Very Red and Very Blue Districts
Naturally the question arises as to whether the common ground positions in the country as a whole are also found in very red or very blue districts. To find out the samples were divided six ways according to the partisan index (from Cook’s ratings) for the Congressional districts where the respondent lived. In every case the direction of majority opinion was the same as for the national sample and in nearly all cases the views in very partisan districts differed from overall national sample less than the national sample of partisans, and in the few cases where they did it was by no more than the margin of error.
This leaves open the possibility that some specific districts might differ from the national sample more than the national partisan sample, but that is likely to be quite rare and it is extremely unlikely that the direction of majority opinion would differ from the national sample of partisans. It is also possible that the partisan sample within a very partisan district might differ from that national partisan sample, but based on analysis of numerous cases we have generally found that partisans in very partisan districts are no more deviant from the overall national sample than is the national partisan sample.
Examining the views of the demographic subgroups—including race, age, gender and education—in a very small number of cases a demographic sample was divided while the overall sample had a majority position. In only one case did the majority of a demographic group diverge from the national majority position and, surprisingly, this was a case in which African-Americans were not supportive of a tax increase on the wealthy.