Back to Articles
    Leadership & Strategy

    Should Your Nonprofit Lease or Buy? Using AI to Model the Decision

    Almost every nonprofit eventually faces a version of this question, and almost every nonprofit answers it emotionally. Owning feels like permanence, stability, and proof that the organization has arrived. Leasing feels temporary, like money handed to someone else. Neither feeling is analysis. The lease-versus-buy question is a capital allocation decision with a twenty-year tail, and it deserves the same rigor a board would apply to launching a new program or taking on debt for anything else.

    Published: September 10, 202616 min readLeadership & Strategy
    A nonprofit building viewed as a strategic and financial asset rather than an operations problem

    This article is about a decision, not about a building. That distinction matters, because we have written twice before about space and both of those articles were about something else. Our guide to AI in nonprofit facility management covers energy optimization, predictive maintenance, and the day-to-day work of running a property you already control. Our guide to space and facility planning covers how to configure offices, classrooms, shelters, and community hubs so they serve programs well. Both assume you have space and are trying to run it better.

    This one assumes nothing. It is about whether your organization should hold real estate at all, and on what terms. That is a capital question, a governance question, and a risk question before it is ever an operations question. The consequences of getting it wrong do not show up as a higher utility bill. They show up as a mortgage payment that squeezes program budgets for fifteen years, a board that discovers it has become a landlord, or a lease renewal signed under time pressure at terms nobody modeled.

    The financial comparison at the heart of the decision is also widely misunderstood. Organizations compare a monthly rent figure against a monthly mortgage payment, notice that the mortgage is similar or lower, and conclude that buying is obviously better. That comparison is wrong on both sides. It understates leasing by ignoring escalators, operating expense pass-throughs, and the cost of a fit-out you will walk away from. It understates owning by ignoring capital reserves, deferred maintenance, insurance, staff time, and the opportunity cost of the equity you just locked into a building.

    What follows is a walk through the decision in the order a leadership team actually encounters it: recognizing when the question is genuinely live, building an honest total cost of occupancy model, applying the affordability tests a board should insist on, understanding the tax consequences of ownership including the landlord problem, modeling under real uncertainty about headcount and program mix, evaluating a donated property offer, assembling the board packet, and planning the exit before you sign anything. AI is genuinely useful in several of these places and genuinely unhelpful in others, and the article is explicit about which is which.

    When the Question Is Actually Live

    Lease-versus-buy is not a question you should be asking continuously. It is expensive to analyze properly and the analysis goes stale. It becomes live at specific moments, and recognizing those moments early is most of the advantage, because every one of them has a deadline attached and the organizations that lose money are the ones that start thinking about it too late.

    The most common trigger is a lease approaching renewal. The mistake here is timing. Executives tend to engage about ninety days out, which is precisely when they have no leverage, because the landlord knows you cannot realistically relocate a program in three months. Serious lease decisions should begin twelve to eighteen months before expiration for anything larger than a small office suite. That window gives you time to tour alternatives, obtain real competing proposals, and negotiate from a position where walking away is actually credible. A renewal negotiated with credible alternatives in hand routinely produces terms a renewal negotiated under deadline pressure never will.

    Growth is the second trigger, and it is the one most likely to produce an expensive mistake, because growth is exciting and excitement is a poor input to a twenty-year commitment. The question to ask is whether the growth is structural or funded. Structural growth reflects a durable shift in demand and a diversified revenue base. Funded growth reflects a large multi-year grant that will end. Buying a building on the strength of funded growth is how organizations end up with an asset sized for a program that no longer exists.

    A donated or below-market property offer makes the question live immediately and usually on someone else's timeline. So does a merger, where two organizations suddenly hold two leases, or a lease and a building, and must decide what the combined entity keeps. The real estate question in a merger is frequently the largest single line item in the integration plan and the one least analyzed during negotiations, which is why property should appear explicitly in merger due diligence rather than being treated as a detail to work out afterward.

    Finally, the question becomes live when a program ends or shrinks materially. This is the trigger organizations handle worst, because nobody wants to be the person who says the building is now too big. Space is sticky. A nonprofit that has lost a third of its headcount over four years and is still paying for the footprint it had at peak is paying a program cost disguised as an occupancy cost, and that money is coming from somewhere.

