Recently, prediction markets have become very popular. I have a new concept that might support some cool experiments.
The idea isn’t originally mine; it comes from a very mind-bending paper. The author is one of the “grandmasters” of crypto—Ralph Merkle. In a radical proposal, he suggested using prediction markets to govern a country. Surprisingly, this paper was published in the journal Cryonics.
When I first read it, I thought the concept was interesting but impractical—zero feasibility. But upon re-reading recently, I realized that if the scenario isn’t limited to national governance, it could actually be a versatile, operational framework.
If you don’t remember who Merkle is, he’s one of the co-inventors of “asymmetric encryption” (public-private key cryptography) and the inventor of the “Merkle Tree.”
Every on-chain transaction depends on public-private keys. And each Bitcoin block is stamped with a Merkle root—used for efficiently proving that all transactions within the block are complete and unaltered.
Background on the Paper
Merkle was quite blunt: he believed that “one person, one vote” democracy is fundamentally flawed. This system forces most ordinary people—who often lack understanding of economics, political science, or sociology and are misled by media—to vote on extremely complex legislation.
This isn’t fair and inevitably leads to mediocre or poor decisions. The paper describes a governance machine (Merkle calls it DAO Democracy) that operates completely differently from traditional voting systems.
Traditional voting is “decide first, see results later” (vote for A, then bear the consequences). Merkle’s machine is “predict first, decide later.” Its operation relies on two core components:
1. The sole goal: Citizens’ “Happiness Index”
The system has a single, unchangeable ultimate goal (protected by DAO contracts), called the “Happiness Index.”
This index is determined by citizens’ post-facto ratings. Every year, all citizens rate the past year on a scale from 0 (worst) to 1 (best). The average of all scores becomes that year’s “Annual Happiness Index.”
This score is the system’s only metric of success.
2. The decision engine: Prediction Markets
With a single goal, decision-making becomes straightforward. When someone proposes a new law (e.g., “Should we build a new high-speed rail?”), instead of voting, the system opens two parallel prediction markets:
- Market A: Predicts, “If the law passes, what will the long-term Happiness Index be?”
- Market B: Predicts, “If nothing is done, what will the long-term Happiness Index be?”
The system then waits for the prediction period to end and compares the prices of A and B.
If Market A’s price exceeds Market B’s (say, 0.72), the system automatically approves the law. Otherwise, it vetoes it.
The Cleverness of the Design
This design is brilliant because it shifts decision-making from a biased, populist “political problem” to a rational, information-driven “prediction problem.”
In prediction markets, reckless bets (“I don’t care, I just hate high-speed rails!”) will lose money. Those who profit are the ones who most accurately predict whether the law will make the majority happier in the future.
It cleverly leverages “greed” to let rational voices, rather than the loudest, dominate decisions. Of course, the actual mechanism is more complex than I described—interested readers can check out the paper themselves.
Bringing It Back to Reality
Personally, I think using this system to govern a country is practically impossible.
Merkle himself acknowledged many challenges: how to prevent the system from pursuing high scores at absurd costs—like “giving everyone hallucinogens”—or how to handle laws with a 10% chance of causing apocalyptic outcomes.
Beyond technical hurdles, political friction makes it unlikely any existing regime would adopt such a scheme.
But if we look at narrower domains, with appropriate abstractions and carefully crafted conditions, I believe there could be feasible pathways.
A Simple Example
Community homeowners’ association decision-making: “Face-saving” members want to spend 100,000 yuan to build an unnecessary fountain. “Practical” members want to use that money to fix a leaking roof.
In traditional voting, this ends up being decided by “who has the loudest voice,” not “who’s right.”
Applying the “Merkle Machine” concept:
- Goal: Annual resident satisfaction.
- Proposals: Two options, with prediction markets to set their prices:
Market A: Predicts, “If we build the fountain, what will the satisfaction score be at year’s end?”
Market B: Predicts, “If we repair the roof, what will the satisfaction score be?”
Residents whose homes are leaking (the real experts) only have one vote in traditional voting. But in this system, they can confidently bet on Market B, knowing fixing the roof will improve satisfaction. If Market B’s predicted satisfaction is higher than Market A’s, the system automatically approves fixing the roof.
At year’s end, residents rate their satisfaction. Those whose homes are no longer leaking give high scores. The people who bet on fixing the roof win the bets, earning the money from those who bet on the fountain.
The actual implementation would be more complex, but the core idea is the same.
Essentially, this approach turns subjective, community-wide decisions—often fraught with bias—over to a transparent, market-driven prediction system. Democracy’s “one person, one vote” isn’t eliminated but transformed into a different form, enabling the entire mechanism to operate smoothly.
This concept could even evolve into a “governance-as-a-service” platform. The platform itself doesn’t set KPIs or policies; it provides neutral tools—like DAO contracts, prediction markets, and oracles.
Any organization, from homeowners’ associations to open-source communities, could register, input their specific KPIs (like “satisfaction” or “downloads”), and propose initiatives.
The platform’s role is to run the markets and deliver the optimal decision. It acts as a neutral referee, offering a plug-and-play decision-making machine for organizations facing tough, transparent choices.
