The American people are often assured that the census is a neutral, constitutional count of those living in the United States—as the Constitution calls it, an “actual Enumeration.” It is supposed to be a simple act of counting persons for representation.

Until the most recent census, that is what every previous census at least attempted to do. But in 2020, something fundamentally changed. For the first time since 1790, the federal government did not just count people. It altered the count. It engineered the data. And it did so by applying an algorithm to the numbers under the sanitized, academic label of “differential privacy.” 

According to the Census Bureau:

Differential privacy is a scientific framework for processing data to protect the identities and personal information of the people in the data. It works by adding statistical noise—small, random additions or subtractions—to every published statistic so that no one can reidentify a specific person or household with any certainty using any combination of the published data.

In practice, however, differential privacy deliberately falsifies real population data by adding random errors, meaning the numbers used for political representation are not fully accurate. The algorithm scrambles local data in ways that shift political power, disproportionately benefiting urban, left-leaning areas while diluting the influence of rural and conservative communities.

In plain terms, the government took real communities and replaced them with statistical fiction. And then it told states to draw political maps based on that fiction.

It is important to call this what it is. Differential privacy is a deliberate system of data manipulation. It is not a safeguard. It is not a minor technical adjustment.  The more random noise it adds, the less accurate the data become. This is not a bug; it is the system’s signature design feature.

The Census Bureau applied this algorithm nationwide: Every state, every locality, and every census block is now under its umbrella. And while the Bureau preserved raw statewide totals to avoid obvious detection, it intentionally distorted the granular data that govern representation by shifting where people are counted, even if they don’t live there. 

That is where the fraud lies.

Redistricting does not happen at the state level. It happens at the block level. It depends on knowing how many people live in a neighborhood, a precinct, a voting district. That is where political power is allocated. And that is precisely where differential privacy does its damage. It shifts populations between blocks. It scrambles demographic characteristics. It makes accurate map-drawing impossible. 

A Harvard University analysis found that differential privacy transfers population across geographies and makes it impossible to comply with the one person, one vote principle as it is currently understood. 

The consequences were predictable and profound. When you distort block-level data, you distort representation. When you obscure citizenship characteristics, you prevent meaningful distinctions between citizens and noncitizens in the redistricting process and any ability to uncover those distinctions. And when you shift or inflate populations within states, you quietly shift political power. 

Differential privacy, combined with counting errors, tilted the political scales toward Democrats for both apportionment and redistricting. And this was not just about maps. It was about seats in Congress and political power in this country.

The Census Bureau later admitted major counting errors across multiple states. Interestingly, the errors were not neutral. They resulted in a net shift of congressional representation. Florida and Texas, fast-growing and more conservative states, were undercounted and lost seats they otherwise would have gained. Meanwhile, blue states like Minnesota and Rhode Island held onto seats they should have lost, and Colorado gained one it did not deserve. This is a reallocation of political power at the national level.

And it did not stop there. Differential privacy created something even more insidious: plausible deniability.

Before 2020, if a locality identified a census error, it could challenge it through established processes called “Count Question Resolutions.” After 2020, that became impossible. No one could tell whether a bad number was a real mistake or algorithmic noise. The data was intentionally obscured. The inputs were hidden, and the system could not be audited. 

Even worse, the algorithm effectively nullified the possibility of ever determining citizenship at the level needed for lawful redistricting. It masked the very characteristics that matter for political representation. And because large urban areas tend to have higher noncitizen populations, this masking effect increased their representational weight while diluting the influence of rural and conservative regions. That is the real-world effect. Power moved away from rural America and toward dense, progressive urban centers.

All of this was done in the name of privacy. But privacy was the pretext, not the outcome. The outcome was instability in the data, the breakdown of verification mechanisms, and the inability to ensure that representation reflects actual communities. 

The Constitution requires an actual enumeration. Not an estimate. Not a model. Not an algorithm that rewrites reality after the fact. The census is supposed to count people, not simulate them. Yet a simulation is exactly what we got.

The defenders of differential privacy argue that it is necessary to protect individual identities. But that argument collapses under scrutiny. The federal government managed to conduct every prior census in American history without resorting to wholesale data falsification. What changed was not technology. What changed was the willingness to manipulate outcomes in the name of process. And the Bureau is already making preparations to continue this manipulation in the 2030 census. The need for redress has never been more apparent.

This is why President Donald Trump and institutions like the Center for Renewing America call for the 2020 Census to be republished. The objective is not to correct technicalities but to repair the Constitution. Because if the data underlying representation are corrupted, then representation itself is corrupted. 

Representation in Congress; allocation of electoral votes; distribution of federal funds; the drawing of every legislative district in the country—all of it flows from the census. If the census is manipulated, then the entire system of self-government is built on a compromised foundation.

The 2020 Census did not simply make mistakes. It introduced a system that made mistakes untraceable and corrections impossible. It replaced transparency with opacity, accuracy with approximation, enumeration with manipulation.

Americans were told they were being counted. Instead, they were being adjusted.

That is why the following question is not partisan, but patriotic: Did you know the 2020 United States Census defrauded the American people?