It is sometimes startling to see how much public opinion on marijuana legalization has changed in my lifetime. Between 1995 and 2016, support for legalization jumped from 25 percent to 60 percent. There are weed-related ballot measures up for a vote in five states today. Increasingly, it’s more normal for a state to have some sort of legalization — even if just for medical purposes — than for the drug to be completely illegal. What do we know about the impacts of these changes? Honestly, not a lot that’s definitive, according to a 2016 review of the literature. For instance, marijuana use seems to be higher in states that have legalized than those that haven’t — but those states also had higher rates of use to begin with. Another example: Marijuana is the most common drug reported as being involved in motor vehicle accidents — but there’s also a lot of evidence that this is because risky drivers like to get high, rather than that the pot itself causes car accidents. Or this: Marijuana use has historically been associated with increased use of opioids — but there’s also evidence that medical marijuana could reduce the number of people taking opioids for pain.
The whole review is full of things like this, seemingly contradictory research that probably feels at least a little truthy either way. Ultimately, I think what we’re seeing here is an interesting example of how we don’t always have the answers to what changes in law will do before we make them. It’s in making the changes that we start to be able to collect enough data to know. Everything is an experiment.