Key Takeaways
- New Jersey’s Cannabis Regulatory Commission seeks to address cannabis lab shopping by hiring a contractor for statistical audits of testing data.
- Lab shopping occurs when brands select labs based on favorable results rather than accuracy, creating trust issues in cannabis testing.
- The RFP outlines detailed requirements, including a technical report that examines market integrity and lab performance.
- Potency inflation in cannabis affects medical dosing and consumer fairness, leading to potential risks for patients relying on accurate labels.
- New Jersey’s proactive data analysis approach contrasts with California’s method of retesting products on dispensary shelves.
Cannabis testing has always relied on trust. A grower submits a sample, a lab runs it, and a Certificate of Analysis tells everyone downstream what’s actually inside the package. That system only works if the lab has no reason to bend the numbers. New Jersey just admitted, in writing, that it isn’t fully confident that’s the case.
The state’s Cannabis Regulatory Commission (NJCRC) published a Request for Proposals last week, seeking a contractor to conduct ongoing statistical audits of testing data. Buried in the background section is a sentence that cuts straight to the point: “NJCRC need[s] to ensure that cannabis lab shopping is not taking place.” That’s not a hypothetical concern. It’s a direct acknowledgment that the current system has a blind spot, and the state wants a data-driven way to close it.
This puts New Jersey in an interesting position relative to other cannabis states. Rather than waiting for a whistleblower, a lawsuit, or a viral news story to expose bad actors, the NJCRC is trying to build a repeatable detection system from the numbers already sitting in its database. It’s a quieter approach than California’s, but arguably a more proactive one.
What Is Cannabis Lab Shopping, and Why Does It Persist?
Lab shopping happens when a cannabis brand routes its products to whichever testing lab consistently returns the most favorable results, whether that means higher potency numbers or fewer contamination failures. It’s rarely provable in a single instance. A brand can always claim it switched labs for turnaround time, price, or convenience. But across a market, the pattern shows up in the data even when no single case gets flagged.
The incentive structure explains why it keeps happening. Dispensaries want to stock high-THC flower because customers often shop by percentage. Producers need labs willing to report those numbers. And testing labs operate as businesses competing for the same client base. A lab that consistently reports conservative, accurate results risks losing accounts to a competitor willing to report higher ones. Nobody has to conspire for this to happen. The market pressure does the work on its own.
Research backs up how widespread the gap can be. A study out of CU Boulder, published in Scientific Reports, found labeled cannabis flower potency claims running as high as 39%, even though the highest value researchers actually measured in the lab was 33%. That’s not a rounding error. It’s a six-point gap between what’s printed on the label and what independent testing actually found.
What New Jersey’s RFP Actually Asks For
The scope of work in New Jersey’s RFP is more detailed than a typical compliance check. The winning contractor will need to produce four things twice a year: an executive summary for leadership and legislative briefings, an academically rigorous technical report, a reproducible analysis schema with annotated code, and a regular check-in schedule with NJCRC staff.
The technical report itself has to cover ten sections, including laboratory-level and market-level analysis, statistical signal detection, and market integrity indicators. That last phrase, market integrity indicators, is doing a lot of work. It shows that the state isn’t just asking “did this lab follow procedure.” It’s asking whether the overall pattern of results across the market looks statistically normal or suspiciously convenient.
Eligible applicants must have demonstrable experience analyzing cannabis lab data across multiple states, which narrows the field considerably. The contract is capped at $50,000, a modest budget for a project the NJCRC itself describes as essential to consumer trust and medical dosing accuracy. Proposals are due October 22, 2026.
What Statistical Red Flags Could an Auditor Actually Look For?
Without seeing the awarded contractor’s methodology, it’s worth knowing what this kind of analysis typically looks for in practice:
- Labs whose average reported potency sits consistently higher than the market mean for comparable strains, especially if that gap widens over time
- Sudden shifts in which lab a brand uses, particularly right after a batch fails or returns a lower-than-expected result elsewhere
- Failure rates that vary dramatically between labs testing similar product types, which can suggest inconsistent standards rather than genuine differences in product quality
- Statistical distributions that cluster suspiciously close to legal thresholds, a pattern sometimes called “threshold bunching”
None of these signals prove wrongdoing by themselves. A lab could legitimately serve a client base with better cultivation practices. But when these patterns show up together and persist across reporting cycles, they become much harder to explain as coincidence.
