General Comments on the New York Times Polls
First, we have to note that they have to call about 25,000 people to get the standard 500 responses. That is 2%. If you don’t know or don’t reveal why 98% don’t answer, it raises the problem of serious biases. However, when you look at their maps, for those that answered, it implies that they all answered the survey. I find that odd. It wasn’t stated how the callers identified themselves.
I shouldn’t speculate, but I will. If you call a Trump/Fox News indoctrinate, they know that the NYTimes is “fake news” and the “enemy of the people”. So they would refuse to talk to you, and you could seriously undercount the Trump support voters.
Or, if conservatives watch the regular news, they know that their backing of Trump is illogical, and looked down on by liberals. So they don’t want to be embarrassed by revealing their choice to a presumed liberal interviewer.
On the other hand, finding out why the 98% don’t answer, or don’t talk to you, could be as revealing as the detailed demographics of those who do take the poll.
This could be why Trump won the Presidency, despite the polls showing Clinton the clear winner. Actually, it is not. Trump won the three crucial rust belt states by less than a percent, which put him over the electoral college count. One percent is too small to detect by polls with errors in the four percent range.
Then there is the problem of the 10% or so that are “undecided”. The NYTimes also admits that they don’t know if some are just decided but “decline to state”. That could be for the same reasons as above.
About 80% of their responders are on cell phones. The rest are on land lines. What does this imply?
Four demographics previously seen in 2016 election polls stand out.
- There is a 90% party loyalty. So the crucial votes are the Independents. Since Democrats need to overcome Republican district leans, the independents are very important. They do lean Democratic.
- Older people are more Republican and younger people more Democratic. But the youngest have a smaller probability of voting.
- White males favor Republicans, minorities and women favor Democrats. But minorities and women don’t overly favor Democrats as much as the 100% you would think that they should.
- The more education people have, the more those people favor Democrats.
Other article have shown that about 20% of those who claim that they are going to vote do not vote.
The polls have about a 5% error on each party candidate, and quote a 10% error on the difference between the candidates.
Their is also heavy rain expected in the Central to the Eastern US on Monday and Tuesday. Since Republican run states make city Democratic voters stand in line forever, this will favor Republicans.
Unfortunately, the NYTimes did not ask how many people are having nightmares about the hordes of women and children about to INVADE the US from Honduras. And, if that was affecting their voting.
