ICE Removals Originating in Georgia

Limitations

Years are CALENDAR years of the departure date (Departed Date, with a month column), not federal fiscal years. ICE and the Deportation Data Project report by fiscal year (October 1 to September 30), so a figure here will not match a published 'FY2026' number. Coverage is 2022-10-01 through 2026-08-06: calendar 2022 is October-December only and 2026 stops on 6 August, so both end years are partial and a trend line dips at each end for that reason alone. GEORGIA MEANS REMOVALS WHOSE REMOVAL RECORD NAMES GEORGIA AS THE APPREHENSION STATE (Apprehension State = GEORGIA), not people removed from Georgia or Georgia residents. That field is empty on 71.7% of the 1,250,386 removals nationwide (896,039 rows, 829,606 of them labeled as CBP arrests), and those removals are in no state's count. The Georgia figure is therefore a FLOOR, and a low one. Linking each removal to the same person's ICE arrest record (the Deportation Data Project's arrests file, matched on the anonymized person ID; measured on the 2026-08-06 release) finds 7,930 more removals, 57% more than the 13,923 counted here, whose ICE arrest was recorded in Georgia - 48% to 65% more in every full calendar year. 7,026 of them have no Apprehension State and 5,838 are labeled CBP on the removal record, although the linked arrest is an interior ICE arrest in Georgia, usually on the same date. The publisher documents the person ID only between arrests and detention, so this link is inferred; and the arrests file starts on 2022-10-01, so the shortfall is understated for 2022 and 2023. The CBP label is therefore not a reliable sign of a border arrest: nationally, 7.7% of CBP-labeled removals departing before 2025 link to an interior ICE arrest, against 57.4% of those departing from July 2025. Of the 13,923 Georgia removals counted here, 13,913 are labeled ICE and 10 CBP, while nationally 66.6% are labeled CBP. The arresting agency is NOT published as a breakdown (it would be a column that is 'ice' in 99.9 percent of cells). Every Georgia figure here is interior (ICE) enforcement; do not compare it against a national total that reads CBP-labeled removals as border removals. ice_program is the ICE program that produced the removal (the Criminal Alien Program - jail and prison handoffs - Fugitive Operations, the two 287(g) models, docket control, Alternatives to Detention, Border Patrol, other), from the source's 'Final Program Code' column, whose header is swapped with 'Final Program' in the release (the transform reads the pair by content). Reliability caveat from the publisher: the Deportation Data Project states that it "continue[s] to have questions about the reliability of the removals data" and that recent releases differ from the late-July-2025 release it believes was most reliable; this dataset is built from the later 2026-08-06 release. The four source workbooks hold 1,250,386 rows against the roughly 1.05 million records DDP's page advertises; the observed count is treated as authoritative and no rows were dropped to reconcile. Rows identical on all 35 source columns are collapsed before counting (870 nationwide, 0 among the Georgia rows). The source has no county field, so this dataset is STATE-ONLY: county_fips is NULL on every row and there is no county breakdown (Apprehension City is mostly empty and mixes detention towns such as Lumpkin with arrest locations). State rows are a complete grid of every month x category combination, so a 0 means no removal matched that cell (zero is real; no suppression is applied at state grain). month = 'all' is the calendar-year total; every categorical's 'all' value is its total, so filter each to 'all' except the one you are breaking out, or the same removal is counted on several rows. demographic is ICE's own gender determination (male/female), not self-reported; removals recorded with gender Unknown count toward 'all' only, so male + female is slightly below all. citizenship_country names the 22 countries most frequent among Georgia records across the ICE and immigration-court sources; every other country is 'other' and an unrecorded country is 'unknown'. criminality is ICE's own case classification as recorded at removal. Removals are counted, not people: a person removed twice in the window is two removals. NOT THE SAME COUNT AS `eoir_case_outcomes`.num_removal_orders: that topic counts removal ORDERS an immigration judge issued, this one counts deportations ICE carried out. The two do not reconcile at state level and neither checks the other. Many removals here never passed through immigration court (reinstated prior orders, expedited removal), and an order there may never be executed (an in absentia order against someone not in custody, for example); `year` here is the departure year, there the filing year; and Georgia here is the state where the originating arrest was recorded, there the address of record (the detention center for detained respondents, including people arrested in other states).