The NY Times has an interesting piece on statistical studies of the deterrent effect (if any) of the death penalty. For those who want to get straight into fact-free debate, the bottom line is that the evidence is too weak to allow a firm conclusion one way or the other. What’s interesting to me, though is the way in which debates within different disciplines proceed, and the lags in transmission between them. Here I think the NYT story, while excellent in many respects, is quite misleading, presenting a story of deterrence-hypothesis economists facing off against legal critics.
That was pretty much the way things stood in the 1970s, after the publication of Isaac Ehrlich’s study in the American Economic Review claiming that one execution deterred 7 or 8 homicides. Ehrlich used multiple regression analysis (quite difficult and computationally demanding in those days, and correspondingly highly regarded) in an attempt to control for other factors affecting homicide rates and isolate the effect of the death penalty.
Over the next decade, economists learned a lot about the limitations of regression analysis. With limited amounts of data, it’s impossible to avoid mining the data for patterns which are then used to fit the model. And if you try enough specifications on weak data, you can get just about any result you want. A classic exposition of this point was Ed Leamer’s 1983 article “Let’s take the con out of econometrics” which pointed out the fragility of regression analysis on time-series data and picked, as an example, the deterrent effect of the death penalty.
As is the way, there was plenty of back and forth after that, on Leamer’s general claims, his proposal of Extreme Bounds Analysis as a remedy, and on the specific question of the deterrent hypothesis. Still, I don’t imagine any economist who’s been paying even moderate attention for the last twenty-five years would have been surprised by the finding of Justin Wolfers and John Donohue, published in the Stanford Law Review in 2006 that “the death penalty – at least as it has been implemented in the United States – is applied so rarely that the number of homicides that it can plausibly have caused or deterred cannot be reliably disentangled from the large year-to-year changes in the homicide rate caused by other factors. … Sampling from the broader universe of plausible approaches suggests not just ‘reasonable doubt’ about whether there is any deterrent effect of the death penalty, but profound uncertainty – even about its sign”
The other point made in the NYT article is that economists tend to believe that people respond to incentives, and no doubt the favorable initial reception of Ehrlich’s work reflected this. But once economists started thinking about actual magnitudes, the story fell apart. The probability of being executed as a result of committing a homicide is tiny, and pales into insignificance compared to, say, the risks of being a street crack dealer as estimated by Levitt. In fact, I recall, but can’t verify a claim that the mortality rate for prisoners on death row is actually lower than that for prisoners in general.
By contrast with this, I was struck by this passage from Cass Sunstein and Adrian Vermeule also in the Stanford Law Review, and quoted by Wolfers and Donohue
More recent evidence, however, has given new life to Ehrlich’s hypothesis. A wave of sophisticated multiple regression studies have exploited a newly available form of data, so-called “panel data,” that uses all information from a set of units (states or counties) and follows that data over an extended period of time. A leading study used county-level panel data from 3054 U.S. counties between 1977 and 1996.23 The authors found that the murder rate is significantly reduced by both death sentences and executions. (scare quotes in original)
The reverence with which panel data studies (pretty old hat to economists) are described suggests that not much in the way of critical analysis is going on here. Wolfers and Donohue point out that the study cited here (by Dezhbaksh et al) doesn’t even have fixed year effects, a strategy Wolfers and Donohue describe as “a clear outlier in the literature”. They might have gone on to point out that it’s an outlier because such an omission would lead to immediate rejection at most econ journals.
Like Wolfers and Donohue, I don’t think statistical evidence on the deterrence hypothesis is ever going to get us far. On the other hand, US experience has shown beyond doubt that innocent people are regularly sentenced to death, and it’s virtually certain that some have been executed (the legal system doesn’t provide any real way of establishing this, except in the rare cases when the real killer is convicted subsequently). That’s not the only argument against the death penalty, but it’s sufficient in my view
{ 2 trackbacks }
{ 68 comments }
abb1 11.20.07 at 11:41 am
The argument against the death penalty, just like the argument against torture, is that it’s too barbaric.
stostosto 11.20.07 at 12:16 pm
abb1
Barbaric, maybe, but if it worked as an effective deterrent, you’d have to consider the trade-off.
Of course, there is another factor: The barbarians. That is, the use of the death penalty as an instrument to satisfy the desire for revenge – an ancient and “barbarian” form of justice.
And of course, both arguments also apply on torture.
DB 11.20.07 at 12:32 pm
Downloadable link to Leamer paper at http://www.international.ucla.edu/burkle/article.asp?parentid=2504
abb1 11.20.07 at 12:49 pm
Nah, sto, we would only have to consider the trade-off if it were somehow critically, irreplaceably effective, which is obviously not the case since a lot states without the death penalty have lower murder rates than many of those with it.
A number of barbaric practices could potentially be effective: cutting hands, tattooing ‘thief’ on foreheads. We are not considering those. Well, not yet anyway.
harry b 11.20.07 at 1:43 pm
John — suppose the death penalty could be shown without doubt to have a strong deterrent effect, saving, say, 15 lives for every execution. And suppose we could establish that one out of every 10 executees was innocent. Would you still oppose the death penalty?
