In exchange for a deal on young immigrant Dreamers, the White House is demanding that Congress reduce legal immigration by ending the diversity visa lottery and almost all family sponsorship categories. On Fox News last week, Attorney General Jeff Sessions made the case for these changes by stating that he wants legal immigrants to “have the education and skill level to prosper in America.” He asked rhetorically, “What good does it do to bring in somebody who is illiterate in their own country, has no skills, and is going to struggle?”

But this generalization about diversity and family-sponsored immigrants is wildly inaccurate. Not only are many of them educated, they are generally much better educated than U.S.-born Americans are. Nearly half of all diversity and family-sponsored immigrants who arrived in 2015 had college degrees. Diversity and family-sponsored immigrants were 62 percent more likely than U.S.-born natives to have graduated college. At the same time, they are no more likely to have dropped out of high school than natives.

Table 1 provides the educational attainment for natives and immigrants by type of entry. As it shows, refugees, asylees, and unauthorized immigrants are among the least educated. Employer-sponsored immigrants are by far the most educated. But diversity lottery winners and immigrants sponsored by U.S. family members are right in the middle, and generally better educated than their U.S.-born counterparts.

Table 1: Adult Education Attainment of U.S.-Born and Immigrants Entering in 2015 By Method of Entry

Sources: Office of Refugee Resettlement; U.S. Department of Homeland Security; U.S. Department of Homeland Security; American Community Survey 2015 (5-Year Sample);

Because the Department of Homeland Security (DHS) fails to ask legal immigrants their education level when they enter, the only way to make these estimates was through other sources. To obtain the family and diversity education levels, it was necessary to work backwards from the overall foreign-born educational attainment figures for foreign-born adults over the age of 18 from the Census’s American Community Survey (ACS). Then it is possible to subtract out the other categories of immigrants from the overall figures. The Office of Refugee Resettlement, for example, reports the education level of adult refugees and asylees who entered in 2015.

Almost all employer-sponsored primary applicants in the first, second, and third preferences are required by law to have college degrees, and EB-5 millionaire investors and EB-4 broadcasters and religious workers likely do as well. I assumed that all spouses in EB-1, EB-2, and EB-5 categories had college degrees, and that 75 percent of those in the EB-3 and EB-4 categories do as well. Only 1.3 percent of the adults entering in EB categories were EB-3 “unskilled workers” and their spouses who work in jobs that do not require a college degree, and I assumed that half of these immigrants had no high school degree.

The Center for Migration Studies used ACS data to estimate the educational attainment of all adult unauthorized immigrants. For this analysis, I assumed that new border crossers have the same level of educational attainment as the overall illegal population. According to DHS, about 170,000 border crossers made it into the United States illegally. I assume three quarters were 18 years old or over. While this figure ignores visa overstays who make up roughly half of all new entrants to the illegal population in 2015, illegal immigrants are underrepresented in the ACS, so I only included 127,500. Note that assuming more unauthorized immigrants would result in diversity and family sponsored immigrants appearing more highly educated than they already do.

With these estimates, it is only necessary to subtract refugees-asylees, employment-based workers, and illegal immigrants from the overall foreign-born figures in the ACS to obtain estimates for diversity and family-sponsored immigrants, who represented 96.2 percent of all other adult legal immigrants in 2015. The other 3.8 percent primarily includes Iraqis who worked with the U.S. government in the war and illegal immigrants who received cancelation of removal or U visas as victims of crimes. It is unlikely that these populations are skewing the results in any way.

Table 2 provides the absolute figures used in this analysis. Column 1 uses DHS figures for each legal immigration category subtracting the children who entered in those categories, as well as three quarters of the illegal border crosser figures estimated by DHS in 2015. The educational distribution is based on the shares in Table 1.

Table 2: Adult Education Attainment of U.S.-Born and Immigrants Entering in 2015 By Method of Entry

Sources: Office of Refugee Resettlement; U.S. Department of Homeland Security; American Community Survey 2015 (5-Year Sample); U.S. Department of Homeland Security

No matter what assumptions you use, however, it is just impossible to significantly change the broad conclusions. That’s because the family and diversity categories make up 70 percent of all adult legal immigrants to the United States, and refugees and asylees make up half of the rest and are the least educated group of legal immigrants. We know that immigrants generally are the highest educated that they’ve ever been, so given the size of the family and diversity population, it is just impossible that they are the unskilled group that the attorney general describes them to be.

Diversity and family sponsored immigrants are more highly educated than U.S.-born Americans. There are good arguments for changing the selection process for legal immigrants, but claiming that they lack any skills or education simply is not one of them.

Notes on temporary visa holders: The Census ACS data exclude most temporary visa holders—all tourist and temporary business travelers, which are 81 percent of the total. However, the ACS does include some temporary residents who plan to stay for an extended period, although they are likely underrepresented in the sample. Based on a review of temporary visa categories, however, these individuals would not change the result of this analysis. If anything, they could imply that family and diversity immigrants are slightly better educated than the results in Table 1 imply.

F-1 students make up the largest share of all these other categories (23%), and the vast majority of these students initially enter as high school graduates pursuing U.S. undergraduate degrees. I assume a roughly 90-10 split between undergraduates and graduate students (while the stock is closer to 50-50, I assume most graduate students initially entered as undergrads). Based on the programs under which they enter, J-1 exchange visitors (20%) are also mainly college students (about 84 percent) with some lesser-skilled au pairs (10 percent) and a few higher-skilled workers (5 percent).

Then H-1B, L-1, and E-2 high skilled workers and their spouses (22%) are roughly 95 percent college graduates. H-2A and H-2B seasonal farm and non-farm laborers and their spouses (15%) mainly lack high school degrees (about 70 percent have no high school degree). These categories make up 80 percent of all relevant adult nonimmigrants.

Making reasonable assumptions about the other categories produces the estimates in Table 3, which—while they are educated guess work—are unlikely to be far off. Thus, it appears that new temporary residents are probably a little less educated than the average immigrant in 2015, and so, to the extent that temporary residents skew the figures in Table 1 at all, they make family and diversity immigrants appear slightly less educated than they would otherwise.

Table 3: Adult Education Attainment of Temporary Residents

Source: Authors’ estimates based on U.S. Department of State; Migration Policy Institute; U.S. Citizenship and Immigration Services; Economic Policy Institute; Institute for International Education