Bureaucracy Times - The Red Tape Roundup

Editor’s Note

The past few weeks have been crazy here at Bureaucracy Times. As I write this, I’m hopeful that this edition of The Red Tape Roundup will be published tomorrow. If it does, it will be the first one in the last few weeks.

Two weeks ago, we had our normal meetup where we dug through the data from the previous week, identified our top five most read articles, and meticulously represented our FOIA records to keep you, the reader, up to date on our activities. It’s an accountability that we hold dear, even if we are but a few writers.

But that edition from two weeks ago was never published. And neither was last week’s edition. The reason? A scheduling error.

You see, we are a small operation so we use AI tools to help us with certain tasks, like scheduling. We’ve even disclosed as much on our website. The primary AI bot we use is the WordPress Agent, which is available to any paid WordPress plan. I’d estimate that we lean on the WP Agent for 80% of our non-generative workload for things like scheduling, internal linking, grammar checks, and managing our behemoth collection of categories and tags.

Even though the WordPress Agent is not as seamless as some of the other platforms we’ve used, it does integrate better than any of them with our WordPress website. Sure, it gets our time zone confused every week, defaulting to UTC time rather than our US Eastern timezone. That makes scheduling fun, but that wasn’t the issue that torpedoed our last two newsletters. No, that was something more confusing.

Two weeks ago, after getting all the pieces of our newsletter together, and after passing it around our editorial team for final approval, we handed the product to the WordPress Agent and asked it to schedule it for a Sunday release, as we usually do. The WP Agent acknowledged the task and confirmed that it was scheduled, as it usually does.

But the newsletter never went live. It was never sent to subscribers. It was never scheduled. But why?

We chose to schedule the newsletter for Sundays because that would give us a weekly day off. Saturday’s are fairly light duty, usually involving nothing more than research for the upcoming articles we write. But Sundays are magical days that we use to unplug from our workloads, spend time with our kids, and just do life. Sometimes, we even travel, which was the case two weeks ago.

Because our team is geographically separated, and because we all hold jobs outside of this publication, we often don’t link up again until Monday afternoon. Which is exactly when we noticed that the newsletter was not sent.

I went back to the Agent and asked for the status of the newsletter: “unpublished”.

I asked the Agent why it didn’t publish: “unscheduled”.

I submitted a screenshot of the confirmation that it had been scheduled: “unknown error”.

Wow. Not even WordPress’s own agent could identify the problem. But by then, we’d already missed the publication window, so we let it go and planned to address it during the next newsletter. Which also didn’t launch as it should have, despite our best efforts.

So, what’s the point of recounting this to you, our dear readers? Well, the primary reason is to offer accountability to you, which is something we take seriously as accountability journalists. We wouldn’t have much of a platform to stand on if we professed to hold the federal government accountable but refused to be open and honest about our own shortcomings. We failed to publish, and that’s on us. We could blame AI, but we are the ones who are ultimately responsible for Bureaucracy Times.

Secondarily, we explain all of this to highlight the continued need for human intervention and action in all aspects of our professional lives. It’s easy to justify using automated tools because we’re a small publication; but it’s harder to admit that we had become so complacent using those tools that we didn’t even bother to manually check the status of the newsletter…which is, and always has been, an option.

So, we’re growing up. We’re keeping our heads in the game. We’re not trusting the machines as thoroughly as we have before. Like us, they make mistakes. Like us, they forget to do certain things. And like us, they forget what time it is.

This newsletter will be scheduled by hand, online, on WordPress, like the cavepeople did thousands of years ago. Until next time!

FOIA Watch

Another week of asking federal agencies for records, explanations and occasionally the courtesy of identifying which clock they believe is running. The paperwork continues to move. Whether it is moving toward disclosure remains a separate question.

DOE has acknowledged our nuclear-waste records request. It still hasn’t decided whether the request is properly submitted. On November 3, we received a letter from the Department of Energy concerning FOIA HQ-2026-02565-F, our request for transaction-level records associated with approximately $12.2 billion in Judgment Fund payments arising from the federal government’s failure to fulfill spent nuclear fuel disposal obligations.

The letter, signed by contractor FOIA analyst Mark Vees of Wits Solution Inc., states that DOE received the request September 20 and is still reviewing whether it satisfies the agency’s requirements for a perfected request. DOE also invoked an unusual-circumstances extension, citing records outside the sole jurisdiction of its Office of General Counsel.

That leaves a peculiar procedural question. The agency has acknowledged a request, invoked additional processing time and simultaneously indicated that it has not completed its determination of whether the request is perfected.

On November 5, we sent a clarification seeking the effective receipt and perfection dates, the revised response deadline and identification of any DOE components handling the records.

The request itself remains unchanged. We are seeking the financial trail behind decades of federal nuclear-waste liabilities, not a tutorial in administrative timekeeping. Nevertheless, the tutorial appears to be included at no additional charge.

CMS requested personal information before proceeding with a FOIA. We asked it to identify the request. On November 2, CMS FOIA personnel advised that a pending submission needed a complete mailing address and telephone number before processing could continue.

