Chapter 6 — The Twentieth-Century High-Water Mark and Two Outliers

Oklahoma to Hawaii, 1907–1959

“Most people don’t realize they are the beneficiaries of this trust. It’s for their children and grandchildren. So, a parent’s role is to ask their school board and their legislators: ‘How are our school trust lands being managed? How much money are we getting from them?’ A regular citizen can be a watchdog.” — Margaret Bird, August 2025 video transcript

What this chapter does. This is the chapter in which the federal architecture project ends. After the two states admitted in 1959, the country stops installing the section-grant template — stops trying to write the school-trust idea into admission acts at all. What I want the reader to leave with is a clear sense of what the federal government built when it tried hardest, where it gave up, and who, going forward, is going to have to hold the architecture from underneath. The chapter that follows this one used to be a Comparative Atlas; that work is now part of the volume’s Conclusion. From this point on, what we have is the operating record.

The architecture project that began in the Confederation Congress in 1785 — section sixteen of every township reserved for schools, a rule written into a survey grid before any state existed to receive it — was still being built a hundred and twenty years later. Builders changed. Materials changed. The arid West got four sections per township instead of one. “In trust” language got bolted on. Permanent funds got walled off behind irreducibility clauses. But the shape of the thing held, more or less, through every state admitted between 1803 and 1907.

Then five states arrived in fifty-two years that broke the pattern apart.

Oklahoma came in first, in 1907, still inside the template but already showing strain — half the new state had no available federal sections to grant, because the lands had already been promised, parcel by parcel, to citizens of the Cherokee, Choctaw, Chickasaw, Creek, and Seminole Nations. Then New Mexico and Arizona, in 1912, received the strongest school-trust language Congress would ever write — the federal high-water mark, in a single paragraph, in a single statute that admitted two states. Then a forty-seven-year silence. And then, in a single year, 1959, Alaska and Hawaii — neither of them fitting the section-grant template, neither of them a school trust in the Cooper v. Roberts sense, each broken away from the template in a different direction.

I want to slow down at the chronology because it is the spine of the chapter. The architecture project does not end because the country stops admitting states. It ends because the country stops building the same structure. Two of the five states in this chapter — New Mexico and Arizona — represent the federal template at its maximum strength. The other three — Oklahoma, Alaska, Hawaii — represent three different ways the template was strained, supplemented, or set aside altogether. After 1959, no future state would receive the sectional pattern that 1785 had set in motion. The country had stopped building the same thing.

That is the architectural fact I want under the reader’s hand for the rest of the volume.

I. Oklahoma: the template under strain

Oklahoma’s 1906 Enabling Act looks, at first glance, like the standard late-century template.¹ Sections sixteen and thirty-six in every township were granted “for the use and benefit of the common schools.” The state ratified its constitution in 1907 with a permanent-fund clause at Article XI § 2, an irreducibility commitment, and an inviolable appropriation of the income to the schools. A board of trustees — the Commissioners of the Land Office — was constitutionally established under Article VI § 32. By the standards of any Atlantic-seaboard state from the original thirteen, Oklahoma’s school-trust architecture in 1907 was a marvel.

But the standard nineteenth-century template did not survive contact with Oklahoma’s geography intact. The state was stitched together out of two territories: former Oklahoma Territory, which had been federal public domain and therefore section-able in the Northwest Ordinance sense; and former Indian Territory, where the lands had already been allotted under the Curtis Act and the Atoka Agreement to members of the Five Civilized Tribes.² There were no available sections sixteen and thirty-six to grant on the Indian Territory side, because the federal government had already granted those lands to specific human beings.

So Congress, in section 7 of the 1906 Act, did something it had never done before and would never do again: it paid cash. Five million dollars, “for the use and benefit of the common schools of said State,” in lieu of the section grants the federal government no longer had to give.³ The cash was to be held and invested by the State expressly “in trust” for the schools — a stronger formulation, interestingly, than the bare “for the use and benefit” language that governed the section grants on the western side. Oklahoma’s school architecture was, from the day of admission, bifurcated along a line invisible on a modern map but present in every page of the Commissioners of the Land Office’s books — section grants on one side, a cash settlement on the other, both feeding into one permanent fund.

The state-side architecture filled in what the 1906 Act left thin. The 1907 constitution accepted the federal grants as a “sacred trust” in Article XI § 1, and Article XI § 2 declared the Permanent Common School Fund “forever inviolate,” with the State expressly obliged to reimburse the fund for any loss.⁴ I want to put a marker on that reimbursement clause, because as someone whose career has been spent inside the technical legal architecture of these funds, I recognize what Oklahoma did there as unusual. Most state irreducibility clauses make a promise — corpus shall not be diminished — and leave the question of what happens after a diminution to future legislative discretion. Oklahoma’s clause makes a guarantee: the corpus shall not be diminished, and if it is, the State itself will make it whole. That is a stronger fiduciary commitment than most public-land states wrote, even when the federal text under them was relatively weak.

