11. Democracy 2.0? Leveraging the power of the markets to align incentives
Today we will introduce the concept of Futarchy, which was originally proposed by economist Robin Hanson. Futarchy can be viewed as a form of government in which elected officials define measures of national well-being, and prediction markets are used to determine which policies will have the most positive effect.
Prediction markets
Prediction markets simply refer to an open marketplace where individuals can predict real-world events such as those relating to politics and sports i.e. betting on the outcome of a future event, such as the winner of the Champions League final or the US presidential election.
"Vote on values, but bet on beliefs"
This is a popular quote by Robin Hanson that encompasses the underlying principles on which futarchy is based. The beauty of prediction markets is that individuals possess a personal stake in deducing the likelihood of a future event arising, thus, ‘putting their money where their mouth is', creating a cost to getting an answer incorrect.
Historical performance of betting markets relative to polls
"Back in the era before scientific public opinion polls, election markets worked remarkably well, somehow the smart money knew" - Erikson & Wlezien (2012)
Several papers looked into the performance of betting markets compared to polls in determining the outcome of an election, with the majority finding the former to perform better, albeit not without its flaws. Even though the focus slowly shifted towards polls being predominantly used to predict elections, prediction markets are now making a comeback once again.
Erikson & Wlezien (2012) ended their paper by raising an interesting proposition, "given that historically election results were so accurate on the eve of an election, could we say the same thing about public opinion on topical issues?" This is in part because prediction markets have two critical advantages over polls.
Firstly, prediction markets are designed to align incentives amongst participants i.e. they are financially motivated to make accurate predictions, thus, providing a more reliable 'pulse check' on stakeholders’ confidence. Secondly, prediction markets are resistant to traditional forms of manipulation, so unlike with polls, bad actors are unable to control the narrative to try to swing an election their way.
Incorporation of futarchy within democracies
Futarchy is an ideologically neutral and untested form of government. It is one of many ideas that try to tackle the flaws currently present within our democracies. In this case, while our traditional understanding of democracy continues to focus on 'what we want', betting markets will guide us as to 'how to get it'.
.
Elected representatives will be responsible for formally defining and managing the measurement criteria, while market speculators will decide which policies are most likely to achieve the aforementioned criteria. Robin Hanson stated that for futarchy to work well, we need to accept the following 3 assumptions:
Futarchy step-by-step implementation
One of the most comprehensive analyses regarding futarchy was written by Vitalik Buterin ("Vitalik"), co-founder of Ethereum, on the Ethereum Foundation blog back in 2014. Vitalik summarises this governance model via 7 core pillars, in which a high-level example will be provided to simplify matters.
Scenario: The year is 2008, during the financial crisis and the US must decide whether banks should be bailed out or not.
What does futarchy offer over and above our existing systems?
Public trust in government officials and the 'experts' is at an all-time low. Technological advancements, such as those within blockchain technology facilitate electronic voting more cheaply and securely, maintaining an immutable record.
These solutions can help address voter apathy and partisan bias prevalent amongst most of the population. Futarchy derives several benefits including:
Decentralised autonomous organisations ("DAOs")?
These principles, while still in their infancy, have begun being applied within a few DAOs. For context, a DAO is a community-led entity with no central authority. At any point, the proposals, voting, and even the underlying code itself can be publicly audited.
DAOs serve to challenge the notions surrounding traditional organisations that all follow a hierarchical structure. While the concept of futarchy can be applied within traditional organisations, its true value is best expressed in the context of a DAO, which is dependent on the ability of the protocol to align incentives amongst its participants and users.?
One example relates to Gnosis, which in November 2020 announced its intention to move towards a futarchy-based model. They intended to incorporate prediction markets to guide decisions relating to the development, support, and governance of its ecosystem. Furthermore, every successful proposal is required to pass through three phases before being approved, with the only requirement being that the community members must hold at least 1 GNO to vote.
Martin Koppelmann, co-founder of Gnosis, highlighted that for the time being, decisions will be made purely by voting, with the voters only being informed by the prediction markets. Before more weighting (and ultimately decision-making power) can be trusted in these markets he added that:
While this idea did not take off as once hoped for, due to a lack of sufficient interest, it still serves as a good example of one potential implementation of futarchy within the context of DAOs.
