Faulty Studies

The BBC highlight a study by the British Medical Association, and demosntrate why doctors probably shouldn't do economics. For instance:

It did not recommend how large the tax hike should be, but pointed out a 10% rise could reduce alcohol-related deaths by nearly 30%.

Yet alcohol is one of the best and most well documented examples of a good that is price inelastic. Price elasticity is a measure of how greatly the demand for a good is changed by rises and falls in price. If a price elastic good, like restaurant meals, sees a price change of 10%, demand is likely to change by more than 10%. A price inelastic good, on the other hand, seeing a 10% change in price, would see a less than 10% change in demand.

Of course, very few consumers of alcohol die from it, those that do tend to be 'problem drinkers'. But for whom is alcohol the most price inelastic? Problem drinkers! Thus, a rise in the price of alcohol would reduce alcohol related deaths by much, much less than the change in price, less even than the change in demand for alcohol overall. Given the numbers we're dealing with, it probably wouldn't reduce alcohol related deaths by even one.

What it would do is increase the price of alcohol for all, which creates an interesting possibility. Poor people tend to consume more alcohol than rich people (in terms of volume, not value), so it might actually be that by raising the price of alcohol and thus making people poorer, policy makers are increasing rather than decreasing demand for alcohol.

The article points out that:

Alcohol was 65% more affordable in 2006 than in 1980

On its own this is pretty meaningless. What would be more interesting is to see how average weekly individual alcohol consumption has changed since 1980. Has it risen by close to 65%, more than 65%, or less than 65%? Personally I would only look at males, as changing attitudes amongst and towards females has probably resulted in a big increase to bring them more into line with men, independent of price changes.

The BBC article is also contradictory. It states:

Alcohol consumption has been rising steadily for the past 15 years

But at the bottom they have a graph showing otherwise:



I added a pair of horizontal lines so it's easier to see that alcohol consumption for both men and women was the same in 2006 as it was in 1997. That's 9 of the past 15 years where alcohol consumption has not "been rising steadily", so how can they justify the above statement?