The trustee body created by Article VI § 32 — the Commissioners of the Land Office, also called the School Land Trust — has two distinctive features worth naming. The Superintendent of Public Instruction sits ex officio, putting a beneficiary fiduciary inside the trustee circle, which is rare among the public-land states. And the President of the State Board of Agriculture also sits, reflecting the dominance of grazing and agricultural leases in Oklahoma’s revenue mix.⁵ Whether seating beneficiary-side and industry-side fiduciaries together in the same body is an institutional strength or a built-in conflict is a question the doctrinal record has answered, more or less, in Oklahoma’s favor.

Oklahoma’s doctrinal arc is unusually well developed. State ex rel. Williamson v. Commissioners of the Land Office (1956) held that the Enabling Act’s restrictions are valid federal law and may not be modified by Oklahoma statute or constitutional amendment.⁶ Oklahoma Education Association, Inc. v. Nigh (1982) invalidated a series of statutes that had subsidized incumbent lessees and capped lease rentals on trust property, holding that the Enabling Act and Article XI together create “an irrevocable compact” and a “sacred trust,” and that no statute may diminish the State’s duty to administer the trust for the exclusive benefit of the beneficiaries and for maximum return.⁷ School District No. 23 of Okfuskee County v. Commissioners of the Land Office (1933) held that proceeds from oil and gas leases on sections sixteen and thirty-six and on lieu lands must be paid into the permanent school fund as corpus, with only interest expendable — a rule that has done as much to build the modern fund as any other single Oklahoma decision.⁸ Two later constitutional amendments completed the doctrinal arc: State Question 617 (1988) constitutionalized the lease-management discipline the court had imposed in Nigh, and State Question 665 (1994) installed prudent-investor language into Article XI § 6.⁹

The result, over the long term, is a more modest corpus than its Western neighbors built — roughly $2.5 billion, against Utah’s $3.2 billion SITLA fund, Arizona’s $8 billion Permanent Land Endowment Trust Fund, and the much larger figure that anchors the New Mexico section below.¹⁰ Two structural drivers explain Oklahoma’s relatively slim balance: half the territory had no underlying section acreage at all, and what land it did have was disposed of and leased on terms less disciplined than the late-century template envisioned. Oklahoma also predates, by four years, the federal text’s high-water mark. Its 1906 Act does not contain the express “in trust” language for the section grant, the “null and void” clause, or the federal Attorney General enforcement provision. On the four-axis language-strength scale used throughout this book, Oklahoma scores 1 of 4 on the section grant and stronger on the cash substitute — the same weak federal floor most nineteenth-century states lived on, with a sharper text installed only on the cash side.¹¹

What Oklahoma’s 1906 Act tells us is that the template was already running out of road in 1906. Congress could not section-grant Indian Territory because the lands were spoken for. It compensated in cash, and the cash mechanism — though novel — was a workaround, not an upgrade. The architectural improvement was still ahead.

II. New Mexico and Arizona: the federal high-water mark

It came in 1910, in a single statute that admitted two states.

The New Mexico–Arizona Enabling Act of June 20, 1910 — chapter 310, 36 Stat. 557 — is the single most important paragraph of federal text in the entire school-trust corpus. Section 10 governs New Mexico; section 28 governs Arizona; the substantive language is parallel, and the section-asymmetry has caused some confusion in secondary literature.¹² I want to put section 10’s operative language in front of the reader before I do anything else with it, because everything that follows depends on what is in those words. The text reads, in part:

SEC. 10. That it is hereby declared that all lands hereby granted, including those which, having been heretofore granted to the said Territory, are hereby expressly transferred and confirmed to the said State, shall be by the said State held in trust, to be disposed of in whole or in part only in manner as herein provided and for the several objects specified in the respective granting and confirmatory provisions, and that the natural products and money proceeds of any of said lands shall be subject to the same trusts as the lands producing the same.

Disposition of any of said lands, or of any money or thing of value directly or indirectly derived therefrom, for any object other than that for which such particular lands … were granted or confirmed, or in any manner contrary to the provisions of this Act, shall be deemed a breach of trust.

… Every sale, lease, conveyance, or contract of or concerning any of the lands hereby granted or confirmed … not made in substantial conformity with the provisions of this Act shall be null and void, any provision of the constitution or laws of the said State to the contrary notwithstanding.

It shall be the duty of the Attorney-General of the United States to prosecute in the name of the United States and its courts such proceedings at law or in equity as may from time to time be necessary to enforce the provisions of this Act relative to the application and disposition of the said lands and the products thereof and the funds derived therefrom.¹³

Read that paragraph slowly. Then notice what it does that no prior admission act had done.

It uses the words “in trust.” Not “for the use and benefit of,” which is what every prior act used and which is the sort of language a state can interpret loosely without obviously breaking it. In trust is the operative term of art in fiduciary law. It conscripts every doctrine of trust enforcement that English and American chancery courts have built up over four centuries. It is the difference between asking a state to be a good steward and binding it as a trustee. I have spent my career inside that difference, and I can tell you it is not rhetorical. It is the difference between a hortatory clause and an enforceable one.

