Bureaucracy Times

Congress Asked the Pentagon What the War Costs. The Pentagon Didn’t Pick Up.

There is a particular kind of silence that only a cabinet department can produce. It is the silence of an institution that has decided the question does not deserve an answer. That is the silence the Congressional Budget Office ran into this year when it tried to price out Operation Epic Fury, the American military campaign against Iran that began on February 28, 2026. Ranking Budget Committee Member Brendan Boyle had asked CBO in March for a nonpartisan number. CBO went looking for one. The Department of Defense, according to CBO’s own letter, simply did not respond to the agency’s requests for information.

The Silent Treatment

This is worth sitting with. CBO is not a partisan attack shop; it is the body Congress built specifically so that both parties could agree on a baseline set of facts before arguing about what to do with them. When the agency needed to price out five categories of wartime cost, from expended Tomahawks to increased flying hours, it did not get so much as a courtesy briefing from the department actually fighting the war. Instead, CBO had to reconstruct DoD’s own costs using government databases, public reporting, and a private think tank’s homework, principally a CSIS analysis of munitions inventories that CBO cites as its best available source for how many missiles the Pentagon actually fired. An agency of Congress was reduced to citing a nonprofit’s guesswork about the executive branch’s own arsenal, because the executive branch would not tell it.

CBO is admirably restrained about what this means. It notes, in the driest possible register, that its resulting estimate of $38.1 billion “is subject to considerable uncertainty.” That sentence is doing a great deal of work. It is bureaucratic for “we had to estimate a war.”

Four Different Prices for the Same War

Having built its own number from the wreckage of DoD’s silence, CBO then had the pleasure of comparing it against the three numbers the administration offered instead, none of which agree with each other. In June, the Office of Management and Budget sent House Speaker Mike Johnson a formal request for $87.6 billion in supplemental funding, of which the portion CBO judges directly attributable to the conflict comes to $42.3 billion, about 10 percent above CBO’s own figure. Then, on July 22, Secretary of Defense Pete Hegseth testified before the Senate that the war had cost $37.5 billion. Two weeks after that, without public explanation, DoD quietly revised its own munitions figure upward from $25 billion to $27.7 billion in a briefing to committee staff, nudging its total higher still.

So Congress has, at various points this year, been handed four figures for the price of the same war: CBO’s $38.1 billion, the administration’s implied $42.3 billion, Hegseth’s sworn $37.5 billion, and DoD’s own quiet revision on top of that.

Outside government, independent estimators landed in roughly the same neighborhood without any of them matching either. CBO’s letter notes that nongovernmental groups tend to publish still-higher totals, generally because they extrapolate from the war’s expensive early days or fold in costs CBO deliberately excluded, and separate think tank tallies circulating this year ranged anywhere from the low thirties to the high forties of billions, depending on which week’s fighting a given analyst used as a baseline. None of these numbers are wild outliers from one another, which is almost the most unsettling part; they are close enough to look authoritative and different enough to be unreconcilable, produced by an executive branch that cannot seem to settle on its own arithmetic even while testifying under oath about it.

A Receipt With Twelve Billion Dollars Redacted

The gap between CBO’s number and the administration’s supplemental request is not just noise, though. Of the $67.1 billion the administration requested for DoD, CBO could only trace $42.3 billion to activity plausibly connected to Operation Epic Fury. The remaining $24.8 billion sits in categories CBO’s letter describes as largely opaque to it, including $5.1 billion for “cybersecurity and autonomy,” $4 billion for something called the Airborne Moving Target Indicator and Space Data Network Backbone, and $12.1 billion filed simply as “other, classified.” CBO goes out of its way to note that prior wartime supplemental requests, for Iraq and Afghanistan, never buried this large a share of the ask behind a classification stamp. A Congressional Research Service review of the same request separately flagged that OMB’s letter and its enclosures did not provide enough detail even for Congress to determine whether some of this money duplicates the department’s ordinary FY2027 budget request, meaning lawmakers may be asked to fund the same missiles twice under two different labels.

Classified line items are standard practice. A classified third of a wartime supplemental is not, and it lands at the exact moment Congress is least equipped to examine it, because the one agency built to do that could not get DoD on the phone.

What the Ledger Won’t Say

CBO’s letter is, in its understated way, a small monument to institutional failure operating at two levels simultaneously. At the surface level, a department at war would not answer basic cost questions from its own legislature’s scorekeeper. Underneath that, the numbers the administration did produce cannot be reconciled with each other, let alone independently verified, and a meaningful fraction of the request has been placed beyond scrutiny entirely. Congress is being asked to appropriate tens of billions of dollars on the strength of a request that its own budget office had to partially reverse-engineer from a think tank report, while a rotating set of official figures drift a few billion dollars apart from each other every few weeks.

Nobody involved has lied, exactly. Nobody involved has really told the whole truth either, and the agency built to notice the difference had to work the case with the lights half off. That is not a small failure of paperwork. It is what happens when the one branch of government legally entitled to ask “what did this cost” gets treated like a nuisance caller, and the answer it eventually publishes has to carry a footnote admitting it is mostly an educated guess about a war still being fought.

Sources: Congressional Budget Office, Cost of Operation Epic Fury (September 2026); Rep. Boyle’s March 2026 CBO request letter; CSIS, Last Rounds: Status of Key Munitions After the Iran War Ceasefire; OMB June 24, 2026 supplemental request to Speaker Johnson; Congressional Research Service, IN12700. See also: The Pentagon’s Own Gas Station Is Losing Money on Every Gallon
Fediverse reactions
The Federal Budget Works Until 2056, Provided Social Security Keeps Paying Benefits After Its Trust Fund Runs Dry

The Congressional Budget Office has published its latest look at the federal government’s finances over the next three decades.

The numbers do not improve with scrutiny.

Under CBO’s February 2026 long-term projections, federal debt held by the public rises from roughly 101 percent of gross domestic product in 2026 to 120 percent in 2036. By 2056, it reaches 175 percent of GDP.

That is an extraordinary number. Yet buried inside the spreadsheets supporting the projection is another problem that arrives much sooner.

Social Security’s largest trust fund runs out of reserves in 2032.

The combined Social Security trust funds follow in 2033.

And CBO’s long-term federal budget projection generally assumes scheduled Social Security benefits continue to be paid anyway.

That does not make the projection wrong. It does, however, make understanding what CBO is actually projecting rather important.

CBO describes its long-term budget outlook as an illustration of what federal finances and the economy would look like over the next 30 years if current laws generally remained unchanged. It is a baseline, not a prophecy.

The distinction becomes particularly important when Social Security enters the spreadsheet.

The Revenue Line Barely Moves

CBO published a separate workbook containing its 2026 long-term Social Security projections as a supplement to its broader Long-Term Budget Outlook data.

Reading the two together provides a clearer picture than either does alone.

One of the most striking findings is what does not happen.

Social Security tax revenues do not collapse.

In 2026, CBO projects Social Security tax revenues equal to approximately 4.51 percent of GDP.

By 2030, the figure is approximately 4.53 percent.

In 2036 it remains about 4.52 percent.

Even in 2056, Social Security tax revenues are projected at roughly 4.46 percent of GDP.

The revenue line is remarkably stable.

The spending line is not.

Scheduled Social Security outlays equal approximately 5.26 percent of GDP in 2026.

They rise to about 5.61 percent in 2030.

By 2036, that figure reaches 5.91 percent.

By 2056, scheduled outlays remain around 5.79 percent of GDP.

That produces a persistent structural gap between dedicated Social Security tax revenues and scheduled benefits.