    Triggers that put the question on the board agenda

    Each carries a deadline, and late engagement costs money

    • A lease expiring within eighteen months
    • Sustained growth that has outrun the current footprint
    • An offer of a donated or below-market property
    • A merger that leaves the combined entity with duplicate space
    • A major program ending, shrinking, or moving to a different model
    • A landlord selling the property or proposing a substantial rent increase
    • A capital campaign under discussion where a building is a candidate use

    Total Cost of Occupancy: The Only Honest Comparison

    The right unit of comparison is the total cost of occupying space over the period you realistically expect to need it, expressed in present value terms so that money spent in year twelve is not treated as equivalent to money spent today. Anything less than that is a comparison of two numbers that do not mean the same thing. Building this model is the single most valuable analytical act in the entire decision, and it is also where most organizations stop short.

    On the lease side, start with base rent and then keep going. Rent escalators compound, and the compounding is larger than intuition suggests. A three percent annual escalator means year ten rent is roughly thirty percent above year one, and over a fifteen year horizon the cumulative effect reshapes the comparison entirely. Then add the operating expense structure. Under a triple net or modified gross lease, you are responsible for some combination of property taxes, insurance, and common area maintenance, and those pass-throughs are frequently uncapped or capped loosely. Tenants can negotiate caps on controllable expenses and audit rights over the landlord's reconciliation, and a nonprofit that signs without either has accepted an open-ended obligation it cannot forecast.

    Add the costs people forget: the tenant improvement allowance shortfall, since the landlord's contribution rarely covers a full fit-out and the balance is your capital; moving costs; technology and cabling; furniture; the restoration obligation at the end of the term if the lease requires you to return the space to its original condition; and any personal guarantee or letter of credit that ties up cash for the life of the lease. Fit-out costs in particular have risen sharply, and an organization budgeting a fit-out from a number it remembers from a decade ago will be badly wrong.

    On the ownership side, the model is longer and the omissions are more dangerous. Debt service is the obvious item and the smallest share of the surprise. Add property insurance, which for a nonprofit occupying and possibly sharing a building may be more complex and more expensive than leadership expects; utilities, which you now pay in full rather than as an allocation; routine maintenance and janitorial; grounds and snow removal; and the staff or contractor time to manage all of it.

    Then add the two items that separate a serious model from a hopeful one. The first is a capital reserve, funded annually, for the systems that will fail on a schedule you can predict even if you cannot predict the exact year. Roofs, HVAC, elevators, parking surfaces, electrical panels, and building envelopes all have replacement cycles, and an ownership model that does not fund them is not a model, it is a deferral. The second is the deferred maintenance you are inheriting at purchase, which a building condition assessment will quantify and which is frequently large enough on its own to change the answer.

    Finally, account for the opportunity cost of capital. A down payment plus closing costs plus immediate capital repairs represents cash that leaves the balance sheet. If that money would otherwise have funded an operating reserve, seeded a new program, or simply remained liquid during a funding disruption, the building has a cost that never appears on any invoice. This is the point where an ownership case that looked comfortable often stops looking comfortable, and it is the point most enthusiastic building committees skip.

    Lease side of the model

    What the monthly rent figure leaves out

    • Base rent with the full escalator schedule compounded
    • Operating expense pass-throughs, taxes, insurance, and CAM
    • Fit-out cost above the tenant improvement allowance
    • Moving, cabling, furniture, and signage
    • Security deposit or letter of credit tying up cash
    • Restoration obligation at the end of the term
    • Renewal rent reset to market, not to your escalated rate

    Ownership side of the model

    What the mortgage payment leaves out

    • Down payment, closing costs, and financing fees
    • Deferred maintenance inherited at purchase
    • Annual capital reserve for roof, HVAC, and building systems
    • Property insurance, liability, and any flood or seismic coverage
    • Utilities, janitorial, grounds, and routine repairs in full
    • Staff or contracted time to manage the property
    • Opportunity cost of the capital now sitting in the building

    The Affordability Tests a Board Should Apply

    A total cost of occupancy model tells you what something costs. It does not tell you whether your organization can carry it. Those are separate questions, and the second one belongs squarely to the board. The useful framing for a board is not "can we make the payment in a normal year" but "what happens to this obligation in a bad year," because a lease and a mortgage are both fixed obligations that do not shrink when revenue does.

