Mycelial Economics

The network beneath the forest

A Douglas fir seedling germinating in the shade of its parent cannot photosynthesize enough to survive. It should die. Often, it does not. Instead, carbon flows to it through a network of fungal threads so fine — two to twenty micrometers in diameter — that a single tablespoon of healthy forest soil contains miles of them. The seedling is being subsidized by trees it cannot see, via an intermediary it could not name.

This is not a metaphor. In a landmark 1997 paper in Nature, ecologist Suzanne Simard and colleagues tracked carbon isotopes through Douglas fir and paper birch stands in British Columbia, demonstrating measurable net transfer between species through their shared mycorrhizal networks. The birch was exporting carbon to the fir. The network was routing it. No central coordinator existed. No one had decided that fir seedlings were a priority investment. The allocation emerged from the structure of connections and the chemistry embedded in those connections.

Simard went on to identify what she called "mother trees" — older, larger individuals connected to hundreds of neighbors through the fungal web, functioning as hubs. When these hub trees were felled, neighboring seedling survival rates declined measurably. The network had topology: some nodes mattered more than others, and their removal degraded the whole system's resilience. Her long-running Mother Tree Project has since expanded this work into managed forest systems, asking what happens to the network when logging practices sever the hubs.

The picture that emerges is not of a harmonious cooperative. It is of a complex, dynamic allocation system. Trees under stress pull resources toward themselves through chemical signals in root exudates. Trees with surplus — in good light, with adequate water — become net exporters. The directionality is driven by gradient, not charity. The fungus is not altruistic either: it is trading, taking carbon from plant roots in exchange for phosphorus and nitrogen it has scavenged from soil. Every connection in the network is a negotiated exchange, running at scales and speeds that no market mechanism we have ever built can match.

Biological markets: trade without price signals

Toby Kiers, an evolutionary biologist at Vrije Universiteit Amsterdam, has spent years making the economics of mycorrhizal networks precise. Her framing — "biological market theory" — is deliberately rigorous and deliberately unsentimental. Fungi behave as economic agents. They allocate phosphorus preferentially to plant partners offering the most carbon in return. When the supply of phosphorus increases suddenly (a "boom"), fungi lower their effective "price." When it drops (a "crash"), price rises. The system runs supply-and-demand dynamics without prices, without a market, without any agent that has modeled the whole.

The mechanism is what matters. Both plant and fungus can detect the quality and quantity of what their partner is providing, and both adjust their own allocation accordingly. This is not pre-programmed reciprocity. It is responsive, continuous, bilateral negotiation mediated by biochemistry. Kiers' lab has tracked this at the level of individual hyphal branches, watching nutrients move faster through high-flow connections and slower through low-flow ones — the network literally remodeling its own infrastructure in real time based on what is working.

This is the architectural insight. Our economic systems route resources through abstracted price signals — a representation of value that can be manipulated, hoarded, and decoupled from the underlying need. A forest floor routes resources through gradient chemistry — a direct encoding of need in the signal itself. When a tree is phosphorus-stressed, its root exudates change in ways that the fungal network reads as demand. There is no intermediary who profits from knowing the demand before everyone else does.

A note of precision here, because the biology has been oversimplified in popular accounts: a 2023 paper in Nature Ecology & Evolution found that positive citation bias has inflated claims about how deliberately and reliably mycorrhizal networks distribute resources to the needy. Some older studies cited as evidence of intentional "helping" were actually studies of fungal distribution that never measured nutrient transfer at all. The phenomenon is real; the degree of directed benevolence is contested. What remains solid: the network routes resources, the routing responds to chemical signals encoding need and surplus, and the system achieves distributed allocation without central coordination. The precise mechanism is more market-like and less altruistic than the popular "wood-wide web" narrative suggests — which makes it more interesting as an economic model, not less.

