Sort the evidence by how settled it is, rather than by how often you hear it, and the argument changes shape. Most of what gets fought over is not actually in doubt. And most of what is genuinely in doubt barely gets mentioned.
Which Immigration
Asking whether immigration helps or hurts is like asking whether spending is good. Break the question apart and the pieces point somewhere specific, and not where the argument usually goes. The country has been getting richer from it the whole time the politics got angrier, which tells you the money was never what people were reacting to.
The first piece on this site that recommends anything, which is a departure worth flagging. This framework normally reads conditions and stops there. Here it is asked what would follow if its readings are right. Treat the measures as consequences of the analysis rather than as advocacy, and treat the section listing what would prove them wrong as the important one. Companion to the aid and migration comparison. Not policy advice.
The fourth band is the honest one. Whether immigration pushes down wages for Americans without a high school diploma is not settled. The disagreement between Borjas and Card is real, not partisan noise, and this piece is not going to pretend otherwise.
If those effects are real, then some of the gains are landing on exactly the families the Toilet Paper Theory is about. That is a serious objection, not a talking point.
The bookkeeping problem underneath everything
The third band holds the finding that explains almost everything about how this argument gets conducted. And it is barely about immigration at all.
The costs — schools, emergency rooms, local services —
land on counties and states, right away
The revenue — income tax, payroll tax, growing with time —
lands on the federal government, over decades
A county commissioner and the Congressional Budget Office
look at the same person and both do the math correctly.
land on counties and states, right away
The revenue — income tax, payroll tax, growing with time —
lands on the federal government, over decades
A county commissioner and the Congressional Budget Office
look at the same person and both do the math correctly.
That is not a failure of communication, and it is not a public that has been lied to. It is a bookkeeping problem. And bookkeeping problems can be fixed without changing a single admission number. Any analysis that treats local anger as irrational has skipped the only part of the arithmetic a local government can actually see.
The one thing this framework knows that others do not
Everything above is standard economics. What this framework adds comes from somewhere else: the hours ceiling.
The method here is always to ask what has actually run out. Applied across recent pieces, the answer has kept coming back real rather than fake. 36,000 artillery shells against a target of 100,000. Roughly 140,000 skilled shipyard workers who do not exist. Electricians named by Microsoft as the single biggest thing slowing data center construction. Transformers on 144-week waiting lists.
Those are hours ceilings. They do not answer money on the timescale money is being spent.
Immigration is the only thing that relieves an hours ceiling faster than a training pipeline, because a trained welder can arrive next month and a new one takes years. That is not a moral argument. It is the same arithmetic that made the munitions shortage real instead of fake.
Which means this framework, applied honestly, is friendlier to immigration in the skilled trades than most commentary, and it gets there without any appeal to values at all. It also says something narrower than the usual pro-immigration case. The argument is strongest exactly where a documented physical bottleneck exists, and much weaker where one does not.
Six things that follow
Reimburse local costs federally.
Washington collects the money eventually, so Washington should pay the bill now. This changes how many people come by exactly nothing, and removes most of the sensible local objection. It is the cheapest fix for the loudest complaint, and it is pure bookkeeping.
Compress the delay before work authorization.
Every month between arriving and being allowed to work turns a future taxpayer into a present cost, and helps nobody at all. It is the biggest single reason the early years look expensive, and we do it to ourselves.
Allocate a channel against documented occupational shortages.
Not by diploma and not by where someone was born, but against the actual bottlenecks: welders, electricians, linemen, transformer technicians. The trades whose absence is physically stopping construction right now. The hours ceiling is what makes this different from just preferring skilled people.
Buy predictability rather than a number.
School districts and employers can handle almost any steady number and cannot handle a wave. A surge costs the most in its first year, when capacity is fixed. A boring, announced figure beats a larger one that arrives all at once.
Make the lawful channel beat the smuggler on price and time.
What actually drives people to cross illegally is not a lack of enforcement. It is that the illegal route is faster and cheaper than the legal one for most people. A legal route that wins on price and time moves people without enforcing anything.
Cut the cost of sending money home.
Sending money home costs 6.36 percent worldwide, against a 3 percent target. Closing that gap puts about $20 billion a year back into families abroad, at no cost to America, and it is roughly thirty times the aid budget it would replace.
Notice what those have in common. Five of the six work by moving costs to where the benefits already are, or by removing a delay. Not by changing how many people arrive. The number is what the argument is about, and it is very nearly the only lever this analysis does not reach for.
What is actually on the table
Two things define where Congress currently is. The Secure America Act, passed in June 2026, sends $69.5 billion to immigration enforcement through 2029. The Dignity Act, introduced in July 2025 by Salazar and Escobar with support from both parties, is the serious comprehensive proposal still waiting.
