Stewardship Takes Staff

Across the country, communities are turning to community land trusts to keep homes affordable for good. The promise is permanent, and the people keeping it do remarkable work with lean teams. At the typical CLT, three full-time staff carry that promise for every home in the portfolio.

That figure comes from the 2022 Census of Community Land Trusts and Shared Equity Entities, published by the Lincoln Institute of Land Policy with Grounded Solutions Network researchers. The census counted 314 organizations holding an estimated 43,900 homes, including about 15,600 shared-equity homeownership units. The median organization reported three full-time employees, and 59% reported five or fewer paid staff. Those same people also develop homes, raise funds, support their boards, answer the phone and much more.

Stewardship drives lasting affordability

Stewardship is the part of the model that keeps every other part working. The CLT Technical Manual describes it as four kinds of ongoing work: disclosure, monitoring, support and approval. Each new home adds to that work, and the obligation lasts as long as the adopted ground lease.

The research shows the payoff. At the end of 2010, a year of record foreclosure filings, a national survey of 62 CLTs found 1.30% of CLT homeowners’ mortgages seriously delinquent. The Mortgage Bankers Association reported 8.57% across the market. Foreclosure proceedings ran 0.46% for CLT homeowners and 4.63% market-wide. The study (Thaden, 2011) credited affordability and stewardship together: CLTs that review mortgages, educate buyers before and after purchase, and intervene early when payments slip.

A 2019 study of 58 shared-equity programs and 4,108 homes found an average annual move rate of 2.6%, compared with 6.9% for all U.S. homeowners. Nearly three in four owners stayed six years or longer.

Stewardship also takes time

The same research shows how much of that work rests on a few people. In the 2011 survey, 45% of CLTs reported devoting more staff time to delinquency and foreclosure prevention after 2008. About a quarter held a one-on-one meeting with each homeowner at least once a year. In the 2022 census, nearly a third of programs did not know or did not track whether their homeowners had refinanced. These numbers speak to capacity, never to commitment.

Three people can carry every resale calculation, monitoring letter and board memo by hand for only so long, and the annual check-in is often what gets squeezed. The work that gets squeezed is the work funding rarely covers. Lease and transfer fees are expected to carry stewardship once development money stops, yet 64% of programs cover 20% or less of their operating budget from internal revenue. That puts obligations written into every ground lease and covenant at risk: monitoring can fall behind, approvals can go undocumented, and contact with homeowners can thin out between purchase and resale.

Practitioners know this better than anyone. We believe staff time, alongside land and subsidy, limits how many homes the sector can steward properly. We treat that belief as a premise to test, and the framework below is how we plan to test it.

Where AI fits

The question worth asking is whether AI can return hours to the relationships the research credits for homeowner success. That is why we build Stuard Gneiss (say it nice; Stuard for short), a stewardship system built on Claude and tailored to how each organization actually works. Its entry tier, Stuard Companion, is an AI assistant configured around each CLT’s own ground lease, resale formula and procedures, so it speaks the program’s language from day one. As board policies or formulas change, Stuard changes with them. PermAffai sets it up and keeps it current, so staff never have to build or maintain it themselves.

Our approach rests on three commitments.

1. People decide. Staff review and approve every notice, checklist and homeowner message Stuard drafts before it goes out.
2. Formulas stay fixed. Stuard asks staff for each input, runs the resale formula and purchase option price through tested, version-controlled code that matches the adopted ground lease, and returns a detailed, step-by-step breakdown of the calculation. Staff can move forward with confidence, and every result can be reproduced and audited.
3. Data stays home. Each CLT’s homeowner records stay in that CLT’s own systems. A CLT may choose to share de-identified program data with PermAffai, and PermAffai reports it to sponsors and the field only as combined totals that never identify any single CLT.

Our theory of change

1. Assess. AI readiness assessments map stewardship workflows and record baseline staff hours.
2. Assist. Stuard Companion drafts routine notices, monitoring checklists and policy answers for staff review.
3. Convene. A founding cohort of 10 to 15 CLTs, convened through a sector intermediary once a sponsor commits, with shared baselines.
4. Design. Program design builds stewardship into new programs from the start, with adopted guidelines, SOPs and a locked resale workbook.
5. Outcome. Fewer hours on routine tasks, reproducible resale pricing and more time for check-ins and early outreach.
6. Impact. More permanently affordable homes stewarded well, with homeowners retained and affordability preserved at every resale.

Five metrics we will report

1. Stewardship practice adoption. Board-adopted policies and SOPs, scored against core practices such as mortgage review, lender notice of delinquency and annual homeowner contact.
2. Staff hours saved per deployment. Measured against each organization’s own baseline at 30, 90 and 180 days. We will publish no hours figure before we have measured one.
3. Homes fully stewarded per stewardship FTE. Counts only homes whose obligations are current, alongside all homes per staff member. The gap between the two is the stewardship an organization owes but cannot staff.
4. Resale compliance accuracy and on-time resales. Every resale price checked against the adopted formula, and every closing tracked against the ground lease timeline.
5. Homeowner retention and delinquency. Read against the benchmarks above as a guardrail. Growth that weakens homeowner outcomes does not count as success.

Two of these metrics have no published sector benchmark: resale accuracy and homes per stewardship FTE. Programs track their numbers in different ways, and the Census authors point to broader use of data systems as the way to close that gap. A shared, consented baseline would serve the whole field, whatever tools each organization chooses.

Permanent affordability depends on the people who keep the promise. The homes are built to last. The question now is whether the capacity to steward them will grow alongside them.