In 2026, that difference amounts to approximately 0.75 percent of GDP.

By 2030 it exceeds 1 percent.

In 2036, it reaches approximately 1.38 percent.

The basic problem is therefore not disappearing revenue. It is a system whose scheduled expenditures increasingly exceed the dedicated tax revenues supporting them.

The demographic explanation is not particularly mysterious. In CBO’s 2026 demographic outlook, the agency projects that the population will continue aging while growth slows substantially. Beginning in 2030, CBO expects deaths to exceed births annually. The number and age distribution of Americans directly affect both the workforce paying Social Security taxes and the population receiving retirement benefits.

The economy, meanwhile, continues growing.

CBO’s Social Security tables project nominal GDP increasing from roughly $32.3 trillion in 2026 to approximately $96.5 trillion in 2056.

Taxable payroll rises from about $11.3 trillion to $32.6 trillion.

This is not a projection built around economic collapse.

The economy becomes much larger.

The financing gap survives anyway.

Then the Reserves Disappear

The trust fund numbers make the timeline considerably less abstract.

CBO projects the Old-Age and Survivors Insurance trust fund ratio at 1.53 in 2026. It falls to 1.07 in 2028, 0.60 in 2030, 0.36 in 2031 and 0.12 in 2032.

Then the number disappears.

CBO projects that the OASI trust fund will be exhausted during fiscal year 2032. When the Disability Insurance trust fund is included, combined Social Security reserves are projected to be exhausted in 2033.

Those findings are also laid out in CBO’s March 2026 testimony to the Senate Budget Committee on Social Security’s finances.

Trust fund exhaustion does not mean Social Security suddenly has no money.

Payroll taxes and other Social Security revenues continue arriving after the reserves disappear. The problem is that those revenues would not be sufficient to pay all scheduled benefits.

CBO’s supplemental spreadsheet shows just how large that difference could become.

If benefits after trust fund depletion were limited to amounts payable from available Social Security revenues, combined OASDI benefits would be approximately 23 percent below scheduled benefits in 2034.

The reduction reaches roughly 24 percent during several subsequent years.

The situation is even sharper when OASI is examined independently. CBO’s tables show payable benefits approximately 26 percent below scheduled benefits in 2033, 28 percent below in 2034 and 2035, and 29 percent below in 2036 and 2037.

Those are not predictions that Congress will cut Social Security benefits by those amounts.

They illustrate what the financing mechanism can support if scheduled benefits are compared with revenues available after reserve depletion. CBO’s Social Security testimony specifically describes this as a payable-benefits scenario and emphasizes the uncertainty surrounding long-term projections.

And that distinction leads directly into one of the stranger features of the federal government’s long-term budget projections.

Two Futures Inside the Same Spreadsheet

For purposes of its federal budget baseline, CBO generally assumes that Social Security will continue paying scheduled benefits.

Its Social Security analysis simultaneously calculates what benefits would look like if payments were instead limited by available program revenues after trust fund exhaustion.

Both calculations are useful.

They are also very different futures.

In one, scheduled benefits continue after the reserves financing the existing gap have disappeared.

In the other, payable benefits fall substantially below scheduled benefits.

Congress does not have to accept either outcome untouched. Lawmakers could change taxes, alter benefits, transfer money, restructure portions of the program, adopt some combination of those approaches, or pursue other legislative changes.

CBO is not predicting what Congress will do.

It is displaying the arithmetic of what happens under specified assumptions.

That distinction becomes increasingly important because Social Security is only one part of the government’s larger fiscal problem.

Interest Eventually Costs More Than Social Security

According to CBO’s 2026 Budget and Economic Outlook, federal debt held by the public reaches approximately 120 percent of GDP in 2036.

The annual federal deficit reaches roughly $3.1 trillion that year, equal to about 6.7 percent of GDP.

Then the interest bill becomes a problem of its own.

By 2056, CBO projects net federal interest spending equal to approximately 6.9 percent of GDP.

Social Security spending is projected at roughly 6 percent.

Put differently, under CBO’s long-term baseline, the federal government eventually spends a larger share of the entire national economy servicing its debt than it spends on Social Security.

Interest does not provide retirement benefits, build roads, purchase aircraft, inspect food, or prosecute crimes.

It services previous borrowing.

And as debt accumulates, interest becomes another source of deficits, which can require additional borrowing, which produces additional interest.

By 2056, debt held by the public reaches 175 percent of GDP.

Yet the most consequential decisions may arrive more than two decades before that endpoint.

The Fiscal Collision Arrives Earlier Than 2056

The timelines from CBO’s two workbooks can be placed beside each other.

Social Security already spends more than its dedicated tax revenues in 2026.

The OASI trust fund is projected to exhaust its reserves in 2032.

Combined Social Security trust fund reserves are projected to be exhausted in 2033.

By 2034, payable combined benefits would be roughly 23 percent below scheduled benefits if payments were limited by available Social Security financing.

By 2036, federal debt reaches approximately 120 percent of GDP while the annual deficit reaches 6.7 percent.

Then the debt and interest burden continue climbing for another 20 years.

The Social Security financing decision does not wait for the broader debt problem to arrive.

It gets there first.

That makes the headline number of 175 percent debt-to-GDP somewhat deceptive as a measure of when the problem becomes important.

Washington does not have 30 years before it encounters the arithmetic.

According to CBO’s own projections, one of the largest federal programs reaches its financing crossroads within roughly seven years.

And Then Reality Changed Five Days Before Publication

There is one final wrinkle in CBO’s 30-year outlook that demonstrates just how carefully these projections should be interpreted.

CBO released its long-term outlook data on February 25, 2026.

The agency included an unusual but important qualification.

Its projections did not incorporate the budgetary or economic effects of the Supreme Court’s February 20 ruling concerning tariffs.

Five days.

CBO produced a federal fiscal projection extending three decades into the future while acknowledging that a major event occurring five days before publication had already changed part of the landscape underlying the baseline.

That is not a criticism of CBO. Producing long-term economic projections requires freezing assumptions at some point. Otherwise, the spreadsheet would never leave the building.

It is, however, an excellent demonstration of what a baseline actually represents.

And we no longer have to speculate about whether the omitted tariff developments were financially meaningful.

In an August update to its tariff projections, CBO estimated that trade-policy changes through July 31 would increase cumulative federal deficits during fiscal years 2027 through 2036 by approximately $900 billion compared with the agency’s February baseline.

About $700 billion of that difference comes from larger primary deficits. Another $200 billion comes from additional debt-service costs.

CBO said the increase was largely driven by the removal of tariffs imposed under the International Emergency Economic Powers Act following the Supreme Court decision. The administration subsequently imposed other tariffs, but CBO projected that those would generate less revenue than the tariffs removed after the ruling.

That $900 billion should not simply be pasted onto CBO’s 2056 debt number. Tariff policy affects revenue, prices, trade, economic activity and debt-service costs, and the policies themselves have changed repeatedly.

But the episode provides a useful reminder.

Thirty years is a very long time in federal budgeting.

Apparently, five days can be enough.

The Spreadsheet Is Not the Future

None of CBO’s projections establish that federal debt will equal exactly 175 percent of GDP in 2056.

CBO itself warns that its baseline projections are subject to considerable uncertainty and that future legislation, administrative actions, judicial decisions and economic developments could produce markedly different results.

They do not establish that Congress will allow scheduled Social Security benefits to exceed dedicated financing indefinitely.

Nor do they establish that lawmakers will permit payable benefits to fall roughly one-quarter below scheduled amounts after trust fund depletion.

Those are policy questions that remain unresolved.