Kieran Healy 11.20.07 at 1:58 pm
Here’s a post of mine from a couple of years ago on the Sunstein and Vermeule paper. Me back then:
“[S&V’s threshold claim] is an elegant idea, but trouble with it is that only few states execute anyone in a given year. Most execute no-one. A tiny few—notably Texas—kill a lot of people in some years. As a result, evidence for a threshold deterrent effect depends on a very small number of observations. In a nice analysis of state-level data from 1977 to 1997, Richard Berk shows that just eleven state-year observations out of a thousand drive the deterrent effect. It’s possible to mess around with the specification a bit to get a less strongly skewed measure (by standardizing the number of executions by the number of death sentences, say) or making the data more fine-grained so that you have more observations (using county-quarters as a unit, for instance), but in the end its hard to escape the worry that about 1 percent of the observations are behind the results.
We’re probably witnessing the birth of a dubious stylized fact about deterrence and the death penalty. I don’t doubt that the Sunstein and Vermeule paper raises a bunch of interesting questions, but the empirical results they rely on just don’t seem that robust. This is a bit ironic given their argument that “The widespread failure to appreciate the life-life tradeoffs involved in capital punishment may depend on cognitive processes that fail to treat ‘statistical lives’ with the seriousness that they deserve.†One of these processes is the tendency to latch on to a cool finding a bit too quickly. Negative results (like the ones reported in Berk’s paper) are just not as interesting, unfortunately.”
rea 11.20.07 at 2:40 pm
It’s anecdotal, of course, but you can find some murderers who fantasize about being executed, and for whom the death penalty is not a deterrent but an encouragement.
The one murderer I represented (a very sad, crazy boy) had a well-developed fantasy about the electric chair, and was quite disappointed to learn that our state has no capital punishment.
John Emerson 11.20.07 at 2:48 pm
There’s a whole bloc of Midwestern states where there’s no death penalty: Michigan, Wisconsin, Minnesota, Iowa, and North Dakota. This is an old tradition there, and Wisconsin has only executed one person in its entire history.
Four of these states rank from 34th to 50th in murder rate, with the worst of them (Wisconsin) having a murder rate slightly more than half the national average.
The remaining state, Michigan ranks 18th in murder rate with a rate very slightly above the national average. Michigan is the only one of the five with a significant black population (14%, slightly higher than the national average) and also includes the ruined city of Detroit.
The states with the top ten murder rates all have the death penalty: Alabama, Arizona, Arkansas, California, Louisiana, Maryland, Mississippi, Nevada, Oklahoma, Tennessee (not in order, and these standings are from 1995 whereas my list of death-penalty states is contemporary.)
What these numbers suggest to me is that murder causes the death penalty, and not the other way around. Or perhaps people in Texas believe in killing of all kinds more than people in Michigan do. Race is an obvious factor too, of course.
When arguing with death penalty proponents I also end up concluding that the death penalty has some kind of very deep emotional or existential meeting for them, and that deterrence or reduction of crime are of secondary importance. (The same is true of death penalty opponents, except that I feel generally safer in the company of the latter group than I do in the company of the most devoted supporters of the death penalty).
Now, what I reported just now is statistically naive and partially impressionistic, but I don’t think that it’s erroneous, and these statistically naive results look pretty robust to me. When conclusions reached by sophisticated methods disagree with those reached by naive but valid methods, shouldn’t the naive methods should be favored? Shouldn’t statistical sophistication be used only when no conclusion jumps out at you?
Matt Weiner 11.20.07 at 2:59 pm
Could someone give me a nickel summary of how the panel data for the counties are supposed to work? (To put my priors on the table, it seems to me that with so few executions spread across so many counties the data are bound to be incredibly noisy — is there something about “panel data” that filters that out?)
Jack 11.20.07 at 3:00 pm
Some countries without the death penalty have low murder rates so there are other things that can prevent murder so even if Harry’s hypothesis were established so that executing 51 people including 5 innocents to eliminate non-execution murder in the UK you could not conclude that introducing the death penalty was right. There might be ways of achieving the same result that don’t involve killing.
Also if you assume a high ratio of deterrence and you don’t like killing, only a fraction of murderers should be executed. If that was by a rule then some murderers would not be up for execution and therefore could not be deterred by execution so you would either have to over kill or introduce a lottery.
These are marginal results. It is not true that everywhere and at all times the ratio of deterrence is 7 or 8 or 11.
What other things could be penalised by death with a net saving of lives? Dangerous driving? Smoking? Gun manufacture?
Tom T. 11.20.07 at 3:02 pm
And if you try enough specifications on weak data, you can get just about any result you want.
But why are the results coming out consistently in favor of deterrence? Do all of the researchers want a positive result, and none a negative? One of the professors quoted states that he is opposed to the death penalty. Why would he want a positive result?
