No More Death Penalty in New Mexico

[DaShi]

And nobody has managed to prove that it is NOT more of a deterrent, so that's not a valid argument against it.

It's actually impossible to prove such a thing. Thus, the policy stance is usually based on evidence of it being a deterrent. Yes, "The absence of evidence is not the evidence of absence." However, you don't make policy based on that.

[SlowwHand]

I love how Texas still thinks it's relevant to specify if a black man or white man is about to die.

Racism is alive and well, even if they pretend it's dead.

I don't know why you felt the need to take that shot. I haven't brought up race, and race is just one of the statistics I posted. It's not a Texas site that I referenced.
That's ok though.

[KrazyHorse]

It's actually impossible to prove such a thing.

It's impossible to prove anything outside mathematics, if you want to take the formal meaning of the word.

The way I'm using it (as it's used to a greater or lesser extent in all fields of study OUTSIDE mathematics) is that there is not yet a great preponderance of evidence for one view or the other.

Thus, the policy stance is usually based on evidence of it being a deterrent. Yes, "The absence of evidence is not the evidence of absence." However, you don't make policy based on that.

?

You are FORCED to make a policy decision here. Either the law contains provisions to execute murderers or it does not. I'm not claiming that the lack of proof against a deterrent effect is reason enough for me to strongly support executions. But claiming that there is no deterrent effect and that therefore the death penalty is unjustifiable is ridiculous, given the state of current research.

[KrazyHorse]

Look, my instinctual preference is for there to NOT be a deterrent effect. State-sanctioned murder is not exactly my preferred crime-prevention technique.

But it's morally repugnant to me to allow this preference to get in the way of study of an issue which has the capability to SAVE LIVES.

[MikeH]

Or maybe it doesn't. If it does maybe it's not worth the extra money, other programs might save more lives for less. If you can't quantify it then how can you work out if it's worth it?

[KrazyHorse]

I didn't say that it WILL save lives; I said that the research has the CAPABILITY to save lives. Just like studying the effectiveness of a new drug does.

I'm not saying that the death penalty definitely does deter murders. Nor am I saying that the death penalty is the most cost-effective means of preventing murder. What I am saying is that blanket statements like yours are unfounded. All the sloganeering in the world is no substitute for actual study of an issue.

[MikeH]

Loads of people have studied it though, with the conclusions there's no evidence it's a deterrent. That's the point. I'm not sloganeering.

[DaShi]

It's impossible to prove anything outside mathematics, if you want to take the formal meaning of the word.

The way I'm using it (as it's used to a greater or lesser extent in all fields of study OUTSIDE mathematics) is that there is not yet a great preponderance of evidence for one view or the other.

Thus, the policy stance is usually based on evidence of it being a deterrent. Yes, "The absence of evidence is not the evidence of absence." However, you don't make policy based on that.

?

You are FORCED to make a policy decision here. Either the law contains provisions to execute murderers or it does not. I'm not claiming that the lack of proof against a deterrent effect is reason enough for me to strongly support executions. But claiming that there is no deterrent effect and that therefore the death penalty is unjustifiable is ridiculous, given the state of current research.

Yes, and the logical decision is to not use the death penalty given the evidence, or lackthereof. Why use something that is costly and unproven? However, laws and rules are not always based on logic.

[KrazyHorse]

Loads of people have studied it though, with the conclusions there's no evidence it's a deterrent

This is simply untrue. Some people find no evidence of deterrence. Others find statistically significant evidence of deterrence.

The statement "there is not yet conclusive evidence that there is a deterrent effect", or "different studies find conflicting deterrent effects" is far, far different from the statement you just made here.

[KrazyHorse]

Yes, and the logical decision is to not use the death penalty given the evidence, or lackthereof.

Really? That's the logical decision?



Here's a model which says that in the face of uncertainty between conflicting results for the deterrent value of the death penalty we should apply it:

One study finds that each execution deters 18 murders as a statistical matter. Another study finds that there is 0 deterrent effect. Both are credible on their face, use plausible models and statistical analyses blah blah blah. Say that the cost differential between life imprisonment and execution is 3 million dollars (from what I recall it's actually in the 1-2 mill range, but whatever).

If we do not apply capital punishment, outcomes are independent of which study is correct:

X murders
Y spent on criminal justice

If we apply capital punishment in c cases then outcomes are as follows, depending on which study is correct:

X + c murders (includes the killing of convicted murderers)
Y + c*(3million$) spent on criminal justice

X - c*(18 - 1) = X - 17c murders (includes the killing of convicted murderers)
Y + c*(3million$) spent on criminal justice

(if deterrence study is correct)

The net change from applying capital punishment is as follows:

c extra murders, c*(3 million$) extra money

(no deterrence)

17c fewer murders, c*(3 million$) extra money

(deterrence)

If we value a human life at 1 million dollars (extremely low; lawsuits value lives at 2-3 million, while safety regulations value lives at 6 million or so) then we merely have to believe that there is a 20% chance that deterrence exists and is of the scale that the pro-deterrence study found in order to make out in expected value.

In actuality, we have no real basis for knowing whether OR NOT we can cause significant deterrence by applying the death penalty. And in the face of such ignorance we cannot make a logical decision for whether or not to apply it without appealing to our Bayesian priors (i.e. prejudices).

[KrazyHorse]

By the way, if we use more realistic numbers for the cost of capital trials/appeals and human life then all we require is a 7% chance that the pro-deterrence study is correct in order to justify capital punishment.

So in terms of expected value, your statement that ceasing to execute murderers is the logical thing to do is exactly equivalent to stating that you're 93% sure that a study which is academically on par with any on the subject is completely wrong.

Very logical indeed.

EDIT: messed up a calc

[DaShi]

Sorry, I glanced over it and saw the phrase "if we value human life at $$$" and stopped reading.

[Kuciwalker]

KH, there's a cost you don't account for: the cost keeping the DP as a politically relevant issue, which it will remain as long as we continue to use it. Once we stop using it, it will start to disappear from political consciousness.

[KrazyHorse]

KH: There's a cost you don't account for: the cost keeping the DP as a politically relevant issue, which it will remain as long as we continue to use it. Once we stop using it, it will start to disappear from political consciousness.

Yes, but the benefit from keeping it politically relevant is that it encourages further study, and we may be able to push those uncertainties in one direction or the other and make a more informed decision at some point in the future. A complete ban in the US would effectively stop the debate in the Western world, and would stop us from getting any more data.

[KrazyHorse]

Sorry, I glanced over it and saw the phrase "if we value human life at $$$" and stopped reading.

Then you're an idiot. Because I wasn't the one who first brought up the issue of cost.

MikeH:

If it does maybe it's not worth the extra money

You:

Why use something that is costly and unproven?

****.