What reliable estimates exist for the net number of unauthorized immigrants who entered and remained in the U.S. during 2017–2020?

1. What the question really asks and how experts answer it

As framed, the user wants the net number who entered and remained during 2017–2020, which is a flow measure (arrivals minus departures) distinct from the stock (total unauthorized population); most authoritative public estimates do not publish a single "net entries during 2017–2020" number but instead produce annual or point‑in‑time stock estimates using the residual method, and researchers infer flows from changes in those stocks combined with administrative data on apprehensions, expulsions, deaths and emigration (the residual framework used by DHS, Pew, CMS and others) ^{[4] [1] [3]}.

2. What leading estimates report for the relevant years

Major residual‑method estimates for the unauthorized population cluster in a narrow range: Pew/Passel and Cohn put the population at about 10.5 million in mid‑2017 (and reported the 2021 figure was also roughly 10.5 million), DHS’s Office of Immigration Statistics estimated about 11.4 million as of January 2018, and the Center for Migration Studies’ series and DHS tabulations likewise report figures near that band for the late 2010s — not dramatic year‑to‑year jumps in 2017–2020 ^{[1] [2] [3] [5]}.

3. What the stock changes imply about net entries 2017–2020

Given these stock estimates, the most defensible conclusion is that net entries (new people who entered and remained) during 2017–2020 were modest in aggregate: residual estimates indicate the total unauthorized population in 2017 and in the immediate post‑2019 period remained near 10.5–11.5 million, meaning flows in and out roughly offset each other and produced only small net increases or decreases rather than multi‑million swings over 2017–2020 ^{[1] [2]}. Complementary indicators show border encounter volumes and at‑the‑border apprehension rates that also constrain likely successful entries: average estimated unauthorized entries fell from earlier decades to ≈190,000 per year in 2013–2020, and higher apprehension rates during 2018–2020 imply fewer successful crossings than in the 2000s ^[6].

4. Why estimates differ and where uncertainty remains

Differences between groups reflect methodological choices: data sources (ACS vs CPS), how analysts adjust for survey undercounts, how they model visa overstays versus entries without inspection (EWI), and how they account for deaths, departures, and DHS enforcement records; for example, Warren/CMS emphasizes visa overstays and produces a series that can differ by a few hundred thousand from Pew/MPI panels, and DHS’s own modeling choices yield yet other point estimates ^{[4] [5] [7]}. The residual method itself produces a relatively stable band—typically 10.5–12 million in the 2010s—but its precision for short multi‑year flows (like 2017–2020) is limited because inflows, outflows, undercount adjustments and administrative changes can mask net flows of several hundred thousand ^{[1] [4]}.

5. Bottom line — the best, reliable answer

Using established, peer‑reviewed residual estimates and DHS administrative context, the best reliable statement is this: there is no evidence from leading estimates of a multi‑million net addition to the unauthorized population during 2017–2020; instead the stock remained in the roughly 10.5–11.5 million range, implying net entries across 2017–2020 on the order of hundreds of thousands (positive or negative depending on the series and assumptions), with meaningful uncertainty tied to methodology and undercounts ^{[1] [2] [6]}. If a precise flow number is required, researchers must combine a chosen stock series with DHS flow/aprehension data and explicit assumptions about departures — an exercise that will produce a range, not a single undisputed count ^{[3] [6]}.