Follow two groups over time
In a fictional example, a harmful outcome falls from 20% to 10% in Group A and from 10% to 4% in Group B. Both groups improve.
Initially, the absolute gap is ten percentage points and the risk ratio is two. Later, the absolute gap is six percentage points and the ratio is 2.5. The gap is smaller in percentage points but larger as a ratio.
| Time | Group A | Group B | Absolute gap | Ratio |
|---|---|---|---|---|
| Earlier | 20% | 10% | 10 percentage points | 2.0 |
| Later | 10% | 4% | 6 percentage points | 2.5 |
Neither measure is an arithmetic error. Each preserves a different relationship between the groups.
What does “inequality increased” mean here?
It means the relative comparison increased, if that is the chosen measure. It would be incomplete to leave out that both groups' risks and the absolute difference decreased.
Likewise, announcing only the smaller absolute gap would omit the larger proportional difference. Reporting both is often more informative than treating one as the definitive account.
The measurement is not the explanation
The CDC's ratios-and-proportions material provides the basic arithmetic distinction. It does not tell us why the groups differ. Changes in population composition, exposure, access, measurement, or other conditions may matter.
The example also does not determine whether an observed difference is unfair or what policy response is appropriate. Those judgments require additional evidence and explicit values.
For a public report, a clear sentence could say: “Risks fell in both groups; the percentage-point gap narrowed while the ratio increased.” That wording preserves a complicated result without forcing it into a single direction of progress.
