On the inefficacy of limiting instructions

The rationale for allowing into evidence a defendant's criminal record asserts that such evidence can be used for the limited purpose of impeaching a dèfendant witness's credibility and, in accord with judges' instructions, will not be used to assess likelihood of guilt. The effect that the defendant's prior record has on mock jurors' assessments of credibility and guilt was tested in a two (cases) x four (type of prior conviction) factorial design. Adults' ratings of the defendant's credibility did not vary as a function of prior record and were consistently the lowest of the credibility ratings of all witnesses. Conviction rates did vary by prior record, however, with the highest conviction rate occurring when the prior conviction was the same as the present charge and the lowest conviction rate occurring in the no-prior-conviction condition. Defendants with a previous conviction for perjury or a dissimilar crime were convicted at an intermediate rate. We concluded that the risk of prejudice to the defense under existing policy is greater than the unrealized potential benefit to the prosecution.Portions of this research were presented at the 90th Annual Convention of the American Psychological Association in Washington, D.C., August 23–27. 1982. The authors wish to thank Norman Berkowitz, Edward Krupat, and Marianne LaFrance for their comments on an earlier draft of this paper, and Hon. Robert J. Hallisey for providing us with sample instructions and for his counsel. Of course,the authors accept the responsibility for what they did, found, and concluded.