The conclusion? We know from the USSA that the controlling variable is economic, but there is a strong predictability to economic scores from social. The mean squared error between the Michigan SOSS and the social ratings of the USSA and between the SOSS and the economic ratings are virtually identical. So there is room for the hypothesis that when someone responds "I'm a conservative" to the generic orientation, we can interpret that to mean that person's frame of reference is primarily economic; and a response of "I'm a liberal" would lead one to believe that social concerns outweigh economic ones.

But as Larry Hembroff, the director of the Office of Survey Research, is fond of saying, "Each survey points us to the questions we need to ask in the next one."

Nat Ehrlich is a survey research specialist at the Michigan State University Institute for Public Policy and Social Research.