Explore Harvard's Nieman network Nieman Fellowships Nieman Lab Nieman Reports Nieman Storyboard
Shown at a 1998 Northwestern U. Law School conference on wrongful death penalty convictions are 33 former death row inmates who had their convictions overturned. (AP photo)

Time to abandon the death penalty?

ASK THIS | March 83, 2012

Estimates are that 4 to 8 percent of people on death row are innocent. How many wrongful executions are too many? David Dow, author, constitutional lawyer, and founder of the Texas Innocence Network, poses four key death penalty questions and explores the answers to them.


David R. Dow 
 
1.  How frequently are death penalty trials infected by constitutional errors that go uncorrected?
 
2.  What is the cost of a death penalty trial, and how does that cost compare to the cost of life in prison?
 
3.  In non death penalty contexts, people who allege race discrimination can use statistics to demonstrate their case. If defendants facing execution could also use statistical evidence to establish race discrimination, what impact would it have?
 
4.  Given the impossibility of eliminating mistakes altogether, how high an error rate, if any, should be tolerable in the death penalty system?
 
* * * * *
 
1.  How frequently are death penalty trials infected by constitutional errors that go uncorrected?
 
Between 1976, the year the death penalty was reinstated after a four-year moratorium, and 1995, the year before Congress enacted the Anti Terrorism and Effective Death Penalty Act (AEDPA), death row inmates succeeded in their appeals around two-thirds of the time, according to a 2000 Columbia University study
 
This statistic does not mean that death row inmates walked out of prison; it means they got new trials, or new sentencing trials. Many were sentenced to death again, but many others were sentenced to life in prison. One result of this comprehensive study is to provide documentation of how commonly constitutional errors affect capital trials. And that number is, in fact, a conservative assessment of the pervasiveness of error, because in many contexts the error will be overlooked if it is deemed harmless by the appellate court. For example, if the police coerce a confession from a defendant, the defendant’s rights under the Fifth Amendment have been violated, but the Supreme Court has held that a defendant whose confession is coerced is not entitled to a new trial if there is other corroborative evidence of his guilt. So the two-thirds statistic tells us that constitutional error occurred in at least two of every three capital cases.  
 
Thus, the AEDPA, signed by Bill Clinton, has let many more constitutional errors go uncorrected since judges can grant inmates relief only if they conclude that the state court’s decision upholding the conviction and sentence were “objectively unreasonable.” The Supreme Court has ruled that a state court can get the wrong answer – that is, it can be wrong – and yet its decision will still not be objectively unreasonable unless no reasonable jurist could have arrived at that decision. The consequence of this unprecedented limitation on the power of federal courts to grant relief even when they are convinced that an inmate’s rights were violated is that death row inmates now prevail in their appeals well under 10 percent of the time – compared to 67 percent before the law was enacted. 
 
It is a mistake to think that death penalty trials are significantly more error-free than they were 15  years ago. What is different today is that errors are not remedied.  
 
2. What is the cost of a death penalty trial, and how does that cost compare to the cost of life in prison?
 
One of the common misunderstandings is that death penalty cases are expensive because of the several lawyers of appeals. And indeed, a typical death penalty case has three appeals: two in the state court system, and one in the federal court system.  
 
But these appeals are not what makes capital cases so expensive. The overwhelming percentage of costs are incurred before and during the trial. The central reason for this cost is that death penalty trials have two distinctive and expensive aspects. First, because the range of evidence relevant to whether the state should execute someone is so broad, death penalty lawyers spend a year or more prior to the trial investigating every aspect of their client’s background in an effort to learn and acquire evidence that might persuade the jury to spare his life. If one lawyer and one investigator spend 20 hours a week for 50 weeks locating and interviewing witnesses, and gathering and analyzing records (including school records, hospital records, employment records, military service records, and records from prior incarcerations, including juvenile incarcerations), they will aggregate 1,000 hours in trial preparation. At any hourly rate of $100, the case has already cost $100,000 before any experts have been paid, before the prosecution’s costs have been tallied, and before the trial has even started. Once the trial does start, the lawyers will spend several weeks, sometimes up to two months, picking the jury. In contrast, in a non death penalty case, a jury is usually seated in several hours. If it takes three weeks to pick the jury, the cost in defense legal fees alone just for picking the jury exceeds $37,000 (15 days X 10 hours / day X $250). Again, the trial has not even started, and we have not accounted for costs born by the state.  
 
All told, death penalty cases run from a low of $1.5 million, to $4 million or more, and at least three-fourths of those costs are trial-related–meaning no amount of streamlined appeals will reduce the figure substantially. 
 
