> Very strong" is not a very useful description. Quantitative explanations are important. Unfortunately, your linked paper is not available. This one is:

> https://www.academia.edu/download/3521147/Pratt___Cullen_200...

Err, not for me, it isn't.

So I looked up the Pratt/Cullen 2005 paper (https://sci-hub.se/10.1086/655357), and right off the bat, the abstract doesn't seem to agree with you:

> Indicators of "concentrated disadvantage" (e.g., racial heterogeneity, poverty, and family disruption) are among the strongest and most stable predictors.

Also, what makes you think your counterexamples are unrelated to poverty? I would expect "unemployment length" and "poverty" to be HIGHLY correlated, and unemployment is #2 on the list.