The response did not clearly identify the underlying case.

On November 5, we replied requesting confirmation that the correspondence would be associated with the original submission, together with its tracking number, perfected status and effective receipt date.

Those details matter because the date on which an agency considers a request perfected can determine when statutory processing deadlines begin. We are not assigning the exchange to a particular CMS investigation until the agency confirms the association.

A request for records should not require an investigation to establish which request the agency is discussing. Yet here we are.

The VA Inspector General appeal remains open. Our October 28 administrative appeal concerning VA OIG FOIA 27-00005-FOIA is awaiting a confirmed acknowledgment or determination.

The original request sought records concerning generative AI use in the Inspector General’s reporting and oversight work. The appeal challenges the adequacy of the agency’s final response.

The underlying question remains whether the office maintains sufficient documentation to demonstrate how AI-assisted material was prepared, reviewed and verified. We are not alleging that its reports are inaccurate. We are asking for the records that would establish the controls used to protect their accuracy.

The FDA has requested more time, and we agreed. On November 5, FDA press specialist April Grant contacted us concerning our inquiry into the agency’s final guidance on prescription drug use-related software.

The original November 4 deadline had passed. Rather than treat the request for additional time as an excuse to begin another round of correspondence, we granted an extension through November 11 at 5 p.m. Eastern.

This is a media inquiry rather than a FOIA request, but it illustrates why we track the two together. An agency that provides a substantive written explanation may eliminate the need for a records request, or at least tell us which records to request.

GAO has provided a useful answer about LNG safety standards. On November 3, Brody Garner, GAO’s Director of Media Relations and Operations, responded to our inquiry concerning recommendation five of GAO-20-619, which addresses technical standards incorporated into federal liquefied natural gas regulations.

GAO has not yet reviewed PHMSA’s October 28 proposed rule. It plans to examine the proposal and update its recommendation tracking as appropriate.

The important clarification is that PHMSA elected to update all eight incorporated standards. GAO indicated that a final rule updating those standards would be needed to close the recommendation on that basis. The proposed rule alone does not complete the work.

GAO also reported that PHMSA had not explained its earlier rulemaking delays to the watchdog.

We sent an acknowledgment November 5. The correspondence is answered, but the regulatory question remains open.

The week’s correspondence brought several investigations closer to a reporting decision. The Alaska Department of Health supplied detailed material concerning Rural Health Transformation Program funding, prompting a three-question reconciliation follow-up November 5. The Department of Labor provided substantive information about the Yenkin-Majestic workplace-safety case, and the Chemical Safety Board answered questions about five outstanding recommendations.

The FDA extension, GAO response and Alaska clarification demonstrate the value of giving agencies a reasonable opportunity to explain the record before resorting to formal disclosure demands.

The coming week brings several important checkpoints. FDA’s revised deadline is November 11. November 12 is the requested response date for numerous newer investigations, including territorial population data, federal election-crime statistics, the Provider Relief Fund audit and Amazon’s Washington, D.C., delivery restrictions. The Millennium Challenge Corporation has until November 13 to address our questions concerning Ghana’s debt-default eligibility history.

The filing cabinets remain undefeated. But at least a few of them have started answering questions.

From the Investigative Desk

Alaska’s Rural Health Money Has Three Different Numbers. We’re Trying to Find Out Why.

A government funding announcement usually arrives with a large dollar figure, a collection of promised benefits and enough enthusiastic language to suggest that the difficult part is already finished.

Our examination of Alaska’s Rural Health Transformation Program began when the numbers refused to cooperate.

The federal government awarded Alaska approximately $272.2 million for the first year of the program. State announcements subsequently described approximately $249 million supporting 254 projects.

Those figures are not necessarily inconsistent. A statewide award can include administrative costs, reserves or funds that have not yet been distributed through individual project awards.

But the difference requires an explanation.

So we began reconstructing the funding process, working backward from the public announcements into the award documentation, project allocations and administrative arrangements.

The investigation quickly acquired a third number.

Materials supplied by the Alaska Department of Health contained a reference to $372,174,855.72, approximately $100 million above the separately identified $272,174,855.72 budget figure.

That does not establish that Alaska received an additional $100 million, that the state overspent its award or that any money is missing.

It establishes that the documentation contains figures requiring reconciliation.

The distinction matters.

An investigative newsroom can produce an alarming headline simply by subtracting one large government number from another. Producing an accurate account requires determining what each number represents before performing the subtraction.

That is where most of this week’s work occurred.

On November 4, the Alaska Department of Health responded to our October 29 inquiry with two attachments and directed us to its public Rural Health Transformation Program materials, including a funding dashboard and first-year award summary.

The response gave us considerably more information than a routine acknowledgment. It also allowed us to narrow the remaining questions.

On November 5, we sent a cordial follow-up to Public Information Officer Mirna Estrada seeking clarification of the competing dollar figures, the legal obligation status of several unsigned awards and related allocation details.