It runs the trust through to natural products and money proceeds, foreclosing the argument that once timber is severed or oil is produced the proceeds escape the trust. That sentence is doing a quiet but enormous amount of work. Without it, every dollar of mineral royalty and every board-foot of timber leaves the trust at the moment of severance, and the state is free to treat the proceeds as general fund. With it, the trust runs all the way through.

It calls the violation by its true name: breach of trust. It is one thing for a court, decades later, to look at a transaction and conclude that it was a breach. It is another thing entirely for Congress, in 1910, to write the breach-of-trust label forward into the words of the grant itself. That label is the doctrinal hook every subsequent enforcement case has hung on.

It declares non-conforming dispositions void, “any provision of the constitution or laws of the said State to the contrary notwithstanding.” The federal floor here is not a default rule the state can override by amending its own constitution. It is supreme. A state legislature that authorizes a below-market trust-land lease has not merely behaved badly; it has performed a legal nullity. And it provides for federal enforcement — the U.S. Attorney General is expressly authorized, indeed directed, to bring suit. Almost every other school-trust provision in the federal corpus relies on the state-as-trustee to police itself, or on private beneficiaries who must first establish standing in state court. The 1910 Act gives the federal sovereign a direct enforcement hook against the state sovereign.

Five protections stack: express in trust language, products-and-proceeds reach, breach of trust naming, null and void for non-conforming dispositions, and federal Attorney General enforcement. On the four-axis language-strength scale, New Mexico and Arizona both score 4 of 4. They are the only two states to do so. This is what the federal government was capable of doing when it tried hardest. The high-water mark sits where it sits.

III. Same federal text, different state outcomes

But the 1910 Act, however maximal, is only half of the architecture. The other half is the state-side construction each state built on top of it — and that half is where the two states diverge sharply.

New Mexico ratified its constitution in 1911 and made one structural choice that no other public-land state has made. Article XIII § 1 vests trust-management authority in a single named, statewide elected fiduciary:

The commissioner of public lands shall select, locate, classify and have the direction, control, care and disposition of all public lands, under the provisions of the “Enabling Act” and such regulations as may be provided by law.¹⁴

Not a board, not an ex-officio committee of executive officers. One elected person, four-year term, two-term limit, accountable directly to voters. Article XXI § 9 — the compact-acceptance provision — closes the loop by ordaining the Enabling Act provisions “fundamental law” of New Mexico, enforceable in state courts as state law as well as in federal courts as federal law.¹⁵ The architecture is constitutional, not statutory. The trustee is named, visible, and answerable.

I want to slow down at the New Mexico governance choice because, as someone who has spent years thinking about who actually runs school trusts in this country, I think it is the most important state-side decision any 1910-Act state made. When you put one human being’s name on the trustee role, you give the public an address. When you elect that human being statewide, you give the press, the courts, and the voters a single accountability target. When the office holder runs again in four years, the trust lands are a campaign issue, not a buried agency line item. New Mexico’s framers in 1911 understood something about the political economy of trusteeship that most other states did not: a corpus is only as durable as the constituency that watches it, and a constituency cannot watch what it cannot see. The single elected commissioner is what makes the trust visible.

Today that single elected office presides over the Land Grant Permanent Fund — corpus around $30 billion, annual distribution above $1.3 billion to a roster of twenty-one beneficiaries that runs from K-12 common schools through the state’s universities and hospitals to the Miners’ Hospital and the institution the 1910 Act called the State Hospital for the Insane.¹⁶ It is among the largest sovereign-style permanent funds in the United States. A century of comparatively disciplined administration, on top of the federal high-water mark, on top of a constitutional governance structure that put one human being’s name on the trustee role: that is what produces $30 billion.

The federal Attorney General clause has, importantly, not been left dormant. In Ervien v. United States, 251 U.S. 41 (1919), the United States itself sued New Mexico’s Commissioner of Public Lands, Robert P. Ervien, for spending three percent of trust-land income on advertising “the resources and advantages of the state.” The argument that better advertising would attract settlers and indirectly enrich the trust was rejected as a matter of plain reading. Justice McKenna’s opinion — short and unsparing — held that the trust purposes “must be regarded; they must be performed; they admit of no substitution; they admit of no addition.”¹⁷ Ervien is the only documented occasion on which a sitting U.S. Attorney General has activated the section 10 enforcement clause against any state. It happened in 1919. It produced a Supreme Court holding that has anchored every subsequent enforcement case in the 1910-Act states. And then, on the available record, it never happened again.

