Election Polls: sensitive to methodology
Reuters/IPSOS did at least 2 polls in the past week, to show how the exact wording of the choices affects the results.
Given the choice of Clinton, Trump, and “Other”, it was: Clinton 41, Trump 35, Other 25
Given the choices appearing on ballots, it was: Clinton 37, Trump 37, Johnson 5, Stein 1
(which leaves an additional 20% unaccounted for)
A USC polling series has an intention-to-vote result as: Trump 47, Clinton 42
Some graphs in the “Characteristics of Candidate Support” in the USC link above show unexpected trends. [Note that the y-axis scales in those graphs vary, making some of the trends look more exaggerated than others !!]
Trump is strongest among middle income voters, winning that $35-$75K household income block by 15 points!!! Clinton is ahead significantly in the lower and upper income groups. This is an interesting development I wasn’t aware of…
[via king of the alt-media dumpster dive, Zerohedge . Their article obsesses about the removal of the “Neither” option in favor of “Other”, but as usual if you follow the links there is novel and valuable information to be found]
UPDATE – another poll by PPP, 7/30, has Clinton +5 in both 2-way and 4-way races, and has lots of good info on favorability, but sadly limited demographic information.
Progressives, Vote Jill Stein 2016! Say No to Iraq/Libya style warmongering! Say No to TPP!