A Center for Immigration Studies analysis of federal household-survey data estimates that the U.S. foreign-born population declined by 2.9 million between January 2025 and July 2026, while the organization preliminarily calculates that the illegal-immigrant population fell by about 2.3 million during the same period.
By yourNEWS Media Newsroom
The U.S. foreign-born population has fallen by an estimated 2.9 million people since President Donald Trump returned to office in January 2025, according to a new Center for Immigration Studies analysis of the Census Bureau’s monthly Current Population Survey.
The analysis places the foreign-born population at approximately 50.5 million in July 2026, down from about 53.3 million in January 2025. Because the monthly figures are rounded, CIS calculates the overall decline at 2.9 million.
The foreign-born category is considerably broader than the illegal-immigrant population. It includes everyone living in the United States who was not a U.S. citizen at birth, including naturalized Americans, lawful permanent residents, long-term temporary residents such as foreign students and guest workers, and immigrants living in the country illegally.
CIS separately estimates that the illegal-immigrant population may have declined by approximately 2.3 million, from an estimated 15.8 million in January 2025 to 13.5 million in July 2026. The organization describes that figure as preliminary because the Current Population Survey does not directly identify respondents by immigration status.
“The best data we have shows an enormous decrease in the total foreign-born population,” said Steven Camarota, CIS director of research and lead author of the analysis.
CIS attributes most of the decline to non-citizens, particularly Latin American non-citizens who reported arriving in the United States in 1980 or later, a population that substantially overlaps with the illegal-immigrant population. The organization said the number of naturalized citizens, by comparison, increased by approximately 680,000 between January 2025 and July 2026.
The reversal followed a four-year period in which CIS calculates that the foreign-born population increased by approximately 8.3 million between January 2021 and January 2025. The organization described that as the largest four-year increase recorded in the United States and said the subsequent decline represents a sharp change in the direction of immigration-related population growth. Its analysis of the four-year increase forms part of a broader series examining immigrant population trends.
The 2.9 million reduction amounts to roughly 35% of the preceding 8.3 million increase, although the two figures represent net population changes rather than a count of specific people entering and subsequently leaving the country.
The Center argues that a decline of this magnitude would mean net migration — arrivals minus departures — turned negative during the first 18 months of Trump’s second administration. CIS described that as the first such period since the 1930s, while also acknowledging substantial statistical uncertainty around the precise size of the decline.
Separate Census Bureau population estimates point in the same direction in terms of a sharp migration slowdown, though not to negative net international migration. The Census Bureau reported that net international migration peaked at approximately 2.7 million in 2024, fell to about 1.3 million in 2025 and was projected to decline to approximately 321,000 in 2026 if prevailing trends continued. Census attributed the decrease to a combination of fewer people entering the United States and more people leaving.
That distinction is significant. The Census Bureau’s January 2026 projection still anticipated positive net international migration, while CIS says the more recent household-survey results through July imply that migration became substantially negative.
CIS calculates that, if the 2.9 million decline is taken at face value and approximately 450,000 deaths among foreign-born residents are incorporated, net migration would have been about negative 2.45 million between January 2025 and July 2026, equivalent to an annualized rate near negative 1.63 million. The organization discusses that calculation and its limitations in its net-migration analysis.
The uncertainty surrounding those figures is substantial. CIS notes that the monthly foreign-born estimate has a wide confidence interval. Using the lower end of the January 2025 estimate and upper end of the July 2026 estimate, the organization says the decline could theoretically be as small as 763,000. Under that scenario, it calculates net migration at approximately negative 313,000 for the entire period, or about negative 208,000 annualized. The report addresses that confidence-interval scenario.
Long-term Census Bureau historical figures show that the foreign-born population previously declined for several decades during the middle of the 20th century. It stood at approximately 14.2 million in 1930, 11.6 million in 1940, 10.3 million in 1950, 9.7 million in 1960 and 9.6 million in 1970.
The 1930-to-1940 decline amounted to approximately 2.6 million over a decade, followed by a roughly 1.3 million decrease from 1940 to 1950. If the CPS-based 2.9 million estimate for January 2025 through July 2026 accurately reflects an actual population decrease, the numerical reduction would exceed the decline recorded during any single one of those decades despite occurring over only 18 months.
Signs of the reversal appeared considerably earlier. Reporting in August 2025 cited CPS figures showing that the foreign-born population had fallen from 53.3 million in January to approximately 51.9 million in June, a decrease of about 1.4 million. By October, the estimated decline had reached approximately 2.2 million.
