Hard cases may make bad law, but the Harcrow case makes important law

Harcrow is an uncommonly interesting opinion.

The central question in the case is whether a state crime lab report prepared at the behest of law enforcement officers concerning a particular identified criminal suspect are testimonial for Crawford purposes. A majority of Judge Erdmann, Chief Judge Effron and Judge Ryan says yes. Judge Stucky, joined by Judge Baker, suggests the answer isn't clear -- or at least wasn't before the majority's opinion, thus precluding the military judge's admission of the evidence without objection from being considered plain error.

The majority and the separate concurrence agree that plain error is to be measured at the time of appeal, not the time of trial. And Judge Ryan writes a thoughtful concurrence to observe, in part, that applying this rule piles legal fiction upon legal fiction to hold that a military judge erred by failing to follow case law that didn't yet exist when, in reality, the MJ likely would have applied the correct case law and the government would have abided by that case law. (She also observes that documents prepared in anticipation of litigation don't qualify as business records and that because the lab report at issue appears to have been prepared in anticipation of litigation, it might not have been admissible in the pre-Crawford world either.)

The principal point of distinction between the majority and the concurrence is what precisely was the state of law before Harcrow itself. According to Judge Stucky, at the time of trial a lab report was plainly admissible without Confrontation Clause difficulties under the well-established business record hearsay exception. Crawford threw that clear state of the law into confusion, but a confused state of the law cannot give rise to plain error, since the governing legal rule wasn't plain at all.

The majority, on the other hand, finds the requisite clarity in CAAF's Magyari opinion, which held that a run-of-the-mill urinalysis lab package wasn't testimonial. See United States v. Magyari, 63 M.J. 123 (C.A.A.F. 2006). In Magyari, CAAF had indicated that lab reports wouldn't always be admissible without Confrontation Clause implications. Judge Erdmann's majority opinion explains:

Our reasoning in Magyari that "[b]ecause the lab technicians were merely cataloging the results of routine tests, the technicians could not reasonably expect their data entries would 'bear testimony' against [a]ppellant at his court-martial" does not apply here. Here the laboratory tests were specifically requested by law enforcement and the information relayed on the laboratory reports pertained to items seized during the arrest of an identified "suspect."

United States v. Harcrow, __ M.J. ___, No. 07-0135/MC, slip op. at 13(C.A.A.F. Mar. 13, 2008) (internal citation omitted).

Judge Stucky counters that CAAF dicta isn't sufficient to establish the requisite clarity to result in an error being plain.

In the end, though, none of this helps LCpl Harcrow himself, since CAAF holds that his confession was the principal evidence against him and it was sufficiently corroborated even without the lab report. Thus, the majority concludes that the confrontation error was harmless beyond a reasonable doubt.

So Harcrow is extremely important in holding that lab reports prepared at the behest of law enforcement agents looking for evidence against an identified suspect are testimonial. The dispute over whether this particular error was plain is not terribly significant, since it will be unlikely to apply to any post-Crawford cases--cases in which, presumably, all of the trial participants would be attuned to the importance of a lab report's testimonial status and in which the defense should have preserved a Crawford issue.

But there is one additional fascinating aspect of Judge Erdmann's majority opinion. He begins with a lengthy discussion of forfeiture vs. waiver, which includes a helpful discussion of United States v. Olano, 507 U.S. 725 (1993). This discussion is instrumental in getting to the plain error analysis issue in the case. I expect to see this discussion cited in many briefs and many opinions to come.