After reading Kevin Carson’s Anarchist Notes on the Theory of Money, Credit, and Capital, I was struck by two things:

1. We can do this.
2. Where the fuck is reproductive labor in this?

Here’s where my thoughts have led me so far. If you’re not familiar with mutual credit, I’ve added a summary at the end you can read first – What IS Mutual Credit?

Table of Contents:

• Legibility and Credit Extension
• Co-ops and Ecological Stewardship
• Reproduction as Commons
• Cash and Credit Across Transition
• Rivalrous and Non-Rivalrous Exchange
• What IS Mutual Credit?

Legibility and Credit Extension

In Anarchist Notes, Carson quotes Thomas Greco on the proper basis of credit issuance:

“It is crucial that the quantity of the exchange medium be balanced with the flow of goods and services coming into the market, and that it be self-adjusting…. If money is properly issued, there will never be any problem of under-supply or over-supply. The quantity of money will always be just the right amount to purchase the goods and services which it represents…. The proper basis of issue is the transfer of value, as it is being exchanged, from a producer to another (potential) producer.”

And later, on what qualifies a participant to issue credit:

“A participant is qualified to put into circulation an amount of ‘money’ up to the amount of his/her sales over a 2 or 3 month period.”

Both formulations are doing something similar. They treat market activity — value transfer between producers, sales turnover over a recent window — as the natural anchor for credit issuance. The unit of account is calibrated to the flow of legible exchange. This is sensible as far as it goes, but it leaves two substrates unaccounted for, and both matter.

The ecological substrate. “The transfer of value” treats value as a closed loop between producers, but every transfer is also a draw on ecological systems whose regenerative capacity is finite. Issuing credit against value that depends on irreversibly depleting non-substitutable natural capital is issuing claims against a wealth base that may cease to exist. The proper basis of issue isn’t transfer of value — it’s transfer of value within the limits of what ecological systems can regenerate. Without that constraint built in, the “self-adjusting” property Greco describes is self-adjusting only with respect to the market, not with respect to the substrate the market depends on. The currency floats free of the conditions that make it possible.

The reproductive substrate. Sales-turnover-based credit qualification builds in a structural bias toward productive labor and against the labor that makes productive labor possible in the first place. Every adult capable of generating sales is the product of decades of reproductive work, such as feeding, teaching, protecting, integrating into community life, that doesn’t itself generate sales but without which there is nothing to sell and no one to sell it. A credit system that calibrates issuance to recent sales volume treats this entire substrate as exogenous, the way classical economics treats the household. Would one need to ‘rent’ their creditworthiness from someone in the productive sector, reinforcing hierarchies?

We already see the capitalist version of this bias in lending discrimination against women-owned businesses, where the same pattern of “what counts as legitimate productive activity” produces predictable exclusions. Carrying that calibration into a mutual credit system would carry the bias with it.

Both omissions point to the same underlying issue: the proper basis of credit issuance has to make legible what the system actually depends on, not just what the system most easily measures. Sales volume is easy to measure. Ecological regenerative capacity is harder. Reproductive labor may be harder still. But the difficulty of measurement isn’t a justification for treating the substrate as if it weren’t there.

This is what I mean by the legibility problem. Credit extension is, structurally, a decision about whose contribution counts as economic and who is recognized as a participant capable of issuing claims against future flows. When the basis of legibility is sales turnover, the people whose labor doesn’t pass through markets become economically invisible, even though the markets depend on their labor. They become illegible in the social field, with no material agency that the system recognizes. Someone in that position may reasonably wonder whether someone like them is allowed to dream of a different life — because the system has already answered, in its choice of what to count, that they’re not the kind of person economic flows are organized around. In time, the coercion of the system becomes internalized.

A mutual credit system worth building has to extend legibility to ecological stewardship and to reproductive labor, both, and it has to do so structurally rather than as a supplement. Not “credit issuance based on sales, plus a separate program for caregivers, plus an environmental tax.” The basis of issue itself has to internalize what capitalism made external. The rest of these notes work out what that may look like, although I’m still exploring functional implementation.

Co-ops and Ecological Stewardship

After working through several alternatives — price multipliers, separate commons ledgers, ecological pools — I’ve landed on a simpler structure: ecological care and restoration labor should be credited by the co-ops using the natural resource.

