Mike the Mad Biologist

Every week, I try to summarize the D.C. crime stats, because we only seem to talk about crime in cities when the numbers get worse, not when they’re getting better. As I point out, D.C. has seen large decreases in four categories: homicides, thefts of cars, thefts from cars, and robberies (muggings). Homicides seem to be part of a multi-year decline, as in the last two years, homicides have dropped by a third each year, and D.C. seems on pace to do that again; likewise, robberies also have dropped consistently over the same period. The car-related categories have experienced huge drops, which would be expected with a bunch of armed guardsmen walking around. Few people are going to try to boost a car when there are more eyes on the street.

But that has been mostly supposition on my part. A recent analysis by the Niskanen Center seems to support my supposition (boldface mine):

We argue that MPD’s improvement was driven not by headcount but by a tactical shift toward proactive, upstream enforcement — though the department left significant gains on the table by failing to concentrate that enforcement where and when crime was most severe, or to adjust dynamically as crime patterns shifted. We also examine the August 2025 National Guard deployment, which added roughly 2,000 uniformed personnel to D.C.’s streets virtually overnight, and find that it produced a real but narrow improvement: a 24 percent reduction in opportunistic property crime in the first six months, with no measurable effect on violent crime…

Washington, D.C.’s recent crime history contains three lessons that speak directly to the national debate about policing.

…police headcount is not all that matters. Crime fell in D.C. as MPD shrank to its smallest size in half a century. This is counterintuitive only if we assume that the effectiveness of a police department is mostly a function of officer count. It is not. It works through how officers are deployed, what they do when on duty, and whether their activities are concentrated in the places and times where they can actually prevent crime.

The second is that MPD figured this out. The shift toward upstream, proactive enforcement between 2022 and 2025 — fewer officers, more arrests, different kinds of arrests — represents a strategic pivot that contributed to a decline in crime. This is not visible in standard staffing statistics, but it is visible in arrest composition data. Giving credit where it is due requires looking beneath the headline numbers.

The third is that the National Guard deployment demonstrated both the promise and the limits of presence-based policing. It worked — but on the wrong kind of crime, in the wrong places, and at enormous cost relative to what a targeted approach could achieve. Property crime deterrence through visible uniformed presence in public spaces is real. But violence in D.C.’s most disadvantaged neighborhoods requires something more targeted, more sustained, and more strategically aligned with the actual geography of harm.

One thing worth noting is that the Bloomberg/Bowser school of public safety is predicated around making areas that are visited by out-of-city workers and tourists appear safe because police scare away ‘undesirables.’ This can have some beneficial side effects (e.g., fewer car-related crimes), but it doesn’t necessarily make the rest of the city safer without other policy changes. This is one reason why we see the National Guard patrolling some of the safest neighborhoods in D.C.

Anyway, food for thought. And get the Guard out of the D.C.