Every society that allows credit or land tenure to exist runs into the same structural fact eventually: claims compound faster than the real activity underneath them can sustain, and something eventually has to give. What differs, case to case, era to era, is whether “something has to give” arrives as a designed, scheduled event that everyone can see coming, or as a crisis that arrives on nobody’s calendar and gets resolved under duress. This is a survey of the different attempts, across roughly four thousand years and a half-dozen unrelated legal traditions, to answer that one design question: how do you let credit and land tenure function without letting either compound into a permanent, unaccountable claim?
The modern diagnosis
The starting point is contemporary rather than ancient. Ray Dalio’s How Countries Go Broke describes what he calls the Big Debt Cycle: a long-run pattern, playing out over fifty to a hundred years, in which borrowing outpaces income growth until debt service becomes unsustainable, forcing some combination of default, restructuring, or currency debasement. Dalio has studied thirty-five such cases across the last century and argues, with some justification, that nobody has adequately mapped the underlying mechanics — despite the fact that, in his own words, virtually every reserve currency and monetary order in history has eventually broken down this way.
What Dalio’s own proposed fix — a cap on deficit growth relative to GDP — actually is, once you look closely, is a flow-rate constraint. It slows the rate at which debt accumulates. It contains no release valve, no mechanism for what happens once the debt has already accumulated past serviceable levels. It assumes the reset, when it comes, will keep coming the way it always has: through crisis, negotiated under duress, rather than through anything scheduled or rule-based.
That gap — a diagnosis of compounding claims with no designed exit — turns out to be a very old problem, and it has an equally old set of attempted answers.
The jubilee, properly understood
Leviticus 25 is usually summarized as “every fifty years, debts are cancelled and land goes back to its original owners.” That summary hides the more interesting mechanism underneath it. Land sales in the text are priced not as sales at all, but as a number of remaining harvest-years until the jubilee — which is to say, as a lease, priced by term, from the outset. There was never a freehold for the reset to act on. The formula is baked into the transaction from day one: everyone party to the sale, buyer and seller both, knows in advance exactly how many years of value they’re transacting, because the clock is public and fixed.
The text is also more pragmatic than the popular version suggests. A parallel provision allows a kinsman, or a seller who has since prospered, to redeem land before the jubilee arrives, at the same term-based formula. And crucially, houses in walled cities are explicitly carved out — permanently exempt from the reset if not redeemed within a year of sale. That’s not an oversight. It reads like the drafters already understood that an unbounded jubilee was unworkable against concentrated urban and trade wealth, and built the exemption into the law rather than pretending the law was airtight.
Whether ancient Israel ever actually ran a fifty-year jubilee cycle in practice is genuinely unclear — there’s little archaeological or extra-biblical evidence for it, and most historians treat it as more aspirational than executed. The seven-year sabbatical release (shmita) has better evidence of being taken seriously as an ongoing practice, and it has a documented history of being deliberately circumvented: Hillel the Elder’s prosbul, a first-century-BCE legal instrument, let a creditor pre-declare a debt to a court before the sabbatical year, reclassifying it as a court debt exempt from cancellation. Interestingly, the workaround is usually defended as pro-poor rather than anti-poor — without it, lenders simply stopped extending credit as the reset date approached, since any outstanding loan would be wiped, which hurt exactly the population the law was meant to protect. It’s an early, clean illustration of a pattern that recurs throughout this whole survey: any reset defined by an enumerated scope creates a boundary, and sophisticated actors will find the boundary and work it, sometimes with genuinely mixed effects on the people the reset was meant to help.
Usury, and why it was abandoned rather than refuted
Leviticus’s own usury law is narrower than commonly assumed — it bans interest between Israelites specifically, while explicitly permitting it on loans to foreigners. It’s an in-group rule, not a general claim that interest itself is illegitimate. The more absolute prohibition came later, through Aristotle’s argument that money is sterile and interest an unnatural breeding of money from money, formalized by Aquinas into scholastic doctrine and then into outright medieval Christian canon law.
