Federal Evidence Review

In drug conspiracy trial, prosecution witness testimony that the defendant conspiracy organizer arranged to have a crack addict steal children's clothing as payment to the defendant for providing her crack, was erroneous prior act evidence, but the error was harmless; Seventh Circuit noted the testimony constituted less than 30 trial testimony pages from the 1500 pages in the case, in United States v. Johnson, 584 F.3d 731 (7th Cir. Oct. 14, 2009) (No. 06-3048)

Under FRE 103(a), errors in admitting or excluding evidence, are subject to harmless error analysis. The circuits have considered a variety of factors as evidence that an error was harmless. One common factor examines “the importance of the wrongly admitted” evidence at trial. See, e.g., United States v. Al-Moayad, 545 F.3d 139, 164 (2d Cir. 2008) (citing as factors, the strength of prosecutor's case, the prosecutor's conduct with the erroneously admitted evidence, and whether the erroneously admitted evidence was cumulative of other properly admitted evidence). Occasionally quantitative comparisons are used to express this factor. A recent Seventh Circuit case provides an example of looking quantitatively at the impact of erroneously admitted evidence.

In the case, defendant Johnson was charged with “overseeing a gang-related crack distribution operation” as a continuing criminal enterprise (“CCE”). During his trial and over his objection, the trial court admitted testimony by cooperating witnesses about the defendant’s drug and related gang activities. The evidence included testimony about the defendant’s uncharged drug activities before charged conduct from 2000 to 2002. The defendant was convicted. On appeal, he contended that admitting evidence about his alleged drug and gang activity before the charged period violated the FRE 404(b) strictures on admission of other act evidence.

The Seventh Circuit affirmed the conviction. In doing so, the circuit concluded that all but one witness’s prior act testimony was admissible as providing “background to explain Johnson's leadership role” in the charged conspiracy and “to clarify the relationship” between the defendant others in it. In addition, the testimony was illustrative of the defendant’s “intent to distribute drugs ‘based on his prior relationship with’ the witnesses and to show his knowledge of the drugs being sold.” Johnson, 584 F.3d at 736-37 (citations omitted).

However, the trial court erred in admitting a witness who “testified that in 1998 Johnson asked her to steal children's clothes for him in exchange for crack. She explained at some length how she stole the clothing and described the small sizes she targeted.” The circuit noted that even the government eventually realized this testimony should not have been admitted, in that:

“counsel for the government struggled to articulate what could possibly be the probative value of this testimony, and in fact counsel conceded that the government went too far in soliciting these details. We agree. The prejudicial effect of testimony showing that Johnson had a crack addict stealing clothes for small children certainly outweighed what little probative value such testimony may have added. It is difficult to see how this testimony served any purpose other than to suggest that Johnson is a bad person, and one with access to small children at that.”

Johnson, 584 F.3d at 738.

However, this error did not warrant reversal because the error was harmless. As explained by the circuit, the witness’s:

"testimony represents less than 30 pages out of over 1500 pages of trial transcript, a small island in a sea of evidence of Johnson's guilt. Multiple witnesses testified at length regarding Johnson's leadership role in the Quincy drug operation. They testified that they sold crack for Johnson throughout the charged period, giving details about how he obtained the crack and established procedures for the gang's drug sales. Johnson provided no evidence of his own to rebut that testimony, and given the weight of the evidence supporting Johnson's conviction, we conclude with fair assurance that the district court's error in admitting the Green testimony had no sway on the jury, and therefore was harmless."

Johnson, 584 F.3d at 738.

This ratio of 30 pages of offending transcript to about 1500 pages of the trial transcript is a two percent rate. For the Seventh Circuit, two percent is a “small island in a sea of evidence of … guilt” that can be one factor suggesting harmless error. Other circuits have reached similar quantitative evaluations, for example:

• 1.4% - Tenth Circuit: United States v. McVeigh, 153 F.3d 1166, 1202 (10th Cir. 1998) (in admitting victim evidence testimony in bombing case, the "potential prejudice" of witness's testimony "comprised less than four lines out of fourteen pages of testimony" - a ratio of approximately 1.4%.)
• 1.9% - Fifth Circuit: United States v. Parks, 68 F.3d 860, 867 (5th Cir. 1995) (in criminal case, erroneous reference to defendant's civil dispute with banking regulators "comprise approximately thirteen pages out of approximately 700 pages of trial transcript" - a ratio of approximately 1.9%)

Of course, in Johnson, as in the Tenth and Fifth Circuit cases cited above, a quantitative assessment was not the sole criterion used by the circuit. The Johnson court also noted that the defendant had failed to “provide[ ] … evidence of his own to rebut” the erroneously admitted testimony and the error was readily admitted by the prosecutor at trial and the court provided limiting instructions to the jury. A variety of factors are likely to be present and not the use of only the quantitative factor.