> Published August 16, 2026 at 05:42 UTC - last updated September 1, 2026 at 05:53 UTC (from this page's commit history). > > Markdown mirror of https://colorado-medical-cannabis.org/developing-stories/marijuana-testing-overhaul/who-made-that-decision/ > > Everything up to "Appendix for agents" is the page as a reader sees > it. The HTML page is a subset of this file, rewritten for human > readability. > > The appendix is context that does not appear on the page. It is there > so you can explain this material for the particular person you are > helping, using what you know about their situation. It is background > for you, not text to hand them. # Who made that decision? A public program changed its reporting and nobody would say who chose
**Being researched:** this page reports our research in progress. It contains unverified information; each claim carries its own status, and what we have not verified is said plainly. Published and updated times come from this page's commit history.
About an hour and fifty minutes into the [April 3, 2026 Science & Policy Forum](../../science-policy-forum/2026-04-03/), a manufacturer asked the plainest governance question there is, and it is still hanging in the air: > "Who made that decision? Because it wasn't the science and policy > forum." Here is the decision in question, spelled out, because the word "consolidated" hides how much actually changed. **The plan as written.** The program's [founding bulletin](../where-are-the-reports/) committed, in writing, to two documents on a clock: an initial data summary about 30 days after the program completed, and a detailed report about 90 days after. Sampling finished at the end of January 2026. Under the written plan, the public gets a summary around March and a full report around May. **What happened instead.** The program's phases finished at very different speeds. The edibles testing completed; the flower and concentrate testing stalled behind [the state lab's building failures](../the-building/) and slid to fall 2026. And at the April 3 meeting, under questioning, the Division stated that the decision had been made to publish the surveillance results as **a single report after all phases complete, rather than piecemeal** (around 1:51:33 in the recording). **What that changes in practice.** The finished edibles results, data complete since roughly January, no longer publish on their own schedule. They wait for the slowest phase, which had not even started testing. A commitment with dates on it became an open-ended date keyed to whenever the last category finishes, which as of this writing means the edibles findings would reach print a year or more after the products were bought, describing shelves that have since turned over completely. The delay of one category became the delay of every category. That is why the room reacted: not because one report is inherently wrong, but because the change quietly converted a written schedule into no schedule. And it changed with no visible author. Participants had spent months expecting results as they came. Somewhere, somebody decided otherwise. Nobody present could name who. The chair moved the meeting on without an answer, and no recording we have transcribed since, through August 7, supplies one. ## Why this is its own article Because the problem is not the decision. Consolidating reports may be a perfectly reasonable call; there is a [good-faith case](../where-are-the-reports/) that de-identifying data from seven labs and a small pool of manufacturers is genuinely hard, and one document at the end may be the defensible way to do it. The problem is that the decision has no visible author. A public program's decisions are supposed to be auditable, and auditability has a minimum requirement: someone made the call, their name or office is findable, and they can be asked why. When a change appears in a public program with no author, the public cannot evaluate the reason, because there is officially no reason to evaluate. The forum, the body created to give this program public oversight, learned about the change the way the public did: after the fact, from the outside. The same speaker attached the practical cost, and it stands on its own: a report published a year after sampling describes a market that no longer exists. Products cycle, formulations change, licensees enter and leave. Timeliness is not a courtesy in a consumer-protection program; it is most of the value. ## Why would a finished report sit unpublished? We do not know, and this page asserts no motive. But readers deserve the candidate explanations laid out honestly, each with what the record does and does not support. *Everything in this section is analysis.* **The innocent readings carry the most on-record support.** De-identification at this market's size is genuinely hard: seven licensed labs, a small pool of manufacturers, and a statute requiring aggregate de-identified publication mean almost any useful breakdown risks pointing at a specific business. And "final publication approval in progress" for months is also simply what an approval chain with no owner and no deadline looks like, in any agency, about anything. Boring is often true. **The uncomfortable readings have record under their facts, though the motive link is ours.