Explainer
Why rounded percentages do not always add up to 100%
The short answer
Rounding each component independently can make the displayed percentages sum to 99.9% or 100.1% even when the underlying shares sum to exactly 100%. Calculate from the underlying values and disclose presentation rounding.
In this article
A table’s three category shares add to 99.9%. That is a reason to inspect the inputs, not an automatic proof that a category is missing. The discrepancy may result from rounding each row before adding the displayed values. Reconstruct the arithmetic before proposing an explanation.
Equal thirds are enough to produce a difference
Illustrative example
| Category | Count | Exact share | Displayed share |
|---|---|---|---|
| A | 1 | One third | 33.3% |
| B | 1 | One third | 33.3% |
| C | 1 | One third | 33.3% |
| Total | 3 | 100% | 99.9% when displayed rows are added |
The exact shares sum to one. Each displayed share omits the repeating decimal after 33.3, so their displayed sum is 99.9%. The total can correctly be presented as 100% when it is calculated from the counts. A note should explain that components may not sum to the total because of rounding.
This distinction follows the relationship between counts and relative frequencies described in OpenStax’s frequency-table explanation. The table here is an original arithmetic example, not a reproduced exercise or an observed market distribution.
A displayed number represents a range
Under ordinary rounding to the nearest tenth, a displayed 12.3% can represent values from approximately 12.25% up to 12.35%, with the precise treatment of boundary ties depending on the rounding convention. It does not provide unlimited precision for a later calculation.
If two releases both display 12.3%, their unrounded values may still differ. Likewise, subtracting two rounded percentages can yield a slightly different result from subtracting the original values and rounding once. Keep the most precise permitted source values in the calculation layer and round the final presentation.
Small counts create large percentage steps
With a denominator of eight, one additional count changes a share by 12.5 percentage points. A result displayed to two decimal places cannot remove that coarse underlying resolution. More decimal places describe formatting; they do not create more observations or establish greater measurement accuracy.
A useful table may therefore show both the count and the percentage. That lets a reader recognize whether a dramatic-looking percentage change comes from a large underlying population or a small number of records. Explain the denominator before drawing an industry-wide conclusion.
When rounding is not a sufficient explanation
- Check whether the categories are mutually exclusive and exhaustive.
- Check for a hidden “other,” unknown or unclassified category.
- Confirm that every row uses the same denominator and period.
- Look for totals that mix currencies, units or reporting populations.
- Read the source’s stated rounding and suppression conventions.
If categories overlap, their shares need not sum to 100% at all. If the denominator changes by row, adding shares may be meaningless. A rounding note should not be used to excuse a discrepancy larger than the published precision can plausibly explain.
