Fixing the death penalty system

In this post, I will set forth my views on the appropriateness of capital punishment, the problems it faces, and what should be done.

First of all, there is strong evidence that the death penalty does, in fact, deter murder. A blogger who writes under the name of "Engram" has marshalled this evidence. I have nothing to add to his excellent presentation.

To me, this evidence is reason enough to keep the death penalty. However, our death penalty, as it stands today, has many problems, and these need to be addressed.

It is indisputable that innocent people have been sentenced to death on several occasions since the Supreme court allowed death sentences to resume in 1976. One example is Kirk Bloodsworth. Mr. Bloodsworth was sentenced to death based on eyewitness identification, and was later exonerated by DNA. As a result of this case and others like it, we don't sentence people to death based on eyewitness identification anymore. This is surely a step in the right direction.

The anti-death-penalty people have posted a list of exonerations here. Exonerations come in various levels of certainty. In some cases, such as Mr. Bloodsworth's, it is clear that the exonerated person is completely innocent. In others, it is not clear if the exonerated person is innocent or not. In still others, it looks like the exonerated person is probably guilty -- but not beyond a reasonable doubt, based on admissible evidence. The pro-death penalty people have a list of cases they say are in this category here. In this debate, the claims of both sides should be viewed with caution.

One topic of debate is whether any innocent person has been executed in this country since the Supreme Court decision Gregg v. Georgia (1976) allowed executions to resume. My personal belief is that this has happened. The reason is that several people have been exonerated because of the development of DNA matching technology. And if DNA matching technology had not been developed, many of these people would have had no way to exonerate themselves. I believe the rate of false convictions, pre DNA, should have been the same in cases where DNA evidence was later found to be available as in cases where it was not. Since several of the convictions where DNA was later available proved to be wrongful, it is likely that the same is true in cases without DNA evidence. And, with no DNA, the defendant may have had no way to exonerate themselves later.


There have also been many non-DNA exonerations. But this does not affect the validity of the above argument. The point is that many people who were exonerated by DNA would not have been able to do so without it. Therefore, there are almost certainly innocent people who were convicted in non-DNA cases, and the defendants were simply stuck after that.

To sum up the situation, we have the following:

(1) The death penalty almost certainly does deter murder.
(2) Innocent people have almost certainly been executed.

Note the symmetry of the above. There is evidence both in favor of and against the death penalty. In both cases, a close examination of this evidence shows it is almost certainly correct. Unfortunately, advocates on each side of this debate tend to either ignore or try to negate the opposing point.

In addition, there is the problem of death penalty appeals. These tend to be enormously expensive and time-consuming. It is not unusual for someone to spend 20 years on death row before they are finally executed. It is common for people to leave death row by dying of natural causes. These delays consume enormous resources. They are painful for the families of the victims. And they make a mockery of the death penalty itself. The Department of Justice keeps statistics on the reasons people leave death row. Here are the statistics for 2005 -- the most recent year for which they are available:

Under sentence of death at start of year: 3320
Executed: 60 (1.8%)
Suicide: 3 (0.1%)
Died of other causes: 22 (0.7%)
Left death row for other reasons (i.e, sentence commuted, guilty verdict overturned, death sentence overturned, exonerated): 109 (3.3%)
Sentenced to death during 2005: 128
Under sentence of death at end of year: 3254.

Notice that very few of the people under sentence of death during 2005 were actually executed during that year.

Looking at this another way, here is what happened to people sentenced to death during the three-year period from 1984-86, i.e, approximately 20 years before the 12/31/05 cutoff date for inclusion in the report:

Sentened to death 1984-6: 846*
Left death row:
Executed: 177 (21%)
Other death: 48 (6%)
Other or unknown reason (mostly legal reasons): 401 (47%)
Still under sentence of death as of 12/31/05: 220 (26%)

* Those who were sentenced to death during this time, had their sentences or convictions overturned, and were later resentenced to death are not included.

It is notable that only approximately 21% of those sentenced to death approximately 20 years earlier had been executed during that time. And the ones whose sentence was still under some kind of appeal outnumbered those who had been executed!

The main reason for these endless appeals is that our system does make mistakes. Innocent people do get sentenced to death. The appeals process is society's way to try to make sure those mistakes are rectified before the defendant is executed.

I would like to propose a new way to handle the death penalty. I believe that overall, it would increase the number of death sentences, thereby decreasing the murder rate. I believe it would also make the death penalty less error-prone. It is this:

The primary way to determine in a murder case whether or not the defendant should be sentenced to death should be whether or not there is overwhelming evidence of his or her guilt. Not merely proof beyond a reasonable doubt. But evidence so strong that it admits no other explanation than the defendant's guilt.