    Start with occupancy as a share of total expenses, and look at the trend rather than the point. An occupancy burden that rises steadily as a share of the budget is squeezing programs whether or not anyone has named it. Then look at debt service coverage, the ratio of cash available from operations to annual principal and interest. Lenders will run this calculation regardless, and many commercial lenders want to see coverage comfortably above one, with a cushion. A board should run the same calculation under a revenue decline scenario rather than accepting the ratio at budget, because the ratio that matters is the one in the year you lose a major contract.

    Liquidity is the test that catches organizations most often. Months of operating reserve, measured after the down payment and initial capital repairs have left the balance sheet, is the number that determines whether a building becomes a source of stability or a source of fragility. An organization with four months of reserve that spends three of them on a down payment has not acquired an asset, it has acquired an obligation with no margin behind it. Lenders frequently impose covenants requiring minimum reserve levels precisely because they have watched this happen, and a board should apply the constraint before the lender does.

    Revenue concentration deserves a specific look. A building financed on the strength of government contract revenue carries the risk profile of those contracts, including payment delays, reimbursement timing, and non-renewal. Working through cash flow forecasting under realistic delay assumptions is a more honest test of whether an organization can carry debt than an annual budget ever is, because annual budgets smooth away exactly the timing problems that cause trouble.

    Finally, be clear about what a capital campaign actually commits you to. A campaign that raises the purchase price does not make the building free. It makes the acquisition free and leaves you with every ongoing cost in the ownership column, permanently, funded from operations. Campaigns that fund a building without also funding an endowment or reserve for its maintenance are common and they transfer a recurring expense onto an operating budget that was not sized for it. If a building is a candidate use of campaign proceeds, run the feasibility analysis on the full obligation rather than on the sticker price, and say plainly in the case for support what the ongoing costs will be and where they come from.

    Tests to run before committing

    Each run at budget and under a downside scenario

    • Occupancy cost as a share of total expenses, with trend
    • Debt service coverage under a material revenue decline
    • Months of operating reserve remaining after closing
    • Concentration of the revenue that services the debt
    • Whether any loan covenants constrain future borrowing or spending
    • Source of funds for ongoing costs after a campaign closes

    Signals to slow down

    Patterns that precede a difficult ownership experience

    • The case rests on rental income from tenants not yet identified
    • No capital reserve line appears in the ownership model
    • The building condition assessment has not been commissioned
    • A single funder or contract services most of the debt
    • The timeline is set by the seller or donor rather than by the board
    • Nobody has modeled what happens if the campaign underperforms

    Property Tax Exemption and the Landlord Problem

    Ownership brings tax consequences that leasing does not, and they are more consequential for organizations that intend to rent out part of the building, which is a very common element of the ownership case. The plan is usually some version of "we will occupy two floors and lease the third to cover the debt service." That plan is workable, but it changes your tax position in at least two distinct ways and neither is automatic.

    The first is property tax exemption, which is a matter of state and local law rather than federal law, and which varies considerably across jurisdictions. Exemption generally attaches to the charitable use of the property, not to the tax status of the owner. Some states require exclusive charitable use; others permit partial exemption apportioned to the share of the property used charitably. An organization that owns a building and leases part of it to a commercial tenant should expect the assessor to apportion the exemption accordingly, and should expect that failing to report the change can result in a retroactive reassessment with penalties. This is a question for local counsel before purchase, not a question to resolve after the first assessment notice arrives.

    The second is unrelated business income tax at the federal level. The general rule is favorable: under Internal Revenue Code section 512(b)(3), rent from real property is normally excluded from unrelated business taxable income. The exceptions are where organizations get caught. Rent may lose the exclusion where substantial personal services are provided to the tenant, where more than half the rent is attributable to personal property rather than real property, where the tenant is a controlled entity, or where the property is debt-financed.

