There are few actual works I remember as being foundational to my current opinions than this experiment on rats where in any given group, some would end up forcing others to bring them food. It was done with enough different rats that they were able to redo the same experiment by picking only "bossy" rats on one hand, and only "submissive" rats on the other.
However, the results were identical: among the "bossy" group some started ferrying food for others, and among the "submissive" group some started bossing around others to bring them food. This showed to me that social roles aren't innate and can change if positions open up in the social food chain.
Of course rats probably can't conceive this abstract social organization and are doomed to just reproduce it over and over, no matter which rats are selected to comprise it. But we humans can, which means we also have the power to change our own social organization.
This means to me that crime isn't so much a function of actual people who somehow would be inherently evil, but of a social position that exists because we let it exist. Crime itself seems to be better predicted related to wealth, either very low or very high, than by any other inherent human trait. A straightforward answer to crime then seems to be reducing at least economic inequality, which we definitely can act upon, even if we generally don't.