III. The Drift: A Typology of How Forever Promises Come Apart
Opening
Figure 2 (Drift), at the head of this section, names the forces that compound to push every aspirational trust off course. The categories that follow are what those forces look like in operational practice — and where the simple word “drift” stops covering what is happening.
The framers of the school compact wrote a stronger institution than the country was prepared to operate. They reached, with the legal tools available to them, for the most rigid form they could imagine — the perpetual charitable trust — and they put that form into the geometry of national settlement. They did the work that a generation of constitution-makers can do.
What they could not do was install the ordinary, tedious, generation-by-generation enforcement that any institution requires to actually function. That work fell to whoever happened to be in office on a given afternoon in 1872, in 1887, in 1953, in 2014, in 2024. And the through-line of the next two hundred years is that the men and women who happened to be in those offices were, on average, too distracted by the present to defend the future from themselves.
The fifty-state record is in. The companion volume — LOOKING BACK · Schools of the Republic — works through the architecture’s evolution era by era and the cases as they apply, and what it shows, when the comparative view is laid out in one place, is that the pathology comes in distinct categories. Calling all of it “drift” hides the differences. It is worth taking the differences apart.
Two of the categories below are forms of drift in the strict sense — slow, structural, and unintentional movement of an aspirational asset away from its purpose. They are the gradual failure mode. Three of the categories are forms of directed seizure — affirmative acts by particular people in particular rooms to convert the trust into something else. They are the sudden failure mode. Both modes show up in the same fifty-state record. They run on perpendicular axes. The two phenomena are not the same, and the next section will show why they require different remedies.
Drift, in the strict sense, is corrected by attention. Directed seizure is corrected by litigation, by constitutional amendment, or by mobilization sufficient to reverse the act of conversion. A constituency built only to remedy drift will not stop a directed seizure when one comes. A constituency built only to fight named seizures will not maintain the slow attention that drift requires.
Margaret Bird’s working reframe captures the distinction in the language the rest of this volume uses: drift and directed seizure. The five categories below are how the reframe organizes the evidence. This section is the empirical floor for the gradual mode in particular; the sudden mode shows up here too, because the historical record does not always sort itself by axis, and where the evidence cuts both ways the text will say so.
III.A — Drift by omission: the architecture that was never built
The first category is the simplest. Some states never received the fiduciary architecture in the first place, and the asset moved away from its purpose because nothing structural was in place to hold it.
The clearest case is California. The 1850 Admission Act and the 1853 Survey Act granted school sections without the “in trust” compact language the Northwest Ordinance template had carried since 1803, and the California Supreme Court read the omission as decisive in Sherman v. Buick (1876). The companion volume’s Chapter 4 carries the historical record; the diagnostic point belongs here. By 2026, California retains approximately 8–10 percent of the original 5.5 million-acre school grant. Comparable retention rates in Arizona, Montana, and Wyoming run between 75 and 85 percent. The proceeds, where they remain, flow not to common schools in the founders’ sense but to the California State Teachers’ Retirement Fund — a beneficiary substitution that is not a betrayal (CalSTRS retirement obligations are a real public claim) but that is also not the original dedication.
California is the lower bound of the structural diagnostic. When the architecture is largely absent at the founding, the asset shrinks early; the revenue stream is redirected at the deposit rule; there is no constitutionally protected beneficiary class with standing to challenge any of it. The schools, in California, have been moved from beneficiary status to something closer to payee — the fund’s residual claimants where there is anything left to claim, rather than the protected class the architecture was supposed to dedicate the corpus to.
The other case at this end of the spectrum is the Original 13 — the states that predated the federal template. Volume I’s Chapter 1 carries the structural fact: the Thirteen built state-derived school architectures of varying strength; the dominant accountability story in those states is not trust-breach litigation but adequacy-and-equity litigation. Drift by omission does not require malfeasance. It requires only that the architecture not be there to begin with.
III.B — Drift by attrition: the slow sell-off
The second category is the textbook drift the field’s vocabulary was originally built for. The architecture is in place. The fiduciary language sits in the constitution. And the asset goes anyway, sale by sale, decade by decade, through the ordinary inattention of generations of officeholders. This is the gradual failure mode in its purest form — what the framers could not engineer against because there was no single moment to point to and say, that is the breach.
Oregon’s first half-century is the most documented case in the country, and Volume I treats it at length. The architectural starting point was strong: the 1859 Admission Act granted sections 16 and 36 in every township; Article VIII, Section 2 of Oregon’s 1857 constitution declared the Common School Fund “separate and irreducible.” On paper, the architecture was as strong as anything else in the public-land system.