The dormancy of active federal litigation since 1919 is itself a finding the rest of this book picks up — whether it reflects exemplary state-level stewardship that obviated the need, or federal under-investment in trust enforcement, or both, is a question worth asking. My own working judgment, formed over thirty years inside this architecture, is that some of both is true. New Mexico’s record under the single elected commissioner has been good enough most of the time to keep the federal AG off the docket. But the absence of active enforcement is also a measure of how much the federal government has withdrawn from the role section 10 reserves for it. A clause that goes unused for over a century starts to feel like an instrument that has been set down. Whether it can be picked up again is a separate question; whether it should be picked up — when, and in which jurisdictions — is the question this book ends on.

[BOXED-INSERT placeholder: chapter 6]

Arizona ratified its constitution a year later, accepted the same federal text in Article XX (the compact-acceptance Ordinance), mirrored section 28’s language verbatim in Article X, and made a different state-side choice on the trustee.¹⁸ Arizona’s State Land Commissioner is a gubernatorial appointee under statute (A.R.S. Title 37), heading the Arizona State Land Department, which the legislature created by statute three years after admission.¹⁹ The trust commitment is constitutional; the trustee is not. Accountability runs through the governor rather than directly to voters. The five-member Board of Appeals — itself gubernatorially appointed — must approve sales and commercial leases, but the central reporting line runs from Commissioner to Governor, not from Commissioner to electorate.

The operational record reflects the difference. Arizona’s grazing-lease history is a recurrent fiduciary critique that no court has fully cured. A 1997 Auditor General performance review found that Arizona’s grazing-lease rates were among the lowest in the West; in fiscal year 1996, 8.5 million acres of grazing land generated only $2.6 million — roughly thirty cents per acre.²⁰ A 2024 Auditor General review of agricultural leases, including those held by the Saudi-owned firm Fondomonte, found that the State Land Department had not adjusted rental rates for seventeen years, had not required lessees to pay for or report the volume of groundwater used to irrigate alfalfa for export, and had set agricultural rental rates at fifty percent below market value as a matter of policy.²¹ Section 28 makes the natural products of trust land subject to the same trust as the lands themselves, and the Arizona Supreme Court has held — in Farmers Investment Co. v. Pima Mining Co., 111 Ariz. 56 (1974) — that groundwater beneath trust land is a product of the land.²² The unresolved question of whether the failure to price groundwater consumed under agricultural leases is a continuing breach of section 28 sits squarely on Arizona’s docket as this book goes to press.

I want to make the contrast as plain as I can. Same federal text. Same admission date, within a month. Same quadrupled grant of sections 2, 16, 32, and 36 of every township. Arizona’s Permanent Land Endowment Trust Fund stands today at roughly $8 billion — about one quarter of New Mexico’s, on a comparable acreage base.²³ The Common Sense Institute’s 2025 retrospective estimates that Arizona’s K-12 land trust has produced approximately $5.8 billion in cumulative distributions over a century, and argues that under a more aggressive early-liquidation-and-investment strategy the fund might today be worth on the order of $163 billion.²⁴ Other observers point out that New Mexico’s corpus advantage is driven primarily by oil-and-gas royalty inflows from Permian Basin trust lands rather than by any difference in liquidation timing. Whatever the precise mix of causes, the order of magnitude holds: same federal text, same admission date, same quadrupled grant — different outcomes by a factor of about four.

This is the chapter’s central observation, and I want it to land. A maximal federal trust commitment, fed into a state governance structure that does not name and elect a fiduciary, can still drift. The federal floor is necessary but not sufficient. The architecture project was always two layers, and the upper layer was always the state’s. Congress can write the text in 1910. Whether the state lives inside it in 1925, 1965, or 2025 is a question that depends on who the state has put in the trustee chair and on what the public — the actual beneficiaries — knows about that chair and about the person in it.

Arizona is also the originating jurisdiction of the modern Supreme Court’s most important restatement of trust-fiduciary obligations under the enabling acts. Lassen v. Arizona ex rel. Arizona Highway Department, 385 U.S. 458 (1967), held that Arizona must compensate the trust at full appraised value when trust lands are used for highway rights-of-way, even for an undeniably public purpose; the presumed-enhancement theory the Arizona Highway Department had operated under since 1929 was rejected as too speculative to satisfy the Enabling Act’s intent.²⁵ Lassen construes section 28 directly and binds New Mexico equally as the other party to the 1910 Act. It is the doctrinal anchor for any modern enforcement claim under the federal high-water mark, and it is the case to which every breach-of-fiduciary-duty argument anchors. The doctrine was forged in Arizona. It governs everywhere the 1910-Act template reached.

