The case for drug legalization has always looked tight to me: make it all legal, unclog prisons of people who were either having fun or a health problem, leave the first alone if they don't bother their neighbours or run people over with their car, and get treatment for the others. In the process, take criminal networks out of the equation, avoid gang violence, and use the police for really serious stuff. On top, you get regulated products, avoiding problems of purity and possibly better conditions of consumption as people don't have to hide to do drugs.
Now, if the goal is to reduce human suffering, a new study of Florida by the US Centers for Disease Control suggests that things may not be that simple, even if suffering is defined, in the narrowest possible way, as avoiding death.
The study shows a death rate for prescription drugs of 13.4 per 100,000 people, compared with 3.4 per 100,000 for cocaine and heroin taken together (there was no significant number of marijuana-related death). Now, to put this in perspective, remember that the homicide rate for the US is about 5 per 100k, for Canada 1.6, and for France 1.3. In fact, up and until President Calderon launched his War on Drugs in Mexico, that country's homicide rate was hovering around 10 per 100k.
In other words, legal prescription drugs may be more lethal than illegal ones--even when both trafficking-related and overdoses are considered together.
Fortunately for legalizers, a number of caveats warrant mention: 1) marijuana, the drug most commonly and realistically contemplated for legalization, is not particularly dangerous, and possibly less damaging--though it depends on consumption rates--than alcohol or cigarettes; 2) the study only regards Florida and some factor may make things worse there (the study mentions that similar trends were found in Kentucky though and a Toronto Star investigation sees a similar explosion of oxycontin-related deaths in Ontario...); 3) during the period covered by the study (2003-2009), prescription drug deaths had increased by 84%, from 7.3 to 13.4 per 100k, i.e. current rates do not reflect some kind of stable status quo; and 4) much of the increase is driven by three drugs: oxycodone (+264% during that period), alprazolam (+ 233.8%) and benzodiazepines (+ 168%), which implies that measures specifically targeting this trio could have a large impact on the overall death rate for prescription drugs.
The case against legalization is not closed, in other words, but there is still much to mull about.
[Thanks to Freakonomics for the tip on the study...]