Data-Driven Analysis

The Immigration Threat

Seven claims. Exposed to evidence. What survives contact with the data — and what doesn't.

Corey Petty February 2026 ~15 min read
Scroll to begin

There's a story roughly half the country believes. It goes: Democrats are importing voters through mass immigration. The immigrants are fiscal parasites, freeloading off welfare while contributing nothing. The Census inflates blue-state power by counting non-citizens. The system is rigged.

I wanted to test each of these claims against the best available data — peer-reviewed research, government statistical agencies, and analyses from think tanks across the political spectrum. What I found wasn't a simple debunking. Some of the underlying observations are real. But the framework they get plugged into collapses under the weight of the very evidence that's supposed to support it.

Every claim below includes its source, that source's institutional bias, and an evidence grade. If you want to dig deeper on any data point, the Source Analysis & Evidence Foundation document catalogs every source with full bias assessments and cross-reference notes.

"They're Importing Voters"

The central theory: immigrants vote overwhelmingly Democratic, so more immigrants means permanent Democratic power. From 1996 to 2020, there was a real trend supporting this — Obama won naturalized citizens by ~30 points in 2012.

Then came 2024.

2020 Biden +21
2024 Harris +4
Democratic
Republican
Naturalized citizen vote
Source: Pew validated voter study
0
point swing among naturalized citizens from 2020 to 2024. The "reliable Democratic constituency" split 51–47.
Pew Research — Validated Voter Study Grade A

Among Hispanic naturalized citizens specifically, 51% voted Trump. The entire theory's core prediction failed its empirical test.

The pipeline problem

Even if immigrants did vote reliably Democratic, the naturalization timeline makes "voter manufacturing" extraordinarily inefficient: median 7.5 years from green card to citizenship (10.9 for Mexican nationals). Only 49% of the foreign-born are citizens. Only 54–62% of those turn out. Mexican naturalization rate: 42% — the lowest among major groups. From unauthorized entry to first possible vote: 15–25 years.

Dig deeper: Why the 2024 shift matters structurally
Fraga, Velez & West (APSR, 2025) found Latinos are sorting by ideology, not ethnic identity — the opposite of what the "emerging Democratic majority" thesis predicted. Ruy Teixeira, who co-wrote The Emerging Democratic Majority (2002), now works at the conservative AEI and explicitly argues his own thesis was wrong. The California shift from red to blue was caused by Republican Prop 187 backlash (1994), not IRCA amnesty (1986) — the chronology contradicts the conspiracy. Source bias note: Pew is center/nonpartisan; APSR is the field's top peer-reviewed journal. Evidence grade: A.

"Non-Citizens Inflate Blue-State Representation"

Elon Musk popularized the claim that counting non-citizens in the Census shifts "20 seats" to blue states. The 14th Amendment says "whole number of persons" — citizen-only proposals were explicitly rejected during its drafting. But what about the actual number?

20
Seats claimed to shift to blue states
Musk, 2024 — no methodology cited
vs
2
Seats that would actually shift
PNAS Nexus, 2025 — peer-reviewed

Off by a factor of ten. And the partisan direction is ambiguous: Texas and Florida — both GOP-trending — would lose seats under a citizens-only count. The net partisan effect in any election since 1980: no more than 3 Electoral College votes shifted between parties.

Dig deeper: The constitutional and legal history
The Trump administration attempted both a citizenship question on the 2020 Census (blocked by SCOTUS in Dept. of Commerce v. New York, 2019) and an executive memorandum to exclude undocumented from apportionment (data never delivered). Both failed because the legal framework is clear. At the 2007 peak unauthorized population (12.2M), the maximum apportionment shift would have been 5 seats. Source: PNAS Nexus (academic, peer-reviewed); Census Bureau. Evidence grade: A.

Follow the Money

The "free-rider" framing depends on the assumption that undocumented immigrants don't pay taxes. Here's what the IRS, SSA, and independent tax analysts actually found:

$0
Total taxes paid by undocumented immigrants in 2022.
Institute on Taxation and Economic Policy, 2024 Grade B+
$25.7B
Social Security payroll taxes — for benefits they can never receive
ITEP • Grade B+
$6.4B
Medicare payroll taxes — for benefits they can never receive
ITEP • Grade B+
$2.15T
Cumulative wages in SSA's Earnings Suspense File — unmatched W-2s, largely from undocumented workers
SSA • Grade A
8.9%
Avg effective state/local tax rate paid by undocumented immigrants — higher than the top 1% (7.2%)
ITEP • Grade B+

The programs they're barred from

Undocumented immigrants are categorically ineligible for: Social Security, Medicare, Medicaid (except emergency), SNAP, TANF, SSI, federal housing assistance, ACA marketplace plans, federal student aid, unemployment insurance, and the EITC. Emergency Medicaid — which reimburses hospitals, not individuals — totaled $3.8B in FY2023: 0.4% of Medicaid spending.

