Timothy P. O'Neill

Chicago Daily Law Bulletin
Volume: 153 Issue: 9

January 12, 2007

Type of Evidence May Outweigh the Strength

Comedian Bob Newhart was born in Oak Park, raised in Austin, and graduated from St. Ignatius College Prep and Loyola. He then attended Loyola Law School. He never finished. He either ''left school'' or ''flunked out.'' In characteristic Newhart style, he says ''I have my explanation for what happened, and the school has theirs.'' Bob Newhart, I Shouldn't Even Be Doing This! (Hyperion, 2006), p. 34.

In discussing the move from law to stand-up comedy, he offers this final thought: ''I always figured that I would have pursued trial law because trial lawyers are actors. They stand in front of judges and juries and entertain them with borderline preposterous stories — not unlike those told by stand-ups, come to think of it.'' (35).

One view of the criminal jury trial visualizes the prosecutor's burden as presenting sufficient evidence to move a ''needle'' on the ''evidence meter'' from zero to ''proof beyond a reasonable doubt.'' Yet a growing body of literature conceives of the criminal trial as a battle between the prosecution and the defense to convince the jury, ala Newhart, to buy its own particular ''story.''

Under this theory, it is not enough for the defense to merely poke holes and contend that the prosecution has not moved the needle into the ''beyond a reasonable doubt'' range. Rather, the defense must offer its own ''story'' of why the jury should be comfortable in its decision to acquit. See, e.g., Nancy Pennington & Reid Hastie, A Cognitive Theory of Juror Decision Making: The Story Model, 13 Cardozo Law Review 519 (1991); Richard Lempert, Telling Tales in Court: Trial Procedures and the Story Model, 13 Cardozo Law Review 559 (1991).

A New Zealand law professor, Kevin Jon Heller, has recently made an important contribution to this literature. In addressing the problems that juries have in processing circumstantial evidence, Heller has written an article that is a must-read for anyone who tries criminal jury cases. Kevin Jon Heller, The Cognitive Psychology of Circumstantial Evidence, 105 Michigan Law Review 241 (2006).

First, Heller lays the groundwork for his argument. ''Direct evidence'' is evidence that ''proves a fact without an inference or presumption and which in itself, if true, establishes that fact.'' (248). The best examples are confessions and eyewitness identifications. ''Circumstantial evidence,'' on the other hand, is ''evidence from which the fact-finder can infer whether the facts in dispute existed or did not exist.'' (250). This would include all forensic evidence, such as DNA and fingerprints, as well as nonforensic evidence that does not per se prove the defendant's guilt.

Both kinds of evidence are, of course, admissible at a criminal trial. Yet statistics indicate that direct evidence is much less reliable than circumstantial evidence. For example, some studies have shown that eyewitness identifications are mistaken more than 58 percent of the time, while less than 1 percent of DNA matches are erroneous. (244). And the famous Bedau and Radelet study on wrongful convictions in capital cases showed that 68 percent resulted from problems with direct evidence, while only 9 percent resulted from circumstantial evidence problems. (244).

Despite this, Heller notes that studies have consistently shown that juries significantly undervalue circumstantial evidence and just as significantly overvalue direct evidence. His article offers an explanation why this is true.

The traditional rationalization is that juries simply do not understand the actual probative value of circumstantial evidence. Yet a famous study by Gary Wells disproved this. Wells created a hypothetical civil case concerning the possible liability of a bus company. He presented the basic facts to two groups, but with one crucial difference: One group was presented with direct evidence showing that there was an 80 percent chance of liability, while the other group was presented with circumstantial evidence that there was an 80 percent chance of liability. Wells found that the group presented with the direct evidence was much more likely than the circumstantial evidence group to find liability.

Did this mean that the circumstantial evidence group did not understand the probative value of the evidence? Absolutely not. Interestingly, the mean subjective probability of guilt found by the circumstantial evidence group was remarkably similar to that found by the direct evidence group — 70 percent. Despite this, the direct evidence group was four to nine times more likely to find the bus company liable. Wells concluded that there is a fundamental psychological distinction between the ways jurors consider direct and circumstantial evidence that has nothing to do with mere probabilities.

