Which datasets (Census, ACS, DHS, other) did Camarota rely on in his 2024 analyses?
Executive summary
Steven A. Camarota’s 2024 analyses in the Center for Immigration Studies repeatedly rely on U.S. household survey public-use microdata — primarily the Current Population Survey in its monthly and supplement forms — and on public-use products produced or distributed by the Census Bureau; his team explicitly cites the CPS (including the January–March three‑month average and the August CPS) and the 2024 Annual Social and Economic Supplement in multiple 2024 and 2024–25 CIS reports and testimonies [1] [2] [3]. The publicly available snippets do not show a reliance on Department of Homeland Security administrative datasets in the quoted materials, so any DHS use is not documented in these sources [4] [1].
1. The backbone: Current Population Survey (CPS) monthly microdata
Camarota’s 2024 writings and testimony lean heavily on the CPS — the U.S. household survey run jointly by the Bureau of Labor Statistics and the Census Bureau — using monthly and short-run averages of public‑use CPS files to estimate changes in the foreign‑born and noncitizen populations (he and Karen Zeigler analyzed “raw data” from the CPS for at least one prominent 2024 piece) [3] [2]. His pieces reference combined three‑month averages (January–March) and specific monthly snapshots such as August 2024, showing that his empirical claims are driven by the CPS public‑use series rather than by opaque administrative tallies [2] [1].
2. The Annual Social and Economic Supplement (ASEC) of the CPS
For income, earnings, and some demographic breakdowns Camarota cites the 2024 Annual Social and Economic Supplement (ASEC) — commonly described in his materials as the “public use 2024 Annual Social and Economic Supplement” — indicating use of the CPS ASEC microdata for cross‑sectional measures and supplement‑based estimates in 2024 analyses [1]. The ASEC reference is explicit in testimony about housing and earnings and is presented as the source for particular tabulations released by CIS [1].
3. How Census Bureau public‑use data is described and used
Camarota repeatedly frames his statistics as coming from “Census Bureau data collected earlier this year” when reporting counts of voting‑age noncitizens and related state‑level estimates, reflecting his use of Census‑distributed CPS public‑use files rather than proprietary datasets [2]. Several CIS pieces and congressional testimony explicitly call out “public use” CPS and ASEC files as their underlying data, underscoring that the analyses depend on survey microdata accessible to outside researchers [1] [2].
4. What the sources do not show: DHS, ACS, or administrative tallies
Within the provided 2024 materials and snippets, there is no explicit citation of Department of Homeland Security administrative datasets, such as immigration enforcement or interior enforcement records, nor an explicit citation of the American Community Survey (ACS) in the quoted passages; therefore, use of DHS administrative data or ACS microdata is not documented in these excerpts and cannot be asserted from them [4] [1] [2]. The absence of DHS/ACS citations in the supplied sources means the question of whether Camarota used those datasets in other 2024 work is unanswered by the available reporting [4] [1].
5. Limitations, interpretation and competing perspectives
CIS’s transparency about using public‑use CPS and ASEC files means other researchers can replicate many calculations, but the CPS has known limits — sample size, undercount of certain immigrant subgroups, and timing sensitivity when comparing short monthly windows — issues raised in broader methodological debates about immigrant population measurement; critiques of Camarota’s conclusions often point to those survey limitations and to CIS’s institutional advocacy for lower immigration as context for interpretive choices [1] [3]. The sources supplied document CIS’s dataset choices but do not adjudicate contested methodological adjustments (for example, undercount corrections) or whether administrative datasets were consulted outside the cited materials, so those remain open questions beyond the cited documents [1] [2].