What CBO’s spreadsheets do establish is the scale of the arithmetic confronting whoever eventually has to resolve them.

Social Security’s dedicated revenue remains relatively stable while scheduled expenditures exceed it. Trust fund reserves absorb the difference for a while. Those reserves eventually disappear.

At roughly the same time, the broader federal government is running persistent deficits, accumulating more debt and devoting an increasingly large share of national economic output simply to paying interest.

CBO can project that process all the way to 2056.

Congress does not have the luxury of waiting that long to decide what happens next.

Sources: CBO, The 2026 Long-Term Budget Outlook (February 2026); CBO, The Budget and Economic Outlook: 2026 to 2036; CBO, The Demographic Outlook: 2026 to 2056; CBO, Social Security’s Finances: March 2026 Senate Budget Committee testimony; CBO, August 2026 update on tariff revenue effects.
Fediverse reactions
DOJ Warns About Ubiquitous Surveillance. We’d Tell You What Table 2 Says, But It’s Redacted

The Justice Department has a ubiquitous technical surveillance problem.

It does not have a department-wide definition for that problem. Until recently, it had no coordinated strategy for managing it. No individual or office had explicit responsibility for coordinating the response. And when auditors went looking for employee training specifically addressing the threat, they found none.

Then there is Table 2.

We would tell you what Table 2 says, but the Justice Department has blacked out essentially the entire thing.

That would ordinarily be little more than an irritating footnote in the world of federal records. In this particular report, however, the black boxes inadvertently provide a rather good demonstration of exactly what the Inspector General is warning about.

Bureaucracy Times received the report the decidedly unglamorous way investigative journalism occasionally happens. We entered an email address on a government website and subscribed to updates from the Department of Justice Office of Inspector General.

No classified records request. No FOIA lawsuit. No source slipping an envelope under the door.

We asked the government to send us its newsletter.

And the government sent us a report containing a table it apparently could not show us.

The digital trail everyone leaves behind

The September 2026 report, Audit of the Department of Justice’s Efforts to Mitigate the Effects of Ubiquitous Technical Surveillance, examines what DOJ calls UTS.

The concept is remarkably broad.

According to the OIG, ordinary use of technology and online services creates digital trails and discoverable connections. UTS involves capturing and analyzing that information to connect people with things, events or locations. The FBI divides the threat into five vectors: visual and physical surveillance, electronic signals, financial information, travel information and online activity.

A cellphone can reveal location information. Financial transactions contain identifiers. Hotel stays and airline reservations create records. Browsing and social-media activity generate advertising data.

Individually, those records may appear mundane.

Combined, they can tell a story.

Sometimes a deadly one.

The OIG’s previous 2025 audit of the FBI documented a drug-cartel investigation in which the cartel hired a hacker who exploited mobile phones and other electronic devices to obtain call and geolocation information. According to the Inspector General, that information was used to intimidate and, in some cases, kill potential sources or cooperating witnesses.

In another organized-crime investigation, call logs on an employee’s cellphone exposed communications with a known FBI-issued phone. The FBI lost the source.

These aren’t hypothetical privacy concerns dreamed up for the annual cybersecurity PowerPoint.

People can die.

Investigations can be compromised.

Sources can disappear.

And the OIG concluded that the problem extends well beyond the FBI.

DOJ had a threat without a strategy

Following its FBI audit, the Inspector General expanded its examination to the rest of the Justice Department.

What it found was not reassuring.

The auditors concluded that, outside the FBI, DOJ had not taken sufficient steps to manage the UTS threat. The deficiencies, according to the OIG, increased risks to Department employees, investigations, operations, sources and witnesses.

DOJ didn’t even have a common department-wide definition of ubiquitous technical surveillance.

The OIG found no individual or office explicitly responsible for coordinating DOJ’s response. There was no policy or guidance establishing how components should coordinate on UTS. And there was no single mechanism ensuring that information about emerging threats and mitigation practices moved throughout the Department.

Of 16 DOJ components and units interviewed outside the FBI, the auditors found that only the Drug Enforcement Administration’s Defensive Counterintelligence Unit had developed its own definition of UTS.

The DEA defined it as the widespread and continuous collection of data through technologies, often involving ordinary devices and activities, that can be connected back to an individual.

The FBI, meanwhile, was considerably further along.

Following the earlier OIG audit, the bureau designated UTS a “Tier 1 Enterprise Risk,” conducted an FBI-wide assessment, began developing a mitigation plan and training program, and continued working on an enterprise-wide UTS strategy.

The rest of DOJ was playing catch-up.

Forty-six courses. Zero dedicated to UTS.

Perhaps the cleanest measurement of that gap appears in the Department’s training records.

DOJ identified 46 training courses offered during fiscal years 2024 and 2025 that it believed might be related to ubiquitous technical surveillance.

The Inspector General checked.

None was UTS-specific.

Some courses touched subjects related to the threat, but according to the auditors, none actually described the UTS threat, explained how it might be exploited to DOJ’s advantage or taught personnel how to protect the Department from it.

Only seven of the 46 courses were mandatory.

Participation in the remaining voluntary courses among non-FBI DOJ personnel was less than 1 percent.

The mandatory offerings included general DOJ information-technology training, DEA’s Foreign Travel Briefing, ATF’s Counterintelligence Awareness and Security Training, and U.S. Marshals Service courses covering operations security.

The OIG concluded that DOJ had neither baseline UTS awareness training for its workforce nor specialized UTS training for employees working in particularly sensitive areas such as surveillance, source handling and counterintelligence.

There is a substantial difference between not recognizing a theoretical risk and failing to organize around a threat that has already materialized.

DOJ was dealing with the latter.

Immigration judges were already being targeted

During the current audit, DOJ told investigators about UTS-related incidents involving immigration judges overseeing cases handled by the Executive Office for Immigration Review.

Individuals had gathered publicly available but sensitive information about judges from disparate sources and published it in ways that could be used for intimidation or other purposes.

The response required coordination among DOJ security personnel, the U.S. Marshals Service and, sometimes, the FBI.

EOIR also employed a privacy and data-removal service to try to remove judges’ personal information from the internet.

That episode is particularly important because it demonstrates the defining characteristic of UTS.

The information does not necessarily have to begin as secret.

The danger comes from aggregation.

Take pieces of information that are individually public or innocuous, connect them, and something considerably more sensitive can emerge.

Which brings us back to Table 2.

About that table

On page three of the publicly released report sits a table.

Its title is redacted.

Its column headings are redacted.

Its row labels are redacted.

Its substantive entries are redacted.

Even the source printed beneath the table is redacted.

What remains visible is essentially its skeleton: four principal rows, two columns and a collection of bullet points hidden beneath black rectangles.

The OIG explains on the cover of the report that the complete version contains information the Justice Management Division considered law-enforcement sensitive and that those portions were removed to produce the public version.

There is nothing inherently scandalous about that.

Law-enforcement agencies possess information that shouldn’t be publicly released, and Inspector General reports routinely undergo redaction for legitimate reasons.

But the placement of this particular table was irresistible.

Immediately before Table 2, the OIG explains that it interviewed personnel from 12 non-law-enforcement components and subcomponents and four law-enforcement components, encompassing another 32 subcomponents. The auditors found a general lack of awareness and understanding of UTS across DOJ, although law-enforcement components were generally more aware of the threat than their non-law-enforcement counterparts.

Then comes the blacked-out table.

Immediately afterward, the report concludes that DOJ had taken no enterprise-level action to coordinate its components or actively manage department-wide vulnerabilities.

Naturally, we wanted to know what was underneath all that black ink.

So we tried to reconstruct it.

