How this instrument works
A percentage point is the unit you reach for when both figures on the table are already rates — an interest rate, an approval rating, a tax bracket, an unemployment reading — and you simply want to know how far apart they sit. The formula is points = p1 − p2: no ratio, no reference value, nothing in a denominator. That absence is the entire identity of this instrument. A relative change and a relative difference both exist to turn a comparison into a fresh percent figure, which means both of them divide by something. This measure refuses that step, because dividing one rate by another to manufacture a second percent figure is exactly the move that causes the mix-up this page exists to prevent.
The reasoning behind leaving out the division is that a rate expressed as a percent is already normalized, sitting on a fixed 0-to-100 ruler by definition. Treat two such readings as marks on that ruler and the natural question is simply the distance between the marks — ordinary subtraction, the same operation used for two marks on any ruler. Turning that gap back into a fresh percent of one of the readings adds a second, unrelated layer of relativity on top of a figure that was already relative, and the two layers do not commute: 45 minus 40 is a 5-point gap, but 45 relative to 40 is a 12.5% jump, because that second calculation divides by 40 instead of just subtracting from it.
Two structural facts set this measure apart from its two cousins in this batch. First, sign under a swap: reversing p1 and p2 flips the sign of the result and nothing else — the size never changes, unlike a relative change, where swapping the two inputs alters both the sign and the size of the answer because the denominator switches too. Second, there is no forbidden input and no ceiling: p1 and p2 can be zero, negative, or over 100, and the subtraction still runs cleanly, whereas a relative change breaks entirely at a zero base and a relative difference tops out at 200%. A rate can fall from 2 to −1, a genuine 3-point drop, with no division anywhere to object to it.
- Enter the first rate into Percentage 1 — any figure, including zero or a negative reading.
- Enter the rate you are comparing it against into Percentage 2.
- Read Difference in percentage points: a positive number means Percentage 1 sits higher on the scale, a negative one means it sits lower.
- Swap the two entries to see the point spelled out directly — the figure keeps its size and only the sign reverses, since this is a straight subtraction.
Worked example — a rate moving from 40% to 45%
Set Percentage 1 to 45 and Percentage 2 to 40. Difference in percentage points is 45 − 40 = 5.0 exactly — five points, a plain gap between two figures that already live on the same scale. No fraction is formed and no reference value is chosen; the answer is the subtraction itself, reported with the digit that was entered.
Compare that 5-point gap to the same move read as a relative change instead: (45 − 40) ⁄ 40 × 100 works out to 12.5%, since that second calculation treats 40 as the base and asks what fraction of it the 5-unit rise represents. A headline that reports 'rates rose 5%' when the true move was 40% to 45% has quietly swapped a 5-point gap for a 12.5% relative jump — two correct numbers describing the same event, attached to the wrong label. This is one of the more common mix-ups in everyday reporting on interest rates, poll numbers, and tax changes.
Questions
What is a percentage point, exactly?
It is the plain unit of distance between two figures that are already rates. If a value moves from 40% to 45%, it has risen 5 percentage points — the literal subtraction 45 − 40 — as distinct from a relative change, which would divide that same 5-unit gap by the starting value 40 and report 12.5%.
Why do reporters and analysts get this wrong so often?
Because 'percent' and 'percentage point' sound interchangeable but are not: dropping the word 'point' silently swaps a subtraction for a division. Saying a tax rate 'rose 5%' when it actually moved from 20% to 25% is wrong — it rose 5 points, which in relative terms is a 25% increase, a very different-sized claim.
How is a percentage point gap different from percentage difference?
Percentage difference divides the gap between two raw values by their average, so it works on any pair of numbers and produces a fresh percent figure. This measure only makes sense when both inputs are already rates, and it never divides at all — it is p1 minus p2, full stop, with no averaging step to compute.
Can the result be negative, and what does that mean?
Yes — a negative result simply means Percentage 1 is smaller than Percentage 2. Entering 10 for Percentage 1 and 90 for Percentage 2 returns −80, meaning the first rate sits 80 points below the second one on the scale, with no other interpretation needed.
Why doesn't this calculator ever refuse an input?
Because subtraction has no forbidden values. A relative change is undefined when the starting figure is zero, since it would require dividing by zero, and a relative difference is undefined when the two figures sum to zero. This gap has no denominator to break, so zero, negative rates, and readings above 100 all subtract cleanly.
Does swapping Percentage 1 and Percentage 2 change the size of the answer?
No — only the sign changes. Because points = p1 − p2 is a plain subtraction, reversing the two inputs negates the result exactly: 45 − 40 gives 5, and 40 − 45 gives −5. The size stays fixed at 5 either way, a simpler relationship than a relative change shows under the same kind of swap, where both the sign and the size can shift.