Well fans, Election Day draws near. Biden is gradually becoming, if not popular at least less unpopular, Fivethirtyeight.com says the Democrats are favored to hold onto the Senate and the House, though still likely to switch to GOP control, will remain closely divided. And underlying all the Leftieness? Abortion and the Dobbs decision. I have elsewhere predicted that abortion as an issue was already baked in so to speak and that it wouldn't significantly alter the dynamics of the election. Once again, yet again, I appear to be utterly and absolutely WRONG. Somebody should cash in on my enviably consistent record!
I think abortion, narrowly, wasn't/isn't the issue. It's the whole gestalt. There are two world views being offered, and one of them is really oppressive. I think voters are reacting to that.
Sad story about a pro-life lady in Texas that wanted a baby but the fetus was not viable. Her life was in danger and she had to go to New Mexico for an abortion. Now her Texas doctor is warning her not to get pregnant again while she lives in Texas. https://www.cnn.com/2022/09/09/health/abortion-restrictions-texas/index.html Stories like this will have to turn more of the public against braindead anti-abortion laws? I heard that in many areas up to 70% of new voter registrations are female.
State Trooper (Texas DPS): Alright, little lady. Pull that uterus over to the curb. Where ya goin' all so hot fire there, honey?
I pay little attention to slanted and biased polls regarding a president's popularity. If someone could point to an impartial poll, I'd appreciate it. Right now, I use: https://projects.fivethirtyeight.com/biden-approval-rating/
Polls are samples. That means they're subject to random error. Because pollsters want to be very sure their polling reflects accurately what the population thinks, they set a high bar. Most use the .05 confidence level. That means that there is a 95% chance the poll reflects the population. That means they set a "margin of error" that reflects that 95%. So, when a poll says a result is, say, +/- 3%, that means they're 95% sure the sample result is within that range of accuracy compared to the population. One way to narrow the confidence interval (margin of error) is to increase the sample size. But, quickly, this becomes both expensive and inefficient. The number just don't move as much as the sample size grows. That's why we can tell what a 100 million people feel about a question by asking 600. You can double or even triple the sample size without doing much to the confidence interval. Another way to decrease the margin of error is to do multiple samples, or to aggregate polling results. That's what FiveThirtyEight does. When aggregated, one particular poll--even one with "outlier" results--doesn't have much impact on the margin of error. Polling got a bad rap from the 2016 election, but that wasn't fair. Polls--especially aggregated polls--provided results that were within the margin of error regarding the popular vote. This despite some very fast-moving, last-minute developments. What the polling didn't do was predict the distribution of those results that, by quirk of fate, had Donald Trump winning by small margins in states that mattered--states he wasn't expected to carry. There really is no such thing as an "impartial poll." Humans build polling questions, and humans are subjective. This is why standardized exams are inherently subjective--even math ones. Bias is everywhere. But.... You have to ask not if the poll is impartial, but instead does the poll produce reliable results? You can only truly know this in retrospect. In 2016, everyone was distracted with the flabbergasting outcome to go back and see just how accurate the polling truly was.
