Study: Marijuana Legalization Boosts State Agriculture Jobs by 9%

August 15, 2026

Photo by alexandr Kornienko via Unsplash.

Every legalization campaign leans on the same pitch: tax revenue, tourism dollars, a new industry sprouting up where prohibition used to sit. What gets less attention is the farm labor side of the equation -- the people actually growing the plant, not just selling it at the counter. A new analysis out of Texas Tech is putting numbers behind that gap, and the number is notable: recreational legalization boosted agricultural employment by roughly 9% in the states that enacted it.

The study was presented at the Agricultural & Applied Economics Association's 2026 meeting and leans on Bureau of Labor Statistics data covering about 95% of all U.S. jobs, tracked from 1990 through 2024 across 45 states. That's a long runway and a wide net, which gives the findings more weight than a lot of the back-of-envelope economic-impact reports that circulate around legalization ballot measures. It also adds to a small but growing body of research specifically tying legalization to farm-sector labor demand -- a narrower and more concrete question than the usual headline stat of how many people work in cannabis, full stop.

What the Texas Tech Study Actually Found

Photo by Lukas Blazek via Pexels.

The research comes out of Texas Tech's Davis College of Agricultural Sciences and Natural Resources, led by Tuhinur Rahman Chowdhury, whose work focused specifically on agricultural employment and wages in states that legalized recreational marijuana. Rather than just comparing legal states to illegal ones before-and-after a law change, the study used a synthetic difference-in-difference estimator -- a method that builds a weighted comparison group out of non-legal states to approximate what a given legal state's labor market would have looked like had it never legalized. That's a meaningfully more rigorous approach than a simple pre/post comparison, since it tries to control for the fact that states differ in industry mix, population trends, and baseline farm economies.

The underlying data came from the BLS Quarterly Census of Employment and Wages, a dataset that covers roughly 95% of U.S. jobs and is generally considered one of the most reliable sources for state-level employment and wage tracking. Chowdhury's team ran the analysis across four separate outcomes: agricultural employment, agricultural wages, total employment across all industries, and total wages across all industries. That structure matters, because it lets the researchers isolate whether any effect is specific to farm labor or just a reflection of broader economic activity.

The baseline result was fairly clean: agricultural employment rose about 9% in states after they legalized recreational cannabis, while none of the other three measures -- agricultural wages, overall employment, or overall wages -- showed a statistically significant change. In other words, legalization looks like it pulled more people into agricultural work specifically, without producing a detectable ripple effect on pay or on the rest of the state's job market.

Why Jobs Went Up But Wages Didn't

After recreational marijuana legalization, agricultural employment rose about 9%, while agricultural wages and all-industry employment and wages showed no measurable change.

A 9% jump in agricultural employment with no corresponding wage bump might sound like a contradiction, but the researchers frame it as exactly what you'd expect from an elastic labor supply in agriculture. In plain terms: when demand for farm labor increases, agricultural labor markets tend to respond by pulling in more workers rather than by paying existing workers more. That's a familiar dynamic in seasonal and manual-labor-heavy sectors, where there's typically a pool of available or underemployed workers who can be absorbed into new production without triggering a bidding war for their time.

Applied to cannabis specifically, the implication is that legal cultivation functions as a labor-intensive addition bolted onto an existing agricultural labor market, rather than a shock that reshapes compensation across the sector. Cannabis farming -- particularly the more manual aspects like trimming, harvesting, and processing -- draws on the same general labor pool as other specialty crops. States that already had agricultural infrastructure and workers accustomed to that kind of seasonal, hands-on work seem to have simply redirected some of that labor supply toward cannabis, rather than needing to import workers at a premium.

The authors are careful to note that this pattern isn't automatic or uniform. How much a state benefits appears tied to when it adopted legalization and what its local labor market looked like at the time -- a point that becomes central once you look at how the effect breaks down by legalization cohort rather than as a single national average.

Not All Legalization Cohorts Saw the Same Boost

Photo by Jupilu via Pixabay.

