Before You Share a Statistic, Check the Denominator

A percentage describes a relationship between quantities. Without knowing what is being counted and what it is divided by, readers cannot tell whether two numbers describe the same thing.

Name the population behind the rate

Ask who or what belongs in the numerator and denominator. For example, BLS defines the unemployment rate as unemployed people divided by the labor force, expressed as a percentage. It is not a percentage of the entire population.

Before comparing rates, check the definitions, location, time period, and exclusions. Two figures with similar labels can measure different populations.

Factory photograph cataloged by the Library of Congress as 1936. A historical illustration, not a depiction of today's labor market. Photo: Harris & Ewing. Source · Public domain; no known restrictions. Image unmodified.

Distinguish percentages from percentage points

Consider a hypothetical rate that rises from 4% to 6%. The increase is 2 percentage points. Relative to the original 4%, it is a 50% increase. Both calculations describe the same change, but they answer different questions.

A headline using only the larger-sounding figure can leave the starting level unclear. The before and after values belong beside the change so the scale stays visible.

Keep counts and rates together

In a hypothetical example, incidents rise from 10 among 1,000 cases to 15 among 2,000 cases. The count increases, while the rate falls from 1% to 0.75%. Neither number alone describes both the workload and the frequency.

Choose the measure that matches the question. Total incidents may matter for staffing. The share of cases affected may matter for comparing groups. Explain the distinction rather than selecting whichever measure delivers the stronger reaction.

Check the comparison before the conclusion

Ask whether the series uses the same definitions and collection method across the period. Read notes about estimates, revisions, and missing observations. A graph's starting point also affects the story a reader sees.

Use a final checklist: What is counted? Out of what total? Compared with when? Under which definition? Then state what the statistic shows and what further evidence would be needed to explain why it changed.

Statistics & Data Literacy

For a broader guide to categories, denominators, baselines, timeframes, rates and chart framing, read How to Read Political Statistics: Baselines, Denominators, Rates and Timeframes.