Beside, showing that drinkers earn more does not show that drinking is an overall economic boon as the first sentence implies. It could as easily mean that drinkers simply get a bigger slice of the pie.
They are not claiming causation. They are hypothesizing correlation between drinking and sociability/social network size, and causation from that to higher earnings.
"We hypothesize that drinking leads to higher earnings by increasing social capital."
The hypothesis seems to be that drinking makes people like you and that gets you better jobs.
As opposed to being social getting you better jobs and making you drink more, or having a better job allowing you to drink and be social, two alternate explanations which seem a lot more likely to me.
right, except for the bit where they suggest that we should encourage drinking to boost earnings. It is the absolute definition of the correlation/causation mistake. They also got a positive coefficient on the Divorced variable - I guess we should encourage divorces to boost the economy?