Percent Change vs. Percentage Points: The Mistake That Sneaks Into Real Decisions

A One-Letter Error That Survives Editing

Open almost any week of financial news and you will find some version of this sentence: "The central bank raised interest rates by 0.75 percent today." The sentence is wrong, and the wrongness is not pedantic. The central bank raised rates by 0.75 percentage points, which on a rate moving from 1.75 to 2.50 is a 42.9 percent increase. Those are different facts about the world, and confusing them turns a routine policy move into either a yawn or a calamity depending on which version you read.

The percent-versus-percentage-point distinction is one of those rare statistical errors that survives careful editing, that appears in serious newspapers, and that a smart reader can correct in their head once they know the shape of it. This piece is a working explainer: what each term actually measures, where the swap shows up most often, and the small mental habit that catches it every time.

The Two Quantities, Stated Cleanly

A percentage point is an absolute difference between two percentages. If unemployment goes from 4% to 5%, that is an increase of one percentage point. The subtraction is straightforward: 5 minus 4 equals 1. The unit is the percentage itself, treated as a number on a scale from zero to one hundred.

A percent change is a relative difference, expressed as a fraction of the starting value. The same move from 4% to 5% unemployment is a 25 percent increase, because the new value is one quarter larger than the old one. The calculation is (new minus old) divided by the absolute value of old, times one hundred.

Both descriptions are true at the same time. They are not in tension; they answer different questions. "How much did the number change in absolute terms?" gets you percentage points. "How much did the number change relative to where it started?" gets you percent change. The error is not in choosing between them. The error is in using the word "percent" when you mean "percentage points," or vice versa, which destroys the reader's ability to tell which question you answered.

This is the single most useful sentence to keep in your head: when both numbers are themselves percentages, the difference is in percentage points; the ratio is in percent. If you can hold onto that, the rest follows.

Where the Confusion Actually Lives

The confusion is concentrated in domains where the underlying numbers are themselves rates or shares: interest rates, polling, tax brackets, unemployment, conversion rates, click-through rates, market share. In every one of those domains the headline value is already a percentage, which means a change in the headline value can be described either way, and which way matters.

Interest Rates and the Fed

The cleanest example is monetary policy. Through 2022, the Federal Reserve raised its target federal funds rate from a range of 0.25 to 0.50 percent in March up to a range of 3.00 to 3.25 percent by September, in a sequence of moves that included several 75 basis point increases. A "basis point" is one hundredth of a percentage point, which is the conventional unit precisely because the percent-versus-percentage-point confusion is so easy to make in financial reporting. A 75 basis point hike is a 0.75 percentage point increase. It is not a 0.75 percent increase. On a target rate moving from 2.50 to 3.25, those translate to either a routine quarter-of-a-point move or, on the percent-change reading, a 30 percent jump in the headline rate. The first is what the Fed actually did. The second is the sentence you will read in a careless wire story.

The convention of quoting central bank decisions in basis points exists to dodge the ambiguity. When a Fed governor says "we moved seventy-five," everyone in the room knows they mean seventy-five basis points, which means 0.75 percentage points. The journalist translating that for a general audience has to make a choice, and the wrong choice happens often.

Polling and Margins

The same confusion sits underneath every misread of a political poll. When a candidate's support goes from 42 percent to 45 percent, that is a three percentage point gain, which is roughly a seven percent relative gain. If the poll has a margin of error of plus or minus three points, the change is at the edge of significance — the apparent movement could be entirely noise. A reader who hears "seven percent gain" and "three percent margin of error" mashed together loses the ability to make that judgment, because they are now comparing two numbers that are not measuring the same thing.

Polling aggregators that report movements in points rather than percent are doing the reader a favor. The convention preserves the reader's ability to compare the size of a move to the size of the noise floor without having to mentally translate units.

Conversion Rates and A/B Tests

The percent-versus-points confusion is endemic in product analytics. A landing page test reports that the new variant has a conversion rate of 4.5 percent against the control's 4.0 percent. The honest description is "0.5 percentage points higher, or about 12.5 percent relative lift." The marketing deck almost always picks the bigger number — "12.5 percent lift" sounds like a real win, "half a percentage point" sounds like noise. Neither framing is wrong, but a product manager making a roadmap decision needs both, because the absolute change tells them how many additional customers per thousand visitors the test produced, and the relative change tells them how much room there might still be on the same lever.

The same trap shows up in retention numbers, churn, click-through rate, and any other rate that already lives on a zero-to-one-hundred scale. A churn drop from 8 percent to 6 percent is two percentage points, or a 25 percent relative reduction. The first number is what shows up on next quarter's revenue. The second is what shows up in the deck.

Tax Brackets and Marginal Rates

Tax policy is a third place where the confusion does damage. A proposal to "raise the top tax rate by three percent" can mean two completely different things. If the current top marginal rate is 37 percent, a three percentage point increase moves it to 40 percent. A three percent relative increase moves it to 38.11 percent. The political fights around tax policy frequently rest on whether the audience hears one or the other, and the headline writer's choice between "by three percent" and "by three points" can swing the reader's sense of the move by an order of magnitude.

The Mental Habit That Catches It

The single habit that catches this error is simple: every time you read or write a sentence in which a percentage changes, ask whether the change is in percentage points or in percent, and refuse to let the sentence stand until you have decided. If you are reading, the sentence might be ambiguous; figure out which one is mathematically plausible given the underlying values and assume the writer meant that. If you are writing, pick the right word and stick to it.

When the underlying numbers are small (a polling shift from 42 to 45, an unemployment move from 4 to 5), the two framings differ by a large multiple, and the difference is consequential. When the underlying numbers are large (a literacy rate moving from 95 to 96), the two framings are numerically close, and the difference is small but still real. The habit is the same in both cases.

If you want to keep yourself honest while drafting, our percentage calculator has both modes side by side — "X is what percent of Y" for ratios and "percentage change from X to Y" for relative differences — and showing yourself the two numbers next to each other for a few examples is the fastest way to internalize the distinction. After you have done it five times, the language drops into place on its own.

Basis Points, the Workaround

The financial industry got tired of explaining the difference and just made up a new unit. A basis point is one hundredth of a percentage point, used to describe small changes in interest rates, bond yields, and credit spreads. "The ten-year yield is up twelve basis points" is unambiguous in a way that "the ten-year yield is up 0.12 percent" is not, because the second sentence could be misread as a 0.12 percent relative change, which on a yield of 4 percent would be 0.0048 percentage points, a movement nobody would ever bother reporting.

The convention is borrowed sporadically into other domains. Polling sometimes uses "points," tracking polls sometimes use "tenths of a point." The underlying purpose is the same: detach the unit of change from the unit of the underlying quantity so the reader cannot misread one as the other.

What Honest Reporting Looks Like

The best version of a sentence describing a change in a rate gives both framings explicitly, in the order that matches the reader's likely question. "The unemployment rate rose to 5 percent, a one percentage point increase from last month, or about 25 percent in relative terms." The two clauses prevent each other from being misread. Most copy desks consider this too wordy, which is why the error survives.

If you are writing for an audience that includes anyone who will make a decision based on the number — investors, voters, product managers, policy people — the wordiness is the right call. The reader who only wanted one framing will skim past the other. The reader who needed both will thank you. The reader who would have misread a single-framing version is now safe.

The percent-versus-percentage-point distinction is, in the end, a small piece of statistical literacy that pays off every time you encounter a number-about-a-number in the wild. It costs nothing to learn, it survives whatever changes the news cycle throws at you, and it gives you a permanent edge on readers who have not done the work. That is a good return on a few minutes of attention.