Immigration

He ran against illegal immigration. He is shutting down the legal kind too.

Donald Trump promised to secure the border and deport criminals. By mid-2026 he had done that, and far more: cut refugee admissions to the lowest level on record, stripped legal status from over a million people, made a green card slower to get than at almost any time in history, and presided over the first year of negative net migration in roughly half a century. This is a tour of what was promised, what was done, what it costs, and why the case for it rests on claims the evidence does not support.

How to read this

Labels, not adjectives.

Immigration runs hot, so this page is built to be checked, not to shout. Each load-bearing claim is tagged by how solid it is:

Documented a policy action or official number on the record.   Strong evidence a robust research finding or consensus.   Mixed a genuine tradeoff where the honest answer has two sides.   Contested experts disagree or it is unsettled.

The myths section gives each claim its strongest version and its kernel of truth before weighing it against the evidence, because a debunk that ignores what is real about a fear does not persuade anyone who holds it. Where immigration carries genuine costs, this page says so plainly.

First, the terms

"Legal" and "illegal" are blurrier than the slogans.

Almost every immigration argument collapses two different things, so it is worth being precise before anything else. The roughly 52 million foreign-born people in the United States are not a single group, and most are not here illegally.

Most are here legally
About 73% of the foreign-born are here lawfully: roughly 46% are naturalized US citizens, 23% are green-card holders, and 4% hold temporary visas (students, workers). About 27% are unauthorized. (Pew Research, 2025)
"Illegal" is not one thing
A large share of unauthorized immigrants, by some estimates around 40%, never crossed a border illegally. They entered on a valid visa and overstayed it. And being present without status is generally a civil violation, not a crime; improper entry is a misdemeanor and illegal reentry a felony, but simply overstaying is not a criminal offense. (Pew Research; Center for Migration Studies; 8 U.S.C. 1325-1326)
Status is a spectrum, not a switch
Millions live in lawful-but-precarious categories: asylum seekers awaiting a hearing, holders of Temporary Protected Status or humanitarian parole, DACA recipients. They are here legally, until a policy change says otherwise.

That last point is the hinge of this whole story. When the administration terminated Temporary Protected Status and parole programs, it did not, by itself, deport anyone. It converted more than a million people who were here legally into "illegal" immigrants, with the stroke of a pen. The line between legal and illegal is not only something migrants cross. It is something the government draws, and redraws. Documented

Both slogans flatten the same reality. "They're all illegal" is false, since most immigrants are here lawfully. And "just come legally" assumes a line that, for most people, does not exist (Part 2 returns to this). The honest picture is a set of narrow, capped, and shifting categories, in which legal and illegal are less a moral binary than a status that policy can grant or strip. This page keeps the two distinct, and flags which one each claim is actually about, because most of the crackdown described next fell on the legal side.

Part 1 · The promise and the program

What he said he would do, and what he actually did.

The campaign was about the border and about criminals: stop the crossings, deport the dangerous. By one measure, the border part worked. Southwest border encounters fell from about 1.53 million in fiscal 2024 to about 238,000 in fiscal 2025, the lowest in any year since 1970. (Pew Research, CBP) Documented

But the program reached far past illegal immigration. The same administration moved, piece by piece, against the legal system: refugees, asylum seekers, temporary-status holders, work-visa holders, green-card applicants, and even birthright citizenship. Here is the record by mid-2026.

