[boken]

This is addressed at length - and quite early on - in the article. The correlation stands, apparently.

[daveguy]

Thanks for the heads up. They say there is both a youth element and an income element and claim the effect is present even compensating for youth. However, there are exactly zero statistics (alpha, p, confidence, nothing) to justify the conclusion.

[boken]

Which is why I wrote 'apparently.'

This is an interview with NPR, not an academic paper. You may as well question whether the correlation exists at all as hone in on this one point.