We failed

There are important distinctions here.

Bureaucracy Times did not attempt to remove the PDF’s redactions, exploit hidden document layers, recover deleted text or otherwise circumvent the government’s decision to withhold the material.

Instead, we tried something considerably less exciting.

We searched publicly available government records.

The theory was simple and, appropriately enough, almost identical to the threat described in the report itself: perhaps information appearing separately in other public DOJ, FBI, DEA or OIG records could be combined to independently establish what Table 2 contained.

We got surprisingly far.

The public report identifies the organizations auditors interviewed, including ATF, the Bureau of Prisons, DEA, U.S. Marshals Service, National Security Division, Criminal Division, Executive Office for United States Attorneys, Organized Crime Drug Enforcement Task Forces, Civil Division and EOIR, among others. The FBI was excluded from this particular audit because the OIG had already examined it separately.

We could establish that DEA’s Defensive Counterintelligence Unit had developed a UTS definition.

We could establish that EOIR had experienced the immigration-judge incidents.

We could establish that the Marshals Service participated in responses to those incidents.

We could establish that DEA, ATF and USMS offered training touching areas relevant to UTS.

What we could not establish was which organizations corresponded to the four visible rows of Table 2 or precisely what its individual bullet points said.

There were plausible candidates.

Plausible isn’t proof.

So we stopped.

The black boxes remain black.

Which rather proves the point

There is something wonderfully circular about the exercise.

DOJ’s Inspector General published a report explaining that seemingly ordinary pieces of information can be collected from disparate sources, connected together and used to reveal relationships that aren’t obvious when each piece is viewed independently.

DOJ then redacted information from that report because it was considered law-enforcement sensitive.

We received the redacted report because we had subscribed to a public government email list.

Then we attempted to combine other publicly available government records to determine whether the missing information could independently be reconstructed.

And we discovered that the public record gets you surprisingly far.

Just not far enough to responsibly fill in Table 2.

That may be the most appropriate ending imaginable for an Inspector General report about how much information can be reconstructed from seemingly ordinary pieces of data.

DOJ agrees changes are needed

The Justice Department isn’t contesting the Inspector General’s central findings.

The Justice Management Division’s Security and Emergency Planning Staff concurred with all three recommendations.

DOJ says it will develop a coordinated UTS strategy and governance framework, establish a structured process for identifying and monitoring risks, maintain a UTS risk portfolio and implement tiered training.

The Inspector General lists the recommendations as “resolved.”

That does not mean they’re finished.

The first recommendation will remain open until the OIG receives evidence that DOJ has actually established the coordinated strategy and assigned responsibility for managing its UTS risk portfolio. The second requires evidence that risks have been identified and mitigation plans developed.

The third will not close until DOJ demonstrates that it has created and implemented mandatory basic UTS awareness training across the Department, along with advanced training for employees occupying higher-risk positions.

That gives the Inspector General something concrete to measure later.

It gives us something to watch, too.

And if the Department eventually publishes a follow-up explaining how everything turned out, Bureaucracy Times is already subscribed to the newsletter.

We’ll check Table 2 first.
Fediverse reactions
IRS Warns About Fake ‘Tribal Tax Credits.’ The Public Record Already Names the Promoters.

The IRS says the credits do not exist and warns taxpayers about unnamed “promoters.” Congressional investigators and SEC filings have been considerably less shy.

The Internal Revenue Service has a warning for taxpayers: Someone is selling fake “Tribal Tax Credits.”

The agency’s September 18 alert says promoters are marketing nonexistent federal tax credits under names including “Tribal Tax Credits,” “Native American Tax Credits” and “Sovereign Tribal Tax Credits.” The IRS says no such federal credits exist and warns that taxpayers who claim them could face penalties, interest and even criminal consequences. IRS warning.

There is just one conspicuous omission from the IRS warning.

Names.

The IRS repeatedly refers to “promoters” without identifying the companies or individuals involved.

Fortunately, other parts of the federal government have been less mysterious.

In April 2025, Senate Finance Committee investigators publicly identified White River Energy Corp., Nepsis Inc., Nepsis Tax Advisors LLP, Quartermaster Tax, E3 Family Office, Lifetime Advisors and Lifetime Navigators in connection with the promotion or distribution of purported tribal tax credits.

The committee said the IRS had already confirmed to its investigators that the credits did not exist. Senate Finance Committee.

That makes the IRS’s decision to issue an anonymous scam warning more than a little curious.

The Numbers Are Not Small

White River’s own filings with the Securities and Exchange Commission provide a window into just how ambitious the operation became.

In December 2023, White River told investors that it had received another $5 billion in federal income tax credits supposedly issued to a Native American tribe. The company claimed the credits had been placed under White River’s name and taxpayer identification number at the U.S. Treasury. SEC filing.

By February 2024, White River reported an agreement contemplating the acquisition of another $13 billion in purported credits in exchange for $3 million in cash and 10 million restricted shares of White River stock. The company said it had already paid a $1.25 million deposit. SEC filing.

Then came the sales operation.

In April 2024, White River announced plans to sell $5 billion of purported credits in three tranches. The first $2.5 billion would sell for 50 cents on the dollar. Another $1.5 billion would go for 55 cents. The final billion would cost 60 cents on the dollar.

White River also said it possessed a legal opinion from an unidentified law firm supporting the validity of the credits. The company deemed the opinion a “trade secret” and said it would disclose it only to certain prospective purchasers and representatives who signed nondisclosure agreements. SEC filing.

The IRS now says references to unverifiable legal opinions and demands for nondisclosure agreements are themselves warning signs of the scheme. IRS warning.

The People Behind the Sales

White River’s SEC filings also identify executives who stood to benefit directly from selling the credits.

In April 2024, the company’s board increased annual base salaries for CEO Randy May and CFO Jay Puchir to $1 million each. Both executives also became eligible for royalties equal to 1 percent of gross proceeds from tax-credit sales. SEC filing.

Six months later, White River doubled those royalties to 2 percent each. Directors James Cahill, Danny Hames and Greg Landis were each granted royalties equal to 0.33 percent of gross proceeds from future sales. SEC filing.

The distribution network extended beyond White River.

Senate Finance investigators said a whistleblower identified Lifetime founder Jon LaCasse and his daughter, Payton LaCasse, as personally pitching the credits. Investigators also independently confirmed that Lifetime’s website hosted Nepsis promotional materials concerning the credits. Senate Finance Committee.

The Senate investigation later obtained a recording of a call between Puchir and more than 100 investors. According to the committee, White River could not provide investors with a definitive government document or government contact capable of authenticating the credits. Senate Finance Committee.

By January 2026, Senate Finance Committee Ranking Member Ron Wyden said White River was under a DOJ criminal tax investigation involving its sales of the purported credits. The committee also accused the company of failing to disclose that investigation in required SEC reports. Those are congressional allegations; a criminal investigation itself is not a finding of guilt. Senate Finance Committee.

So Why Doesn’t the IRS Name Them?

That brings us back to September 2026.

The IRS warns taxpayers that promoters may claim secret government agreements exist. It warns about credits sold at steep discounts. It warns about unverifiable legal opinions. It warns about nondisclosure agreements. And it warns that previously accepted tax returns do not prove the credits were legitimate.

The agency even provides instructions for reporting abusive promoters.

What it doesn’t provide is the identity of any promoter.

That might be understandable if federal investigators were still trying to determine who was selling these products. But the public record tells a different story.

Congress publicly identified companies connected to the promotion of the credits in April 2025. SEC filings identify White River executives and directors who received compensation tied to their sale. Congressional investigators identified individual sales activity. And by January 2026, a Senate committee was publicly discussing a Justice Department criminal investigation.