What’s notable is how that prohibition was eventually dismantled: not by refutation, but by redefinition. Calvin’s reworking of the doctrine — interest legitimate when it functions as return on productive capital rather than exploitation of the desperate — is generally credited as the theological hinge that let interest-bearing credit become respectable in Protestant commercial culture, feeding directly into the joint-stock, credit-based capital formation that underwrote the rise of capitalism. The thing modern capitalism is built on is, historically, the exact thing every prior tradition treated as restricted or morally suspect. It wasn’t disproven. It was reclassified.
The lesson this offers for jubilee-style design is a genuinely useful one: banning interest outright kills capital formation, because you cannot price risk or the time-value of money without some return mechanism — every tradition that tried it eventually drove lending underground into other forms rather than eliminating the underlying dynamic. Capping the compounding horizon, by contrast — allowing interest, but guaranteeing it cannot accumulate past a known, fixed point — is what jubilee actually does. It’s a governor on the mechanism, not a ban on the mechanism itself. Islamic finance built an entire modern commercial system on precisely this distinction: ijara (lease) and diminishing musharaka (co-ownership transferring gradually) price financing through rental value and declining ownership share rather than interest on a principal balance, solving under real commercial pressure the same constraint jubilee was solving under religious law.
Working models: land held by the many, occupied by the one
Several real, currently operating systems have converged on some version of jubilee’s underlying structure without necessarily citing it.
Switzerland’s Baurecht separates land ownership from building ownership entirely. The land is typically held by a municipality, church, or foundation — a civic body, not a profit-maximizing landlord — and leased on terms of thirty to a hundred years. The resident owns the building outright and pays ground rent for the land beneath it. The single most important design feature, and arguably the cleanest solution found anywhere in this survey to the renewal-capture problem discussed below, is that the compensation formula for the end of the term is agreed at the moment the lease is signed, not negotiated under pressure decades later when one party may have far more leverage than the other.
Singapore’s HDB system proves the same underlying idea can function as the default tenure arrangement for an entire national population rather than a niche product: roughly eighty percent of Singaporeans hold 99-year state leaseholds on their flats rather than freehold. Its weak point is instructive. The state’s periodic early-buyback mechanism (SERS) is discretionary rather than formula-triggered — some estates get selected for redevelopment, others don’t, on criteria that have drawn criticism as opaque — which reopens exactly the “whoever controls the trigger controls the outcome” problem a fixed formula is supposed to close. There’s also an unresolved lease-decay problem: value runs toward zero as a flat’s 99-year term approaches its end, with no automatic right of renewal, and Singapore has had a live domestic policy debate for the better part of a decade over what this means for retirees whose main asset is a decaying lease.
Israel’s Land Authority, holding roughly ninety-three percent of the country’s land and leasing it on 49- or 98-year terms, is the one modern system to adopt Leviticus 25’s language directly rather than converging on the same structure independently. In practice it has drifted toward near-automatic renewal over the decades — the same erosion of a reset mechanism’s teeth seen in Canberra’s own long-running Crown leasehold system, where routine renewal has, in practice if not in law, become close to indistinguishable from freehold.
Hong Kong’s New Territories leases (99 years, issued from 1898) and China’s 70-year urban land-use rights (issued from 1990, with the first tranche now approaching expiry) show the two different failure modes that arise when a large number of leases are issued in the same era and approach expiry roughly together. Hong Kong’s expiry became a single, enormous geopolitical negotiation folded into the Sino-British Joint Declaration rather than a distributed, rule-based process. China’s is a currently live and unresolved policy question, watched closely because nobody yet knows which way it will break.
The negative control: what happens without any reset mechanism at all
Two very different systems, usually placed at opposite poles of political ideology, turn out on closer inspection to share the exact same structural defect: no reset mechanism whatsoever, once a claim is granted.
