Statistical look at crime in Columbia Heights

For anyone interested, I looked up some information on crime in Columbia Heights.

Here's a breakdown. These figures are percent change from (2007-2008) to (2008-2009). These are yearly figures compared, so you get an idea of trending.

EDIT: You can get an idea of crimes that have occured in the past two years. I am by no means a statistical expert. I am an expert in saying "crime has not decreased since 2007."

For Police Service Area 304, which encompasses the area between 16th St. and Georgia Ave, Florida Ave. and Harvard St:

Homicide: +200%
Sex Abuse: -62%
Robbery (no gun): +12%
Robbery (gun): +6%
Assault w/Dangerous Weapon (no gun): +20%
Assault w/Dangerous Weapon (gun): +75%
Total Violent Crime: +10%

For Police Service Area 302, which is between 16th St. and Park Pl., and Harvard St. and Rock Creek Park:

Homicide: +100%
Sex Abuse: -23%
Robbery (no gun): +25%
Robbery (gun): +14%
Assault w/Dangerous Weapon (no gun): -13%
Assault w/Dangerous Weapon (gun): -53%
Total Violent Crime: +5%

In both PSA 302 and 304, violent crime is up. Homicides are up. Assault with Dangerous Weapon is down. I don't know if that's because of better policing, or that the shooters are more accurate and have turned more ADWs into homicides. Sex abuse is down across both PSAs. That's good too. Plus one for Law and Order: SVUMPDC. But PSA 302 is in bad shape. And now, you've got figures and not just a blogger making unfounded assertions.

Look at those stats for PSA 302 and tell me if a rowhouse for half a million is a "Good Deal or Not."

Here's a Craigslist ad for you, "Now with 200% more homicide!"


  1. WTF? a post with real data and little hate? odd. it is nice to see some statistical data supporting the general feeling that columbia heights is hole.

  2. Yo, smart guy.

    1) Comparing two years does not a trend make. "Metro deaths up 100000000% since last year!! OMG!!" Would you care to report for the last five years? Or ten?

    Or maybe that doesn't support your "conclusion."

    2) Where's a reference to the data?

    3) At least a third of PSA 302 is Petworth, not Columbia Heights. Including the location of the shootings the other night.

    Typical crap from someone who knows nothing about how to interpret statistics.

  3. @Jamie

    PSA302 might include Petworth, hence why I included the map. Crime is actually worse in 304 than 302, so I don't know what point you are making.

    Data is from the MPD web site. Data was not available from periods prior to 2007 (as far as I could get the site to work).

    The data shows that crime has not decreased in 302 or 304 from 2007-2009. How else would you interpret that?

  4. As I said. One year does not a trend make. Isn't that obvious from my example?

    The other thing that makes this completely irrelevant is that the numbers (of homicides, particularly) are obviously very small. If you could post a link to the data (since I couldn't find it at that level yesterday) I could comment on that. But because you're talking about "100%" and "200%" what are we talking about here, from 1 to 2? From 2 to 4?

    The statistical significance of such changes are almost completely irrelevant. A single event in which a few people were killed would account for such an increase. The absence of such an event in one year would account for a similar decrease. This is not a "trend."

    I am not going to try to explain all this here because it would take too long. Go read a primer on statistics. But I hope you can understand at least the basic point. In order for something to be a trend you need:

    1) A pattern of more than a single year
    2) Enough data that it is not extraordinarily skewed by a small number of events
    3) A comparison to a baseline - that is, crime everywhere else

    What you have posted here means nothing. It could mean, "there was a really bad gang fight in 2008 in which 4 people were killed."

    If crime dropped in 2009 by 200%, because such an event didn't happen, would you say "Columbia heights, biggest drop in crime in history! Miraculous turnaround!" Of course not.

    Finally, you are ignoring every other factor that makes this information meaningful. Did crime drop citywide in whatever period you're looking at? Guess what - we entered a recession in 2007. Would it surprise you to know that crime generally increases in a recession because people lose their jobs? Did crime in Columbia Heights increase more in any given analysis period than it did everywhere else in the city?

    It's like the Texas pro-gun people saying "gun crimes in texas dropped 70% in the ten years following the concealed carry law! Woo hoo, we were right!" They conveniently ignore the fact that in the same period, on average, gun crimes decreased even more in the rest of the country. So this piece of information actually means that the concealed carry law caused an INCREASE in gun crimes relative to everywhere else that did NOT enact a concealed carry law.

  5. Your statistical analysis doesn't say a terrible amount w/o tending over a broad number of years, actual data (not %), and comparisons to 'control' neighborhoods.

    "0 homicides one year and 1 the next!? THEY INCREASED BY INFINITY PERCENTAGE!?!? ZOMG!"