The operating disposition was not. F. G. Young’s 1910 Oregon Historical Quarterly paper described the period without softening it: “it has been Oregon’s unwavering aim to get its state lands as rapidly as possible into private hands and under cultivation … every acre as soon as the state had title, and quite too frequently even when it had only slight basis for expecting to secure it, was on the market.” The 1887 amendment Young called The Infamous Act removed the auction mechanism and required the State Land Board to sell at the uniform price of $1.25 per acre.
Governor George Chamberlain, in his 1907 message to the legislature, supplied the bottom-line measurement of the drift: “But for the policies which have been adopted by the Legislature from time to time with respect to its school lands, the irreducible school fund might have been five or six times as large as it is at present.” Five or six times as large. Read as an arithmetic statement, Chamberlain’s sentence is the cleanest possible measurement of drift by attrition: the trust corpus, after a half-century of state administration, was worth between a sixth and a fifth of what the framers’ architecture would have produced if it had been honored. The federal land-fraud trials of 1904–1910 produced twenty-one convictions of state and federal officials.
What is striking, reading the Oregon record, is not that things went wrong. It is that they went wrong in exactly the ways the seven structural defenses Section V will name would have prevented. There was no enforceable fiduciary code; the constitutional duty in Article VIII, Section 2 was a purpose statement, and the future statute that was supposed to install a prudence rule, a loyalty rule, and a beneficiary remedy never installed any of them. There was no asset restoration. There was no individual accountability — the constitutional officers were nominally trustees, the clerks did the work, and the speculators harvested the gap. There was no independent beneficiary advocate. There was no specialized legal defense.
The doctrinal floor was, in fact, articulated cleanly when the courts had occasion to look at it. The Oregon Supreme Court called the Common School Fund a “trust of the highest nature” in Eagle Point Irrigation District v. Cowden (1931). Oregon’s own Attorney General, in 1992 and again in 2003, concluded that “the school lands granted to the State of Oregon are a trust for the benefit of public education” and that the State’s “obligations are binding. They cannot be disregarded.” The architecture, as the State’s own legal officers periodically restated it, was correct. What did not happen was the operational defense the architecture assumed.
Drift by attrition is what happens when the architecture exists on paper and the constituency who would enforce it does not.
III.C — Drift by capture: the institution begins to harvest itself
The third category is the modern form of drift, and the one Section IV will diagnose most carefully. The architecture is in place. The constituency, if there ever was one, has dissipated. The agency charged with administering the trust has, over time, come to view the trust corpus as the asset on which its own continuity depends. Decisions that improve the agency’s administrative position, the dominant industry’s position, or the legislature’s general-fund position get made, one at a time, with defensible explanations in the moment. The aggregate effect, viewed across two and a half decades, is the trust drifting into something else.
Oregon’s post-1990 Elliott State Forest story is the case study.
The Elliott had been, for most of the twentieth century, the workhorse asset of Oregon’s Common School Fund — 93,000 acres of Coast Range Douglas-fir managed for sustained-yield timber production, generating reliable annual revenue. By the mid-2010s, the State Land Board had concluded the Elliott was a net drain on the fund. Litigation over endangered-species protections had reduced harvestable acreage; the political costs of timber harvest had outrun the school revenue. The Board’s judgment, defensible inside the agency’s frame, was that the trust would be better served by selling the asset and reinvesting the proceeds.
In 2017, the State Land Board voted to sell the forest in a single transaction to a private buyer, then reversed under public pressure. In 2019, the state issued $100 million in bonds to compensate the Common School Fund for the Elliott’s reduced timber revenue. By 2022, the legislature had approved a plan to convert the Elliott into a research forest managed by Oregon State University, with bonds and other appropriations standing in for the timber receipts the trust would otherwise have received.
Each step had a defensible rationale. The aggregate is the structural pattern. An asset reserved in 1859 for common schools became, between 1990 and 2022, an asset whose use was decided by a coalition of state agencies, university administrators, environmental advocates, and legislators — none of whom were the trust’s beneficiaries, and most of whom had institutional interests other than maximizing common-school funding. The bond compensation is real money, but the bonds are general-obligation debt of the state, paid back from general revenue. The trust’s claim on a particular working asset has been converted to the trust’s claim on a particular line item in the state budget. That is not the same thing.
The Oregon AG’s office had, as it happened, ruled on exactly this question forty-five years before the Elliott decoupling. In 1977, asked whether income-producing Common School Grant Land could be designated as a Natural Area Preserve under ORS 273.562, the Attorney General answered yes only “when such designation would adversely affect the financial contribution of such land to the Common School Trust Fund” — meaning, in operational terms, no. The 2022 State Land Board did the categorical move anyway. The 1977 doctrine was not amended; it was bypassed.
The 2024 OASTL case — Advocates for School Trust Lands v. State of Oregon, 24CV38372 — and the Oregon Court of Appeals’ January 28, 2026 opinion granting standing to schoolchildren plaintiffs is the moment when, after 167 years, an organized beneficiary constituency finally appeared inside the courthouse and asserted the trust against the state as trustee.
