Earlier quoted context omitted.
32%->27%->28%->30% appear to be just rounding errors, especially considering the number of individuals concerned. If anything, such numbers indicate to me that homeownership rates haven't meaningfully changed and the only thing that did change is the advent of noisey narratives.
> appear to be just rounding errors, especially considering the number of individuals concerned Isn't it in fact the opposite? The bigger population, the easier to find a significant signal? Those are variances of millions of people. We wouldn't treat a 5% swing in unemployment as a rounding error, would we?
If unemployment keeps hovering around 29% +/-3% over basically a century, that variance is a rounding error.