Refugees
The refugee program was suspended in January 2025, and the annual ceiling was set at 7,500, the lowest in the program's history, down from 125,000. The priority group named was white South Africans. (White House EO 14163; Federal Register; Global Refuge)
Legal status stripped
Termination of the CHNV parole program (about 500,000 people) and Temporary Protected Status for several countries together canceled the legal status of more than 1.5 million people who were here lawfully (Venezuela ~600,000, Haiti ~330,000, Afghanistan ~17,000). The Supreme Court let the parole terminations proceed. (DHS; WLRN; SCOTUSblog)
Asylum
A January 2025 proclamation suspended asylum at the southern border and shut down the CBP One appointment app, canceling roughly 30,000 scheduled appointments. (NBC News; Democracy Forward)
Work visas
A September 2025 proclamation imposed a $100,000 fee on certain new H-1B skilled-worker petitions. A federal court vacated it in June 2026; a conflicting ruling means its status is unsettled. (American Immigration Council; Al Jazeera)
Travel ban
A June 2025 proclamation barred entry from 12 countries fully and 7 partially, later expanded. (CRS; American Immigration Council)
Citizenship itself
An executive order sought to end birthright citizenship for children of unauthorized or temporary residents (enjoined by every court to rule, with a Supreme Court decision expected around mid-2026), and the Justice Department made stripping citizenship through denaturalization a top civil priority. (SCOTUSblog; NPR)
Green cards
Green-card approvals were reported cut by roughly half over the year, with DHS moving to stop granting adjustments except in extraordinary cases and pushing applicants to apply from abroad. Employer-sponsored processing hit an all-time high of about 3.4 years. (Cato Institute; NPR)

The throughline: this was not only about people crossing illegally. It was a broad contraction of legal immigration and a challenge to citizenship by birth. The clearest single proof is the bottom line below.

Negative
net migration in 2025, the first time in roughly 50 years (Census estimates; Brookings)
~ -220,000
CBO's projected net migration for 2026, more people leaving than arriving
7,500
refugee ceiling, an all-time low (was 125,000)
Part 2 · The deportation machine

A historic budget, a marginal result, a much wider net.

Enforcement became the centerpiece, so it is fair to judge it on its own numbers. Start with the baseline: ICE's annual budget had been around $10 billion, roughly $3.9 billion of that for detention. The 2025 budget law then poured about $170 billion into immigration enforcement, adding roughly $75 billion to ICE over four years and about $45 billion for detention alone, several times its prior level. ICE is now the highest-funded law enforcement agency in the federal government. (American Immigration Council, 2025; Brennan Center; NPR, Jan 2026) Documented

For all of that money, actual removals rose only about 7 percent over the previous year. The spending did not buy a surge in deportations. It bought a surge in detention: the population held roughly doubled in a year to a record of about 70,000 by January 2026. (TRAC, Nov 2025; Stateline, Feb 2026) Documented

Mostly not criminals

The promise was "the worst first," violent criminals. The government's own data show the opposite. By late 2025, about 74 percent of people in ICE detention had no criminal conviction, and only about 5 percent had a violent one. Nearly all of the growth in detention came from people with no criminal record at all. (TRAC Syracuse, Nov 2025; Cato Institute, Dec 2025) Documented

~74%
of ICE detainees had no criminal conviction (TRAC, late 2025)
~5%
had a violent conviction (Cato)
~70,000
in detention by January 2026, a record

DHS counters that 70 percent of arrests are of people "convicted or charged" with crimes, but that figure folds in people merely accused and routine immigration violators. Both can be partly true: some of those removed are genuine, serious offenders, and DHS publicizes them. But the conviction-based numbers are clear that the net widened far beyond the "worst of the worst." The Laken Riley Act, the first law Trump signed in 2025, made detention mandatory for people merely charged with minor offenses like shoplifting. Mixed

How they are found

The dragnet reaches into ordinary life. ICE ran sweeps at Home Depot parking lots, car washes, and worksites, and made arrests at courthouses and at immigrants' own check-in appointments. The administration rescinded the longstanding policy keeping agents away from schools, churches, and hospitals. It expanded agreements deputizing local police to more than 1,000 agencies. And it moved to feed ICE data the government had never used for enforcement, including IRS tax records and Medicaid enrollment files. (DHS, Sep 2025; ACLU; KFF, Jan 2026; courts have blocked parts of the data-sharing) Documented

There is a bitter irony in the tax piece. Undocumented immigrants paid an estimated $97 billion in taxes in 2022, much of it filed voluntarily through IRS-issued taxpayer numbers. The administration then sought to hand that very tax data to ICE to locate them. The compliance the system invited became the map for the raids.