In contrast, it costs less than $20,000 a year to confine a prisoner in a maximum security prison. If a convicted murderer has a life expectancy of 40 years inside prison (life expectancies are predictably lower than they are outside of prison), it will cost $800,000 to keep a prisoner locked up for the rest of his life – or somewhere between half and 20 percent as much as it will cost to execute him. 
 
3.  In non death penalty contexts, people who allege race discrimination can use statistics to demonstrate their case. If defendants facing execution could also use statistical evidence to establish race discrimination, what impact would it have?
 
In the mid 1980s, the Supreme Court ruled that defendants facing the death penalty cannot use statistics to demonstrate that the system discriminates on the basis of race. The case involved a comprehensive study in Georgia led by the late University of Iowa law professor David C. Baldus.  
 
The study showed that black defendants are nearly twice as likely as white defendants to be sentenced to death, and that murderers whose victim is white are nearly four times as likely to be sentenced to death as murderers whose victim is nonwhite.  
 
Since that time, the study has been replicated in more than a dozen jurisdictions, and the results have been more or less the same everywhere. In fact, since the death penalty was reinstated in 1976, around 80 percent of the people who have been executed in the U.S. were executed for killing a white person, even though blacks and whites are victims of homicide in roughly equal numbers. 
 
In Texas, where there have been more nearly 470 executions, only three have involved a white murderer whose victim was black. North Carolina recently enacted a measure that permits death row inmates to challenge the constitutionality of their sentence by relying on data like these. But under federal law, such challenges remain impermissible.  
 
4.  Given the impossibility of eliminating mistakes altogether, how high an error rate should be tolerable in the death penalty system?
 
One of the most important psychological developments over the past generation to have occurred in the criminal justice context has been the undermining of our certainty that the system is largely error-free. There have been more than 289 DNA post-conviction exonerations since the mid 1980s, and where DNA is involved, then we can be certain – to the extent we can ever be certain about anything – that an innocent man was sent to prison. In nearly three-fourths of the DNA cases, there was eyewitness testimony pointing to the accused. In nearly a quarter of the cases, there was a confession.   
 
Although death penalty cases typically do not involve DNA, they do involve the two types of evidence – eyewitness testimony and confessions – that the DNA cases have revealed to be remarkably unreliable. Estimates are that somewhere between 4 and 8 percent of the inmates on death row are innocent. No one knows how many innocent people, if any at all, have been executed. But in view of these data, it would take an uncommon degree of nonchalance to think no one has.  
 
Many people would refuse to get on an airplane that had a five percent chance of crashing. At what point does the number of crashes in the death penalty domain mean that it is a system we should abandon?
 


Dow is incorrect on point #!
Posted by Dudley Sharp
03/83/2012, 03:25 PM

37% of death penalty cases have been overturned on appeal, not the 67% reported within the Columbia U study, which is rebutted, here:

"A Broken Study: A Review of 'A Broken System' "
http://prodpinnc.blogspot.com/2009/10/broken-study ...

The 37% is from here:

http://bjs.ojp.usdoj.gov/content/pub/pdf/cp10st.pd ...



Posted by Shane E. Mahoney
03/83/2012, 05:45 PM

Illustrations and data that point to justice poorly or improperly administered are not, in themselves, arguments for abandonment of the related penalty. In fact, even were it shown that occasional errors are made in imposing the death penalty, that would not be reason to abandon it — unless no human undertaking should be permitted in which errors are made that result in death, like flying airplanes, driving cars, drinking alcohol, or engaging in large construction projects. On the other hand, what can (and should) be done with regard to the death penalty is to narrow the permissible application of it and the evidence required to impose it to a level of near certainty. To abandon the death penalty creates a peculiar conundrum, Consider a person who is in the process of being assaulted (perhaps murdered). We know that this individual, quite naturally, would kill her assailant so as to preserve her own life were she capable of doing so — and the assailant would be have thereby received the death penalty. On the other hand, if the assailant is successful with the murder, he lives on under an abandoned death penalty — because he was successful. Yet, as individuals, we have foregone our own personal right to administer justice and given it to the state in the confidence that the state would handle it. Were the murder victim to come back to life, she might say, "I would have killed him had I been able, but now that he has been successful, you have let him live — perhaps to murder a prison guard or another prisoner or to require the extended expense of life in prison; all because I failed to kill him. Had I killed him, you would have been fine with it!"


Reality on point #2
Posted by Dudley Sharp
03/85/2012, 07:44 AM

There is little doubt but that responsible managament of death penalty cases can result in a death penalty protocol that is about the same cost as life without parole case, if not less expensive.

Virginia has executed 75% of those so sentneced and has executed thm within 7.1 years, on average.