    That last exception deserves particular attention because it interacts directly with the financing plan. Under Internal Revenue Code section 514, income from debt-financed property held to produce income is subject to unrelated business income tax in proportion to the debt on the property. Property substantially all of whose use is substantially related to the organization's exempt purpose is excluded from the debt-financed rules, which is why the portion you occupy for programs is generally fine. The portion you mortgage and then rent to a commercial tenant is a different analysis, and an ownership model that projects rental income without accounting for potential tax on that income is projecting a number that may not survive contact with the return.

    None of this makes renting out space a bad idea. It makes it a decision with a compliance tail. Someone at the organization will need to track the rental relationship, file what needs filing, maintain the apportionment documentation the assessor will want, and understand the lease well enough to know whether the services you provide the tenant have crossed a line. That is real recurring work, and it belongs in the model as a cost rather than being assumed away.

    A Building Is a Second Business

    The financial model captures cash. It does not capture what ownership does to an organization's attention, and for small and mid-sized nonprofits that is frequently the larger cost. A building generates a continuous stream of decisions that must be made by someone with authority, and in an organization without a facilities function that someone is the executive director.

    Consider what actually arrives. Contractor selection and oversight. Vendor contracts for elevators, fire systems, and landscaping. Insurance renewals with coverage questions nobody on staff is equipped to evaluate. Code compliance and accessibility obligations. Permits. Snow. A water intrusion event at eleven at night. A tenant complaint. An HVAC failure during a heat wave in the middle of a program. None of these are catastrophic individually. Collectively they consume executive attention that was previously going to fundraising, program quality, and staff development, and the displacement is rarely measured because the displaced work simply does not happen.

    Becoming a landlord compounds this substantially. A program-driven organization that leases space to tenants has taken on a genuinely different business with its own skill requirements: marketing vacant space, screening tenants, negotiating and enforcing leases, collecting rent, handling a nonpaying tenant, managing build-outs, and absorbing vacancy when a tenant leaves. Boards approve this plan on the strength of a projected rent roll and rarely ask who is going to do the work or what happens during the six months a space sits empty. The rental income in the model is gross; the work behind it is not free, and vacancy is not hypothetical.

    There is also a governance effect worth naming. Once an organization owns a building, the building acquires constituents. Board members become attached to it. Donors who gave toward it have expectations about it. Staff have preferences about their space. Decisions that should be made on program grounds start being made on building grounds, and the question "should we still be in this location" becomes progressively harder to ask. Organizations report that the building outlived the strategy far more often than they report that the strategy outgrew the building.

    None of this is an argument against owning. Ownership delivers real advantages: cost predictability over long horizons, protection from displacement in a market where mission-aligned space is scarce, control over configuration, the ability to build purpose-specific facilities, and an asset on the balance sheet that can support future borrowing. The argument is that these advantages should be weighed against a fully specified version of the costs, including the ones that never appear on an invoice, and that the weighing should happen before anybody falls in love with a specific property.

    Modeling Under Uncertainty, Not Modeling a Forecast

    The deepest problem with lease-versus-buy analysis is that it requires a fifteen-year view of an organization that cannot confidently describe its next three years. Headcount is uncertain. Hybrid work patterns are still settling. Program mix shifts with funding. Nobody knows what their in-person attendance will look like in 2034. A model that produces a single number is quietly pretending otherwise, and boards read single numbers as predictions.

    The fix is to build the model once and run it many times. Define three or four coherent futures rather than varying one input at a time. A contraction scenario where a major funding source ends and headcount falls by a quarter. A stable scenario reflecting the current trajectory. A growth scenario where a new contract adds staff and service hours. A remote-heavy scenario where the organization moves to distributed work and needs a fraction of its current desk count but more meeting and convening space. Then ask which option is least bad across all of them, rather than which option is best in the expected case. Our guide to scenario planning with AI covers how to construct scenarios that are genuinely distinct rather than variations on a theme.

    The hybrid work variable deserves specific handling because the planning basis itself has changed. Organizations used to size space against total headcount. The relevant figure now is peak concurrent occupancy, meaning the number of people on site on the busiest day, which for many organizations is substantially lower than headcount and which behaves differently as the organization grows. A nonprofit that sizes a purchase against total headcount when attendance has settled into a three-day pattern is buying space it will heat, insure, and maintain for years without using. Utilization data from your existing space, even rough badge or sign-in data, is worth more here than any benchmark.

