Abstract:We study the effect of public tallies on online elections, in a setting where voting is costly and voters are allowed to strategically time their votes. The strategic importance of choosing \emph{when} to vote arises when votes are public, such as in online event scheduling polls (e.g., Doodle), or in blockchain governance mechanisms. In particular, there is a tension between voting early to influence future votes and waiting to observe interim results and avoid voting costs if the outcome has already been decided.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: in the presence of voting costs and public vote - counting results, how do voters strategically choose the voting time in online elections to maximize their own interests or influence the election results. Specifically, the research focuses on the following points:
1. **Strategic nature of voting time**:
- Voters can choose when to vote, which is especially important when the voting cost is high and the voting results are public.
- Early voting can influence the voting decisions of other voters, while delaying voting allows for a decision to be made after observing some results, thus avoiding unnecessary voting costs.
2. **"Free - rider" effect**:
- Some voters may choose to wait and observe the voting situation of other voters to determine whether they need to vote, thus saving costs.
- This behavior may lead to the "last - minute" voting phenomenon, that is, all informed voters choose to vote at the last moment.
3. **Bandwagon Effect**:
- When the interim results are made public, voters who vote later are more likely to support the leading candidate. This phenomenon is called the "Bandwagon Effect".
4. **Application in governance mechanisms**:
- Especially in the blockchain governance mechanism, reaching consensus quickly is crucial. If the voting time is not properly selected, it may lead to decision - making delays, which in turn will affect the security and efficiency of the system.
To answer these questions, the paper constructs a new model, analyzes the behavior of informed voters and uninformed swing voters in different voting stages, and proves the existence of an equilibrium state where early voting and late voting co - exist under certain conditions. In addition, the paper also explores the influence of different parameters (such as voting costs) on these equilibrium states.
### Model overview
The paper considers a competitive election scenario with two candidates \( F \) and \( A \), and the voters are divided into three equally - sized groups:
- **Informed voters**: Support a specific candidate, with a voting cost \( c \), and obtain a utility of \( + 1 \) or \( - 1 \) depending on the victory or defeat of the candidate.
- **Uninformed swing voters**: Have no clear preference and vote according to the current interim vote - counting results, tending to support the leading candidate.
The election is carried out in rounds, and the interim vote - counting results are announced after each round. Informed voters can choose to vote immediately, wait for the next round, or abstain, while uninformed voters arrive and vote randomly according to the probability distribution \( \vec{q} \).
### Key contributions
1. **Existence proof**: Proves that under certain conditions, informed voters can adopt a mixed voting strategy between early and late voting.
2. **Equilibrium analysis**: Characterizes several particularly important equilibrium types, especially the "last - minute" voting behavior related to blockchain governance.
3. **Numerical simulation**: Evaluates the average cost thresholds in different types of equilibrium states through numerical simulation.
In summary, by introducing the strategic choice of voting time and public vote - counting results, this paper provides a new perspective for understanding voting behavior in online elections and a theoretical basis for designing more effective voting mechanisms.