4.2  Total Value of Externality

The externalities of alcohol consumption include tangible and intangible costs. Devlin et al (1997) have estimated the direct cost of the resource use associated with alcohol consumption. These estimates can be used as a proxy for the value of the tangible external costs. As discussed, as quantification is difficult for externalities that do not have observable monetary values this paper does not include an estimate of all possible externalities. Instead this paper uses the Devlin et al estimate as an estimate of the minimum plausible bound of the total externality, focusing on tangible costs.

The Devlin et al (1997) estimate calculates the net social cost of alcohol consumption in New Zealand in $1991. The external costs estimated were mainly fiscal costs, being: hospital resources, ACC, Ministry of Transport, police, penal institutions, community sentencing, periodic sentencing and court costs. The social costs of lost production were also estimated.

The health costs estimated were the health care resource use from alcohol related injury and illness. It was assumed alcohol abusers did not use more primary health care resources or require more pharmaceutical products. ACC expenditure was estimated both including ($56 million) and excluding ($29 million) payments for non-economic loss - compensation for impairment, pain and suffering, mental suffering, disfigurement, loss of capacity to enjoy life and other points of significance. Much of the payment for non-economic loss would accrue to the drinker and thus be internal. For this reason only 50% of this cost is included in the upper range of the estimate of external costs.

Production losses calculated arose from: excess unemployment of drinkers, reduced efficiency of drinkers, excess sick leave and absenteeism, premature death of drinkers and losses due to incarcerations. Over 85% of these costs come from excess unemployment and reduced efficiency. Assume that 5% of these costs are external and that most of the cost is borne by the individual, e.g through lower wages. A conservative assumption is used as this paper seeks to estimate the minimum plausible bound of the externality. The estimated range for external and internal production costs was $703,583,000-$3,389,005,000. This is large in relation to the estimated range for health care and justice service costs ($345,194,000-$592,858,000), thus the conservative assumption does materially alter the results.

The estimate of the external cost of use of public services will overestimate the external costs of alcohol consumption to the extent that cost recovery was not taken into account. However, as not all the costs of alcohol consumption are estimated this is a conservative estimate.

Significant external costs not included in this estimate include: the costs resulting from property damage that cannot be recouped through the court system; the cost to the government of care for victims of alcohol related incidents, such as the care of children of alcoholics; the cost of third party loss of life and injury from alcohol related incidents that were not compensated through ACC and the costs of alcohol damage to the foetus. The Land Transport Safety Authority reports that for 1999 drunk drivers were responsible for killing 40 of their own passengers and 18 other drivers, passengers, cyclists and pedestrians.[33]

The following table presents the costs as estimated by Devlin et al:

Incorporating the above assumptions, the Devlin et al estimate puts the external tangible costs at $376 - $772 million per year ($1991). The range results from using different assumptions as to the alcohol attribution factor for criminal activity (.29-.68) which was used to calculate production losses due to incarceration and the costs associated with the criminal justice system; the use of different discount rates in calculating production losses (5%-10%) from the premature death of abusers; the use of different excess unemployment rates used to calculate production losses from the excess unemployment of abusers (10%-30%) and the inclusion of non-economic costs associated with ACC.

The external costs are dominated by the costs of the police and the criminal justice system. Even under the conservative assumption of the role of alcohol in criminal offending these costs make up around 65% of costs. Further, the cost of accidental falls and road traffic fatalities make up around 47% of the cost of hospital treatment.[34] This suggests that acute alcohol consumption gives rise to significant costs.

The external benefits estimated by Devlin et al include the reduced risk of: coronary heart disease, cardiac dysrhythmias, heart failure, cholelithiasis and the production losses averted due to the beneficial effect of alcohol on health. The reduced use of hospital services due to these benefits was estimated to be $27 million. As the incidence of fatal ischaemic stroke is about one third of the incidence of fatal CHD, I estimate the hospital resource savings from the reduced risk of ischaemic stroke would be around $8 million per year. External production losses averted due to the reduced incidence of CHD, assuming 5% of the production cost is external, are between $1.3 and $1.9 million. Thus the estimate of the external benefit is $36 - $37 million per year. The external benefits of alcohol consumption are small relative to the external costs.[35]

These estimates suggest that the net external cost of alcohol consumption is large, being significantly more than $340 million annually ($1991) and most likely more than $735 million annually ($1991). This is $385 million - $831 million ($1999).