For instance, a DNA match alone would not be overwhelming evidence. The DNA analyst could be lying. Or the DNA could come from contamination. But if, for example, you have DNA, and the defendant's fingerprints are all over the gun, and the defendant kept changing his story and later confessed, that would be overwhelming evidence. The above would not, however, be overwhelming evidence if one person testified to all of it. Because that witness could be lying. Ideally, the witnesses should come from seperate organizations -- so that corruption in one organization could not allow someone to be framed.

Another example of overwhelming evidence would be if you have clear surveillance video of the defendant committing the murder. Technology in this area has advanced, but video remains difficult to fake.

Testimony from someone who is receiving a benefit from the criminal justice system (i.e, a reduced sentence for another crime) should never be considered as a part of overwhelming evidence. Because such witnesses are inherently unreliable. The same is true of eyewitness identification by anyone who did not know the defendant prior to the crime.

Additional safeguards should be put in place. In particular, both police work and analysis of forensic evidence should itself be videotaped. And these videotapes should be made available to the defense, or, even better, posted on the web. In this era of inexpensive webcams, and 470GB of CDs retailing for $25, it should be possible to do this without spending a lot of money. Furthermore, any technical work (i.e, fingerprint matches, hair or fiber matches, and DNA matches) that is going to be used in a murder trial should be subject to random, blind audits. That is, the person doing the work does not know if their work is going to be audited, and if so, by whom. The auditing process can be computerized and randomized, so the auditor could be any specialist in the same field in the country.

With all of the above safeguards, and the requirement of overwhelming evidence to sentence a defendant to death, I believe there would be very few wrongful death sentences. The reason can be found on the innocence project's website. There, they list the causes of wrongful convictions

Eyewitness Misidentification
Unreliable or Limited Science
False Confessions
Forensic Science Fraud or Misconduct
Government Misconduct
Informants or Snitches
Bad Lawyering

All of the above causes, except for bad lawyering, would be greatly reduced by the above safeguards. For instance, with police work and lab work being videotaped, government misconduct would become much riskier and more difficult, as would forensic fraud, and so on. And with the additional requirement of overwhelming evidence, it just wouldn't happen often.

But wrongful death sentences would still happen occasionally, due to bad defense lawyering, for instance. And so there should be one final safeguard. Whenever everyone is sentenced to death, there should be one automatic appeal, with a different lawyer if the defendant so chooses. And in this appeal, the lawyer should argue to a panel of three judges why the evidence is not overwhelming. They would be allowed to introduce any new evidence, without limitation. And the judges would have the power to overturn the death sentence. BUT they would be limited to some fixed percentage of their cases where they are allowed to do that. (20%, for instance).

If this appeal fails, the defendant should be allowed to appeal to the United States Supreme Court (skipping all the intermediate steps which exist today). If the Court either rules against the defendant, or decides not to grant cert, the death sentence should be carried out within a few weeks. Appeals that the manner of death would be too painful should not be allowed. This could be accomplished either by the Supreme Court ruling that, under the above system, the death penalty is no longer "unusual", or by a Constitutional amendment if for some reason the Court is unwilling to make that ruling.

I believe that in cases of premeditated murder, where there is overwhelming evidence that the defendant wanted to kill someone and did so (even if the person they killed was not the person they wanted to kill), the death sentence should be the general rule. There could be exceptions in cases where there are extenuating circumstances, such as if the victim was abusive towards the defendant (but not so abusive that the killing could be considered self defense). Other exceptions would be if the defendant was less than 18 years old, or was severely retarded. But if none of the above applies, the jury should routinely be asked to sentence the defendant to death. (And usually, they do so when they are asked). And with the above relatively brief appeals process, the sentence would typically be either overturned or carried out within a year or two.

Under this system, some people who are currently sentenced to death would not be. Scott Peterson, for instance, would likely escape the death sentence, as the case against him, while strong, was not overwhelming. John Couey and Joseph Smith would both likely have been sentenced to death, however, and Smith might well already have been executed. Additionally, may people who today are not under sentence of death would likely face this sentence.

Lastly, government agents who engage in fraud need to be routinely prosecuted. (Currently, they frequently are not.) And the charges should be more serious than they are today. If they try, but fail, to fraudulently give someone the death penalty, the charge should be attempted murder. And if, despite all the safeguards, someone is executed because of their fraud, the charge should be first degree murder.

I believe if the country adopted this system, the following would happen:

* A larger portion of murders would lead to death sentences.
* These death sentences would be less subject to error than they are today.
* Death sentences would be carried out relatively soon after the trial, making them a real deterrent. And as a result,
* The murder rate would drop.