How To Calculate A Percentage

The word "percent" means per hundred, and that single fact generates all three question types. If percentages are always measured against something, then every problem has three participants — the part, the whole, and the percentage — and asking for any one of them is just algebra on the same identity P = W x x/100. Learning to name those three roles before reaching for a calculator prevents nearly every error people make with percentages.

Type one is "what is x% of W?" You know the base and the rate, and you want the slice. Multiply: 18% of $450 is $450 x 0.18 = $81.00. This is the form used for tips, sales tax, commissions and any situation where a rate is applied to an amount. The only real failure mode here is forgetting to divide the percentage by 100, which produces an answer exactly one hundred times too large.

Type two is "P is what percent of W?" You know both amounts and you want the rate. Divide the part by the whole: $63 out of $180 is 63/180 = 35%. This is the form behind test scores, survey results and every "what share of the budget went to this line" question. The rule worth memorising is that the denominator is always the thing that follows the word "of" — individuals again and again get this backwards and produce 285% instead of 35%.

Type three is "P is x% of what?" You know the slice and the rate and you need the whole, which is the hardest one to recognise in the wild and the easiest to get wrong. Divide rather than multiply: $72 being 24% of something means that something is $72 / 0.24 = $300. This form shows up when a tax-inclusive total needs splitting back into price and tax, when a discount is quoted off an unknown list price, and in nearly every "we grew 8% last quarter" inference.

Percentages are not additive, and this is where confident arithmetic quietly produces wrong answers. Two successive changes multiply rather than add. Increase $250 by 20% and you reach $300; reduce the result by 20% and you arrive at $240, not $250, because the second 20% acts on a larger base. The net movement is -4%, and it is always exactly minus the square of the single change whenever the two moves are equal and opposite.

The consequence generalises further than people expect. Discounts stacked at 30% and then 20% on a $200 item give $200 x 0.70 x 0.80 = $112, whereas a single advertised 50% off would have given $100. The order never matters — multiplication commutes — but neither does the total ever behave like 30% + 20%. Whenever you see sequential percentage arguments, reach for multiplication factors rather than addition.

Reversing a percentage increase is the inverse failing of intuition. A gain of 25% to reach 100 from 80 requires a fall of only 20% to undo, because 20 of the new 100 is one fifth, while 20 of the old 80 was one quarter. The further the original move, the more extreme the asymmetry: a 90% collapse demands a 900% rise to return to the starting point. This single relationship explains a great deal of confident nonsense in financial commentary.

Finally, percentage points and percentages are different units and conflating them changes conclusions. A rate moving from 2.0% to 2.5% has risen half a percentage point, but it has risen twenty-five percent of its own value. Both statements are correct and they answer different questions; whenever someone reports a change in a rate, ask which unit they mean before acting on it.