  6. "Long Term" trending you can't get from 2 years worth of data.

    However, if you lived in the neighborhood from 2007-2009, and had been told that "it's very safe" I think these figures dispute that.

    Is it safer than 2001? I don't know, MPD doesn't have that data. You can go to their site and type in the PSA you want to see the data from.

    So what if there was a gang fight in which 8 people were killed, 8 people were killed at it made your neighborhood less safe. Is it really 10,000% less safe than the year previous, maybe not, but the fact still remains that more people were killed.

    For comparison, if you run the stats for the same period of time in Mt. Pleasant, all categories of crimes were down.

  7. Dave you keep missing the point.

    Two years of data involving a few incidents MEANS NOTHING! The comparison to Mt. Pleasant means nothing.

    How about this? What if I pointed out that there had been no homicides on 11th Street in five years, whereas there were two on Mount Pleasant Street in the same period.

    Would you therefore conclude that 11th Street is safe and Mt. Pleasant Street is not? Of course not. Small numbers of incidents mean nothing. Comparing small numbers of incidents means nothing. Comparing two years of data means nothing.

    Did you think Georgetown was suddenly "unsafe" after the Starbuck's killing? Do you think Metro is unsafe because there has been a single accident involving loss of life in many years?

    You have an opinion that Columbia Heights is less safe or no safer than it was. You are trying to support that opinion with data that you (erroneously) believe proves something.

    It proves nothing. You don't know what you're talking about. It's that simple. Please go read a statistics primer, you just don't get this.

    For the record, I never said columbia heights was "very safe." You said "it's about the same as it was five years ago." You've said nothing that supports that idea here, and I've given you plenty of data that refutes it, including the obvious (that there are WAY more people living and walking around the streets, while violent crime generally in 3D is down in that period).

  8. Jamie,

    Here's my point:

    Many, many, people consider Columbia Heights to be safe. You know better than this. I know better than this. However, there are a lot of people moving to DC and renting apartments in PSA 304 who are told that it is a safe area, that was "scary a few years ago" but now that DCUSA opened it's safe.

    The data from MPD disputes this claim. Reading headlines disputes this claim. Is the 2007-2009 "uptick" a trend? I don't know. This data can't tell me that. I've since edited the post to clarify my misuse of the word "trend."

    I took a few statistics classes about 5 years ago. I hated statistics but I still have the books around somewhere. That's really beside the point.

    My point (which is my point, and not your point, and not anyone else's point) is:

    In the last two years since Columbia Heights has become a "trendy" place to live, there has not be a decrease in violent crime. Whether these are "isolated" incidents or not, the fact remains that from 2007-2009, there's been no significant drop in crime. Over a period of 1999-2009, I'd imagine there has been a drop in crime. But in the past two years things have not magically changed as many people think.

  9. Dave, this is the point. The data does not support your conclusion. No matter how many times you say the earth is flat, it's still round. I have tried to explain this to you by example, but either you just can't understand it or you choose not to. Either way, it's true, and you really don't know what you are talking about.

    "In the last two years since Columbia Heights has become a "trendy" place to live, there has not be a decrease in violent crime."


    "In the last two years, the number of deaths on Metro train collisions has skyrocketed, from zero to 9"


    "From 1999 to 2000, the number of people losing their lives in airplane crashes doubled."



    Each of those observations is identical to the observation you have made here. None of them means a goddamn thing.

    "Over a period of 1999-2009, I'd imagine there has been a drop in crime."

    First of all, it probably has. Did you not notice the figures I posted the other day showing five years of data from 3D? But even if you think that's not important, what have you put forward to say anything at all about that time period?

    Secondly, why would you expect crime (in absolute numbers) to decrease when suddenly there are tons of people --- with money -- moving into and spending lots of time on the streets in a transitional neighborhood? There are a lot more targets. These thugs probably used to do their fishing in Mt. Pleasant, but now they can do it at home.

    Frankly, I'd be surprised if it didn't go up. But pure numbers of crimes mean nothing. What's also gone up?

    - The population of columbia heights. Has the crime per number of residents gone up?

    - The amount of foot traffic. Has crime compared to the number of people on the street at any given time gone up?

    It should be obvious to ANYONE that given the small change (if any) in crime the overall SAFETY has gone way up. You are far less likely to be a crime victim than you were a few years ago if you walk around or live in columbia heights.

    I'm not going to post again. If you still don't get it, I can't help you. I'm guessing you didn't do too well in that statistics class.

  10. I'm not 100% convinced that your comparison to the Metro accident or plane crashes are applicable.

    The uptick in crime isn't because of merely one incident that caused the percentages to change so much. I get that, I get that you can't say flying on an airplane is less safe because a plane crashed.

