Suppose you and your spouse are both voters with opposite political views. As a result, on election day you both show up to the ballot box and simply cancel each other out. Regardless of how you both valued the alternative outcomes, if one always cancels out the other while each incurring the cost of voting, this is a net social loss. If you and your spouse could simply agree to not vote, you would not incur said cost and improve social welfare.
If you vote, or plan to vote in the next election, you could similarly improve social welfare by finding another voter who would otherwise cancel your vote out and find a way to similarly ensure that you both abstain from voting in the next election. Perhaps you could both unregister to vote at the same time or maybe write a contract that requires your political opposite to compensate you if they register to vote.
If these transaction costs are lower than the costs of voting, you have yourself a net increase in social welfare. The less likely you are to be the median voter, the greater the social savings will be. In other words, the further in advance you know who you're voting for, the more benefit you'll confer to the world by just staying home with your new friend.
Perhaps a entrepreneur can arrange a website to lower the transaction costs and help political yangs find their yins.