Get your Tour de France fix
Cats have nine lives and 72-hour days

Staggering excess

Business Insider calls this stacked bar chart "staggering" (link).  Maybe they are referring to its complexity.


Is there a reason to include all the fine details? The details serve little purpose other than to shout at readers that there is a lot of data behind this chart. It is impossible to compare the different drugs on the individual harmful effects based on reading this chart. All we can see is the comparison of total harm. (By the way, I can't explain why anabolic steriods rated 10 would be sandwiched between khat and ectasy both rated 9.)

It turns out there is an easy way to fix this chart. We turn to the original Lancet paper which contained this chart, by David Nutt (link). The 16 categories of harm are nicely organized into a tree structure:


Instead of taking data from the right side of this tree, we can take data from the aggregated levels. The sacrifice in detail comes with a major benefit in clarity. In the original paper, Nutt produced a chart that aggregated everything to two levels:









It's amazing how much more we learn from this chart even though it has less data than the previous one. (I'd still remove the data labels since they are redundant when one has the axis labels.)

Similarly, we can plot a chart at the level of physical, psychological, social, etc., and it would still be much more readable than the "staggering" one.

PS. Apparently, David Nutt is a controversial character. See Wiki (link).


Feed You can follow this conversation by subscribing to the comment feed for this post.

Ian Watkins

David Nutt is not a controversial character - his sacking by the government was controversial. He is, to my knowledge, well respected in the drugs research field.
He advocates evidence-based policy, whereas politicians prefer policy-based evidence.
I guess combining these 2 now make him controversial!


Hi Kaiser,

thanks for your nice demonstration on that chart, and more generaly thanks for your website, which I have been following for a while now.

May I ask you if there's the data somewhere, as I'd like to play around with that one and can't find the source numbers behind the chart (I assumed that you have a source and didn't guess the numbers from the chart) ?



FK: I wasn't able to find the data. I didn't recreate this chart, I just pulled it from the original source. This one would be a nice exercise because the data is quite complex but it has some structure to play with.


Hi Kaiser,

i've done some digging, and it seems that most of the data come from the study "Development Of A Rational Scale to assess the harm of drugs of potential misuse" ("lancet 2007 ; 369 : 1047-53") of David Nutt, that can be found on the net.

I haven't had yet the time to extract the data from there, but if you're interested, i'll come back to you once it's done.


Thanks for the chart.

The comments to this entry are closed.