In a recent article in this journal, Johnston and Pattie argue that th
ere are regional differences in class polarization within Britain and
that these differences have changed over the last 30 years. They claim
that class polarization has remained fairly strong in the north, but
has declined dramatically in the south, so that regional differences i
n class polarization are greater than they were in the 1960s. Like Joh
nston and Pattie, we use log-linear and related models to analyse data
from the British Election Surveys, but we come to very different-almo
st opposite-conclusions. We find that class polarization has declined
in Wales and possibly Scotland, but has shown little or no trend in En
gland. The hypothesis that class dealignment has been most rapid in th
e south is decisively rejected. Regional differences in class polariza
tion are now smaller than they were in the 1960s. Reasons for the diff
erences in conclusions are discussed.