Drift by capture is the hardest form of drift to oppose, because no individual decision in the chain looks like betrayal. Every actor in the room has a defensible reason. The drift happens at the level of the room itself.
III.D — Directed seizure by stealth: the concealed conversion
The fourth category is a different phenomenon. It is not drift in any sense. It is a coordinated political move by identifiable people, framed in language calculated to obtain consent from beneficiaries who would not have consented if the underlying mechanism had been disclosed. This is the sudden-failure-mode axis the framers also could not engineer against — and it runs perpendicular to the slow attrition of III.B.
Utah’s Project Bold is the cleanest available demonstration.
The story belongs to Scott M. Matheson, a Stanford-trained lawyer who became Utah’s first Democratic governor in twenty years when he took office in January 1977. Matheson understood trust law in the technical sense most governors do not. He had practiced it before he entered politics. He inherited a school-trust portfolio of approximately three million acres, scattered across the state in the section-16-and-36 reservation pattern that the 1894 Utah Enabling Act had produced. The scattered geometry was, as a matter of asset management, a problem.
Project Bold, the policy initiative Matheson developed and pushed publicly between roughly 1982 and 1984, proposed a solution that on its face was sound: aggregate the scattered parcels through land exchanges with the federal government, blocking up the trust holdings into manageable contiguous areas where economies of scale could be realized. The State School Board was approached for support on the rationale that consolidation would substantially increase the trust’s revenue per acre.
Margaret Bird’s account adds a detail the academic literature does not address. According to her contemporaneous recollection — she was on staff at the Utah State Office of Education during the relevant period, and later led the school-trust-lands program there — Project Bold contained, on page 124 of the proposal, language quietly requesting Congress to amend the Utah Enabling Act so that the federal grant of school trust lands would flow “to the state” rather than “for public schools.”
If Margaret’s account is correct, the consequence of the page-124 language would have been categorical, not incremental. A change in the Enabling Act’s grantee from “public schools” to “the state” would have converted the school trust into a state asset. The fiduciary architecture — the rule against self-dealing, the duty of undivided loyalty, the prohibition against converting trust assets to non-trust purposes — would have ceased to apply. The lands would have remained in state hands; the constraints that made the lands a trust would have lifted. The State School Board, whose support was being solicited on the consolidation argument, was, on Margaret’s account, not informed that the proposal also contained the request that would extinguish the trust character of the lands the consolidation was being argued to preserve.
The proposal did not pass Congress. Matheson left office in 1985. The blocking-up concept survived without him and was eventually realized in the 1998 Leavitt-Babbitt exchange and subsequent acts; those exchanges preserved the school-trust character of the swapped acres and did not alter the Enabling Act grantee. The trust survived Project Bold. It survived in part because the State School Board, even if not warned in the form Margaret describes, mobilized in time to defeat the more consequential elements of the proposal.
Directed seizure by stealth is what happens when a sophisticated actor inside the system attempts to convert the trust without the beneficiaries noticing. The remedy is not “more attention to drift.” The remedy is a beneficiary advocate who reads page 124. Utah did not have one in 1982. It built one — the Land Trusts Protection and Advocacy Office, codified in Title 53D — partly in response.
III.E — Directed seizure by floor vote: the open conversion
The fifth category is the same phenomenon as III.D, conducted in the open. The legislature votes; the governor signs; the trust loses; everybody who was paying attention can see it happening. The disclosure is not the problem. The political response that should have stopped it is the problem. This is the sudden-mode failure in its plainest form: the lights are on, the cameras are running, and the trust still loses.
Two cases. Twenty-two days before this draft was written, the governor of Nebraska signed a bill that moved $40 million from a school-trust mechanism to a discretionary K-12 spending fund. Forty-three years before this draft was written, the Utah legislature withdrew $37.5 million from the State School Fund permanent corpus, dropping the fund from $53.5 million to $18.6 million in a single year. Both maneuvers were authorized by ordinary statute. Both were public. Both were preceded by warnings from people who understood what was being done. Both happened anyway.
Nebraska first, because it is live. The bill is LB 1072. The governor is Jim Pillen. The bill became law on April 7, 2026. The mechanism was a cash-fund sweep: LB 1072 transferred the cash-fund balances of the Board of Educational Lands and Funds — Nebraska’s school-trust administrative entity, which holds approximately $3.047 billion in trust assets — into the Education Future Fund, a discretionary K-12 spending vehicle. The technical category is “cash fund sweep by ordinary statute,” not “raid on the constitutional corpus.” Nebraska’s Permanent School Fund corpus is “forever inviolate” under Article VII, Section 1 of the state constitution. The sweep targeted earnings and balances rather than the corpus itself.
