The social capital penalty paid by teetotallers.

Social capital is important and helps protect health and reduce loneliness. Governments worldwide are pursuing policies to reduce the amount of alcohol consumed to protect public health but alcohol consumption remains a prevalent feature of social interaction in the UK. Previous studies have identified a strong relationship between alcohol and social capital which varies in direction depending on the dimension of social capital studied. Using a large nationally representative longitudinal dataset for the UK, we apply an outcome-wide longitudinal design for causal inference, adjusting for covariates, as well as lagged values of outcome and exposure, to investigate if drinking less alcohol or not drinking alcohol at all is related to five binary social capital outcomes: socialising, being active in an organization, feeling lonely, number of close friends, and a bridging social capital score. We use two drinking exposures, binary drinker status, and categorised drinking frequency. We find that not drinking alcohol is negatively associated with socialising. Analysis using the frequency of drinking alcohol exposure finds drinking alcohol monthly or less is negatively associated with being active in an organisation. We find little evidence of any relationship between drinking alcohol and feelings of loneliness, number of friends and bridging social capital. Our results suggest that non-drinkers face barriers to some forms of social capital including socialising, which could be due to alcohol being a social norm in the UK. However, our results also suggest that high-frequency drinkers can reduce their drinking with minimal impact on their social capital. Our findings suggest more needs to be done to make socialising easier for non-drinkers. Furthermore, our findings support the implementation of policies to reduce high-frequency drinking.