"Facts are stupid things"

September 29, 2004 9:31 PM

There are some questions that social scientists should be able to answer. Either executing people cuts the homicide rate or it does not. A fascinating look into statistics and the death penalty.
posted by LimePi (32 comments total)

Mad props to homunculus, whose previous link to CSICOP tipped me off to this article.
posted by LimePi at 9:39 PM on September 29, 2004

I don't know, the only deterrent effect I know for sure that the death penalty has is that the criminal being put to death is not going to kill anymore people.

And, for better or worse, depending on your point of view, that's the reason I am for capital punishment. Deterring others is incidental to fully punishing someone undeserving of continued life.

I did enjoy the numbers though. How is it possible that not a single person in the US died of radiation in 2001? AND no one got crushed to death by a reptile?

That and its downright scary that I've got a 1 in 17,000 chance of dying by self harm inflicted with a gun!
posted by fenriq at 10:11 PM on September 29, 2004

That's exciting. Those states really do prove that a gun in the home is more likely to kill a family member than a criminal:

Accidental Firearms discharge, W32-W34 -- 802
Assault by firearm, X93-X95 -- 11,348
Intentional self-harm by firearm, X72-X74 -- 16,869

Heheh. I suppose it's just the US government's way of culling the herd.
posted by shepd at 10:15 PM on September 29, 2004

Whoops, missed this one:

Legal intervention involving firearm discharge, Y35.0 -- 323

I still win.
posted by shepd at 10:16 PM on September 29, 2004

I'm ambivalent about this article. I strongly suspect it's right in this instance, wrong in its theme. There are quite a few natural systems, very complex, which yield to multivariate statistical analysis. This article seems like an indictment of econometrics, excepting when in "flatland" (an analogy that seemed inapt to me). All in all, there's an undercurrent of academic/scientific politics: defend social science from the math geeks!
posted by Ethereal Bligh at 10:21 PM on September 29, 2004

Surely the issue is not whether it's effective or not: it's more about the integrity of the criminal justice system in a given country. From what I hear, no state - esp. a state that will not hear new DNA evidence - has yet reached that standard. We may get there at some point soon, though.

Over here in the UK, its an unpopular view to support the death penalty: however, I would support the ultimate penalty on murderers, if it could be shown that
- without question the defendant had top-of-the-line representation,
- that no-one fell asleep in court
- appeals were thoroughly exhausted,
- and there were no circumstances which indicated that the defendant could at some point be safely rehabilitated.

To me, it's not even about deterrence: it's about community safety. All the studies in the world could show that no other murderer was deterred by hanging (or whatever the method chosen, according to national choice) a different murderer; a murderer is not around to kill anyone else. That's good enough for me. I don't see why I should have to keep some killer alive for the rest of his life - who knows, another 50+ years - behind bars, when we aren't at all squeamish about the carnage we create in other ways.
posted by dash_slot- at 10:25 PM on September 29, 2004

The article is itself an example of what it condemns, since the author seems to favor naive descriptivism because that leads to the conclusions he wants to reach.

I don't get the offense he takes at ceteris paribus -- it doesn't make any sense to ask whether or not Texas has a lower or higher homicide rate than other states, because so many factors feed into homicide rates. Doing this would lead to the conclusion that all states should do what North Dakota does, which is almost certainly to be small, rural, and largely homogeneous and little or nothing to do with their justice system.

It would make sense to ask whether Texas has a clearly lower, clearly higher, or neither homicide rate than it would if it did not execute with such Vader-esque enthusiasm. Which is to say, whether or not executions lower homicide rates -- guess what -- ceteris paribus.

Also, he ignores the real problem in work like this -- endogeneity. Does Texas have a high homicide rate because it uses an ineffective death penalty, or is there greater demand for a death penalty because of a higher homicide rate? Teasing these things apart is *hard*.
posted by ROU_Xenophobe at 10:35 PM on September 29, 2004

how could the possibilty of death be a deterrent when death penalty cases are situations in which one is
not thinking/capable of grasping the choice
doesn't care/value life
chosen death

it always seemed more about some form of social justice
posted by ethylene at 10:36 PM on September 29, 2004

Yeah, I share a lot of your thoughts, dash_slot-. Ultimately, I come down against capital punishment. But I don't rule out the idea that capital punishment could be the just response to a crime. I'm not sure I agree that lifetime incarceration is somehow more humane than execution. I don't think it's a deterrent. (And, I don't think that's really the main rationale of most of those that support it, either. They see it as punitive, mostly.) I am very concerned that it is a mistake that can't be corrected, if an innocent person is executed. Although, if I'm not convinced that incarceration is more humane than execution, then the same would apply there. And you couple that with the fact that I don't think the system in the US is comparatively fair between, most importantly, whites and minorities...well, there's a big problem. The system here in the US is badly broken. Capital punishment is just the tip of the iceberg.
posted by Ethereal Bligh at 10:43 PM on September 29, 2004

MetaFilter: Mad props to homunculus.

