The costs of tobacco, alcohol and illicit drug abuse to Australian Society in 2004/05
8. Comparability with previous social cost estimates
The present authors have previously estimated the social costs of drug abuse in Australia for the calendar years 1998 and 1992, and for the financial year 1998/99. It is tempting to try to calculate the rate of change of these costs over time by comparison between the four sets of estimates. However, this is an exercise which should be approached with caution.
Estimates of the social costs of drug abuse can change for a variety of reasons. The main reasons are:
- Changes in the underlying available epidemiological information, as indicated by the attributable fractions (AFs). The AF for a particular illness or injury indicates the proportion of such cases with that condition in the population that can be causally attributed to consumption of the drug under consideration. The fraction has two components: the strength of the causal relationship between the drug consumption and the condition (the ‘relative risk’), and the prevalence of the consumption by the community of the drug under review. Changes in either or both components will affect the size of the AF, which will feed through to changes in the estimated costs. The estimated relative risk represents the current state of knowledge concerning the causal relationship between the drug consumed and the condition under review, and may change as new research evidence emerges. This may result from development of improved research knowledge rather than from a change in the actual cause of the condition under study. On the other hand, changes in prevalence rates will represent real changes in population exposure to the risk in question.
To illustrate this point, changes in the tobacco AFs for the present study are more a result of reductions in smoking prevalence than a change in information on smoking relative risk. On the other hand, a significant change in the relative risk information for alcohol is the major reason for the substantial increase in the estimated social costs of alcohol misuse.
- Changes in the scope of estimates. As an illustration, a major reason for the increased social costs estimated for 1998/99, compared with earlier years, was the comprehensive estimation for the first time of the costs of crime attributable to the consumption of alcohol or illicit drugs. With some relatively minor exceptions, the scope of the present study is similar to that of the 1998/99 research.
- Changes in the information available for the estimation of some categories of cost. One of these changes in the present study relates to availability of improved estimates of workplace absenteeism attributable to alcohol consumption. The result is that the estimated costs of alcohol-attributable workplace absenteeism are much higher than in the 1998/99 study.
- Changes in the costs and effectiveness of prevention and treatment programs. There are many instances of these types of changes. For example, there is now a much wider availability of both government-subsidised prescribed pharmaceuticals and over-the-counter nicotine therapies. Also, there has been increasing state expenditure on treatment programs, most of which have not been subject to economic evaluation. Rapid changes in the patterns of consumption of illicit drugs (see Table 6) have, as a public policy reaction, led to changes in prevention strategies.
- Changes in the general level of prices and costs. During the period from 1998/99 to 2004/05 the Australian National Accounts implicit price deflator for gross domestic product rose by 21 per cent (an indication of the change in the general level of prices and costs over this period). Comparison should be made between real changes (adjusting for the effects of inflation) rather than changes in aggregates expressed in current price terms.
If movements in estimated costs resulted from changes in one or more of
- the prevalence of drug consumption
- the costs or effectiveness of prevention and treatment
- the general level of prices and costs,
On the other hand, if the changed social cost estimates resulted from changes in one or more of
- information on relative risk
- the scope of the estimates
- information availability,
In relation to the comparability of the 2004/05 estimates with those for 1998/99, there has been a significant change in the relative risk information for alcohol. Generally, the scope of the estimates of the social costs of the three drug categories has not changed greatly but the availability of information has improved in some areas, particularly for alcohol-attributable absenteeism and ambulance usage.
In summary, the 1998/99 and 2004/05 estimates for tobacco and illicit drugs are, after taking account of the increase in the general price level, broadly comparable. Those for alcohol are not directly comparable (this issue is discussed further in Section 9 below).
Accordingly, the basis for comparison of the 1998/99 and 2004/05 estimates of the social costs of tobacco is provided in Table 49 below. The same comparison for illicit drugs is provided in Table 50. No meaningful comparison of alcohol costs is possible. To eliminate the effects of the general increase in prices over the period, the 1998/99 cost estimates in these two tables are adjusted to 2004/05 prices by application of the change over that period in the Australian National Accounts gross domestic product implicit price deflator.
Top of Page
Table 49, Comparison of constant price estimates of the social costs of tobacco, 1998/99 and 2004/05, at 2004/05 prices
Tobacco 1998/99 ($m)
Tobacco 2004/05 ($m)
Tobacco per cent change
The real social costs of tobacco abuse are estimated to have risen during the period 1998/99 to 2004/05 by 23.5 per cent (consisting of a 30.9 per cent increase in tangible costs and a 19.3 per cent increase in intangible costs). Although smoking prevalence has been falling steadily (the percentage of the population aged 14 years and over who are daily smokers falling from 21.8 per cent in 1998 to 17.4 per cent in 2004) and smoking-attributable mortality has also fallen very significantly (from 19,429 deaths in 1998/99 to 14,901 deaths in 2004/05), the lagged effects of past smoking both on healthcare and on the workforce have meant that the overall social costs of smoking continue to rise. However, the period between the decline in smoking prevalence in Australia and the subsequent decline in mortality appears to be much shorter than that indicated by models of the “tobacco epidemic” (see Lopez et al., 1994). This suggests that the social benefits of policies designed to reduce smoking prevalence are likely to be realised far sooner than indicated by the tobacco epidemic models. Earlier realisation of these benefits will result in a higher social rate of return from expenditures on anti-smoking programs.
As the lagged effects work their way through the system, and assuming that smoking prevalence continues to decline, real smoking costs (adjusting for the effects of inflation) should eventually fall very significantly.
The lag between the decline in smoking prevalence and the consequent reduction in social costs is an illustration of the fact that the extent to which costs are avoidable depends very much on the time period under consideration. The longer the time period, the greater will be the proportion of costs which are avoidable (until an irreducible minimum is reached).
Top of Page
Table 50, Comparison of constant price estimates of the social costs of illicit drugs, 1998/99 and 2004/05, at 2004/05 prices
Illicits 1998/99 ($m)
Illicits 2004/05 ($m)
Illicits per cent change
During the same period the real social costs of illicit drug use are estimated to have risen by 11.3 per cent (consisting of an 11.8 per cent increase in tangible costs and an 8.7 per cent increase in intangible costs). Between 1998/99 and 2004/05 the structure of the illicit drugs market changed very substantially, with a decline in the prevalence of use of some drugs as to whether the increase in the real social costs of the abuse of illicit drugs resulted from an increased level of usage of illicit drugs or from other causes.
Top of Page