Monday 16 November 2009

Nonsense upon stilts

At Offsetting Behaviour Eric Crampton posts an article he wrote for NORML on the New Zealand Drug Harm Index. See here. Eric opens by saying,
If the New Zealand Drug Harm Index is nonsense, then Detective Senior Sergeant Scott McGill’s claim that recent police action against marijuana prevented $379 million in social harm is nonsense upon stilts. Social cost measures provided in reports like the “New Zealand Drug Harm Index” and “Costs of Harmful Alcohol and Other Drug Use” are very good at providing very big numbers related to harms but are not terribly useful for public policy purposes. We’ll here contrast the method used in these kinds of reports with proper economic analysis.
There is a real issue here, in that some of the economics underlying these types of "social harm" indexes is just plan bad. When discussing the infamous BERL report on the cost of alcohol Eric writes,
In sum, social cost measures like those derived in BERL’s report of the costs of harmful alcohol use differ substantially from economic notions of social costs and consequently are of very limited policy relevance.
Eric then goes on to discuss the Drug Harms Index. He points out that there are common features between the alcohol cost report and the drug harm report. He concludes that
In sum, the Drug Harms Index mixes together costs of drug use with costs of prohibition and costs borne by drug users with costs imposed by drug users on others so opaquely that it is difficult to say to what possible policy use their numbers can be put. This is truly disappointing; a proper measure that provided the different categories of harms could be very useful in helping to decide which drugs would best quickly be legalized and which might be of lower priority. If it were the case that the vast majority of the costs of marijuana use were really costs associated with prohibition and if external costs were relatively low, a very good case could be made for legalization combined with an excise tax to defray those external costs. If it were the case that methamphetamine were associated with very high relative costs, with most of those costs being due to drug use rather than to prohibition, and with a greater proportion of those costs falling on external parties, that would provide a reason to legalize marijuana while keeping P prohibited. But we simply cannot say anything useful from the numbers as presented.
and
It seems that the main use of the report is in agitprop: providing big scary numbers with a sciency feel that can help to justify ongoing prohibitionist policies. After all, if the costs of drugs are over a billion dollars, we’d be crazy to even consider legalization, wouldn’t we? Well, maybe not if those costs accrue more to prohibition than to drug use. Unfortunately, from the report as it stands, we just cannot tell.
Bad economics can be used to justify almost anything, which is why reports such as those on the social costs of alcohol and the New Zealand Drug Harm Index need to be examined very carefully. People should not take them at face value. Read all of Eric's piece and you will come away with a healthy scepticism towards such measures.

No comments: