Academics Eric Crampton of the University of Canterbury and Matt Burgess of Victoria University have performed a huge public service (not commissioned by any outside party) by exposing the flaws in a study of the social costs of alcohol.

The study was undertaken for the Ministry of Health and ACC by Business and Economic Research Ltd (Berl). It was cited approvingly by Sir Geoffrey Palmer, SC, president of the Law Commission, in a speech relating to the liquor review which the commission is undertaking.

Berl calculated the annual social costs of alcohol to be around $4.8 billion. Crampton and Burgess show that this is a gross over-estimate.

Much of the critique is technical, focusing on the erroneous assumptions and methodology of the Berl study.

For the purposes of policy, the key issue is the external costs of alcohol consumption in the form of crime, publicly funded healthcare and road accidents.

Other costs, such as lost production and lower incomes sustained by drinkers, are "internalised"; that is to say, they are borne by consumers themselves.

Crampton and Burgess estimate that the policy-relevant net external costs amount to $146.3 million, less than 5 per cent of Berl's headline $4.8 billion figure.

Berl also ignores the benefits to consumers of alcohol consumption. These benefits are not the savings in public expenditure on healthcare due to moderate consumption, which have been documented in medical research. These are external benefits.

Rather, they are simply the benefits people derive from the enjoyment of alcohol, and are approximated by what they are willing to pay for alcohol products.

The Berl report should not have passed the smell test at the Law Commission. It is well known that some past studies have made similar mistakes.

Crampton and Burgess state that "The Berl report is wholly inadequate for use in assisting policy development".

A senior Treasury official has rightly commented that the Law Commission's reputation is at risk if it relies on it. Unless Berl can refute significant criticisms, the chief executives of the Ministry of Health and ACC should also be held accountable for such a poor use of $135,500 of taxpayers' money. Indeed they should be demanding their money back.

Many of the social trends associated with liquor consumption have improved with liberalisation starting 20 years ago.

Drinking habits are regarded by many New Zealanders as even better in some European countries which have fewer restrictions overall.

For example, many countries allow spirits to be sold in supermarkets. Why should this not be allowed here?

Liquor is in many ways not special. Hundreds of products - matches, detergents, electricity, pharmaceuticals, motor vehicles and firearms, for example - cause problems if misused.

Nevertheless, there are external social costs, such as drink driving, which give rise to legitimate concerns.

The challenge for policy is to target these problems with effective interventions (and enforcement of existing laws), not to penalise with regulations or taxes the vast majority of responsible drinkers.

As one commentator has noted, "Raising taxes on alcohol to prevent problem drinking is akin to raising the price of petrol to prevent people from speeding."

The last government's official tax review recommended in 2001 that excise taxes on alcohol should be scrapped now that a comprehensive GST is in place.

The Law Commission needs to engage with this analysis and follow the Generic Tax Policy Process for any recommendations on tax.

Similarly, it should follow the required Regulatory Impact Statement process for any recommendations on regulations in its forthcoming discussion paper.

That process requires a demonstration that the benefits of any recommendations or regulations exceed the costs. Competent analysis requires benefits and costs to be quantified, not just asserted, otherwise serious public policy errors could be made.

It is highly unlikely that proposals to restrict liquor outlets, for example, would meet a cost-benefit test.

Instead, the Law Commission should focus on ways of internalising the external costs of alcohol abuse. For example, why should those who injure themselves in alcohol-fuelled assaults or burglaries enjoy generous ACC benefits?

Many foreigners would regard such treatment as ludicrous. Will Sir Geoffrey Palmer, one of the "fathers" of ACC, be open-minded enough to look at such an obvious remedy?

Similarly, if we are willing to confiscate the vehicles of boy racers, why should we not confiscate the vehicles of serial drink drivers?

There are limits to what can be achieved with alcohol-specific interventions.

Research suggests that problems of destructive behaviour often go back to family, childhood and individual experiences.

The Regulatory Impact Statement framework requires "the range of feasible options" for meeting a public policy objective to be explored.

The Law Commission must range broadly and look at all options in a rigorous and dispassionate way. It should not compound the flawed Berl analysis with substandard work of its own.

* Roger Kerr (rkerr@nzbr.org.nz) is the executive director of the New Zealand Business Roundtable.