The famous number from the Legal Services Corporation’s 2022 study is that low-income Americans get no or insufficient help for 92% of their substantial civil legal problems. It is the number everyone quotes, and it is the wrong one to design against, because it mixes together people who never sought help, people who did not know they had a legal problem, and people who asked and were refused.
The number to design against is this one, from the same study:
LSC-funded organizations “must turn away 1 out of every 2 (49%) requests they receive due to limited resources.”
An estimated 1.4 million civil legal problems brought to those organizations each year go unresolved — 71% of everything presented to them.
Those people found legal aid. They got through. They asked. And half of them were told no, not because their case lacked merit, but because there was no one to give it to.
That is not an awareness problem or a demand problem. It is a capacity problem, which means it is a problem with a supply chain — and supply chains can be engineered. This piece proposes a specific mechanism, in enough detail to build, and is honest about where it breaks.
“Limited resources” is true and useless
Every discussion of this ends at “legal aid needs more money,” which is correct and has produced approximately no change in the turn-away rate for decades. The reason is that money is necessary and not sufficient. To become capacity, money has to convert into two things, and the way we currently give it is bad at producing either.
Input one: salaried attorneys, which requires predictable money
A legal aid organization cannot hire a permanent staff attorney against a one-year grant. Nobody can. Hiring an FTE is a multi-year commitment — recruitment, training, a caseload that outlives any single funding cycle — and an executive director looking at funding that may or may not renew in eleven months will not make it. They will spend the money on contract work, on short-term projects, on anything reversible.
So episodic philanthropy converts into permanent capacity at a terrible exchange rate. A firm’s $200,000 check does not become a lawyer. Five consecutive years of $200,000 would have, but nobody promised the five years, so it did not.
The form of the money matters as much as the amount. This is the single most important fact in legal aid finance and it is almost never the thing anyone tries to fix.
Input two: supervision in the specific practice area
The second constraint is the one that surprises people, and it inverts the usual assumption about volunteers.
A volunteer lawyer from outside the practice area is not free capacity. They need training, malpractice coverage, a mentor to call, and someone competent to review their work — and that someone is the organization’s scarce housing or family or benefits specialist. Every hour that specialist spends supervising is an hour not spent on cases.
Which means adding untrained volunteers to a supervision-constrained organization can reduce net output. Legal aid directors know this and are too polite to say it to the firms offering help. It is why the enthusiastic pro bono program sometimes produces less than everyone expected, and why “we sent forty associates” is not the gift it sounds like.
Supervision in the relevant specialty is the actual bottleneck. Money buys it only if the money is stable enough to hire someone senior, which returns us to input one.
Why the current mechanism fails at both
Rule 6.1 of the Minnesota Rules of Professional Conduct says a lawyer “should aspire to render at least 50 hours of pro bono publico legal services per year,” that the responsibility “is not intended to be enforced through disciplinary process,” and — in the comment — that a lawyer “may discharge the pro bono responsibility by providing financial support to organizations providing free legal services to persons of limited means.”
So the sanctioned path for a firm that cannot staff the work is to write a check. Fine. Now look at what that check is:
- Annual and revocable. No commitment beyond this year. Cannot be hired against.
- Unrestricted or loosely restricted. Funds general operations, which is genuinely useful and completely invisible.
- Specialty-blind. The firm’s own expertise — the thing it actually has in abundance — is not transmitted at all.
- Unverifiable. Neither the firm nor anyone else can say what was delivered as a result.
The firm gets a thank-you letter. The organization gets money it cannot build with. The specialty knowledge, which is the scarce input, stays at the firm.
The mechanism
A forward, specialty-tagged, delivery-verified purchase commitment, bundling capital with supervision.
Concretely, a participating firm signs a contract that says:
Over 36 months, we will purchase up to 1,500 hours of delivered legal service in housing defense, at $85 per delivered hour, funds escrowed quarterly and released on verified delivery. We additionally commit 120 hours per year of supervising-attorney time from our own employment and litigation partners, to be scheduled with the provider.
Five design elements, each solving something specific.
1. It is forward and multi-year — this is the whole point
A provider holding a signed, escrowed, 36-month purchase commitment has something a check is not: a bankable asset. They can show it to a board, to an auditor, to a lender. They can hire against it, because the revenue is contracted rather than hoped for.
This single change is what converts philanthropy into FTEs. Everything else in the design is secondary to it.
2. Funds are escrowed and release on verified delivery
The money is committed up front and held. It transfers as service is delivered and attested.