The cambium model: growth without accumulation

The vascular cambium is a single cell layer in a woody plant — a cylinder of undifferentiated meristematic cells running the full length of every stem and root. Each year, these cells divide outward (producing new phloem, the tissue that transports sugars) and inward (producing new xylem, the tissue that transports water). This is secondary growth: the thickening of existing structure rather than the extension of new structure.

What is remarkable about cambium is its spatial logic. Growth is distributed uniformly around the circumference. The cambium layer does not decide that the south-facing side of the tree is most important and thicken preferentially there. Growth is regulated by local hormone signals — primarily auxin flowing downward from actively growing shoot tips — and each section of cambium responds to its own local concentration. The result is that the whole organism thickens proportionally. No section grows at the expense of adjacent sections. Structural integrity is maintained across the whole.

There is a corrective mechanism too. When a tree is subject to persistent lateral loading — wind from a prevailing direction, the weight of a heavy branch — the cambium on the compression side upregulates production of "reaction wood," denser fiber tissue that compensates for the load. The growth allocation responds to mechanical stress. The organism routes additional structural investment to where the environment is applying pressure.

This is categorically different from how capital accumulates in human economies. Capital does not respond to need-gradient signals; it responds to return-gradient signals. It flows toward wherever returns are highest, which is typically wherever capital is already concentrated — because concentration enables the scale effects, risk diversification, and market power that generate superior returns. The feedback loop is the opposite of cambium's. The more you have, the more flows to you. The less you have, the more flows away. No organism built on cambium logic would survive, because it would perpetually starve its stressed sections while over-thickening its comfortable ones until it became structurally absurd and then collapsed.

Where extractive economics structurally fails

The pathologies of extractive economic systems are not accidents of bad actors or insufficient regulation. They are structural. The feedback loops that produce them are the same loops that produce short-term efficiency, which is why they are so difficult to correct from within the system.

Wealth concentration has a simple mechanism: returns to capital exceed returns to labor over long periods (this is Piketty's r > g, but you do not need the formal argument — just observe that owning productive assets generates compound returns while selling time does not). Concentration increases market power, which increases returns, which accelerates concentration. The positive feedback loop is not a bug. It is the natural attractor state of a system that routes investment toward return-maximizing rather than need-maximizing signals.

Resource depletion follows the same logic. A fishery held as a commons is over-fished when individuals can capture all the return from extraction but share the cost of depletion across all users. Garrett Hardin called this the Tragedy of the Commons in 1968, and it became the canonical argument for privatization as the solution. Elinor Ostrom won the Nobel Prize in Economics in 2009 for demonstrating that this framing was empirically wrong: communities that manage commons successfully do so through polycentric governance — local rules, locally enforced, with graduated sanctions and shared monitoring. Neither the state nor the market is necessary. A third option — commons governance — works, and has worked for centuries in Swiss mountain pastures, Japanese forests, and Spanish irrigation systems.

What Ostrom discovered, without framing it this way, was a human implementation of mycelial-style governance: distributed, locally responsive, with feedback loops that route pressure toward over-extraction before it becomes irreversible. The commons institutions she documented were not romantic or utopian. They were precisely calibrated enforcement mechanisms, evolved through centuries of trial and error. They worked not because the participants were altruistic, but because the rules created local gradients that made sustainable use individually rational.

What has actually been tried

The history of attempts to build mycelial-style economic systems is longer and more varied than most people know, and the record is genuinely mixed.

The WIR Bank (Wirtschaftsring) is the longest-running example. Founded in Switzerland in 1934 by sixteen entrepreneurs during the liquidity crisis following the 1929 crash, it created a parallel currency — the WIR franc — that circulates only within its member network of small and medium businesses. No interest is charged. Credit is issued to members against pledged assets. As of the mid-2000s, it had grown to 62,000 members and handled approximately 6.5 billion CHF in annual transactions. Crucially, WIR usage is countercyclical: when Swiss franc liquidity dries up in recessions, WIR transactions increase, functioning as an automatic economic stabilizer. The system has operated continuously for over ninety years without a central bank or government mandate. It is not radical. It is unremarkable infrastructure that quietly reduces volatility.