John Emerson 11.20.07 at 3:02 pm
Of the other states without the death penalty, Vermont, Rhode Island, and Maine rank below #40 in murder rate. Hawaii, Massachusetts, and West Virginia rank from #31 to #39. New York and Alaska are about where Michigan is, at #18 (tied) and #14 respectively. So the top 13 murder states are also death-penalty states.
D.C. is the jewel on the crown of the death-penalty advocates: D.C. would be #1 in murder rate if it were a state, and it has no death penalty.
Tom T. 11.20.07 at 3:04 pm
8. Wouldn’t another hypothesis be to consider cause and effect the other way around; that the states with the higher murder rates believe that there is a greater need for the death penalty, while the states with lower murder rates believe they don’t need it?
Bloix 11.20.07 at 3:13 pm
Puh-leez. DC is not a state, it’s a city, with a population of 580,000. In murders per capita, it ranks seventh in cities with populations over 250,000.
John Emerson 11.20.07 at 3:19 pm
More naive stats: seven states have actually executed 4 or more people total in the last 2 years, with Texas executing more than the whole rest of the U.S. combined: Texas, Oklahoma, Ohio, Alabama, North Carolina, Virginia, and Florida.
Of these states Oklahoma and Alabama rank in the top ten for murder, Texas and North Carolina in the next ten (and above the national average), with Virginia, Florida and Ohio ranking from 22nd to 28th, slightly better than the national average. Only the last three of these have lower murder rates than the most murderous of the non-death-penalty states (New York, Michigan and Alaska).
Actual executions.
As for the motives of the statistical perps: lots of moderates and conservatives really need terribly to disagree with weenie liberals about everything, and virtuoso statisticians really want to show their stuff and are willing to ignore naive data coming from long-standing natural experiments.
But then, as Sebastian is about to say, “D.C.! D.C.! D.C.!”. That’s their natural experiment.
lemuel pitkin 11.20.07 at 3:22 pm
Over the next decade, economists learned a lot about the limitations of regression analysis. With limited amounts of data, it’s impossible to avoid mining the data for patterns which are then used to fit the model. And if you try enough specifications on weak data, you can get just about any result you want. A classic exposition of this point was Ed Leamer’s 1983 article “Let’s take the con out of econometricsâ€
Hey, thanks, this looks really good.
John Q. or others: any recommendations for the best couple of books/articles/web thingies on the limits of regression analysis (and related techniques), the basic requirements for them to be useful, and so on?
I have my own priors in this area (e.g. that cross-country regressions are worthless) but it would be nice to see them backed up by (or refuted, I guess; that would not be as nice) by the experts.
Thom Brooks 11.20.07 at 3:46 pm
A few points are worth mentioning. First, if I recall correctly, the worry many had with Ehrlich’s study was that the greatest deterrent effects of capital punishment appeared during the years immediately following Furman v Georgia when the death penalty was temporarily banned in the United States.
Secondly, I am surprised there has been no mention of Bowers and “the brutalization effect”: this is the claim that murder rates tend to rise slightly, not fall, after executions. There has been some work on this and I am surprised little has been said.
Third, more generally, some may be interested in this piece I wrote on why retributivists should reject capital punishment.
Kieran Healy 11.20.07 at 3:47 pm
Richard Berk’s book Regression Analysis: A Constructive Critique is accessible and good.
David in NY 11.20.07 at 4:19 pm
The “people respond to incentives” notion applied to this situation is naive to the point of idiocy. There are incredible disincentives to murder even without the death penalty. Murder is subject to life imprisonment, often without parole, just about everywhere. The idea that there’s a murderer somewhere saying to himself, “Gee, I’ll go ahead and commit this murder because the worst that can happen is I’ll spend the rest of my life in prison; thank God there’s not a death penalty, because then I wouldn’t,” or “Gee, I won’t commit this murder because of the death penalty, although I would if I were only facing life in prison,” is nuts. Just nuts. And if economists think murderers think like this, they should get out and meet a few more murderers. Doesn’t economics have a way to deal with actors that are not, almost by definition, “rational” in the conventional sense? Because that’s what we’ve got here.
David Kaib 11.20.07 at 4:56 pm
Here’s my fundamental problem with this whole endeavor – these models include measures of the perception, by potential criminals, 1) that they will be arrested if they commit a murder, 2) that they will be convicted if arrested, 3) that they will be sentenced to death if convicted, and 4) executed if sentenced to death, based on the state in which they commit the crime. These perceptions are measure by the actual likelihood of these things happening, in a particular state. We are asked to assume that the actual likelihood of each of these things is a reasonable measure of the perceptions of potential criminals about those likelihoods, based on their personal knowledge or through the media. But it’s not reasonable, and in fact, if spelled out, is colossally implausible. This isn’t just an issue of false assumptions, but of plain bad measures.
goatchowder 11.20.07 at 5:07 pm
Anyone who believes in the fairytale of the “rational actor” has obviously never sold anything for a living, nor paid close attention to politics or to mammalian ethology in general, for that matter.
