This "wonder" you find yourself with is a result of the mismatch between observable data, statistics, and our intuition. On the news we may read about Arson (o

This "wonder" you find yourself with is a result of the mismatch between observable data, statistics, and our intuition. On the news we may read about Arson (or other "angry actions" by individuals) but this is highly negative selection bias - the actual incidence rate is near zero. Because we have a lot of data like this (negative news that you are exposed to) we tend to over-estimate the incidence rate of arson (e.g. 1% instead of 0.001%) So, when you ask the question "Why isn't $action provoking more $arson here", the answer is simple: statistically speaking, each individual is extremely unlikely to do arson (or similar), so even though they've been perturbed, the incidence doesn't really go up. Another way to say it: Everyone knows the ROI on arson is negative, no matter how upset you are, crime is never worth it as a response, because you have zero upside. So almost everyone (overwhelming majority) don't do this (but we might think they would, again because we're exposed to negative bias news data)