The basic public health approach to alcohol assumes that drinkers get no benefit from drinking. That assumption underlies most of the big “social cost of drinking” figures you hear. The social cost of cell phones would be huge too, if you assumed nobody enjoyed using them and then counted any spending on more than a bargain model as being a social cost.
Dunbar et al do some thinking about the benefits of social drinking:
Alcohol use has a long and ubiquitous history. Despite considerable research on the misuse of alcohol, no one has ever asked why it might have become universally adopted, although the conventional view assumes that its only benefit is hedonic. In contrast, we suggest that alcohol consumption was adopted because it has social benefits that relate both to health and social bonding. We combine data from a national survey with data from more detailed behavioural and observational studies to show that social drinkers have more friends on whom they can depend for emotional and other support, and feel more engaged with, and trusting of, their local community. Alcohol is known to trigger the endorphin system, and the social consumption of alcohol may thus have the same effect as the many other social activities such as laughter, singing and dancing that we use as a means of servicing and reinforcing social bonds.
These kinds of effects have previously been suggested as solution to what’s sometimes called the alcohol-income puzzle: drinkers earn more. Peters and Stringham suggested that drinking together builds social capital.
The Dunbar et al paper finds drinkers report higher happiness, better perceived worthwhileness of life, and better life satisfaction. They conclude:
The aim of this study was to ask whether there was any evidence that alcohol consumption has social benefits beyond a simple hedonic ‘high’ or anxiolytic effect. Because alcohol triggers an endorphin response, we hypothesised that it might increase the degree of social bonding (feelings of social closeness: see Dunbar and Shultz 2010; Roberts et al. 2014) and this might have implications for how happy and socially engaged people become. The evolutionary significance of this lies in the fact that our social networks provide us with the single most important buffer against mental and physical illness (House 2001; Fowler & Christakis 2008; Holt-Lunstad et al. 2010; Tilvis et al. 2012; Kim et al. 2016). We asked whether the frequency of social alcohol consumption (indexed by the frequency of drinking in pubs) or the type of venue (‘locals’ vs bars) influenced people’s social experiences and their wellbeing.
The survey data suggest that respondents who have a ‘local’ that they visit on a regular basis are more socially engaged, feel more contented in their lives, and are more likely to trust other members of their community. On some, but not all of our social measures, those who drink ‘casually’ were more socially engaged than those who didn’t drink at all, suggesting that there are independent effects due to being a drinker and having a regular drinking venue. …
Aside from direct health benefits that might arise from up-regulating the endorphin system, the principal benefit of the social consumption of alcohol may thus be that it acts much like the many other endorphin-stimulating activities that we use for social and community bonding (notably laughter, singing, dancing, and even storytelling: Dunbar et al. 2012; Tarr et al. 2015, 2016; Pearce et al. 2015; Dunbar et al. 2016). This is not to suggest that alcohol consumption is an adaptation in the formal biological sense, but rather that we discovered how it could be used to trigger a mechanism (the endorphin system) that is an adaptation for social bonding. Indeed, there is now a widespread view among archaeologists that cereal cultivation was first started in order to brew beer rather than to provide food (Dietrich et al. 2012; Hayden et al. 2013). We suggest that, like these other social bonding activities, the consumption of alcohol, once it had been discovered, came to be adopted as part of the complex set of activities and rituals associated with bonding our (by monkey and ape standards) large social communities.
They don’t establish causality here, and they acknowledge that (unmeasured) extroversion, for example, might lead people both to going to pubs more and to having better social networks.
But I think there’s more going on than drinking being an arbitrary coordination point for social bond reinforcement.