[Drumline]

> Murder is rare, but it has a very high impact on society and the surrounding people.

How do you know? What metrics are you using to form this opinion? Or are you using the same gut impression that the article alludes to and the GP specifically states?

By my math each murder could affect 10 people as to leave them in a state equivalent to death and still not come anywhere close to the impact as smoking.

Let me put it another way. Odds are that you've never known anyone that was murdered but there is a pretty damn high you've known someone that died from smoking. BTW, the number of people who die from second hand smoking alone dwarfs the number of people who are murdered.

[interpol_p]

I'm just using my gut impression. Sorry for not being more clear.

My main point was that seeing the raw numbers may also present a skewed perception of the how impactful those deaths were on society. Especially when you see them side by side, and one looks so much more "important" than the others by virtue of being an order of magnitude larger.

You might be right that a death from smoking related disease is more impactful on society than a murder, I have no idea about that and my speaking about it probably confused the point.

[Drumline]

I respect the general thrust of your position in that we as a society must evaluate things like death on all axes including social impact. But until thing like social impact become less nebulous and more quantifiable I will continue to base my opinions on the the "raw" numbers because at least I can compare apples to apples.

[interpol_p]

I'm not suggesting we ignore "raw" numbers. Just that we should realise they are not directly comparable, as they are not all measuring sudden or violent terminations of life.

[Drumline]

They are directly comparable as they both measure preventable deaths. If some types of death weigh more heavily on society than others, if there is some sort of X factor to be considered then how do we determine what it is?

Should we say deaths from murders are 1.5x important as smoking deaths? 2.5x?

This line of reasoning, while a tad ghoulish, is important because decisions have to be made, governance has to be picked, and taxes have to be spent. If it turns out we have no rational systematic way of determining a multiplier I believe it's reasonable to go with the information we have which are the basic facts about what's killing us.

[interpol_p]

We may some methods for quantifying how impactful a death is. A brief search led me to [1, 2] which uses "years of potential life lost before 75 years (YPLL75)"

I think it's quite important not to treat all deaths as equal. It's just a gut feeling on my part, as I don't have the qualifications to make any serious statements about this. But I feel the long awaited death of an elderly relative is vastly less impactful than the sudden death of a child. I feel as though a murder has far more negative consequences for society than the death of a smoker from disease (one obvious one is the amount of resources that must go into finding and prosecuting, incarcerating the murderer.)

So when we compare ~16,000 murders to ~500,000 deaths from smoke-related disease, we also have to compare the toll of 16,000 murder investigations. Thousands of prisoners being entered into the prison system. Thousands of appeals and court cases. As well as the cultural and social impact of having so many humans bear witness to murder and the stress of sudden, violent, loss of loved ones; this is much harder to quantify but should not be ignored.

[1] http://uwphi.pophealth.wisc.edu/publications/other/measuring... [2] http://uwphi.pophealth.wisc.edu/publications/issue-briefs/is...