Southerners, that is.
We were discussing a Phillips Curve paper in macro last night that was sort of an institutionalist approach to the Phillips Curve (which makes sense - unemployment is related to the price level through wage bargaining and labor market institutions). He had two indexes he was looking at - an employment insecurity index and a social bargain index. One of the elements of the employment insecurity index was the change in the manufacturing employment differential between the Midwest and the South (the whole plant movement thing). That went positively into the index (interregional shifts increase employment insecurity).
What a weird way to think about things!
After all, that's an increase in employment security for the South! Sure the jobs could leave some day, but at the frequency he was looking at it, it is unambiguously an increase in employment security.
He tested the elements of the index separately too, and that one - not surprsingly since it's really ambiguous in its relationship to employment security - was insignificant.