Yet when the IRS issued its nationwide scam warning eight months later, the people it warned Americans about became simply “promoters.”

There may be a legal, investigative or procedural reason the IRS declined to identify them. The agency’s alert does not explain one.

The IRS has now told taxpayers what to look for.

The public record tells them where some of those warning signs have already appeared.

Bureaucracy Times will continue examining that record.
Fediverse reactions
The Postal Service Knew. It Filed a Meeting Agenda Instead.

The United States Postal Service lost $9 billion in fiscal year 2025. Somewhere between $3.6 billion and $5 billion of the surrounding rot, spread across counterfeit shipping labels, counterfeit stamps, and a payment system nobody bothered to fraud-test before launch, was not a mystery to the agency while it happened.

It was, by the Postal Service’s own admission, something people inside the building had flagged, discussed, and in at least one case warned about in writing more than a year before anyone shut anything down. This is not a story about an agency that failed to notice a problem. It is a story about an agency that noticed, formed a working group, and kept mailing.

A Timeline That Reads Like a Confession

Start with the number that should embarrass everyone involved. From March 2024 through February 2026, the Postal Service’s own Office of Inspector General says counterfeit package labels drained $3.1 billion in unrecoverable revenue, with another $520 million projected through 2028. Ninety seven percent of counterfeit labels identified in that window made it all the way to delivery.

The detection system built to stop them, the Counterfeit Postage Intercept Process, caught the labels it was designed to catch in 2023 and then sat still while counterfeiters moved on, the institutional equivalent of reinforcing the door after the windows had been open for a year.

The Enterprise Payment Account fraud embedded in that mess has its own paper trail, and it is the most damning document in the pile. Corporate Treasury warned Postal Service stakeholders about the vulnerability in January 2025. The agency did not officially identify the fraud until May 2025.

The Inspector General discovered it independently that August, during unrelated site visits, which is its own small humiliation; the watchdog found the fire before the fire department did. The agency did not shut down the roughly 2,800 fraudulent accounts responsible until January 2026, a full year after Treasury’s first warning, by which point December alone had bled $125 million in a single month, 14 percent of that month’s entire label revenue.

The Inspector General’s own words on the agency’s response were blunt: “the measures have not been timely or robust enough to counter this fraud scheme.” OIG’s audit traced the root cause to something administratively poetic. Treasury, the office responsible for catching fraud, had no consultation rights and no decision authority over the very payment system fraud was flowing through. The system was never tested for fraud prevention before it went live.

A Machine That Cannot See What It Was Built to See

If the label fraud is a story about warnings ignored, the companion audit on counterfeit stamps is a story about equipment deliberately blinded. The Postal Service’s primary stamp authenticator, the Advanced Facer Canceller System 200, can run in a mode that catches nearly everything or a mode that catches almost nothing, and management chose the second one to save processing time.

When the Inspector General ran its own controlled test, feeding the machine counterfeit stamps in its low-revenue protection setting, the machine caught none of them. Zero. Management’s defense amounted to a cost argument, and a fair one on its own terms, except that nobody then bothered to implement any interim check, such as basic sample testing, to find out how much fake postage was sailing through in the meantime. That gap alone, OIG estimates, put $337 million at risk in a single fiscal year.

The stamp fraud also has an online dimension that makes the agency look almost quaint. A contractor hired to scan the internet found 8,895 instances of counterfeit stamps being marketed or sold in under two months. The Postal Service takes over 30 days to disable an online threat once found; the contractor’s industry benchmark is eight.

The Inspector General points out, not unkindly, that the agency already uses the Lanham Act aggressively to sue people selling unlicensed eagle-logo T-shirts and baseball caps, and applies it far less often to people selling fake postage using the same logo. Priorities, evidently, are a choice.

Ninety Seven Cases, Nine Convictions, One Percent Recovered

Here the story stops being about detection and becomes about consequence, or the near total absence of it. Between 2020 and 2025, the Postal Inspection Service presented 44 counterfeit label cases to the Department of Justice for prosecution.

Nine were accepted. All nine won convictions, a perfect record on the cases federal prosecutors bothered to take, which only sharpens the question of why 35 others were turned away and why the agency still cannot say. The Inspector General recommended a simple fix, a structured data field to track why DOJ accepts or declines a case, so patterns could inform future referrals.

The Postal Service disagreed, arguing the information “is not provided” by DOJ in the first place, an answer that concedes the problem while declining to solve it.

Of the $57 million in losses the Inspection Service actually identified through investigation over that same five year window, it recovered $572,250. About one percent. The delay between a search warrant and a seizure warrant runs four to five months, plenty of time for a suspect whose bank has already tipped him off to walk the money elsewhere.

The agency had two untapped options available to it. One was the Revenue Deficiency Process, an administrative billing mechanism. The other was simply referring cases to state prosecutors. It has used neither with any consistency. When OIG recommended exploring both, management disagreed with that too, and pointed to a prior report about stamps as though it had already answered a different question.

The Working Group That Was Not a Program

None of this happened for lack of meetings. The Postal Service assembled a “Fraud Analytics and Mitigation” team, held regular cross-functional sessions, and produced slide decks. What it did not produce, across two years and two separate audits, was a signed program charter, a clear line of decision authority, or meeting minutes documenting what was decided and why.

Management’s own response to the recommendation that it fix this was that formal governance amounted to unnecessary overhead given how fast fraud moves, an argument that would carry more weight from an agency that had, in fact, moved fast.

The Postal Service agreed with eight of the eleven recommendations in the label audit and rejected three, all in the section on prosecution. It agreed to all four recommendations in the stamps audit and all five in the payment account alert, mostly after the money was already gone.

Every implementation date attached to a real fix lands sometime between April 2026 and August 2027. The Inspector General will keep filing reports. The agency will keep filing agendas.

The $3.62 billion did not wait for either.

Sources: Efforts to Mitigate Counterfeit Postage (USPS OIG, Report 25-072-R26, September 10, 2026); Counterfeit Stamps (USPS OIG, Report 25-121-R26, June 16, 2026); Management Alert: Enterprise Payment Account Fraud (USPS OIG, Report 25-072-2-R26, February 10, 2026)
Fediverse reactions
The Judge Gets Her Sixty Votes, the Country Gets a Quorum Call It Didn’t Need

The Senate spent Tuesday doing what the Senate does best, which is confirming a lifetime appointment by a vote count that would embarrass a homeowners association. By 52 yeas to 43 nays, Senators invoked cloture on the nomination of Angela Veronica Colmenero to be United States District Judge for the Southern District of Texas, which means Colmenero has effectively already won a seat on the federal bench for as long as she cares to keep it, with the formal confirmation vote merely a formality due by Friday. Nobody on the floor bothered to explain why they were voting the way they were. Nobody had to. That is what cloture is for.

The Rest of the Menu

If judicial confirmations are the entree, everything else Tuesday was garnish. Senators also advanced cloture, 70 to 21, on an amendment to the Protect College Sports Act, a bill that exists because Congress has apparently decided that name, image, and likeness litigation is a more pressing use of floor time than anything involving an actual federal agency. Two bills passed by unanimous consent, one modifying the boundary of the Vicksburg National Military Park and one conveying a reversionary land interest in Sacramento; both are the legislative equivalent of clearing out a junk drawer, and neither will trouble a headline writer. The Senate also agreed to a pair of commemorative resolutions, because a body that cannot pass appropriations bills on time can always find the votes to recognize a heritage month.