The Dignity Act deserves credit where this analysis agrees with it, and it agrees more than expected.
| Measure | Dignity Act | Secure America Act | Divergence |
|---|---|---|---|
| Federal reimbursement of local costs | Absent | Absent | The biggest item here appears in neither bill. Nothing on the table touches the mismatch that drives the local anger. |
| Speed to work authorization | Substantially met | Not addressed | Dignity lets people work as soon as they enroll and decides asylum cases in 60 days. This is the clearest agreement, and the second biggest item here. |
| Visas allocated against occupational shortages | Absent | Absent | Dignity fixes the visa math: spouses and children not counted, per-country limits raised from 7 to 15 percent, easier visas for science doctorates. All of that changes who waits and for how long. None of it aims a visa at a shortage of welders. |
| Predictability over level | Partial | Not addressed | Faster decisions reduce the swings in the backlog, which helps. Neither bill commits to a steady announced number. |
| Lawful channel priced against the smuggler | Partial | Enforcement only | Dignity makes the legal route faster. The $69.5 billion of enforcement makes the illegal route dearer without making the legal one cheaper. One blade of the scissors. |
| Remittance transfer costs | Absent | Absent | This is a question about payments rules, and it sits outside immigration bills entirely, which is exactly why the biggest available lever keeps going unpulled. |
| The low-skill wage objection | Addressed | Not addressed | The American Workers Fund, paid for out of the $7,000 each participant owes, aims to retrain one American for every person enrolled. That is a real answer to the strongest objection, though a one-time fee is a thin basis for a permanent program. |
Three things worth changing
Three differences matter. The rest is detail.
First, nobody is paying the counties. No bill on the table sends federal money to the local governments carrying the early costs, and the federal government is the one collecting the eventual revenue. It is the cheapest available fix for the loudest available grievance, and not one person has proposed it. That is strange enough to be worth saying twice.
Second, the visa reforms answer a different question. Raising per-country caps from 7 to 15 percent, and not counting spouses and children against the total, are sensible repairs to a queue. They are not a way of aiming visas at anything. A country short of linemen does not need a shorter line for the same applicants. It needs a door marked linemen. This is the one place where this framework produces a proposal not already in circulation, and it comes from the hours ceiling rather than from any opinion about immigration.
Third, $69.5 billion bought one blade of a pair of scissors. Enforcement makes the illegal route more expensive. It does nothing to make the legal route cheaper. If what actually decides the matter is which route is faster and cheaper, then spending on only one side moves the ratio less than splitting the same money across both. And this framework’s standing question applies to that appropriation exactly as it applied to the shells: the money was found, and the question is whether the thing it was buying shows up.
The country is arguing about how many. The analysis says almost all the recoverable gains are in which, when, and who pays. Those are three separate questions, and not one of them is on the ballot.
What none of this settles
This framework can say three things. The country comes out ahead overall. The distribution is a choice, not a fact. And certain measures make both sides better off on the evidence available.
It cannot tell you how to weigh a large gain to a migrant against a small loss to an American worker. That is a values question. No amount of statistical work resolves it, and anyone whose spreadsheet answers it has smuggled a moral premise past you.
What would prove this wrong
| Claim | Measured by | Falsified if | Horizon |
|---|---|---|---|
| Local anger tracks who pays, not how many arrive | Local opinion in places that get their costs covered, against similar places that do not. | Covering those costs leaves local opposition exactly where it was, meaning something else is driving it. | by 2032 |
| The wait for a work permit drives most of the early cost | Whether each arrival group costs or contributes, sorted by how long they waited to work legally. | Groups allowed to work quickly cost about the same as groups made to wait. | by 2031 |
| Immigration fills a skills gap faster than training can | How long it takes to fill known trade shortages by visa, against how long by apprenticeship. | Home-grown training closes those gaps about as fast as bringing people in does. | by 2031 |
| Waves, not numbers, drive the cost of absorbing people | Local cost per arrival where the flow is steady, against places that got the same total all at once. | The cost per person is the same whether they arrived steadily or all at once. | by 2031 |
| Enforcement alone moves things less than fixing the legal route | Illegal crossings, against how the price and the waiting time compare between the two routes. | Enforcement spending cuts illegal crossings with no change at all in what the legal route costs. | by 2030 |
Three warnings, worst first. This piece recommends things, which this framework normally refuses to do. Discount it accordingly. Reading conditions and suggesting action are different jobs with different failure rates.
Second, the wage evidence in the fourth band is genuinely unresolved. These measures survive that uncertainty better than a simple case for more immigration would. They are not immune to it. Third, the bills were read from published summaries rather than from the legislative text, so a detail may have moved.
The treatise asks what has actually run out. Here the answer points two ways at once, which is unusual.
The shortage of skilled hands is real, and immigration is the only thing that fixes it inside a decade. The shortage of money in the counties absorbing new arrivals is fake, because the revenue exists and is simply turning up at a different address.
One of those gets fixed by letting people in. The other gets fixed by a transfer nobody has proposed. Neither one gets fixed by arguing about the number.