Averaging all legal states together produces that headline 9% figure, but the cohort-level breakdown tells a more uneven story. When Chowdhury's team split states out by the year they legalized recreational marijuana, the employment gains showed up most clearly among states that legalized in 2017 and 2019. States that adopted legalization in other years produced weaker effects, and in several cases the results weren't statistically significant at all.

That's an important caveat for anyone tempted to treat 9% as a number every future legalization state can bank on. The researchers' own conclusion is that adoption timing and regional labor market conditions matter a great deal -- a state legalizing cannabis into a tight labor market, or one without existing agricultural capacity to absorb new cultivation work, may see a much smaller bump, or none at all. Policymakers citing this study to project job growth in their own state should treat the 2017 and 2019 cohort results as a best-case scenario rather than a guaranteed baseline.

This isn't the first research to land on agriculture as a specific beneficiary of legalization, either. A 2023 NBER working paper from researchers at San Diego State University and Bentley University -- later published in the Journal of Population Economics in May 2025 -- found similar links between legalization and agricultural job growth in California, Colorado, and Oregon, three of the earliest and largest legal markets in the country. Two separate research efforts, using different methods and different time windows, arriving at the same general conclusion gives the farm-labor angle real credibility rather than treating it as a one-off statistical curiosity.

How This Fits the Bigger Cannabis Jobs Picture

Photo by Elsa Olofsson via Unsplash.

It's worth keeping this agricultural finding in proportion to the cannabis industry's overall job numbers. Vangst and Whitney Economics' U.S. Cannabis Jobs Report put direct marijuana industry employment at roughly 412,500 workers as of early 2026 -- a figure that spans cultivation, processing, retail, testing labs, and ancillary support roles across all the states with some form of legal market. The agricultural employment gains identified in the Texas Tech study represent a narrower slice of that total: specifically the farm-sector jobs tied to growing the plant, not the dispensary staff, budtenders, or compliance officers layered on top.

That distinction matters for how this data should actually be used. It's tempting to fold every cannabis-adjacent job statistic into one big number, but a state weighing legalization needs to know which part of that job growth is realistic for its own economy. A state without much existing farmland or agricultural workforce infrastructure isn't going to replicate California's or Colorado's cultivation-driven job gains just by legalizing retail sales.

Legal status, licensing structures, and labor rules for cannabis cultivation still vary considerably from state to state, and in some cases county to county within a state. Readers considering any of this from a business, employment, or policy angle should confirm current local law and licensing requirements rather than assuming what applies in Oregon or Colorado applies at home. What the combined research does suggest, though, is that cultivation licensing -- not retail alone -- appears to be the more meaningful driver of rural and agricultural job creation tied to legalization.

Put the Texas Tech findings next to the earlier NBER-linked research on California, Colorado, and Oregon, and a fairly consistent picture emerges: legal cannabis cultivation adds real agricultural jobs, but it does so by pulling more workers into the sector rather than raising what any individual worker gets paid. That's a more precise and more useful claim than the vague economic-development language that usually accompanies legalization campaigns, and it gives agricultural economies something concrete to point to beyond projected tax revenue.

It's also a finding with real limits attached. The cohort analysis makes clear that timing and local labor conditions did a lot of the work in producing the 2017 and 2019 gains -- states legalizing now, in a very different labor market and with a much more saturated national cannabis supply chain, shouldn't assume they'll see the same 9% lift just by passing similar legislation. The farm labor markets that absorbed cannabis cultivation in the late 2010s aren't necessarily the same ones a new legal state will be working with today.

Still, as more states debate legalization -- and as some existing markets debate expanding cultivation licensing -- this research shifts the conversation in a useful direction. Instead of arguing legalization in the abstract, states with agricultural economies now have a more specific case to make: cultivation licensing tends to create farm jobs, that effect shows up in more than one independent study, and it's distinct from the general cannabis employment figures that get cited most often. Whether that translates into policy that actually prioritizes cultivation licensing over retail expansion is a separate question, but the data at least gives that argument something to stand on.