This makes the instrument work in both directions. The funder is not writing a blank check into the future — funds release only against delivery. The provider is not relying on a promise — the money already exists and is held for them. A commitment that is escrowed is a commitment a lender will lend against; a pledge is not.
3. Delivery is attested by two parties, never one
An hour is minted as deliverable only when both the delivering attorney and the provider organization attest to it, with the organization confirming the client met the income threshold. Self-attestation counts for nothing. Supply is hard-capped by real service.
4. It is tagged by practice area — and the tag does two jobs
It lets the firm bundle supervision, which is the actual fix. This is the part I think is genuinely underexploited. An employment firm that funds wage-claim capacity can also supply the supervising attorney for that work. It is not asking its people to practice housing law; it is contributing exactly the expertise it has, into the exact constraint that limits output. Capital and supervision arrive in the same practice area, from a firm that has both.
The exchange rate here is favorable in a way people underestimate. One hour of expert supervision unlocks several hours of competent volunteer or junior-staff work. A firm giving 120 supervision hours a year may be adding more capacity than its dollars do.
And the clearing prices become a map. If housing credits clear at $110 and consumer-debt credits at $70, that is information nobody currently has: capacity in housing is scarcer relative to demand. Providers, funders, and law schools can all act on it. We currently allocate legal aid by grant cycle and institutional history, using survey data that is years stale. A price is faster and denser, and it is generated as a byproduct of transactions that were happening anyway.
5. It is non-transferable and retires on delivery
A commitment may be assigned to another funder only with the provider’s consent — so a firm that loses its budget can hand off rather than default. It cannot be resold at a markup, held for appreciation, or traded.
That constraint is doing legal work, and it is worth being explicit about why.
The securities question, answered
Under SEC v. W. J. Howey Co., 328 U.S. 293 (1946), an investment contract requires an investment of money, in a common enterprise, with an expectation of profits, derived from the efforts of others. All four elements must be present.
The first, second, and fourth are arguably met here. The third is not, and cannot be, by construction: an instrument that cannot be resold, cannot appreciate, pays no yield, and is consumed on delivery is not bought in expectation of profit. It is bought the way a prepaid service contract is bought. No investment contract, no registration obligation.
This is why non-transferability is a design constraint rather than a preference. Add a secondary market and every element lines up, and whoever structured it has conducted an unregistered offering. Any version of this with tradeable, appreciating instruments is unlawful, and the lawyer who built it has a problem much larger than a failed project. The non-speculative version is the only lawful version, and it is also the version that works — speculation would add nothing to a mechanism whose entire purpose is to convert money into hired staff.
One more rule to clear. Rule 5.4 prohibits sharing legal fees with a nonlawyer. No fee is being divided here: the delivering attorney’s compensation is unaffected by whether a commitment is purchased, and the payment runs to an organization as a subsidy. Structurally this is a grant with delivery conditions, and conditional grants to legal aid have never been treated as fee-sharing. That analysis is sound but it is not free — get an advisory ethics opinion before the first dollar moves.
Does it need a blockchain? Here is the test
I am going to answer this decisively rather than leaving it open, because a proposal that will not commit is not a proposal.
No — not at first. Yes — possibly at scale, for exactly one function.
The function is multi-party escrow without a central custodian. Run this in one state with three providers and five firms and the escrow is trivial: a title company or a bank holds a few million dollars in a segregated account and releases on written attestation. Boring, cheap, well-understood, and completely adequate.
Now scale it. Forty firms, thirty providers, twelve states, $50 million of committed forward purchases. Somebody is holding that float. Whoever holds it has fiduciary exposure, earns the interest, controls release timing, and becomes a single point of capture — and every candidate for the role (a bar foundation, an access-to-justice commission, a nonprofit intermediary) has institutional interests in the numbers they would be certifying. That is a genuine structural problem, and programmatic escrow with multi-signature release is a genuine answer to it: funds are committed and released by rule, and no participant holds anyone else’s money.
So the threshold test: below roughly one state and a few million dollars, use a database and an escrow agent. Above it, when the custodian problem becomes real, the ledger earns its place.
Build the database version first. Not as a compromise — as the correct sequencing. Every part of this proposal that matters (forward commitments, escrow, practice-area tags, two-party attestation, clearing prices, supervision bundling) works identically either way. The ledger is a scaling decision about custody, and you should not make scaling decisions before you have scale.
How it breaks
Four failure modes. Three have controls. The fourth is the hard one.
Provider under-delivers. Capacity forecasts miss. The contract needs shortfall terms up front: unused escrow rolls into the next period, or refunds pro rata at the funder’s election, and under-delivery is not a default unless it is chronic. Get this wrong and providers will not sign, because no executive director will accept a contract that punishes them for a hiring market they do not control.