Where people are sent

Often to their home countries, but increasingly to third countries that are not theirs at all. In March 2025 the government flew about 238 Venezuelan men to El Salvador's CECOT mega-prison, most under the Alien Enemies Act of 1798, a wartime law invoked in peacetime, and paid El Salvador several million dollars to hold them, with no hearings. Membership in the gang it cited was frequently assigned on the basis of tattoos and clothing that experts on that gang call unreliable. Groups were also sent to South Sudan and Eswatini, countries to which most had no connection. Courts pushed back: the Supreme Court held that even under that 1798 law, people are owed notice and a chance to challenge their removal. (NPR, Dec 2025; SCOTUSblog, Jul 2025; CNN, Oct 2025; Supreme Court, May 2025) Documented

Removal first, review later

Immigration court is civil, not criminal, so there is no right to a government-appointed lawyer; people facing deportation, including children, must pay for counsel or face the system alone. On top of that thin baseline, the administration expanded "expedited removal" nationwide in January 2025, letting officers deport someone in as little as a day, with no hearing before a judge, unless that person can immediately prove more than two years' continuous presence (a federal court paused the policy in August 2025). (NAFSA, Jan 2025) Documented

The net also reached people with legal status. Green-card and student-visa holders were detained, in some cases over political speech, and judges ordered several released on First Amendment grounds. In at least one instance the government deported a man in direct violation of a standing court order, then conceded the removal was an error. The recurring pattern is removal first, and review, if it comes at all, only afterward. (PBS NewsHour, 2025; CNN, May 2026) Documented

Deaths in custody

The surge in detention has been measured in lives. At least 32 people died in ICE custody in 2025, up from 11 the year before, the most in roughly two decades, even as the agency's own safety inspections fell sharply. The deaths did not stop with the new year: an 18th was reported within the first four months of 2026, putting the year on pace for a new record. About three-quarters of the people being held had no criminal conviction. (NPR, Oct 2025; POGO, 2025; CBS News, 2026) Documented

32+
people died in ICE custody in 2025
11
died the year before
~20 years
since custody deaths were this high

The shape of it: a record budget and record detention produced only about a 7 percent rise in actual removals, while the people swept up were, by the government's own data, mostly not criminals, and in some cases here legally. This was less a targeted hunt for dangerous people than a wider and far more expensive net.

Part 3 · The consequences

What it costs, in jobs, prices, and growth.

Immigration is woven into the parts of the economy most Americans rely on without seeing. Pull the thread and things unravel in predictable ways.

The work that depends on immigrants

Roughly half or more of hired crop farmworkers are foreign-born. About 27 percent of direct-care workers (the people who care for the elderly and disabled) are immigrants, and 42 percent of home health aides are foreign-born. Construction is about 23 percent foreign-born. (USDA; PHI 2024; American Immigration Council) Documented

When that labor is removed, the result is not a wave of natives filling the jobs. Studies of past enforcement (the Secure Communities program) found that crackdowns reduced the employment and wages of U.S.-born workers too, because immigrant and native labor are often complements, not substitutes. Modeling of mass deportation projects higher consumer prices, on the order of 1.5 percent to 9 percent by 2028 depending on scale, in food, housing, and care. (Journal of Labor Economics; Peterson Institute) Strong evidence magnitude uncertain

The growth left on the table

The nonpartisan Congressional Budget Office estimated that the 2021-2026 immigration surge would raise GDP by about $8.9 trillion over a decade and reduce federal deficits by roughly $0.9 trillion, because immigrants work, spend, and pay taxes. Run that in reverse and restriction shrinks the workforce and slows growth. (CBO, July 2024) Strong evidence

