Cannabis revenues as temptation goods

Rob MacCounStanford Law School

Jon’s talk: Some key points

National legalization, probably on the alcohol model, is very likely

Long-run consumption may (at least) double

Prices matter

Quantity consumed matters more than how many consume

Lots of policy levers – are we using them to learn what works?

5/18/2016MacCoun 2

Nat’l legalization inevitable?

Still 46 states to go (45.5 since DC halfway there)

President Cruz might put a stop to it. So might President Ryan. Even President Clinton. Or Republican-led House and Senate.

CO and WA and OR and AK could “blow it”

If consumption or harms start “spiraling out of control”

Headline: “Carload of Utah high school football players in fatal crash after drug holiday”

5/18/2016MacCoun 3

5/18/2016MacCoun 4

True, but boomer’s support

dropped a lot in 1980-1990

5/18/2016MacCoun 5

Easing budget pressures could reduce push for pot revenues

5/18/2016MacCoun 6

Effects in CO, WA?Caveat emptor

If you want to dismiss opponents’ claims, ought to hold yourself to same standard of evidence

May take many years to assess effects

Retailers have many steps to get licensed

Legal production has to reach scale

Short term “nostalgia” and curiosity purchases

Prices may fall over time (perhaps a lot)

Dutch marijuana use told different stories in 1979 vs. 1992 vs. 2008

5/18/2016MacCoun 7

On doubling

With respect to prevalence, we know what a doubling looks like: 1979

My students are usually stunned to learn this.

With respect to consumption, would be entirely new territory

My students already have seen a lot more daily and heavy use than my cohort saw.

Jon suggests we might see 40 billion additional hours of intoxication

5/18/2016MacCoun 8

Fundamental tradeoff

Legalization reduces average harm per dose (medical, crime, etc.)

Legalization increases number of doses consumed

Net effect (harm/dose x doses) is unknown

5/18/2016MacCoun 9

Legalization and average harm

In Drug War Heresies, Reuter and I predicted that after legalization, the average user would be a less risky user.

Influx of users who wouldn’t use because it was illegal.

No longer sure we were right.

We were focused on prevalence (who uses)

Might be swamped by increasing riskiness of existing users and new products (vaping, dabs, edibles)

5/18/2016MacCoun 10

5/18/2016MacCoun 11

What strains with the High Times Cannabis Cup?

Cannabis Cup

5/18/2016MacCoun 12

All finalists exceed federal potency estimates

Trend toward higher THC

1st place winners have higher THC and lower CBD

SimSemilla model

Users (count)• Occasional• Monthly• Weekly• Daily

Grams/day(per type)

Days-useddistribution

Grams/yeardistribution

Harm incidents per year:• Overall• by User type• Gini & skew

Grams/daydistribution

p(harm|dose)

Dose-harmfunction:• Ceiling• Shape• Threshold

5/18/2016MacCoun 13

5/18/2016MacCoun 14