Walt 11.20.07 at 5:21 pm
tom t.: They clearly don’t all give the same result. If you put John E.’s stats in a regression, you would get that the death penalty increases the murder rate.
John: I think about the question of naive versus sophisticated statistics pretty frequently. It’s tempting to think that naive stats should be preferred to sophisticated statistics, but it’s easy to imagine situations where the naive statistics give you an answer that’s the exact opposite of the truth. At the same time, sophisticated methods tend to be more fragile (there are exceptions), and tend to be more sensitive to the truth of your underlying assumptions.
Jim Harrison 11.20.07 at 5:23 pm
Economists do badly whenever the boundary conditions are more important than the equations. At the margin, it may be the case that increasing the likelihood of capital punishment in a particular area will decrease the murder rate there in the next couple of years, but the endorsement of barbarous punishments is part of the culture that sets the murder rate at a high level in the first place. Texans are legal and illegal killers because, like the citizens of other southern states, they are enthusiastic fans of violence, vengeance, and intolerance. This is a mystery?
lemuel pitkin 11.20.07 at 5:26 pm
Anyone who believes in the fairytale of the “rational actor†has obviously never sold anything for a living, nor paid close attention to politics or to mammalian ethology in general, for that matter.
The rational actor is a good approximation provided that (a) there’s a mechanism providing for differential survival and growth/reproduction and (b) that survival/growth is directly based on whatever it is the actors are regarded as maximizing.
So to regard firms as profit-maximizing is a useful approximation, tho it breaks down in all kinds of situations where success becomes decoupled from profitability. But it’s not at all useful for other kinds of human behavior.
That our dominant social science proceeds on the basis that human beings are all just little capitals is about as perfect an example as could get of the idea of hegemony.
Walt 11.20.07 at 5:27 pm
Lemuel: Kennedy’s A Guide to Econometrics talks about the implicit assumptions and limitations of regression, as well as some other methods.
Sebastian Holsclaw 11.20.07 at 5:40 pm
“The death penalty “is applied so rarely that the number of homicides it can plausibly have caused or deterred cannot reliably be disentangled from the large year-to-year changes in the homicide rate caused by other factors,†John J. Donohue III, a law professor at Yale with a doctorate in economics, and Justin Wolfers, an economist at the University of Pennsylvania, wrote in the Stanford Law Review in 2005. “The existing evidence for deterrence,†they concluded, “is surprisingly fragile.â€
Gary Becker, who won the Nobel Prize in economics in 1992 and has followed the debate, said the current empirical evidence was “certainly not decisive†because “we just don’t get enough variation to be confident we have isolated a deterrent effect.â€
But, Mr. Becker added, “the evidence of a variety of types — not simply the quantitative evidence — has been enough to convince me that capital punishment does deter and is worth using for the worst sorts of offenses.—
I’ve seen a lot of progressive programs turn on less evidence than “we aren’t certain we have isolated a detterent effect. Want to revamp the entire economic system on happiness studies anyone?
“The other point made in the NYT article is that economists tend to believe that people respond to incentives, and no doubt the favorable initial reception of Ehrlich’s work reflected this.”
Well actually, most people in the world believe that people respond to incentives. There may be academic sub-specialties that disagree for some reason but as a general rule people do respond to incentives. The discovery of the incentives of punishment and reward date back to pre-history. There are indeed border-cases where people don’t respond as we would expect to incentives, but those are border-cases—treating them as THE RULE is going to get you in trouble analytically. So if the very small number of death penalty cases seems to show a small effect, it isn’t a great argument to say that “The probability of being executed as a result of committing a homicide is tiny, and pales into insignificance compared to, say, the risks of being a street crack dealer as estimated by Levitt.†This argues as much for increasing the probability of being executed as a result of committing a homicide as it argues against the death penalty.
Katherine 11.20.07 at 5:47 pm
I would like to submit this to the debate:
http://www.amnesty.org.uk/videos.asp?id=118
It represents, fairly well I think, a response to the utilitarian argument for the death penalty.
And on an entirely emotive note – in 2006, 91 per cent of all known executions took place in China, Iran, Pakistan, Iraq, Sudan and the USA. To steal the words of the West Wing – that’s a club that the US wants to be a member of, right?
abb1 11.20.07 at 6:10 pm
21: these models include measures of the perception, by potential criminals, 1) that they will be arrested if they commit a murder, 2) that they will be convicted if arrested,…
Why, this could be some kind of a crime of passion, where the arrest and conviction are pretty much guaranteed. As you squeeze the throat of your unfaithful lover you’re thinking: “yeah, this sure feels good, but what about that Sizzlin’ Sally? Nah, it ain’t worth it.” So, there you go, a life is saved.
On the other hand, if you’re a career criminal who’s already committed a capital crime, you have the incentive to be as ruthless as necessary to avoid being captured, because your life is at stake and you have nothing to lose.
