Jobs Americans Will Do: Just About All of Them

If “immigrants do jobs that Americans won’t do”, there should be occupations in which the workers are overwhelmingly foreign-born. However, among hundreds of occupations identified by the Census Bureau, natives outnumber immigrants in all but a handful, and in none of them do illegal immigrants constitute a majority. The willingness of natives to work a broad range of jobs is even more apparent in low-immigration localities. “There are jobs Americans won’t do” is clearly not a strong argument for immigration.

Overall findings:

• Of the 525 civilian occupations identified in Census Bureau data, only five are majority immigrant (either legal or illegal) — with just one, “manicurists and pedicurists”, exceeding 60 percent.
• The five majority-immigrant occupations account for only 0.6 percent of the civilian U.S. workforce. Moreover, native-born Americans still comprise 40 percent of workers in these occupations.
• Many occupations often thought to be overwhelmingly foreign-born are in fact majority native-born:
  • Maids and housekeepers: 51 percent native
  • Construction laborers: 61 percent native
  • Home health aides: 61 percent native
  • Landscaping workers: 66 percent native
  • Janitors: 71 percent native
• About half of agricultural workers are immigrants, but all agricultural workers — natives and immigrants together — constitute less than 1 percent of the U.S. workforce.
• There are 65 occupations in which 25 percent or more of the workers are immigrants. However, these occupations are still held by about one in every nine native-born workers — 16 million natives in total.

Illegal immigrants:

• There are no occupations in which illegal immigrants in the data constitute more than one-third of workers.
• Illegal immigrants work mostly in construction, maintenance, food service, and agriculture. However, the majority of workers even in these occupations are either native-born or legal immigrants.

Low-immigration metropolitan areas:

• The cities and surrounding suburbs of Pittsburgh, St. Louis, Richmond, Nashville, and Columbus are examples of relatively low-immigration areas with relatively high per capita incomes. In these places, the willingness of natives to work stereotypically immigrant jobs is even more apparent:
  • Taxi drivers: 67 percent native
  • Painters: 73 percent native
  • Maids and housekeepers: 76 percent native
  • Dishwashers: 87 percent native
  • Janitors: 88 percent native
• Among the 431 occupations with sufficient data to analyze in these five low-immigration areas, just 13 are at least 25 percent immigrant.

Discussion

This report uses American Community Survey (ACS) data covering the years 2019 through 2023. Respondents are included if they are civilians who are at least 16 years old and “in the labor force” — that is, either working or looking for work. The immigrant share of an occupation is the fraction of workers in that occupation who were not U.S. citizens at birth. The margins of error around the estimated immigrant shares are calculated with replicate weights, which account for the ACS’s complex sampling design.

The claim from activists that “immigrants do jobs Americans won’t do” normally has two purposes. First, it implies that natives need not worry about the wage-depressing effects of immigration, since the two groups supposedly do not compete in the labor market. Second, it suggests that some jobs vital to our economy will simply not get done without immigrants.

This report demonstrates that the activists’ claim is wrong — natives are willing to do just about any of the occupations identified by the Census Bureau. Native-born Americans constitute a majority of the workers in 520 out of the 525 occupations, and they are a significant minority even in the remaining five. When employers assert that Americans will not do particular jobs, or when they raise the related specter of a “labor shortage”, what they really mean is that not enough Americans accept the low wages that employers prefer to pay. In the absence of immigration, the market wage rises and more natives become open to jobs they supposedly wouldn’t do.¹

Select MSAs. The readiness of natives to work “immigrant jobs” is further evident when considering metropolitan statistical areas (MSAs) with relatively low levels of immigration. An MSA is a region that includes both a core city and its surrounding suburbs. This report takes the top 50 MSAs by population, then selects the 25 MSAs from that list with the highest per capita incomes. Of these relatively high-income MSAs, the five with the lowest foreign-born population shares are Pittsburgh (4 percent foreign-born), St. Louis (5 percent), Richmond (9 percent), Nashville (9 percent), and Columbus (10 percent). As the spreadsheet indicates, even an occupation like taxi driver, which is majority immigrant nationwide, is two-thirds native in these places.

There is still an immigrant presence in all five of the select MSAs, of course, as the foreign-born shares listed above indicate. Given record levels of immigration to the U.S. as a whole, any large MSA — particularly a prosperous one — will tend to attract immigrants. Nevertheless, for various historical and economic reasons, the five select MSAs have relatively small immigrant populations despite their size and wealth. In terms of generating wealth, Pittsburgh is especially praiseworthy for weathering the decline in manufacturing and transitioning to a service-based, post-industrial economy. It has done so with a remarkably low level of immigration, and yet there are still taxi drivers there, along with construction workers, maids, landscapers, and so on.

One drawback of studying a small group of MSAs is that the sample size is much smaller than in the national analysis. Occupations with margins of error greater than 10 percentage points have been omitted from the MSA analysis, as have all of the illegal share estimates.

Illegal Immigrants. Illegal immigrants are present in survey data, but the Census Bureau does not explicitly identify them.² To determine which foreign-born respondents are most likely to be illegal, we start by eliminating those who are almost certainly legal: spouses of natural-born citizens, veterans, people who receive direct welfare payments (except Medicaid in certain cases), people who have federal jobs, Cubans, immigrants who arrived before 1980, immigrants who arrived at age 60 or older, people who earn more than $150,000 per year, people in certain occupations requiring a license or background check, and people likely to be on student visas.³ The remaining candidates are weighted to replicate known characteristics of the illegal population — e.g., population size, age, sex, region or country of origin, state of residence, and length of residence in the United States.⁴

Inferring the legal status of Census respondents is not an exact science. It is subject not only to conventional sampling error, which is reflected in the confidence intervals around the immigrant shares in the spreadsheet, but also to imprecision in the estimation method itself. Because of this non-sampling uncertainty, we cannot calculate confidence intervals for the illegal shares.

The Biden immigration surge has complicated matters even further. During the last administration, millions of immigrants received “parole” — a legal gray area in which the recipients have no visas and are “inadmissible”, but still have temporary permission to live and work in the U.S. Distinguishing these parolees from ordinary legal immigrants in the data is challenging. Furthermore, the sheer scale of the surge could lead to undercount of both legal and illegal recent arrivals. In fact, the Census Bureau recently doubled its initial estimate of net international migration for 2022 and 2023.⁵

The illegal shares in this report are based on our count of 10.3 million illegal immigrants in the 2019-2023 ACS data. The true number was almost certainly greater than what the Census Bureau captured in those surveys, and it is definitely greater now. Nevertheless, given that no occupation in the present data shows an illegal share greater than one-third, it is implausible that illegal immigrants could dominate any occupation today.

End Notes

¹ In fairness to employers, recruiting idle Americans back into the labor force is a challenge, but the challenge will be met only when immigration no longer provides a convenient alternative. For a longer discussion of this topic, see Amy Wax and Jason Richwine, “Low-Skill Immigration: A Case for Restriction”, American Affairs, Winter 2017.

² “About the Foreign-Born Population”, Census Bureau, December 16, 2021.

³ Medicaid covers childbirth and some post-partum care for illegal immigrant mothers. About a dozen states now allow illegal immigrant children to receive Medicaid. California, Illinois, and the District of Columbia also cover certain adult illegal immigrants. As for occupations requiring a license, most states will not license illegal immigrants to be K-12 teachers, lawyers, doctors, etc. However, California, Illinois, and Nevada are notable exceptions. During the 2019-2023 period that this study covers, many states were revising their policies on Medicaid access and occupational licensing. Rather than attempt to account for every individual exception, we have adjusted our illegal-identification method based only on the clearest and most long-standing policies that states have had in place.

⁴ The source of the known characteristics is “Estimates of Undocumented and Eligible-to-Naturalize Populations by State”, Center for Migration Studies.

⁵ Jason Richwine, “New Data Reveal the Scale of Biden’s Immigration Surge”, National Review, March 13, 2025.