This is a proof-of-concept, not a finished bill. Large parts of it were drafted collaboratively with Claude, an AI model, across many rounds of revision, argument, and correction — a practice I've disclosed throughout this series. I am not a demographer, an immigration attorney, or a member of Congress. I am looking for feedback, for the holes, for the places the arithmetic doesn't hold. Tell me where it breaks.
Confronted with the possibility of real abuse — an unreviewable Secretary imposing a fifty-year "temporary" designation, or ending one on a coin-flip — Justice Alito didn't dispute it could happen. He held that courts aren't the body to police it, and that Congress "would have ample means" to, including through appropriations.
He named the failure mode. The judiciary removed itself. The remedy has been assigned, by name, to Congress.
Challenge accepted.
Image generated by ChatGPT
The arithmetic is the argument. Where it isn't, I've said so plainly, in a section built for exactly that purpose.
Securing the Border by Opening It: An Immigration Framework
A claim that sounds like a contradiction: the fastest way to secure the border is to let far more people cross it legally. The fastest way to make deportation workable is to make it rarely necessary. The fastest way to fix the asylum system is to leave the asylum statute almost untouched.
Each of these sounds like a trick. It isn't. Nearly every hardship commonly attributed to immigration — an overwhelmed border, an abused asylum system, a permanently "temporary" protected-status program, wage suppression, an unassimilated underclass, a fiscal drain — turns out, on inspection, to be caused not by the migration itself but by a mismatch: the gap between what the legal immigration system permits and what the economy, the labor market, and separated families actually demand. When that gap is wide, as it has been for decades, the excess demand doesn't disappear. It routes around the law. It becomes the border crossing, the asylum claim filed as a workaround, the TPS renewal that never ends.
Close the gap and the symptoms recede together, because they share one cause.
Everything is easier if you do it right.
This is not an argument for open borders, and it is not an argument from charity. This is an argument that the disorder everyone is reacting to is a design failure, and that the design is fixable. It is also an argument with a hard conditional built into it. "If you do it right" is not throat-clearing — it's the entire content of the proposal. Everything below is sequential. Read in isolation, half of this is naïve. Read as a system, it resolves seven fights most people think are unrelated to a single fix.
Part 1: The Problem, Quantified
The case for reform usually opens with the border, because the border is what people see. This opens with arithmetic instead, because the border is not, in the end, the largest number in this story.
The U.S. total fertility rate has been below the 2.1 replacement level since 1972 — over half a century. Without immigration, the U.S. working-age population would have started shrinking in 2012. Births can't rescue this math: a baby born in 2026 doesn't become a taxpayer until roughly 2044, well after every entitlement trust fund in this piece hits crisis. Immigration is the only lever that adds working-age, tax-paying adults on a timeline that matters.
CBO projects potential GDP growth averaging 1.8%/year through 2055, down from 2.4% over the prior 30 years — and attributes roughly five-sixths of that decline to slower labor-force growth, which is the immigration-addressable part. Modeling the cost of continuing at 1.8% against a restored-immigration counterfactual:
| Scenario | Cumulative lost output (30 yrs) | Per household/yr, year 30 |
|---|---|---|
| Conservative | ~$61 trillion | ~$35,800 |
| Moderate | ~$104 trillion | ~$61,500 |
| Full CBO gap | ~$127 trillion | ~$74,800 |
Even the conservative case costs more than two years of current U.S. GDP. The cost is invisible early and brutal late — which is exactly why the status quo feels free and isn't.
The entitlement math doesn't need modeling; the government already did it. The 2026 Social Security Trustees Report moved OASI depletion up to Q4 2032 — an automatic 22% cut to ~70 million people — and the 75-year shortfall jumped $4.2 trillion in a single year, a deterioration the trustees attribute partly to a lowered immigration assumption. Medicare's trust fund depletes even sooner, in 2033. The worker-to-beneficiary ratio goes from 2.9-to-1 today toward 2.2-to-1 by mid-century.
Downstream of the fiscal numbers: immigrants are ~28% of the direct-care workforce staffing the eldercare system the boomers are aging into. Immigrants founded or co-founded ~55% of U.S. unicorns and ~46% of Fortune 500 companies. The chain runs GDP shortfall → Social Security cut → Medicare cut → debt spiral → unstaffed eldercare → unbuilt housing → strangled innovation → a slow national ossification. Each survivable alone. The claim here is that they compound, share a root cause, and yield to one lever.
Part 2: The Front Door — How Many, and How Set
There's no single "right number" — it depends on the target. Keeping the working-age population from shrinking (the floor) takes roughly 450–500K net immigrants/year; current policy, at a projected 321K for 2026, sits below that floor. Sustaining 2010s-era growth takes ~1M or more. Holding the old-age dependency ratio constant would take on the order of 10 million a year — which is not a policy option, it's proof that immigration is necessary but not sufficient. It buys two to three decades. It has to be paired with a higher retirement age, productivity gains, and whatever fertility recovery is achievable.