Terminus Est 11.20.07 at 7:15 pm
Putting aside the atrocity of dispassionate, regularized murder by the government for a minute…the government-sanctioned murder, via death penalty, of a single innocent individual is unacceptable to the point of rendering any other argument moot.
Here’s the follow-on question for those who might disagree: If the execution of “only” one innocent person out of 10 is OK if it prevents/deters 15 murders, does that apply if that one person is YOU? Or your wife, boyfriend, husband, son, daughter…?
I think we all know the answer.
John Emerson 11.20.07 at 7:17 pm
The evidence of a variety of types — not simply the quantitative evidence — has been enough to convince me that capital punishment does deter and is worth using for the worst sorts of offenses.
This guy certainly seems to trim his methodology to fit the conclusions he wants. I guess that’s how you win Nobels.
It would seem to me that the comparison New York vs. Texas, two country-size states, should count as data. The “not enough executions” problem doesn’t hold, because Texas executes more than 20 inmates a year, and New York none.
In 2006 New York had 4.8 murders per 100,000 people and Texas 5.9. The national rate was 5.7.
According to the following link, every non-death penalty state except Michigan has a murder rate less than the national average: Link.
It’s true that if you’re smart enough, you can find ways to make the data seem inconclusive if you want to. That’s how science works — you just analyze the material until it’s unintelligible.
Detlef 11.20.07 at 7:25 pm
Sebastian,
if people respond to incentives, wouldn´t it make more sense to try and raise the US clearance rate for murder? Instead of just executing some (in a worst case including innocents) and hoping for the best (some deterrence)?
Clearance rate for murder in the USA 2006: 60.7%
Clearance rate for murder in Germany 2006: 95.5%
USA: 5.7 murders per 100.000 inhabitants 2006
Germany: 1.0 murders per 100.000 inhabitants 2006
(In both cases murders and nonnegligent manslaughter)
Sources:
FBI Uniform Crime Reporting Program
German BKA Federal Crime Statistic
Seems to me that actually getting arrested is a much larger deterrent. Not to mention that if an innocent person is actually convicted of murder, in Germany that person will be still alive…
Sebastian Holsclaw 11.20.07 at 7:30 pm
“if people respond to incentives, wouldn´t it make more sense to try and raise the US clearance rate for murder?”
This could just be that I’ve missed some reporting phraseology, but this is the second time in a week I’ve seen ‘clearance rate’ for crimes after never having seen the phrase before. What makes it different from ‘conviction rate’?
Bloix 11.20.07 at 7:59 pm
“Clearance rate” is a statistic kept by the police, and means the ratio of crimes “cleared” divided by the crimes reported. “Cleared” means ending in an arrest. Prosecution and conviction are handled by different authorities who keep different statistics. So a crime can be “cleared” without ever resulting in a conviction.
This may seem unreasonable but the police use the statistic as a measure of their own performance – eg what percentage of each kind of crime is cleared, how long does it take to clear different kinds of crimes, what is a precinct’s clearance rate, etc.
Barry 11.20.07 at 8:09 pm
For example, if charges were brought against somebody, and the person was acquitted, the case might still be marked as ‘cleared’. I’m not sure what else goes into it.
The main thing is that the 65% clearanced rate for the USA doesn’t mean a 65% conviction rate, so things are even worse than one might think.
Barry 11.20.07 at 8:21 pm
Please ignore my previous comment – bloix got there firster and better)
I gotta love Becker – honesty is not something that he’s encumbered by. Detlef has noticed that the death penalty advocates are strenuously working for a deterrant which is so weak that it might or might show up in US data, using sophisticated methods, while they could put their energies into increasing the clearance rate, which would be unopposed.
John Emerson 11.20.07 at 8:24 pm
From the Times:
According to roughly a dozen recent studies, executions save lives. For each inmate put to death, the studies say, 3 to 18 murders are prevented.
The effect is most pronounced, according to some studies, in Texas and other states that execute condemned inmates relatively often and relatively quickly.
So Texas had 1384 murders in 2006, with 24 executions. Each execution prevented at 3 to 18 murders, so without the executions Texas would have had somewhere between 4000 and 24000 murders. The latter figure would be about 100 murders per 100,000 people, more than any country in the world (more than twice as many as Jamaica, Venezuela, South Africa, or Colombia.)
Someone show where I’m wrong. Some very sophisticated studies have come up with some results which don’t seem to fit with other easily-available information, and it’s being taken very seriously.
Often at CT I’m just trolling, but in this case it seems that we really are dealing with “rational fools” with axes to grind who should be retired from whatever business they’re in.
http://www.disastercenter.com/crime/txcrime.htm
http://en.wikipedia.org/wiki/Murder_rate
http://www.deathpenaltyinfo.org/article.php?did=186&scid=
John Emerson 11.20.07 at 8:27 pm
37: The second paragraph is also from the Times piece and should be italicized.