A note on Utah, since the chapter sits in the cohort of late-template arid-West admissions and Utah is the natural comparator. Utah’s 1894 Enabling Act had given the state four sections per township — sections 2, 16, 32, and 36, the same quadrupled grant New Mexico and Arizona would receive sixteen years later — and Utah had built, between 1894 and the early 1990s, a permanent fund whose corpus and discipline were comparatively respectable but whose annual distribution was structurally too small to register inside the school system. The reform arc of the 1990s and 2000s — which I lived inside, and which Volume II treats at length — turned that around: the operating revenue from the lands rose from roughly $5 million to roughly $150 million annually, the permanent fund was repositioned to grow every year rather than be drawn down, and the distribution of those funds was rerouted through the school community councils rather than through the legislature’s general appropriations process. The point I want to make in the architectural register of this chapter is small but I think important: Utah’s reform was not a fix to the 1894 federal text. The text was always there. The reform was a fix to the state-side architecture sitting on top of the federal text — the trustee, the investment discipline, the distribution mechanism, and the constituency. Same pattern as Arizona vs. New Mexico, just the inverse case: in Utah, a comparatively quiet federal floor got picked up and built upward by state-side institutional choices that brought the fund and its constituency to operational scale. Federal text alone never builds a trust. State-side architecture does. Constituency holds it.

IV. The silence, and then the outliers

After Arizona was admitted in February 1912, the country added no states for forty-seven years. The continental section-grant project was finished. When the next two states arrived, in a single year, neither of them sat on continental public domain, and neither of them inherited the section-grant template. The country tried two different paths in 1959 — and then, having tried them, did not repeat them.

Alaska: the constitutional opt-out

Alaska arrived first, on January 3, 1959.²⁶ The Alaska Statehood Act of July 7, 1958 — Pub. L. 85-508, 72 Stat. 339 — broke from the township-and-range template entirely. Instead of granting numbered sections of every township for school support, section 6(b) authorized Alaska to select, within a thirty-five-year window, up to 102,550,000 acres “from the other public lands of the United States in Alaska which are vacant, unappropriated, and unreserved at the time of their selection,” together with up to 400,000 additional acres from national forest land, “for the purposes of furthering the development of and expansion of communities.”²⁷ Add the small confirmed territorial school-land reservation under section 6(k), and the lump grant runs to 104.45 million acres — the largest single state land grant in U.S. history.²⁸

What is missing from section 6(b) is at least as important as what is there. The clause does not contain the words in trust. It does not name common schools, or any other beneficiary class, as the object of the conveyance. It does not impose a price floor, an inviolable-corpus rule, or a federal Attorney General enforcement mechanism. It does not even confine selection to revenue-producing land. On the four-axis language-strength scale, Alaska scores 1 of 4 — equal-footing compact form only, with every trust axis empty.

This was not because Alaska’s framers and Congress could not draft trust language. The same Statehood Act, in section 202(e), conveyed approximately one million acres for the establishment and support of a Mental Health Trust, with express trust language and beneficiary protection that would later prove enforceable in State v. Weiss and produce a $200 million restoration settlement in the 1990s.²⁹ The grant for the University of Alaska, similarly, was made on terms the Alaska Supreme Court would later treat as binding trust obligations.³⁰ When Congress wanted to install fiduciary architecture in 1958, it knew how. For the 102.55-million-acre block, it chose not to.

The state constitution that took effect at admission deepened the choice rather than softening it. Drafted at the Alaska Constitutional Convention in 1955–56 by fifty-five delegates, ratified by territorial voters in April 1956, and made effective on admission, the Alaska Constitution is widely regarded as one of the best-drafted state charters of the twentieth century — a self-consciously modernist document influenced by progressive municipal-reform models.³¹ Article VII § 1 commits the legislature to maintain a public school system but does not commit any particular revenue stream or land base to that system. Article VIII frames natural resources as a resource for “the people” of Alaska — the population at large, not schoolchildren or any other dedicated class.

The most distinctive provision is Article IX § 7, the Dedicated Funds Clause, which prohibits the dedication of state revenues to any particular fund or purpose, with narrow exceptions enumerated in the constitution itself. This is the structural inverse of the section-sixteen-state irreducibility clause. Where Oregon’s Article VIII § 2 declared the Common School Fund “separate, and irreducible,” Alaska’s Article IX § 7 declared, in effect, that no fund shall be set apart from the general fisc except by constitutional amendment. The framers did not merely decline to install a school trust; they wrote a constitutional rule against the broader category of dedicated funds, of which a school trust would be one species. That rule has been ratified by every subsequent generation that has either declined to amend it or amended it to create only narrowly enumerated exceptions.

I want to slow down here, because Alaska is the case that most challenges the framework the rest of this volume runs on. The pattern of every prior chapter has been: federal compact installs a fiduciary obligation; state constitution layers on irreducibility; doctrine and practice either honor it or drift from it; the remedies are equitable, the beneficiaries are schoolchildren, and the architecture either holds or does not. Alaska steps off that train. There is no compact, no in-trust language, no irreducible corpus, no school-purpose dedication. Article IX § 7 actively forbids the kind of dedicated fund the section-sixteen states are built around. It is not a state that failed to install the architecture. It is a state that, as a deliberate constitutional choice, declined to install it. That distinction matters to how we read the rest of Alaska’s record.

