When P(vote affects outcome) x (Value of outcome A - value of outcome B) < emotional_benefit_of_doing_whatever_you_want. You observe this effect in national po

When P(vote affects outcome) x (Value of outcome A - value of outcome B) < emotional_benefit_of_doing_whatever_you_want. You observe this effect in national politics all the time - for instance, the emotional benefit of voting for a cool guy like Trump drastically exceeds P(vote effects outcome) x (Delta in outcomes). Last time I checked, I had to approximate P(vote effects outcome) by hand because it was too small to represent with double precision floating point numbers.