[TrumpDebunked]

The Internet has been flooded recently with reports saying non-citizens commit 42.7% of federal crimes despite being only 7% of the US population:

https://www.google.com/search?&q=non-ci ... OR+42.7%29

This is an invalid comparison based on questionable metrics, beginning with the "7%" of US population being "non-citizens". According to many reports, it is based on "the U.S. Census Bureau’s American Community Survey" with that reference linked not to a Census Bureau web page but rather to this:

https://www.kff.org/other/state-indicat ... ip-status/

There it states that the data source is "Kaiser Family Foundation *estimates* based on the Census Bureau's American Community Survey, 2008-2017". And estimates they must be, because the US Census Bureau does NOT (yet) collect data on citizenship status:

https://www.scotusblog.com/case-files/c ... -new-york/

The closest analog might be data on "foreign born", which they define to include "naturalized U.S. citizens, lawful permanent residents (immigrants), temporary migrants (such as foreign students), humanitarian migrants (such as refugees and asylees), and unauthorized migrants" - such data NOT including legal status:

https://www.census.gov/topics/populatio ... about.html

So much for the "7%". Now let's look at the "42.7%":

That statistic is drawn from Table 9 (page 52) of the 2018 Annual Report and Sourcebook of Federal Sentencing Statistics:

https://www.ussc.gov/sites/default/file ... cebook.pdf

Of the 69,425 federal crime cases analyzed in this report, the citizenship status of 522 "offenders" was not known. Of the remaining 68,903 cases, 29,445 involved non-citizen offenders, and dividing 29,445 by 68,903 yields 42.7%. So yes, 42.7% of these federal crime cases had non-citizen offenders, but that statistic cannot be used in an apples-to-apples comparison with the 7% estimate (or any other overall population metric) because, as is disclosed on page 202, "[a] single offender may account for more than one case if the offender was involved in more than one sentencing event during the fiscal year". In other words, although a single individual is counted only once in the "population", a single individual may be counted multiple times in this report's "offender" statistics - hence apples and oranges.

And how serious is the distortion? For "Immigration" offenses, quite serious. Consider, for example that "...in more than 21,000 cases last year, [Border Patrol] agents apprehended border crossers who already had been caught six or more times":

https://abcnews.go.com/ABC_Univision/Po ... d=19244006

That particular distortion explains not only why 42.7% of "offenders" were non-citizens but also why 54.3% were Hispanic. Furthermore, broad statistical assimilations like these suggest that a hungry Mexican crossing the Rio Grande in search of work is an offense as serious as a cold-blooded murder, which is far from rational.

So much for the 7% - and the 42.7%:

"Non-Citizens Commit 42.7% Of Federal Crimes, Despite Being Only 7% Of US Population" - DEBUNKED.