Soviet housing is popularly remembered as a system with no private property. That’s true as a legal matter and misleading as a practical one. Urban land and nearly all housing stock were nationalized from 1917 onward; allocation ran through municipal and workplace waiting lists against a hard threshold of living space per person, with waits commonly running fifteen to twenty years and a sixth of the Soviet population still living in communal apartments or with no fixed residence as late as 1989. But once a family finally received an apartment, they held it for life, passed it on by inheritance, and could not practically be evicted — under an instrument called a propiska, permission to occupy rather than a lease or a title, with no market and no price attached, only a direct barter-style swap available if a family wanted to relocate. Strip away the legal language and what remains is something families held in every functional sense a family holds a home — permanent, exclusive, inheritable — without ever being granted a name, a market value, or an exit ramp in law. A system with unbounded individual holding and no price mechanism to force turnover produces exactly what it produced: chronic, worsening shortage, because existing stock essentially never returned to circulation regardless of how much new stock got built.
Freehold ownership, sitting at what’s normally understood as the opposite end of the political spectrum from Soviet housing, shares the identical defect from the opposite direction. It is fully tradeable — which distinguishes it from the Soviet case, since a functioning market keeps some stock recirculating — but tradeable is not the same property as bounded. A freehold sale transfers the same unbounded claim to a new permanent holder; nothing about the transaction returns it to any shared pool or subjects it to periodic redistribution. Given enough time and enough generations, nothing in the instrument itself prevents freehold claims from concentrating, for the same structural reason a Georgist economist identifies in land rent generally: there is no governor on it.
The two systems fail in mirror-image ways — total stasis from having no turnover at all, versus slow, price-driven concentration from having turnover with no ceiling — but they land in the same category on the axis that actually matters here, which is not who holds the claim (state or individual) but whether the claim is bounded. That reduction is worth carrying forward past this survey: the state-versus-individual axis that usually organizes housing-policy argument turns out to be close to irrelevant to whether a system stays healthy over time. What matters is whether a term, a formula, and an enforced trigger exist to return the claim to circulation.
Eastern Bloc post-1989 privatization offers a third data point, distinct from both: a one-time, unconditional transfer of state-held claims into full freehold, essentially overnight, with sitting tenants given title at minimal or no cost across most of Central and Eastern Europe. It was driven less by economic design than by political necessity — incumbent governments used the giveaway as a shock absorber to defuse unrest, the mirror image of the elite-resistance dynamic seen elsewhere in this survey, since here relinquishing a claim, rather than protecting it, was what bought political survival. The consequences are instructive: public housing collapsed to a residual share of the market almost immediately, with no periodic mechanism to ever recover any of it, and unclear responsibility for shared building infrastructure left many large housing estates deteriorating for decades. Once a population holds unconditional freehold, reintroducing any form of bounded tenure afterward has proven, across the region, close to politically impossible — direction of a one-time transfer, it turns out, matters as much as the mechanism itself.
Why rule-based resets keep losing, even when they’re formally proposed
It would be easy to conclude from all this that the fix is obvious: legislate a rule, make it automatic, done. Two documented failures suggest it’s harder than that.
In 2001–2003, the IMF’s Anne Krueger formally proposed a Sovereign Debt Restructuring Mechanism — in effect, a standing bankruptcy court for nations, allowing orderly restructuring rather than ad hoc crisis negotiation. It was blocked, primarily by the US Treasury and the private financial industry, on the argument that a standing mechanism would make default more predictable and therefore raise borrowing costs generally. That argument has real economic merit, but it also protects something else: once a reset mechanism is legitimized, it constrains all future claims, not just the current negotiation, and concentrated creditor interests have both the standing and the sophistication to resist that constraint even when the immediate deal on the table wouldn’t cost them anything. Resistance, in other words, isn’t only about the direct financial hit — it’s about the precedent that periodic constraint on accumulation becomes a normal, expected feature of the system rather than an emergency measure.
Even where a rule does get legislated, Goodhart’s Law tends to bite: any target that becomes the basis for behavior stops being a reliable measure of the thing it was meant to track. The EU’s Stability and Growth Pact deficit limits are a clean real-world test case — France and Italy have breached the 3% threshold repeatedly for decades through negotiated waivers and creative accounting, because a voluntary cap enforced by the actors it constrains isn’t really a constraint, it’s a norm, and norms get renegotiated whenever the short-term cost of holding to them exceeds the cost of breaking them. Modern capital markets add a second, sharper layer on top of this: financial engineering’s entire comparative advantage is constructing instruments that sit just outside whatever boundary a rule defines, at institutional speed, as a competitive necessity rather than an occasional lapse. Every major postwar financial regulation has generated a documented arbitrage response of exactly this kind.
