Is Reasonable Doubt Self-Defining?, 64 Villanova L. Rev. __ (forthcoming, 2019)

In Wisconsin, our pattern burden of proof instruction in criminal cases concludes by telling jurors "not to search for doubt" but "to search for the truth."  It is obvious to everyone who is not a prosecutor or former prosecutor that such language lowers or diminishes the reasonable doubt (RD) standard.  And if it wasn't obvious, my coauthor Larry White and I twice demonstrated the instruction's burden-lowering effect in controlled studies published in journals at Richmond and Columbia.  (And to to their credit, more than twenty trial court judges now recognize the pattern instruction's defect and have modified the language.)

But what about states that don't attempt any definition of RD?  States that decline to do so contend that RD is self-defining and therefore needs no further explanation.  On the plus-side, jury instructions that leave the term undefined can't do anything blatantly unconstitutional, such as telling jurors "not to search for doubt."  But is RD really self-defining?  Published studies demonstrate that it is not, as jurors fail to distinguish between it and the two lower, civil burdens of proof.  But these studies are old and used small, unrepresentative samples.  Therefore, Larry White and I conducted a new study with more statistical power to see if RD, when left undefined, offers defendants any more protection than two lower standards of proof, preponderance of the evidence (POE) and clear and convincing evidence (CCE).

Our article is forthcoming in the Villanova L. Rev., and you can find a pre-publication draft on SSRN or on my website's articles page.  Read the abstract, after the jump, which includes our findings.  (And scroll to the bottom of the post for a bonus video.)  Spoiler alert: the earlier studies were confirmed.  RD is not self-defining, i.e., when left undefined, there is no statistically significant difference in conviction rates between test participants who received a RD jury instruction, a POE instruction, or a CCE instruction.

ABSTRACT:

Many courts believe that reasonable doubt is self-defining and, therefore, do not explain the concept to their juries. The empirical evidence, however, suggests otherwise. Controlled studies demonstrate that mock jurors do not distinguish between reasonable doubt, clear and convincing evidence, or even preponderance of evidence standards when reaching their verdicts.

This Article presents our empirical study in which we sought to (1) conduct a more powerful test by remedying the methodological weaknesses of earlier studies, and (2) determine whether, instead of following their burden of proof instruction, mock jurors use a simple heuristic or rule of thumb regarding the quantum of evidence necessary to convict.

Our first finding is consistent with previous findings: we found no significant differences in conviction rates between groups that received different burden of proof instructions. Second, the data also revealed what we call “the 60/65 rule.” That is, nearly all study participants either (a) said that less than 60% of the evidence favored the State and voted not guilty, or (b) said that more than 65% of the evidence favored the State and voted guilty.

These findings demonstrate that reasonable doubt is not self-defining. Not only do mock jurors in multiple studies fail to distinguish between reasonable doubt and the two lower, civil burdens of proof, but they are also willing to convict criminal defendants on a quantum of evidence (approximately 65%) that is much lower than what judges expect and the Constitution requires.

Given these findings, we recommend that courts use a relative, comparison-based definition of reasonable doubt to properly convey to jurors the high burden the government must satisfy before depriving a person of life, liberty, or property.

Villanova Bonus Video: