Immigrants own roughly 20% of U.S. employer firms, not 20–25%, and make up 13.9% of the population, not 15%

The claim holds that immigrants own between 20% and 25% of all U.S. businesses while making up 15% of the population — framing immigrants as dramatically over-represented among business owners. The verdict is partially false: the underlying contrast is real, but both numbers in the claim are stretched beyond what the primary data support.

Start with the firmest number. The U.S. Census Bureau's Annual Business Survey, covering 2021 data released in 2023, places immigrant-owned employer firms at roughly 19–21% of all employer firms nationwide. The 2017 ABS, published in 2020, found the figure at 20.7%. Those results are consistent and credible — they come from the Census Bureau's own survey infrastructure, not advocacy estimates. So the lower bound of the claim, approximately 20%, is grounded in real data.

The 25% upper bound is where the claim breaks down. No Census employer-firm dataset supports it as an economy-wide figure. What does reach or exceed 25% is immigrant ownership in specific sectors: the New American Economy Research Fund's 2015 report found immigrants owned 28% of Main Street businesses — retail, accommodation, and food services — despite being about 13% of the population at the time. That sector-specific peak has apparently been laundered into a general economy-wide range, which is a classic cherry-pick. The strongest version of the claim would acknowledge the sector distinction; instead it presents the high-end outlier as the ceiling of a normal range.

The population figure has the same problem, just in the opposite direction. According to the U.S. Census Bureau's American Community Survey 2022 one-year estimates, the foreign-born share of the U.S. population is approximately 13.9%, not 15%. Rounding up by more than a full percentage point may seem minor, but it directly widens the gap between population share and ownership share, making the contrast sound more dramatic than the data warrant.

To be fair to the claim's spirit: the core finding — that immigrants own a disproportionately large share of U.S. businesses relative to their population share — is genuinely true and well-documented. SBA-commissioned research by economist Robert Fairlie in 2012 found immigrants owned about 18% of all U.S. businesses (employer and non-employer combined), with self-employment rates running higher among immigrants than among native-born workers. The National Foundation for American Policy documented in 2022 that immigrants or their children founded 44 of the top 91 Fortune 500 companies. The disproportionality is real. The specific figures cited in the claim are not.

The manipulation pattern here is number-blending: take a defensible average (20%), attach an unsupported ceiling borrowed from a specific sector (25%), overstate the population denominator (15% instead of 13.9%), and present the whole package as a single unified statistic. Each individual distortion is small enough to seem like rounding; together they make the contrast appear sharper and more dramatic than the evidence justifies. When you see a range like '20–25%' cited without a source distinguishing economy-wide from sector-specific data, that's your signal to ask which dataset the upper bound actually comes from.