January 13, 2005
Confidence Intervals and Reasonable Doubt
January 13, 2005 12:18 AM
Linking to Dana Mulhauser's analysis of Booker, Will Baude wonders whether she is correct in saying that the beyond a reasonable doubt standard "is usually taken to mean that the jury must be at least 95 percent certain of its decision."
My second Criminal Law class is in nine hours, so I'm guessing, but I think this figure may be tied to statistics. I took statistics for economics, but that was in another century and besides the grade was bad, so take my explanation with a grain of salt. Confidence intervals are usually calculated so that we are 95% certain of our results; that is, the margin of error is roughly two standard deviations. And if it's good enough for political polling and the Census, I guess it's good enough for convictions.
The CrimLaw textbook* says that when a trial judge tried to explain degrees of proof on a scale of zero to ten, and said beyond a reasonable doubt was seven and a half, he was reversed by the Nevada Supreme Court (McCullough v. State, 1983), which said, "The concept of reasonable doubt is inherently qualitative. Any attempt to quantify it may impermissibly lower the prosecution's burden of proof, and is likely to confuse rather than clarify." U.S. v. Walton (2000) says the same: the trial judge had refused to define reasonable doubt even when the jury requested that he do so, and Fourth Circuit affirmed with the same "confuse rather than clarify" rhetoric.
* So how worthless are the sentencing sections of our textbooks now?
Psychologists who study jury decision-making usually place reasonable doubt at 85% certainty. I'll try to dig up something from the psych lit when I have the time and energy.
I trust that Armen's study will focus not just upon the jury's decision as a whole but also as to the individual jurors. The jury's decision may be a compromise. So if the 85% factor applied to the jury, then perhaps some jurors were at 100% and some perhaps under 50% (especially with a 12 person jury).
"Proof sufficient to establish in the mind of the trier of fact the truth to a moral certainty of the facts alleged."
I've heard 95% bandied about, but I think it's hard to equate numbers with subjective certainty.
I wrote a paper about jury confusion re: reasonable doubt when I was in law school. I recall seeing a study mentioned where the respondents thought that "clear and convincing" was a higher standard than BRD. It's a total mess. And any criminal lawyer would tell you that if jurors thought they had to be 95% sure to convict someone, we'd have a lot more jury trials.
Admittedly, it was a somewhat crude figure; I've seen it anywhere from 90% to 99%. I wanted a number, though, for the rhetorical effectiveness of having something to hold up against 51%. I figured that whether it was 91% or 97% didn't much matter, given that I wasn't using it for any sophisticated calculations.
(In my opinion, jurors actually use a much lower standard for conviction, but that's from observation working for a PD, and not from any sort of study. But what I wanted for purposes of the article was the difference in judicial standards, not empirical ones.)
As promised, Koch and Devine, "Effects of Reasonable Doubt Definition and Inclusion of a Lesser Charge on Jury Verdicts" in Law and Human Behavior, V. 23, No. 6 (1999) [Note: This comment program is not letting me link to the url because it contains the words online and com, so I'll have to improvise. The full addy is http://ipsapp009.kluweronline (insert a dot here) com/IPS/content/search/ext/x/J/4968/I/25/A/3/type/PDF/article.htm, but you probably need a proxy server authentication from an academic institution to have access, if not I'll be more than happy to e-mail a copy in pdf format] offers a good overview of the concept in its intro. I can't dig up the Wrightsman articles he references electronically, but Wrightsman's book might be too general.
Sweet lord that was difficult...it wouldn't even let me do the -dot-com thing.
What's interesting to me is that most of us will likely agree that beyond a reasonable doubt is a relatively high percentage - at the very least above 60%. Yet to pass a statute defining a crime, a legislature need only have a 51% vote (plus the exec/gov needs to sign).
So, have you committed murder "beyond a reasonable doubt" when there's 100% certainty that you aided a terminally-ill patient who was in excruciating pain committ suicide, but when the statute that made it a crime was passed by only the slimmest of majorities? We know beyond a reasonable doubt that the act was committed, but what about the classification of the act as a crime? Clearly there is a reasonable doubt that what you've done is a crime per se.
Yeah, at the risk of stating the obvious, DC, the argument doesn't work at all. It isn't the case that there is "reasonable doubt" about whether "what you've done is a crime". It is a crime-- even the losers in a vote usually at least admit that they have in fact lost.
There is "reasonable doubt" about whether it ought to be a crime, but that plays only into the nullification question.