'In view of the existing quantomania, one may be forgiven for asserting that there is more to be said for rough estimates of the precise concept than precise estimates of economically irrelevant concepts.'Some of our calculations are necessarily a bit rough and can be improved upon; however, each one is our fair best approach given the data available to us. A somewhat rough estimate of the policy-relevant external costs we view as more useful than BERL's estimate of an economically irrelevant concept.
We considered only the costs that BERL attributes to alcohol alone. There's no particular reason that similar fisking couldn't apply to their costs of other drug use, or combined use of alcohol and other drugs, but we simply didn't have time to take on the whole thing before the Law Commission was due to report back with policy recommendations about alcohol use.
The biggest source of difference between our result and BERL's is that we focus on costs that are external to the drinker. Economists generally view the state as having a legitimate role in mitigating the external costs of individual behaviour. If I undertake an activity that imposes costs on you, I might well decide to do more of it than I would if I were bearing the costs. So, we say that these externalities have efficiency consequences: they distort behaviour. The government may well have an interest in preventing individuals from doing themselves harm even if there is no cost to anyone else; however, paternalistic policies of this sort really are best justified by arguments coming from outside of economics.
BERL argues that all consumption by men who drink more than two pints of beer per day (one pint for the ladies) is harmful and consequently has only costs. They add up these costs - including the costs of alcohol consumed - to get a very large figure. It's obvious that many who drink two or three pints of beer per day derive at least some enjoyment from their consumption.
If BERL had produced a cost-benefit analysis that weighed the benefits to drinkers, or if BERL had focused only on external costs, there would be much less distance between our result and theirs. But by bringing in privately-borne costs with no offsetting benefits, and deeming all such costs to be "social costs", BERL essentially has begged the question: they chose a method that could do nothing but produce a very large number. And so I'm reminded of the old joke with which I prefaced my earlier post.
In subsequent posts, I'll walk through exactly how we approached each cost item reported by BERL, in the order in which they're treated in our paper. On deck: the costs of lost output.