In two weeks, I’m giving a FEE lecture on “Public Opinion for Libertarians.”  I’ll probably start by asking for a show of hands: “Who wants the depressing version of today’s lecture?  Okay, who wants the really depressing version?”  The sad truth is that the status quo is quite popular, and even moderate libertarian reforms like abolishing the minimum wage are persistently abhorrent to the overwhelming majority of the population.

Still, there’s one amazing exception: marijuana legalization.  For decades, public support was stuck at about one-in-four.  I’d decided legalization was hopeless, except maybe through a medical marijuana loophole.  But yesterday I learned that pro-legalization sentiment exploded during the Naughts.  Gallup’s been asking since 1970.  The latest news:

Gallup’s October Crime poll finds 44% of Americans in favor of making
marijuana legal and 54% opposed. U.S. public support for legalizing
marijuana was fixed in the 25% range from the late 1970s to the
mid-1990s, but acceptance jumped to 31% in 2000 and has continued to
grow throughout this decade.

What’s behind the change?  The biggest shifts between 2005 and 2009 are among the young and women.  The young were always relatively supportive, but now they’re absolutely so.  Women, on the other hand, were historically less supportive, but the gender gap has almost vanished. 

One of my colleagues jokingly credited Harold and Kumar.  Their first movie premiered in 2004, so the timing’s about right.  But then why didn’t Cheech and Chong get the job done thirty years ago?  At least to me, the sudden swing in public opinion remains a mystery.  Ideas?  And if persistent opposition can finally crack for marijuana, how about the minimum wage?

Update: In the comments, Gabriel writes:

it looks like cohort replacement based on a quick and dirty analysis of the GSS
variables “grass” “age” and “year”. cohorts born since 1950 are much
more tolerant of weed than older cohorts and over time they’ve become a
much greater proportion of the population.

Yes, but why did cohort replacement make so little difference between 1977 and 1996?