Undocumented Immigrant Population Characteristics

Limitations

A MODELED ESTIMATE, not a count of people: the Center for Migration Studies of New York (CMS) estimates the undocumented population from American Community Survey microdata (logical edits that remove likely legal residents, population controls by country, and an undercount adjustment; method: Robert Warren, Journal on Migration and Human Security, 2021). CMS labels the release data year 2024 and does not state the underlying ACS vintage. CMS publishes no margin of error; the figures are uncertain by an unquantified amount. The 'all' row (502,000) is the same CMS figure served as estimator = 'cms', year = 2024 in undocumented_population_estimates, which compares CMS with Pew, MPI and DHS. Of the four estimators' files collected here, only CMS's state file carries breakdowns, so this topic has CMS alone; MPI publishes state profiles with differently banded breakdowns on its website, which are not collected. For the size of Georgia's undocumented population, and how the estimators differ (340,000-502,000 for the latest years), use undocumented_population_estimates. BREAKDOWNS ARE NEVER CROSSED: CMS publishes each one separately (sex, age, region of birth, country of birth, years in the U.S.). A breakdown row carries 'all' in the other four columns, so filter one breakdown at a time, keep the rest at 'all', and never sum across two breakdowns (there is no 'women from Mexico' cell). Rounding: CMS rounds every cell to the nearest 1,000, so each value is within 500 of CMS's unrounded estimate, and a breakdown's parts need not add to the total: the age brackets sum to 501,000 against 502,000. On birthplace rows CMS also withholds values under 1,000 (below), and it does not say whether it applies that rule before or after rounding, so a published birthplace 1,000 means roughly 500-1,499 or 1,000-1,499. region_of_birth uses CMS's eight regions. CMS's north_america is Mexico and Canada, with central_america and caribbean as separate regions - unlike the DHS OHSS north_america in the naturalizations and new_permanent_residents topics (which includes Central America and the Caribbean) and the ACS regions in the census topics. country_of_birth lists the 31 countries CMS publishes, chosen by CMS nationally; it is NOT exhaustive: the listed countries sum to 472,000 of 502,000, and there is no 'other' row (a residual would carry up to +/-16,000 of rounding). 'korea' is CMS's unsplit 'Korea'; 'china' is CMS's 'China' (CMS does not say whether Hong Kong and Taiwan are included). CMS does not show a country or region with fewer than 1,000 undocumented residents in a state ('--' in its file, NULL here); no Georgia cell is withheld in the 2024 release. years_in_us is years since arrival in the United States at the data year: under_5 (184,000) matches CMS's own 'arrived 2020 onward' count. demographic carries sex (female, male) as the ACS records it; age_bracket follows CMS's eight age bands. Not included: CMS's race and Hispanic-origin breakdown (CMS does not say whether Pacific Islanders fall in 'Asian non-Hispanic' or 'Other non-Hispanic'); its liminal-status flags (DACA, TPS, parole, SIJS, U/T visa, pending asylum), whose categories overlap (Georgia's sum to 253,000 against 193,000 people with any liminal status) and so cannot be one breakdown; its period-of-entry table (an undefined smaller universe); its year-of-immigration table (the same information as years_in_us); and its socio-economic tables (marital status, English ability, language, school enrollment, education, computer access, health insurance, employment, occupation, poverty, family size, income). Georgia statewide only: CMS publishes states, not counties, and the other states and the national row are not included. Every value is CMS's published aggregate; this platform applies no suppression or adjustment. CMS's term is 'undocumented'.