Early voting at “permanent locations” has been going on a while, but none of the locations are convenient to me. Other locations open up on Monday, including one right down the street from work, and that’s when I’ll vote.
Any sense of whether an uptick in early voting favors a particular party?
1 Like
I simply do not pick up unless I know the number. If it is important enough and they know me, they will leave a message. If they know me, they likely know my cell phone number and will text me. I do get a shit-load of calls. Downside of WFH, I guess, and a first-world problem.
1 Like
Democrats are more likely to vote early than Republicans, so most people are assuming more early voting is good for Democrats.
1 Like
The deck of cards has millions of cards. If I poll two random people in the entire country, you can bet your ass that the outcomes are independent.
So you’re claiming there is a strong selection bias? In the sense that the people who respond to surveys are more likely to vote for Harris or Trump? Very doubtful.
Unfortunately that is not what is happening. It is closer to:
The deck of cards has millions of cards. Suppose Clubs, Diamonds and Spades never answer…
1 Like
Unless I’m job hunting I never ever answer a number I don’t know unless the caller ID identifies it’s my kids school or something
2 Likes
What’s your point? Unless you can prove that such behavior is correlated with political beliefs, then you have no point.
That has nothing to do with independence. I can poll from a biased pool and still get a random sample. It’s just that the estimate that I get is biased.
I heard Charlie Kirk on the radio pushing his listeners to early vote. The unspoken undercurrent seemed he was worried Trump would lose more voters between now and election day. I heard the talking heads on a different religious channel talk about churches needing to have early voting parties, which I guess some have, where everybody comes in and fills out their ballots and votes there. This seems like the religious rights way of keeping their women and kids in line. I do not know how prevalent this is.
1 Like
That was definitely true for 2020 and 2022 and the Republicans bad mouthed early voting so much back then it probably has some carryover to this cycle even though Trump is now trying to encourage his supporters to vote early and even by gasp mail.
As @now_samantha touched on, I was looking at stats earlier this morning, it was something like 45% of early voters (in person and mail-in) were registered Democrats and 30.5% Republicans, with the remainder other or no affiliation.
That doesn’t mean Republicans couldn’t have a sudden surge in early voting, but unlikely.
I think the polls are accurate to the extent that people dont lie. Trying to model the selection effects is pretty hard and probably pointless. I don’t think that republicans (or democrats) are more likely to answer a survey, so it doesn’t matter. It’s possible, however, that new voters lean a certain way, like with Trump in 2016.
Well that willingness to respond to rando calls is definitely correlated with age, which is predictive of voting intentions
2 Likes
Am I crazy in thinking most of the statistical work of political surveying is dealing with that exact problem? You’ve got an unrepresentative sample. They don’t just say oh well, they stratify and make judgments on how to weight it based on forecasts of ultimate voting proportions.
1 Like
This is not exclusive to churches that vote R.
1 Like
I was mostly responding to Steven’s claim that the sampling isn’t independent. I think that’s basically impossible. Now what you are describing is another matter. You are saying that the pool of respondents is simply biased so that the estimate that I get is simply not representative of the entire voting population. If I randomly called people in the united states among a list of registered voters, you are sayiig that the people who respond have a different probability of voting for Harris than the general population? Please convince me.
P[Vote Trump | Old] > P[Vote Harris | Old].
But P[Respond to poll | Old] > P[Respond to poll | Young]
results in a skew for Trump in raw response counts.