Buried in the “New Public Laws” notice, easy to miss and clearly meant to be, was Public Law 119-110, better known as S. 307, addressing sexual harassment and sexual assault of Bureau of Prisons staff by the people BOP is paid to incarcerate. It was signed on September 16 and only logged in Tuesday’s Congressional Record, which tells you something about the pace at which Congress documents its own work. That a law protecting federal correctional employees from assault by inmates needed to be written at all is the kind of institutional admission that deserves more than a passing line item, and it is the kind of story this publication intends to come back to once the implementation record exists to judge it against.

The Meeting That Isn’t Explained

The single most interesting sentence in the entire Daily Digest was also the least detailed. The Committee on Homeland Security and Governmental Affairs will hold a business meeting Wednesday “to consider the request to direct Senate Legal Counsel to file an application seeking immunity.” Immunity for whom, in what proceeding, and why the request needed to be phrased with the specificity of a ransom note, the digest does not say. Congress is under no obligation to tell the public what its committees are up to before they do it, and on Wednesday it will not. We will be watching to see who walks out of SD-342 needing a lawyer less than they did walking in.
Fediverse reactions
Federal Register Watch: September 23, 2026
President Trump extended the $100,000 fee on new H-1B visa petitions for another year, and the same day’s docket also revived a shelved drug-scheduling plan, quietly relabeled a foreign-aid advisory council, and locked in a Colorado military land withdrawal for another decade and a half.
- Executive Office of the President — H-1B fee extended, program-integrity order issued alongside it. Proclamation 11069, signed September 18 and published today, extends through September 21, 2027 the rule that H-1B petitions for specialty-occupation workers must be accompanied by a $100,000 payment, with only narrow national-interest exceptions. The administration cited what it called a 92 percent drop in H-1B filings from major IT outsourcing firms since the fee took effect as evidence it worked. It paired the extension with Executive Order 14431, directing agencies to coordinate more closely on policing the program. Renewing the fee for a second year, rather than letting it lapse, is a stronger signal that it is becoming a permanent feature of the visa system rather than a temporary emergency measure.
- HHS — an interim final rule that took effect the moment it was published. The new rule pauses registration for ACA marketplace agents and brokers who lack a current Plan Year 2026 agreement with the federal exchanges, blocking them from registering for Plan Year 2027 until February 1, 2027. HHS invoked authority to skip the advance notice-and-comment process that precedes most rules, meaning the policy bound brokers before anyone outside the agency had a chance to object; the public only gets to comment after the fact, which limits how much the process can still change before it takes hold.
- Interior/BLM — a 15-year renewal of the Army’s hold on Fort Carson and Piñon Canyon. Public Land Order No. 7971 keeps roughly 5,650 acres of public land and more than 141,000 acres of federal mineral rights in Colorado off-limits to the mining, grazing, and homesteading uses normally allowed on public land, reserving them for Army training and live-fire exercises. A “withdrawal” of this kind removes land from the ordinary rules that let private parties stake claims or apply for permits on it; renewing one for 15 years, rather than a shorter term, locks in military control of the long-contested Piñon Canyon Maneuver Site well into the 2040s.
- Millennium Challenge Corporation — a council “reestablished” months after it was renewed. MCC published a notice reestablishing its Advisory Council for the 2026–2028 term, less than three months after a July 2 notice announced that same council’s charter had already been renewed for that identical term. Agencies typically renew an advisory committee’s charter once per term rather than following a renewal with a reestablishment of the same body soon after, so the change in language is the kind of “why now” wrinkle worth a second look. Neither notice explains what changed between July and September, or why “renewal” stopped being the operative word.
- DEA — reviving a scheduling plan it shelved four years ago. The agency proposed placing five tryptamine hallucinogens — 4-OH-DiPT, 5-MeO-AMT, 5-MeO-MiPT, 5-MeO-DET, and DiPT — into Schedule I, the most restrictive drug category, reserved for substances found to have no accepted medical use, that carries the strictest handling requirements and criminal penalties. DEA proposed the same scheduling action back in 2022 but withdrew it to seek updated scientific evaluations from HHS. Its return now, with the same five substances and no public explanation of what new evidence closed that gap, is the sort of resurfacing worth flagging.
- OMB/DOD/GSA/NASA — paperwork notices tied to a “Revolutionary” rewrite of federal contracting rules. Three linked notices (FAR Case 2026-011, 2026-006, and 2026-010) seek public comment on the paperwork burden of a sweeping Federal Acquisition Regulation overhaul proposed on September 18 under Executive Order 14275, spanning twelve coordinated proposed rules that touch contractor qualifications, contract types, patent and data rights, bonding, and dozens of other subjects. The FAR governs how the government spends hundreds of billions of dollars a year on contracts, so a rewrite at this scale reshapes federal procurement well beyond the routine paperwork notices that formally introduced it today.
This edition draws on the full Federal Register docket for September 23, 2026: 98 documents in all. Excluded as routine: roughly two dozen NAGPRA repatriation and inventory-completion notices, about ten SEC self-regulatory-organization rule filings, routine FERC pipeline and hydroelectric project notices and combined filings, antidumping/countervailing-duty determinations, bank holding company formation and control notices, standard Paperwork Reduction Act information-collection notices unrelated to the FAR overhaul, and routine advisory-committee and agency meeting notices.
The Dots Say What Warsh Would Not

A Chairman Who Declines to Own His Forecast

At his press conference on September 16, Fed Chair Kevin Warsh offered the public a tidy piece of humility. The central bank, he said, “cannot affect any individual price,” whether at the gas pump or the grocery checkout, but it could work to keep any change in relative prices from broadening out, from producing “second and third order effects on the economy.”

It is a reasonable description of monetary policy’s actual reach. It is also, conveniently, a description with no falsification clause attached. If prices broaden out anyway, the Committee can say the hike softened what would otherwise have been worse. If they do not, the Committee takes the credit. Either outcome confirms the theory, which should always make you suspicious of the theory.

There is a second detail worth sitting with, buried in the fine print of the Summary of Economic Projections released the same afternoon. Of the 19 seats on the Committee, only 18 participants submitted a projection for 2028 and 2029, and Warsh himself is understood to have again declined to submit his own dot for the federal funds rate, a habit he established at his very first meeting as chair back in June.

A central banker who will tell reporters, at length, what the Committee is trying to accomplish, while declining to put his own number on paper for where rates should go, has found a genuinely elegant way to be accountable for the rhetoric and insulated from the record.

What the Dots Actually Say

The rhetoric was calm. The dots were not. In June, the median FOMC participant projected the federal funds rate would end 2026 at 3.8 percent, consistent with the one quarter point hike that had already been substantially priced in. By September, with that hike delivered and the range sitting at 3.75 to 4 percent, the median projection for year end had moved up again, to 4.1 percent.

Read plainly: the Committee did not simply catch up to where it said it was headed. It moved the target further away, in the same direction, in the same meeting where it congratulated itself for a “timelier return” to 2 percent inflation.

The inflation numbers explain why. Median PCE inflation for 2026 came in at 3.7 percent, versus 2.0 percent, the Committee’s own longstanding target, a gap that would embarrass most institutions into using the word “elevated” more than once in a press release. Core PCE, which strips out food and energy on the theory that doing so reveals the underlying trend, sat at 3.4 percent.

Seventeen of eighteen participants said the risks to inflation were weighted to the upside. Seventeen of eighteen said their uncertainty about inflation was higher than normal. This is not a committee that is confident inflation is cooling. This is a committee that voted to raise rates while telling you, through its own survey data, that it expects to be wrong in the direction of more inflation, not less.