Two such exceptions define modern Alaska. The 1976 Permanent Fund amendment added Article IX § 15, requiring “at least twenty-five percent of all mineral lease rentals, royalties, royalty sale proceeds, federal mineral revenue sharing payments and bonuses received by the State” to be placed in a permanent fund.³² The 1990 Constitutional Budget Reserve Fund amendment added section 17. The Alaska Permanent Fund’s corpus today runs to roughly $80 billion — larger than New Mexico’s Land Grant Permanent Fund — and, since 1982, a portion has been distributed annually as the Permanent Fund Dividend, a per-capita check to every eligible Alaska resident, with aggregate distributions on the order of $1.7 billion a year.³³ Its beneficiary class, by constitutional and statutory design, is “all Alaskans,” not common schools.

The Alaska Supreme Court has confirmed that even the flagship fund does not carry section-sixteen-state-style irreducibility protection. In Wielechowski v. State (2017), the court held that Governor Bill Walker’s 2016 partial veto of the PFD appropriation was constitutional: the dividend is paid from a legislative appropriation subject to the normal appropriation-and-veto process, not from a constitutionally protected entitlement.³⁴ In Dunleavy v. Alaska Legislative Council (2022), the court invalidated a legislative attempt to forward-fund school districts, holding that appropriating future revenues for future public-education budget cycles violated Alaska’s implied annual-appropriation model.³⁵ Where the section-sixteen states sometimes find their school trusts under-enforced, Alaska finds its constitution actively enforcing the rule against fund dedication — even against the legislature’s own attempts to dedicate.

Alaska’s small statutory Public School Trust Fund — the surviving cash-fund descendant of the 1915 territorial school grant — distributes roughly $15 to $30 million a year to Alaska school districts; the Department of Revenue reported its market value at $945.6 million on November 30, 2025.³⁶ The dominant funding mechanism for Alaska public schools is the Foundation Formula under AS 14.17, an annual general-fund appropriation operating squarely within the Article IX § 7 anti-dedication regime.

Treating Alaska as “scoring low” on a section-sixteen-state rubric would mistake the architectural choice for a failure within a paradigm Alaska did not adopt. The honest framing is different: Alaska is the cleanest twentieth-century example of the road not taken. Whether Alaska schoolchildren are better or worse off than they would have been with a section-grant architecture is an empirical question the architectural lens alone cannot answer. The Permanent Fund is enormous; the school system is funded by general appropriation; the architecture is internally consistent. Treating it on its own terms — rather than as a deficient school-trust state — is what honesty requires.

Hawaii: the multi-purpose trust

Hawaii arrived seven months later, on August 21, 1959.³⁷ It broke the template differently. Hawaii had never been federal public domain in the Northwest-Ordinance sense; there were no sections to reserve because there had been no township-and-range survey of a sovereign Hawaiian Kingdom.

The pre-statehood history bears stating cleanly because the legal characterization of it remains contested, and the chapter’s job is to describe rather than adjudicate. Until January 17, 1893, the Kingdom of Hawaii was an independent constitutional monarchy whose lands, after the Great Mahele of 1848, were divided into Crown Lands, Government Lands, and lands held by individual chiefs and commoners. On January 17, 1893, the monarchy was overthrown by a small armed group of American and European businessmen with the participation of U.S. Minister John L. Stevens and U.S. Marines from the USS Boston. In 1898, the Newlands Resolution annexed the Republic of Hawaii to the United States and accepted cession of approximately 1.8 million acres of former Kingdom lands. The 1900 Organic Act established the Territory of Hawaii.³⁸ In 1993 Congress formally apologized to Native Hawaiians, in Public Law 103-150, for the United States’ role in the 1893 overthrow.³⁹ Federal and state legal sources describe the 1893–1898 transfer in terms of cession and annexation; Native Hawaiian sovereignty advocates describe it as an uncompensated seizure. The events themselves — overthrow, annexation, territorial period — are not in dispute. The legal character of the transfer remains contested, and this chapter names the contestation rather than picking sides.

The Hawaii Admission Act of March 18, 1959 (Public Law 86-3, 73 Stat. 4) governed statehood. Section 5(b) transferred to the new state the federal title to the former Kingdom lands. Section 5(f) imposed the trust:

The lands granted to the State of Hawaii by subsection (b) of this section … together with the proceeds from the sale or other disposition of any such lands and the income therefrom, shall be held by said State as a public trust for the support of the public schools and other public educational institutions, for the betterment of the conditions of native Hawaiians, as defined in the Hawaiian Homes Commission Act, 1920, as amended, for the development of farm and home ownership on as widespread a basis as possible[,] for the making of public improvements, and for the provision of lands for public use. Such lands, proceeds, and income shall be managed and disposed of for one or more of the foregoing purposes in such manner as the constitution and laws of said State may provide, and their use for any other object shall constitute a breach of trust for which suit may be brought by the United States.⁴⁰

