How to force a conclusion from a study [updated]

So a study commissioned by the alcohol industry demonstrates that banning alcohol adverts will be a bad thing.  Consider me unsurprised and cynical.

I don’t know that the impact would necessarily be positive. I don’t know how much of an impact banning ads will have on consumption, but it does seem that the alcohol industry lobby, cunningly called the “Industry Association for Responsible Alcohol Use”, which is a bit of ridiculous PR spin, doesn’t seem shy to manipulate each part of the debate in their favour.

From the article:

“Qualitative and quantitative research by Econometrix shows that alcohol advertising is not a significant factor in determining consumption and has little or no effect on alcohol consumption per capita in South Africa,” said Jeffrey at the press briefing.

And then, some harder stats on the economic impact

  • Gross domestic product: An estimated reduction of 0.28% of GDP was to be expected.

  • Employment: 11 954 jobs was estimated to be cut.

  • Fiscus: Total tax income will decrease by R1 783-million; and

  • Trade: Exports would decrease by R225-million and imports would decrease by R304-million.”

There’s almost certainly an inconsistency here.  Let’s take that third bullet first. How is tax income decreasing? Since there is “no impact” on alcohol consumption from advertising, there will be no lower volumes of booze sold and therefore no loss of sin taxes. Ceteris paribus, alcohol industry companies will make higher profit, resulting in higher income taxes. Nowhere is there mention of higher profits to shareholders, greater returns on investment or reduction in prices. So I don’t see how the quantitative work here talks to the earlier conclusion that there won’t be a reduction in volumes consumed.

Again, it’s not that I alone possess perfect knowledge of what will happen to alcohol consumption, but it’s disingenuous to claim that it won’t lower consumption but then model the impact of lower consumption!

The other bullets are also interesting. Lower GDP and lower employment says something important and subtle or important and wrong about our economy. If consumers simply spend less, their savings will naturally increase and imports will decrease (based on the mix of imported versus locally produced alcohol). An enhanced savings rate is a key ambition of National Treasury. It’s not clear that this scenario is based for our economy at all. We haven’t yet talked to the higher profits to shareholders and investors, or more likely, price reductions in booze, resulting in an off-setting increase in consumption but possibly even greater savings, depending on the price elasticity of booze. Again, potentially further good.

Now in reality, it’s very likely that consumers will substitute at least a portion of their alcohol consumption with consumption of other goods and services. And what we’re saying is that whatever these things are, they will employ fewer people and contribute less to the economy that the production of alcohol. This just seems a stretch that really needs to be justified.

It’s also fairly clear that the analysis has not considered an increase in labour productivity through lower alcohol consumption, which will increase GDP.  Additionally, I see no analysis of the possible savings in healthcare costs associated with alcohol consumption (both medical things like heart disease and accidental death and injury).

The thing is, until the report is released in detail, all we have is an attention grabbing headline with no way of evaluating whether it’s actually carefully thought through and objectively evaluated, or whether it’s just massively disappointing work.

Update: Full report is available here

While the report is quite extensive and covers some of the points I’ve raised, it reads clearly as a piece of propaganda rather than objective research.