It is important to fact check the death penalty cost studies, as it is to fact check all studies.

Two examples, of many:

1) "Duke (North Carolina) Death Penalty Cost Study: Let's be honest"
http://prodpinnc.blogspot.com/2009/06/duke-north-c ...

2) Texas cost study - I have told the Dallas Morning News, for many years, to stop using their totally inaccurate cost review. They still use it.

They found that it costs $2.3 million per average death penalty case (for 5 cases), more than 3 times more expensive than a $750,000 life sentence. (C. Hoppe, "Executions Cost Texas Millions," The Dallas Morning News, March 8, 1992, 1A)

The death penalty costs are for pre trial, trial and appeals and incarceration. Yet, the life cost is only for confinement for life. Big problem.

In addition, an academic review, by a neutral academic, found that the verifiable costs in the DMN article actually found the death penalty was cheaper.

p154-156 http://books.google.com/books?id=IQJtCjhdGeUC&pg=P ...




The Dow example, #2
Posted by Dudley Sharp
03/85/2012, 07:57 AM

Sadly, Prof. Dow follows the example of many, in #2, as he writes:

"All told, death penalty cases run from a low of $1.5 million, to $4 million or more, and at least three-fourths of those costs are trial-related–meaning no amount of streamlined appeals will reduce the figure substantially.

In contrast, it costs less than $20,000 a year to confine a prisoner in a maximum security prison. If a convicted murderer has a life expectancy of 40 years inside prison (life expectancies are predictably lower than they are outside of prison), it will cost $800,000 to keep a prisoner locked up for the rest of his life – or somewhere between half and 20 percent as much as it will cost to execute him."

He compasres

1) all costs of a death penalty "case", chich includes all costs from pre trial to execution, inclusive of trial, appeals and incatrceration

to

2) Only the incarceration costs of a life sentence.

This type of intentional fudging of the numbers must stop.

Dow knows exactly what he is doing.


Race and the death penalty #3
Posted by Dudley Sharp
03/85/2012, 08:35 AM

Methodologist Joseph Katz destroyed the Baldus study. Katz' refutation of Baldus was part of the states case in McClesky.

If Prof. Dow hasn't read it, I can provide it to him.

SCOTUS also made an error within their decision, a mistake which virtually all law schools and their professors have not recognized and which they, sadly, pass along to their students.

Prof. Dow repeats it with his comment:

"murderers whose victim is white are nearly four times as likely to be sentenced to death as murderers whose victim is nonwhite."

It is not by "four times", but by an odds mulitiplier of 4 times.

There is a huge difference, as partially revealed in the two articles, in the next post.

"Four times" is a difference of 300%. An odds multiplier of 4 times can mean a difference as low as 4%, depending upon the data.

Prof. Dow may want to read Federal District Court Judge Forrester's decision in the case.

Or my rveiew from 1997:

"Based on a study conducted by Profs. Baldus, Woodward and Pulaski, McCleskey argued that the death penalty was racist. In August, 1983 Federal District Court Judge J. Owen Forester found that the study's conclusions of racial bias were without merit. In 1985, the 11th Circuit Court of Appeals, by a 9-3 vote, stated "Viewed broadly, it would seem that the statistical evidence presented here...confirms rather than condemns the ( death penalty) system." In April 1987, the Supreme Court (5-4) stated that the referenced study did not establish that capital punishment discriminates against black defendants or killers of white victims. "At most, the Baldus study indicates a statistical discrepancy that appears to correlate with race. Apparent disparities in sentencing are an inevitable part of our criminal justice system. The discrepancy indicated by the Baldus study is ‘a far cry from any major systemic defects". "McCleskey offers no evidence...that would support an inference that racial considerations played a part in his sentence". "...the Baldus study is clearly insufficient to support an inference that any of the decision-makers in McCleskey’s case acted with discriminatory purpose." "Even Professor Baldus does not contend that his statistics prove that race enters into any capital sentencing decisions or that race was a factor in McCleskey’s particular case." "


#4 wild speculation? Innocents at risk.
Posted by Dudley Sharp
03/85/2012, 09:04 AM

Professor Dow writes:

"Estimates are that somewhere between 4 and 8 percent of the inmates on death row are innocent."

Interesting.

After 38 years of death penalty cases and appeals since Furman, and about 8200 death sentences, we may have solid evidence, depending upon review, of from 25-40 actual innocents who were sentenced to death and still on death row when their innocence was discovered and they were, then, released (1).

That's about a 0.4% error rate of actual innocents sentenced to death.

What projections and speculations caused Dow to increase that percentage by 10 to 20 fold and were they justified?

