thegagline: "Really? I just read the 19 pages of Fox poll results and methodology and it looks like 42% of those polled were Democrats and 43% were Republicans."
Fox does use a left-leaning pollster, estimated as skewing roughly 2% towards Harris, which means Fox is actually calling the popular vote "even" as of today.
More conservative pollsters still give Trump the edge:
Finally, whether a pollster is left-leaning or conservative, the electoral college math works out that Harris will need at least +3% in the popular vote to achieve an electoral college win.
In other words, Harris +3% means the electoral college vote will most likely be tied, so she needs more than 3% to win the election.
If I remember correctly, in 2020 Biden was +7% as of September, and in 2016 Clinton was +5%, so Harris is trailing both of them.
So, Trump is still slightly favored, imho.
I agree with your comments.
Early in my career, I was an independent marketing research specialist, conducting surveys, polls, and performing the statistical analysis for very large corporations
The greatest variable that is hardest to control is sample selection and trying to create a sample that reflects the target population. It is virtually impossible where there is a large population and a small sample.
As it relates to statistical probability, there is a truth that prevails, “Statistics don’t lie, but liars use statistics. “
There is no such thing as left leaning or right leaning.
There will be turnout models that differ. They may differ to result in varied results, but they are all making a call about turnout.
Some may use the Census to define D/R/I party mix in turnout, and they make their 1000+ phone calls and toss out respondents if they already have enough D or R or I. The Census defined that, not the pollster, and the Census is 2021, quite old now. But other pollsters may use a different source of D/R/I. Maybe actual checking of Secy of State sites for states — and of course that will run into problems with the several states who do not register by party.
When pollsters hire spokespeople from both sides, you will get interpretation text that is slanted, but the data itself derives from turnout model. If they expect extra black turnout because of Harris, if they expect extra women turnout because of Harris, these are not unreasonable expectations. They may be wrong, but not unreasonable.
Which is why these polls are pretty tightly clustered. None are coming in with 15 point leads in both directions. They are almost all within MOE.