Platform cooperatives represent the attempt to apply cooperative ownership models to digital platforms. Stocksy (photographer-owned stock photography), Up&Go (worker-owned home services), and the Drivers Cooperative (ride-share owned by its drivers) are operating examples. The model is structurally sound: the people generating value on the platform own the platform, so surplus flows back to contributors rather than to external shareholders. The adoption barrier is capital — launching a cooperative platform requires upfront investment that conventional venture capital will not provide because the governance structure prevents the extraction that makes VC returns work.

Blockchain and DeFi are the cautionary tale. The promise was a protocol-level, trust-minimized system for routing value without intermediaries — genuinely mycelial in aspiration. The execution largely recreated speculative finance on a new substrate. A 2025 study found that on average, the top 0.23% of DeFi wallet addresses hold 92.29% of total token supply — concentration that exceeds even Bitcoin's, which itself is severely concentrated. The European Central Bank found the top 100 addresses control more than 80% of governance voting power in major DeFi protocols. The system built the infrastructure for decentralization and then immediately concentrated along the same gradients that concentrate everything else, because it embedded the same return-maximizing signals that produce concentration in all markets. The architecture was mycelial; the incentive structure was extractive. Incentive structure won.

Mutual credit networks — systems where members extend interest-free credit to one another, creating a currency that expands when there is trading capacity and contracts when debts are settled — are perhaps the closest structural analogue to mycelial resource flows. The currency is created at the point of exchange rather than extracted from a scarce supply. There is no interest, so there is no compound accumulation. Credit flows to where there is productive capacity to use it. The Community Exchange System, operating across dozens of countries, and the Sardex network in Sardinia (which processed over €60 million in business-to-business exchanges annually at its peak) are working implementations. They are not replacing national currencies. They are filling gaps the national currency does not reach — exactly as WIR francs fill liquidity gaps in Swiss recessions.

The blueprints technology makes possible

For most of human history, the gap between how ecosystems allocate resources and how human economies do was not just a matter of will or institutional design. It was a matter of information infrastructure. A mycelial network achieves real-time, continuous, distributed signal propagation through direct chemical contact. Human economies had no equivalent. Price signals were the closest approximation — an abstraction that encodes supply and demand across large geographic distances. The abstraction introduced lag, opacity, and the possibility of manipulation. We used prices because we had nothing better.

We now have something better. Not better in every dimension — but better in the specific dimension that matters for building mycelial-style allocation systems.

Real-time visibility into need and surplus is now technically feasible at large scale. Community land trusts in several cities use live occupancy and income data to price units dynamically against actual housing need rather than speculative market value. Open food networks connecting farms to buyers route produce against real order demand rather than against what the distributor thinks will sell. These are small, but they are proofs of concept for a different signal architecture — one where the resource flow is calibrated to gradient rather than to return.

Distributed compute already routes workloads mycelially: a packet does not travel a predetermined path from source to destination, it finds its way through whatever network topology is currently functional, rerouting around congestion and failure. The internet is a working implementation of mycelial routing at the information layer. The question is whether equivalent logic can operate at the resource and value layer.

The cambium model suggests a specific design pattern: local signal response, with growth distributed proportionally and stress-routed preferentially. In economic terms, this translates to systems where investment allocation responds to need-gradient signals rather than return-gradient signals, and where over-extraction triggers automatic countermeasures rather than requiring regulatory intervention after the fact. Ostrom's design principles for successful commons governance — clear membership, rules that match local conditions, graduated sanctions, nested governance structures — are a specification for how to implement this in human institutions. The technology to instrument and automate those principles now exists.

The limits of the metaphor

The biological analogy is productive, but the map is not the territory. It is worth being precise about where it breaks down, because the breakdowns are where the design work actually lives.

Trees cannot defect. A tree cannot decide to stop sending root exudates, fake a stress signal to pull more resources from the network, or route its carbon to a private stockpile rather than the shared system. Human economic actors can and do all of these things. Any mycelial-inspired economic system has to solve the defection problem that the biological system does not face. Ostrom's commons institutions solve this with monitoring and graduated sanctions — expensive mechanisms that the mycelium does not need. This cost is real and should not be elided.

