Arbiters solve disputes in order to collect arbitration fees. They are incentivized to vote honestly because, after a dispute is over, arbiters whose vote is not coherent with the group will lose some reputation, and receive less share of the arbitration fee [Schelling Point].
The distribution mechanism of arbitration fee and reputation updation after the dispute has been resolved is inspired by the Schelling Point Theory. Arbiters receive tokens proportionate to the consistency of their vote with others arbiters, and get their reputation score increased or decreased accordingly. The tokens are divided between the coherent parties proportionally to their weight. Parties are considered coherent if they voted as the majority.
To disincentivize non-participation and lack of revealing of one’s vote, the penalty for not revealing one’s vote, the penalty is twice as large than the penalty for voting incoherently. This incentivizes arbiters to always reveal their vote.
One could argue that those decisions being subjective, no Schelling Point would arise. In the informal experiments run by Thomas Schelling [Schelling, T. C. The strategy of conflict. Oxford University Press, 1960], it showed that in most situations a Schelling Point plebiscite by all parties does not exist. But Schelling found that some options were more likely to be chosen than others. Therefore even if a particularly obvious option does not exist, some options will be perceived as more likely to be chosen by others parties and will effectively be chosen. We cannot expect arbiters to be right 100% of the time. No arbitration procedure could ever achieve that. Sometimes, honest arbiters will lose reputation. But as long as overall they lose less than what they gain, the system will work with maximum efficiency.