How does crime in DC compare to other cities?

Wading into the mess that is local crime data

Some readers have asked how DC’s crime situation compares to other American cities. It’s a reasonable question that should be easy to answer; but with America’s decentralized law enforcement system it’s actually pretty difficult to answer accurately. As best as I can tell, DC’s crime rate is rather typical of American cities but we have an above-average and rising homicide rate:

• Different cities report crime statistics in incredibly different ways so timely comparisons across many cities are very difficult

• DC’s property crime rate is pretty typical of American cities while its homicide rate is above average

• DC’s rising homicide rate in 2023 is an unfortunate outlier; as most American cities have seen decreasing homicide rates this year

  • This increase in homicide has been concentrated in Ward 8

• Crime in DC is incredibly uneven with different Wards having crimes rates that are comparable to wildly different cities

  • Wards 7 and 8 have homicide rates most similar to New Orleans, LA while Wards 1 and 2 are more comparable to Salt Lake City, UT

• Republican-run large cities are rare in America and generally have less crime than most Democratic cities; but have more crime than top-performing Democratic cities like Boston and New York City

  • Common perceptions of “high crime” cities often wildly differ from per-capita crime rates. New York City for example is enormous with over 8 million residents so there’s always some crime occurring. But on a per-person basis it is one of the safest cities in America.

This entire post could be one long caveat about the vagaries of how cities report crime statistics so I want to quickly lay out how I pulled this data together. I will also note that this “local crime data quality” problem has bedeviled researchers for years and that other writers have certainly done this better:

• The most consistently reported crime is homicide which I pulled from local government websites, media reports or AH Datalytics’ excellent dashboard

• I also tried to pull together crime statistics that correspond to the property crimes and violent crimes that are tracked in https://crimecards.dc.gov/ for a comparable “total crimes” metric

  • Notably, some cities report different categories of crimes than DC. I did my best to only include cities where it was an apples-to-apples comparison of the same types of crimes but this was a manual process and errors are possible.

• I tried to find data for every city that readers requested and did an over-sample of the (relatively rare) cities with Republican mayors due to a reader request

• All 2023 crime rates are per capita (based on the most recent Census “QuickFacts” table) and annualized to control for differences in population and when the cities published their data

  • Per capita only captures residents and not commuters and this impacts different cities’ crime stats in different ways

Starting with overall crime rates in 2023 YTD we can see that DC is pretty typical. The most comparable cities to DC on this metric are Seattle, WA and Atlanta, GA. Property crimes are ~86% of total crimes so violent crime rates have relatively less impact on the overall crime rate:

I’ve shaded DC and each ward red in these graphs to highlight the District’s relative position and how different some of the wards are. It’s surprising that the District’s relative position wasn’t worse given the 30% YTD increase in property crime in DC. In terms of validity, DC being “better than Baltimore, worse than Richmond” seems plausible. I will warn that property crime data has the most variation in reporting which is why I also looked at homicide data specifically:

Here we can see that DC’s overall homicide rate is above average with especially high rates in wards 8 and 7. Those two wards suffer from homicide rates that are higher than some of the most dangerous cities in America like New Orleans, LA or Memphis, TN. As an aside, those two cities are much more than just the touristy areas like the French Quarter or Beale Street and have very serious gun violence problems that dwarf DC’s. Regionally, both Baltimore, MD and Richmond, VA also have above average homicide rates. As many, many Washingtonians have pointed out, this terrible violence in wards 8 and 7 takes place just a few miles from wards with dramatically lower homicide rates and that is evident in the graph. Ward 3 hasn’t had any homicides in 2023 so far and wards 1 and 3’s homicide rates are more comparable to Salt Lake City, Utah.

One way in which DC unfortunately stands out from other cities is that our homicide rate in 2023 has increased YTD. According to AH Datalytics,11 of 18 cities with the most homicides in 2022 have seen decreases (so far) in 2023. Across all 75 cities tracked by AH Datalytics, homicides in 2023 are down in 63% of them:

It’s notable how much city-level crime rates differ from common perceptions of cities. Chicago, IL is often used as a shorthand for murder but less-discussed cities like St. Louis, MO have objectively much higher homicide rates. In general, very large American cities like New York City and Los Angeles tend to have lower crime rates despite reputations as high-crime areas. I think this is driven by a few factors:

• These cities are so populous that there’s always some high profile crime happening. While each person’s risk of being a victim is much less, everyone can see crime happening either in person, on TV or on social media

• These cities often do have higher crime rates than many of their surrounding suburbs. So from a local perspective they are “high crime” even if they have less crime than other American cities.

• There are significant political motives to portray all large cities as “crime ridden”

With those caveats, here are the cities that have the most similar crime rates to each of DC’s wards based on 2023 YTD data:

The “repeats” in that table are due to some wards having very similar crime rates and the limited sample of cities for which I could pull together 2023 data. To address reader inquires about a broader sample of cities (including some governed by Republicans) I had to switch to NeighborhoodScout’s summary of FBI data. The problems with this dataset are that 1. It is from 2021 and 2. It does not match DC Crime Cards data. But it allows us to compare a broader sample of cities on a dataset that is at least internally consistent.

When looking at 2021 overall crime rates from NeighborhoodScout; DC again appears to be pretty typical. While somewhat reassuring, no one should take that as high praise (and again, this is 2021 data). Policymakers in DC should be interested in how we can close the safety gap with Boston or New York City:

Looking specifically at violent crime rates (note this is broader than the homicide rate we looked at previously), we see that DC performs worse than average. The general positioning of DC as “fairly typical but with above average violent crime” that we saw in the 2023 dataset seems to hold here as well:

The question of partisan control of cities and its impact on crime is obviously pretty contentious. Republican-run large cities are pretty rare in America and they aren’t very comparable to the many Democratic-run cities. The Republican-run cities (noted with an *) tend to be much smaller, less dense and just very different from many of these Democratic cities. That said, there are a few broad observations that one can make:

• Most Republican cities have lower crime rates than most Democratic cities. I don’t particularly think this is causal (given the underlying differences) but it is generally true.

• The safest Democratic cities Los Angeles, New York City and Boston have lower crime rates than almost all of the Republican-run cities. While pointing out these three cities may seem like “cherry-picking” it’s also true that they combine for more population than every Republican-led city (on this chart) combined.

• The variation in crime rates within Democratic cities is a lot larger than the differences between Democratic and Republican cities as a whole. This seems like an under-discussed topic and a place where DC could likely learn from other cities.

Cross-city crime comparisons are very messy for a number of reasons. Different cities document and organize crime data in very different ways. Different municipalities represent different shares of their respective metro areas, they can have wildly different daytime and nighttime populations and there’s often local factors (like DC’s dependence on federal agencies for many criminal justice functions) that drive local crime trends. But this exercise does reinforce a few “common sense” findings for DC:

• DC’s crime challenges aren’t unique and we should try to learn from other cities that seem to be having more success like New York, Boston and Los Angeles

  • As staunchly-Democratic cities, they are likely to have policies that could be palatable to DC’s electorate

  • New York and Boston in particular are competitors for some key white collar industries; given that they are rival highly-educated metros in the Northeast Corridor

• Homicide rates in DC are incredibly unequal and are another manifestation of how DC is divided.

  • Where DC stands out negatively from other cities is its homicide rate, and that problem is disproportionately concentrated in wards 8 and 7. They are the wards that deserve increased policing and prevention resources.