Quadratic Funding

The Mechanism

The mathematics underlying quadratic funding is over two centuries old and is one of the most widely deployed techniques in modern measurement. Least squares optimization, developed by Gauss and Legendre in the early 1800s to handle noisy astronomical observations, solves a specific problem: given many partial, noisy signals about something you cannot directly observe, how do you combine them into the best stable estimate of what is actually there. The answer is to find the estimate that minimizes the sum of squared differences between the estimate and each individual signal. Squaring is the load-bearing operation: a signal wildly off the consensus contributes a much larger error term than one only slightly off, so the math naturally pulls the answer toward where most signals agree, while outliers get suppressed without being ignored.

This is the math underwriting GPS. Your phone is in contact with several satellites at once, each producing a noisy estimate of your distance from it, and no single satellite knows where you are. The phone finds the location that minimizes the sum of squared distance errors across all available satellites, and that location is what shows up on your map.

Quadratic funding applies the same operation to a different kind of signal. Where GPS treats satellite readings as noisy estimates of your location, quadratic funding treats individual contributions as noisy estimates of how much a group collectively wants something. The mechanism was formalized in 2018 by Vitalik Buterin, Zoë Hitzig, and Glen Weyl, building on decades of mechanism-design theory. It works through a matching pool: contributions to each option get matched in proportion to the square of the number of contributors rather than the total amount raised. A project supported by 100 people at $10 each receives a much larger match than a project supported by one person at $1,000, because the squared-breadth term dominates the linear-depth term. The principle is identical to GPS: weight the consensus, suppress the outliers, recover the stable estimate of what the group actually wants.

The Structural Principle

Aggregating the preferences of many people into a single collective decision faces a problem that does not arise when aggregating GPS signals: the participants have unequal resources to begin with, and most existing aggregation methods count those resources linearly. One-dollar-one-vote treats a single $10,000 contribution as equivalent to 10,000 contributions of $1 each. One-person-one-vote refuses to weight intensity at all, treating a participant who barely cares as equal to one for whom the outcome is existential. Hierarchical decision-making concentrates the signal further still. Each method makes a different tradeoff, and each produces a different failure mode.

The failure mode that quadratic funding specifically addresses has a name in current parlance: whale capture. A small number of disproportionately resourced participants — a few wealthy donors in a foundation grant round, a few large stakeholders in a participatory budget, a few organized factions in a community vote — end up determining outcomes that the process is nominally inviting everyone to shape. The word is new but the dynamic is ancient. Roman patrician families ran exactly this play on the participatory institutions of the Republic for centuries, using patronage networks and inherited resources to swing assembly decisions far beyond their numerical share, while the formal structures continued to call themselves participatory. The capture is older than the vocabulary; the vocabulary is just the latest name for it.

Quadratic funding is one mathematical response to whale capture, not the only one and not free of its own tradeoffs. Collusion resistance is a live problem — what stops a whale from splitting their contribution across many sock-puppet accounts to harvest the matching pool? Identity verification is a precondition the math itself does not solve. The matching pool itself has to come from somewhere, which means the structure works best when paired with an existing funder willing to underwrite it. But within those constraints, the mechanism does something previous aggregation methods cannot: it distinguishes broad support from concentrated support, in the signal rather than around it, mathematically rather than procedurally.

Where This Could Land

The principle points to any institution that allocates resources or makes collective decisions across many participants with uneven intensity of preference, and that wants the outcome to reflect breadth rather than depth of resource. The candidate test is whether the institution is trying to be participatory — whether it would consider whale capture a failure mode rather than a feature. Where the answer is yes, quadratic funding becomes available as a tool.

In public governance, participatory budgeting programs at the municipal scale are the most direct fit. Cities running such programs already invite resident input on resource allocation, and most current implementations measure that input in ways that favor either the most organized groups or the most well-funded campaigns. A quadratic structure would distinguish a project supported broadly across a neighborhood from one supported intensely by a smaller faction. Grant-making foundations face an adjacent version of the same problem when distributing across many candidate recipients, and so does any public consultation process that solicits but then has to aggregate community input.

In commercial governance, the candidates are anywhere participatory decision-making coexists with stakeholders of unequal resource: cooperative ownership structures choosing strategic direction, employee voice mechanisms in mid-sized firms, professional association governance, member-driven cultural institutions deciding what to program. The pattern is consistent — many participants, uneven intensity, currently aggregated through methods that lose one signal or the other.

The activation gap is unusually wide. The people who understand quadratic funding are concentrated in mechanism-design academia and cryptocurrency-adjacent communities, where the conversation is largely about crypto-native applications. The people who could deploy it — municipal staff running participatory budgeting, foundation program officers, cooperative organizers, governance committees in professional bodies, anyone who has watched a participatory process get captured — sit in entirely different rooms and largely do not know the mechanism exists. The crypto vocabulary that surrounds current discussion of the tool is itself a barrier. For the connection to take, someone running a participatory process needs to encounter the mathematics in language that does not assume crypto fluency, and a mechanism-design specialist needs to be willing to leave the room they currently sit in.

Rubedo's Interest

Rubedo's network spans 57 treaty corridors, each potentially holding a research thread, a slate project, a documentary pairing, a student-led investigation, an institutional contact. The recurring operational question is signal detection: across that breadth, where is genuine collaborative traction forming versus where is one persistent voice creating an illusion of momentum. Pure top-down direction loses the local knowledge that lives only in the people working a specific corridor. Pure linear aggregation of internal interest — counting expressions of enthusiasm at face value — is vulnerable to exactly the dynamic the mechanism addresses, where one well-resourced or well-positioned advocate can swing attention away from corridors with broader but quieter support. A quadratic structure for surfacing internal interest, where a corridor with five different independent signals gets weighted more heavily than one with a single persistent advocate, is a direction we may build toward as the network grows.