The community-owned industrial capital stack — a design problem

The community-owned industrial capital stack

Doctrine page (design problem, not a solved recipe). The capital-without-capture thesis closed the financing question (industrial-scale capital can be raised as non-equity money under an asset lock) but left the category’s second constraint — genuine community ownership — unresolved at scale. This page takes that up as a design problem: which ownership vehicles deliver broad, democratic community control of a $5M–$50M industrial asset, how each survives the financing event, and where the honest state of the art runs out. It proposes a composite; it does not claim a built precedent exists, because none does (see No built precedent).

The problem the thesis handed down
The distinction that does the work: three kinds of “owned”
The vehicles, scored on the two exposures
The recommended composite
No built precedent — the honest state
The lethal joint: the financing event
What would move this from design to doctrine
See also

The problem the thesis handed down

The thesis verdict was qualified yes on financing, unresolved on community ownership. Two exposures survived both its adversarial passes, and they define this page’s job:

Exposure A — the community-ownership gap. Every famous “raise-and-hold without capture” structure (Patagonia, Newman’s Own, Bosch, Zeiss) bought durability by concentrating control in a family or insider trust. They prove the asset-lock, not community ownership. The category requires broad, democratic control with no dominant founder — and that is the unproven half.
Exposure B — the debt-service / foreclosure back door. “Debt ≠ capture” holds only while the borrower performs; on a DSCR default the workout hands a lender governance leverage that can override the no-exit lock. So the ownership vehicle must also be the one that survives a workout.

A community-owned capital stack is therefore not one instrument but an assignment problem: route each exposure to the vehicle that handles it best, and make sure the financing event (the moment most likely to dissolve community control) passes through a layer engineered to absorb it.

The distinction that does the work: three kinds of “owned”

The single most clarifying move is to stop saying “owned” and name who votes. Three distinct animals get conflated, and only the first is the category’s:

Kind	Who controls	Examples	Why it is / isn’t the category
Community-owned	open membership, democratically elected, no dominant founder	(CLTs; member public-benefit nonprofits) — but none at industrial production scale	This is the category’s requirement. Also the rarest.
Producer/worker-member-owned	a closed class — the farmers/workers who supply or staff it	American Crystal Sugar, Organic Valley, Mondragon, Evergreen co-ops	Democratic among members, but exclusionary by design; can turn against labor (American Crystal locked out its own ~1,300 workers, 2011–13) or demutualize and sell (Dakota Growers Pasta, 2002–04).
Founder/steward-trust-owned	a family or self-appointing insider trust	Patagonia, Bosch, Zeiss, Newman’s Own	Strong asset-lock, but the opposite of broad control — durability purchased by concentrating control.

The category lives only in row one. Most large, durable, democratically governed industrial assets that actually exist live in row two — and the moment a row-two structure has to capitalize a big plant, it tends to close its membership (to producers/workers/anchors) or take on equity, which are precisely the doors community control is lost through. That recurring move is the heart of the problem.

The vehicles, scored on the two exposures

(Each individually proven; sourced in the 2026-06-21 thesis research. Scored on: does the canonical form deliver broad community control (A), and does community governance survive a workout (B)?)

Community Land Trust (CLT), land-only, ground-leasing the plant. The only vehicle where broad community election is the default — the classic tripartite board (⅓ leaseholders / ⅓ community residents / ⅓ public-interest) is explicitly anti-faction. Strongest on B: the ground lease makes the CLT a third party to every mortgage with cure rights and workout standing; CLT foreclosure rates ran far below market in 2007–09. Caveat: the lock covers the land, not the building — a completed foreclosure on the building can strip its restrictions, so keep the building in a member-nonprofit and the land in the CLT.
Member-governed public-benefit nonprofit. The most direct statutory route to “members elect the board,” and the strongest no-exit asset-lock of any vehicle (charitable assets can’t pass to insiders even in dissolution — they go to another charity). Failure mode: nonprofits default to self-perpetuating boards, which quietly delete the member vote — broad control must be deliberately locked open, not assumed.
Multi-stakeholder / solidarity co-op (workers + community + consumers as classes). Broad only if a community-supporter class holds real board seats; the defining weakness is faction capture between classes, and seat formulas are bespoke (a warning sign). No special foreclosure protection. Real examples (Weaver Street, Fifth Season) are food/retail, not heavy manufacturing.
Perpetual Purpose Trust with a community-elected steward council. Capable of genuine community control but essentially never used that way — Patagonia’s is family-guided and opaque. Highest founder-syndrome risk; requires a bespoke charter making the steward committee community-elected with no founder override. Only DE/NH/WY/ME have suitable trust law.
Golden share / community veto. A ~1% share (held by a community-governed nonprofit) vetoing sale or mission change. A complement, not a standalone — it prevents exit but does not create control, and courts have declined to enforce it against bankruptcy/creditors, so it is weak exactly where Exposure B bites.
Blended nonprofit-parent + worker-co-op (the Cleveland/Evergreen model). The best-matching real governance blend for a production asset (nonprofit parent over worker co-ops, debt-and-grants-only, “no equity relinquished”). But Green City Growers is food production at single-digit millions, and the co-ops lack full independent control — the parent “controls strategy,” which can reintroduce top-down capture.

