There’s a news item on The Drinks Business website reporting a study in the USA, presented at the AAAS meeting (one of the USA’s most important cross-disciplinary scientific conferences) showing a link between the number of off-trade alcohol outlets and violence in any particular neighbourhood.
The conclusion? We’re presumably supposed to think that this link represents some sort of causality. That is, the more outlets selling alcohol, the more violence. So if those outlets are closed down, would there be less violence?
Not necessarily. There’s nothing here to show causality. It could simply be that rougher neighbourhoods have higher rates of violence, and are also able to support more alcohol outlets. People living in nicer neighbourhoods with lower levels of violence simply might be choosing to shop for their alcohol in other ways than popping down the off-licence to pick up a six pack of beer.
Of course, there may be a real association between off-licence density and violence that is in fact causal. But we’re going beyond the data if we make that connection. And that’s bad science.