Friday, 23 April 2010

Alcohol - let's review

TVHE reminds us of a few things we ought remember in alcohol policy.

I'll summarize and add a few more.

First, LC reportedly recommends increasing the alcohol excise tax by 50%. Recall that the price elasticity of demand for heavy drinkers is -0.28; for moderate drinkers, -0.44. That means that if the price of a unit of alcohol goes up ten percent, moderate drinkers buy 4.4% less alcohol while heavy drinkers by 2.8% less. Moderate drinkers enjoy consuming alcohol, both because it's inherently enjoyable, and because there are health benefits. Heavy drinkers also enjoy consuming alcohol, but let's grant the healthists' premise that they drink more than they would were they fully rational and fully informed.

Now, without some tedious digging into national income accounts, I can't put precise numbers on it. But let's ball-park things. A half-litre of 5% beer has $0.63 in tax; low end beer sells for about $1.90 for a half-litre. So the tax currently is about a third of the price of low end beer. If the tax rises from 0.63 to $0.94 and there's full pass-through, tax rises to 42% of the price of beer, now priced at $2.21 per half-litre. The price of beer then rises by about 16%. At stated elasticities, moderate drinkers reduce their beer consumption by 7%; heavy drinkers, by 4.6%. If there were one harmful drinker for every moderate drinker, and if we normalize the costs to the moderate drinker of lost consumer surplus to 1, then the heavy drinker needs to enjoy benefits of at least 1.5 for the policy to pass cost benefit. If there are two moderate drinkers for every harmful drinker, the ratio rises to 3.

Recall further that the benefits to heavy drinkers from reducing consumption, even granting the assumption that they're irrational or ill informed, is only the area below the marginal cost curve and above the marginal benefit curve for the portion of consumption that's forgone with the tax increase.

An increase in the excise tax will reduce consumption. But, it will reduce consumption by more among moderate drinkers than among heavy drinkers, and will almost certainly fail cost-benefit analysis.

Second, recall that the estimates of the costs of harmful alcohol use are grossly overstated. The estimates use a method designed to inflate the costs of harmful use by counting all private costs as social costs, by attributing to alcohol all of the costs involved in activities to which alcohol was only inframarginal rather than decisive, by assuming that heavy drinkers enjoy zero benefits by their consumption, and by assuming that an alcohol-induced absent or unemployed worker can never be replaced by another worker. It's complete nonsense. A rough, but methodologically sound, accounting puts the external costs of alcohol consumption slightly below the tax revenue. Anybody wanting to cite BERL's shonky old number on the costs of harmful alcohol use would do well to re-read our critique, their reply, and our rejoinder.

If the government wants to tighten up alcohol regulation, I hope that they'll use an honest paternalistic argument rather than a dubious economic one.

Farrar in the NBR (ungated) faults the Law Commissioner:
The crusading was not restricted to New Zealand. Sir Geoffrey even went to Australia, and spoke at an Australian Drug Foundation conference. Not the Minister., ot the Director-General of Health, but the Law Commission President. The 68 year old Sir Geoffrey decried the fact that people put photos from parties up on Facebook. He wants an end to people getting drunk – an endeavour that would be as likely to succeed as prohibition succeeded in the 1930s.
I've only talked once with Sir Geoffrey. He didn't come across as a teetotaller wanting to eliminate everyone's fun but rather as placing very high weight on the costs imposed by a tiny minority of drinkers and next to no weight on the benefits of moderate drinking to moderate drinkers; consequently, reduced consumer surplus among moderate drinkers doesn't enter his social welfare function with the same weight as reduced harms from the few bad guys. I worry that his overestimating of the harms and discounting of the benefits leads him to advocate welfare-reducing policies. I hope Parliament is sensible enough to weight his recommendations appropriately.


  1. With all due respect Eric, I suspect this sort of cost/benefit argument will fall on deaf ears. It's simply not sexy. Your average voter out there is more easily swayed by emotive headline-grabbing sound bytes about the ills caused in society by "binge" drinkers. And since recent govts seem to be poll driven I'd be very surprised if govt doesn't adopt most of the LC's recommendations unless there is a huge public outcry against them.
    Or to paraphrase, I'd argue that govt wants to be seen to be doing something positive to confront these troubling issues, and it doesn't have to be an effective response, it just has to appease the great unwashed.
    Having said that, I agree with your assessment. I'm also becoming increasingly disheartened by the authoritarian nature of some of my fellow left-wingers. It's almost enough to make a man consider voting Act :)

  2. I don't much disagree. But the prior LC issues statement suggested that there were good economic arguments for raising the tax rate. I disagree rather strongly. There are plenty of paternalistic reasons, but not economic ones. If they're going to do silly things, they oughtn't claim that economic analysis says those things are sound.

  3. FYI: I'm currently recruiting for BLO - the Beer Liberation Organisation. You can take my Pale Ale from my cold, dead hand!