Psychologists call this the ''Wells Effect,'' i.e., the fact that jurors are likely to acquit in a circumstantial evidence case even when they know that there is sufficient evidence to convict. Wells concluded, ''Psychologically, there seems to be a difference between saying that there is an 80 percent chance that something is true and saying that something is true based on [direct] evidence that is 80 percent reliable.'' (258). That is, juries are more likely to acquit in the former category of circumstantial evidence cases.

But why? Keith Niedermeier tested this by giving two groups different kinds of circumstantial evidence about a bus accident that killed a dog. The first group was told that the tire tracks found on plaintiff's dog clearly matched 80 percent of the defendant's buses and 20 percent of another company's buses. The second group was told that the tire tracks partially matched defendant's buses and partially matched the other company's buses. But this ''partial match'' circumstantial evidence still resulted in an 80 percent probability that one of defendant's buses was involved, and a 20 percent probability that the other company was responsible.

The result? Despite the equal probability, the ''clear match'' group was twice as likely to find for the defendant. Why? Niedermeier concluded that it was easier for those in the ''clear match'' group to actually picture a bus from the other company being involved; after all, there was a clear, definite match with at least one bus from the other company. He concluded that Wells was correct; a jury is more likely to acquit when there is an 80 percent chance of a defendant's guilt than when the evidence of defendant's guilt is 80 percent reliable. He referred to this as the ''ease-of-simulation hypothesis'': a jury is more likely to acquit even in a strong circumstantial evidence case if it is actually able to visualize a scenario in which the defendant is innocent.

Why is the ability of the jury to picture a scenario in which the defendant is innocent so important? Heller argues it is because as jurors hear evidence they engage in a ''simulation heuristic.'' Jurors do not simply ''weigh'' evidence along a zero to 100 continuum. Rather, they use the evidence to construct a coherent mental picture — a story — of what actually happened. The first mental picture will be based, of course, on the prosecution's case, which Heller refers to as the ''factually inculpatory scenario.'' After the prosecution rests, the issue then becomes whether the defense can provide sufficient evidence for the jury to construct a ''factually exculpatory scenario.'' The stronger the FIS, the tougher it will be for the jury to create an FES. In other words, the stronger the FIS, the more difficult it will be for the jury to create an FES of ''minimally acceptable imaginability.'' Only if the FES is acceptably imaginable will an acquittal be a possibility.

It is easier for the prosecution to help the jury imagine an FIS through direct evidence. Direct evidence offers a vivid, structurally coherent story for the jury. Evidence such as eyewitness testimony and confessions paints bold pictures that produce a convincing story. Direct evidence travels in a single step to the ultimate issue of guilt or innocence. An FIS based on direct evidence makes it very difficult for a jury to construct an FES.

Contrast this with an FIS constructed from circumstantial evidence. By its very nature, circumstantial evidence is abstract, rhetorical, and probabilistic. It does not provide the vivid pictures found in an FIS based on direct evidence. Moreover, it is easier for juries to construct an FES in a circumstantial, rather than a direct, evidence case. A circumstantial case is by definition based on probabilities; by definition it will simultaneously provide evidence of both guilt and innocence. Even where the defense offers no evidence itself, in a circumstantial case the prosecution is providing the jury with evidence to create an FES.

Heller concludes by positing that juries are not interested in issues of legal guilt based on probabilities; they are interested in factual guilt. It is not enough for jurors to merely believe that probabilities indicate that a defendant committed the crime. Rather, as Heller says, jurors ''need to feel confident in their gut that the defendant actually committed it.'' (303).

The results of this are extremely troubling. ''Because gut-level certainty is a function of ease of imagination, not mechanical probability calculations, there is no necessary or predictable relationship between gut-level certainty and the objective or subjective probability of the defendant's guilt. And … it is all too easy for guilty defendants to appear innocent in circumstantial cases and innocent defendants to appear guilty in direct cases.'' (303). Heller concludes by suggesting that jury instructions could be designed to combat these tendencies in jurors. But until then, his analysis provides a sobering wake-up call to attorneys who try criminal cases before juries. Prosecutors and defense lawyers alike should consider the possibility that it may not be the strength of evidence, but rather the type of evidence, that may be the difference between guilt and acquittal.