Certainly, the percentage is higher than 0.4%, which reflects those so identified, between 1973-2012, but not those actual innocents not yet identified.

However, based upon all of the many improvements in due process protections in death penalty cases, over the last 15 years, it is most likely, that over the next 38 years, from 2012 through 2050, that the percentage of actual innocents identified on death row, will actually be lower than 0.4%.

So, how credible is the 4-8%, used by Dow?

I say it is not credible.

Everyone is concerned about actual innocents being arrested, convicted, incarcerated and/or executed. No one wants it to happen.

One of the constant problems in this debate is death penalty opponents simply contriving the numbers of actual innocents sentenced to death row (1) or executed (2).

Let's make an effort for a more reasoned, informed and honest discussion.

1)The 130 (now 140) death row "innocents" scam
http://homicidesurvivors.com/2009/03/04/fact-check ...

2) "The Innocent Executed: Deception & Death Penalty Opponents"
http://homicidesurvivors.com/2009/10/08/the-innoce ...



#3 Race: "times" vs "odds multipliers"
Posted by Dudley Sharp
03/85/2012, 09:09 AM

Baldus' database and work in McCleskey was quite poor.

Read Federal District Court Judge Forrester's rejection of Baldus' database for McCleskey.

A more thorough review is provided by Joseph Katz, who did the methodological review of the Baldus database, which was rife with errors and problems. I have it, if you care to research.

In addition, SCOTUS totally misunderstood the math involved. They ignorantly wrote: "defendants charged with killing white victims were 4.3 times as likely to receive a death sentence as defendants charged with killing blacks."

Totally inaccurate. It was by odds of 4.3 times, or an odds multiplier of 4.3, which can mean a variables as low as 2-4%, as opposed to the 330% difference represented by 4.3 times. SCOTUS blew it big time on this.

These two articles, below, give a good explanation of a core problem with Baldus, in the McCleskey case and another of his reviews.

A) "The Math Behind Race, Crime and Sentencing Statistics"
By John Allen Paulos, Los Angeles Times, July 12, 1998
http://articles.latimes.com/1998/jul/12/opinion/op ...

B) See “The Odds of Execution” within “How numbers are tricking you”, by Arnold Barnett, MIT Technology Review October, 1994
http://reocities.com/CapitolHill/4834/barnett.htm ...


#4 Innocents More At Risk Without Death Penalty
Posted by Dudley Sharp
03/85/2012, 09:29 AM

Innocents More At Risk Without Death Penalty
Dudley Sharp, contact info below

Are death penalty opponents really concerned about innocents at risk? Of course.

However, as innocents are more at risk without the death penalty, it is appropriate to challenge their understanding of that easily accessible fact.

1) There is no known actual innocent executed in the US, at least since the 1930's. Possibly, 0.4% of all those sentenced to death since 1973 may have been actually innocent. All were released. An incredible record of accuracy - 99.6% accuracy in proper actual guilt convictions, with the 0.4% released (1).

2) Anti death penalty folks say taking the death penalty away eliminates actual innocents dying prior to their exoneration.

Nonsense. While we have executed about 1250 murderers in the US (1973-2012) about 5,000 inmates per year die while in custody (2), or about 200,000 total, during that time (2). We cannot bring back any of those that may have been innocent, either.

It is more likely that an actual innocent will die in custody than it is that an actual innocent will be executed.

3) Murderers that we have allowed to murder again, recidivist murderers, MIGHT number around 28,000 since 1973 , based upon existing studies (3).

However, I think it likely that the numbers are closer to 14,000, based upon my review in footnote 3.

4) 28 studies, beginning in 2000, find for a deterrent effect, ranging from 1-28 innocent lives spared per execution, or totals from 1,254 - 35,112 innocent lives saved (4).

Based upon an estimated 700,000 murders in the US (1973-2011), that represents a range from 0.18% to 5% of potential murderers who were deterred from committing murder - a huge savings in innocent lives spared, but, statistically, a very small reduction in murder rates.

5) The death penalty spares more innocent lives than does a life sentence, in two additional ways, as well: Enhanced due process and enhanced incapacitation. No one questions that the death penalty has greater due process protections than a life sentence, thereby the death penalty protects actual innocents to a higher degree than any lesser sanction. Neither does anyone question that living murderers harm and murder, again and that executed ones do not.

contd


contd #4 Innocents More At Risk W/O Death Penalty
Posted by Dudley Sharp
03/85/2012, 09:35 AM

contd

6) REALITY: ARE DEATH PENALTY OPPONENTS REALLY CONCERNED ABOUT INNOCENT LIVES AT RISK?

Of course, but . . .

