# Secret Census Algorithm STOLE House Seats from Red States! Source: https://www.youtube.com/watch?v=SEf7qD5lLlY Recap page: https://rapidrecap.app/video/SEf7qD5lLlY Generated: 2026-02-24T15:05:37.572+00:00 --- ## Quick Overview The 2020 Census was deliberately rigged using a secret, unvetted algorithm called 'differential privacy' that inaccurately allocated seats and billions in federal funding, disproportionately harming Republican-leaning states and benefitting Democratic-leaning states, leading to a shift of nine House seats away from Republican-leaning states and costing them over $86 billion in federal funds. **Key Points:** - The 2020 Census was marketed as an 'actual enumeration' but used a secretive, unvetted algorithm called 'differential privacy' by Census Bureau's Chief Scientist, John M. Abowd. - This algorithm intentionally injected mathematical noise into every block count, turning the census from a headcount into a model, which resulted in significant coverage errors, with 14 states showing significant errors. - Republican-leaning states were widely undercounted, leading to a shift of nine House seats away from them toward Democratic-leaning states. - Florida's undercount was roughly three-quarters of a million people, and Texas's undercount was on the order of half a million, both of which should have gained congressional seats. - The error cost states billions in federal funding; states like Florida, Minnesota, and Rhode Island kept seats they would have had under an accurate count, while Colorado gained a seat it did not deserve. - The Supreme Court's opinion in Department of Commerce v. House of Representatives (2000) made clear that statistical sampling for apportionment is illegal on statutory grounds, but Abowd's algorithm amounted to statistical sampling by another name. - The core lever in the privacy budget, Epsilon, is not a philosophy but a number with consequences, with an Epsilon of 15.81 resulting in the effective erasure of the order of 30 to 95 people in small areas. ![Screenshot at 00:00: Andrew Branca, Scotus Bar Attorney, begins the discussion by introducing the topic of the Census being weaponized against the American people.](https://ss.rapidrecap.app/screens/SEf7qD5lLlY/00-00-00.jpg) **Context:** The discussion centers on the alleged manipulation of the 2020 US Census data by Census Bureau Chief Scientist John M. Abowd through the implementation of a new, opaque methodology called 'differential privacy.' The speaker argues this was done to intentionally alter population counts, impacting congressional apportionment and federal funding allocation, which disproportionately favored Democratic-leaning states over Republican-leaning states. ## Detailed Analysis The 2020 Census was intentionally corrupted by the Census Bureau's Chief Scientist, John M. Abowd, using a secretive algorithm called 'differential privacy.' This system, which should have been an actual enumeration, instead injected mathematical noise into every block count, effectively turning the census into a flawed model. Evidence surfaced in March and May 2022 showing that 14 states experienced significant coverage errors, including overcounts and undercounts. The errors disproportionately harmed Republican-leaning states, resulting in a net shift of nine House seats away from them toward Democratic-leaning states. Florida's undercount was nearly 750,000 people, and Texas was undercounted by about half a million; both states should have gained seats. Colorado wrongly gained a seat, while Florida and Texas missed out on seats they deserved. The Electoral College was affected, and these states lost over $86 billion in federal formula funds. The practice is illegal because Abowd's algorithm constituted prohibited statistical sampling for apportionment, as established by the Supreme Court's opinion in Department of Commerce v. House of Representatives (2000). The core privacy lever, Epsilon, was set too high (15.81), meaning that blocks of 105 residents could have their count published as 95, 105, or even further off, depending on post-processing, thus obscuring the data and becoming statistical alteration that mandated counting people, not blurring them. ### Census Rigging Allegations - The 2020 Census was not an actual enumeration but used a secret algorithm called 'differential privacy' - This system injected mathematical noise into block counts, creating errors - Abowd, the Chief Scientist, pushed this opaque methodology. ### Impact of Errors - 14 states showed significant coverage errors; Republican-leaning states were widely undercounted, leading to a shift of nine House seats away from them - Florida was undercounted by 750,000, Texas by half a million, resulting in lost seats and over $86 billion in lost federal funds. ### Legal Violation - The Supreme Court previously ruled statistical sampling for apportionment illegal, and Abowd's algorithm constituted statistical sampling by another name - The court cited the drumbeat of doubt and the blocked question process as evidence of bad faith. ### The Epsilon Variable - Epsilon (15.81) determined the level of privacy noise; this noise caused blocks of 105 residents to be published as 95, 105, or even further off, leading to statistical alteration, not just blurring. ### Simulation Results - Researchers simulating the algorithm found differential privacy moves people around, leading to communities looking less diverse and changing political makeup - The system made mixed neighborhoods look more uniform, and its design is biased against counting people, not blurring them. ![Screenshot at 00:00: Andrew Branca, Scotus Bar Attorney, begins the discussion by introducing the topic of the Census being weaponized against the American people.](https://ss.rapidrecap.app/screens/SEf7qD5lLlY/00-00-00.jpg) ![Screenshot at 00:54: A screenshot of an article titled 'The Algorithm That Rigged the Census: How One Bureaucrat Stole the House and Billions in Funding,' providing context for the claims being discussed.](https://ss.rapidrecap.app/screens/SEf7qD5lLlY/00-00-54.jpg) ![Screenshot at 01:43: The speaker references the article, asking how the census got corrupted, showing the article text on screen.](https://ss.rapidrecap.app/screens/SEf7qD5lLlY/00-01-43.jpg) ![Screenshot at 06:52: The speaker discusses the illegality of Abowd's statistical sampling for apportionment, citing the Supreme Court's stance on statutory grounds.](https://ss.rapidrecap.app/screens/SEf7qD5lLlY/00-06-52.jpg) ![Screenshot at 08:59: A map graphic illustrating the differential impact of the algorithm across states, showing states shaded based on the effect of the noise allocation strategy.](https://ss.rapidrecap.app/screens/SEf7qD5lLlY/00-08-59.jpg)