Raghav 11.20.07 at 8:34 pm
I haven’t had a chance to read it yet, but Dezhbakhsh and Rubin have written a response to the Donohue and Wolfers Stanford Law Review article. The final summary and footnotes at the end accuse Donohue and Wolfers of impropriety, which is interesting given that Rubin’s Senate testimony was partly cribbed from a paper by Joanna Shepherd.
Sebastian Holsclaw 11.20.07 at 8:41 pm
“For example, if charges were brought against somebody, and the person was acquitted, the case might still be marked as ‘cleared’.”
This seems like a fairly dubious cross-national comparison statistic (and for that matter I wouldn’t want to compare California to Mississippi on that basis either). Arrests are not the same as ‘solving’ a crime and jurisdictions vary in their standards from New York to Berlin to DC.
John Emerson 11.20.07 at 8:47 pm
If a very high proportion of murders are not cleared, however, you know for certain that a high proportion of murderers escaped completely, even if most of the murderers charged were convicted. IE, a high clearance rate might not be a good thing if police are faking it, but a very low clearance rate is unquestionably a bad thing.
lemuel pitkin 11.20.07 at 8:59 pm
So Texas had 1384 murders in 2006, with 24 executions. Each execution prevented at 3 to 18 murders, so without the executions Texas would have had somewhere between 4000 and 24000 murders. … Someone show where I’m wrong.
24 x 18 = 432. 1384 + 432 = 1816. That’s where.
John Emerson 11.20.07 at 9:05 pm
Too good to be true. Plugged in the wrong number.
John Emerson 11.20.07 at 9:13 pm
It would seem that something as drastic as the death penalty would have to be justified by more than Becker’s guy feelings, and more than sophisticated statistical manipulations. If the deterrent effect doesn’t jump out at you, perhaps it should be ignored.
Unless you have presupposed that the abolition of capital punishment is the drastic step, and that the reinstituion of capital punishment is the default.
John Emerson 11.20.07 at 9:22 pm
“gut feelings”
abb1 11.20.07 at 9:34 pm
Seems to me if what you want is the deterrent effect, you have to go much more drastic than the way it’s done now. Something like lynching or old-western-style hang-’em-high instant justice with no appeals, no nothing. That’s my gut feeling.
abb1 11.20.07 at 9:38 pm
Nah, come to think of it, that wouldn’t work either; this is how gangs operate and it doesn’t deter them.
geo 11.20.07 at 10:04 pm
terminus est @#30: “If the execution of “only†one innocent person out of 10 is OK if it prevents/deters 15 murders, does that apply if that one person is YOU? Or your wife, boyfriend, husband, son, daughter…?”
But won’t those fifteen unmurdered people be other people’s wives, boyfriends, husbands, sons, daughters … ?
Dan Karreman 11.20.07 at 10:48 pm
Sebastian,
I suspect that reasonable doubt may explain the difference between clearance and conviction. For example, the murder of Swedish prime minister Olof Palme yielded one conviction that was overturned on appeal on procedural grounds. There is no other suspect – though a lot of grassy knoll theories – and the Swedish police unofficially considers the murder “polisiärt uppklarat” which roughly translates to cleared. The OJ Simpson case is roughly an American equivalent, I presume. The JFK murder is also a case that can be considered cleared without any conviction.
On a related note, I’m surprised than no one has made the conservative case against the death penalty – the government screws up all the time and thus can’t be trusted to executive people. Or is capital punishment an exception from this provision? Sebastian?
Dan Karreman 11.20.07 at 11:14 pm
“Executive” in #49 should of course be “execute”
Sebastian Holsclaw 11.20.07 at 11:43 pm
But if cleared is ‘arrested’, I’m not likely to think they are good comparisons.
And I’m perfectly open to the argument that I ought to have a seamless garment of government incompetence.
derrida derider 11.21.07 at 12:29 am
But sebastian, surely you should then be arguing that arrest, convictions and executions should be privatised? I’m sure Blackwater would be happy to take a no-bid contract for it.
David in NY 11.21.07 at 12:59 am
Exoneration Using DNA Brings Change in Legal System
By SOLOMON MOORE
Published: New York Times, October 1, 2007
“In a 2007 study, Professor [Samuel R.] Gross [of the U. of Michigan Law School] analyzed 3,792 death sentences imposed from 1973 to 1989 and found that 86 death row inmates, or 2.3 percent, had been exonerated through 2004.”
This is a pretty high error rate.
And I’m sorry, the notion that an execution saves 3 to 18 murders seems to me simply absurd. Anyone who believes it simply has not met as many murderers as I have.
John Emerson 11.21.07 at 1:22 am
53: Even if an innocent were executed in every single case, economics tells us that on the net we would be ahead by at least two and as many as seventeen wrongful deaths per execution, whereas each wrongful execution avoided would meand that anywhere from three to eighteen innocents would be murdered. Science doesn’t lie.
David in NY 11.21.07 at 1:34 am
Well, john, I’m relieved you pointed that out. On reflection, I’m sure all those innocent folks getting lethal injections will be thinking, “My only regret is that I have but one life to give for my country.” In fact, maybe death penalty proponents will volunteer for execution as a sign of their patriotism and devotion to their fellow man? I mean, it’s not often that one has the opportunity to save, maybe, 18 lives!
