I am going to give up and fall back on my initial protest that there are better things to discuss. Think about it scientifically, and ask: What question does this analysis answer? What are we trying to learn? We are not trying to answer any of these questions: Are people doing better or worse today than before? Is life getting better or worse for the poorest people? Are typical individuals in the population seeing improvement? The question we are trying to answer is more like this: How have the membership counts in arbitrary demographic blocs changed over time, relative to each other? We are discussing changes in the shape of the bell curve, while ignoring the fact that the bell itself is moving and growing. As if someone asked you about the home team's season, and you said "compared to last year, they scored more points in the first and last quarters of their games, and fewer near halftime."