    But your example of the Metro would be apt if the accident was caused purely by random chance. But it wasn't. Are you infinitely less safe on the Metro than last year because of one crash, no.

    Why is it incorrect to say that violent crime has not decreased? I didn't say violent crime per capita or violent crime per foot-mile walked, or anything like that.

  11. "Why is it incorrect to say that violent crime has not decreased?"

    It's not incorrect. I just said that in my previous post.

    What is incorrect is to say that this means safety in Columbia Heights is "about the same" as it was five years ago. Because pure numbers of crimes don't mean anything in a vacuum. They depend on how many people live there and how many people are on the streets.

    Your metro observation shows you're not a complete moron. Maybe the crash is indicative of underlying safety problems due to aging of the system.

    However, all you've said about Columbia Heights is "crime went up in one year" to support your original statement that safety is about the same as it was five years ago. And that's all I said about the metro. But hey! You think it means something. But you explained that by providing other relevant information. You continue to ignore the other relevant information about CoHi - the population and foot traffic.

  12. Fair enough. Firstly I'll retract my statement that things haven't changed in 5 years. I was terrified to go to someone's apartment that wasn't even really in Columbia Heights (it was on 16th St, I believe, probably near Park).

    I'll continue to stand by my statement that the neighborhood has not dramatically improved in the last 2 years. I'll continue to seek out data that proves this in a more accurate way.

  13. OOOOMMMMMMGGGGGG SHUT UP! SRSLY! Jaime get over it. You aren't cool because you live in CoHi.

  14. ZOMG! OK srsly i'll sht up cz u sd so, A. U r my hro.

  15. I'm reposting this comment from @Denizen of Tennallytown because I feel like Jamie hasn't seen it yet. It is perfect:

    Denizen of Tennallytown said...


    Don't get your panties in a twist because the recession put the brakes on your dream of buying a dilapidated rowhouse in the ghetto and riding the gentrification wave to make a quick buck, or short of that, at least allow you to live in a neighborhood where, you know, you don't get stabbed on your morning walk to get a cup of coffee from the corner liquor store sellng singles to winos.

    "Cohi" still isn't a safe place to be. Four square blocks of insta-downtown do not a safe place make. Try walking down one of the sidestreets bordering the DC USA complex (lamest name by the way) after dusk. You'll get some friendy stares from the locals sitting of their stoops. As the recession worsens and the summer days get hotter, expect a number of these friendly locals to become even friendlier. I'm sure they could also use a brand new electronics gizmo from Best Buy.

    I stayed at my friend's place on 15th St near Meridian Hill Park. Virtually Adam's Morgan. I'm awakened in the middle of the night by a thump and the breaking of glass. Apparently, somebody with a gun was running from the cops and tried to ditch the heat by throwing it through my friend's window. There's safe for you.

    But keep sanding those floorboards! It's bound to turn around soon, with prices even more inflated than the last bubble... right?!

  16. Oh Jaime! You and I talk in internet abbrevs just like the white girls from PA who live in Cohi do! Sisters!

  17. @A:

    I've lived in DC for 19 years. When I bought a house in Mt. Pleasant, we had this guy called the "shotgun stalker" who scared the crap out of everyone and made people afraid to live in the city. I bought a house for $200K and sold it 9 years later for $900K.

    I really wish I listened to all those people who takled about all the "porch monkeys" and "crackheads" and "car breakins" in Mt. Pleasant. I might still be renting an apartment in Viginia like you. I'd be $700,000 poorer, but I wouldn't have risked running into any threatening "locals sitting on their stoops"

    Laughing all the way to the bank.

    -- Jamie

  18. This comment has been removed by the author.

  19. This comment has been removed by the author.

  20. I said "like the girls from PA". I could tell you were a guy by your aggro commenting style and that you lived here awhile by your deep knowledge of the area. No disrespect to you or your 700 grand intended.

  21. Oh it's all gone now, i lost it all in the stock market.

  22. This comment has been removed by the author.

  23. In the cosmic scheme of things, we'll all be dead, all matter will cease to coalesce into forms that can sustain and create basic physical principles like gravity and light, the universe will collapse, with time accelerating faster than matter around it. Every possible permutation of time will create and recreate itself, we will all live our lives again and again, in perpetuity(are we living them right now?), as time accelerates, faster and faster, with the universe collapsing behind it, chasing the dragon, ever in search of falling into that dark, cold impossibly small spot of singularity. everything. collapsed.

    so technically, the entirety of existence is infinitely dangerous. Your 500k fixer upper on Park Rd was probably a good buy, given its relative safety.