I like it.
posted by homunculus at 11:16 PM on September 29, 2004

When the state executes a man, it becomes a murderer. When the states sends a man to jail, it becomes a gay dungeon master.

(from some guy's .sig out on the intarweb somewhere)
posted by Captain_Tenille at 11:29 PM on September 29, 2004

and there were no circumstances which indicated that the defendant could at some point be safely rehabilitated.

That sounds oddly close to killing children because they'll never be rehabilitated from their disease. And, oddly enough, a lot of Canadians agreed with him, too.

dash slot-, I'm betting you could find plenty of murder cases where it looked crystal clear that someone committed the crime, but a decade later, it turned out the person didn't commit it.

Death Penalty supporters, weep for these names:

Guy Paul Morin. Donald Marshall Jr. David Milgaard. James Driskell. Romeo Phillon. Thomas Sophonow.

If we had Capital Punishment, we'd have murdered them.

How about a moment of silence for those who have been on death row, but managed to survive through innocence.

Capital Punishment is dead wrong. There's no possible, rational, way to deny someone their right to an honest appeal. And that's what Capital Punishment does.

The only reasonable way to support Capital Punishment is that if the case is overturned after you've killed the defendant, that you may execute the prosecution and judge in the case. They deserve it. They murdered someone. Didn't they?

Even one murder by a government is a disaster to the moral foundation of any country founded on rationality.
posted by shepd at 11:33 PM on September 29, 2004

Interesting in its own right but of interest here: Canada's Murder Rate Lowest in Three Decades in 2003

The national homicide rate dropped to its lowest level in more than 30 years last year, says Statistics Canada.

The rate, which has generally been declining since the mid-1970s, fell by seven per cent, to 1.73 victims per 100,000 population, the agency said Wednesday.

Canada's rate was about one-third the 5.69 per 100,000 people in the United States. It was also lower than England and Wales at 1.93, but slightly higher than France at 1.65 and Australia at 1.63.

Police reported 548 homicides in 2003, 34 fewer than in 2002.

548 homicides in the whole country. Man.
posted by y2karl at 11:35 PM on September 29, 2004

damage is done and no one is trying to mitigate the continuing ripples of said damage--
which is more humane is moot, human experience too subjective
were they to tailor punishments, we'd stray into the extreme and horrific and no one would ever get tried
entering the system-- the subject is too vast
large scale change hurts and drags no matter how pure the end product goal
all i can think of beside what people already do as point of course is to work at a local level, because there are usually stages before it gets all Oz, and if the only sense of community people can find is "inside"--

*can't go on, sings a certain chorus from west side story*
total tangent: been wondering if people can truly move on, if only mental blocks and dissociation distance one enough from trauma, of what they've done or been through--
posted by ethylene at 12:48 AM on September 30, 2004

dash-slot-, and I had you down as a liberal! Miscarriages of justice continue to occur, even when the basis for finding someone guilty is "beyond all reasonable doubt". Why is that? If we unable to come up with a criminal justice system that cannot follow that simple maxim, should we not be putting our efforts into fixing the criminal justice system?

It would be nice to find an answer to the original question. One side or the other would have to work a little harder to justify their position. However, I suspect the question is unanswerable. There are just too many factors at play and we can't even be sure whether A causes B, B causes A, or if there is no correllation between them.
posted by salmacis at 1:14 AM on September 30, 2004

A self comdemning article indeed , as the author discovers models never perfectly describe "reality" yet uses models weakness to show models aren't perfect ! Whoa, I'd rather hear why a model doesn't accurately approximate "reality" instead of hearing that it does because it does.

It sounds much like creationism under a glassing of relatively obscure maths, with the implict assumption 1. given that model A doesn't "exactly" represent "reality" 2. therefore God.
Whoa !

On a related tangent: I can't accept death as a penality as there's is no know way to resurrect "innocents" while there is at least one way to find an "innocent" guilty : reaching the common conclusion that , no matter what, he is guilty.
posted by elpapacito at 2:52 AM on September 30, 2004

Maybe it depends on whether you count executions as homicide.
posted by Pretty_Generic at 3:44 AM on September 30, 2004

In the end, it seems to me that the question is whether or not you think "justice" means "revenge" or not. We've always known that there's no deterent value to the death penalty. Well, the stats have been available forever, at least. In the end, some people want an eye for an eye.