What survives that pressure better, on the limited evidence available, is a rule that acts automatically rather than being invoked at anyone’s discretion — closer to how unemployment insurance spending rises and falls without a vote each time, than to a target politicians renegotiate annually — combined with a rule that measures something that has already happened (realized debt-service payments as a share of actual collected revenue, say) rather than a forward-looking, more easily manipulated proxy.
Extending the frame: superannuation, resource rents, and infrastructure monopolies
The same design vocabulary turns out to transfer cleanly into contemporary questions well outside land tenure in the narrow sense.
Australian superannuation funds are, right now, becoming landlords at scale through build-to-rent partnerships, and some are experimenting with letting members effectively lease housing their own fund owns. This creates a real structural tension: a fund’s fiduciary duty is to maximize member returns, which pulls toward rent-maximizing behavior, while its role as housing provider to those same members pulls toward tenant-protective behavior. The Swiss Baurecht split — one body holding land and bearing a public-interest mandate, a separate body owning and operating the building under a fixed-cost ground lease — offers a structural way through that tension: it isolates the fiduciarily clean commercial layer (efficient building operation) from the public-interest layer (land cost and access), rather than asking one entity to serve both masters at once.
A Georgist distinction between earned wealth (value created by labor and investment) and unearned wealth (value accruing purely from possession of something scarce and non-produced) turns out to organize a surprising range of modern infrastructure questions once you look for it — orbital slots and lunar polar sunlight in a space-settlement context; leading-edge chip fabrication capacity and grid power in an AI-infrastructure context; the collectively-produced corpus of human writing an AI model trains on, versus the genuinely earned engineering that builds the model around it. In each case, the design question is the same one this whole survey keeps returning to: is the claim on something scarce and non-produced, which argues for treating it as a rent to be captured for the common benefit, or is it a claim on something genuinely built, which argues for leaving it alone and regulating only the monopoly position it may acquire once built, the way utility regulation lets a company earn a return on the grid it built while still capping what it can charge to access that grid.
Where the framework runs out
It’s worth being explicit about where this whole design vocabulary stops applying, because it’s tempting to reach for it everywhere once it starts explaining a lot.
Every functioning case in this survey — Leviticus, Baurecht, Singapore, even Israel’s own domestic system — governs claims within an already-agreed political community, under a shared court and a shared authority both parties recognize. None of it has anything to say about where the boundary of that community itself is drawn, or about resolving a contested-sovereignty dispute between two peoples who share no such common authority. Applying a domestic-tenure-reset framework to that kind of question is a category error; the genuinely applicable literatures there are restitution law and post-conflict property settlement, which are real, serious, and — on the available evidence from Central Europe’s own post-1989 restitution processes — reliably slow and difficult even under far more cooperative conditions than most such disputes enjoy.
The reduction worth keeping
Strip away the historical costume and what’s left is a small number of genuinely distinct design moves, tried independently by people who mostly never read each other: bound the term instead of banning the underlying transaction (jubilee, over an outright interest ban); fix the exit valuation formula at the start of the relationship instead of negotiating it under pressure at the end (Baurecht, over Hong Kong- or Canberra-style renewal); trigger the reset automatically rather than leaving it to anyone’s discretion (automatic stabilizers, over Singapore’s SERS or a voluntary deficit target); and separate whoever controls essential, non-substitutable infrastructure from whoever captures the rent on genuinely scarce resources, so the same actor never holds both levers at once.
None of them is sufficient on its own, and none of the working examples surveyed here gets all four right simultaneously. But every failure documented in this survey — Soviet ossification, freehold’s slow concentration, the SDRM’s defeat, Singapore’s lease-decay problem, Eastern Europe’s irreversible giveaway — traces back to one of these four elements being missing, not to the underlying idea being unworkable. That’s a more useful way to hold the whole subject than sorting it by which political tradition produced which system, because the political tradition, on the evidence gathered here, turns out to predict almost nothing about whether the resulting system actually works.