    Sensitivity analysis is the companion to scenarios and answers a different question: which assumptions actually drive the outcome. Vary the discount rate, the escalator, the maintenance reserve percentage, the holding period, and the terminal value of the property, and observe which ones move the answer. Frequently the decision turns almost entirely on two or three inputs, and knowing which ones tells leadership where to spend its remaining diligence budget. If the entire case depends on an appreciation assumption, that is worth knowing before the board votes, because appreciation is the assumption nobody can substantiate.

    One structural point is worth stating plainly. The longer you expect to occupy a space, the more the economics favor owning, because acquisition costs amortize across more years and escalators have more time to compound against the lease. The shorter and less certain your horizon, the more valuable flexibility becomes, and flexibility has a price you pay in rent. That is the real trade being made. Framing it that way to a board is more useful than presenting a net present value comparison that appears to settle the question arithmetically.

    Scenarios worth running

    Test each option against all of them, not just the expected case

    • Contraction: a major funder ends, headcount falls materially
    • Stable: current trajectory continues with normal variation
    • Growth: new contracts add staff, hours, and service volume
    • Distributed: remote-heavy staffing, fewer desks, more convening space
    • Stress: a capital system fails early and the reserve is short
    • Vacancy: the projected tenant income does not materialize

    The Donated Building, Which Is Sometimes a Liability

    A donor offers your organization a building. The instinct is gratitude and acceptance, and the instinct is sometimes wrong. Real estate is the one gift category where saying yes can cost more than saying no, and the organizations that have learned this learned it expensively. The correct posture is warm appreciation combined with a documented diligence process that runs on your timeline, not the donor's.

    Begin with the question of why the property is being given away. There is often a reason, and a donor unwilling to explain it is telling you something. Deferred maintenance, environmental contamination, a difficult location, unresolved title issues, restrictive covenants, deed restrictions on use, an existing tenant with rights, or simple unmarketability are all common. A property that would sell readily usually gets sold, and the proceeds donated, which is frequently the better outcome for everyone including the donor.

    Diligence on a real estate gift looks much like diligence on a purchase, and should not be shortened because the price is zero. A title report. A survey. An environmental site assessment, since contamination liability can attach to an owner regardless of who caused it and remediation costs can dwarf the property value. A building condition assessment quantifying deferred maintenance and remaining useful life of major systems. A review of carrying costs including taxes, insurance, utilities, and security for a vacant building. Confirmation of any deed restrictions, easements, or reversionary interests. And an honest assessment of marketability if you intend to sell rather than occupy.

    Carrying cost is the trap that catches organizations that accepted a property intending to sell it. A vacant building generates expenses immediately and revenue never, and if it takes eighteen months to sell in a soft market you have funded eighteen months of taxes, insurance, utilities, security, and maintenance on an asset producing nothing. Ask specifically what happens if the property does not sell within a year, and have an answer before you accept rather than after.

    This is what a gift acceptance policy is for, and a nonprofit that does not have one should write it before a real estate offer arrives rather than during one. The policy should state that real property gifts require board approval, specify the diligence required, name who has authority to decline, and reserve the organization's right to decline any gift without explanation. Some organizations require that property be transferred into a single member limited liability company to insulate the organization from liabilities attached to the property, which is worth discussing with counsel. The policy is most valuable as cover: it lets a development director say the process requires these steps rather than saying the organization is suspicious of the donor's motives.

    A below-market purchase offer or a long-term nominal-rent lease from a partner organization deserves the same scrutiny, with one addition. Favorable terms from a partner frequently come with informal expectations about programming, shared space, or governance influence. Those expectations should be written down in the agreement, because unwritten expectations attached to a property arrangement have a way of surfacing several years later at the least convenient moment.

    Diligence before accepting real property

    Run the full process regardless of the purchase price

    • Title report, survey, and confirmation of liens or easements
    • Environmental site assessment before title transfers
    • Building condition assessment with system-by-system remaining life
    • Deed restrictions, reversionary clauses, and use limitations
    • Zoning confirmation that your intended use is permitted
    • Twelve to twenty-four months of carrying cost if it does not sell
    • A written gift acceptance policy that permits declining

    Where AI Actually Helps, and Where It Does Not

    AI is useful in this decision for a specific reason: most of the analytical work is structured modeling, document comprehension, and drafting, and all three are things current tools do well when a person supplies the inputs and checks the output. What AI cannot do is supply judgment about a local market or absorb liability for a wrong answer, and the difference between those two categories is the difference between a tool that saves a month of work and a tool that produces a confident mistake.