The aging math

The country is getting older and the ratio of workers to retirees is falling. Immigration is the main thing slowing that decline; it modestly improves Social Security's long-run balance and adds prime-age workers. Cutting it accelerates the squeeze on the programs older Americans depend on. (Penn Wharton Budget Model; SSA Trustees) Strong evidence a tailwind, not a cure

The "line" that does not exist

The crackdown also made the legal path, the one critics tell people to use, harder. For most people who want to immigrate, there is no line to get into. One analysis estimates that over 99 percent of people who want to immigrate legally have no eligible category at all. For those who do, the waits are measured in decades. (Cato Institute; American Immigration Council) Documented

~25 years
current wait for a sibling of a US citizen from Mexico (State Dept. Visa Bulletin, June 2026)
~1.8M
people in the employment green-card backlog; new Indian applicants face a projected wait longer than a lifetime
<1%
odds in the diversity lottery: ~20 million entries for ~55,000 visas

Leaving the country to "do it the right way" can itself trigger a 3- or 10-year bar on returning, a catch-22 that traps people who would gladly use a legal route. (American Immigration Council) Documented

Part 4 · The claims

The talking points, weighed against the evidence.

The case for restriction rests on a set of claims about immigrants. Here they are, each given its strongest version, then measured against the research.

Immigrants commit more crime.

The kernel of truth
Individual crimes by people here illegally are real and sometimes horrific, like the 2024 murder of Laken Riley, and they receive heavy coverage.
What the record shows
Immigrants, including the undocumented, are incarcerated and convicted at lower rates than native-born Americans. Cato's analysis of census data put the 2024 incarceration rate at 674 per 100,000 for illegal immigrants versus 1,195 for the native-born. In Texas, the one state that tracks status, undocumented immigrants were convicted of crimes at roughly half the native-born rate, including for homicide. A peer-reviewed study in PNAS found the same using arrests and convictions. A single tragedy is real, but it cannot set a group's crime rate, which requires dividing offenses by population.
Not supported by the evidence. Strong evidence
Cato (2026); Light, He & Robey, PNAS (2020); Texas DPS; NIJ

They take jobs and lower wages for Americans.

The kernel of truth
This one has a real exception. For the workers most interchangeable with low-skill immigrants, native-born adults without a high school degree and earlier immigrants, there can be small negative wage effects in the short run. That is a genuine cost, not a myth.
What the record shows
The number of jobs is not fixed: immigrants are workers and consumers and founders who expand the economy. The National Academies of Sciences found that over 10 or more years, the effect on the wages of native-born workers overall is "very small," with little effect on employment. Immigrants often do work that complements native workers, letting natives move into higher-paid roles. On average, immigration slightly raises native wages over the long run.
False as a sweeping claim; true for one narrow group in the short run. Mixed
National Academies (2016); Ottaviano & Peri; Peri & Sparber; the Card/Borjas Mariel debate

They don't pay taxes and drain welfare.

The kernel of truth
First-generation immigrants are, on average, a net cost at the state and local level, mainly because local governments pay to educate their children. Emergency care and public schooling are real public expenditures available regardless of status.
What the record shows
Undocumented immigrants paid an estimated $96.7 billion in taxes in 2022, including about $34 billion into Social Security, Medicare, and unemployment insurance, programs they generally cannot collect from. They are barred from SNAP, non-emergency Medicaid, the ACA, and SSI. Their average state and local tax rate (8.9 percent) exceeds that of the top 1 percent (7.2 percent). And the state-level cost of educating their children flips to a gain in the next generation, who are among the strongest fiscal contributors in the country.
The "drain" framing is not supported; the state and local cost is real but front-loaded. Mixed
ITEP (2024); National Academies; NILC

We had to do all this. The border was wide open.

