# The Deregulatory Morrill Act

"America as Access Provider" closed with a prediction. The Morrill Act for AI gets written, or the country stops being the country it has been. The historical record was the most credible prior available. The polity has resolved its self-contradictions by widening access infrastructure, not flattening outcomes. The 1862 mechanism was the live machine.

The directional prediction is structurally correct and the mechanism is wrong. The 1862 mechanism inverts somewhere between 1965 and 2008. What does the inverted mechanism look like, and why would a federally-funded Morrill Act for AI written under 2026 fiscal and regulatory conditions deepen the access stratification it intends to heal?

## The chart Andreessen pointed at

Marc Andreessen walked through Mark Perry's chart of the century with Joe Rogan in May 2026 (#2501), compressing it to a world where the college degree costs a million dollars and the flatscreen TV costs a hundred.

The chart is from the American Enterprise Institute and indexes US consumer-price changes against the overall CPI across a fixed basket of categories, January 2000 forward. The pattern is the chart's value. Through 2022: hospital services +220%, college tuition +178%, college textbooks +162%, medical care services +130%, child care +115%. CPI overall +74%. Wages +99%. On the other side: TVs -97%, toys -72%, computer software -70%, cell-phone service -41%. The categories that went up are the ones with the heaviest federal funding, licensing, and regulatory capture. The categories that went down are the ones closest to free-market discipline.

The mechanism behind the chart is not a mystery. Federal subsidy of demand without supply-side response inflates prices. Federal licensing of supply restricts entry. Both produce the same outcome from different sides. Higher education and healthcare ran both moves in parallel beginning in 1965 (Higher Education Act, Medicare and Medicaid) and running uninterrupted through the present. The 60-year empirical record is the experiment. Perry's chart is the experiment's readout.

## Where the 1862 mechanism worked

The historical sequence the access-infrastructure argument leans on contains five events: Homestead 1862, Morrill 1862 and 1890, GI Bill 1944, Highway Act 1956, Pell Grant 1965. Four of the five operated under conditions the 2026 sector does not share.

The Morrill Act of 1862 worked because higher education was not yet a regulated sector. The fifty-six state and federal land-grant institutions it seeded competed with private classical colleges in a market where the federal floor was new ground, not displaced incumbents. Tuition was low because cost structures were low. The marginal student could not access higher education at any price, because the institutional supply was thin and tilted toward the patrician class. Adding federally-financed supply democratized access by enlarging the producer base. The mechanism is supply-side expansion in a non-saturated market.

The Homestead Act of 1862 worked under the same condition. Public land was held by the federal government and unallocated. Distribution mechanisms were claim-based. The Act enlarged the producer base directly by converting state-held inputs to private productive ownership. The marginal homesteader could not access farmland at any price because the institutional supply was held off-market by federal title. Releasing it expanded the producer base.

The GI Bill in 1944 partially worked because higher-education capacity had expanded since 1862 but was still nowhere near saturation, the country had a war-justified compensation logic, and the regulatory layer atop higher education was thin. The mechanism is demand-side subsidy in a still-unsaturated supply environment.

The Pell Grant in 1965 is the inflection. The Higher Education Act and its Pell mechanism subsidized demand in a sector whose supply had been substantially saturated and whose regulatory layer was thickening. Per Mark Perry, college tuition began its detachment from CPI within a decade of the 1965 program and has not reattached since. The same inflection is visible in the medical-care series after the 1965 Medicare and Medicaid acts. The 1862 mechanism, applied a century later in a regulated-and-saturated sector, did not democratize access. It inflated price. The chart of the century is the record.

## Why the 2026 sector inverts harder

The 2026 education sector is past saturation, past regulatory thickening, and past the point at which marginal federal access funding produces marginal access gain. Three structural differences from 1862 make the inversion sharper than the 1965 case.

First, the binding constraint on amplification has moved. In 1862 the marginal student could not access the technology of higher education at any price. In 2026 the marginal student can access an OpenAI free tier, Anthropic's Claude.ai free tier, Google's Gemini free tier, and a public-library terminal that gets her to any of them. The technology is not the bottleneck. The bottleneck is the complementary skill stack of literacy, numeracy, judgment, and taste that converts access into amplification. That stack is produced by the broken K-12 layer the historical-pattern argument leaves unnamed.

Second, the fiscal envelope has closed. The federal deficit ran at $1.8 trillion in fiscal year 2024 and CBO projects a primary deficit averaging 3.5% of GDP through 2034. The Pell Grant program ran $31 billion in FY2023 against a higher-education sector whose 1965 cost structure was roughly 5% of current per-pupil expenditure in real terms. A Morrill Act for AI scaled to the AI build-out's capital intensity would need to be an order of magnitude larger than Pell and would arrive into a fiscal environment in which mandatory-spending growth in Medicare, Social Security, and interest service is already crowding out the discretionary capacity historical access infrastructure relied on.

