That would seem to be the implication of a new study just published online by the journal Drug and Alcohol Dependence. But the study's authors aren't so sure.
The study measured drinking patterns in individuals who enrolled in treatment for marijuana dependence as part of a study designed to test different treatment methods. Participants greatly reduced their marijuana use, but 73 % also increased the number of days on which they drank alcohol by at least 10%. Most also increased the amount they drank on those drinking days. This seems like prima facie evidence of a substitution effect -- alcohol being substituted for marijuana.
The researchers, surprisingly, don't draw that conclusion, based on the fact that drinking behavior did not seem to change in proportion with marijuana use. Instead, they write, "We are left with a mystery."
It seems to me that, in the absence of another plausible cause, substitution of booze for marijuana still looks like the most likely explanation, though more research is absolutely needed. Given what's known about the much more serious health risks of alcohol as compared to marijuana, this ought to cause at least some unease regarding the 140,000-plus Americans forced into treatment for alleged marijuana problems by the criminal justice system each year.