The kernel of truth, and it is a big one
This is the strongest point on the list, because it is largely correct. The 2021-2024 surge was real: border encounters topped 2 million a year and genuinely strained border towns, shelters, and the asylum system. Securing the border was a legitimate goal, and the enforcement push brought crossings to roughly a 55-year low. By that measure, it worked. Credit where it is due.
What the record shows
But "the border" and "legal immigration" are two different things, and the crackdown used the first to justify gutting the second. Bringing crossings down did not require cutting refugees to a record low, stripping legal status from more than a million people already here lawfully, charging $100,000 for a skilled-worker visa, or halving green cards. Those are choices about legal immigration, made under the banner of a border emergency that has already been brought under control. (On the "no one is vetted" version: legal entrants are heavily screened, and refugees are the single most-vetted category, an 18- to 24-month process.)
The border crisis was real and is now controlled. Using it to justify cutting legal immigration is the sleight of hand. Mixed
Pew Research (2026); CBP; USCIS refugee screening

They're all here illegally.

The kernel of truth
The unauthorized population is large in absolute terms, around 11 to 14 million, and it grew during the recent surge.
What the record shows
About three-quarters of the roughly 52 million foreign-born residents are here lawfully: about 46 percent are naturalized citizens, 23 percent are green-card holders, and 4 percent hold temporary visas. Roughly 27 percent are unauthorized. Most immigrants are, by a wide margin, here legally.
Not supported. Strong evidence
Pew Research (2025); Congressional Research Service

They don't assimilate or learn English.

The kernel of truth
Many first-generation adults never become fully fluent, and the US naturalization rate is lower than in peer countries like Canada, which does slow political integration.
What the record shows
Linguistic assimilation across generations is one of the most consistent findings in the field. Two-thirds of immigrants report speaking English well; by the second generation proficiency is high; by the third, English dominance is near-universal. The National Academies concluded immigrants come to resemble the native-born over time across language, income, and occupation.
Not supported across generations. Strong evidence
Migration Policy Institute; National Academies (2015); Cato

Birthright citizenship is a loophole for "anchor babies."

The kernel of truth
Unconditional citizenship by birth is not the global norm, and the precise meaning of the 14th Amendment's "subject to the jurisdiction thereof" is a genuine legal question now before the Supreme Court.
What the record shows
The 14th Amendment grants citizenship to those born on US soil, and the Supreme Court affirmed in 1898 (United States v. Wong Kim Ark) that this does not depend on the parents' status. The "anchor baby" pathway is also a weak mechanism: a citizen child cannot sponsor a parent until age 21, and an unauthorized parent typically still faces bars. Every court has so far blocked the executive order to end it.
Constitutionally grounded; the legal future is unsettled pending the 2026 ruling. Contested
Wong Kim Ark (1898); National Constitution Center; SCOTUSblog
Part 5 · What we are giving up

Immigrants are not a burden the country tolerates. They are the engine.

Step back from the line-by-line and look at what the policy is actually trading away. The American economy's defining strength, its ability to start and scale world-changing companies, runs disproportionately on immigrants.

46%
of 2025 Fortune 500 companies were founded by immigrants or their children (231 firms, $8.6T revenue, 15.4M jobs)
59%
of US billion-dollar startups have an immigrant founder, worth ~$5 trillion, averaging 833 jobs each (NFAP, 2026)
40%
of US science Nobel Prizes since 2000 went to immigrants

This is not a story of immigrants taking a fixed pie. Immigrants found businesses at roughly twice the rate of the native-born, and a peer-reviewed study of more than a million firms concluded they act more as job creators than job takers. They are 16 percent of US inventors but produce about 36 percent of US patents once collaboration is counted. The names are familiar: the founders of Google, Nvidia, Intel, Tesla, Stripe, and Moderna's parent were immigrants; Amazon and Apple trace to children of immigrants. (NFAP; NBER; American Immigration Council) Strong evidence

And the fiscal picture rewards patience. A college-educated immigrant contributes more than $800,000 net over a lifetime; the children of immigrants are among the strongest net contributors in the entire population. The state-and-local cost of the first generation is the down payment on that return. (National Academies, 2016) Strong evidence

The talent does not vanish. It moves to our rivals.