Why Potency Inflation Is a Consumer Protection Problem
It’s tempting to treat inflated THC numbers as a marketing quirk rather than a safety issue, since nobody overdoses from cannabis the way they might from a contaminated pesticide load. But that framing misses the actual harm.
Medical cannabis patients dose based on the numbers printed on the label. If a product labeled 28% THC actually tests closer to 22%, a patient managing chronic pain or a seizure disorder isn’t getting the dose their treatment plan assumes. That’s a real clinical problem, not a cosmetic one.
There’s also a straightforward fairness issue. Consumers are paying premium prices for high-potency products based on numbers that may not reflect what’s inside the package. A market where the label doesn’t match the product isn’t functioning as an honest market, regardless of whether anyone experiences a physical health scare.
How Does New Jersey’s Approach Compare to California’s?
California just took a very different route to the same underlying problem. Earlier this month, Governor Newsom signed AB 1965, which gives the state’s Department of Cannabis Control (DCC) the authority to buy or collect products already sitting on dispensary shelves and run independent, off-the-shelf lab testing on them. That closes a different gap: a batch can pass its original pre-sale compliance test and still end up different from what a customer actually buys, whether from mislabeling, degradation, or a sample that wasn’t representative to begin with.
AB 1965 also gives the DCC authority to require blind proficiency testing and round-robin testing among labs, checking whether different facilities return consistent results on identical material. That’s a direct response to a real enforcement record. Since December 2023, California has pulled licenses from four labs tied to potency inflation, including Verity Analytics, which was accused of overstating THC content by as much as 32%, per department notices cited by Beard Bros Pharms.
Put the two approaches side by side, and a clear distinction emerges. New Jersey is mining historical testing data for suspicious statistical patterns before a product ever reaches a shelf. California is physically retesting the finished product after it’s already there. One approach is retrospective and analytical, catching patterns a human reviewer would likely miss. The other is direct and physical, catching problems that only show up in the actual product, not the paperwork trail. A mature regulatory system may eventually need both.
What Comes Next for Cannabis Testing Oversight
New Jersey’s approach won’t produce results overnight. The first assessment cycle still needs a contractor, a data pipeline, and enough historical testing records to establish what “normal” looks like before anyone can flag what’s abnormal. But the model itself, a recurring, reproducible, code-documented audit, is a meaningful shift from relying on tips or media investigations to catch bad actors.
If other states are watching, and they should be, the real test will be whether NJCRC acts on what the data shows. A statistical signal is only useful if it leads to license reviews, retesting requirements, or public disclosure when a lab’s numbers stop making sense. Otherwise it’s just a very well-documented report sitting in a drawer.
Frequently Asked Questions
Cannabis lab shopping is when a brand or cultivator chooses which testing laboratory to use based on which one consistently returns favorable results, such as higher potency numbers or fewer contamination failures, rather than based on accuracy or reliability.
The NJCRC wants a biannual, reproducible statistical assessment of its cannabis testing program to detect lab shopping and other market integrity issues, ensuring that Certificates of Analysis accurately reflect product safety and potency for both recreational and medical consumers.
New Jersey is analyzing historical testing data for statistical anomalies that suggest lab shopping. California’s AB 1965 instead gives regulators authority to physically retest products already sitting on dispensary shelves, catching discrepancies between the original lab result and the finished product.
Potency inflation affects medical dosing accuracy for patients managing conditions like chronic pain, and it means consumers may be paying premium prices for products that don’t match their labels, which is a consumer protection and market fairness issue independent of physical harm.
The NJCRC’s Request for Proposals caps the contract at $50,000 for a biannual analysis that includes an executive summary, a technical report, reproducible analytical code, and regular progress meetings with agency staff.
