Hmm, following the link to look at more charts of the underlying data I'm confused about how to interpret this. First, the definition of "middle class" they're

Hmm, following the link to look at more charts of the underlying data I'm confused about how to interpret this. First, the definition of "middle class" they're using is literally average Chicago income, not fixed to some inflation adjusted amount or compared to the U. S. as a whole. I didn't see at a glance if that average income changed over the time. If the average income rose, but the distribution changed, that's not necessarily the middle class being "almost gone". Also, there's a lot on uncontrolled confounders here. Age is a big one. It mentions that half of census regions saw their average income decline by 20% but that of those regions the majority of people there are under 19 (!). So one interpretation of the data could be that the city as a whole got richer but a few small areas stopped having kids.