You can’t fix a problem you don’t understand, and it’s very clear that the Federal Communications Commission under Donald Trump doesn’t want to understand its failure to make affordable broadband available to all Americans.

During a pandemic when Americans are forced to work, learn, and get their health care online, the FCC’s refusal to accurately measure US broadband connectivity gaps has quickly shifted from administrative farce to outright tragedy.

Once each year, the law requires the FCC to determine whether broadband is “being deployed to all Americans in a reasonable and timely fashion.” If not, the agency is supposed to take concrete steps toward fixing the problem.

While the Trump FCC pantomimes these obligations, the reports it issues rely on poor methodology, unsound logic, and flawed data, resulting in a distorted view of America’s broadband problem and government policy based on little more than magical thinking.

The FCC’s 2020 Broadband Deployment Report, released last June, claims the number of Americans without access to broadband sits somewhere around 18.3 million. But third-party reports have suggested it’s closer to 42.8 million. Tens of millions more Americans are trapped under a broadband monopoly, a tally experts say is also undercounted by the FCC.

At the heart of the problem sits the data that internet service providers (ISPs) are required to submit to the FCC. The agency’s methodology permits an ISP to count an entire census block—which can encompass hundreds of square miles in rural areas—as “served” if it is capable of providing broadband service to just one resident of that census block.

There’s no requirement that any home in that census block actually have connectivity—so long as the ISP could provide service, that’s enough for the FCC to count an entire census block as served. Given larger census tracts can include anywhere between 1,000 and 8,000 people, such methodology opens the door to a significant overestimation of broadband deployment.

FCC commissioners have repeatedly admitted that the agency’s data and methodology is flawed. Congress has also repeatedly demanded the FCC do better, recently culminating in the passage of the bipartisan Broadband Data Act, which requires the FCC to collect more granular data on broadband connectivity.

Despite years of criticism, a congressional mandate, and a pandemic showcasing the urgency of bridging the digital divide, the Trump FCC continues to double down on misleading data. The agency kicked off the process for gathering information for its 2021 Broadband Deployment Report in August. The FCC acknowledges that its data is flawed, but proceeds to rely on that data anyway to make broad, inaccurate declarations about the state of the industry.

“More Americans than ever before now have access to the benefits of broadband as the Commission’s policies have created a regulatory environment to stimulate broadband investment and deployment,” the Trump FCC crows in its latest report.

Except industry data shows that to be simply false. Several studies have found that in the wake of the FCC’s hugely unpopular attack on consumer protections like net neutrality, there was no broadband investment spike. With 16.9 million children lacking access to broadband during a health crisis, using bad data to declare “mission accomplished” is policy malpractice.

In addition to relying on flawed data and shoddy methodology, the FCC continues to cling to its aging, cripplingly insufficient 25-Mbps downstream, 3-Mbps upstream definition of “broadband,” helping it further downplay America’s affordable broadband crisis.

Last updated in 2015, this 25/3 definition doesn’t come close to meeting the modern needs of American families in the video-streaming, cloud-storage era. This is especially true during a pandemic that has forced multiple members of a household to work, learn, and play from home over what’s often a single, shared connection.

While the FCC likes to focus on the download speed of 25 Mbps, the upload speed of just 3 Mbps is ridiculously inadequate to handle the kind of high bandwidth applications used for work, school and telehealth. FCC Commissioners and independent researchers alike have advocated for a standard closer to 100 Mbps downstream, 25 Mbps upstream.