    Building and stress-testing the occupancy model. This is the highest-value use. Describe the lease terms, the purchase terms, the financing, and the assumptions, and have a model build a full cash flow schedule across both options with present value comparison. The genuine advantage is not speed, it is completeness. A model prompted to build a comprehensive total cost of occupancy comparison will include line items your spreadsheet forgot, because it is working from a general structure rather than from what your organization happened to think of. Ask it explicitly what is missing from the model you have built. The answer is regularly a capital reserve line or a restoration obligation.

    Generating scenarios and sensitivity ranges. Once the model exists, producing four coherent scenarios and running the comparison under each is fast, and so is identifying which assumptions the answer actually depends on. Asking for a tornado analysis, meaning a ranking of inputs by how much they move the outcome, turns a spreadsheet into a decision tool and tells leadership where to concentrate remaining diligence. This connects directly to the broader planning work described in our guide to building a strategic plan with AI, since a real estate decision that contradicts the strategic plan is usually the real estate decision that is wrong.

    Reading lease documents and flagging clauses. A commercial lease runs sixty pages of dense language and most nonprofit executives read it once, quickly. Having a model summarize the document, extract every economic term into a table, and flag the clauses that most commonly disadvantage tenants is a substantial improvement over reading it alone. Ask specifically about escalation mechanics, operating expense definitions and caps, audit rights, assignment and subletting, restoration obligations, renewal options and how renewal rent is set, early termination rights, casualty and condemnation, and any personal guarantee. Then take the flagged list to an attorney. The tool is doing triage so that expensive professional time goes to the clauses that matter, not replacing the review.

    Drafting the board memo and the hard questions. Synthesizing the analysis into a decision memo that a board can actually read is work that otherwise delays the decision by weeks. More valuable still is the adversarial use: ask the model to write the five questions a skeptical, financially literate board member would ask about this recommendation, then answer them before the meeting. This is a cheap and genuinely effective way to find the weak point in your own case. It pairs well with the preparation practices in our guide to building board meeting packets with AI.

    Researching comparable market rates. AI can help you assemble and interpret published asking rents, understand how rates are quoted in your market, and convert between full service and triple net quotes so you are comparing like with like. Treat this as orientation rather than valuation. Asking rents are not transacted rents, published data lags, and concessions such as free rent periods and improvement allowances do not appear in headline numbers.

    What it cannot do. AI does not read a market. It does not know that the block is about to change, that the landlord is motivated, that the seller has a deadline, or what comparable space actually traded for last quarter, and it will nonetheless produce a plausible-sounding paragraph if asked. It is not a substitute for a tenant representation broker who has no interest in the landlord, a real estate attorney licensed in your state, an appraiser, a building inspector, or an environmental consultant. It cannot tell you your organization's risk tolerance. And most fundamentally, the model is only as good as the assumptions the leadership team supplies, which means a well-built model resting on an optimistic headcount projection is a well-built wrong answer delivered with more confidence than a rough estimate would have carried.

    Genuinely useful AI tasks

    Structured work you can verify against a source

    • Full cash flow model across lease and purchase options
    • Identifying cost lines missing from your existing model
    • Scenario generation and sensitivity ranking of assumptions
    • Lease abstraction into a table of economic terms
    • Flagging tenant-unfavorable clauses for attorney review
    • Drafting the board memo and the skeptic's questions

    Keep with professionals

    Local knowledge, licensure, and liability

    • Valuation and what comparable space actually transacted for
    • Negotiating the deal and reading the other side's motivation
    • Legal review of lease, purchase, and financing documents
    • State and local property tax exemption analysis
    • Building condition and environmental assessment
    • The organization's tolerance for fixed obligation and risk

    The Board Packet, and the Exit You Plan Before You Enter

    A board asked to approve a multi-million dollar, multi-decade commitment deserves a decision packet that supports a real decision rather than ratifying one already made. The packet should present at least three options, and one of them should be staying put or doing nothing, because a comparison against a single alternative is not a comparison. It should show the full occupancy cost of each option over a stated horizon in present value terms, with the assumptions listed explicitly and separately so that a director can disagree with an assumption without having to reverse-engineer the spreadsheet.