Third, the polity has polarized along the regulation-vs-deregulation axis the chart of the century names. The Trump administration's December 2025 executive order on AI and the March 2026 National Policy Framework propose federal preemption of state AI laws and explicit deregulatory posture. The Biden 2023 EO 14110 was the inverse: federal regulatory expansion. AI policy is now a partisan football, and any major federal access program is read by half the polity as the regulatory-capture mechanism the chart documents and by the other half as the access infrastructure the historical pattern names. The historical Morrill Acts passed in conditions of low polarization and broad institutional trust. The 2026 polity has neither.

## What the deregulatory candidate looks like

A Morrill Act 2 that would actually heal access in 2026 would not look like the Pell Grant. It would look like the unwinding of a Pell Grant. The structural moves the chart of the century predicts would help share a common shape: each removes a piece of the post-1965 demand-subsidy-and-credentialing scaffolding that the chart documents as the cost-spiral driver.

Strip federal credentialing requirements from the labor markets adjacent to the AI build-out, including nursing assistants, paralegals, K-8 tutoring, and accounting clerks. The structural reason this move is the right shape is the Baumol-disease wedge — the principle that labor-intensive sectors with restricted entry inherit the rising-wage trajectory of competing sectors without the offsetting productivity gain. AI-equipped operators competing with credentialed incumbents at the entry layer is the wedge that introduces productivity gain into the sector for the first time since the credential layer was built. The 1862 supply-side expansion mechanism returns, by way of removing the entry restriction rather than adding new supply.

Replace the federal-student-loan-as-guaranteed-revenue structure with means-tested vouchers redeemable at any accredited provider, including AI-native institutions that did not exist when the current accreditation regime was written. The structural reason this move is the right shape is the supply-side response the original Pell Grant did not produce: guaranteed revenue with inelastic supply produces price inflation, full stop. The redirection of the same federal dollars through a competitive supply environment produces a different price trajectory. The amount of federal money does not have to change. The shape of the federal money does.

Open the K-12 layer to AI-native curriculum providers via state-level charter expansion. The structural reason this move is the right shape is that the binding constraint named in the first inversion-condition (the broken complementary-skill stack) is produced by the K-12 layer. Federal access programs that route around K-12 deliver amplification tools to students whose complementary-skill stack does not support amplification. The federal lever here is funding-conditional preemption, the same instrument the March 2026 EO proposes to use on AI safety rules. The asymmetric prediction is that this lever gets used on AI safety preemption and not on K-12 curriculum, because the political coalition for the first exists and the coalition for the second is structurally weaker.

None of these is a Morrill Act in the 1862 mechanism. Each is a deregulatory move in the opposite mechanism. The historical-pattern argument is right that the polity will respond, and right that the historical record favors response. The amendment is that the response that would actually heal looks structurally inverted from the response the parent essay predicts. The 1862 supply-side mechanism became the 1965 demand-side mechanism became the 2026 mechanism of removing the regulatory-and-incumbent-protection scaffolding that 1965 demand-subsidy and credentialing built up over sixty years.

## The compressed self-contradiction

The country resolves its self-contradictions, in the long run, because it has to. The access-infrastructure argument holds. The amendment is that the resolution machine the country uses changes shape across centuries. The 1862 machine was the federal land-grant. The 1944 machine was the federally-financed credentialing pipeline. The 1965 machine was the demand-subsidy through a credentialing pipeline already at capacity. The 2026 machine has to be the one that removes pieces of the 1965 machine without removing the country's commitment to the open kingship the historical pattern named.

Whether the polity can run a deregulatory access expansion is the open question. The historical record contains one direct precedent: airline deregulation 1978, signed by Carter after the Civil Aeronautics Board had become bipartisan-suspect, with consumer-price evidence visible inside the trucking and natural-gas adjacent deregulations of the same period. The political coalition was trust-busting Democrat plus Reaganite Republican, the same shape Reagan-era telecom deregulation and Clinton-era welfare reform also produced. The structural feature that made the move possible was producer-side regulatory consensus losing its bipartisan cover under consumer-price pressure the polity could no longer absorb. The 2026 chart of the century is the consumer-price evidence already visible. The producer-side regulatory consensus on higher education has not yet lost its bipartisan cover; this is what the next political cycle is about.

The Morrill Act for AI will not be written in the form the access-infrastructure argument expects. It might be written in the inverted form. The country needs the inverted form to keep the open kingship reachable from the access-stratified bottom decile. The country may or may not produce it. The historical record gives one precedent and several centuries of confidence in the resolution machine. The next several years are the experiment.

provenance · first_seen 2026-05-21T01:49:50Z · drafted 2026-05-21T01:49:50Z · published 2026-05-21T11:09:10Z · edited 2026-05-24T16:30:57Z