Here is the part the crackdown forgets: the scientists and founders it turns away do not stop existing. They go somewhere else, and that somewhere is increasingly a competitor that has openly declared it intends to surpass the United States.

The recruiting is already aggressive. When Canada opened a work permit for H-1B holders in 2023, its cap of 10,000 filled in about a day. The European Union launched a "Choose Europe for Science" initiative with 500 million euros to draw researchers, and France stood up programs specifically to host scientists leaving the US. The UK, Australia, and the UAE all run fast-track talent visas. China runs aggressive returnee programs and has a stated national goal to lead the world in artificial intelligence by 2030. (Government of Canada; European Commission; China State Council) Documented

And the outflow has started. New international student visas fell about 36 percent in the summer of 2025; applications from US-based scientists to Europe's main research funder roughly tripled; a Nature poll found about three-quarters of responding US scientists were considering leaving (a measure of intent, not yet of moves); and departures of China-born scientists had already jumped 75 percent after a 2018 crackdown. (Inside Higher Ed; Nature; PNAS) Mixed

Keep it in proportion: the United States still hosts around 60 percent of the world's top AI researchers, a lead built almost entirely on foreign-born talent. This is not a mass exodus. It is the slow draining of the inflow that produced the lead, and leads do not announce the moment they start to slip. (MacroPolo) Strong evidence

The irony is sharp. America became the world's scientific superpower in large part by welcoming the talent other countries cast out, the refugees who fled 1930s Europe and went on to build the Manhattan Project and decades of American advantage. A policy that turns that talent away, toward the very rivals racing to displace the United States, is not protecting American primacy. It is exporting it. Strong evidence

We have run this experiment before

When the United States slammed the door with the immigration quotas of the 1920s, the earnings of native-born workers in the most-affected areas did not rise; they fell, as employers turned to machines and other labor. Restriction did not deliver the gains its backers promised. (Abramitzky, Ager, Boustan et al., NBER) Strong evidence

The other side

The strongest case for restriction, fairly stated.

An honest page owes the other view its best argument.

A nation can choose who enters.
Border control and a functioning legal system are legitimate, and the 2021-2024 surge was a real strain on cities, shelters, and schools. Wanting order is not bigotry.
Some workers do lose out.
The research is clear that low-skill native workers and earlier immigrants, the closest substitutes, can see lower wages in the short run. Dismissing that cost is its own form of dishonesty.
Local budgets feel real strain.
The fiscal payoff is long-run and largely federal, while the cost of schooling and emergency care lands now, on states and towns. People pointing at present strain are not wrong about the timing.
Scale and speed matter.
Even immigration's defenders concede that absorbing very large inflows quickly, without enough housing or processing capacity, produces backlash and hardship. Pace is a legitimate concern.

Where the strain landed

This deserves more than a line, because for some places the cost was heavy and immediate. During the 2021-2024 surge, a handful of destination cities absorbed most of it. New York City spent about $1.9 billion in its first 15 months and projected up to $10 billion or more over three years, housing more than 200,000 arrivals under its right-to-shelter law and enrolling roughly 30,000 children in its schools. Chicago spent hundreds of millions and, before one winter, signed a $29 million contract for heated tents to move people off police-station floors. Denver, hit hard for its size, cut hours at motor-vehicle and recreation offices to pay for its response. And Massachusetts, the only state with a legal right to shelter, watched its family-shelter system swell toward a billion dollars a year, forcing the governor to cap it for the first time in four decades. Documented

Smaller communities felt it most sharply. Springfield, Ohio drew an estimated 12,000 to 20,000 Haitian immigrants, most of them legally present under Temporary Protected Status, into a city of under 60,000, straining clinics, classrooms, and rents, with one hospital reporting up to $50,000 a month in translation costs. A 2023 school-bus crash caused by an unlicensed immigrant driver killed an 11-year-old boy, a real death that politicians seized on until the child's father asked them to stop. The viral rumor that immigrants there were eating pets was false. The strain on local services, and that child's death, were not. Documented

