TL;DR
This paper tackles a crucial limitation in civic crowdfunding: the assumption that all participants have positive valuations and symmetric beliefs about a project. The authors introduce novel mechanisms that can handle agents who dislike a project (negative valuations) and those with uncertain or biased beliefs about its success, making crowdfunding more realistic and effective for diverse public goods.
What Problem Does This Paper Address?
Civic crowdfunding, where citizens collectively fund public projects like libraries or parks, has become a powerful tool. However, existing mechanisms largely assume that everyone involved wants the project to succeed and has a similar understanding of its likelihood of success. This is a significant oversimplification. For instance, some community members might actively oppose a proposed project (e.g., a new landfill in their neighborhood), meaning they have a negative valuation. Additionally, individuals might have varying degrees of confidence or skepticism about a project's success, leading to asymmetric beliefs. These scenarios can lead to strategic behavior, like free-riding, and prevent mechanisms from accurately reflecting collective preferences. This paper aims to bridge this gap by designing crowdfunding mechanisms that can accommodate both negative valuations and asymmetric beliefs, independently.
Methods
The authors propose two main extensions to existing civic crowdfunding mechanisms:
- Handling Negative Valuations: They adapt two established mechanisms, Provision Point Mechanism with Refund (PPR) and Provision Point Mechanism with Securities (PPS), to create new versions called PPRN and PPSN. The core idea is to introduce a second "market" or target, not just for project provision but also for its rejection. Agents can contribute to either market, and the project is provisioned or rejected based on which target is met first. The challenge is to design refund schemes that incentivize agents to contribute according to their true preferences, even if they oppose the project. They achieve this by carefully designing refund structures that make contributing to the "anti-provision" market more attractive for those with negative valuations.
- Handling Asymmetric Beliefs: For agents with positive valuations but varying beliefs about project success, they introduce a novel "Belief Based Reward" (BBR) scheme. This scheme uses a peer prediction mechanism (specifically, Robust Bayesian Truth Serum - RBTS) to incentivize agents to truthfully report their beliefs. The BBR rewards agents based on their reported beliefs and a shared budget. This is then integrated into modified versions of PPR and PPS, named PPRx and PPSx, to create a two-phase mechanism: a Belief Phase (BP) and a Contribution Phase (CP). The rewards from the BP influence contributions in the CP, encouraging agents with strong beliefs (positive or negative) to participate rather than free-ride.
Datasets
Not discussed in the paper.
Key Insights and Ideas
- Dual Markets for Negative Valuations: The introduction of a "rejection point" alongside the "provision point" is a clever way to model and incentivize participation from agents who oppose a project. This allows for a more nuanced aggregation of preferences.
- Belief-Based Rewards: The BBR scheme is a key innovation for handling asymmetric beliefs. By rewarding truthful belief reporting, it combats the free-riding problem that arises when agents are overly confident in a project's success.
- Independent Handling: The paper wisely tackles negative valuations and asymmetric beliefs separately, acknowledging that combining both simultaneously is a significant challenge that remains an open research question.
- Equilibrium Guarantees: The proposed mechanisms (PPRN, PPSN, PPRx, PPSx) are designed to ensure that at equilibrium, the project is provisioned or rejected based on the aggregated preferences and beliefs, leading to more robust outcomes.
Key Results
The paper presents theoretical results demonstrating that their proposed mechanisms (PPRN, PPSN, PPRx, and PPSx) achieve equilibrium outcomes. Specifically:
- For PPRN and PPSN, they prove that at equilibrium, either the provision point or the rejection point is reached, reflecting the collective sentiment of agents with negative valuations.
- For PPRx and PPSx, they show that agents with stronger beliefs towards the project's provision contribute more than those with weaker beliefs. This leads to the project being provisioned at equilibrium, driven by the incentivized contributions from agents with asymmetric beliefs.
- The analysis shows that for agents with positive valuations and asymmetric beliefs, their contributions are higher when they have a stronger belief in the project's success.
Limitations and Future Work
- Combined Negative Valuations and Asymmetric Beliefs: The authors explicitly state that their current mechanisms do not simultaneously handle both negative valuations and asymmetric beliefs. Combining these two aspects into a single, robust mechanism is identified as a significant challenge and a direction for future work.
- Budget Allocation: The discussion section notes that the Project Maker (PM) needs to carefully allocate budgets for the mechanisms. Inadequate budgets might disincentivize participation, while overly large budgets could lead to agents contributing just enough to receive refunds.
- Cost Function for PPS: The PPS mechanisms rely on specific properties of the cost function used in prediction markets. The paper assumes these properties hold but doesn't delve into the practicalities of designing such functions for real-world scenarios.
- Experimental Validation: While the paper provides theoretical proofs, it does not include empirical evaluations or simulations to test the mechanisms in practice.
The authors conclude by emphasizing the importance of future work in exploring the feasibility of combining negative preferences and asymmetric beliefs into a unified framework for civic crowdfunding.
Summary updated