Those figures followed record growth earlier in the decade. The foreign-born population was estimated at 49.5 million in December 2023 after increasing by roughly 4.5 million during the first three years of the Biden administration. By March 2024, it had climbed to approximately 51.6 million, representing about 15.6% of the U.S. population.
The current CIS report relies on the Current Population Survey, commonly known as the household survey, which is conducted each month by the Census Bureau for the Bureau of Labor Statistics. Approximately 110,000 people are included in the monthly survey.
For July 2026, CIS says the estimate of 50.5 million foreign-born residents carries a margin of error of plus or minus approximately 1.038 million at a 90% confidence level. That leaves room for significant month-to-month variation even when the longer-term trend in the survey appears substantially lower.
The Census Bureau itself identifies the CPS among its surveys providing foreign-born population information and publishes annual foreign-born tables derived from the survey. CIS cites those uses in explaining why it selected the CPS for its analysis.
Other government agencies and research organizations have also used CPS data in immigration research. The Congressional Budget Office has incorporated the survey into immigration analysis, while the Social Security Administration has used CPS data in work estimating the population living in the United States illegally.
The Bureau of Labor Statistics publishes monthly CPS-derived employment information for foreign-born workers, and the Federal Reserve’s economic database maintains foreign-born population series derived from the household survey. Pew Research Center also used monthly CPS figures when examining the immigrant-population decline in 2025.
CIS nevertheless acknowledges an important methodological warning. A 2024 Census Bureau working paper cautioned against relying on the CPS to estimate the size of the foreign-born population when better data are available.
The American Community Survey is substantially larger and generally preferred for detailed foreign-born population estimates, but it is released with a longer lag. CIS argues the CPS is particularly useful for studying the rapid policy changes of 2025 and 2026 because no comparably current demographic survey is available. The report also points to immigration research by the Migration Policy Institute and Brookings Institution as examples of the broader use of CPS-related population data.
Another methodological issue involves the weights used to turn CPS survey responses into national estimates.
CIS has previously examined the implications of CPS population weighting. The Census Bureau recalibrates the survey weights each January based on updated estimates of births, deaths and net international migration. The Bureau of Labor Statistics provides details on those annual population-control adjustments, while the underlying national population projections are published by the Census Bureau.
The January 2026 adjustment produced an upward shift of approximately 874,000 in the estimated immigrant population compared with December 2025, according to CIS. The organization discusses the effect in its analysis of the January reweighting.
Citizenship and nativity, however, are not themselves variables that Census statisticians force to predetermined population totals. CIS emphasizes that point in its discussion of CPS controls.
Looking separately within years to avoid the January reweighting, CIS says the foreign-born population declined approximately 1.7 million between January and December 2025 and another 2 million between January and July 2026. The organization argues those within-year movements support the existence of a substantial decline independent of annual weighting changes.
The survey also appears to show the U.S.-born population increasing approximately 4.7 million between January and July 2026, but CIS says that number should not be interpreted as an actual measured increase. Because the CPS is weighted to a predetermined total U.S. population, a lower estimated foreign-born population mechanically produces a larger estimated U.S.-born population. The survey can therefore identify changes in population composition without independently measuring the overall national population total.
CIS also examined the labor-market composition of the decline.
Foreign-born employment fell by slightly less than 1.3 million, while the foreign-born population not working declined by about 1.6 million. CIS calculates that workers accounted for approximately 44% of the overall decrease even though about 60% of foreign-born residents were employed in January 2025.
The remainder consisted primarily of children, elderly people, full-time caregivers and other non-workers. CIS argues that the disproportionate reduction among non-workers, combined with reported employment gains among U.S.-born workers, may have limited the immediate effect on the overall workforce.
The separate BLS establishment survey, which measures nonfarm payroll employment through employers rather than households, has continued to show job growth. CIS discusses the relationship between those data and the decline in immigrant workers in its labor-market analysis.
Estimating the illegal-immigrant population requires an additional layer of analysis because the CPS does not ask respondents whether they are lawfully present in the United States.
CIS says its estimation method generally follows the residual approach used by the Department of Homeland Security’s Office of Homeland Security Statistics. Researchers first estimate the legally present immigrant population and then subtract that number from the total foreign-born population after accounting for other demographic factors.