4.3  Consumption Patterns

The distribution of alcohol consumption is highly skewed. A small number of people consume very large amounts of alcohol. The vast majority of drinkers consume a relatively small average volume of alcohol. The 1995 National Survey of drinking in New Zealand estimated that the top 5% of drinkers consumed 34% of the alcohol, the top 10% of drinkers consumed 49% of the alcohol, the top 25% of drinkers consumed 74% of the alcohol and the top 50% of drinkers consumed 93% of the alcohol.

This distribution of alcohol consumption implies that a small percentage of consumers generate externalities related to adverse long-term health effects. However the costs of binge drinking, rather than the costs resulting from long-term alcohol use, make up the major portion of external costs. The National Survey indicated that a significant portion of alcohol is consumed in single sittings. The survey asked males how frequently they consumed six or more 15 ml drinks in a drinking session and females how often they consumed four or more 15ml drinks in a drinking session.[36] For males, 41% said at least monthly and 21% at least weekly. For females, 22% said at least monthly and 8% at least weekly. This implies that a large number of consumers consume alcohol in a way that carries a significant risk of generating external costs.

4.4  The Rate of the Uniform Excise

We can use information on the incidence of damage from alcohol and the distribution of alcohol consumption to make inferences about the relationship between the total externality and the amount of revenue that would be raised by an optimal uniform tax in the case of alcohol.

Consider figure 3 in the case of alcohol. As illustrated, the expected external costs of alcohol increase with the average level of consumption for both long term and acute consumption.[37] The optimal uniform tax has been imposed on units of alcohol (dotted line). Assuming elasticities are roughly equal across consumers at different consumption levels[38], heavy drinkers will have a higher weighting in the optimal uniform tax.[39] This will increase the tax relative to the situation of equal weighting. Consumers have the same marginal externality function but different demand curves.

Figure 3: Revenue and Externality    Alcohol

 

For heavy drinkers the revenue gained from their consumption is likely to be less than their externality. The number of heavy consumers is relatively small, given the skewed nature of the distribution of alcohol consumption. However the externality they generate and the revenue they pay is relatively large. For a number of binge drinkers the revenue raised from their consumption will also be lower then the externality they generate. The majority of consumers consume low average levels of alcohol. Thus for most consumers the revenue collected will exceed the externality they impose. The lower weighting of lighter consumers in the calculation of the optimal excise, all else equal, will increase the excess of revenue over externality for lighter consumers and decreases the excess of externality over revenue for heavy consumers.

Given that for the vast majority of consumers under an optimal uniform tax the revenue collected from them will exceed the externality they generate it is reasonable to assume that the revenue from the tax should be at least as much as the total externality. If the revenue is around the same value as the externality consumers will pay the social costs of their consumption “on average”. A small number of heavy consumers will pay the highest tax per person but this will not be enough to cover the externalities they create. Similarly a small number of acute drinkers will pay less in tax than they generate as externalities. Lighter consumers will pay tax in excess of their externality. However, this cost will be spread over a large number of people. The lack of information makes it difficult to be more precise than this.

In 1999/00 the amount of revenue collected from the tax on alcohol was $580 million ($1999/00 including GST on the excise and the duty collected on imported alcohol).[40] This is near the mid-point of the estimated bound of the external tangible costs of alcohol ($608 million). Thus the current rate of excise tax can be justified on externality grounds. As total external costs include a number of other intangible costs the total external costs are likely to be significantly more than the revenue collected. A case to increase the excise would need to show that these costs are significant and that tax was the best tool to address these costs.