The federal trust language is, on three of the book’s four axes, unusually express: public trust, breach of trust, and federal-government enforcement standing are all installed in operative text. Hawaii scores 3 of 4 on the language-strength axis — strong, not maximal, and the fourth axis (a textual restoration mechanism) could plausibly score higher on a fuller reading of the breach-of-trust pathway. By any reading, Hawaii’s federal trust text sits in the country’s top tier.⁴¹

But the structural complication is unmissable: the trust has five purposes, and schools are one of five. The school share of a 1.4-million-acre five-purpose trust is structurally diluted by the four other beneficiary classes, and Hawaii does not separately allocate ceded-lands revenue across the purposes by statute or by constitutional formula. Management is dispersed across multiple agencies — the Department of Land and Natural Resources for most ceded lands, the Department of Hawaiian Home Lands for the 1920 Hawaiian Homes Commission Act carve-out, the Office of Hawaiian Affairs as the constitutional trustee for the Native Hawaiian pro rata share under Article XII §§ 5–6, and the Hawaii State Department of Education as the operational K-12 entity.⁴² No single body sits as fiduciary for the entire trust. There is no constitutional trustee for the school share.

The case-law history of section 5(f) is dominated by the Native-Hawaiian-betterment beneficiary class, not by the public-schools class. Pele Defense Fund v. Paty (1992) recognized standing for Native Hawaiian trust beneficiaries to seek prospective enforcement under section 5(f).⁴³ Trustees of OHA v. Yamasaki (1987), Office of Hawaiian Affairs v. State (the OHA I and OHA II rulings of 2001 and 2006), and the 2008 OHA v. HCDCH moratorium decision all turn on the Native Hawaiian share.⁴⁴ Rice v. Cayetano, 528 U.S. 495 (2000), invalidated on Fifteenth Amendment grounds Hawaii’s restriction of OHA trustee voting to Native Hawaiian voters.⁴⁵ Ching v. Case (2019) is the clearest modern statement that section 5(f) imposes operational fiduciary duties — the Hawaii Supreme Court held that the State had breached its trust by failing to monitor a sixty-five-year, one-dollar lease of trust land to the U.S. Army at the Pohakuloa Training Area.⁴⁶ In none of these cases is the public-schools beneficiary class the plaintiff or the central beneficiary. School-trust enforcement under section 5(f) is comparatively under-litigated, perhaps because no one has ever been able to point to the school slice of the revenue stream and call it a slice.

Hawaii is also the only state with a single statewide unified school district. There are no county districts. The Hawaii State Department of Education is itself the sole school district, governed by a State Board of Education that a 2010 constitutional amendment changed from elected to gubernatorially appointed.⁴⁷ Whatever the section 5(f) trust does or does not do for schools, it does or does not do at the level of the entire state at once.

Hawaii does not have a permanent school fund of federal-grant origin. There is no Hawaiian analog to Oregon’s Common School Fund, Texas’s Permanent School Fund, or New Mexico’s Land Grant Permanent Fund. The closest measurable Hawaiian fiduciary corpus is the OHA Native Hawaiian Trust Fund, on the order of $600 million to $800 million — but that fund serves the Native-Hawaiian-betterment beneficiary class, not the school class. The OHA pro rata share itself was statutorily fixed at $15.1 million per year by Act 178 of 2006.⁴⁸ No comparable dedicated school distribution exists. The Hawaii State Department of Education’s annual budget — roughly two to three billion dollars in recent years — is dominated by general-fund appropriations and federal aid, not by ceded-lands trust distributions. Whether and how section 5(f) ceded-lands revenue actually reaches the schools is a question the public record does not cleanly answer.

Hawaii is the country’s clearest example of the proposition that the strength of fiduciary text and the practical claim of schoolchildren on that text are not the same thing. A 3-of-4 federal text, a five-purpose trust with no allocation rule, and four agencies sharing pieces of the management — that combination produces, for the school class, almost nothing measurable. The architecture is real on the page. The slice the schoolchildren get is, in practice, structurally invisible.

V. What the high-water mark left behind

By 1959 the section-grant project was over. It had run for a hundred and seventy-four years, from the 1785 Land Ordinance through the admission of Hawaii. It produced fifty states with deeply uneven school-trust outcomes — from Texas’s $58 billion Permanent School Fund and New Mexico’s $30 billion Land Grant Permanent Fund at the top, down through the Western public-land states in the middle, down through the original thirteen and the early-republic admissions where no federal grant ever arrived, and out to the two structural outliers at the end where the architecture was deliberately not built.

