Survey data need not underestimate alcohol consumption.

BACKGROUND: There are two main ways to assess alcohol consumption in a population: per capita estimates, usually derived from data on taxable alcohol available for consumption, and population-based surveys. Population-based survey estimates of alcohol consumption are often compared with estimates based on taxable alcohol available for consumption as a measure of validity. Discrepancies between these two measures occur, with the majority of population-based surveys substantially underestimating taxable alcohol available for consumption. METHODS: This article argues, however, that high proportions of taxable alcohol available for consumption can be accounted for by population-based surveys and reports a method of data collection for a national alcohol survey that has accounted for 94% of the taxable alcohol in New Zealand. RESULTS: The ability of the survey methodology to account for this proportion of taxable alcohol is likely due to the within-location beverage-specific alcohol consumption measures used in the survey, the process of recording the quantity of alcohol consumed, the use of a computer-assisted telephone interview system, and the population coverage achieved. CONCLUSIONS: Population-based surveys using the methodology outlined in this article may account for, if not exceed, the high proportions of taxable alcohol available for consumption estimates for some populations.