A state-run survey of 37,000 middle and high school students in Washington state finds that marijuana legalization there has had no effect on youngsters’ propensity to use the drug.

The Washington State Healthy Youth Survey found that the 2016 rate of marijuana use was basically unchanged since 2012, when the state voted to legalize marijuana for recreational use. In the survey, researchers used the measure of “monthly use,” asking students across all grade levels whether they’d used the drug within the past month.

The survey’s numbers show that neither the vote for legalization nor the opening of pot shops in 2014 have had any measurable effect on the rate of marijuana use among teenagers in the state.

Concerns about adolescent pot use have been one of the chief drivers of opposition to legalization campaigns in Washington, Colorado and elsewhere. Attorney General Jeff Sessions recently articulated the view when he told reporters that “I don’t think America is going to be a better place when people of all ages, and particularly young people, are smoking pot.”

The concern is that people who start using the drug at a young age are more likely to become addicted to it later. And like any other drug, marijuana use during adolescence – particularly heavy use – can have negative effects on children’s mental health and school performance.

But the data coming out of Washington and Colorado strongly suggest that those states’ legalization experiments, which began in earnest in 2014, are not causing any spike in use among teenagers. Teen marijuana use in Colorado decreased during 2014 and 2015, the most recent time period included in federal surveys. A separate survey run by the state showed rates of use among teenagers flat from 2013 to 2015, and down since 2011.

The picture in Washington has been a little more mixed. The federal survey showed no significant change in teenage marijuana use in the most recent period. But a separate study released last year did find evidence of a small uptick in marijuana use among eighth- and 10th-graders in the state.

But the Washington state findings in that study were derived from a national data set that wasn’t intended to produce representative samples at the state level, said Julia Dilley, the principal investigator on a separate federally funded study investigating the effects of marijuana legalization in Washington and Oregon.

That doesn’t make those earlier numbers incorrect, necessarily, but it does limit how accurate they can be for an individual state such as Washington. The state’s own survey, administered to tens of thousands of students and designed to be representative of the entire state, is “more likely to be accurate for reporting state estimates, in my opinion,” Dilley said.

All in all, these findings are good news for policymakers in California, Massachusetts and other states looking to start recreational programs. They suggest that legal weed has not had much of an effect on teenage drug use.

Even federal authorities, longtime skeptics of the merits of marijuana legalization, are starting to come around to the findings.

“We had predicted based on the changes in legalization, culture in the U.S. as well as decreasing perceptions among teenagers that marijuana was harmful [and] that [accessibility and use] would go up,” Nora Volkow, director of the National Institute on Drug Abuse, told U.S. News and World Report late last year. “But it hasn’t gone up.”