This is, of course, the final polling update of the 2016 cycle
as tomorrow is election day.We have a small confession to make. We have used
RCP averages throughout this entire cycle because frankly the site is easy to
use, but that ease has come at a price. As we have watched more carefully we have
noticed a minor skewing of their averages.
RCP’s clear inclusion of narrative skewing polls at the state level
means that if we continue through another cycle, we will be doing our own
average. With that caveat, here are the
numbers.
The races looks ever so marginally
tighter this week, but that is somewhat deceptive. Clinton leads by 47.2 to 44.3 down from 48.0 to 44.9 last week. This
was basically because pre-Comey 1 and pre-tightening polls were included last
week and dropped from average this week. To give you an idea of the stability
here are the current polling numbers with + for Clinton and a Minus for
Trump, (+3, +4, +1,+3,+4 +6,+ 4,-5 ,+5,+7,+2,+4,
tie). Obviously, one of these thing is
not like the others. If you drop best for polls for each side,
you end up with Clinton by 3.3.
The Four-Way Data does almost nothing to change this math. We should also note that almost all of the
differences in these margins come from differences in Non-White, performance or
margin.) Except for the LA- Times
outlier, the White margins moved between 12-18 points in polls for which we
could find cross-tabs. Clinton did not do much better in the aggregate even
when she got within 12 of Trump with Whites. So we stand by our point from yesterday. This is a close race, but not one where
there is much disagreement about the outcome. For a different outcome the polls
have to be wrong en masse.
0 comments:
Post a Comment