In a mutual credit system organized around co-ops, every productive activity is, at some point in its supply chain, an interaction with natural systems. Even labor that looks purely cognitive or relational is downstream of food, shelter, and ecological conditions that themselves depend on stewarded land. This means resource-engaged co-ops sit at the foundation of the value stack — they’re the point at which the natural world becomes economic activity, and everything else is a transformation of what they bring into that exchange.

Pricing ecological stewardship into the cost of resource engagement is therefore pricing it into the base of value itself. The cost of replanting, monitoring, fire management, watershed protection, and all the labor of maintaining the substrate gets carried forward through the entire chain of transformations as part of what those resources cost in the first place. There’s no other place it could honestly sit. If stewardship were funded downstream, those goods would already have been produced from inputs whose stewardship costs were excluded from their pricing. The correction would be running on top of an uncorrected base. Funding stewardship at the point where co-ops engage the substrate means stewardship is part of what value is, not an adjustment to it.

This is why the cost of tracking and stewarding the commons falls on the co-ops regardless of whether the resource use is theirs or households’. Complete tracking is what the resource condition requires, and the co-ops are positioned to do it because they’re the ones in direct relationship with the substrate. The entire structure of value for the entire community is made of those resources. The co-ops aren’t being penalized because the value of their work is what the value is made of to begin with.

Reproduction as Commons

Reproductive labor is commons stewardship and it should be funded the same way ecological stewardship is funded.

Parents and other caregivers, aunts and uncles and grandparents, communal carers, teachers, and the whole web of adults involved in raising children, are stewards of the human commons in a way that’s structurally analogous to how forest stewards are stewards of the timber commons. The next generation isn’t anyone’s private possession. It’s the community’s substrate, and the labor of sustaining it is stewardship of the commons that all future community life depends on.

The forest yields lumber, fuel, and materials for the community’s life, but only if it’s stewarded — replanting, thinning, fire management, monitoring. Without stewardship, it’s drawn down and eventually fails.

The human community yields workers, caregivers, scientists, theorists, builders for its own continuation, but only if its children are stewarded — fed, taught, loved, protected, integrated into community life. Without stewardship, the community isn’t reproduced and eventually fails.

If reproductive labor is commons stewardship, then the same funding logic that handles ecological stewardship handles this too. Co-ops engaged with the human commons fund the labor that maintains it, and the cost gets carried into the price of what they produce — which, for the human commons, means everything that uses adult human labor, because adult human labor is what reproductive stewardship produces.

If an apple costs 5 credits for the co-op’s labor alone, 7 to include ecological stewardship, and 10 to include reproductive stewardship, people buying the apple would only ever know the apple as costing 10. All other labor adjusts in response. The 10-credit apple is what an apple costs when nothing has been externalized. The 5-credit version is what an apple costs when ecological and reproductive labor have been silently extracted to fill the gap.

This makes the price structure carry information about the human commons that capitalist prices systematically exclude. If people are underprovisioned, the apple price goes up to meet the need. If caregivers are too individually burdened and the community needs to absorb more of the load through communal arrangements, the apple price also goes up. Either way, the apple price is telling you something about the state of the human commons. The signal is built into the unit of account.

A small note on accumulation: as prices vary across these conditions, mild decay on idle credits may work as a light precautionary measure against credit accumulation during periods of price adjustment.

Cash and Credit Across Transition

A reasonable concern about mutual credit is that even in a community ready to use it, people still have obligations that demand cash, like rent, utilities, taxes, and debt servicing to capitalist institutions. If credits can’t meet those obligations, the system has a hard time getting established because participants need to keep one foot in cash regardless of how well the credit network is functioning internally.

The Mumbuca system in Brazil offers a workable pattern for this. Cash flows into the system at par — bring in a dollar, get a credit — but cash flows out only with deliberate friction, typically a small conversion fee (somewhere in the 1–10% range) or a waiting period, or both. The friction itself funds the cooperative entity that operates the conversion boundary. This creates an asymmetry that would make a transition system workable: low friction to enter, enough friction to encourage staying.

The mechanics for a participant: you do labor, you earn credits, you need cash for rent, you go to the cooperative conversion entity, you hand over 100 credits, you get back $90 in cash. The $10 difference funds the cooperative running the conversion service.

