G: Yes. Many undocumented immigrants pay taxes, and in some cases they pay substantial federal, state, and local taxes. The answer depends on how they work and how they're paid.

Here's the breakdown:

1. Payroll taxes (Social Security and Medicare)

If an undocumented immigrant works "on the books," taxes are often withheld from each paycheck just like for any other employee.

This can happen because they may:

• Use an Individual Taxpayer Identification Number (ITIN) for tax filing (though ITINs generally are not used for employment eligibility).
• Work using a false or borrowed Social Security number.
• Have an invalid or expired Social Security number that nevertheless results in payroll withholding.

In many of these cases:

• Federal income tax is withheld.
• Social Security tax is withheld.
• Medicare tax is withheld.

If the Social Security number doesn't match a valid worker, those Social Security contributions generally cannot be claimed as retirement benefits by that worker. Billions of dollars have accumulated in what's called the Social Security Administration's "earnings suspense file" because wages couldn't be matched to a valid account.

2. Federal income taxes

Many undocumented immigrants voluntarily file federal income tax returns.

The IRS issues ITINs specifically so people who are not eligible for Social Security numbers can comply with tax laws.

Reasons include:

• Following the law.
• Building a record of tax compliance, which can sometimes be relevant in future immigration proceedings.
• Claiming refunds of over-withheld taxes when eligible.
• Paying taxes owed to avoid penalties.

3. State income taxes

In states with an income tax, many undocumented workers also file state tax returns using an ITIN.

4. Self-employment taxes

Undocumented immigrants who are self-employed may file taxes using an ITIN and pay:

• Federal income tax
• Self-employment tax (which covers the equivalent of Social Security and Medicare contributions)

How much do they pay overall?

Multiple independent analyses have estimated that undocumented immigrants collectively contribute tens of billions of dollars annually in federal, state, and local taxes. A significant portion comes from payroll taxes and income taxes, not just sales or property taxes (including property taxes paid indirectly through rent).

A common misconception

People sometimes say either:

• "Undocumented immigrants don't pay taxes." — This is false.
• "All undocumented immigrants pay taxes." — This is also false.

The reality is mixed:

• Those working entirely in the cash economy without reporting income may pay little or no income or payroll tax.
• Those working through employers with withholding, or who voluntarily file tax returns, often pay income and payroll taxes similarly to other workers.

So the accurate statement is that many undocumented immigrants do pay federal and state income taxes and payroll taxes, while others do not, depending on how they are employed and whether they report their income.

G: That's a good question, and the answer is broader than undocumented immigration. The IRS didn't create ITINs as a workaround for undocumented immigrants—they created them because the U.S. tax system and immigration system have different jobs.

The IRS's job is to collect taxes. Immigration agencies' job is to determine who may legally live and work in the United States. Congress has generally kept those functions separate.

An Individual Taxpayer Identification Number (ITIN) exists because there are many people who have a legitimate legal obligation to file U.S. taxes but are not eligible for a Social Security number. Undocumented immigrants are one category, but far from the only one.

Examples include:

• Foreign investors who own U.S. assets and owe U.S. taxes.
• Nonresident aliens with U.S.-source income.
• International students or researchers in situations where they need to file taxes but aren't eligible for an SSN.
• Dependents or spouses of certain visa holders.
• Other foreign nationals who have U.S. tax reporting obligations.

So the principle is:

If U.S. tax law says you owe taxes, the IRS needs a way to identify you, even if you can't get a Social Security number.

That raises the obvious question: Why doesn't the IRS simply refuse taxes from undocumented immigrants?

There are several reasons:

• Tax liability doesn't depend on immigration status. If someone earns taxable income in the U.S., the tax code generally says they owe tax regardless of whether they are lawfully present.
• The government wants compliance. From the IRS's perspective, collecting taxes from someone is preferable to driving that income completely underground.
• It separates enforcement functions. The IRS is intended to administer tax law, not serve as an immigration enforcement agency. That separation encourages compliance by people who otherwise might never file.

This separation is actually quite old. It reflects the idea that a person can simultaneously:

• violate one area of law (immigration), and
• still have obligations under another (tax).

That's not unique to immigration. For example, income from illegal activities is still taxable under U.S. law. If someone illegally runs an underground gambling operation or sells illegal drugs, the income is still subject to federal income tax. The government doesn't treat paying those taxes as legitimizing the underlying conduct.

The same logic applies here: paying taxes does not legalize someone's immigration status, nor does possessing an ITIN grant any immigration benefit or work authorization. An ITIN is simply a taxpayer identification number used for tax administration.

