That's the point, whether the killers are in prison or free, everyone will die. But with the killers in freedom, people will have a decreased life expectancy an

That's the point, whether the killers are in prison or free, everyone will die. But with the killers in freedom, people will have a decreased life expectancy and that decrease - simply on average or also taking into account the distribution - is a meaningful way to measure the impact. But that analogy is not really good, in case of the killers you can pretty much quantify how many people died because of them and therefore you could say not releasing the killers saved two million lives. [1] In case of some additional air pollution you are not able to say who died because of it, you can only observe a change in life expectancy. [1] But even that is only meaningful to a certain extend - if all the killers decided to only kill people seconds before they died from another reason, that two million lives saved really means two million times ten seconds saved.