The working band: a statutory floor of ~500K, a target of 1.2–1.5 million, and a ceiling near 2 million — set not by ideology but by physical absorptive capacity. Canada ran the natural experiment in 2023: population grew 3.2%, mostly from immigration, and the result wasn't a values debate, it was arithmetic — housing demand outran construction, per-capita GDP fell, and public opinion flipped so hard the government cut targets 20%. Overshooting the ceiling doesn't just strain services. It destroys the political coalition for immigration itself.
Don't legislate a number — legislate a mechanism, the way Congress gave the Fed a mandate instead of setting interest rates directly. A standing, independent levels body reads housing starts, labor-market data, and processing throughput, and sets the annual number inside a statutory floor and ceiling Congress writes. We set the guardrails and the machine. The machine sets the number.
Part 3: Selection — Who, and On What Basis
A points-and-needs hybrid, every channel clearing the same universal floor first: a criminal-and-terror screen, and a civic/behavioral screen — does the applicant accept the constitutional order and a pluralistic society, tested on conduct, not private belief. Above the floor, scoring runs on age (a 28-year-old delivers ~40 years of contribution before drawing benefits; a 55-year-old delivers ~12 — this is the dependency-ratio fix, stated as a formula), job offer, shortage-occupation skills, English proficiency, education. Notably absent: national origin, race, religion — not merely on principle, but because none of the three predicts contribution.
On country caps: the current 7% per-country limit was written in 1990 to prevent discrimination. Its actual effect is the purest version of it. India generates over half of employment-based demand and gets the same allocation as Liechtenstein; the backlog runs 12–13 years, with new applicants facing 50-to-80-plus-year waits, and more than 400,000 people in the queue are expected to die before they're reached. The system isn't malfunctioning — it's functioning exactly as designed, and the design is the discrimination. The fix is to eliminate hard per-country ceilings and score individuals; origin isn't a field on the form. Short-run results will skew toward India and China simply because that's where the backlog is — that's a feature of fixing the discrimination, not a new one.
Criteria get more statutory protection than the annual number, not less — because criteria, not volume, are where discrimination hides. Congress writes a closed list of permissible factors and bans origin outright; a standing body only sets the point values inside that list, with weights and outcomes published annually.
Part 4: The Strategic Case — A Race for a Shrinking Pool
Global fertility is 2.25 and falling; 71% of humanity already lives below replacement; by 2100, a Lancet-affiliated projection puts 97% of countries below replacement. This means two things at once: migration pressure from traditional sending regions should decline over the coming decades on its own, and the world's defining mid-century competition flips from "keep people out" to "attract a shrinking pool of working-age people," with every aging rich nation bidding for the same shrinking supply.
China is the clearest illustration of what happens when a nation can't compete for that pool. Its fertility rate sits near 1.0. Its population has been shrinking since 2022 — the first sustained decline since the Great Leap famine — and its worker-to-retiree ratio goes from roughly 8-to-1 today to 2-to-1 by 2050, aging before it's gotten rich. And it structurally cannot fix this with immigration: its foreign-born population is about 0.1% of its total, against 14% in the U.S., because Chinese citizenship is essentially blood-based and a 2020 proposal to loosen it was shelved after nationalist backlash. China can rent talent. It cannot absorb it. Its own best founders and students disproportionately leave — for the United States.
China's demographic collapse isn't bad luck running parallel to an unrelated slowdown. It's the direct, foreseeable consequence of an ethnic-nationalist conception of belonging — a door closed by its own definition of itself. The single most consequential unforced error the U.S. could make in this competition is adopting the same instinct.
Part 5: TPS — The Trigger for This Document
This is placed last among the substantive sections because it's the reason the rest of this piece exists. On June 25, 2026, the Supreme Court decided Mullin v. Doe, 6–3.
The Court held that the TPS statute bars judicial review of DHS's decisions to terminate or extend a designation, and rejected an equal-protection challenge from Haitian TPS holders — Justice Alito writing that the record didn't establish the termination was race-based; Justice Kagan's dissent argued the racial motivation "fairly shouts." Work authorization for roughly 350,000 Haitians and 6,000 Syrians terminates on the administration's timeline, and the ruling clears the path to end designations for Venezuela, Somalia, and Ethiopia.
The diagnosis this piece opened with — that there's nothing temporary about a program renewed for a quarter-century — is no longer contested. It's the government's own stated justification: DHS's general counsel has said outright that "the T in TPS stands for TEMPORARY, yet many of these designations became de facto amnesty." The question isn't whether TPS is broken. It's what a humane remedy looks like — and Mullin supplies the sharpest possible argument for why that can't be left to executive discretion.
Confronted with the possibility of real abuse — an unreviewable Secretary imposing a fifty-year "temporary" designation, or ending one on a coin-flip — Justice Alito didn't dispute it could happen. He held that courts aren't the body to police it, and that Congress "would have ample means" to, including through appropriations.
He named the failure mode. The judiciary removed itself. The remedy has been assigned, by name, to Congress.
Challenge accepted.
