I'd be interested in looking at that study from the 70s. The question I was left with was, "Are regional variations in the cost-of-living being accounted for?" In other words, the Southerners may have been making less money in absolute terms, but since cost-of-living and purchasing power varies in different parts of the country (just look at gas prices), were they really worse off, comparatively? I would suspect the variation would have less to do with the presence/absence of unions and more to do with the presence/absence of industry.