David in NY 11.21.07 at 1:39 am
Or at least they will be saying, “‘Tis a far, far better thing …”
ken melvin 11.21.07 at 1:57 am
Hell, the penalty doesn’t even prevent the murder of those executed.
SG 11.21.07 at 3:54 am
shouldn’t Texas just randomly execute 450 people a year? To get full deterrent effect they’d probably have to have been near the crime when it happened, but in most cases it isn’t that hard. At the minimum deterrent effect of 3 innocents per execution, that would mean that the rate of actual murders would drop to 0. So Texas would have cut the murder rate by 67%.
It’d be kind of a state-run execution-based murder-deterrent system, and the great thing is that there is no chance of government incompetence preventing its success. The only thing the government can do wrong is accidentally randomly kill someone who actually committed the murder, in which case presumably that execution saved 4 lives instead of 3. This would have to be the only govt program to promote a social good (less murders) which cannot be ruined by mismanagement – even Sebastian would be happy. And if the deterrent effect is more like 18 lives per execution, by executing less than 100 innocent people Texas could reduce their murder rate by nearly 95%. Brilliant!
sure, government-subsidized execution-based murder-prevention does smack a little of social engineering, but all the utilitarians out there would have to be happy with the results. Practical libertarians would even be able to give up all those guns which they have only been keeping in the house for vital self-defence. Think of the boost to the economy of people spending savings previously sequestered for personal defence on productive economic activity.
It’s a win-win! I might suggest it to Rudy Giuliani to help him win the presidency…
Tom T. 11.21.07 at 4:50 am
Doesn’t #30 boil down to the old moral/philosophical conundrum as to what to do when the choice is to take an action that will kill one person, or take no action and allow six people to be killed? I.e., the bus driver has fainted and the bus is heading toward a group of six children, but if you grab the wheel and swerve away from them, the bus will run over an old man. One can vary the relationships (e.g., the old man is your father), of course.
John Emerson 11.21.07 at 4:57 am
Don’t get me started on trolley-car philosophy! Just don’t!
Thom Brooks 11.21.07 at 10:57 am
Well, here’s a new problem to consider:
*If* capital punishment can be demonstrated to have a deterrence effect, *then* why limit capital punishment only to murderers?
I say this noting a few items. First, from what I have seen, the evidence is far from conclusive anyway. Secondly, –as highlighted by Dan Kahan in “The Secret Ambition of Deterrence” Harv. L. Rev.— people may *claim* that a view one way or another on deterrence supports why they support or oppose the death penalty, but when asked if the empirical evidence were different virtually no one changes their position. His view was then that support or opposition was linked, not to the empirical evidence, but retributivism, etc. which served as a primary justification for whichever position a person held.
This might suggest that those who support capital punishment for murderers alone on the grounds that it deters might actually be supporting capital punishment primarily on the grounds that murderers, not rapists or traitors, *deserve* death. The argument then is about desert, *not* what is/is not deterrable. If one is a retributist, then why not simply admit it…?
John Emerson 11.21.07 at 2:14 pm
Dead thread I suppose, but sometimes an execution has the effect of assuring people that everything’s OK again — attaining “social closure”. The villain is dead, the police are on top of things, and there’s no more danger. But this kind of closure can work even when the suspect is innocent (only requiring that the same criminal not commit the same kind of crime again in the same place). And it can also be used by crooked policemen to protect one of their own or a criminal associate.
In the extreme case it can be like a purposeless kind of public drama of human sacrifice, like sacrificing a scapegoat to the rain gods.
I think that a lot of the retributivists are at a pretty primitive level this way. And I also think that “closure” for society or the victims is a treacherous justification for any kind of punishment.
Tom T. 11.21.07 at 3:03 pm
Re: #61. The Supreme Court has previously taken capital punishment off the table for any crime other than murder, except that it left open the possibility that the death penalty could be applied in the case of sexual abuse of a child. There is a case currently percolating in which that issue will be decided.
As to #62, don’t the same observations hold for imprisonment?
John Emerson 11.21.07 at 4:01 pm
They do to a lesser degree, because imprisonment isn’t a dramatic spectacle. Executions, especially public executions, try to make the punishment as vivid as the crime.
Sebastian Holsclaw 11.21.07 at 4:56 pm
“But sebastian, surely you should then be arguing that arrest, convictions and executions should be privatised?”
I presume that is tongue in cheek, but in case it isn’t, no. The insight that the government is awful at certain things does not automatically imply that privatization makes it work either. To my knowledge, the government sucks at faster-than-light travel, but I don’t hold out much hope for the private market on the score either. We are talking about the limits of human knowledge here. Torture might be defensible in certain cases of absolute certainty and high pressure time intensive necessity that don’t actually occur in the world.
