NDARC Technical Report No. 195 (2004)
Effects of the heroin shortage
In 2001, a heroin shortage began in New South Wales that was accompanied by a reduction in heroin injecting among injecting drug use (IDU), in favour of an increase in cocaine and possibly methamphetamine injecting. Fatal and non-fatal heroin overdoses substantially decreased, more so in younger than older IDU and there were no offsetting increases in non-fatal overdoses involving cocaine, methamphetamine or benzodiazepines. There was a sustained decline in injecting drug use and a short-lived increase in property crime in NSW, followed by a sustained reduction in such offences. Treatment seeking among older heroin users and treatment compliance improved slightly during the heroin shortage but neither of these changes was sustained.
Were there changes in economic costs?
An obvious policy question is: what were the economic implications of these changes brought about by the reduction in heroin supply? Measuring the economic impact of the reduction in the availability of heroin use, while complex, is only part of the answer. We first need to agree on: (1) the program that produced the shortage and estimate; (2) what it cost by comparison with (3) with the costs of the program that existed status quo ante. We would also need to estimate the costs of the consequences of heroin and other drug use; and the benefits of the decreased availability of heroin.
There are a number of types of economic evaluations that could be used to undertake this task. These can, primarily be classified into two types of analysis: cost-effectiveness analyses and cost-benefit analyses. Other types of evaluation, such as cost-utility and cost-minimisation can be regarded as special cases of cost-effectiveness analysis.
Framework for an economic evaluation
The key steps in conducting an economic evaluation of the heroin shortage are identifying the programs that are being compared and clearly specifying the changes in the costs and outcomes. We also need to indicate from whose perspective the costs and benefits are being quantified. A full evaluation of social costs for instance, would include direct, indirect or intangible costs affecting a number of different groups including drug users, non-users, or government. The costs and outcomes included in the evaluation must reflect the chosen perspective.
The next steps are to identify the type of resources used in each alternative program, to measure them, and to cost the resources used. We also need an outcome measure that represents either a final or intermediate outcome from the resources used. Examples of possible outcome measures include lives (or life years) saved, cases prevented, drug seizures, or some other measure of decreased heroin availability.
Evaluating the heroin shortage from an economic perspective
In this section of the report, we outline one possible approach to measuring the costs and outcomes of achieving the observed decrease in heroin supply. We also discuss some of the issues involved in undertaking and the likely limitations of such an evaluation.
The major difficulty in evaluating the economic impact of the shortage is in identifying the program/s that produced the heroin shortage. An increased budget for drug law enforcement efforts to reduce the importation of heroin may have had an impact on heroin supply but other events may also have played a significant role. These other factors are difficult to identify and quantify; if they cannot be identified, any evaluation runs the risk of omitting these costs.
The choice of a perspective for the economic evaluation depends on what program we believe was responsible for the heroin shortage. A narrow departmental perspective (for instance, from the perspective of the Australian Federal Police) would exclude the costs of any resulting changes in drug use and treatment costs. A governmental perspective would be slightly broader, and a societal perspective would be broader still. If the direct causes of the Australian heroin shortage included events that occurred beyond Australia’s borders, then even a societal perspective would be too narrow.
A second challenge is identifying and estimating the costs of the alternative policy, that is, status quo in the absence of a heroin shortage.
Other major challenges in assessing the economic impact of the heroin shortage include: specifying the length of the shortage; identifying the full range of consequences of drug use; noting which of these consequences we can measure (such as emergency department admissions for the consequences of various drugs); choosing a level (eg: local or state) at which to measure these consequences; and noting whether the price of drugs has also changed.
After specifying the framework for an economic evaluation we found that we were not able to quantify all of the consequences of drug use. In some cases we were able to estimate some outcomes (for example, the costs of hospital admissions for overdose and drug-induced psychosis), but we are unable to make estimates for areas where the data were too sparse or not available (for example, risky injection practices and their consequences). Even those consequences that were measurable may be subject to bias so that comparisons could not be clearly made between the situation resulting from the heroin shortage and that occurring in the status quo ante. An example of this is the fact that existing health surveillance data are biased towards providing better information about harms caused by heroin rather than psycho stimulant use in Australia in a period in which the use of psycho stimulants increased.
The major obstacle to an economic analysis was the impossibility of estimating which programs were responsible for the reduction in heroin supply and what they cost. If we assume that drug law enforcement (DLE) was at least partially responsible for the reduction in heroin supply, then this is one of the costs to factor in. Even estimating the costs of drug law enforcement in Australia was problematic because of the complex relationships between state and Australian Government law enforcement agencies and the difficulty in apportioning the costs of each services activities to the reduction in heroin supply. We were also unable, to estimate: (1) the opportunity costs of DLE, that is, costs and benefits of alternative interventions that could have been implemented with the funding that DLE received; and (2) the costs and benefits of other factors that may have played a role in reducing heroin supply. For all these reasons we did not attempt an economic evaluation.
Citation: Shanahan, M., Degenhardt, L. and Hall, W. (2004) Estimating the economic consequences of a reduced heroin supply in Australia 2000-2003. Sydney: National Drug and Alcohol Research Centre.