In 2020, Shayne Coplan launched Polymarket as a decentralized platform for trading on real-world outcomes, allowing users to wager on political elections, economic data, geopolitical conflicts, and scientific breakthroughs. The mechanism is straightforward: participants buy and sell shares representing yes or no outcomes, with prices converging toward probability estimates. When an event resolves, smart contracts settle all trades in USDC at the winning outcome’s price. The platform operates on Polygon, a Layer-2 network, offering near-zero transaction fees and rapid settlement. What began as a tool for hedging and arbitrage has become a genuine test of a decades-old economic hypothesis: that markets aggregate dispersed information more reliably than institutions, and that prices can serve as a superior mechanism for collective decision-making.

The philosophical appeal is considerable. Markets reduce complex judgments to a single number—a price—that reflects the aggregated beliefs of all traders. No committee, expert panel, or democratic vote needs to debate whether inflation will exceed 4 percent next quarter or whether a particular geopolitical escalation is likely. Instead, the price embedded in a prediction contract embodies a consensus that emerges without explicit coordination. This idea has deep intellectual roots in twentieth-century economics, championed most forcefully by Friedrich Hayek, and it has inspired a small but growing movement around «futarchy»—governance systems in which policy decisions are made by allowing voters to bet on outcomes. Yet the question remains unanswered: could Polymarket-style markets actually replace, inform, or constrain democratic institutions? And what does the existing evidence suggest about the limits of price-based truth-finding?

The Knowledge Problem and Hayek’s Defense of Markets

Friedrich Hayek’s central insight in «The Use of Knowledge in Society» was deceptively simple: no single decision-maker, no matter how intelligent or well-informed, can possess the totality of knowledge dispersed across millions of individuals. A farmer knows the specific conditions of his soil and weather; a merchant knows local demand and customer preferences; a factory manager understands production bottlenecks. This knowledge is tacit, distributed, and constantly changing. Centralized planning—or centralized governance—cannot aggregate it efficiently because it must be transmitted, distilled, and summarized at every step.

Markets solve this problem through price signals. When wheat becomes scarce, its price rises. No central authority needs to order farmers to plant more or consumers to ration. Individual incentives align automatically. A farmer sees profit opportunity and responds; a consumer sees rising costs and adjusts consumption. The price mechanism encodes dispersed knowledge into a single variable that everyone can observe and act upon. Hayek argued that this distributed coordination was more efficient than any alternative—not because markets are perfectly rational, but because they operate with less loss of information and require fewer errors in central decision-making.

Applied to prediction markets, Hayek’s logic suggests that Polymarket and similar platforms should outperform expert panels, opinion polls, or government forecasts. Traders with money at stake have a direct incentive to be accurate. Those with private information—a political operative aware of a campaign’s internal polling, a supply chain manager sensing future commodity demand, a researcher who has seen preliminary findings—can profit by trading on that knowledge. Their actions move prices toward a true probability. The result should be a market price that reflects all available information, weighted by the confidence and capital of those who believe they possess it.

The Wisdom of Crowds Versus the Winner’s Curse

The appeal of aggregated market wisdom rests on a specific mathematical reality: if individuals hold independent beliefs drawn from a distribution around a true value, the average of those beliefs often approaches the truth as the number of participants grows. This phenomenon underlies the famous Francis Galton experiment in which a crowd estimated the weight of an ox, and the median guess proved more accurate than most expert estimates. Markets, the argument goes, are crowds with a financial incentive to be right and a mechanism for everyone’s beliefs to influence a price.

Yet the conditions required for this aggregation effect are stricter than they first appear. Participants must hold genuinely independent beliefs. If traders receive the same news at the same time and anchor on the same initial price, independence erodes. They must also have symmetric information costs and access; a trader with a data terminal and algorithmic execution has an advantage over someone trading on a hunch. The crowd effect depends on heterogeneity of opinion; if everyone agrees, there is no trade and no price adjustment.

Market prediction offers another complication: the winner’s curse. A participant willing to bet at a particular price believes that price underestimates or overestimates the true probability. But if many traders have this belief, the market price should already reflect it. A trader who consistently finds opportunities to profit is either (a) better informed than others, (b) taking bigger risks than others, or (c) benefiting from market inefficiency that others haven’t yet corrected. None of these guarantees that the market price is moving toward truth; the most aggressive, overconfident, or informed traders can move prices in their favor regardless of correctness.

Polymarket’s actual trading patterns illustrate this tension. During the 2020 US election, some prediction markets assigned significantly higher probability to certain outcomes than aggregate polling data suggested. Some of these prices proved wrong in hindsight. In other cases, markets were more accurate than traditional forecasts. The platform’s liquidity—how easily traders can buy or sell without moving the price—determines how much consensus costs to test. If a market has thin liquidity and a few large traders with strong convictions, the price may reflect their beliefs rather than aggregated dispersed knowledge.