What follows is the statute Congress could write: a replacement status with a hard statutory clock (three to five years, not executive-renewable) and a forced binary at the terminus — either conditions have genuinely improved and the protection ends with a funded return, or the situation is durably permanent and the holder converts automatically into the standing front-door channel from Part 2. No third option of indefinite renewal. It works only bound to that front door — without one, "convert to the standing channel" is a meaningless instruction, and the program becomes TPS again under a new name.
(One honesty note: Alito's "ample means" line was a brief rejoinder, not a considered endorsement of this specific reform. The accurate claim is that he pointed at the correct branch — this is what a serious answer from that branch looks like. That's a stronger claim than saying the Court asked for this, and it's the one this piece makes.)
Current long-tenure holders — twenty-plus years, spouses of citizens, parents of citizen children — convert to permanent status upon clearing the same universal screen applied to every other channel in this framework. Not forgiveness without scrutiny — actual scrutiny, applied to people the system left in renewable limbo for a generation.
Part 6: The Hot Buttons, Resolved by the Same Design
Asylum can't be eliminated — non-refoulement is a binding treaty obligation, and ending it is treaty withdrawal, not a policy tweak this piece proposes. But the defect was never the standard; it's the sequencing. A ten-to-twenty-year legal wait makes a weak asylum claim the only way to live and work here in the meantime, burying genuine claims under opportunistic ones. This is the sharpest instance of the whole thesis, because the "tough" response and the correct response point in opposite directions: cracking down punishes the genuine refugee and the opportunistic filer identically, because enforcement can't tell them apart fast enough. Opening the front door removes the incentive to file a weak claim at all — the backlog drains toward the population the statute was written for, and what's left gets adjudicated fast and fairly. The restrictionist and the humanitarian both get what they're actually asking for, and the fix never touches the asylum statute.
Enforcement conduct — the raids, the wrongful detentions, the due-process shortcuts that dominate this debate — isn't addressed here through new oversight rules, because under this design there's little left for them to do. An apparatus tasked with chasing millions of people whose only violation is working without legal status will overreach; scale demands it. Shrink the unauthorized population to the genuine bad-actor residual and the machinery shrinks with it. This is a volume problem wearing the appearance of a conduct problem.
Part 7: The Population Already Here
"11 million illegals" is a fiction of aggregation. 80% have lived here at least five years; 45% for twenty years or more. Fourteen million U.S. citizens and green-card holders share a household with an unauthorized immigrant. 6.3 million children live with an unauthorized parent — all but a million of those kids are U.S. citizens. More than 40% already hold some form of legal protection.
Run this population through the same screen everyone else in this framework faces. Those who fail the criminal-and-terror screen are the small, well-defined population enforcement should already be concentrated on. Those who clear it sort by the same logic as the front door: long-tenure residents convert to legal status; the "twilight" millions already holding TPS, parole, or pending claims resolve into permanent status instead of renewable limbo. This is adjudication, not amnesty — the same scrutiny applied to a new applicant, applied for the first time to people the system simply left unexamined.
The 1986 amnesty failed because it legalized the existing population without fixing the channel feeding future demand — the shadow population refilled within a decade. Resolve this population and open the front door at the same time, and there's no unmet demand left to refill it.
How This Fails
A framework built entirely on favorable assumptions isn't an argument — it's advocacy wearing arithmetic as a costume. Stated plainly:
The cost model depends on admitted immigrants being working-age and employed; a system that admitted dependents would deliver none of it. Immigration buys two to three decades against the age-structure problem — it doesn't repeal it. A levels body can be captured or overridden, the way the Fed contends with political pressure constantly. The timing of the global fertility inversion is uncertain, even if its direction is about as solid as demographic projections get. Even origin-neutral factors like English proficiency correlate with origin — bounded by transparency and funded acquisition, never fully eliminated. Every statutory forcing function here is a statute, and a future Congress can unwind any statute — "harder to undo" is not "impossible to undo," it's just a materially higher bar than the unreviewable executive discretion Mullin just confirmed governs the status quo. And every piece of this is sequential — none of it works as a standalone reform.
Summary
This piece opened with a claim that sounds like a contradiction — that security, humane treatment, and fiscal responsibility aren't competing priorities to trade off, but downstream effects of one design choice, made correctly or made poorly.
The evidence isn't rhetorical. It's arithmetic, assembled across eight nominally separate fights: a $61–127 trillion cost of continuing as we are; a trust fund six years from an automatic cut, with the government's own trustees naming reduced immigration as a contributing cause; a Supreme Court ruling that just handed Congress, by name, the job this piece takes up; an asylum system that gets more accurate and more humane by being left alone; a cap that produces the exact discrimination it was written to prevent; and a rival whose demographic collapse is the direct consequence of the same door-closing instinct now being proposed here in response.
None of it depends on generosity, and none of it depends on austerity. It depends on getting the design right.
The obstacle was never the arithmetic. Every piece of this proposal already exists in some bill somebody in Congress has already written. The obstacle is political — a compromise this comprehensive gives no single faction full credit, and full credit is what the current incentives reward.
The arithmetic is the argument. The obstacle is political, not arithmetic.
The full policy document — with sourcing, the complete selection framework, and the "how it fails" section in full — is linked below.
