Is alcohol consumption in Australia increasing or decreasing?

The consumption of alcohol is embedded in many western cultures, including Australia. While alcohol is typically consumed in social situations and has been shown to have benefits if consumed in moderation (Heath, 2007), excessive alcohol consumption and the misuse of alcohol can have serious and far-reaching effects (Schütze et al., 2011). Campaigns, interventions and policies aimed at reducing harmful alcohol consumption have been implemented. Although the media often highlight the negative aspects of alcohol consumption and the problems associated with alcohol misuse, contrary to popular belief, consumption in Australia seems to be decreasing.

The Australian Bureau of Statistics (ABS) and Australian Institute of Health and Welfare (AIHW) have both recently released reports highlighting the decrease in alcohol consumption figures. The ABS, using ‘sales’ data, has shown a decrease in alcohol available for consumption between 2009/10 and 2010/11 (ABS, 2012a). This was the first decrease since 2001–2002. This data source includes ‘excise data on alcohol produced for domestic consumption and data on imports’. Decreases were reported for beer and wine, while spirit consumption increased. These figures now equate to 10.0 litres of pure alcohol per Australian aged 15 years or older being consumed per year (ABS, 2012a) and compares to a high of 13.1 litres per person in 1974–1975 (ABS, 2011).

In the Australian Health Survey (2011–2012) (n=20,000+), 82.4% of adults reported consuming alcohol in the previous 12 months (ABS, 2012b), a decrease from 85.5% in 2007–2008 (ABS, 2009). There was also a decrease in the proportion of those aged 18 years and over consuming more than the recommended two standard drinks per day (from 20.9% in 2007–2008 to 19.5% in 2011–2012) (ABS, 2012b).

AIHW figures have shown a decrease in consumption as measured by a population-based survey (n=26,000+). Consumption decreased from 83.6% in 2004 to 82.9% in 2007 to 80.5% in 2010 (AIHW, 2011). It should be noted that a change in the method of data collection occurred over the later survey (from a combination of computer assisted telephone interviewing (CATI) and drop and collect in 2004 and 2007, to drop and collect only in 2010).

Recent work undertaken in South Australia has also reported decreases in alcohol consumption (Taylor et al., 2013). Using data collected over a 10-year period from a state-based (CATI) risk factor surveillance system, trends in prevalence estimates of alcohol consumption overall and in conjunction with a wide range of demographic variables, were decreasing. In total, there were 34 different statistically significant trends across the demographic and overall assessments.

So the question remains – is alcohol consumption in Australia decreasing? While ‘yes’ may be the answer, methodological considerations must be taken into consideration.

There are a myriad of ways to measure alcohol consumption including the population surveys and industry figures cited above. Within each of these data collections there is also a plethora of ways in which alcohol consumption can be assessed, centred on quantity and frequency. Compounding this issue in interpretation is the definition of a standard drink, with industry/policy perspectives markedly different to most consumers of alcohol.

In terms of frequency of alcohol consumption, one of the common problems relates to the time period or reference period that is used, with daily, weekly, monthly, last month and last year all regularly used. Recall is a problem with data collection based on self-report but especially with alcohol consumption (possibly related to the mind- and mood-altering effects of alcohol itself). Questions related to the frequency of consumption require details on all alcohol consumed in a day or a week, which can be time-consuming to respond to, and not always a reflection of the average consumption pattern.

Determining the quantity of alcohol consumed is complicated and made more difficult by the type of alcohol, with wine, beer and spirits having different volumes of pure alcohol per drink. Compounding this issue are drinks such as coolers and pre-mixed ‘alcopop’ drinks and different types of beer (mid-strength, full-strength and light).

The measurement of alcohol sales and shipments also has inherent problems, with homemade alcohol and illegal sales not counted (Dawson, 2003). Alcohol used for cooking, discarded, stored or cellared wine and duty-free purchases are additional miscounting issues (ABS, 2012a). This is the case with the ABS ‘sales’ figures provided previously. As stated by the ABS, the alcohol ‘sales’ figures should be shown as alcohol ‘available’ for consumption rather than consumption per se (ABS, 2012a). It has been reported that self-reported estimates of consumption are often as little as 50% of sales data (ABS, 2012c) but the known hard-to-reach populations (e.g. homeless, those living in institutions) are often not surveyed and are often those who are likely to consume higher quantities of alcohol. As such, survey responses should always be taken as under-estimations.

The mode of data collection is another issue. To link consumption with socio-economic, demographic and health outcomes generally requires population surveys. Also possible to assess using population surveys are alcohol-related attitudes, beliefs and knowledge. An additional advantage of population surveys is the possibility of controlling for confounders in any associations and the ability to track individuals or use population level data. CATI is common but, as with all survey modes, socially desirable responses can be an issue. In addition, with telephone interviewing, the use of prompt cards is not feasible. Consequently, the definition of a standard drink is more complex to describe than if a show card is used in a personal face-to-face interview or a description can be provided in a mail-out questionnaire.

Notwithstanding these methodological issues, which are inherent in any or all of these data sources, Australians are drinking less is the consistent story being reported. However, while the population may be drinking less there remain many problem areas and priority issues that need to be addressed, including alcohol consumption among Indigenous Australians, young Australians, pregnant women and those with alcohol use disorders, and these should be addressed as a priority. Further research into these areas would be of interest and relevance for mental health clinicians and policymakers.