Ecosystems are not just. The forest does not have preferences about fairness. A tree in bad soil gets less. An organism that cannot participate in the exchange network simply dies. Biological distribution optimizes for system resilience, not for the welfare of every participant. Human economic systems have additional constraints that the biological model ignores: we care about outcomes for individuals, not just system-level stability. The mycelial model needs to be augmented with explicit mechanisms for inclusion, not just efficient allocation.

Ecosystems are slow. Mycorrhizal network formation takes years. The feedback loops operate on seasonal timescales. Human economies are subject to political cycles, technological disruptions, and crises that require response times the biological analogy does not address. A system well-optimized for the gradual, patient allocation logic of a forest is potentially fragile to fast-moving shocks — unless it is specifically designed for dynamic rebalancing, which requires layers of mechanism the biological analogy does not provide.

There is no designer. Ecosystems are the product of billions of years of selection acting on variation without foresight. Successful patterns were retained; failed ones died. We do not have billions of years. We have to design deliberately, which means we have to understand what we are selecting for and why — something evolution never had to articulate. The biological model is a source of patterns. It is not a source of goals.

What needs to be built

The question is not whether mycelial-style economic systems are possible. The WIR Bank has been running for ninety years. Ostrom documented polycentric commons governance working for centuries before that. The question is whether these patterns can be implemented at the scale and speed that the current economic transition requires — and whether the builders who are actually constructing digital infrastructure understand the signal architecture well enough to embed the right gradients.

Three specific things need to be built, and all three are technically tractable.

Need-gradient infrastructure. The dominant digital economic infrastructure routes resources toward return. Search algorithms surface content that maximizes engagement. Recommendation systems surface products that maximize conversion. Lending systems route credit toward borrowers most likely to repay rather than toward uses most likely to generate community value. Every one of these systems makes a choice about what gradient to follow. Building alternatives means building the instrumentation to make need-gradients legible — which requires data infrastructure that is held in common rather than controlled by the entities profiting from return-gradient routing.

Commons-layer protocols. The open source software stack is the clearest existing model: Linux, Apache, PostgreSQL, and their descendants are infrastructure held as commons, which anyone can use and contribute to, governed by rules that prevent enclosure. The economic value created on top of this commons vastly exceeds the economic value captured by it — which is both evidence that the model works and evidence of why it is perpetually underfunded. The next layer of commons infrastructure is the data and identity layer: protocols for attesting to credentials, tracking contributions, and routing value back to contributors that cannot be enclosed by any single platform. Projects like Verifiable Credentials (W3C standard) and ActivityPub (the protocol underlying the federated social web) are early implementations. They are not ready. They need serious investment and serious builders.

Countercyclical credit mechanisms. The WIR Bank's most important property is countercyclicality: it creates liquidity precisely when conventional liquidity dries up, preventing the collapse of productive economic activity during downturns. This is the cambium model in financial form — routing growth-enabling resources toward stressed sections of the economy at the moment of stress rather than away from them. Mutual credit networks at regional scale, designed with modern tooling, could implement this for small-business ecosystems across geographies where conventional banking has already retreated. The technical barrier is near zero. The institutional barrier is real: it requires communities with enough trust and shared interest to establish the network before they need it.

The biological blueprints are real and specific. The mechanism is not mysterious: local signal response, distributed allocation, feedback loops that route toward stress rather than away from it, and network architecture that makes hub removal survivable rather than catastrophic. Implementing these in economic systems requires solving problems the biology does not face — defection, inclusion, speed — but those problems have known solution patterns. What they require is builders who think at the level of signal architecture rather than product features.

The forest has been running this system for hundreds of millions of years. We have had the observational tools to understand it for about thirty. The gap between understanding the blueprint and building from it is entirely a matter of what we decide to build.