So I wouldn't characterize ITINs as "we don't like illegal immigrants, but we'll let them pay taxes anyway." A more accurate characterization is:

"Anyone who has a U.S. tax obligation needs an identifier, regardless of why they have that obligation."

The IRS is essentially saying, "Whether you're allowed to be here is for immigration authorities to determine. If you owe taxes under the tax code, we're going to collect them."

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G: I think it's fair to separate two questions here:

1. How widely known is it that many undocumented immigrants pay taxes using ITINs?
2. Could access to IRS data change the relationship between tax compliance and immigration enforcement?

For the first question, my impression is that the general public is not very aware of it. Surveys over the years have consistently found that a substantial fraction of Americans believe undocumented immigrants pay little or no tax beyond sales taxes, despite decades of reporting that many do file income tax returns and pay payroll taxes. I don't know of a high-quality poll asking specifically, "Do you know what an ITIN is?", but I'd be surprised if recognition were anything but quite low outside tax professionals, immigration attorneys, accountants, and policy specialists.

For the second question, you're describing a concern that has been at the center of a real legal and political controversy.

Historically, one reason the IRS has maintained strict taxpayer confidentiality under Internal Revenue Code § 6103 was precisely to encourage voluntary compliance. The longstanding understanding was that, with limited statutory exceptions, tax return information would not be used as a general immigration enforcement tool. citeturn0search3turn0search0

Beginning in 2025, the Treasury Department and DHS entered into an agreement allowing certain taxpayer information to be shared with ICE under specified legal circumstances. Supporters argued it fit within statutory exceptions for certain criminal investigations; critics argued it undermined decades of taxpayer expectations and would discourage tax compliance. The agreement was challenged in court, and subsequent litigation found significant problems with aspects of the disclosures, with courts restricting or blocking portions of the data-sharing program. citeturn0search2turn0news19turn0news18

One of the principal arguments made by critics—including former IRS officials, tax experts, and some members of Congress—was almost exactly the mechanism you described:

If people come to believe that filing taxes or obtaining an ITIN increases the likelihood of being identified for immigration enforcement, some will stop filing taxes altogether.

That concern isn't merely speculative; it has been explicitly raised in litigation, congressional correspondence, and analyses of the policy. citeturn0search12turn0search13turn0news25

Whether that concern ultimately proves correct is an empirical question that economists and tax administrators will continue to study. But the concern itself is well established and was one of the major objections to expanding information sharing.

One subtle point I'd add is that an ITIN by itself is not a list of "people here illegally." Many perfectly lawful non-citizens also have ITINs because they're required to file U.S. taxes but aren't eligible for Social Security numbers. So an ITIN database isn't synonymous with an undocumented immigrant database. It could, however, become more informative if combined with other government records or investigative information—which is why access to tax data has long been treated as highly sensitive.

G: I think you've identified a policy tension that has existed for a long time, and it's worth separating the empirical part from the inference about motives.

The empirical part is straightforward:

• The tax system works best when people believe that complying with it won't unnecessarily expose them to unrelated government actions.
• That's one of the reasons Congress enacted strong taxpayer confidentiality rules in the first place. The idea wasn't merely privacy for privacy's sake—it was to encourage honest reporting and maximize tax collection.
• Many tax scholars, former IRS officials, and immigration experts argued that expanding the use of IRS data for immigration enforcement could undermine that incentive structure.

That's a coherent institutional argument regardless of one's views on immigration.

Where I'd be a little more cautious is the last step—that because ITINs disproportionately include undocumented immigrants, there is "really" no other reason to seek access to the data. There are legitimate law-enforcement reasons one might seek tax information in individual cases—for example, investigating tax fraud, identity theft, or other crimes where Congress has authorized disclosure. The legal debate has largely been about whether the newer uses fit within those statutory exceptions, not whether there could ever be any legitimate use.

That said, your broader point about incentives is one many critics made. If the government says:

"We want people to comply with the tax laws."

and later says:

"The information you provided to comply with the tax laws may now be used for immigration enforcement,"

then it changes the incentive landscape. Even if the disclosures are lawful under some interpretation of the statute, people may rationally conclude that compliance carries additional risks it previously did not.

That's why the issue generated such concern among tax administrators. They weren't necessarily arguing that immigration laws shouldn't be enforced; many were arguing that the IRS is a uniquely effective institution precisely because taxpayers generally believe its information will remain confidential except in narrowly defined circumstances. Once that expectation changes, rebuilding it can be difficult.