And the costs are not shared evenly, which is the heart of the grievance. The National Academies finding is blunt: first-generation immigrants are, on average, a net cost at the state and local level, mainly because local governments pay to educate their children, even as they are a net gain for the federal budget and over the long run. The bills (shelter, emergency rooms, classrooms) arrive now and locally; the returns (taxes, the high-earning second generation) arrive later and nationally. And Washington reimbursed only a small fraction of what cities actually spent. A family in a school district absorbing hundreds of new students is not wrong to feel the cost is real and the help is somewhere else. Strong evidence

And it is a solvable problem

The good news, easy to lose in the noise, is that this is not intractable. The acute crisis eased for two distinct reasons, and it is worth keeping them separate. The first was simply that far fewer people were arriving: border crossings fell to a multi-decade low, so the buses largely stopped. That is the one place the enforcement-first approach plausibly helped, though it came partly by shutting off the legal right to seek asylum. With the inflow slowed, cities were no longer filling shelters faster than they could empty them. Documented

The second reason is the one that cleared the people already here, and it had nothing to do with deportation: letting them work. Once asylum seekers and Temporary Protected Status holders received work authorization, they left shelters for jobs. In New York, more than 70 percent of eligible adults in the city's care had received or applied for work permits, and most moved toward self-sufficiency; the city closed 62 emergency shelters between mid-2024 and mid-2025, more than $5 billion in projected savings. Massachusetts closed all of its hotel shelters and ended its emergency ahead of schedule; Chicago and Denver wound down. Federal reimbursement through FEMA's Shelter and Services Program helped at the margin, and a bipartisan 2024 Senate bill would have funded faster processing before it was killed for political reasons. Documented

These points are real, and they argue for a faster, fairer, better-managed system: more capacity, quicker processing, and the work authorization that measurably eased the strain. They do not argue for cutting refugees to a record low, stripping status from people already here legally, or making the legal path nearly impossible. The uncomfortable part is that the current policy is dismantling the very tools that were working, work authorization revoked, refugee resettlement suspended, federal reimbursement terminated and partly clawed back, answering a real problem at the border by gutting the legal immigration that was never the problem.

The words from the top

And the loudest answer has been to dehumanize.

Everything above describes a hard, complex problem and real people on every side: communities under genuine strain, workers with real grievances, and immigrants who are overwhelmingly here legally, mostly working, and committing less crime than the native-born. Problems like these respond to humane, practical tools, the work permits, funding, and processing that measurably eased the strain. The dominant national response has been the opposite. For a decade, the country's most powerful voice has described immigrants in sweeping and dehumanizing terms. Quoting him is not editorializing; it is the record.

It started at the beginning. Announcing his first campaign on June 16, 2015, he said of Mexican immigrants:

"They're bringing drugs. They're bringing crime. They're rapists. And some, I assume, are good people."

Campaign announcement, June 16, 2015 (Time, full transcript)

The signature recurring claim is that other countries deliberately empty their institutions into the United States. He has said it at the Republican National Convention in July 2024, on the debate stage in September 2024, and again in his 2025 inaugural address: that immigrants are pouring in

"from prisons and jails, from mental institutions and insane asylums."

Republican National Convention, July 18, 2024; repeated at the September 10, 2024 debate and the January 20, 2025 inaugural address. He has tied it to the fictional cannibal Hannibal Lecter.