That is why getting sucked into a pure ‘torture is always wrong’ fight with people who don’t share that particular moral axiom (and don’t fool yourself, that is a lot of people–including some you otherwise like) is silly when ‘torture as applied ends up being wrong when actual human institutions try to use it’ is much more effective.
Sebastian Holsclaw 11.21.07 at 5:30 pm
“…but sometimes an execution has the effect of assuring people that everything’s OK again—attaining “social closureâ€. The villain is dead, the police are on top of things, and there’s no more danger. But this kind of closure can work even when the suspect is innocent (only requiring that the same criminal not commit the same kind of crime again in the same place).
…
I think that a lot of the retributivists are at a pretty primitive level this way. And I also think that “closure†for society or the victims is a treacherous justification for any kind of punishment.”
You aren’t really grasping the retributivist concept if you frame it that way because you are actually smuggling in retributivist concepts in the discussion on your side. The idea that you could punish an innocent person to get the same ‘effect’ as punishing a guilty person (so long as no one finds out) is a *utilitarian* argument, not a retributive one.
The retributive argument is based on what people deserve as punishment. Your insight that it is wrong to punish an innocent person for the ‘effect’ of the punishment on the rest of the community buys in to that concept. The innocent person does not ‘deserve’ the punishment. It can also serve to temper ‘rehabilitation’ arguments. If for some reason or another you can’t rehabilitate someone from their propensity to make minor legal violations, retributive theory notices that it isn’t just to lock them up until you fix them–they haven’t done enough to deserve that.
See for example C.S. Lewis on “The Humanitarian Theory of Punishment“. (He has a longer essay somewhere, but I can’t find it. Hmm, that essay must have impacted me more than I thought, because it says what I did only much better).
You shouldn’t be put off by the defense of Christianity here, because it applies just as well to government control of other modes of thought that progressives might be more interested in:
“For if crime and disease are to be regarded as the same thing, it follows that any state of mind which our masters choose to call ‘disease’ can be treated as a crime; and compulsorily cured. It will be vain to plead that states of mind which displease government need not always involve moral turpitude and do not therefore always deserve forfeiture of liberty. For our masters will not be using the concepts of Desert and Punishment but those of disease and cure. We know that one school of psychology already regards religion as a neurosis. When this particular neurosis becomes inconvenient to government, what is to hinder government from proceeding to ‘cure’ it? Such ‘cure’ will, of course, be compulsory; but under the Humanitarian theory it will not be called by the shocking name of Persecution. No one will blame us for being Christians, no one will hate us, no one will revile us. The new Nero will approach us with the silky manners of a doctor, and though all will be in fact as compulsory as the tunica molesta or Smithfield or Tyburn, all will go on within the unemotional therapeutic sphere where words like ‘right’ and ‘wrong’ or ‘freedom’ and ‘slavery’ are never heard. And thus when the command is given, every prominent Christian in the land may vanish overnight into Institutions for the Treatment of the Ideologically Unsound, and it will rest with the expert gaolers to say when (if ever) they are to re-emerge. But it will not be persecution. Even if the treatment is painful, even if it is life-long, even if it is fatal, that will be only a regrettable accident; the intention was purely therapeutic. In ordinary medicine there were painful operations and fatal operations; so in this. But because they are ‘treatment’, not punishment, they can be criticized only by fellow-experts and on technical grounds, never by men as men and on grounds of justice.”
geo 11.21.07 at 6:19 pm
#66: When this particular neurosis becomes inconvenient to government, what is to hinder government from proceeding to ‘cure’ it?
This isn’t such a slam-dunk as Lewis and Sebastian seem to think. No government ever claims, or believes, that it must eliminate some practice because it’s inconvenient; always, rather, because it’s dangerous to public safety. And every such claim has to be examined (with extreme skepticism, of course, given everything known about non-Scandinavian governments). What Lewis is really up here to is a defense of metaphysical free will — what’s essential, for him, is that evil is not caused but chosen. Otherwise, how could the all-loving God condemn sinners to eternal torture?
Sebastian Holsclaw 11.21.07 at 6:42 pm
“What Lewis is really up here to is a defense of metaphysical free will—what’s essential, for him, is that evil is not caused but chosen.”
That is caught up in it, but it isn’t the locus of the argument in this case.
My point, before I got too long-winded, was that it is wrong to claim that the idea of punishing an innnocent person for the effect that would have on the community is a *retributivist* concept. It is a *utilitarian* concept, and in most cases their arguments are strongly opposed to each other.
Thom Brooks 11.22.07 at 12:26 pm
#63: yes, I know what the US Supreme Court has said. Instead, my question is that if capital punishment can be shown to have a deterrent effect (and this is a big “if”), then might it be something we should endorse for crimes other than murder.
My suspicion is that if (philosophically-speaking) all other crimes cannot be punished with death, then Dan Kahan is right and deterrence (whether or not proven) is not doing the work and the position (of whether capital punishment is permissible for murderers) actually boils down to one’s stand on whether a murderer deserves the penalty or not.
Comments on this entry are closed.