Your analogy to collecting a list of Arabic-speaking people gets at a concept used in policy analysis: proxy variables. A dataset doesn't have to perfectly identify a target population to become useful for identifying many members of that population when combined with other information. That's a real phenomenon in data science and privacy research. The important distinction, though, is between recognizing that a dataset functions as a proxy and concluding why someone sought access to it. The former can often be demonstrated empirically; the latter usually requires additional evidence about intent.

So I think there's a strong institutional critique available without having to infer motive:

A tax identifier created to maximize voluntary compliance may lose some of its effectiveness if people come to believe it has become a pathway for unrelated enforcement. Whether or not that policy is lawful or desirable, it creates a tension between the government's goals of collecting taxes and enforcing immigration law, and critics argue that the long-term cost to tax compliance could outweigh the enforcement benefit.

That argument stands on its own, regardless of what one believes about the administration's broader immigration policies.

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G: I think you can make a rhetorically strong comparison, but I'd avoid overstating it by comparing an undocumented worker's lifetime taxes to a billionaire's one-year taxes. A fairer apples-to-apples comparison is annual taxes versus annual taxes.

Here's a defensible way to do it.

Step 1: Estimate the average undocumented immigrant's annual tax contribution

The Institute on Taxation and Economic Policy (ITEP) estimated that undocumented immigrants paid $96.7 billion in federal, state, and local taxes in 2022—about $8,889 per undocumented immigrant on average. That figure includes income taxes, payroll taxes (employee and employer shares in the economic incidence), sales taxes, property taxes, etc. citeturn0search0

If you want to exclude sales tax and focus on taxes people commonly think of as "income/payroll taxes," the report estimates:

• Federal taxes: about $59.4 billion
• Of that, about $33.9 billion were Social Security, Medicare, and unemployment payroll taxes. citeturn0search0

So a reasonable shorthand is:

"The average undocumented immigrant contributes on the order of $9,000 per year in taxes overall."

Step 2: Compare against well-documented billionaires

Here, there are several names you can use without speculation because their federal income tax for certain years has been widely reported from the leaked IRS data analyzed by ProPublica:

• entity["known_celebrity","Jeff Bezos","Amazon founder"]

  • 2007: $0 federal income tax.
  • 2011: $0 federal income tax. citeturn0search2

• entity["known_celebrity","Elon Musk","Tesla and SpaceX executive"]

  • 2018: $0 federal income tax. citeturn0search2

• entity["known_celebrity","Michael Bloomberg","Bloomberg founder"]

  • Paid no federal income tax in at least one reported year. citeturn0search2

• entity["known_celebrity","Carl Icahn","investor"]

  • Paid no federal income tax in 2016 and 2017. citeturn0search2

What that lets you say

Without exaggeration, you can accurately say something like:

"According to ITEP's estimates, the average undocumented immigrant contributes roughly $9,000 per year in taxes. Meanwhile, IRS records reported by ProPublica indicate that several American billionaires legally paid $0 in federal income tax in particular years. So in those years, the average undocumented immigrant appears to have contributed more in taxes overall than those billionaires paid in federal income tax."

Notice the wording. It compares:

• overall taxes for the immigrant,
• to federal income tax for the billionaire.

Those are not the same tax base, and it's important to say so.

A more technically precise version

If someone challenged the comparison, I'd respond this way:

"I'm not claiming the billionaire paid zero taxes of every kind. They likely paid property taxes, sales taxes, and taxes through corporations they owned. I'm making a narrower point: there are documented years in which some of America's richest individuals paid zero federal income tax, while the average undocumented worker was still contributing thousands of dollars in taxes through payroll, income, sales, and property taxes."

That version is, in my view, both rhetorically effective and technically defensible.

One other point I'd make, because it often gets lost: the ProPublica reporting is about legal tax avoidance under the existing tax code, not tax evasion. So if you're using this comparison in discussion, it's strongest if you frame it as highlighting how the tax system can produce counterintuitive outcomes, rather than accusing those individuals of breaking the law. That keeps the comparison grounded in the evidence rather than inviting an unnecessary argument about legality.

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G: I think there's a cleaner and more defensible comparison than trying to estimate a hypothetical "middle-class undocumented worker."

The man you're describing—a person who has been here for decades, owns a home, has children, and has consistently worked—is almost certainly not representative of the average undocumented immigrant. In fact, he's probably above the average in both income and taxes paid simply because he's established enough to own property and support a family.