Every major outlet that checked it (PolitiFact, the Associated Press, FactCheck.org, CNN, Reuters) found no evidence that any country empties its prisons or mental institutions into the US. One immigration analyst called it "a total fabrication." He said it No evidence it is true

The disease and "blood" framing runs alongside it. In 2023 he said immigrants were "poisoning the blood of our country," language the Anti-Defamation League and historians noted echoes a passage in Hitler's Mein Kampf. Trump said he "never knew that Hitler said it." At the September 2024 debate he claimed Haitian immigrants in Springfield, Ohio were stealing and eating pets:

"In Springfield, they're eating the dogs. The people that came in, they're eating the cats. They're eating the pets of the people that live there."

Presidential debate, September 10, 2024. Springfield officials and Ohio's Republican governor said there was no evidence; the moderator corrected it live; PolitiFact named it the 2024 "Lie of the Year."

And the dehumanizing register has only sharpened. He has said of migrants accused of murder that "it's in their genes," warning of "a lot of bad genes in our country," and has called migrants "animals" and "not people" (his defenders say those remarks referred to gang members and convicted murderers, not all immigrants, a distinction worth keeping). As president, the language continued into office: his 2025 inaugural address described "millions and millions of criminal aliens" and "dangerous criminals, many from prisons and mental institutions." Documented

The throughline never changes: immigrants are criminal, diseased, insane, less than human. None of it holds up. Immigrants, including the undocumented, commit crimes at lower rates than native-born Americans, and no country is emptying its asylums into the United States. But accuracy was never the point. This language does not house a single family or relieve a single school. It does the opposite: it makes the humane, workable fixes harder, swaps management for cruelty, and grants permission for a crackdown that tore down the very tools that were easing the strain. These are real people and genuinely hard problems. Dehumanizing them solves nothing, and it makes the problem worse.

Why this is short-sighted

The bill comes due slowly, and it is enormous.

The damage from cutting immigration does not show up at a podium. It shows up in the company that is never founded, the patent never filed, the elderly parent whose home aide never arrived, the farm that planted less, the town that kept shrinking, the Social Security gap that widened a little faster. None of it makes the news on the day it happens. All of it compounds.

The argument for the crackdown rests on claims the evidence does not support: that immigrants drive crime, that they take Americans' jobs, that they drain the system, that the border is still open. The strongest true points, that some workers lose out and that local budgets strain, argue for managing immigration better, not for shutting it down. What is actually being cut is the legal immigration and the citizenship that built American innovation and that will pay for an aging country.

A nation that spent a century becoming the place the world's talent wanted to come to is, in real time, telling that talent to stay home, or to take its companies and discoveries to a rival instead. The people who promised this would make Americans richer and safer are selling the opposite of what the record shows. The cost will be paid not by them, but by a smaller, older, slower country a generation from now, wondering where the dynamism went, and why the rivals it once outran suddenly have it.

And it does not reverse with the next election. A reputation built over a century, that the United States is where the ambitious of the world go to build a life, can be broken in a few years and take a generation to restore. The researcher who has already moved her lab to Munich, the founder who incorporated in Toronto, the family that rebuilt somewhere safer, do not come streaming back the day a policy changes. Once people decide that America is no longer a reliable place to stake a future, that trust is slow, and costly, to win again. Some of what is being lost now is not a tap that can be turned back on. It is a door that, once it closes, the world simply learns to walk around.

Sources

Where this comes from.

Selected primary documents and research. Figures are from the named sources and carry the caveats noted in the text. Where evidence is mixed or contested, the page says so.

Links go to the specific document or report behind each claim. Dates are given where a figure is time-sensitive.

Policy changes (2025-2026)

Economy, jobs, and fiscal effects

Deportation and enforcement

The rhetoric

Crime, myths, and the legal path

Entrepreneurship and innovation

Where the talent goes

Strain on local services

This page summarizes public reporting, government data, and peer-reviewed research as of mid-2026. Figures are estimates and ranges from the named sources; where sources conflict or evidence is mixed, the text says so. Genuine costs and tradeoffs are stated alongside the benefits. "Short-sighted" is a characterization, offered as an argument from the evidence laid out above. Corrections welcome.