The Hayek-Arrow Tension: Markets Versus Democracy

Kenneth Arrow’s fundamental impossibility theorem poses a direct challenge to Hayek’s framework. Arrow proved mathematically that no voting system can satisfy a set of intuitively reasonable criteria simultaneously: respecting individual preferences, avoiding arbitrary dictates, producing transitive social choices, and remaining independent of irrelevant alternatives. His work seemed to suggest that collective decision-making is inherently problematic—not just inefficient, but impossible to perform fairly.

Yet Arrow and Hayek were attacking different problems. Hayek defended markets as a coordination mechanism for resource allocation under dispersed knowledge. He acknowledged that markets assumed property rights and legal frameworks that no market alone could justify. Arrow was examining voting and social choice—the problem of translating individual preferences into legitimate collective decisions when values, not just facts, are at stake. A market price answers the question «What probability do traders assign to this outcome?» A democratic vote answers a different question: «What outcome do citizens prefer?»

This distinction matters for evaluating whether Polymarket could replace democratic institutions. Markets are excellent at revealing probabilistic beliefs about factual outcomes—will it rain, will the GDP grow, will a merger complete? They are not equipped to reveal what outcome a society should prefer when multiple outcomes are feasible. Markets can predict whether a policy will reduce inflation; they cannot tell a society whether inflation reduction or unemployment reduction is more important. A market price represents preferences weighted by wealth and willingness to bet, not preferences weighted by citizenship or moral claim.

The futarchist vision attempts to bridge this gap by proposing that voters choose objectives—a target unemployment rate, a preference for civil liberties—while markets predict which policies will achieve those objectives. Voters decide what they want; prices predict how to get it. Yet this separation is never clean in practice. Predictions about policy outcomes depend on assumptions about implementation, compliance, and counterfactual scenarios. Disagreements about what a policy will cause are often disagreements about values and priorities dressed in predictive language.

Polymarket as Aggregation Engine: What It Does Well

Polymarket’s actual track record suggests that prediction markets excel in narrow, well-defined settings. When the event is binary, objective, and difficult to manipulate, market prices can aggregate information efficiently. The platform’s use of UMA oracles for resolution provides a mechanism for settling contracts without relying on a centralized authority—a genuine advance over predecessors like Intrade, which depended on institutional legitimacy and was shut down by US regulators.

The platform’s technical architecture reinforces information aggregation. Smart contracts operating on Polygon ensure that trades settle exactly as the code specifies, removing counterparty risk and settlement uncertainty. The use of Automated Market Makers (AMMs) instead of traditional order books means that traders can always execute at a mathematically determined price rather than waiting for matching orders. This reduces barriers to entry and participation; retail traders need not compete with market-making professionals for favorable prices. USDC settlement provides immediate liquidity in a stablecoin, avoiding the need to hold speculative tokens.

High-frequency trading capability—enabled by Polygon’s fast finality and near-zero fees—allows informed traders to profit rapidly when they believe prices are mispriced. This arbitrage activity should push prices toward accuracy. If a trader believes a contract is overpriced at 65 cents, they can sell aggressively at lower prices, moving the market down. If they are correct about the true probability being lower, their profit will incentivize others to follow. Over time, this mechanism can correct systematic biases in aggregate beliefs.

The platform’s data also feeds external decision-makers. Policy researchers, election analysts, and investors can observe market prices as one signal among many. A prediction market assigning high probability to a particular outcome adds one more data point to supplement polls, expert forecasts, and other information sources. The question is whether that signal is reliable and whether over-relying on it could substitute for other forms of deliberation.

Limits of Price-Based Truth-Finding: Why Markets Fail at Complex Outcomes

Polymarket’s greatest limitation becomes apparent when outcomes are multidimensional, depend on interpretation, or cannot be definitively resolved. An event like «Will tensions between Country A and Country B escalate significantly?» requires defining «escalate» and «significantly.» Who decides whether the event occurred? The UMA oracle design attempts to solve this by relying on a decentralized panel of token holders and arbitrators, but it still requires judgment. If different people reasonably disagree about whether an event happened, the market price cannot resolve the disagreement; it can only aggregate conflicting interpretations.

Many real policy questions resist clean resolution. «Will healthcare outcomes improve under Policy X?» depends on which outcomes matter, for which populations, over which timeframe, compared to which baseline. A market can price the binary question «Will a policy reduce hospitalization rates by at least 5 percent?» but that market price tells you nothing about whether a 4 percent reduction, improved patient satisfaction with a 3 percent increase in hospitalizations, or a range of other outcomes might be preferable from a societal perspective. The market aggregates beliefs about a specific measurable fact; it does not resolve which measurement matters most.

Information asymmetry also undermines price-based aggregation in high-stakes political and economic outcomes. A government official or major market participant may hold private information about policy intentions, regulatory decisions, or hidden economic data. They can profit by trading on that information before it becomes public—but their trading does not reveal that information to other market participants, it merely prices in their private advantage. The resulting market price may be quite wrong from the perspective of someone without access to that privileged information.

Manipulation presents another challenge, particularly in lower-liquidity markets. If a trader is willing to accept large losses to move a market price upward or downward—perhaps because they benefit from specific outcomes in real life and merely want to influence perception—they can do so on Polymarket as easily as on traditional markets. The blockchain’s transparency does not prevent manipulation; it only ensures that the manipulation is recorded immutably. A large bet by someone with an axe to grind can move prices in their preferred direction without indicating anything about true probabilities.

