The avoidable costs of alcohol abuse in Australia and the potential benefits of effective policies to reduce the social costs of alcohol
7. Summary of results
Table of contents
- List of tables
- List of abbreviations
- Executive summary
- Acknowledgments
- 1. Introduction
- 2. A brief summary of methodological issues
- 3. The aggregate social costs of alcohol
4. Estimating the Feasible Minimum in practice- 5. The intervention choices
- 6. Effective interventions whose benefits cannot be evaluated
- 7. Summary of results
- 8. Conclusions
- 9. Research recommendations
- Appendix A. Methodological issues
- Reference
We proceed to summarise the results of the earlier analyses. Table 27 summarises the results of the exposure-based and Arcadian Normal approaches.
Table 27. Potential percentage reductions in per capita alcohol consumption and alcohol-attributable mortality
| Impact on | Potential percentage reduction |
|---|---|
| Alcohol consumption per capita | 40% |
| Male alcohol-attributable deaths | 47% |
| Female alcohol-attributable deaths | 49% |
| Total alcohol-attributable deaths | 48% |
Since the range of estimates of the impact of alcohol taxation is determined on a different basis than for the other interventions, the social cost estimates for alcohol taxation are presented in a separate table from those for the other interventions. Table 28 presents estimates of the impact of higher alcohol taxation, while Table 29 summarises the estimates of the potential benefits of the other interventions.
Table 28. Potential benefits of higher alcohol taxation in terms of reductions in social costs (expressed in 2004/05 prices), based on performance in Norway, United States and Italy
 | Based on per capita alcohol consumption in | ||
|---|---|---|---|
| Norway $m | United States $m | Italy $m |
| Potential reduction in Australian social costs | 5,940 | 2,190 | 2,810 |
Top of Page
Table 29. Potential benefits of policy interventions in terms of reductions in social costs (expressed in 2004/05 prices)
Intervention | Best estimate $m | Minimum $m | Maximum $m |
|---|---|---|---|
| Individual-based interventions | | | |
| Brief interventions | 5,830 | 3,490 | 8,160 |
| Population-based interventions | | | |
| Greater BAC enforcement | 940 | 780 | 1,090 |
| Reduced BAC level | 280 | 250 | 310 |
| Partial advertising and marketing controls—overall impact | 2,450 | 1,680 | 3,210 |
| Complete advertising and marketing controls—overall impact | 3,860 | 2,500 | 5,150 |
| Partial advertising and marketing controls—impact on road accidents | 310 | 160 | 470 |
| Complete advertising and marketing controls—impact on road accidents | 960 | 690 | 1,210 |
To put the estimates in the above two tables in context, Collins and Lapsley (2008a) estimate that the total social costs of alcohol abuse in the financial year 2004/5 were $15.3 billion.
Interpretation of the results in a study of this type should, for a variety of reasons, be undertaken only with great caution.
The translation of research results into potential reductions in social costs is necessarily imprecise. The determination of the upper and lower boundaries of the estimates is usually based on judgmental decisions by the researchers. Accordingly, in order to prevent a misleading impression of precision, the results presented here are rounded to the nearest $10m.
It transpires, however, that this lack of precision is less significant than it might appear at first glance. Estimates of the potential cost savings for the interventions analysed above are very high in absolute terms even when the lower-bound estimates are chosen. All the interventions considered above have been shown in the literature to be cost-effective.
Top of Page
A second reason for interpretational caution is that, although the potential cost savings are expressed in terms of constant price values (for example, in the present study 2004/05 dollars), these savings would usually only accrue over a period of several years. The expression in constant price terms appears to be the most sensible way to present the results given that, in the current state of knowledge, the time profile of the accrual of benefits is almost always uncertain. Without knowledge of this time profile, it is not possible to estimate the discounted present value of the future time stream of benefits.
The above summary of results presents the estimated benefits of each of the recommended interventions. It is not, however, possible to aggregate these results since there is substantial overlap between the impacts of the various interventions. For example, an increase in alcohol tax rates designed to reduce overall alcohol consumption would almost certainly have the incidental impact of reducing the rate of drink driving. An implication of this interdependence is that a specific intervention, when implemented on its own, would yield a higher individual rate of return than it would if implemented as part of a package of interventions, when the relative rates of return would be disaggregated.
However, there is no reason to believe that the benefits of these individual interventions would be completely mutually exclusive. The estimated total benefits of a package of interventions which were implemented simultaneously would not be the sum of the benefits of the individual interventions but it would certainly exceed the benefits resulting from only the highest yielding of these interventions.
To explain this point, assume a theoretical situation in which both partial advertising bans and a program of brief interventions were introduced, with the best estimate benefits being applied. The total benefits of this program would certainly be less than $8,280m ($2,450m plus $5,830m) but more than the benefits of brief interventions on their own ($5,830m).
Top of Page
