A Clear-Eyed Look at Oregon’s Cannabis Testing Industry

Last year, around this time of year, I conducted an experiment much like the one I wrote about earlier this month in California. In addition to running the experiment, though, I also requested data from the state of Oregon about every THC test that had been performed since 2018. 

Buckle up – just like my results in California, this data shows clear evidence of how rampant THC inflation is occurring in Oregon. This is a huge problem for consumers in the state that the data shows is getting worse over time. Without state intervention there’s no reason to believe it will improve. 

Identifying THC Percentage Inflation

My intent with this was to emulate designs I had seen in medical research: a retrospective review of all of the tests so far, in addition to a prospective review that emulated a proficiency test. As it often is in these cases, my main constraints were financial and logistic resources.  Once I found out that I simply didn’t have the financial resources to manage a fully blind proficiency test, I accepted help from one of the labs to homogenize samples acquired from some friendly farms to the experiment and accepted their help to get them to the participating labs. But this also meant that I didn’t have a clear, personal chain of custody.

In other words, there were several issues with the experiment design, but because I had managed to get so many labs to participate, I really wanted to capitalize and move forward. 

I had three flower samples and one distillate. One of the flower strains, Sample A, is one of the strangest, naturally kiefiest cannabis  I’ve ever encountered.  A regular award-winner for its THC content (which normally tests on-label around 33%), this flower is unlike many others I’ve encountered. The other two flower samples were two different standard strains that test around 18% on average for their grower. The distillate is known to test around 76%.

The results weren’t what I expected, but they prove the point that consumers can’t fully trust what THC percentage on the labels tell them without getting the context of the labs

Sample A, the powerful flower, had a wide range of test results: from 32-42% from the min and max of the range, which is a massive range, especially in light of how close the results were for Sample B.  Sample B  had a relatively tight range of around 5% from 17-22%. Sample C’s range of 17-26% is again quite large at 9%, especially considering most results are clustered around 17-19%.  As D was a distillate and only had a Delta 9 value, it should show less variance than flower does, and it does – a tight range of only 4 %: 72 – 76%.

What was most shocking to me in these results was the range of all of the tests. And while it was surprising to see a 42% value from one of the labs, the most surprising result, to me, was the high value on sample C. While most labs tested that sample around 18, one lab’s test of 26% is shocking. 

Retrospective Records Review

Michael Zoorob, a Harvard-trained researcher, wrote a paper about THC Results in Washington and Nevada that inspired me to contact the state of Oregon and perform a public records request. In his paper, he posited that there was evidence of issues with THC accuracy, and that they were demonstrated by the discontinuation of the normal curve at around 20%. I theorized that Oregon would have the same issue, and emulated the charts that I made after his.  

In 2019, the year that a particularly scathing report from the Secretary of State about the state of Oregon’s labs was released. In it, there were several recommendations on how Oregon should move forward to remediate their program. I requested data that was inclusive of the audit.

2019’s THC data looked relatively normal.

2020 you can start to see the skew form, particularly a large step-down. Look at how ‘spiky’ the data becomes around 21-24% and how there is a large peak forming at 22%.

2021 became more skewed with larger, more aggressive discontinuations of the normal curve. Look at the massive step up between 20 and 21.

The reasons these discontinuities happen is because of the push to get flower above 20%, which is considered a threshold of economic viability. In light of the recent Canadian study of THC results demonstrates that the results fall on a normal curve shape.

As I had made this inquiry in 2022, the data I have for that year was still incomplete. Note also that this curve doesn’t peak at 20, nor should it — the average THC concentration in Oregon is actually at 23.3 for the year of 2021. Which is exactly where this data set peaks. Hence it’s clear to see that the trend is getting as close to the cutoff for selling product at a higher price. This isn’t something that occurs naturally but is intentional manipulation by bad actors. 

I also wanted to use the data to examine what the results looked like on a per-lab basis, particularly high-THC results. I isolated all results of flower that were over 36% and mapped them to labs. It makes a very compelling chart, as it’s very clear that one lab is creating most of the very high results. It’s staggering how evident it is.

I’ve published the results to Tableau, and you can use it to explore the data more. 

What Could the State Be Doing? 

At the time that I ran this experiment, Oregon had not yet implemented rules that tightened the threshold for THC percentage, requiring that the results be within 10% of the label. They’ve additionally implemented rules to begin testing for mold and heavy metals, but have had some issues. OLCC has their hands full with human trafficking and illegal farms in the south and ethical issues causing the governor to start to sweep the agency clean. The best moves the agency could make would be to continue the program they started to test cannabis off the shelf and continue to aggressively recall product that fail. They can continue the legislative work to ensure that disciplinary actions are within the toolbox of both OLCC and ORELAP

OLCC has every piece of information at their fingertips that I do – I gave them what I had. The most important thing that they need to do is act on it, and continuously monitor it

What Does This Mean For Consumers?

Much like our results in California showed, the retrospective study of Oregon’s results coupled with the round-robin style lab testing show that testing in Oregon still is not standardized or reliable enough for consumers to be able to depend on the information. Lab issues are rampant. Consumers are overpaying for product, and the poorly regulated market is putting medical consumers at risk of not being able to appropriately dose their medication. If these laboratories are willing to game THC results to get their clients more money, they would be more than willing to pass product that should have failed for pesticides or mold. Unless Oregon decides to use the large array of tools at their disposal, consumers of regulated cannabis continue to be at risk.