The One Number That Actually Moved in the Fed’s Favor

Credit where it is due. The unemployment rate projection improved, from a June median of 4.3 percent to 4.1 percent in September, and GDP growth ticked up slightly as well. That is the justification buried inside the jargon of “economic activity is expanding at a solid pace.” A labor market that is not cracking gives the Committee room to lean against inflation without immediately worrying about a recession it is causing. It is a coherent policy position.

It is just not the position that got spoken aloud at the podium, where Warsh’s framing stayed centered on prices rather than on the fact that a resilient labor market is precisely what buys the Fed room to keep hiking.

Reading the Distribution Instead of the Median

The median number flatters the Committee with an appearance of consensus that the underlying distribution does not support. Figure 2 in the SEP shows twelve participants clustered around a year end midpoint near 4.125 percent, but four sit higher, near 4.375, meaning four members of the Committee think two more quarter point hikes belong on the table before December, not one.

Two participants sit lower, at 3.875, essentially arguing the September hike should not have happened at all. That is not a unanimous outlook dressed up as a single dot. That is a room with real disagreement about how much further to go, compressed into a median that makes the disagreement disappear.

An institution that wants credit for transparency does not get to publish the distribution and then let the chair speak only in the language of the median. The dots, read honestly, say the Fed expects to keep hiking into a labor market it currently likes and against an inflation number it currently distrusts. Warsh, at the microphone, said something calmer and vaguer, and declined to attach his own name to either number.

The public got the speech. The institution kept the paper trail to itself.

Primary sources: Federal Reserve, September 2026 Summary of Economic Projections; Federal Reserve, September 16, 2026 FOMC statement.
Fediverse reactions
The Ten-Year Plan That Forgot to Have an End Date

In March 2021, the United States Postal Service unveiled Delivering for America, a ten-year plan promising, in essence, that a 246-year-old delivery monopoly could be reengineered into something resembling a modern logistics company, if only it was allowed to consolidate a few hundred facilities, reroute the mail through fewer hubs, and quietly stretch the definition of “on time.”

Five years and roughly $40 billion in projected investment later, the Postal Service’s own Office of Inspector General has produced a report that reads less like a progress update and more like a diagnosis delivered with the studied gentleness reserved for patients who are not going to like the prognosis.

The topline finding, stated with bureaucratic understatement, is this: the Postal Service has cut costs, reorganized its network, and reduced transportation and processing expenses by roughly $2.75 billion combined. Its annual net loss has not moved. It stood at $9.18 billion in FY 2020 and sits at $8.98 billion in FY 2025, a rounding error’s worth of improvement bought with five years of institutional upheaval.

Meanwhile the controllable loss, the metric management prefers because it excludes the pension obligations Congress has never fixed, grew to $2.75 billion. The organization has, in other words, gotten leaner and no less broke.

Rural America Gets the Extra Day

The plan’s central mechanism, the Regional Transportation Optimization initiative, works by no longer whisking mail from small post offices into the processing stream the same day it is collected. Instead, mail from offices more than fifty miles from a processing facility waits, gaining the Postal Service a cheaper truck schedule and gaining the customer an extra day of transit that did not exist before.

When fully deployed, RTO will touch 72 percent of ZIP codes in the contiguous United States, and its distribution is not evenly shared: 86 percent of rural populations are affected against 45 percent of urban ones. Five entire states (Louisiana, South Dakota, Vermont, West Virginia, and Wyoming) will be RTO in full.

The OIG’s examples do the rhetorical work no press release ever will. First-Class Mail crossing Vermont, a state 157 miles long, now takes four days instead of two. A letter from Pierre to Rapid City, South Dakota, 172 miles apart, went from a three-day standard to five.

The report also traces a piece of mail from Providence, Rhode Island, to Akron, Ohio, a route that now detours through Massachusetts and Indianapolis before it can reach Ohio, the postal equivalent of routing a phone call through three area codes to save money on the switchboard. None of this is malfunction. It is the plan working as designed, which is precisely the problem the OIG is pointing at.

The Postal Service says it expected 95 percent on-time delivery under the redesigned network. Performance improved briefly after the changes began, then declined again across every mail class in FY 2024 and stayed down through FY 2025, ticking upward only in the first half of FY 2026.

RTO ZIP codes trail non-RTO ZIP codes by roughly eight percentage points on First-Class Mail, a gap that widened even after management padded the delivery standards in 2025 to make the numbers look better on paper. When the target keeps moving to meet the outcome, the outcome stops meaning very much.

A Ten-Year Plan Without a Program Manager

Here the report shifts from documenting outcomes to documenting something closer to institutional negligence. Delivering for America was sold to Congress, employees, and the public as a structured ten-year plan with $40 billion in investment and $22 billion to $31 billion in projected savings.

In FY 2024, quietly and without much public notice, the Postal Service updated the plan and removed the savings targets and budget goals entirely. Management now describes DFA as a “strategic direction” rather than a plan with numbers attached, which is a remarkably convenient thing to decide once the original numbers stop being achievable.

There is no centralized program office tracking the transformation as a whole, no overarching timeline, and no mechanism to report how much of the ten-year plan has actually been completed.

The OIG had to calculate that figure itself: as of March 2026, fourteen of sixty planned regional processing centers were operating, along with seventy-three of a planned one hundred ninety local processing centers and a hundred fifty-four of four hundred planned sorting and delivery centers. Less than half of the network redesign is finished, five years into a ten-year plan, and nobody at USPS headquarters can currently tell you, in dollars, what finishing the rest will cost.

Management has acknowledged total costs may exceed the original $40 billion estimate. By how much, it will not say, because it no longer tracks the number.

The report is even less forgiving on modeling. USPS built its network redesign, including the Atlanta and Richmond processing centers, on volume projections that ignored trailer counts, loading times, and dock space, resulting in mail trucks idling for up to thirteen hours waiting to unload.

Separately, USPS signed an air cargo contract with volume minimums it then discovered it could not meet with packages alone, forcing it to divert First-Class Mail back onto planes, directly undoing the surface-transport strategy the rest of the plan depends on. An organization that cannot model its own truck bays is now being asked to manage a ten-year, $40 billion transformation on the honor system.

Standardization: The Part of the Plan Nobody Finished

The most damning finding is also the driest. The entire premise of the redesigned network was that new regional processing centers would be functionally interchangeable, standardized boxes that could be measured, compared, and managed from a national dashboard.

Instead, the OIG found that the Atlanta and Indianapolis centers, both purpose-built at a combined cost exceeding $700 million, perform entirely different functions. Atlanta consolidates local mail collection, cancellation, and express mail processing in one building. Indianapolis spreads the same functions across three separate facilities. Table 6 of the report lists fourteen regional centers using five different sorting machine types with no consistent pattern for what gets done where.

The Postal Service built a national network and ended up with fourteen local ones, which is precisely the fragmented, custom-solution architecture Delivering for America was supposed to replace.

Layered on top of that is a persistent scanning problem, flagged across five separate OIG audits, in which manual, error-prone scanning at modernized facilities undermines the very tracking visibility the new network was supposed to deliver. The OIG recommends automation. Management’s response, that full standardization is “impractical” given legacy infrastructure, is the kind of sentence that would have sounded more convincing before the agency spent five years and $40 billion trying to do exactly that.

The Verdict, Such As It Is

To its credit, the Postal Service agreed with all five OIG recommendations, which is the bureaucratic equivalent of nodding solemnly at a doctor while reaching for a cigarette on the way out.

It will build a five-year program schedule by mid-2027.

It will standardize efficiency measurement by 2027.

It will look into automated scanning by 2027.

