Update 15 June: Folks interested in the full argument should hit the BERL tag - that will pull up more of the background work. And stay tuned: a far more thorough examination of the BERL report is coming out soon. My opinion of it has not improved.
Below is the op-ed I wrote for today's Christchurch Press. Unfortunately, they don't put their Perspectives pieces online; otherwise, I'd just link through to there. The version I sent is copied below, but I think they changed the headline.
The costs of everything and the value of nothing
We all know somebody who has a pint of beer with lunch, another pint at dinner, and a small tipple before bed. We might think he drinks a bit too much for his health, but would you believe me if I told you that he doesn’t get any enjoyment at all from his beer? I don’t just mean that, after accounting for what he pays for the drink, he really gets little out of it: instead, I’m telling you that he might as well be running his money through the shredder. I hope you’d tell me to blow it out my ear: our hypothetical drinker might not be leaving his liver in the best shape a couple decades down the track, but surely he gets at least some enjoyment from his daily drinks. Well, Geoff Palmer and the Law Commission, relying on a BERL report on the costs of alcohol, are trying to convince you of exactly the story I’m telling. And it’s nonsense.
The Law Commission was asked by the previous government to review our laws around alcohol usage. The terms of reference asked the Law Commission to deal explicitly with the need to ensure the appropriate balance between harm and consumer benefit. But the BERL report much cited by Palmer in his speech last week is the worst place to start if consumer benefits count for anything. Why? Because BERL explicitly assumes that any male drinking the equivalent of 2 pints plus 4 ounces of beer derives zero enjoyment from his tipple. 495,000 New Zealanders aged 15 and up, about 1 in 6, are considered harmful drinkers by BERL’s accounting. That sixth of the population is defined by BERL as receiving zero benefits from alcohol and are deemed to drink half of the liquor consumed in this country. It’s entirely plausible that a bit over two pints is enough for some people to start seeing some negative health effects from their drinking. But is it really plausible that they get zero enjoyment from those drinks? Hardly.
Why does this matter? Palmer cites BERL’s figure of 5.3 billion dollars in social costs of alcohol and says that includes the costs of crimes committed by drunks. He weighs those social costs against the current excise tax take for alcohol of about $800 million and argues for measures to reduce the difference. What he doesn’t say is that those costs also include the costs of producing and distributing half of the alcohol consumed in this country. A funny thing happens when a product is deemed, by assumption, to have zero benefits. Everything associated with that product then becomes a cost. Economists typically distinguish between private and social costs. Private costs are those borne by individuals through their own behaviour. External costs are those borne by others as a result of someone else’s activity. Technically, social costs are the sum of those two, but in policy discussions “social costs” are often used as shorthand for external costs. So you could be forgiven for thinking that the numbers Palmer cites represent some measure of costs imposed on others when it really mostly counts costs that heavy drinkers impose upon themselves.
If we look carefully through BERL’s numbers, about three quarters of their listed costs fall almost entirely on the drinkers themselves. Heavy drinkers are more likely to be absent from work; their forgone labour income (and income tax revenue for the government) then count as costs. However, heavy drinkers will earn less because they’re less productive: they bear those costs themselves. Heavy drinkers may have some associated health problems. To the extent that government pays for health care, taxpayers bear those costs. But, total excise taxes collected outweigh the health costs to “harmful” drinkers: those health costs are already defrayed by excise taxes on alcohol.
What are we left with after netting out all of the costs that drinkers bear themselves, either directly or via excise taxes on alcohol? Costs on victims of drunk drivers and costs of alcohol-related crime. Those are real external costs of alcohol abuse. It’s important that the Law Commission come up with some real proposals to alleviate these problems. But it’s also important that they weigh up the costs of their proposed solutions on the millions of New Zealanders who reasonably enjoy the moderate consumption of alcohol. Any per unit tax on alcohol aimed at recouping costs generated by a few hoons will always overcharge moderate drinkers while undercharging the folks who generate most of the problems. Geoff Palmer recommends increasing the current tax, but that will only further punish moderate users. Alcohol taxes, as noted in a rather nice 2002 Treasury study by Felicity Barker, approximately cover the external costs generated by heavy drinkers. External costs seem to have increased somewhat since that report. But as it isn’t feasible to implement a nonlinear tax structure (ie charging the hoons more for their drink but not so-punishing the rest of us), an optimal regime will likely combine a linear tax with penalties for the kinds of behaviours that cause harm to others. There’s little point in reducing the drink driving limit if few drivers in the 0.05 to 0.08 range cause problems and if few real penalties are imposed on those currently caught driving over the limit. The Press reported about a month ago on a seven-month pregnant woman caught drink driving at twice the legal limit with seven prior drink-driving convictions and fifteen convictions for driving while disqualified. It seems unlikely that having her license revoked serves as sufficient deterrent; lowering the threshold would similarly have no effect; and, raising the price of alcohol sufficiently to induce her and others like her to forbear would impose massive costs on the rest of us.
The Law Commission’s remit demands that it appropriately balance harms caused by problem drinkers against consumer benefits. It’s not encouraging that Palmer has begun by citing a study that derives inflated cost estimates by assuming that our hypothetical just-over-two-pint-a-day friend derives zero enjoyment from his drink. The BERL report cannot form the basis for any reasonable consideration of the welfare effects of alcohol policy. We cannot derive sensible welfare or policy conclusions from studies that consider the costs of everything but the value of nothing.
Dr. Eric Crampton is Senior Lecturer in Economics at Canterbury. He blogs at http://offsettingbehaviour.blogspot.com