (must be your real address, but will be kept private)

URL: (optional, will be displayed)

You can use some HTML, like <blockquote>quoting another comment</blockquote>, <i>italics</i>, and <a href="http://url_here">hyperlinks</a>. More here.

Your comment:

By submitting a comment, you agree to abide by our comment policy.
Notify me of followup comments via email. (You can also subscribe without commenting.)
Save my name and email address on this computer so I don't have to enter it next time, and so I don't have to answer the anti-spam map challenge question in the future.


Support Us