A two-year interdisciplinary study by UC Berkeley's Cannabis Research Center has found that the single most consistent predictor of reduced unlicensed cannabis cultivation in California is not enforcement or local bans, but the presence of a licensed farm. The finding, which emerged from a combination of satellite mapping, energy consumption data, state track-and-trace analysis, and ethnographic fieldwork across six counties, challenges the prevailing logic that has driven cannabis enforcement policy since legalization.
Michael Polson, director of the Cannabis Research Center and principal investigator on the project, presented the findings alongside five co-researchers in a webinar last week. The study tracked licensed and unlicensed cultivation across California from 2018 to 2024, a period that included a sharp expansion in unlicensed activity, a dramatic price collapse, and a gradual stabilization.
The geographic picture that emerged from satellite data showed a large boom in unlicensed outdoor cultivation between 2018 and 2020, followed by a sharp contraction by 2022, with activity stabilizing by 2024 at roughly 2018 levels. Throughout the study period, the researchers observed a sustained shift away from pure outdoor production toward mixed-light cultivation, a trend that held whether the overall market was expanding or contracting. When there was growth, mixed-light grew faster. When there was contraction, outdoor contracted harder.
© Glass House Brands
On the licensed side, Santa Barbara County emerged as by far the largest producing county in the state during the study period, accounting for a disproportionate share of the 50.4 million kilograms of cannabis produced by licensed farms across California. The explanation, the researchers noted, is largely infrastructural. Santa Barbara had an existing stock of greenhouses previously used for flowers and produce that were no longer in operation, which cannabis companies repurposed for cultivation. Combined with favorable climate and permissive local policy, the county offered a ready-made platform for large-scale mixed-light production that other regions could not easily replicate.
The research team's spatial modeling found that the presence of a licensed farm nearby was the most consistent variable correlated with a decline in unlicensed cultivation across all three time periods studied. Jeremy Sorgan, assistant research professor at Northwestern University Mills campus, described the mechanism as primarily civil rather than criminal: a licensed farm in the area means more code enforcement presence, more foot traffic, and a generally less hospitable environment for operations trying to stay hidden. Unlicensed farms sometimes hide in plain sight, he noted, but in areas with more people and more regulatory activity, they cannot do so for long.
Bans, by contrast, showed no reliable effect. Across the three time periods, ban counties were no more effective at limiting unlicensed cultivation than permit counties, and in the final period from 2022 to 2024, seven of the ten counties with the most growth in unlicensed activity were ban counties, while the six counties with the greatest decreases were all legacy cultivation counties with active permitting programs. Eradication-based enforcement showed a similar pattern: it deterred some growers in the first period, but its effect weakened over time and by the third period had actually reversed, with enforcement reputation correlating with an increase in unlicensed activity rather than a decline.
The researchers described this dynamic, which surfaced repeatedly in ethnographic interviews, as the whack-a-mole effect: growers in ban counties and heavily enforced areas kept moving rather than stopping, pushing into more remote and environmentally sensitive land with faster crop turnaround cycles and less attention to environmental compliance. Civil enforcement tools, by contrast, including fines, landowner liability rules, and environmental regulations, showed growing efficacy over time as they matured and word spread within communities. The key distinction the researchers drew was between enforcement that pushes growers underground and regulation that brings them toward compliance.
The economic data provided perhaps the starkest explanation for who remained in the unlicensed market by the later years of the study. Black market outdoor cannabis prices collapsed from around $1,000 per pound in 2020 to well below $300 per pound in 2021 and 2022, often falling below estimated production costs of $250 to $400 per pound. Cannabis industry insiders in Lake County estimated that 30 to 50% of legacy growers dropped out almost entirely because of market conditions rather than enforcement or policy. Those who stayed were, in the words of one Mendocino County interviewee, experiencing the economy as the best enforcer of the illegal market. They were predominantly people with few alternatives, the researchers found, including low-income rural residents, the elderly, and the disabled whose livelihoods were tied to their land.
Amanda Ryman, who served as co-principal investigator and led the licensed production analysis, found that California's track-and-trace system, while useful as a compliance tool, was too riddled with data inconsistencies and user error to function reliably as a research instrument. Once flower left the farm, harvest IDs were reassigned as batches were combined at distribution and processing centers, making it effectively impossible to track product through the supply chain. Consumer survey data from 2022 added another layer: consumers reported consuming approximately 1,400 tons of cannabis flower that year, compared to 466.6 tons recorded as sold through licensed point-of-sale systems, and 1,700 tons reported as produced by licensed farms, suggesting both significant leakage out of the regulated market and heavy conversion of flower into manufactured products.
