This is a really bizarre graph... 35 in age for example represents a completely different point in history for every one of the generations, so if the gap is bigger, is it:

• Because of racist stuff changing or accumulating? Or...

• Because of home ownership changing nationwide over those intervening years, in general, and that showing up at different points along the X axis for each generation even though it was simultaneous, tricking the viewer?

For example, if Gen X is 20 years younger than Millenial, then 50 for the blue line = the same point in time as 30 for the purple line. The gap at 50 on the blue line is around 28%. The gap at 30 on the purple line is... about 28%

So... gap in home ownership for Gen X and Millenials was actually dead even at the same point in actual time? But not when you spread them out by making the X axis age.

That doesn't mean both things aren't still happening. Just that we can't interpret it from a graph. I think this needs to not be a graph at all that humans have no way to validly make sense of, but instead a multivariate analysis to control out each variable each way and find the most well fitting model.

Unrelated: it would be vastly easier to follow if there was just a "ratio" line for each, not two lines.