Two different questions dressed as one word
"Percentage change" and "percentage point change" both use the word percentage, both describe a number going up or down, and both get printed with a % sign — which is exactly why they are so often mixed up. They answer different questions. A percentage point change is the plain arithmetic difference between two percentages: if a rate goes from 5% to 4%, it has fallen by 1 percentage point. A percentage change asks how big that move is relative to where it started: that same 1-point fall, measured against the original 5%, is a fall of 20%. Both numbers are true at once. They just answer different questions about the same event.
The classic case: interest rates
Interest rates are where this distinction earns its keep, because a small point move is routinely a large relative one. A savings rate cut from 5% to 4% is a 1 percentage point cut — but as a proportion of what savers were earning, it is a 20% cut in their return. A mortgage lender advertising "rates down by 1%" when a rate moves from 5% to 4% is technically describing 1 percentage point, and conflating it with a 1% relative change (which would only take the rate from 5% to 4.95%) understates the real effect by a wide margin. This calculator's own arithmetic shows why: 20% of 50 is 10 — the same one-in-five ratio as 1 is to 5. A 1-point move on a 5% base and a 10-point move on a 50% base are both, relatively, a 20% change.
The classic case: opinion polls
Polling reports are the other place this goes wrong constantly, because "up 8 points" and "up 8%" describe very different swings once the starting share is anything other than 50%. Say a party polling at 64% rises to 72% — a gain of 8 percentage points. Relative to where the party started, that is a much bigger jump: 12.5% of 64 is 8, so an 8-point rise on a 64% base is a 12.5% relative increase in support. Headlines that say "support rose 8%" when they mean 8 points are not describing a small error — a genuine 8% relative rise on a 64% base would only be about 5 points. The two readings diverge further the smaller the starting number: a rise from 4% to 8% is 4 percentage points but a 100% relative increase — the party's support has doubled.
Why the media default to points without saying so
Percentage points are simpler to state and, coincidentally, usually the smaller and more reassuring-looking number, which is part of why they turn up in headlines without the word "points" attached. Unemployment "falling by 2%" almost always means 2 percentage points (say, from 5% to 3%) rather than a 2% relative fall (5% to 4.9%) — the point figure is what statistical releases actually publish, and it gets shortened to "%" in the retelling. The fix is not complicated: a figure described as a change in something that was already a percentage is a point figure unless the source explicitly says "relative" or does the extra division. When in doubt, ask what the original two percentages were and do the subtraction yourself.
The reversibility trap
Relative percentage changes do not cancel the way people expect, and nowhere is this clearer than with a fall followed by a recovery. A value that falls 50% needs to rise by more than 50% to get back to where it started — because the recovery is calculated on the new, smaller base. Take a value of 80 that falls 50%, to 40. To climb back from 40 to 80 you must add 40 — and 40 is 100% of the new base of 40, not 50% of it. The calculator makes the underlying rule obvious on its own terms: 100% of 80 is 80 — adding 100% to any number always doubles it, which is exactly the multiple a 50% fall demands to reverse. A share price, a house valuation, or an investment portfolio that drops 50% in a downturn has not "broken even" once it is up 50% again; it is still 25% below where it started, because that second 50% was calculated on a smaller number than the first one was.
A general rule for reversing a percentage change
The pattern generalises: to fully reverse a fall of x%, the required rise is x ÷ (100 − x) × 100%, always bigger than x itself once x is above zero. A 10% fall needs an 11.1% rise to reverse; a 20% fall needs a 25% rise; a 50% fall needs a 100% rise; a 90% fall needs a 900% rise. The larger the original drop, the more disproportionate the recovery required — which is precisely why avoiding a large percentage loss matters more than chasing an equivalent-looking gain.
Frequently asked
Is a 10 percentage point rise the same as a 10% rise?
Only if the starting figure was 100. Otherwise they are different numbers describing the same event from different baselines — a percentage point rise is the raw difference between two percentages, a percentage rise is that difference measured relative to the starting value.
How do I convert a percentage point change into a relative one?
Divide the point change by the original percentage and multiply by 100. A move from 40% to 50% is 10 percentage points; relative to the 40% starting point, that is (10 ÷ 40) × 100 = 25%.
Why does a 50% loss need a 100% gain to recover?
Because each percentage change is calculated on the value at that moment, not the original value. Halving a number and then doubling the smaller result returns you to the start — but doubling is a 100% increase, not 50%, precisely because the base you are increasing from has already shrunk.
Which figure should I use for the percentage-of-value arithmetic itself?
Our percentage calculator handles the underlying "percent of a value" step behind every example on this page — useful for checking either the point difference or the relative change once you know your two figures. See how we verify every ToolHare tool on the how we build and verify our tools page.