Personally, I have serious issues with a government allowing mourning families to play out their own murder fantasies by killing someone they hate in front of them and calling it justice.
posted by Hildegarde at 4:06 AM on September 30, 2004

If punishment was based on deterring others from committing an offence it would be reasonable to have the death sentence for parking violations, and let terrorists walk free. If you knew that mis-parking your car may result in your execution, how carefully would you park? Yet a terrorist is driven by ideology, and death to them is martyrdom - added bonus! Therefore a death sentence would be ineffective and not a deterrent.

Punishment should be based upon handing an offender a tariff that matches the crime committed; all this deterrent talk is a smokescreen. Who commits a crime thinking that they are going to get caught? Killing people is wrong - no matter who they are or what they have done. It is wrong. It is wrong. It is wrong.
posted by DrDoberman at 4:27 AM on September 30, 2004

Would you bet your life on the fact that the system of justice under which you live is 100% correct in its convictions?

If no then you don't really believe in the death penalty.
posted by i_cola at 4:28 AM on September 30, 2004

Etheral Bligh: "There are quite a few natural systems, very complex, which yield to multivariate statistical analysis."

Well obviously, but the application of econometric modelling to social issues, where the parameters are hugely open to interpretation is due IMHO to the misguided idea that this is a natural science issue. Multivariate analysis is perfect for analyzing, say, gene distributions in particular, controlled, populations, but as the author says:

"this method has consistently failed to offer reliable and valid results in studies of social problems where the data are very limited."

This is not about defending "social science from the math geeks!" Quite the contrary it's about finding the appropriate tools for handling particular data. The conceit. common to many econometricians that they can "mathematicalize" a field by throwing random assumptions at it (which the author points out allows everyone to easily confirm their personal biases), is unfortunately characteristic of a whole "physics envy" mentality in the field.

elpapacito: "the author discovers models never perfectly describe "reality" yet uses models weakness to show models aren't perfect !"
No, the author suggests that certain statistical types of statistical analysis is valid in a particular context, while others are not because, among other things, they can be tweaked to support whatever the analyst wishes them to support.
posted by talos at 4:31 AM on September 30, 2004

Most people commit crimes because they think they can get away with them. The key to reducing crime is in eliminating that belief -- a near-certainty of a 10 year prison term would be every bit as effective a deterrent as a near-certainty of execution, even for a dumb and drug-addled offender (more the rule than the exception).

Vastly increasing policing and surveillance in high crime areas and particularly vulnerable locations, as well as DNA databasing of gang affiliates and people on any initial law enforcement contact, including as juveniles, will do far more to reduce executions than any bleeding heart whining, or (for that matter) any macho, throw-the-book-at-them further increases in mandatory minimums.

There is certainly a small subset of criminals for whom a common jail term is a tolerable, or even perversely honorable, experience. To address that issue we need to make sure that gang affiliates and other people in a subset less likely to fear prison are guaranteed a special sort of jail experience, one that even the most jaded of lifelong criminals will fear to experience.
posted by MattD at 6:17 AM on September 30, 2004

The problem here is probably not that the data are limited -- there's a growing stable of Bayesian techniques that can deal with at least some varieties of limited data, and it seems a reasonably boring-normal cross-section time-series anyway.

The reason we see what we see, with some researchers finding no-deterrence and others finding strong-deterrence, is more likely just that the dependent time-series is simply far too well-known, as are its properties and correlates, and as are the obvious explanatory time series. This makes massaging your data or methodology easier because you can be more confident of what you'll see when you massage it in a given way. Mix this with people have an emotional/ideological/policy bias, and you get lots of biased studies.
posted by ROU_Xenophobe at 6:28 AM on September 30, 2004

a murderer is not around to kill anyone else.

As opposed to someone locked up for the rest of their life?

Shouldn't "retributionists" prefer to have a murderer locked away for the rest of their life to an easy and painless execution? Think about it: your average 30 year-old mass-murderer has 25-30 years of prison rape to look forward to, every single day of their remaining existance. Why let them off so easy?
posted by Civil_Disobedient at 7:14 AM on September 30, 2004

"Well, if I thought I was only going to have to go to jail for the rest of my life, I'd happily kill this guy. But I don't want to die!!!"

The defense of the deterrent argument seems to rest on two very unlikely assumptions:

1) Would-be murderers make some sort of rational cost-benefit analysis.

