The social sub-optimality of competitive elections: comment