The central issue is no longer whether Alaska announced a large healthcare investment. That is established.

We want to distinguish four stages that government announcements often compress into a single word: funding.

There is the amount the federal government awards to the state. There is the amount the state selects for particular projects. There is the amount legally obligated through executed agreements. And there is the amount actually disbursed.

Those are different accounting events.

A project appearing on a funding dashboard does not necessarily mean the recipient has signed its agreement, incurred allowable expenses or received the money.

Similarly, a federal award to the state does not mean every dollar has already reached a rural hospital, clinic or community organization.

These distinctions are particularly important for a program intended to transform healthcare delivery in geographically isolated communities.

The practical value of an award depends on whether recipients can use the funds, what conditions attach to them and how quickly the money becomes available.

We are also examining the role of the Alaska Community Foundation in administering parts of the program and the relationship between state-level allocations and project-level commitments.

The larger accountability question is straightforward.

Can a reader follow the money from the federal award through the state allocation process to an identifiable project and determine its actual financial status?

We have not established improper spending, inflated awards or mismanagement.

The documentation received this week has improved the reconstruction considerably, and the outstanding questions are narrow enough that we can proceed with an article while continuing to seek clarification.

That is an important editorial milestone.

An investigation does not need to wait indefinitely for every agency to answer every question. It does need to distinguish confirmed facts from unresolved discrepancies and explain what the available records support.

The Alaska article has now been drafted and placed in our publication workflow.

The remaining correspondence may sharpen the accounting, but it should not be allowed to prevent publication of the verified findings already in hand.

There is a lesson in that approach.

Sometimes the government’s numbers are wrong. Sometimes they describe different stages of the same process. Sometimes the explanation is sitting in an attachment that nobody thought to mention in the press release.

Our job is to determine which.

And preferably to accomplish that before the next fiscal year produces another set of numbers.

On Our Radar

A few things that haven’t become stories. Yet.

Ghana’s debt-default problem may have changed categories. We want to know what changed legally. Our examination of a U.S.–Ghana Export-Import Bank debt agreement, TIAS 26-812, has developed into a broader question about eligibility for Millennium Challenge Corporation assistance.

Ghana’s earlier debt-default status raised concerns under Section 7012 restrictions, while its appearance in the FY2027 MCC candidate process suggests that its legal eligibility may have changed.

On November 6, we sent MCC three questions about the transition, the August 12 cutoff used in the candidate determination and whether arrears owed by Ghanaian government entities to American companies affect the analysis.

We are not assuming that the agreement itself resolved every legal obstacle. That is precisely what the correspondence is intended to establish. The reply is requested by November 13.

A strategic-projects agreement with Montenegro is producing questions about who will pay for the projects. A recently published U.S.–Montenegro agreement, TIAS 26-813, provides a framework for cooperation on strategic infrastructure and related development.

The agreement’s existence does not establish that a particular project has received American financing.

That distinction prompted separate November 6 inquiries to the U.S. International Development Finance Corporation, Export-Import Bank and U.S. Trade and Development Agency. We are asking what support each institution may provide, whether specific projects have entered its pipeline and what approval processes would apply.

The responses are due November 12. For now, the agreement establishes a diplomatic framework, not a completed financing transaction. We are watching the space between the two.

The Census Bureau’s territorial population projections have reopened an old statistical problem. Our examination of population changes in U.S. territories has led back to the uneven federal data available for infrastructure planning, healthcare delivery and economic analysis.

On November 6, we sent coordinated inquiries to the Census Bureau, Department of the Interior’s Office of Insular Affairs and GAO. The GAO inquiry specifically references GAO-24-106574, which examined territorial data gaps.

The issue is not simply whether territorial populations are declining. It is whether federal agencies have sufficiently consistent, current and comparable data to plan for the consequences.

Population estimates become infrastructure assumptions. Infrastructure assumptions become budget requests. And budget requests eventually become congressional decisions.

When the underlying data are incomplete, the uncertainty does not disappear merely because a spreadsheet produces a number.

A Provider Relief Fund audit has raised another question about federal payment controls. On November 6, we examined HHS OIG audit A-09-23-01006, concerning Phase 3 of the Provider Relief Fund.

The reconstruction has advanced far enough to support a draft article, while a separate inquiry to HHS OIG has been sent and an HRSA media web form submission has been completed.

The reporting question concerns how federal officials administered and verified pandemic-era relief payments and what the audit reveals about the controls used to distribute public money.

We are keeping the agency’s documented findings separate from any broader conclusion about systemic failure. The November 12 response deadline gives the agencies an opportunity to clarify the record before publication decisions are finalized.

Four developing investigations, four different forms of federal accountability: eligibility for foreign assistance, financing for strategic infrastructure, statistical reliability in the territories and controls over emergency healthcare spending.

None requires inventing a scandal.

All require following the underlying records rather than accepting the headline number or diplomatic announcement as the end of the story.

We’ll keep following the paper.

Someone has to read the footnotes.