Method
One structure, applied to every topic.
Most misleading claims contain no false figures. They use a real number on a basis nobody states.
The mechanisms
Long-horizon total
2 published topicsA commitment running for decades is summed into one number and set beside figures that cover a single year.
Over how many years? What is it per year?
Cash, real or discounted
2 published topicsThe same commitment has three legitimate totals — cash as it is paid, restated in today’s prices, and discounted to present value. Each answers a different question, and the gap between them can be an order of magnitude.
Which price basis, and at what discount rate?
Missing baseline
3 published topicsA cost or a change is quoted with no statement of what the alternative was. Without the counterfactual, the number has no meaning.
Compared with what? What was already being spent?
Denominator swap
1 published topicsA share is quoted without saying a share of what, or the denominator changes between the two halves of a comparison.
Per what? Is the bottom of the fraction the same in both figures?
Gross, not net
0 published topicsA headline saving or cost is the theoretical ceiling before offsets, behavioural response, or legal constraint.
What comes off this before it reaches the bottom line?
Definition change
2 published topicsA series is quoted across years in which the thing being counted, or the group it is counted over, changed.
Is it the same measure at both ends?
Chosen window
2 published topicsThe start and end years are picked so the trend runs the required way. A different pair of years reverses it.
Why these years? What does the full series look like?
Out-of-date comparison
2 published topicsA comparison that was accurate when published, quoted years later against a figure that has since moved.
How old is each half of this comparison?
Mis-ranked magnitude
2 published topicsA small item is placed among the largest, or a large one buried, without the rest of the list being shown.
Where does it actually sit in the full ranking?
Stock read as flow
1 published topicsHow many are here and how many arrived this year are different quantities, routinely swapped for one another.
Is this a level or a rate of change?
Perception gap
3 published topicsWhat the public estimates, measured by survey, set against the recorded figure it was scored on.
What was the question wording, and what year?
The rating scale
The data contradicts the claim.
Something real is being described, at several times its actual size.
The figure is accurate on its own basis, but it is being set beside figures built on a different one. This is the most common problem on this site, and it is not the same as being wrong.
Accurate as far as it goes. A material part of the picture — the baseline, the offsets, the denominator, the price basis — is missing.
Correct when it was published, quoted since against a figure that has moved. Common on both sides of an argument, including among people correcting others.
Part of the claim holds, part does not.
The data backs the claim.
No independent evidence exists either way — the claim is a projection or an assertion.
Three parts, one record
- Part 1
The topic
What the issue is actually about: definitions, context and scale. Video and picture explainers follow.
- Part 2
What's being claimed
Measured public beliefs and attributed claims, rated against their sources.
- Part 3
What the facts are
Number construction, honest expressions, the evidence, counterpoints and the bottom line.
Going further
Watch and read, numbered sources, update history and corrections complete the record.