How To Calculate Percent Increase

Percent increase measures how much a quantity grew, expressed as a share of where it began. The formula is short — the change divided by the original value, then multiplied by 100 — and essentially every meaningful question about the topic concerns that denominator. Going from 80 to 100 produces a gap of 20, and 20 divided by 80 is 0.25, so the increase is 25.00%. The new value is simultaneously 1.25 times the original, which is the same fact written as a multiple rather than as a rate. Holding both forms at once makes checking the result almost automatic.

That denominator is the starting value, and the choice is a requirement rather than a convention. A change has to be judged against the situation it changed from, because that is the only baseline that makes the size of a movement interpretable. The same 20-unit gap reads as 25.00% when measured against 80 and as 20.00% when measured against 100, and both are correct answers to two different questions. Substituting the larger figure, the smaller figure or the more recent figure quietly answers a question that nobody asked. When in doubt, ask which value the change happened to, and put that one underneath.

The practical consequence is that the two directions of one move never match. Rising from 80 to 100 is a 25.00% increase; falling back from 100 to 80 is a 20.00% decrease, not 25.00%. Nothing has gone wrong — the second leg acts on a base of 100 rather than 80, and that base is the endpoint of the first leg. Read the gap against the original value and you get 25.00%; read it against the new value and you get 20.00%. The asymmetry widens as the move gets larger, which is why it matters most in exactly the situations where people are most tempted to round it away.

Reversing the calculation is common enough to deserve its own form. If a value rose by 25.00% and reached 100, the original was 100 divided by 1.25, which is 80; the multiplier form new = original x (1 + p) runs in both directions. This is the route to a price before an uplift, a salary before a raise, or a figure before a known adjustment was applied. Checking in that direction catches the denominator error immediately, because 80 x 1.25 lands exactly on 100 while any version built on the wrong base will not. It costs a few seconds and removes the most common failure mode.

Successive increases compound rather than add. Two rises of 10% applied one after the other give a factor of 1.1 x 1.1 = 1.21, which is a single increase of 21.00% rather than the 20.00% that adding suggests. The difference appears because the second rise acts on a base that the first rise has already enlarged. The mechanism repeats across any number of periods, so the gap between adding and compounding grows as the number of steps grows. Anyone summarising a series with a simple sum of percentages is reporting a figure that never described the outcome.

Gains and losses of equal size never cancel, which follows from the same compounding logic. Starting from 100, a rise of 50% reaches 150, and a fall of 50% from there lands on 75, which is 25.00% below where the sequence began. The shortfall is the product of the two swings rather than their difference, so it shrinks quadratically as the swings get smaller. The order does not rescue it either: falling first and rising second reaches the same 75, because the two factors simply multiply in whichever sequence they occur. Statements built on the average of the two percentages describe a position that was never actually held. The honest summary is the net figure measured against the starting point.

There is no upper limit on the result, which surprises people who expect percentages to stay below 100. Doubling a quantity is an increase of exactly 100.00%, because the new value equals the original plus one whole original. Tripling is 200.00%, and each further multiple adds another hundred to the figure. The common slip runs in one direction: describing a doubling as a 200% increase, which overstates the change by a factor of two. What the formula does impose is a floor rather than a ceiling — a genuine increase cannot come out below zero, because the new value is at least the original.

When the quantity that changed is itself a rate, two different units describe the same event. The arithmetic gap between the two rates is measured in percentage points, while the movement relative to the starting rate is a percent increase, and the second figure is always the larger-sounding of the two. Both are legitimate, so the obligation is simply to say which one is being used. Reporting that never distinguishes between them tends to reach for whichever number makes the movement look largest, and that is a choice rather than a calculation.

Three practical details close the picture. First, the two values must share units, because the ratio is dimensionless and the units cancel in the division; comparing 80 of one thing with 100 of another yields a number with no meaning. Second, an original value of zero makes the result undefined rather than infinite, so report the absolute change and the multiple instead. Third, carry decimals through the working and round once at the end, since rounding the ratio before multiplying by 100 shifts the percentage by a visible amount. None of these changes the formula; they decide whether the answer means anything.