Furthermore, over the past few months, work on the project is back underway, with the team exploring the use of AI agents who gather information and trade on prediction markets, with the majority of Gnosis Safe transactions on Gnosis being executed by these AI agents.
Reputation-based prediction markets
An interesting twist to prediction markets involves the use of reputation rather than currency. In such markets, the reputation score acts as a form of currency or credit, influencing the weighting of a participant’s predictions. These markets often utilise blockchain technology to ensure transparency and fairness in reputation scoring.
Such a model will help mitigate issues like groupthink and echo chambers, by holding individuals accountable for their predictions. They also encourage participants to think outside the box, with individuals achieving higher reputation scores being likely to have a track record of thoughtful forecasting leading to more nuanced predictions.
While this idea is still in its infancy and there aren’t many noteworthy examples of it being implemented, offerings such as that by Metaculus comprise a reputation-based online prediction solicitation and aggregation engine. The accuracy of such markets in comparison to those based on some form of monetary currency is yet to be sufficiently explored.
Challenges and other considerations
One of the main problems surrounding prediction markets is the limited number of participants, which potentially leads to the skewing of results. This brings about a lack of liquidity, especially those relating to niche topics, causing wider bid-ask spreads making it more challenging to execute trades at reasonable prices.
Another concern raised by Alexandros Marinos refers to Goodhart’s Law which states that when a measure becomes a target, it ceases to be a good measure. This is explained in the diagram below.
Another concern is that motivated parties, whether these be individuals, companies or nation-states could manipulate the prediction markets using a fraction of their ‘propaganda’ budgets if they were willing to lose their money for what they deem to be the ‘greater good’. This implies that prediction markets are subject to manipulation by rich and powerful entities who are financially incentivised to try to sway a decision one way.
A counterargument raised by Vitalik to this concern was that "as long as the market power of people willing to earn a profit by counteracting manipulation exceeds the market power of the manipulator, the honest participants will win, acquiring the funds from the manipulator".
Moreover, one must not discount the power of the masses, as already highlighted during the GameStop saga. Throughout that event, we witnessed the ability of the masses to mobilise against those whom they determined to be the 'bigger evil' i.e. the short-sellers. Such situations can also occur within a futarchy model, bringing about a heightened level of risk to the perpetrator.
Prediction markets are gaining momentum
Several crypto and non-crypto platforms such as PolyMarket, BetSwirl, SX and Kalshi are providing prediction markets. One also finds other options such as Manifold which do not use real money. The degree to which using non-real vs real money will impact the likelihood of more accurately predicting future events is yet to be seen.
While social media amplifies extreme opinion, such markets instead tend to accentuate educated opinion. Recently, Bill Ackman, a notorious billionaire hedge fund manager shared a few betting markets, included below, relating to the controversy surrounding Harvard and MIT. One example involved a ‘yes’ or ‘no’ choice as to whether:
Harvard application numbers will
University Presidents ousted before 2025
Another current example related to speculation as to whether Sam Altman would be returning as CEO of OpenAI. As can be seen below, on Polymarket, the ‘yes’ side had a 55% chance of being correct at that time. Through such markets, individuals wishing to understand an unfolding story will be able to get a sense in real-time, of how other people are pricing near-term outcomes.
Another example involved the Argentinian presidential election. On 19th November, 2023 prediction markets gave a 95% chance for Javier Milei to win the election. Consider that at this time most news organisations were still unwilling to announce even preliminary results.
This is a great example of prediction markets outperforming traditional media and polls, which both were unable to report anything of value about the race in a timely manner.
Below I included a prediction market for the US Republican Nominee on Polymarket, as of 24th December 2023, whereby if one bet $1,000 on Trump being chosen, they would be estimated to generate a return of $1,190 i.e. 19.04% return should they be correct. Naturally, the more liquidity present in the market the more representative the bets placed would be at predicting a future outcome.
Conclusion
In line with that discussed above, one of the best attributes underlying futarchy and prediction markets as a whole is that it has an in-built self-correcting mechanism that incorporates the wisdom of the crowds, improving over time.
This is because the people who are better at predicting outcomes will be financially incentivised to keep doing so. On the other hand, those that are more frequently wrong than they are right will get tired of losing money. Consider that this is one of many governance models that can be implemented, of which different variations will be explored in other posts.
DISCLAIMER
This is not medical, financial, or legal advice. Please consult a relevant professional prior to commencing anything outlined above, these are simply my own personal opinions.