The United States scored 64 on the Corruption Perceptions Index, the lowest it has ever recorded. That number gets quoted as a verdict on the present. It is not one, and the reason is in the methodology rather than the politics.
Each of the index's thirteen sources reflects perceptions recorded over the two years before publication. The 2025 index was published in February 2026.
So the 64 was built from assessments gathered between roughly early 2024 and late 2025.
Almost nothing from 2026 is in it.
The second instrument reports ranges, not amounts
The obvious correction is to stop asking what experts perceive and read the filings instead. Disclosure is a legal obligation. It should produce facts.
It produces something softer than that.
The federal public disclosure form asks filers to place each asset in a value band. Not to state an amount. And whole categories sit outside it. Cash accounts are not reportable. Treasury securities are not reportable. Holdings inside retirement accounts are not reportable. Assets held within an excepted investment fund are not itemized at all.
So a filing can be complete, accurate and filed on time. And still not tell you what someone is worth, or where it came from.
Which makes the billions a reconstruction
This is the part that cuts against the argument rather than for it, and it belongs here for exactly that reason.
The figures in circulation are large and specific. Roughly $2.4 billion from crypto since 2024. Family crypto holdings up about $7.4 billion since the return to office. More than $1.4 billion in the year ending 30 June 2026.
Those did not come off a form.
They are reconstructions, assembled from token prices, public blockchain activity, company statements and reporting. That work can be good. It is still estimating assets whose prices move daily, through entities whose ownership is layered. And in this market a quoted valuation and a realizable one are different things.
A number can be the best available and still carry an error bar nobody prints.
If this dispatch demands skepticism about a perception index, it owes the same skepticism to the figures that happen to support the uncomfortable reading. The honest position is that both are estimates and neither is an audit.
The third instrument was switched off
There is one mechanism designed to convert an allegation into a finding, and it is not a survey or a form. It is a prosecution.
The Justice Department's Public Integrity Section had roughly 40 full time staff at the start of 2025. It is now reported at two full time attorneys. Other counts put it at four to six. Its open matters went from between 175 and 200 down to about twenty.
Whatever the exact staffing, the direction is not in dispute.
| Instrument | What it is supposed to show | Why it cannot, right now |
|---|---|---|
| Perception index | Expert judgment of public sector corruption | Built from data ending in 2025. Measures belief, not conduct. Not comparable before 2012. |
| Financial disclosure | What an official owns and earns | Value bands rather than amounts, with whole asset classes exempt. |
| Prosecution | Whether conduct broke the law | The unit went from about 40 staff to 2, and from 175 open matters to about 20. |
| Public opinion | What citizens believe | Belief is not evidence. And 21 percent answered that they do not know. |
The fourth instrument measures the wrong thing on purpose
An Associated Press and NORC survey taken in late September found something stark. Forty-one percent of Americans believe the President has done something illegal to profit from the office. Another 18 percent said unethical but not illegal. Eighteen percent said nothing wrong.
Call it 59 percent, which is the figure now being quoted.
It is a real measurement and it measures belief. That is all a poll can do. A majority once believed things that were false, and a majority has often been right, and the poll cannot tell you which case this is.
Notice the number almost nobody repeats. Twenty-one percent said they did not know.
What is actually established
Strip out everything that is estimate, perception or inference, and a smaller set of facts remains.
A senior United Arab Emirates government official acquired a $500 million stake in World Liberty Financial. The President's family co-owns it. Months later the administration agreed to supply the UAE with America's most advanced AI chips.
That is a documented sequence. A sequence is not a transaction, and saying so is not a defense of it. It is a description of what the record supports, and what it does not.
A former White House ethics lawyer, Richard Painter, has said that for any other executive branch official these arrangements would be a violation. That is informed opinion from a qualified source. It is not a finding either.
The scarcinality reading
This site exists to ask one question of any shortage. Is the thing genuinely absent, or is it present and out of reach?
Apply it to evidence.
The conduct happened or it did not. That fact exists. What has gone missing is the capacity to establish which. And that capacity was not lost to accident or to cost. Staff were removed and cases were closed.
It is also, in this framework's terms, a manufactured scarcity of exactly the kind the treatise describes. The evidence is not gone. The means of reaching it were dismantled.
And the consequence runs in both directions, which is worth saying plainly. A person who is innocent of these allegations has also lost the mechanism that would demonstrate it. Nobody gets cleared by a unit with two attorneys either.
What would settle this
Three things, and they are dated rather than rhetorical.
The Corruption Perceptions Index for 2026 publishes around February 2027. It will be the first score built on evidence from this year. If it falls materially below 64, expert assessment has absorbed what the polls already show. If it holds, the perception story is narrower than it looks today.
A restored Public Integrity Section would settle more than any index. Staffing is public and countable. A return toward forty would be a genuine change rather than a signal.
And a disclosure regime reporting amounts rather than bands would end the reconstruction problem entirely. That is a legislative question, not an investigative one. It is also the cheapest of the three.
Two cautions. The scoring window in the figure is inferred from Transparency International's published method, not from a statement about which months were used. Treat its edges as approximate. And this dispatch deliberately reaches no conclusion about whether laws were broken. The instruments that would support such a conclusion are the subject of the piece.