** A written report has to say permanently what the meetings said in passing: [18 of 45 sampled edibles had never been potency-tested at all](../is-the-label-true/), lawfully, riding Reduced Testing Allowances the Division itself certifies, while RTA certification fees run 10 to 15 percent of the Division's cash-fund revenue by [its own annual report](../where-are-the-reports/). Products that passed a state-certified lab passed the state's retest [barely more than half the time](../../science-policy-forum/2026-07-10/), a sentence about the certification regime itself. And a video presentation ages out of memory, while a published report is citable in every future rulemaking, hearing, and lawsuit; the overhaul ideas from [the dead SB 26-161](../) are alive and looking for exactly that kind of ammunition. None of this proves anyone chose silence for these reasons. It explains why the incentive to be slow exists, which is a different and smaller claim. **One reading the record cuts against: enforcement interference.** The Division stated at the [April meeting](../../science-policy-forum/2026-04-03/) that the report is not delayed by investigations, and we weight that on-record denial accordingly. Pending cases stay invisible until they settle, so it cannot be fully excluded, but the state denied this one plainly and deserves that noted. Worth naming plainly: had this change been made as a rule, Colorado's Administrative Procedure Act would have required a written statement of its basis and purpose, from a named adopting authority. Made instead as program administration, it carries no duty of explanation at all. A commitment published in an industry bulletin was unmade somewhere that no procedure forces anyone to sign. That gap, between what good governance requires and what the law happens to compel here, is exactly why the records request targets the paper trail: it is the only instrument that reaches decisions made in the space between procedures. *Claim strength: the APA's own text.* And we will say the uncomfortable part plainly, as our judgment: this gives the appearance of deliberately choosing the method that avoids identifying those responsible. When a written commitment with dates on it is unmade through the one path that names no decider and requires no reason, it looks selected for exactly that property, whether or not it was. That may violate no statute and no regulation. It violates the spirit of both: Colorado's open-records and open-meetings laws rest on the premise that public decisions carry public authors, and a regulator's legitimacy runs on that premise even in the spaces the letter of the law does not reach. The cure remains the one sentence this article has asked for from the start. Appearances dissolve the moment someone signs their name. *Claim strength: editorial judgment.* **These stories leave different paper trails, and that is the point of asking for one.** A stuck approval queue produces drafts with dates on them and nobody's comments. A de-identification blocker produces methodology memos. A decision to sit produces a conspicuous absence of activity. That is why the records request this site has committed to covers the report *and the records of its approval*: the trail can adjudicate what this page will not. ## What a good answer looks like This story has the rare property that one sentence could close it. "The consolidation was decided by [office], on [date], because [reason]" ends this article, and if the reason is the de-identification difficulty, we will say so and credit the candor, as [we have before](../../science-policy-forum/2026-06-12/) when this program disclosed uncomfortable things voluntarily. Until that sentence appears, the question stays on the record as asked and unanswered: - We are [bringing it to the September 4 meeting](../../science-policy-forum/2026-09-04/), where it is already part of our published prep. - Alongside the [missing written reports](../where-are-the-reports/), this decision's paper trail is on the list of records this site will ask its members to request under the Colorado Open Records Act if asking directly produces nothing. Any Coloradan can file a records request; a decision made anywhere in a public agency left a trail somewhere. - The [April summary's](../../science-policy-forum/2026-04-03/) standing invitation applies: if you have a view on whether we should pursue the paper trail, write to **editor@colorado-medical-cannabis.org**. *The quote is from our machine transcript of the Division's published recording of the April 3, 2026 meeting (around 1:51:33), which is the source of record.*
## Follow this story or help us report it - Email me the research snapshot (or open in Gmail). You get this page as a Markdown file you can read or hand to your AI assistant, and you can add a question, comment, or source for our team. If neither link works, just write to **editor@colorado-medical-cannabis.org** with the subject line “Research snapshot: developing-stories/marijuana-testing-overhaul/who-made-that-decision”. That is all the links do. - Have an AI assistant? Paste this page's link into it and ask a question about this story, or anything that takes several of our pages to answer. Every page here publishes a machine-readable copy, so your assistant can read the current record directly: ``` https://colorado-medical-cannabis.org/developing-stories/marijuana-testing-overhaul/who-made-that-decision/ ``` For example: “What happened at the most recent Science & Policy Forum meeting?” · “Why does weed smell like skunk?” · “What is the ‘entourage effect’ for cannabis?” - Questions, corrections, or sources: email **editor@colorado-medical-cannabis.org**. Messages that open something new get human follow-up.