They're paying into a system they cannot draw from. The Social Security Administration itself concluded: "earnings by unauthorized immigrants result in a net positive effect on Social Security financial status generally."

Dig deeper: ITEP methodology and source bias
ITEP is left-leaning on tax policy but their tax incidence modeling is used across the political spectrum. The $96.7B figure relies on modeling (undocumented status isn't recorded on tax returns), using assumptions about compliance rates and income distribution. It's a central estimate with uncertainty ranges. IRS data: 26M ITINs issued since 1996; ~3.8M ITIN returns filed in 2022 reporting $14.4B in taxable income. The payroll tax contribution figure is corroborated by SSA's own Earnings Suspense File, which is direct administrative data (Grade A). Evidence grade for overall tax contribution: B+.

"Immigrants Use More Welfare Than Americans"

This is where institutional bias in the sources becomes most visible. The answer depends entirely on who you ask and how they count.

Per-capita welfare consumption: Immigrants vs. Native-born

All programs combined53% less
100%
47%
Social Security39% less
100%
61%
SNAP (food stamps)35% less
100%
65%
SSI42% less
100%
58%
Native-born = 100% baseline · Source: CATO Institute / 2023 SIPP data Grade A

Why the sources disagree

CATO (libertarian, Koch-funded) uses per-capita, income-controlled comparisons — standard methodology. CIS (restrictionist, Tanton-founded) uses household-level counts that attribute U.S.-citizen children's benefits to immigrant parents, without income controls. Same data, wildly different conclusions. When CATO, the CBO, USDA, and CIS all look at the same data and CIS is the outlier, the methodological burden is on CIS.

This held at every income level. Immigrants in poverty used less welfare than native-born Americans in poverty. The "free-rider" charge is not just unsupported — it's inverted.

Dig deeper: CIS methodology problems
CIS was founded by John Tanton, who had documented connections to white nationalist organizations (SPLC has designated CIS a hate group, which CIS disputes). Their welfare studies count benefits for U.S.-citizen children in immigrant households as "immigrant" welfare use, don't control for income level, and compare low-income immigrant households against the overall native population. It's like comparing pickup truck fuel economy to a fleet that includes hybrids. USDA data independently confirms the CATO finding: noncitizens = 4.8% of SNAP spending while being 6.5% of the population.

The Cost That's Actually Real

Here's where I shift gears. The state and local fiscal burden is the one claim in this entire constellation with genuine, well-documented empirical support. Dismissing it is dishonest.

The problem is architectural. The federal government collects immigration's fiscal benefits. States and localities absorb its costs. That mismatch is real.

$3.75B
NYC asylum seeker costs, FY2024 alone. FEMA reimbursed just $245M.
NYC Comptroller • Grade A
$2.9B
Texas border security spending, FY2023 — up from $0.4B pre-surge
Texas state reports • Grade B
$9.2B
Total net cost on state/local governments nationwide, 2023 — 0.3% of spending
CBO • Grade A
$1.2T
Federal revenue increase from the same immigration surge, 2024–2034
CBO • Grade A

The reframe that matters

This is a federalism problem, not an immigration problem. Washington collects $1.2 trillion in revenue from the immigration surge while states absorb $9.2 billion in costs. The federal government already transfers ~$1 trillion annually to states. Adjusting those transfers — through enhanced SCAAP funding, expanded emergency medical reimbursement, or targeted education grants — addresses the real grievance without blaming immigrants for a structural mismatch they didn't create.