This conversion only runs in one direction for one class of participant. Someone who received care cannot take their credits and cash them out for themselves. The credits issued for care are denominated for circulation within the network, not for extraction. Otherwise the conversion mechanism becomes a way to launder care into capitalist purchasing power, which would undermine the whole point of the transitional system.

Mumbuca has political backing — it’s a municipal program — but a cooperative can run an equivalent mechanism through other means. A worker cooperative where internal labor is denominated in credits and external sales generate cash reserves can do conversion against those reserves. The 10% conversion fee can itself fund the labor that generates the external cash flow. Either way, the cooperative becomes the membrane between the credit network and the surrounding cash economy, with its membrane work funded by the friction of crossing it.

Over time, as more obligations can be met within the network and fewer require cash, the volume passing through the conversion boundary shrinks. The cash economy doesn’t have to be eliminated to be displaced. It just has to become less necessary, transaction by transaction, until the friction at the boundary is mostly handling edge cases rather than the main flow of economic life.

Further along, without a state, we can assume that not every area will participate in using mutual credits. Indigenous polities have their own systems, and purely anarcho-communist areas with no numeric measure of exchange will likely also exist. This is not problematic for the mutual credit system because mutual credit, as I envision it, is grounded in the same kind of relational flow. The actual formalization could disappear and our actions would remain the same as we trade with our friends and neighbors.

Where different systems share commons, such as a watershed, a forest, or a migration corridor, coordination happens through whatever protocols the people involved work out, not through one system’s unit of account governing the others.

Rivalrous and Non-Rivalrous Exchange

How does mutual credit handle things like software?

The question matters because software isn’t the only non-rivalrous good in a community’s economic life. Designs, recipes, written work, music, scientific findings, governance protocols, and accumulated knowledge of all kinds share the same structure. Distributing software to Penelope doesn’t deprive Benjamin the way distributing an apple to Penelope does. The producer hasn’t lost anything material when an additional user uses the same code. So what should the credit accounting look like?

The intuitive answer is that charging per copy or per usage would break the system. If a developer charges credits each time someone uses their software, they accumulate credits indefinitely without any corresponding draw on their own resources or labor. Each debit taken by the user doesn’t represent a loss of anything on the producer’s side. That’s a rentier dynamic in mutual credit form. The developer holds claims on the community’s future labor in proportion to how widely their work spreads, even though the labor that produced the work was bounded and finite. It also reinvents intellectual property by another name, making access to non-rivalrous goods conditional on payment when the underlying logic of “you used my finite resource” doesn’t apply.

For rivalrous goods, per-unit charging makes sense because each unit is a finite allocation. The apple Penelope eats isn’t available to Benjamin, so the credit-debit pair tracks a real material movement.

For non-rivalrous goods, per-unit charging is conceptually awkward because no real resource is being allocated when an additional user uses the same software, idea, or design. But if we credit the producer’s labor without any corresponding debit for usage, credits would be issued without corresponding debits, which means credits would lose their meaning as claims against actual labor flows.

A creative-commons pool might work to support developers, but a pool funding creative work would need a bureaucratic apparatus to decide what qualifies, how much different works should be valued, and who gets to draw against the pool. That apparatus would either ossify into a normative gate (only certain kinds of creative work get recognized) or fragment into endless contestation. Neither is good.

The cleaner answer is to recognize that mutual credit isn’t trying to be a universal valuation mechanism. It’s a tool for accounting for specific kinds of activity — ongoing labor and rivalrous resource allocation — and not everything in a community’s economic life needs to pass through it.

For non-rivalrous creative work, the production itself can be supported outside the credit system. Mutual aid and unconditional support, making the basic conditions of life independent of specific labor, means a creator doesn’t need to monetize their work to live. They can produce it because they want to, because it matters to them, or because the community values having it.

What does pass through the credit system, then, is the labor that’s actually relational and bounded:

Direct support and help. A user has a problem; a developer or support person helps them solve it. This is relational labor between two specific parties. The user receives help and the supporter provides it. Standard mutual credit logic applies cleanly: the user’s account is debited, the supporter’s account is credited, the amount reflects the labor involved. An hour of debugging help is worth roughly an hour of labor. There’s no accumulation problem because each interaction is bounded by the time actually spent.

