Extrapolation of Drug Weight Insufficient Proof of That Element

State v. Carpenter, Minn.Ct.App., 3/27/2017.  Fair warning:  this is a case about statistics.  For those of you who've stuck around, it's also about hubris.

The cops seized sixty-four individually wrapped plastic baggies of suspected heroin.  The state charged Mr. Carpenter with first degree sale, which requires ten grams or more of heroin; and with first degree possession, which requires twenty-five grams or more.  Budgets being what they are, the BCA used something called the "hypogeometric sampling plan".  Check it out.  The BCA tester did whatever this sampling plan said to do and it spit out a number:  19. Test nineteen bags, that's all.  So, that's what she did. And came up short of the required ten grams by almost a gram for the sales count. Not to worry, she pronounced some more statistical words and concluded that she was 95% confident that 90% of the total ninety percent of the total sample contains the same substance as the nineteen tested.  Read for yourself what happens next:

Goldstrand testified at Carpenter’s trial that the hypogeometric sampling plan supports a 95% confidence rate that 90% of the total population size contains the same substance as the 19 test samples. Based on Goldstrand’s testimony, the district court concluded that, of the 31.333 grams of the substance seized, at least 28.1997 grams consisted of heroin—28.17 grams is 90% of the total sample. Because 28.1997 grams is above the 25-gram statutory amount required for first-degree aiding and abetting the possession of a controlled substance, and the ten-gram statutory amount required for first degree aiding and abetting the sale of a controlled substance, the district court found Carpenter guilty of both charges

Mr. Carpenter chose to have a bench trial.  Defense counsel moved for a judgment of acquittal based upon the weight of the heroin. The trial court looked at the rule, Rule 26.03, and concluded that it only applied to jury trials.  Motion denied.

The rule does say, "jury" multiple times, so a literal reading of the rule supports the trial court's conclusion.  To avoid this absurd result, the court of appeals has to reach back all the way to the beginning of the criminal rules, to rule 1, which says that the rules - all of the rules - apply without differentiation between jury and bench trials to prosecutions for felonies, gross misdemeanors, misdemeanors and, and, petty misdemeanors.  The judge got this one wrong.

And the Minnesota Supreme Court has already said, no, to this statistical sampling method for proof of weight.  State v. Robinson, 517 N.W.2d 336 (Minn. 1994).  Extrapolation of weight through random testing is inappropriate when the substance being sampled has a high risk of substitutes; drug dealers mix bogus substances with the real stuff all the time.