2) The analysis works out so that it's worth killing if it means life in prison, but not if it means the death penalty.
posted by callmejay at 8:01 AM on September 30, 2004

"Either executing people cuts the homicide rate or it does not."

That is not an acceptable axiom. First of all, the criminal justice system recognizes multiple *kinds* of homicide, from "First Degree" premeditated murder on one extreme to "Reckless Indifference" on the other.

But even distinguishing between these *kinds* of homicide misses what could be called the two *types* of homicide: against strangers and against family member and close associates.

People who kill family members have a very low recidivism rate: their homicide is a "one shot", and after they have killed they represent almost zero threat to either society or even to other members of their family. As a rule

This is, and should be distinguished from "stranger killers", who, for whatever their motivations, kill or murder people they don't know. There is even an intuitive recognition of the inherent difference, here.

So then, the question really becomes, "Does executing murderers who commit (let's say) 1st and 2nd degree murder, who kill strangers and so have a significantly higher chance of recidivism, cut the homicide rate of these types of murder?"

I have no idea, but I also suspect that it would be a very different answer from the one determined above. (In truth, I don't think so, suspecting that the vast majority of these types of murders are committed by psychologically abberant people, as in "dangerous psychopaths.")
posted by kablam at 8:14 AM on September 30, 2004

I liked reading about the first "exhautive study" of the death penalty, one that involved gathering as much data as possible, and from as many techniques as possible. Why isn't this same comprehensiveness demanded of death penalty research today, especially when the issues involved are incredibly controversial, even critical, and there is such limited information involved?
posted by raysmj at 8:32 AM on September 30, 2004

548 homicides in the whole country. Man.

Canada's lower murder rate is accompanied by a higher rate of sex crimes.
posted by oaf at 11:34 AM on September 30, 2004

Surely the issue is not whether it's effective or not: it's more about the integrity of the criminal justice system in a given country. From what I hear, no state - esp. a state that will not hear new DNA evidence - has yet reached that standard. Just thought I'd say it again, as it seems it was not heard the first time. That makes me anti-state murder, today, pro-state murder when the technology and the social process is perfected. I use that word advisedly. Few citizens would shed a tear at the execution of a Harold Shipman, or a Peter Sutcliffe (about whom there is no doubt re: guilt, btw), and I would not be one of them. No basic harm to society is risked by their death at the scaffold.

Salmacis: ha! not so easy to peg, I guess. The only ideology I subscribe to is that which defends my human rights. I do not respect the human rights of oppressors: not many lambs rise from a nap with the lions.
posted by dash_slot- at 5:31 PM on September 30, 2004

Oaf, can you provide some documentation for that?
posted by Hildegarde at 7:06 PM on September 30, 2004

Peter Sutcliffe is in Broadmoor, not a regular high security prison, because he is certifiably insane. Are you saying you support the execution of the mentally incapable?

I'm also pretty sure that the death penalty would have been absolutely no deterrent in either Shipman or Sutcliffe's cases. So what would execution achieve? Nothing more that bloody vengeance. After thousands of years of civilisation, should we not have risen above that by now?
posted by salmacis at 1:16 AM on October 1, 2004

Actually, I forgot that Sutcliffe was declared mentally unfit. I used him as an example of a serial, or mass murderer, and should have chosen better. No mentally incapable, nor juvenile persons should be executed. Ever. Anywhere*.

It would have been difficult for Shipman's lawyers to demonstrate his mental unfitness, considering his 30+ years practice and all.

Just to reiterate: my opinion is not predicated on a deterrence value of the DP. I don't even see it as vengeance. I just don't care to see him live, when he killed. I also don't want to spend lots of money over lots of years protecting, nurturing and feeding a man [nearly always are men] who attacked, and often worse, women & children. I know that it sounds hard, that's ok - it's what I currently believe. I am open to other POV's, and indeed have moved on some issues recently. I await enlightenment, but to be honest, the challenge is to make me care. I do not care about the fate of an certain killer, like Shipman. So why would I keep him alive?

*I was particularly enraged by the public and hasty execution of a female rape victim in Iran recently: she was 16, and possibly had mental health issues. The judge put the noose around her neck. The rapist went free. Google news alert: "Neka Iran Atefeh Rajabi"

I would do all in my power to change a British gov't who did that. I think our gov't should do whatever it can, short of war, to change that regime & every other similar regime. But thats by the by, and just to express what I feel about judicial killing of juvenile victims.
posted by dash_slot- at 4:12 AM on October 1, 2004