With a sample size in the tens of millions (and hundreds of millions for whites) there's a good chance that a 4% difference is actually statistically significant, too.
And the old adage "statistical significance is not substantive significance" also applies.
Hybrid Mode
