Published on 2026-07-25
Why can 'up 100 percent' sometimes mean almost nothing?
A percentage with no starting number attached can hide a huge change or a tiny one. The fix is one question: how much at the start?
Picture a small workshop that sells handmade candles announcing, proudly, that its customer base grew by 100 percent this year. Sounds like a business on fire. Then you learn the real numbers: last year they had one customer. This year they have two. The percentage is completely honest. The impression it gives is not, and that gap between a true number and the story it tells is one of the oldest tricks in reporting, marketing and even casual conversation.
'Up 100 percent' can mean a company now serves half a million more people. It can also mean a workshop went from one candle buyer to two. Without the number it started from, the two sentences read identically. That missing number, the baseline, is doing all the work, and its absence is rarely an accident.
The same change, two different headlines
Economists have a name for a close cousin of this trick: the base effect. Take a price index that climbs from 100 to 150, then from 150 to 200. Both moves add exactly 50 points, the same absolute change. Yet the first period shows growth of 50 percent and the second shows growth of only 33 percent, purely because the second climb started from a bigger number. The prices moved by the same amount both times. Only the base changed, and the percentage changed with it.
That is why a single rise can be reported as either a shocking spike or a mild uptick depending entirely on which year you compare it to. Inflation reporting runs into this constantly: a modest price rise measured against an already high year looks tame, while the identical rise measured against a quiet year looks dramatic. Neither headline is lying, technically. Both are choosing their starting line.
Percentage points are not percent
Jennifer LaFleur, a data journalism trainer whose tip sheet on math errors is published by the Journalist's Resource at Harvard Kennedy School's Shorenstein Center, uses a health statistic to catch this exact trap. If a county's autism diagnosis rate moves from 1 percent of children in 1981 to 5 percent in 2021, that change is 4 percentage points. It is also a rise of 400 percent, a five fold increase, once you express it relative to where it started. Both descriptions are accurate. They read like two completely different stories, because one keeps the starting point in view and the other buries it inside the word 'percent'.
LaFleur's tip sheet offers a plainer version of the same problem: a state legislature that raises a road safety budget from 30 million dollars to 38 million dollars can honestly say funding is up nearly 27 percent. What makes that sentence trustworthy is not the percentage. It is the fact that both dollar figures are sitting right next to it. Strip those two numbers away, leave only 'funding is up 27 percent', and the sentence could be hiding almost any pair of figures at all, from a rounding error in a massive budget to a fivefold jump in a tiny one.
A percentage without its starting number isn't information yet. It's a shape waiting for you to fill in the size.
What the missing number is usually hiding
Once you start looking for the absent baseline, you notice how often it goes missing on purpose. 'Sales up 200 percent' sounds far more impressive than 'sales went from two units to six', even though it's the same fact. A one off jump in a tiny market gets to borrow the drama normally reserved for genuine, large scale growth, and nobody involved has said anything false. They've simply chosen which half of the sentence to leave out.
The same move works in reverse for anyone trying to bury bad news. A cost that triples from a negligible amount stays negligible in absolute terms, so a report can quietly mention 'costs tripled' and let the reader assume something serious happened, when the honest version might be 'costs went from ten dollars to thirty'. The percentage is technically correct in both directions. It just isn't the whole sentence.
None of this means percentages are dishonest by nature. They are a useful shorthand, built for comparing changes of different sizes on the same scale. The trouble starts only when the shorthand travels alone, stripped of the two plain numbers that gave it meaning in the first place. So next time a figure claims to have doubled, tripled or grown by some triumphant percentage, resist picturing the arrow shooting upward. Ask for the pair of numbers behind it instead: how much at the start, how much now. Handed a good pair, you've learned something real. Handed nothing, you've learned something too.
Sources
- 4 Tips for Avoiding Math Errors in Data Journalism (Education Writers Association) (ewa.org)
- Base effect (Wikipedia) (en.wikipedia.org)