People who reported entering before 1980 are generally treated as overwhelmingly unlikely to be in the country illegally, an assumption CIS explains in its methodological discussion. Its estimate classifies as illegal immigrants foreign-born residents who are not naturalized citizens, permanent residents or valid temporary-visa holders, and includes people with pending asylum applications and parolees, as described in its definition of the estimated population.
Changes in legal immigration also affect those calculations.
CIS cited administrative information and outside analyses suggesting that legal entries declined during fiscal 2026. A Cato Institute analysis, a Migration Policy Institute assessment and an American Immigration Council analysis each examined reductions or processing changes involving legal immigration.
Universities also reported lower international enrollment. The Association of American Universities documented declines among international students, while Bloomberg reporting and Georgetown University reporting described similar trends.
Federal adjustment-of-status data showed quarterly declines in new lawful permanent-resident arrivals during fiscal 2025, according to CIS. Refugee-admissions figures also showed considerably fewer arrivals, with CIS projecting that fiscal 2026 could finish roughly 90,000 below fiscal 2024.
At the same time, the organization cautioned that estimating the full decline in lawful arrivals is difficult because the State Department’s monthly immigrant-visa statistics did not provide the complete 2026 information CIS said it would need. The report summarizes those constraints in its discussion of legal entrants.
Using the information that was available, CIS preliminarily estimated that approximately 31.6 million post-1980 immigrants living in the country in July 2026 were legally present, an estimate detailed in its calculation of the legal population.
The July CPS contained an estimated 44.7 million foreign-born residents who arrived in 1980 or later. Subtracting the estimated 31.6 million legal population produces approximately 13.1 million people CIS classified as illegal immigrants in the survey, according to its residual calculation.
CIS then made the same adjustment for population undercount used in its earlier estimate, producing the preliminary July 2026 total of 13.5 million.
Compared with the organization’s January 2025 estimate of 15.8 million, that would represent a reduction of approximately 2.3 million, or 14.6%. The net change reflects multiple flows simultaneously: new illegal arrivals increase the population, while deportations, voluntary departures, deaths and transitions into legal status reduce it.
The nationality and citizenship composition of the CPS decline is another reason CIS believes much of the decrease occurred among illegal immigrants.
Previous estimates by DHS, Pew Research Center, the Center for Migration Studies and the Migration Policy Institute have estimated that roughly three-quarters of the illegal-immigrant population is from Latin America.
CIS also notes that approximately 80% of border encounters during the earlier migration surge involved people from Latin American countries.
Against that backdrop, CPS data show a sharp increase from 2021 through late 2024 among Latin American non-citizens who arrived in 1980 or later, followed by a steep decline beginning around January 2025. CIS argues that the pattern is consistent with a substantial decrease in the illegal-immigrant population, though the survey itself cannot establish the legal status of individual respondents.
One of the largest uncertainties is whether some immigrants remain in the country but have become less willing to participate in a federal household survey because of increased immigration enforcement.
CIS examined that possibility in an earlier study of CPS participation and said it found no clear indication that foreign-born respondents had suddenly become less willing to answer the survey or its immigration-related questions.
The latest report similarly says the share of respondents declining to answer questions about citizenship or country of birth remains low and has not meaningfully changed. CIS discusses that evidence in its nonresponse analysis.
It also found no surge in non-citizens reporting themselves as naturalized citizens. Instead, the naturalized population increased by roughly 680,000, which CIS said broadly aligns with federal naturalization data.
There is, however, a longer-running decline in participation in the Current Population Survey overall. The Bureau of Labor Statistics publishes the survey’s historical response rates, which have trended downward for more than a decade.
CIS tested whether lower CPS response rates were associated with fewer foreign-born respondents. From January 2021 through July 2026, the organization calculated a correlation of negative 0.89 between overall response rates and the estimated foreign-born population, as detailed in the report’s response-rate calculation.
Because immigration increased substantially from 2021 through 2024 and then declined while response rates were moving lower over the entire period, CIS said the simple correlation offered limited insight.
Researchers therefore performed another analysis using year as a control variable, an approach described in the report’s statistical methodology.
That regression found that each 1 percentage-point increase in CPS response rates was associated with an estimated 110,000-person decrease — rather than increase — in the foreign-born population. CIS presents that result in its regression findings and says a corresponding analysis of the residual relationship did not show a strong connection between survey response rates and the reported immigrant population.