That technical refinement does not soften the structural diagnosis. The earnings of an intergenerational trust belong to the trust’s beneficiaries unless the trust instrument explicitly directs otherwise. The Education Future Fund is not the school-trust beneficiary. The K-12 children of Nebraska are not better served by a sweep than by the orderly distribution mechanism the trust’s design assumes. And the precedent matters more than the dollar amount. Once a state legislature has demonstrated that it can sweep BELF cash-fund balances by ordinary statute, the constitutional “forever inviolate” protection on the corpus itself is a question of how far the sweep can be extended, not of whether the sweep can be undertaken.
The political response is the part that matters most for the next section’s argument. State Senator Jane Raybould of Lincoln offered a floor amendment to strike the $40 million BELF transfer on March 19, 2026. The amendment failed nine to thirty. No litigation has been filed. No State Treasurer objection has been recorded. The trust beneficiaries — Nebraska’s K-12 children, their parents, their teachers — did not mobilize.
The contrast that completes the picture is the Nebraska Environmental Trust. In a parallel maneuver in the same legislative session, Pillen proposed sweeping $40.7 million from the Environmental Trust — a separate, lottery-funded conservation trust — into other state purposes. That proposal triggered public outcry, op-ed coverage, and a former state official’s threat of a constitutional lawsuit. The transfer was reduced from $40.7 million to $11.1 million. The mobilization that forced the Environmental Trust reduction did not appear in the school-trust case. The Environmental Trust has a constituency that organized; the school trust did not. The architectural defenses on each side are roughly comparable. The constituency defense is not.
Utah’s 1983 case is the historical predicate. With Utah in a budget crisis, the legislature withdrew $37.5 million from the State School Fund to cover the year’s educational appropriations. The fund balance fell from $53,500,836 in 1982 to $18,583,248 in 1983. In a single year, two-thirds of the permanent endowment was converted into operational cash. The Utah constitution, Article X, contained the same protection Nebraska’s does: “All public school funds shall be guaranteed by the State against loss or diversion.” On paper, the 1983 withdrawal could not happen. In practice, it did.
What stopped it from staying that way was the recovery — and the recovery is the subject of Section V. The Utah PTA, the Utah Education Association, the Utah State Board of Education, the Utah School Boards Association, and the two principals’ associations educated themselves on the trust over the next two years. By 1991 they had pushed the legislature to create a task force; by 1992 the task force had produced a slate of bills; in 1994 the legislature enacted the School and Institutional Trust Lands Management Act. The pre-reform $50 million fund became, over the next thirty years, the $3.2 billion permanent fund Utah Trust Lands today reports.
That recovery is the most important data point in the two-volume record. The architecture that allowed the 1983 collapse is the same architecture that supported the recovery. Article X did not change. The federal Enabling Act did not change. What changed was the constituency. This is the central thesis the next section will adopt.
Directed seizure by floor vote is not stopped by drift defenses. It is stopped by the people who would have to be willing to sue, to mobilize, to vote out, to raise the cost. In Nebraska in 2026, those people did not appear in time. In Utah after 1989, they did. The five years between collapse and reform organization are the operative span.
Closing: the typology and what comes next
Five categories. Two of drift in the strict sense, three of directed seizure. The categories are not airtight. Real cases blur. Oregon’s Elliott story has elements of capture and elements of seizure. California’s failure-by-omission shades into directed seizure at the deposit-rule level, where a 1984 statute affirmatively redirected school-land revenue to teachers’ pensions. Mississippi’s 1820s record, treated in Volume I, is the clearest example of a single state running on both axes at once.
The categories are useful because they sharpen the question of what to do.
Drift by omission requires building architecture that was never built. California is a constitutional-amendment problem.
Drift by attrition requires installing operational discipline inside an existing fiduciary frame. Oregon’s first half-century is a statutory-reform problem.
Drift by capture requires a beneficiary advocate who is not embedded in the agencies whose decisions she would have to constrain. Oregon’s Elliott is a structural-redesign problem.
Directed seizure by stealth requires a beneficiary advocate who reads page 124. Utah’s Project Bold is a beneficiary-counsel problem.
Directed seizure by floor vote requires a constituency willing to make the political cost of the vote higher than the political benefit. Nebraska 2026 is a mobilization problem.
The companion volume carries the fifty-state evidentiary record. What this section carries is the typology and the architecture-implications that respond to each category — which is what Section IV takes up next.
The pattern is national. The federal admission-act language hardened across the nineteenth century, from Ohio’s single-section grant in 1802 to the maximalist 1910 New Mexico–Arizona language with explicit “in trust” and “deemed a breach of trust” clauses. But the variation in outcomes does not map cleanly onto the variation in language. Same federal text, different states, different outcomes by a factor of four or more. Section IV takes that finding apart.
