Opinion
Only Fry the Certainly Guilty
by Alec Rawls
Senior Staff Writer

Illinois Governor George Ryan claims that the death penalty process fails to properly discriminate between solid cases and cases where arbitrariness may be at work or doubts about guilt may remain. His answer? To commute the death sentences of all 167 death row inmates in Illinois, utterly failing to even try to discriminate between solid cases and cases where arbitrariness may be at work or doubts about guilt may remain. In an act of pure hypocrisy, he has released the most certainly guilty of his state's most depraved murderers into the general prison population where they can never face any additional punishment for any further indulgence of their predatory lusts.

This abomination will lead to at least one good thing: no death-penalty-squeamish candidate will ever be elected governor in America again. But there is a much larger good that can come of it, because Ryan's premise was right. The death penalty process does fail to properly discriminate between cases where guilt is certain and cases where doubts about guilt remain. The proper remedy is not to spurn discrimination, as Ryan did, but to discriminate further.

A governor could do this after the fact, but the more fundamental fix is to have juries discriminate further by asking them to hand down multiple verdicts according to multiple standards of guilt.

At present, juries deliberate on a single less than certainty standard of guilt. Estimates are that our current "beyond a reasonable doubt" standard works out to about an 80% certainty standard in practice, meaning that juries on the cusp will convict an innocent person about one time in five. (Not all juries are on the cusp, so the overall rate of false positives, or type II error, should be less than one in five.)

Alternatively, we could ask juries to rule on both an 80% certainty standard and on a full certainty standard, with the severest penalties reserved for those who are found certainly guilty. The result would be more justice for both the guilty and the innocent.

Death for the certainly guilty would become swift sure. So long as guilt is certain, there is no need for an extensive appeals process. The innocent, on the other hand, while they might be found less than certainly guilty, would never suffer the severest punishments, as they do at present.

To achieve such a system, we first have to figure out how to specify a full certainty standard of guilt, but here we immediately run into a puzzle. Our current "beyond a reasonable doubt" standard is in fact a nominal certainty standard. "Reasonable" in this usage is not a reference to the amount of doubt, but the kind. So long as there is ANY doubt that is reasonable in the sense of not being irrational to entertain, acquittal is required. A synonymous early standard was the "no rational doubt" standard.

If this nominal certainty standard devolves into something far less than a certainty standard in practice, how in the world can we construct a true certainty standard? The solution can be found by noting why our current nominal certainty standard devolves. It turns into far less than a certainty standard in practice because a stand-alone certainty standard is a gross violation of the fundamental principle of moral reason, which requires that all value be accounted.

Type I error and type II error are sides of a coin. To acquit whenever there is any possibility at all of convicting an innocent person is to give no weight at all to the possibility of freeing a guilty person. No moral juror will do that. All value must be weighed. Indeed, the more horrific the crime, the more that moral jurors will lower their effective standard of guilt. Errant conviction of the innocent only destroys one life. Errant acquittal of a mass murderer might cost many times that, so that a full accounting of value in such a case could easily impel jurors to apply less than 50% standard of guilt, regardless of jury instructions.

This phenomenon is clearly demonstrated by post mortems on cases where DNA evidence has revealed errant convictions. The worse the crime the more loathe to acquit a jury will be, convicting in the face of veritable truckloads of reasonable doubt.

What jurors need is an alternative way to account the likelihood, as they judge it, that the defendant is guilty. Adding an explicit less than certainty standard to our current nominal certainty standard would afford this outlet, allowing the certainty standard to be deliberated on in unbiased fashion. In effect, the conflict between jury instructions and moral sense would be shifted from the certainty standard to the less than certainty standard. Confronted with a possible serial rapist, jurors might still be loathe to acquit when deliberating on the 80% standard, but the presence of an explicit 80% standard would relieve the drive to err on the no rational doubt standard.

In sum, the way to achieve a true certainty standard for use in a two verdict system is just to have a two verdict system. Through the simple device of adding an explicit less than certainty standard of guilt to our system of criminal trial we can gain the extra discriminatory power necessary to make the death penalty just for both the innocent and the guilty (and their victims).

Under these conditions, application of the death penalty should be greatly expanded.

At present, we compensate for the fact that the death penalty cannot be taken back, if innocence is discovered, by severely limiting the crimes for which the death penalty is available. Once we start limiting capital punishment to the certainly guilty, execution becomes the proper punishment for every instance of unmitigated murder, and perhaps for many other categories of severe unmitigated violence as well.

We all understand the importance of taking the trash out of society. It is just as important to take it out of the prisons if they are to become humane places where less monstrous individuals can possibly reform. Governor Ryan has taken society's most evil predators and given them a "commit mayhem for free" card.

What they need is to go down the garbage disposer. We just need to be certain first, and all that takes is two verdicts instead of one.

This multiple verdicts scheme is not half baked. To see the fully baked potato, visit www.rawls.org.

Page last modified on Wednesday, 01-Mar-2006 23:51:32 MST.