Percentages: The Five Mistakes That Trip Up Even Careful People

Percentages look simple. Everyone encounters them in middle school and most adults assume they’ve got the concept down. Then a graph in a news article reports that “unemployment rose 5%” and a debate erupts about whether that means 5 percentage points or a 5% relative increase — which mean very different things, and often the article doesn’t make clear which.

This is a practical tour of the percentage mistakes that still catch careful people out.

Mistake 1: percent vs. percentage points

A percentage point is the arithmetic difference between two percentages. A percent is the relative difference.

If an interest rate rises from 4% to 5%:

• The increase in percentage points is 1 (5 − 4).
• The increase in percent is 25% (the new rate is 25% higher than the old).

Both are correct. They measure different things. Using the wrong one in a discussion is a recipe for confusion.

News headlines routinely muddle these. “Unemployment rose 5%” can mean the unemployment rate went up by 5 percentage points (huge, catastrophic) or by 5% relative (a modest quarter-point move from 4% to 4.2%). The responsible way to write it is “rose by 5 percentage points” or “rose by 5%, from 4% to 4.2%” — with enough context for the reader to pick.

A useful mental rule: when comparing two percentages directly, ask “percentage of what.” Five percent of a 4% rate is 0.2 percentage points. Five percent of a 50% rate is 2.5 percentage points. The same “5% increase” is 12× bigger on the second baseline.

Mistake 2: percentage changes aren’t symmetric

A stock goes down 50% then up 50%. Where is it?

Not where it started. Down 50% leaves $100 as $50. Up 50% from $50 is $75. You’re down 25% from the beginning.

To get back to $100 from $50, you need a 100% gain, not 50%.

This asymmetry shows up everywhere:

• A company that lost 30% of its customers needs to gain back 42.9% of its remaining customer base just to return to the original number.
• A diet that lowers a portion size 20% and then increases it 20% ends up 4% smaller than the original.
• Stock market drawdowns: the S&P 500 lost 57% from peak to trough in 2008-2009. Recovering required a 133% gain, not 57%.

The general rule: if something drops by x%, recovering requires a gain of x/(100-x) × 100%. At small x, these are close (a 5% drop needs 5.26% to recover). At large x, they diverge dramatically.

Mistake 3: markup vs. margin

A seller buys an item for $60 and sells it for $100. Pick a measure:

• Markup (percent of cost): (100 - 60) / 60 = 66.7%
• Margin (percent of revenue): (100 - 60) / 100 = 40%

Both describe the $40 gap as a percentage. They use different denominators, and the numbers look very different. Someone saying “40% margin” and someone saying “66.7% markup” can be describing the same deal — but pricing discussions get confused when one person is using markup and the other is using margin, and neither realizes.

Retail and wholesale contexts tend to use markup; finance contexts tend to use margin. Mixing them up is a classic mistake in procurement negotiations.

The conversion:

• Markup → Margin: markup / (1 + markup). A 50% markup is a 33% margin.
• Margin → Markup: margin / (1 - margin). A 40% margin is a 66.7% markup.

Our Percentage Calculator handles the direct “X percent of Y” math; for pricing decisions, always confirm with your counterparty whether they mean markup or margin.

Mistake 4: compound changes don’t add

A price is discounted 20%, then another 10% is taken off the discounted price. What’s the total discount?

Not 30%. It’s 1 - (0.8 × 0.9) = 28%.

The second 10% is taken off a smaller base. Percentage discounts compose multiplicatively, not additively.

This matters in a handful of real contexts:

• Stacked discounts: a “20% off everything, plus 10% off sale items” doesn’t mean 30% off; it means 28% off.
• Inflation-adjusted returns: an investment that grew 10% in a year while inflation was 3% didn’t net 7%. The real return is (1.10 / 1.03) - 1 = 6.8%.
• Sequential tax brackets: a marginal rate of 30% on top of an 8% state rate isn’t 38% combined; depending on whether state is deductible federally, it’s usually a bit less.

Our Discount Calculator handles stacked discounts correctly, including the multiplicative composition.

Mistake 5: base-rate neglect in percentage reporting

A cancer screening test has a 95% accuracy rate, meaning it correctly identifies 95% of cases and correctly rules out 95% of non-cases. Your friend got a positive result. What’s the chance they actually have the disease?

Most people say 95%. The correct answer — without knowing more — is much lower, and depends entirely on the base rate (how common the disease is in the population being tested).

If the disease affects 1% of the population:

• Out of 10,000 people tested, 100 have it. The test correctly flags 95 of them.
• Of the 9,900 without it, the test incorrectly flags 5% × 9,900 = 495.
• Total positive results: 95 + 495 = 590.
• Of those, 95 are true positives. That’s 95/590 = 16.1%.

A 95% accurate test, on a disease with a 1% base rate, produces positive results that are right only 16% of the time.

This isn’t a statistics error in the technical sense — the test really is 95% accurate. It’s what happens when you apply a percentage without considering the base rate. The same math comes up in:

• Security screening: if a breach detection system has 0.1% false positives on a network of a million daily requests, you get 1,000 false alarms a day.
• Forecast accuracy claims: “85% accurate” sounds great until you realize that predicting “no earthquake today” every day would be right 99.9% of the time.
• Marketing attribution: “this channel converts 5% of visitors” is only meaningful if you know what the base conversion rate of any visitor is.

A few quick sanity checks

Some rules of thumb that catch most daily percentage confusions:

• 50% off + 50% off = 75% off, not 100% off. (1 × 0.5 × 0.5 = 0.25.)
• 100% gain = 2×. Zero-to-one is “new thing,” not “100% gain.”
• 200% increase means the thing tripled, not doubled. Confusingly used in news.
• “Up by a factor of 10” = 900% increase, or “10× the original.” Not 1000% increase.
• A 1% increase every month is about 12.7% per year, not 12%, because of compounding.
• Tip % is a percentage of the pre-tax total in most of the US (some restaurants now compute tip suggestions from the post-tax total, which the Tip Calculator lets you toggle between).

When fractions are clearer than percentages

Percentages are convenient for round numbers but can obscure simple ratios:

• “37.5%” is 3/8.
• “66.7%” is 2/3.
• “12.5%” is 1/8.

For discussions about splits, probabilities, or proportions, fractions are sometimes clearer than percentages. “One in three” is immediately meaningful; “33.3%” is a number whose ratio you have to reconstruct.

Our Fraction Calculator handles the conversions and arithmetic when fractions are the natural unit — recipe scaling, for example, where 1/3 + 1/4 is more natural to work with than 33.3% + 25%.

The meta-point

Percentages compress information, and compression throws away context. When someone quotes a percentage, the useful question is: percent of what? Compared to what? Under what assumptions?

These questions aren’t pedantic. They’re the difference between a 5-point unemployment spike (catastrophic) and a 0.2-point tick (normal noise). They’re the difference between a deal with 50% markup and a deal with 50% margin. The percentage itself is often the least informative part of the sentence.

The good news is that most percentage mistakes come from a small number of patterns. Learn the five above, and most of the remaining daily confusions resolve themselves.