Could Futarchy Actually Work? Practical Obstacles

Futarchy proposes that a society vote on objectives—target unemployment of 4 percent, inflation under 3 percent, civil liberties protected—and then allow prediction markets to evaluate which policies will achieve those objectives. The government would implement whichever policies the market predicts will maximize the chosen metric. This design attempts to delegate empirical questions to markets while reserving value judgments for voters.

The first obstacle is that voters’ stated objectives often contradict in practice. Most voters prefer both lower inflation and lower unemployment, but there is a genuine trade-off between them. A market would have to predict the effect of a specific policy on one metric while voters’ true preference is for a weighted combination. The design assumes that objectives are independent and clearly ranked, when in reality they are interconnected and subject to genuine disagreement.

The second obstacle is that predicting policy effects requires controlled experiments or credible counterfactual reasoning that markets alone cannot provide. If a government implements a stimulus policy and unemployment falls, did the policy cause the fall, or would unemployment have fallen anyway? A market price cannot answer this question directly; it can only reflect traders’ beliefs. If traders systematically overestimate policy effectiveness—because of optimism bias, or because they are anchored to the government’s stated intentions—the market will misprice the effect, and futarchy will implement policies based on false beliefs.

The third obstacle is that markets require genuine uncertainty and independent participants. If the outcome of a futarchist vote is that markets should decide policy on unemployment, and the government is known to have vast informational advantages about employment trends, some traders will abandon the market as unwinnable against the information-privileged government. Liquidity dries up; prices become noise. The market price loses legitimacy precisely because it is no longer aggregating independent beliefs.

The Case for Markets as a Complement, Not a Replacement

The most defensible claim for Polymarket and similar platforms is narrower than the futarchist vision. Prediction markets can serve as one source of information within a broader decision-making process, complementing rather than replacing democratic institutions. A central bank considering interest rate policy can observe prediction market prices for inflation alongside labor data, surveys, and model-based forecasts. A government agency evaluating a regulatory intervention can track prediction market prices before and after announcement to observe how informed traders expect the intervention to affect outcomes.

This subsidiary role is consistent with Hayek’s original insight. Markets aggregate dispersed knowledge efficiently; that aggregation can inform decision-makers without substituting for democratic deliberation about values and trade-offs. A prediction market price is one more piece of evidence, valuable precisely because it has been tested by participants with financial incentives to be accurate. But evidence is not the same as decision authority.

The design of prediction markets also matters for their credibility as information sources. Polymarket’s use of USDC and Polygon ensures that settlements cannot be blocked or delayed by a central party. The platform’s openness allows external observers to verify prices and liquidity. Yet these features protect against institutional corruption; they do not protect against the epistemic limits of markets themselves. If a market is thin, manipulable, or resolving a question that admits of genuine interpretation, those structural problems persist regardless of how decentralized the platform is.

A more mature ecosystem might incorporate prediction markets into formal forecasting exercises. Intelligence agencies, central banks, and regulatory bodies already conduct structured forecasting through aggregated expert judgment and prediction tournaments. Markets could extend these methods by incorporating financial incentives and real-time price updates. The result would not be governance by market prices, but rather governance informed by aggregated knowledge about expected outcomes, with final authority remaining with elected representatives and designated officials who can weigh the evidence against other considerations.

Toward a Realistic Assessment of Market-Based Governance

The Hayek-Arrow debate remains unresolved, and Polymarket’s existence does not settle it. Hayek correctly identified that markets aggregate dispersed knowledge more efficiently than central planners. Arrow correctly identified that voting and choice aggregation involve value judgments that markets cannot resolve. Both insights are true simultaneously, and both constrain how prediction markets can function in a democratic society.

Polymarket itself demonstrates the practical value of markets within their appropriate domain. Traders with real money at stake have produced prices that often forecast outcomes more accurately than traditional experts. The platform has attracted institutional investors, quantitative traders, and retail participants, creating genuine liquidity for many events. The technical design—smart contracts, USDC settlement, Polygon scaling—eliminates layers of institutional intermediation that plagued earlier prediction markets.

Yet the platform’s growth has also revealed limits. When outcomes are ambiguous, outcomes are heavily influenced by policy decisions made after the market opened, or when participants have strongly asymmetric information, market prices reflect those problems rather than solving them. Polymarket cannot reliably predict outcomes that depend on interpretation or that are subject to manipulation by large participants. It is not equipped to aggregate value judgments, only factual beliefs.

The path forward is neither to replace democracy with market prices nor to dismiss prediction markets as useless. Instead, decision-makers should observe prediction market signals as one input—valuable specifically because it reflects aggregated judgments tested by financial incentives—while preserving democratic authority over values, trade-offs, and the legitimacy of final outcomes. Markets excel at revealing what traders expect to happen; they do not determine what a society should want to happen. Confusing the two would abandon both the strengths of market aggregation and the foundations of democratic accountability.