All of it is achievable, none of it is new, and all of it should have existed before the agency spent a decade’s worth of political capital and public patience rerouting the mail through more hubs than the routes require.

Delivering for America has, by its own auditor’s account, delivered cost savings without delivering solvency, and standardization without delivering standards. The plan has five years left to prove it was worth the wait.

Rural Vermont, waiting an extra two days for a letter to cross a state you can drive in three hours, is not optimistic.

Source: USPS Office of Inspector General, Delivering for America: Five-Year Assessment (Report 26-042-R26, September 2026).
Fediverse reactions
Washington Helped Build Google’s Ad-Tech Empire. Now It’s Trying to Pry It Open.

The Justice Department is celebrating a federal court order designed to loosen Google’s grip on online advertising technology.

There is just one awkward historical detail.

The federal government reviewed and cleared some of the acquisitions that helped Google build that grip in the first place.

On Sept. 16, the Justice Department announced what it called “substantial relief” in its antitrust case against Google. The U.S. District Court for the Eastern District of Virginia ordered Google to make important portions of its advertising technology work more readily with competing systems, allow publishers greater access to their own data, prohibit certain preferential bidding practices and submit to six years of outside monitoring.

It is the latest development in a case that produced a major government victory in April 2025, when U.S. District Judge Leonie Brinkema found that Google had unlawfully maintained monopolies in publisher ad servers and open-web display ad exchanges.

But getting from a collection of advertising businesses to an unlawful monopoly took years.

And Washington was watching for much of the trip.

The Government Approved the Building Blocks

In 2007, Google agreed to buy DoubleClick for $3.1 billion.

DoubleClick was no minor addition. Its technology included a publisher ad server that would eventually become Google’s DFP and technology that developed into Google’s AdX advertising exchange. Those products later became central to the Justice Department’s monopolization case.

The Federal Trade Commission investigated the DoubleClick acquisition for eight months.

Then it let the deal proceed.

The vote was 4-1.

At the time, the FTC concluded that the transaction was unlikely to substantially lessen competition. Regulators even considered the possibility that Google might combine or bundle DoubleClick’s products with its other advertising businesses.

The commission concluded that DoubleClick lacked the market power necessary to make such a strategy a significant competitive threat.

That prediction has not aged particularly gracefully.

Nearly two decades later, a federal court found that Google unlawfully tied its publisher ad server, DFP, to its AdX exchange as part of conduct that helped maintain its monopoly power.

There is an important distinction. The court did not rule that Google’s acquisition of DoubleClick itself was illegal. The government’s eventual case concerned what Google subsequently did with the market position it had assembled.

Still, Washington had examined one of the central building blocks before Google put it into the building.

Then Came More Pieces

Google acquired Invite Media in 2010 for approximately $81 million.

Invite Media provided technology on the advertiser side of the market and ultimately became part of Google’s advertiser-buying infrastructure. The Justice Department’s 2023 complaint would later identify the acquisition as another step in Google’s expansion across the advertising technology stack.

The transaction apparently fell below the threshold requiring federal premerger notification, meaning regulators did not conduct the same kind of review that DoubleClick received.

Then came AdMeld.

Google agreed to purchase the publisher advertising technology company in 2011 for roughly $400 million.

This time, the Justice Department investigated.

The Antitrust Division gathered information from Google, AdMeld and other participants in the advertising industry. It ultimately closed the investigation without challenging the acquisition, concluding that the deal was unlikely to substantially reduce competition.

One of the government’s assumptions was that publishers could use multiple advertising platforms and shift their business among them.

Years later, the Justice Department revisited that conclusion.

Principal Deputy Assistant Attorney General Doha Mekki described the government’s earlier reasoning in a 2023 speech and delivered an unusually concise bureaucratic postmortem.

“We were wrong.”

Google’s growing scale, DOJ subsequently argued, combined with exclusionary practices to undermine precisely the competitive alternatives regulators had expected would remain available.

Washington Comes Back With a Crowbar

By 2023, the Justice Department was telling a very different story about the advertising market.

Its antitrust complaint specifically traced Google’s expansion through DoubleClick, Invite Media and AdMeld, arguing that Google had established control across critical portions of the technology used to buy and sell open-web advertising.

Two years later, Judge Brinkema found Google liable for unlawfully maintaining monopolies in publisher ad servers and ad exchanges.

The government then sought something much more dramatic: structural relief that included forcing Google to divest its AdX exchange.

The court declined to go that far.

Instead, Google received what might best be described as federally mandated interoperability.

Under the final judgment, Google must create and support integrations connecting AdX and DFP with Prebid, an open-source advertising technology widely used by publishers. AdX must also be capable of submitting real-time bids into competing publisher ad servers.

Publishers must be allowed to access and export their data from Google’s systems, making it easier to migrate to competitors.

Google’s advertiser tools are prohibited from bidding preferentially into Google’s own advertising products simply because Google owns them.

And a monitor and technical committee will oversee Google’s compliance for six years.

The government helped open the door.

Now it is installing a monitor to make sure Google leaves it open for everyone else.

A Monopoly on Fixing Monopolies

There is a tempting joke here about the federal government objecting to monopolies while maintaining something of a monopoly on deciding what constitutes one.

Legally, however, the distinction matters.

Federal antitrust law does not prohibit a company simply because it becomes enormously successful or even obtains monopoly power. The Sherman Act targets acquiring or maintaining that power through unlawful exclusionary conduct.

Nor did the court retroactively declare Google’s major acquisitions illegal.

The more interesting story is institutional.

Federal regulators examined Google’s $3.1 billion DoubleClick acquisition and approved it. DOJ examined the roughly $400 million AdMeld acquisition and approved that too. Invite Media slipped beneath the federal premerger-review threshold.

Those three acquisitions subsequently appeared in the Justice Department’s own reconstruction of how Google built its position across the advertising technology stack.

The FTC made predictions about the competitive consequences of DoubleClick.

DOJ made predictions about AdMeld.

The market developed.

Google developed with it.

Eventually, DOJ acknowledged that at least one of those predictions had been wrong.

Now, almost two decades after the DoubleClick acquisition, the federal government has won a court order requiring Google to connect the resulting advertising infrastructure to competitors, share information with customers and operate under years of technical supervision.

The Justice Department calls that restoring competition.

It may also be something considerably more familiar in Washington.

Fixing yesterday’s regulatory assumptions with tomorrow’s regulatory oversight.

<strong>Sources:</strong> <a href=”https://www.justice.gov/opa/pr/department-justice-again-wins-substantial-relief-against-google” target=”_blank” rel=”noopener”>Department of Justice, September 16, 2026 press release on Google ad tech remedies</a>; <a href=”https://www.justice.gov/atr/case/us-and-plaintiff-states-v-google-llc-2023″ target=”_blank” rel=”noopener”>U.S. and Plaintiff States v. Google LLC, DOJ Antitrust Division case page</a>; <a href=”https://docs.justia.com/cases/federal/district-courts/virginia/vaedce/1%3A2023cv00108/533508/1410″ target=”_blank” rel=”noopener”>U.S. District Court for the Eastern District of Virginia, April 17, 2025 liability opinion</a>; <a href=”https://www.justice.gov/atr/case-document/file/1566706/dl” target=”_blank” rel=”noopener”>DOJ complaint, January 24, 2023 (PDF)</a>; <a href=”https://www.ftc.gov/news-events/news/press-releases/2007/12/federal-trade-commission-closes-googledoubleclick-investigation” target=”_blank” rel=”noopener”>FTC press release closing the Google/DoubleClick investigation, December 20, 2007</a>.
Fediverse reactions