Rather than guessing his tax bill, I'd anchor the comparison to a conservative baseline and then point out that someone like him likely exceeds it.

For example:

"According to the Institute on Taxation and Economic Policy, the average undocumented immigrant pays roughly $8,900 per year in combined federal, state, and local taxes. A long-term undocumented homeowner supporting a family would likely pay substantially more than that average through payroll taxes, income taxes, property taxes (directly or through a mortgage), and sales taxes. Meanwhile, IRS data reported by ProPublica show that some of the wealthiest Americans legally paid zero federal income tax in certain years."

Then you can name names.

Some of the strongest documented examples are:

• Jeff Bezos — Paid $0 federal income tax in 2007 and again in 2011. citeturn0search0
• Elon Musk — Paid $0 federal income tax in 2018. citeturn0search0
• Michael Bloomberg — Paid $0 federal income tax in at least one year covered by the leaked IRS data. citeturn0search0
• Carl Icahn — Paid $0 federal income tax in 2016 and 2017. citeturn0search0
• George Soros — Paid $0 federal income tax for three consecutive years. citeturn0search0

Those aren't estimates; they're the specific examples ProPublica highlighted because they're so striking.

As for the investor you were thinking of, that was Warren Buffett, not William F. Buckley Jr. Buffett famously remarked that his secretary paid a higher tax rate than he did, which became known as the "Buffett Rule." Ironically, Buffett himself wasn't one of the "zero federal income tax" examples in the leaked IRS data—he paid federal income tax, but his wealth increased so dramatically that ProPublica calculated his "true tax rate" relative to wealth growth at about 0.1% over 2014–2018. citeturn0search4turn0search0

One suggestion on rhetoric, though. I think the most compelling comparison isn't actually:

"This undocumented immigrant paid more taxes than Elon Musk."

That invites someone to say, "Well, Musk paid payroll taxes, sales taxes, corporate taxes..." and the discussion veers off into tax accounting.

Instead, I think the sharper comparison is:

"Here's someone who spent decades working, raising a family, buying a home, and paying taxes while lacking legal status. In the same country, there were years in which some of the richest people on Earth legally owed zero federal income tax. Whatever you think about immigration, it's hard to argue that the tax system consistently rewards contribution in proportion to means."

That comparison doesn't require inflating anyone's tax bill or diminishing anyone else's. It simply juxtaposes two well-documented facts: one about long-term taxpaying undocumented residents, and one about how the tax code has allowed certain ultrawealthy individuals to owe no federal income tax in particular years. It's harder to dismiss because each piece stands on its own.

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G: Yes. In fact, I think you can make a stronger point if you avoid the "zero tax" examples entirely. There are several people whose taxes were not zero but were so low relative to either their wealth or income that they become striking comparisons.

Here are the best documented examples.

The one I'd actually use rhetorically, though, is Buffett, because he himself has made the fairness argument.

He wasn't saying he broke the law or that he cheated. He was saying, in effect:

"The law allows me to pay less, and I don't think that's a good outcome."

That makes him unusually difficult to dismiss as simply a partisan example.

One comparison that I think is especially powerful

Suppose you use the ITEP estimate of roughly $9,000 per year in taxes paid by the average undocumented taxpayer.

Now imagine the Houston resident you mentioned—a homeowner with children who had worked for 35 years.

Even without assuming he paid more than average (which he probably did if he owned a home), that's approximately:

• $9,000 × 35 years ≈ $315,000 in total tax contributions.

That's a back-of-the-envelope estimate, not a historical reconstruction, but it's transparent and easy to explain.

Then compare that to Buffett's famous statement that he paid a lower tax rate than his secretary, or to the documented years in which Bezos, Icahn, Soros, and others legally owed no federal income tax. You're no longer comparing "illegal immigrant versus billionaire." You're comparing decades of steady contribution by someone participating in the tax system with how the tax code can produce extremely low liabilities for people at the very top.

I actually think that's the more compelling story. It shifts the discussion away from immigration and toward contribution.

One final suggestion: if you're making this argument publicly, I'd avoid saying someone "paid less taxes than the immigrant" unless the numbers actually support that exact claim. Instead, I'd say something like:

"Here's a man who spent 35 years working, raising a family, paying taxes, and buying a home. Meanwhile, our tax system has produced years in which some of the richest people in the world legally paid no federal income tax or extraordinarily low taxes relative to their wealth. Whatever your views on immigration, that's a striking contrast."

That version is harder to knock down because every piece of it is independently supported.