Attestation fraud. Organizational countersignature, random sampling audits of underlying files, permanent loss of minting authority for inflation, and clawback of released funds. This is the least glamorous part of the system and it is where I would spend most of the administrative budget. An inflated credit is worse than no credit, because it converts inaction into a report of action.
Confidentiality. A credit is evidence that an identified person received legal help. Rule 1.6 has no pro bono exception. The registry records the organization’s attestation, a practice-area tag, an hour count, and a period — never a client, a matter, or anything from which one could be inferred. This means the granularity everyone will want (outcomes, demographics, case results) is exactly the granularity you cannot have, and any version of this that promises outcome data is promising a confidentiality breach.
And the hard one: hours are a gameable unit. If a commitment is satisfied by counting hours, the rational provider maximizes hours, and the cheapest hours are brief-advice contacts, not the two-day trial that actually saved a tenancy. Volume displaces difficulty.
I do not have a clean answer. The partial ones: weight hours by case type against a published schedule; cap the share of any commitment satisfiable by brief advice; require a mix. All of them add administrative complexity and all of them invite argument about the weights. This is the design problem I would most want someone smarter than me to solve before this went past a pilot, and anyone who tells you it is easy has not run a legal aid intake line.
The political risk nobody wants to name
If private money flows in visibly and measurably, a legislature looking for savings may conclude that legal aid needs less public funding.
This is not hypothetical; it is the standard fate of every successful private supplement to a public good. And this mechanism makes it worse in one specific way: by making private contribution legible and quantified, it hands an appropriations committee a number to point at.
The mitigation is contractual and should be non-negotiable: commitments are conditioned on maintenance of effort, and a firm’s obligation terminates if public funding for the covered service falls below a baseline. That makes the private money a supplement by construction rather than by hope. It also means the firms have a direct financial interest in defending appropriations, which turns a constituency that currently writes checks into a constituency that shows up.
The pilot
Enough to test the thesis without betting anything large.
- One state, one practice area. Housing defense — highest volume, clearest outcomes, most acute turn-away rate.
- Three providers, five firms. Small enough that the escrow is a bank account and the registry is a spreadsheet with an auditor.
- 24-month commitments, $1–2 million total.
- Supervision bundled from day one, because if the supervision component does not work, the money component alone is just a better-structured grant.
Success metrics, defined before anyone signs:
- Net new FTEs hired against commitments. The whole thesis. If providers do not hire, the forward structure did not do its job and the mechanism is a rounding error on ordinary philanthropy.
- Turn-away rate at participating providers, against their own baseline.
- Whether participating firms’ total giving rose or merely moved. If firms redirect existing philanthropy into commitments and give no more in aggregate, this improves targeting and adds no capacity — a real but much smaller result, and one worth knowing honestly.
- Supervision hours actually delivered versus committed. Firms over-promise partner time. Measure it.
What would falsify the whole idea: providers sign commitments and still do not hire, because the uncertainty they are managing is not funding uncertainty but something else — talent pipeline, salary competition, physical space. If that is what the pilot shows, then the diagnosis at the top of this piece is wrong, and the right response is to publish that and stop.
Why bother
Because the alternative is another decade of the same sentence. Legal aid needs more money; here is a check; the turn-away rate is 49%.
Money is not converting into capacity, and the reason is structural rather than moral. Nobody in this system is behaving badly. Firms give generously. Organizations spend carefully. The conversion is just terrible, because a one-year unrestricted check is close to the worst possible instrument for buying a permanent staff attorney and transmits none of the expertise the giver actually has.
The proposal here is not charity redesigned to feel better. It is a supply-chain fix: make the money predictable enough to hire against, tag it so expertise travels with it, verify the delivery, and let the prices tell everyone where the shortage is worst.
I do not know that it works. I know the current mechanism does not, and I know precisely which two constraints it fails to relieve. That is enough to justify a pilot, which is all this is asking for.
Sources
- Legal Services Corporation, The Justice Gap: The Unmet Civil Legal Needs of Low-income Americans (2022) (49% of requests turned away for limited resources; 1.4 million problems unresolved annually, 71% of those presented)
- Minn. R. Prof. Conduct 6.1, Rule 5.4, Rule 1.6
- SEC v. W. J. Howey Co., 328 U.S. 293 (1946)
Rules quoted were verified against the Revisor of Statutes rather than cited from memory. This is a design proposal, not legal advice, not securities advice, and not ethics advice. Nothing described here exists, and it is not an offer or solicitation of any instrument. Anyone building it should obtain a securities opinion and an advisory ethics opinion first.