[jerf]

The idea of the broken window theory isn't that the window was broken was any particular way. The idea is that the visual image of a broken window serves as a signal that the degree of social concern/cohesion/enforcement in the area is low, and will thus encourage further behaviors. Also, breaking the second window is psychologically easier than the first.

I say this without endorsement, just explanation. (I'm ready to believe the evidence suggests it doesn't actually work and that the real drop in crime was another cause, but, on the other hand, academics can sometimes just be contrarians too, it's not just an online thing, and I still give some credence to successful engineering efforts such that one ought to at least consider the original theory.)