The recommended composite

The evidence points to a layered structure that assigns each exposure to the vehicle that handles it best — not to any single instrument:

Land → a CLT (tripartite, member-elected), ground-leasing to the operator. Buys: perpetual community land lock + the foreclosure firewall (Exposure B).
Building + operation → a member-governed public-benefit nonprofit (open community membership electing a tripartite-style board). Buys: broad democratic control (Exposure A) + the strongest no-exit asset lock.
A community-held golden share in an independent member-nonprofit, vetoing sale / mission change. Buys: belt-and-suspenders against voluntary exit (scoped to not block the tax-credit unwind).
Tax-credit capital → a single-purpose for-profit QALICB beneath the nonprofit, which the NMTC/HTC investor controls for the 5–7-year compliance window and then puts back at nominal value, leaving clean community ownership. Buys: access to NMTC/HTC without contaminating the community-owned layer (a nonprofit cannot pass the credits directly — it must interpose a for-profit subsidiary).
Service the debt conservatively (low LTV, DSCR headroom) from production surplus — because Exposure B is ultimately a revenue problem: thin mission-priced margins are what trip the covenant.

Each component is individually proven. The composite assigns land-lock to the CLT, broad control to the member-nonprofit, anti-exit to the golden share, and the capture-prone tax-credit window to a disposable subsidiary that unwinds cleanly.

No built precedent: the honest state

There is no documented case of a broadly, democratically community-owned manufacturing or processing facility financed at $5M–$50M and still community-controlled years later. The closest real examples each fall short on a nameable axis:

American Crystal Sugar — large (5 plants), democratic, durable, but a closed producer co-op that locked out its own workers (2011–13).
Organic Valley / CROPP — $1.2B, one-member-one-vote, owns processing — but closed farmer-membership, not community.
Dakota Growers Pasta — a grower-built mill at real scale that demutualized and was sold (the textbook exit failure).
Evergreen / Green City Growers — best community-blend governance, but food production at single-digit millions, co-ops without full independent control.
Mondragon (Ederlan, Maier, Loramendi) — proves worker-governed manufacturing scales, but Spain, and worker- not community-owned.
CHCA — democratic governance at a large enterprise (~1,100 worker-owners elect 8/13 board seats), but home-care services, not a production asset.

Building this category means assembling proven parts into a configuration no one has yet held together at industrial scale.

The lethal joint: the financing event

The pattern across every shortfall is the same: the moment a structure has to capitalize a large physical plant, it tends to close its membership or admit equity — and closure and equity are exactly how community control is lost or sold. The financing event is the joint where the category historically breaks. The composite above is, in effect, a hypothesis about how to pass through that joint without closing or selling: keep equity out (debt + grants + self-exiting tax credits), keep membership open (member-nonprofit + CLT), and make the one unavoidable for-profit layer disposable by design. Whether that holds in a real build through a real downturn is the open empirical question the doctrine now rests on.

(Doctrine note, resolved 2026-06-21: passing through this joint with a self-exiting for-profit scaffold does not violate the “nonprofit structure from day one” invariant — that invariant governs principal ownership and the asset lock, not the financing instruments. Locked in THEORY.md and the-category.md. The for-profit layer is permitted only as a subordinate, disposable scaffold beneath a nonprofit that already owns and controls.)

What would move this from design to doctrine

A built existence proof: one broadly community-owned, member-elected, no-exit industrial production facility at $5M–$50M that serviced its covenants from production surplus through a downturn and unwound its tax-credit layer to clean community ownership. That single case promotes this page from design problem to doctrine.
A counterexample worth hunting first: an obscure open-membership food-processing facility (not a closed producer co-op) would be the most likely place a real precedent hides — worth turning over before claiming none exists.
A falsifier: evidence that open community membership and conservative industrial debt are jointly unsatisfiable — i.e., that broad membership can never assemble or service plant-scale capital without closing or taking equity — would push the category back toward the thesis’s “refuted” branch.