I’m reading a very interesting book by Peter Hitchens about the history of modern policing in Britain. The book is called The Abolition of Liberty and it focuses primarily on the differences from early British policing in the 19th and early 20th century in contrast to the modern policing system. So far, Hitchens has identified two major inflection points in the policing in the 20th century. The first in the 1920s with the start of sweeping prison reforms, and the second in the 1960s with the move from foot patrol (so called “bobby on the beat”) policing to “unit beat” or responsive policing that increasingly put officers in cars dispatched from central stations.

The argument that I found most interesting about this second shift is that it fundamentally changed not just the mode of policing, but the purpose. Police officers went from a primarily preventative organization whose officers were tightly ingrained in the community and whose job was to deter crime in their beat area to being a reactive force more like a fire department who only showed up after crimes had already occurred.

Hitchens makes the argument (though I have a few minor quibbles with the way he reports numbers) that this was in part responsible for the increase in crime in England, and especially London, in the latter half of the 20th century. Arrest rates have risen, but so has total crime, since an important preventative stop-gap has been removed and replaced with a reactionary force that can only ever catch criminals after the fact.

Hitchens’ engrossing explanation for this shift from beat policing to more reactionary policing got me interested in the manpower requirements of the old beat policing system. I decided to try and estimate how much manpower would be required to assign one officer to every quarter square mile in the city, 24 hours a day, assuming 8 hour shifts. I picked quarter square mile beats, since I think this is about the maximum territory that a careful, observant human being can reliably patrol once every hour. This makes for a square space about a half mile or roughly six city blocks on each side.

Based on the land area of Seattle listed in Wikipedia, I came up with a manpower requirement of 1,140 patrol offices, or 160 fewer officers than the SPD currently employs. The 1,300 total officers almost certainly includes non-patrol officers who I would expect to make up the majority of the force. Of course, a main point of Hitchens’ argument is that by placing more officers out on foot or bicycle patrols, you actually need fewer officers in support or investigative roles because of the deterrent effect of having so many officers embedded in the community.

I honestly don’t know if beat policing would work in a city like Seattle. I think it would probably would better in the denser, downtown districts, and many of those are already subject to occasional patrol by officers on bicycles. I do find Hitchens’ argument that getting officers out of cars and on foot in high-traffic areas probably makes for better relations with the public and may well have a deterrent effect, but I do wonder about the scaling and efficacy of 100%, 24-hour coverage.

Still, the simple idea that there’s an officer patrolling every neighborhood would make many opportunistic criminals think twice about committing crime, and may very well reduce response times to crimes in progress.

I should note that, despite a number of high-profile instances of discriminatory policing and undue use of force the past few years, it appears that all three metrics helped by beat policing (police/public relations, crime rates, and response times) are improving, according to this Slog post from last year. The Slog also notes that SPD ” … managed to divert 19 officers to non-emergency beats like foot and bike patrols.”

If Hitchens is correct, this diversion of officers towards regular non-car patrols may have something of a sympathetic effect in which an increase of beat policing reduces crime, freeing up more officers for beat policing, further reducing crime, etc.