Health losses from increased cannabis use disorder after German legalization were projected at 19 times greater than gains from reduced contamination.
Cannabis policy analysts; German health officials; legalization debate participants.
Health losses projected at 19x greater than contamination-reduction gains
What the researchers found
Projected 400,000-800,000 new users and ~2,300 additional severe mental health cases. QALY losses from CUD were 19x greater than gains from reduced contamination. Even a 1% consumption increase would produce net harm.
Why it matters
Germany legalized partly to reduce black market harms. This analysis suggests health costs from increased use may far outweigh contamination gains.
The numbers in context
400,000-800,000 projected new users. ~2,300 additional severe mental health cases. QALY losses 19x greater than gains. 1% increase sufficient for net harm.
How the study worked
Quantitative projection model balancing harm reduction from fewer contaminants against risks from increased consumption using QALY calculations.
What this study cannot tell us
Projection model with inherent uncertainty. Does not account for criminal justice benefits or destigmatization. Consumption estimates may be inaccurate.
How to read the evidence
Quantitative projection model; useful for debate but depends on consumption assumptions.
When this study was published
2025 study
The bigger picture
Challenges a key legalization argument: regulated products being safer doesn't help if total consumption increases substantially.
Questions still open
- Are consumption projections accurate? How should criminal justice savings be weighed? Could regulation limit the increase?
Common questions
Does this mean legalization is bad for health?
How many new users were projected?
Read the original research
A quantitative projection of the net health effects of cannabis legalization in Germany.
PloS one, 20(9), e0330879
Citation
Gandjour, Afschin. (2025). A quantitative projection of the net health effects of cannabis legalization in Germany.. PloS one, 20(9), e0330879. https://doi.org/10.1371/journal.pone.0330879