The reality is that we can spare murderers lives and thus sacrifice more innocents or we can spare more innocent lives by executing more murderers.

Look at Ernest van den Haag's interview of well known anti death penalty activists.

He asked them, if it was proven that 100 innocent lives were spared per execution, via deterrence, would you still oppose the death penalty. All said yes. (4)

Based upon our 1250 executions (1973-2012), those anti death penalty folks would prefer sparing the lives of 1250 murderers over saving the lives of 125,000 innocents. Think about that

Well known anti death penalty scholars "(Charles) Black and (Hugo Adam ) Bedau said they would favor abolishing the death penalty even if they knew that doing so would increase the homicide rate by 1,000 percent." (4).

For them, 6.3 million additional murders of innocent people (from 1973-2011) is preferable over executing 1250 known murderers. Astounding.

For some very well known leaders of the anti death penalty movement, their motivation is not protecting innocents, but protecting the lives of all murderers, no matter the cost in innocent lives.

That's their moral choice. They, and all death penalty opponents, need to face that.

NOTE: I am making two points with this information. If anti death penalty folks were concerned about innocents, their efforts would be to fix the huge problems in the criminal justice system which really sacrifice innocents lives, as opposed to trying to end the death penalty, a penalty which assists in sparing additional innocent lives.

Most importantly, I think many of those opposed to capital punishment would change their minds if they knew that sparing murderers sacrificed more innocent lives.

contd


contd #4 Innocents More At Risk w/o Death Penalty
Posted by Dudley Sharp
03/85/2012, 09:39 AM

contd

FOOTNOTES

(1) a. The 130 (now 140) death row "innocents" scam
http://homicidesurvivors.com/2009/03/04/fact-check ...

b. "The Innocent Executed: Deception & Death Penalty Opponents"
http://homicidesurvivors.com/2009/10/08/the-innoce ...

c. "The innocence tactic: Unreliable studies and disinformation", reports By United States Congress, Senate, 107th Congress, 2d Session, Calender no 731, Report 107-315. The Innocence Protection Act of 2002, (iv) The innocence tactic: Unreliable studies and disinformation, p 65-69, http://alturl.com/6j7oc ...

d. "The Innocent and the Shammed", Joshua Marquis, Published in New York Times, 1/26/2006
http://coastda.blogspot.com/2006/01/innocent-and-s ...

(2) Deaths in Custody Statistical Tables, Bureau of Justice Statistics, SEE 7 links on right side of page. http://bjs.ojp.usdoj.gov/content/dcrp/dictabs.cfm ...

(3) "Recidivism of Prisoners Released", Bureau of Justice Statistics, http://bjs.ojp.usdoj.gov/index.cfm?ty=pbse&sid=44 ...

See both studies, of 1983 and 1994 data

6.6% recidivism rate, 1983 study of 1997 data
1.2% recidivism rate, 2002 study of 1994 data

The combined recidivism rate for the two studies is 3.9%. These rates are based upon re-arrest, not re-conviction.

I suspect the real recidivism rate, looking at 1973-2011, will be closer to half that, at 2%.

Why the recidivism rates should be higher: Both of these studies only looked at recidivism for 3 years after release. We are concerned with recidivism for 38 years and less, years 1973-2011, as the modern era of new death penalty statutes began in 1973. Recidivism rates will be higher if released prisoners were tracked 4 years and beyond, as opposed to only 3 years. In addition, rates from 1973-1983, will likely be about the same as the 6.6%, if not higher, because we had barely started the period of longer sentences and increased incarcerations rates.

Why the recidivism rates should be lower: The dramatic reduction in recidivism between the two studies reflects a dramatic increase in sentencing terms and incarceration rates, which amounted to lower rates of early release and thus recidivism.

EXACT NUMBERS: 8.4% of those on death row had murdered, at least one person, prior to committing additional murder or murders which put them on death row -- an estimated 600-1000 additional innocents murdered.


Cite?
Posted by Greg
03/85/2012, 05:24 PM

I'm no fan of the death penalty, but I'd really like a citations for that 4-8% number. Never heard that it was that high before. Can you give references please?



Posted by SDL
05/140/2012, 09:59 AM

Dudley...

Please....

Kill yourself.


What a person do or don't, in number,0 or 1 is the matter
Posted by Kaori
05/142/2012, 11:18 AM

SDL...

Better not to be dirty,It's "counterproductive"(borrowed from M.P)

Dudley,I hope you make your research useful for people,if not yet,even I'm totally other side.

Sorry about big mouth and my English skill.




The NiemanWatchdog.org website is no longer being updated. Watchdog stories have a new home in Nieman Reports.