The researchers conducted another check using Hispanic respondents because prior research indicates that a large majority of the illegal-immigrant population is Hispanic. A Center for Migration Studies analysis cited by CIS estimates Hispanics comprise roughly three-quarters of the population living in the country illegally.
The weighted Hispanic population in the CPS rises mechanically because Hispanic origin is among the characteristics incorporated into Census population controls. The raw, unweighted number of Hispanic respondents, however, began declining near the end of 2024.
CIS says the unweighted number of Hispanic survey respondents fell 7.4% between January 2025 and July 2026, and 10.1% when December 2024 is used as the starting point. The report discusses those figures in its Hispanic-response analysis.
The organization argues that a declining raw number of Hispanic respondents is consistent with some foreign-born Hispanics leaving the United States rather than merely changing how they describe citizenship or birthplace. It also acknowledges another possibility: some people may remain in the country but no longer participate in the survey.
Federal enforcement data provide additional evidence of increased departures, although CIS says enforcement by itself cannot explain an estimated 2.9 million decline.
U.S. Immigration and Customs Enforcement’s detention and removal data indicate that ICE removals between January 2025 and July 2026 were approximately 260,000 higher than during the same span of the preceding two years, according to CIS. The report explains that comparison in its enforcement-data analysis.
Not all ICE removals can be treated as departures by people who had been settled inside the United States. Some involve individuals transferred from Border Patrol custody, who may never have met the residence requirements necessary to appear in Census population surveys.
More detailed records assembled by UCLA’s Deportation Data Project, including its underlying ICE removal dataset, allowed CIS to focus more closely on interior enforcement.
Using those data, CIS said interior removals more than tripled during the first portion of Trump’s second term compared with the final portion of the Biden administration, increasing by approximately 207,000. The organization describes that calculation in its interior-removal analysis.
Voluntary departures have increased as well.
A Stateline analysis of Deportation Data Project information found 89,494 voluntary departures during the first 16 months of the Trump administration, compared with 11,977 during the final 16 months of the Biden administration.
ICE describes voluntary departure as a process through which eligible individuals leave the United States without a formal removal order under applicable circumstances.
CIS cautions that some of those recorded departures involved people encountered at the border rather than long-term U.S. residents. Even after accounting for that limitation, the organization estimates that enforcement-associated departures from the interior were roughly 350,000 higher during the early Trump period than during the comparable late-Biden period, as discussed in its departure analysis.
That increase is significant but still far smaller than the population decrease implied by the CPS.
For the household-survey estimates to be substantially correct, CIS says voluntary emigration by people who were not formally removed would also have had to rise sharply. The organization argues that increased interior enforcement and the possibility of deportation may have encouraged additional people to leave on their own, but no direct administrative dataset measures such departures comprehensively.
The findings therefore present two conclusions of different certainty.
The first is directly observable within the CPS: the survey’s estimated foreign-born population fell sharply from January 2025 through July 2026, with the long-term movement persisting despite month-to-month sampling fluctuations and the January 2026 population-control adjustment. That decline is also consistent in direction with Census Bureau evidence showing dramatically lower net international migration as both arrivals declined and departures increased.
The second — the estimate that the illegal-immigrant population declined from 15.8 million to 13.5 million — is an inference based on the foreign-born survey data, estimates of the legally present population, adjustments for undercount and information about the demographic groups accounting for the decline. It is not a direct federal count of people residing illegally in the United States.
CIS nevertheless concludes that the available evidence most strongly supports a substantial decrease in that population. The organization estimates the reduction at approximately 2.3 million, or 14.6%, between January 2025 and July 2026.
The report also emphasizes that only 44% of the estimated foreign-born decline consisted of workers, limiting the immediate workforce reduction relative to the overall population decrease.
At the same time, CIS explicitly leaves open the possibility that some unknown portion of the apparent decline results from immigrants becoming less willing to participate in the CPS. The organization says it has found no clear statistical evidence that such nonresponse explains the trend, but acknowledges that the possibility cannot be completely eliminated.
Taken together, the monthly household survey, enforcement records, lower legal and illegal arrivals, increased interior removals and evidence of higher emigration point to a major reversal in U.S. migration patterns after January 2025. The remaining disagreement centers less on the direction of that change than on its precise magnitude: Census Bureau population estimates show net international migration slowing dramatically but remaining positive in its 2026 projection, while CIS interprets the more current CPS data through July as evidence that net migration has already moved below zero.
Original article
Comments