Should social network structure be taken into account in elections?

If the social network structure among the voters in an election is known, how should this be taken into account by the voting rule? In this brief article, I argue, via the maximum likelihood approach to voting, that it is optimal to ignore the social network structure altogether—one person, one vote.