    It should include the affordability tests run under a downside scenario, not only at budget. It should state plainly what the organization gives up in each option, including liquidity, flexibility, and the programs that will not be funded because the capital went into a building. It should name the risks, including the ones that argue against the recommendation, because a packet that contains no counterargument invites a director to construct one from scratch in the meeting. And it should describe the governance implications: new covenants, new reporting obligations, new committee work, and whether the board is prepared to oversee a property.

    Directors also have a fiduciary duty of care that attaches to this decision specifically. Approving a major real estate commitment without documented analysis, without independent professional advice, and without recorded deliberation is a governance exposure independent of whether the decision turns out well. The minutes should show the options considered and the basis for the choice. This is not defensive paperwork, it is the record that demonstrates the board did its job.

    Finally, plan the exit before you commit to the entry. For a lease, that means negotiating for assignment and subletting rights with reasonable landlord consent standards, an early termination option with a defined fee, and clarity on the restoration obligation, all of which are far cheaper to obtain during negotiation than during a crisis. For a purchase, it means asking who the next buyer is. A purpose-built facility with limited alternative use in a weak submarket is difficult to sell, and an organization that may need to unwind the position in a downturn should understand that before it buys. Ask how long comparable properties take to sell, what a distressed sale would realize, and whether the financing permits a sale without penalty.

    Disposal planning also means knowing what happens to a property the organization no longer needs. Real property acquired with federal funds carries obligations under the applicable award terms and property standards, which can include continued use requirements or a federal interest in proceeds on disposition, and that interest can outlive the grant by many years. Organizations that used restricted gifts or grant funds toward acquisition should confirm what strings remain attached before assuming the sale proceeds are unrestricted. Our guide to audit preparation is a useful reminder that these obligations surface eventually whether or not anyone remembered them.

    What the board decision packet must contain

    Enough for a director to disagree intelligently

    • At least three options, including staying put
    • Total cost of occupancy over a stated horizon, in present value
    • Every assumption listed separately and attributed to a source
    • Affordability tests under downside as well as budget conditions
    • Risks stated honestly, including those against the recommendation
    • The professional advice obtained and from whom
    • The exit path and what it would cost to take it

    Conclusion

    There is no general answer to whether nonprofits should lease or buy. There is a good process and a bad process, and the difference between them accounts for most of the difference in outcomes. The bad process starts with a property, works backward to a justification, and presents the board with a recommendation dressed as an analysis. The good process starts with the organization's strategy and time horizon, builds an honest total cost of occupancy model for several options, tests each against futures that include unpleasant ones, and only then goes looking at buildings.

    The specific habits that separate the two are not complicated. Compare total occupancy cost rather than rent against mortgage. Fund a capital reserve in the ownership model or admit the model is incomplete. Run the affordability tests in a bad year rather than a normal one. Treat a donated building as a proposal requiring diligence rather than a gift requiring gratitude. Understand what renting out space does to your property tax exemption and to your unrelated business income position before you count the rent. And ask who is going to do the work, because a building is a second business whether or not the organization has staffed one.

    AI changes what is feasible for a small organization here, and the change is real. A nonprofit that could never have afforded a consultant to build a fifteen-year, four-scenario occupancy model can now build one in an afternoon, abstract a sixty-page lease into a table of terms, identify which assumptions the answer actually depends on, and draft a board memo that anticipates the hardest questions. That is a meaningful shift in analytical capacity for organizations that have historically made these decisions on instinct.

    It is also not the decision. AI does not know your market, cannot negotiate on your behalf, will not catch the environmental issue, and has no view on how much fixed obligation your organization can live with. A model is a structured argument about assumptions somebody chose, and the assumptions are where the answer lives. Use these tools to make the analysis complete and the trade-offs visible, then put the judgment where it belongs, with the people who will still be there in year fifteen explaining the decision.

    Model the Decision Before You Sign

    We help nonprofits build the occupancy models, scenario tests, and board materials that turn a real estate instinct into a decision the board can actually examine.