The 1910 Enabling Act was the federal government’s best work. Five protections stack in a single paragraph: express in trust, products-and-proceeds reach, breach of trust naming, null and void for non-conforming dispositions, and federal AG enforcement. New Mexico has the resulting fund corpus to show for it — though not because the federal text alone was enough. New Mexico has $30 billion because the 1910 high-water mark was paired with a state constitution that named and elected a single fiduciary, a doctrinal series running from Ervien (1919) through Lake Arthur (1921), Walker (1956), Lassen (1967), King v. Lyons (2011), and Moses v. Skandera (2015) that has read section 10 strictly,⁴⁹ and a hundred years of comparatively disciplined administration. Arizona, with the same federal text and a different state-side choice on the trustee, has $8 billion — and a 2024 Auditor General report flagging seventeen years of unadjusted agricultural lease rates and unmetered groundwater extraction that no one has yet sued to remedy. Same federal floor. Different state architecture. Different outcomes by a factor of four.

Oklahoma, with weaker federal text on the section grant, a bifurcated grant, and a state-side architecture that the Oklahoma Supreme Court in Williamson (1956) and Nigh (1982) read aggressively, has roughly $2.5 billion and one of the more developed state-court doctrines in the country. Alaska, with no school trust at all, has the largest sovereign wealth fund in the country going to a different beneficiary class. Hawaii, with strong federal language but a five-purpose dilution, has no isolable school-trust corpus.

The variance across the five states resists a single explanation. New Mexico’s $30 billion corpus rests on the combination of the 1910 federal text, the single elected commissioner of public lands, and a doctrinal series running from Ervien through King v. Lyons that has read section 10 strictly. Arizona, with the same federal text, ended up at one quarter the corpus and a 2024 Auditor General report flagging unmetered groundwater extraction and seventeen-year-stale agricultural-lease rates. Oklahoma’s bifurcated section-grant-plus-cash architecture produced a smaller fund and a stronger doctrinal record. Alaska deliberately wrote a constitutional rule against dedicated funds and built the country’s largest sovereign-wealth fund for a different beneficiary class. Hawaii’s five-purpose ceded-lands trust dilutes the school share into structural invisibility.

Two takeaways follow from all of this, and they are the things this chapter exists to leave the reader with.

The first is that the federal text is necessary but not sufficient. Congress wrote the strongest school-trust language it ever wrote in 1910. Two states received that text on the same day. One of them produced a $30 billion corpus and an enforceable doctrinal record. The other produced a corpus one quarter that size and a 2024 Auditor General report that reads, in places, like an enforcement complaint waiting for a plaintiff. The variable that explains the difference is not the federal floor, which is identical. The variable is the state-side architecture sitting on top of it: the trustee, the governance line, the doctrinal record, and the constituency.

The second is that the constituency is the load-bearing layer no statute can install. Most people don’t realize they are the beneficiaries of this trust. They don’t realize, because no one has told them, that the lands their state holds for the schools are not just public lands but trust lands, with the fiduciary duties that label imports. They don’t realize that they have a right to ask their school board and their legislators how the trust lands are being managed and how much money is reaching the schools. A regular citizen, knowing those facts and willing to ask those questions, becomes a watchdog. Multiply the regular citizens, and you have a constituency. Without the constituency, even the 1910 high-water mark drifts. With the constituency, even a weaker federal text — the Oklahoma cash settlement, the Utah pre-reform fund — can be brought to discipline.

That is the whole proposition. The architecture from 1785 to 1959 left us a federal floor whose strongest expression is the 1910 paragraph, a fifty-state record of state-side choices made on top of that floor, and a doctrine — Cooper, Lassen, Ervien, Williamson, Nigh, King v. Lyons — sufficient to enforce the architecture where the will to enforce exists. The architecture is real. The will is the question. And the will, in the end, lives in the constituency.

Margaret Bird’s anchor — the framing my own working-life inside this architecture has come to rest on — is that the school trust lands are a forever gift to forever schools for a forever democracy. The 1910 Act is the most serious attempt the federal government ever made to write that ideal into law. The state-side architecture either honored it or did not. The administrative discipline either kept faith or did not. Drift is one force; directed seizure is another; the absence of a structure to drift away from, as in Alaska, is a third. All three are present in the twentieth-century admissions.

After 1912, the section-template ended. Congress did not know it had ended. The next admission was forty-seven years away, and when it came, the 1910 paragraph was not what arrived. Alaska and Hawaii represent two paths the country tried but did not repeat. Neither produced a durable school-trust corpus by New Mexico or Utah or Texas standards. Whether the country could revisit the section-grant template, or build something new on top of Lassen and section 10, is not a question for this chapter. It is the question the rest of this volume — and the volume Dave will write next to mine — takes up.

What the architecture project from 1787 to 1959 left behind is fifty states with deeply uneven school-trust outcomes, a federal high-water mark in one paragraph of one statute admitting two states in 1910, and a doctrine sufficient to enforce the architecture where the will to enforce exists. The architecture is real. The will is the question. The constituency is where the will lives. And the constituency — as the next volume will press — has to be built, citizen by citizen, before any of the rest of it durably holds.

The forever gift was given. The forever schools were named. The forever democracy is the assignment. The architecture passed to us; what we do with it is not yet on the page.

— Margaret Bird, Salt Lake City, 2026.

Chapter 6 Footnotes