Maintenance and ongoing development. Bug fixes, security patches, feature work that benefits current and future users. This can be handled in a few ways depending on context. It can be a gift to the commons, supported the same way the original creative work was. It can be funded through a pool mechanism if it serves many individual users or functions as community infrastructure. Or it can be funded by a specific subset of users who explicitly contract for the work. The right approach depends on context, so long as it still represents a relationship and resources.

Infrastructure costs. The energy, bandwidth, and physical hardware that non-rivalrous goods run on are themselves rivalrous. Per-unit charging is appropriate for those, because the underlying resources actually are finite. The non-rivalrous code or content running on top isn’t priced per-use, but the rivalrous substrate it depends on is.

The credit system handles what it’s suited to handle and the rest of the community’s economic life happens through other mechanisms.

What IS Mutual Credit?

Mutual credit is a way of tracking exchange that doesn’t use money.

When Amara does work for Roger, Amara’s account is credited and Roger’s account is debited by the same amount. Nobody had to have money before the exchange.

The credits aren’t a thing — they’re just an account of who has provided value to the community and who has received it. Over time, people work off their debits by providing services to others, and people with credits draw on the community’s labor when they need things. The whole system stays balanced at zero because every credit is matched by a debit somewhere. Credit limits, calibrated to what each person can realistically sustain through ongoing exchange, prevent anyone from going so deep into debit that they can’t realistically work it off. This is the structural protection against runaway accumulation in either direction.

Because there’s no way to accumulate excess credits as transferable wealth, the system structurally prevents several dynamics that produce contemporary inequality:

No wage labor as we know it. Without accumulated capital that needs returns through investment, the dominant form of productive enterprise tends toward worker cooperatives rather than wage-labor firms. Mutual credit doesn’t mandate this, but it removes the structural pressure that makes employer-employee relations the default. Capital can’t be accumulated and deployed to hire others, and workers can access means of production through cooperative organization without needing employment by capital-holders. Existing worker cooperatives demonstrate the viability of this form.

No inheritable class position. Credit positions reflect ongoing contribution and don’t compound into transferable wealth that descendants can use to avoid contributing themselves. What’s inherited is relationships and perhaps community standing, which can offer some advantage, but not the kind of multi-generational extractive position that capital accumulation creates.

No financialization. There’s no secondary market in which credits themselves become objects of speculation, because credits aren’t detachable from the underlying labor relationships they represent. The economy stays grounded in actual exchange between actual people rather than in abstract claims traded among financial intermediaries.

No structural pressure toward growth. Because the system doesn’t require returns on accumulated capital, there’s no built-in pressure to expand production indefinitely. The economy can operate at whatever scale meets actual human and ecological needs, rather than being driven to grow to satisfy capital’s need for returns.

What remains: differential recognition for differential contribution. People who contribute more have stronger credit positions, more standing in their communities, and more discretion in what they take on through the credit system. What disappears is the conversion of that recognition into structural power over others.

Example Ledger:

Starting state: Everyone has a balance of 0.

Iris fixes Marcus’s bicycle. The work takes 1 hour. Iris: +1, Marcus: −1 Overall: Iris: +1, Marcus: −1, Yuki: 0, Sam: 0

Marcus bakes bread for Yuki. 30 minutes of labor. Marcus: +0.5, Yuki: −0.5 Overall: Iris: +1, Marcus: −0.5, Yuki: −0.5, Sam: 0

Yuki teaches Sam piano for an hour. Yuki: +1, Sam: −1 Overall: Iris: +1, Marcus: −0.5, Yuki: +0.5, Sam: −1

Sam helps Iris move furniture. 30 minutes of work. Sam: +0.5, Iris: −0.5 Overall: Iris: +0.5, Marcus: −0.5, Yuki: +0.5, Sam: −0.5

0.5 − 0.5 + 0.5 − 0.5 = 0. Total of all balances is 0. The system stays closed.

Everything else in the post builds on top of this basic accounting. The system is just keeping track of who has done what for whom. No money required.

[Click here to go back to the Table of Contents]

♥︎

I came across this theory as a sex worker whose income doesn’t count as income to capitalist institutions. My labor produces real value at real cost to my body and mental health. These are costs I bear privately. That illegibility translates directly into limited access to housing and credit. I recognize the stakes.

An anarchist economy must be built for us all, dedicated to non-domination.

Lastly, a sleepy fox grows tired of hearing her own voice, so please feel free to
comment, ask questions, or run freely with your own extensions.