> might very well correlate with bad education outcomes

So would dying at age 16.

It's definitely possible (probable?) they've taken the appropriate figures here, but there's definitely a chance for bad statistics here.