Every immigrant-crime segment has the same fingerprint, and once you learn it you cannot unsee it. The host puts a scary-looking statistic on screen, sources it to a federal dataset, and lets you assume it describes crime in America. It does not. The trick is in the sourcing line nobody reads aloud, and it is the same trick whether the segment airs on a cable network or a podcast feed.
The dataset that isn't about crime
Tucker Carlson built years of monologues on the argument that noncitizens commit crime at alarming rates, pointing to data from the US Sentencing Commission. [3] Here is what that dataset actually is: a record of people sentenced in federal court. Federal court handles a narrow, unrepresentative sliver of all crime in the country, and a large share of its caseload is immigration offenses themselves, acts that by definition only a noncitizen can commit, like illegal entry. [3] Count those, attribute them to "noncitizen crime," and you have manufactured a wave out of the act of crossing a border. Even by that federal measure, once immigration offenses are stripped out, noncitizens accounted for roughly 14 percent of cases. [4] The number was never describing your neighborhood.
What the comprehensive data shows
Crime in America is overwhelmingly a state and local matter, so the honest question is what happens when you count all of it. Researchers have, repeatedly, and the answer does not move. The most comprehensive study to date, a National Bureau of Economic Research paper tracking incarceration across 150 years, finds that immigrants have been less likely to be incarcerated than the US-born since 1870, and that in recent decades they are roughly 60 percent less likely. [1] A separate study in the Proceedings of the National Academy of Sciences examined Texas arrest records and found undocumented immigrants were arrested for violent and drug crimes at less than half the rate of native-born citizens. [2] Different decades, different methods, same direction.
Data
| US-born (index) | 100 |
|---|---|
| Immigrants | 40 |
Why the trick keeps working
The federal-data move survives because it is technically true and contextually false, which is the most durable kind of lie. [1][2] The host is not fabricating the sentencing numbers; they are real. The deception is letting a narrow federal caseload, padded with immigration offenses, stand in for crime as a whole, and trusting the audience not to check the footnote. A correction that says "immigrants commit less crime" is fighting a feeling with a fact, and the feeling has a head start. The only counter is to name the specific sleight of hand every time it runs: watch which dataset they reach for, and ask why it is always the one that cannot answer the question they are pretending to ask.
The receipts are not hard to pull. One hundred fifty years of data point one way, the Texas records point the same way, and the scary chart points at a federal courtroom that handles almost none of the crime it is invoked to explain. [1][2][3] The immigrant-crime number has a tell. Now you know where to look.
