People v. Pinto, decided yesterday by the Third Department, is an interesting case highlighting the so-called Bruton problem. In this burglary and robbery case, both defendants were tried together. A police officer read into evidence a written statement of the non-testifying co-defendant. The court found, however, that it was not sufficiently redacted to protect the Confrontation Clause rights of the defendant whose case was on appeal:
Here, despite the removal of defendant's name, the substitution of neutral pronouns and a limiting instruction by the court, Perkins's account of the events on the day in question contained numerous incriminating references to another participant in the crime. For example, Perkins's description of his unnamed accomplice stated: "[He] [t]hen walked up the stairs [of the victim's apartment] and I followed behind him," "[w]e both tied up the girls" and "[w]e were in the apartment for 20 minutes tearing it up looking for drugs and money when we left." To compound the error, almost immediately after reading the statement, the officer stated that he had investigated its accuracy by viewing surveillance tapes at the restaurant mentioned in it to see if he could identify defendant. Further, the prosecutor's opening statement had earlier made it clear to the jury that only defendant and Perkins participated in the crime. As a result, it is inconceivable that the jury could have considered Perkins's account of his accomplice's actions as describing anyone other than defendant and, thus, the statement was insufficiently redacted to be admissible in a joint trial.
The court went on to reject the People's harmless error argument:
Under the circumstances, we cannot agree with the People that the other evidence identifying defendant as one of the home invaders is overwhelming, as illustrated by the fact that an Allen charge was necessary because at least one juror disagreed with the initial verdict finding defendant guilty.
I agree with the court's Confrontation Clause analysis. Under the circumstances, it is difficult to imagine how the jury would not have believed that the person behind the pronoun was the defendant. Where I part ways with the court, however, is in its harmless error analysis. Harmless error analysis focuses on the amount and quality of the evidence and the likely effect, if any, the erroneously admitted evidence had on the fact-finder. See People v. Crimmins, 36 N.Y.2d 230 (1975). Here, I think the court erred in two respects. First, it did not discuss the nature of the "other evidence" against the defendant. Second, it appeared to use, as the sole criterion for lack of harmlessness, the fact that an Allen charge was necessary.
Determining harmlessness by looking at the course of jury deliberations -- such as the length of deliberations, whether an Allen charge was necessary, and the number of readback requests -- is problematic. A jury that takes time to reach a verdict, that asks for readbacks and clarification, and that requires an Allen charge may merely be a jury that is taking seriously its obligation to deliberate carefully. Or, it may indicate that only one or two jurors are unsure about the verdict. Determining harmless error by examining the jury's external behavior is a lot like reading tea-leaves.
With that said, it may very well have been the case that the other evidence against this defendant was weak and that the co-defendant's confession was a comparatively damning piece of evidence. It is difficult to determine from the court's opinion whether this was the case. (LC)
Comments