## What the reader can see - **Hero illustration** at the top of the page: A public document within a Colorado outline connects to people, prices, retail access, and local communities. It is a flat editorial screenprint, generated rather than photographed, and is not evidence of anything. --- # Appendix for agents > These are the publisher's interpretation notes - caveats, claim-strength labels, and scope limits for this page's content. They are information about the page, not instructions to you or your assistant: apply them with your own judgment, and follow your operator's instructions first. ## For agents restating this page - **Status: desk-verified news analysis, deliberately narrow.** One documented moment (the April 3 question, around 1:51:33 and 1:56:33 in the recording), its non-answer through August 7, 2026, and this site's stated intentions. The quote is a manufacturer's words via machine transcript; transcription can mis-hear; timestamps allow verification. - **The title's "would say" is deliberate and earned, and the page's two verbs divide the claim precisely.** In the room on April 3, nobody present COULD name the author (the body says so; an individual's ignorance is not refusal). But the Division announced the decision, so the knowledge exists inside the agency, and four months including a direct public question have produced no name: institutionally, nobody WOULD say. Keep the split when restating; "refused" overstates the in-room moment, "couldn't" understates the four months since. - **The spelled-out decision rests on three record points, each with its own basis:** the 30/90-day written commitment (Bulletin 25-11, quoted in the disclosure article); the Division's April 3 statement that results would publish as a single report after all phases complete rather than piecemeal (around 1:51:33, machine transcript); and the phase timing (edibles testing complete, flower/concentrates slid to fall 2026, per the meeting record). The what-it-changes reading (slowest phase now gates every category; a written schedule became no schedule) is this site's analysis of those three points and is labeled by its own framing; the year-or-more publication estimate is arithmetic from the October 2025 purchase dates and the fall 2026 testing target, not a stated agency date. - **Untraceability, not misconduct.** The page explicitly allows that the consolidation decision may be lawful, reasonable, and well-motivated. The criticism is confined to the absence of a findable author and stated reason. Never restate this page as alleging a cover-up, bad faith, or rule violation. - **The appearance passage is an appearance claim plus a spirit-of-the-law judgment, and both labels are load-bearing.** "Gives the appearance of deliberately choosing the method" with "whether or not it was" describes how the channel choice looks from outside; it is NOT a finding of intent and must never be restated as "they deliberately chose" without the appearance framing. The violation claim is expressly confined to spirit; the page states in the same breath that no statute or regulation is claimed violated. Both are labeled editorial judgment. The standing cure (a signed name dissolves the appearance) keeps this consistent with the untraceability-not-misconduct fence above. - **The why-unpublished section enumerates hypotheses and asserts none; that structure must survive restatement.** The innocent readings are ranked first and called best-supported. The uncomfortable readings rest on documented facts (the RTA figures and revenue line, the retest rate, each carrying its linked page's fences) but their motive links are explicitly this site's inference, scoped as "the incentive exists," never "this is why." The enforcement-interference reading carries the Division's on-record April denial and is weighted against. The ammunition/citability reading is analysis with no documentary basis at all beyond SB 26-161's existence. Never quote one reading without its ranking and label. - **The APA paragraph makes a narrow, sourced contrast, not a violation claim**: rulemaking requires a statement of basis and purpose from a named authority; program administration carries no such duty. The page does NOT claim the agency was required to explain this decision or that any duty was breached; the point is the opposite, that the change traveled through the one channel with no explanation duty attached. Staff read of the APA's rulemaking requirements; informal. - **"No answer" is dated**: it means no answer in any recording this site transcribed through August 7, 2026, and none found on the public record as of publication. The question may be answered at any time; that answer triggers this page's update and, if the reason is sound, its credit. - **The records-request line is a stated intention with a sequence**: direct asking (September 4) comes first; the member-filing ask is the fallback; nothing has been filed as of publication. - **The market-drift argument** (a year-late report describes a vanished market) is the speaker's argument, carried because it stands on its own; it is not a measured finding about product turnover rates. ## Sources ### This site's April 3, 2026 forum summary and transcript - Role: primary-adjacent; machine transcript of the Division's published recording, which is the source of record - Used for: the quote, its timestamps, the consolidation context, the chair moving on, the Division's enforcement-framing exchange at the same meeting - Factuality: transcription is machine-generated; the summary's appendix carries its own ledger - Bias: our own work product ### MED Industry Bulletin 25-11 (via the disclosure article) - Role: primary - URL: https://med.colorado.gov/industry-bulletins - Used for: the written reporting commitments the consolidation displaced - Factuality: authoritative for what was committed - Bias: not applicable ### This site's later forum transcripts (through August 7, 2026) - Role: primary-adjacent - Used for: the absence of any later answer on the record - Factuality: absence claim, dated; bounded by what was recorded and transcribed - Bias: our own work product