Federal vs. state/local: Who captures the fiscal impact?
+$1.2T Federal Revenue
-$0.3T Mandatory
-$9.2B State/Local
Federal revenue gained Federal mandatory spending State/local net cost
CBO projections, 2024–2034 · Grade A · Not to precise scale — illustrates ratio
Dig deeper: The Texas Comptroller study they don't update
The 2006 Texas Comptroller study — conducted under Republican Comptroller Carole Keeton Strayhorn — found undocumented immigrants generated a net positive of $420M for the state government even as local governments absorbed $1.44B in costs. The Perryman Group estimated removing undocumented residents would cost Texas $69.3B in economic activity and 403,000 jobs. Texas has not updated this study since 2006 despite repeated calls to do so — a notable political choice. Florida data: undocumented immigrants = 0.82% of hospital visits despite comprising ~4% of the state's population.

"They Don't Work"

This one is brief because the data is unambiguous to the point where dwelling on it feels like dignifying a claim that doesn't deserve the attention.

Labor force participation rate — BLS, 2024

Men — Foreign-born77.3%
77.3%
Men — Native-born65.9%
65.9%
Overall — Foreign-born66.5%
66.5%
Overall — Native-born61.8%
61.8%
Bureau of Labor Statistics, "Foreign-Born Workers" 2024 Grade A
Immigrants are roughly twice as likely to start businesses as native-born Americans
Kauffman Foundation • Grade B+
46%
of Fortune 500 companies were founded by immigrants or their children
AIC, 2025 • Grade B
70%
of U.S. farmworkers are foreign-born. When Alabama passed enforcement laws, crops rotted.
USDA/BLS • Grade A
38.7%
of U.S. software developers are foreign-born
AIC/Census • Grade B+

Evidence grade for the "non-productive" claim: F — Misleading. Not debatable. Not complicated. Every available labor market dataset contradicts it.

275 Years of the Same Argument

Every claim in this analysis has been made before. Not vaguely — specifically. Directed at different groups, repeated on a roughly 40-year cycle. Each time, the observations were real. Each time, the analytical framework was wrong.

1751
Benjamin Franklin on Germans
"Why should Pennsylvania, founded by the English, become a Colony of Aliens, who will shortly be so numerous as to Germanize us?" He called Germans "swarthy."
↗ 45 million German-Americans today. Largest ancestral group in the U.S.
1850s
Know-Nothings on the Irish
Irish Catholics would deliver America to the Pope. Proposed 21-year naturalization wait. Irish comprised 70% of NYC charity recipients.
↗ 150,000 Irish fought for the Union. Built America's infrastructure. Irish Catholics now lean Republican.
1882
Chinese Exclusion Act
"A race of cheap working slaves." Denis Kearney's Workingmen's Party drove the first race-based immigration ban in U.S. history.
↗ NBER (2024): The Act reduced manufacturing output by 62% in affected counties. It didn't raise native wages.
1900s–1920s
Southern & Eastern Europeans
Stanford president Jordan: "not one in a thousand from Naples or Sicily that is not a burden." Madison Grant: they'd create a "corrupted caricature." Led to the 1924 Johnson-Reed Act.
↗ Italian and Jewish Americans became Supreme Court justices, governors, Fortune 500 CEOs.
2000s–Present
Latin American & Other Immigration
Buchanan: "barbarian invasions." Carlson: "more obedient voters from the Third World." "They won't assimilate. They'll drain the system. They'll change America forever."
↗ Children of immigrants show higher upward mobility at every income level — a pattern consistent across 130 years of data.
+5–6
Percentile points of additional upward mobility for children of immigrants vs. children of native-born at the same income rank. Consistent across 130 years of data and 4 million father-son pairs.
Abramitzky, Boustan & Jácome — NBER/AER Grade A

Real Problems, Wrong Framework

The conservative immigration narrative contains verifiable factual kernels embedded within an analytical framework the data does not support.

What the data supports
• State/local fiscal costs are real and concentrated
• The 1965 Act's effects exceeded its promises
• Biden-era border encounters were unprecedented
• Some state benefit expansions are debatable policy
What the data doesn't
• Deliberate voter-importation strategy (D grade)
• "Free-rider" framing — it's inverted (F grade)
• "Non-productive" claim — every dataset contradicts it (F grade)
• "20 seats" apportionment claim — off by 10× (F grade)
• "This time is different" — contradicted by 275 years (D grade)

The conspiracy theory is seductive because it provides a single, emotionally satisfying explanation for a set of real anxieties. But solving the real problems — the federalism mismatch, the border infrastructure, the ESL teacher shortage — requires an accurate diagnosis. And the dominant narrative is not one.

The data is there for anyone who wants to look at it.