2024 Political Polls

  • Thread starter Thread starter nycfan
  • Start date Start date
  • Replies: 2K
  • Views: 56K
  • Politics 
aGDevil2k said:
The thing Nevada and Georgia have in common is a massive, multi cultural metro area that is completely unlike the rest of the state.

I wonder if that makes them harder to poll well.
Click to expand...
In other words, rural areas with mostly white folks drag down states and the country.
 
Milk and Cookies said:
Have you heard conservatives talk about wanting to change the electoral college to give one vote per county? They say this with a straight face.
Click to expand...
Texas

www.newsweek.com

Texas GOP amendment would stop Democrats winning any state election

A Texas Republican proposal would require anyone holding statewide office to have the support of a majority of Texas counties.
www.newsweek.com www.newsweek.com

"The Republican Party of Texas has voted on a policy proposal that would require any candidate for statewide office to win in a majority of the state's 254 counties to secure election, effectively preventing Democrats from winning statewide positions based on the current distribution of their support."
 
HeelingAg said:
Texas

www.newsweek.com

Texas GOP amendment would stop Democrats winning any state election

A Texas Republican proposal would require anyone holding statewide office to have the support of a majority of Texas counties.
www.newsweek.com www.newsweek.com

"The Republican Party of Texas has voted on a policy proposal that would require any candidate for statewide office to win in a majority of the state's 254 counties to secure election, effectively preventing Democrats from winning statewide positions based on the current distribution of their support."
Click to expand...
I assume the Texas Highest Court has given an advanced go-ahead
 
HeelInTheOzarks said:
Also feels weird to see Nevada that far right of Arizona.
Click to expand...
I've felt all along that Nevada will eventually go Democratic in the end, as it has consistently done since 2008. Nevada is like the GOP version of NC for Democrats - they keep coming close but it's never quite enough. Maybe this year it will change, but I think at the end of the day Harris will win the state by a narrow margin. There have also been other polls recently which show Harris leading in Nevada, so the polls seem more mixed than this article shows.
 
HeelingAg said:
Texas

www.newsweek.com

Texas GOP amendment would stop Democrats winning any state election

A Texas Republican proposal would require anyone holding statewide office to have the support of a majority of Texas counties.
www.newsweek.com www.newsweek.com

"The Republican Party of Texas has voted on a policy proposal that would require any candidate for statewide office to win in a majority of the state's 254 counties to secure election, effectively preventing Democrats from winning statewide positions based on the current distribution of their support."
Click to expand...
This is actually how Georgia used to hold "elections" for most statewide offices, including governor. They had a "county unit" system where each county in GA was given a certain number of votes, and small rural counties were deliberately given greatly exaggerated votes while cities like Atlanta were deliberately given a much lower number of county votes. So a GA county with 10,000 people received about the same number of votes for governor as Atlanta, which back then had 450,000 to 500,000 people. It was about as blatantly undemocratic as you can get, and was definitely designed to give all the power in state government to rural areas at the expense of urban ones. The US Supreme Court finally ruled this system as unconstitutional and overturned it sometime in the 1960s I believe.
 
Mulberry Heel said:
This is actually how Georgia used to hold "elections" for most statewide offices, including governor. They had a "county unit" system where each county in GA was given a certain number of votes, and small rural counties were deliberately given greatly exaggerated votes while cities like Atlanta were deliberately given a much lower number of county votes. So a GA county with 10,000 people received about the same number of votes for governor as Atlanta, which back then had 450,000 to 500,000 people. It was about as blatantly undemocratic as you can get, and was definitely designed to give all the power in state government to rural areas at the expense of urban ones. The US Supreme Court finally ruled this system as unconstitutional and overturned it sometime in the 1960s I believe.
Click to expand...
That story is actually scary given the current makeup of the SCOTUS.
 
WBroylesH said:
I can't understand one word of that memo. Sounds like it was written by someone in a remedial college composition class.
Click to expand...
It's about sampling bias.

1. One difficulty pollsters have in this age of tiny response rates is assembling a representative sample. One way of doing that is age, race and gender -- but there are multiple problems with that. First, the demographic data on race is only an estimate, and there are issues about how people identify. For instance, they might have checked "black" on their driver's license or census bureau form, but now identify as "biracial" or "other" and will tell pollsters that. Gender is considerably more precise but it's still an estimate, like everything produced in the census.

Another way of doing that is party identification. That is, AFAIK, pretty much discredited now, for many reasons. One of the biggest is that party identification is fluid, as many posters can confirm with their life stories. If Trump pushes people to self-identify as Democrats, that might not match how they identified in the past.

2. So an approach often used these days is to try to get a sample that matches vote totals from a previous election (the recalled vote). The virtue there is that it's objective. We know exactly how many votes were cast and for whom in 2020. So then you ask people, as part of the poll, who they voted for in 2020, and then you weight the poll accordingly. You know the electorate was basically 50-50 in Georgia in 2020, so maybe you should be looking for a sample in which half of the respondents voted each way.

There are a couple of problems with the "recalled vote" approach. First, it relies on voters accurately recalling their votes. This might not be as much memory as honesty. If the polls were off in 2020 because Trump voters were shy about admitting it, then you'd think they would also be reluctant to tell pollsters that they voted for Trump. A second problem is that voter enthusiasm and turnout is endogenous within the model. For example, suppose turnout was unusually high on the R side in 2020. The overall registered voter poll in 2020 might have been D +3, but because Pubs turned out en masse, the vote totals were 50-50 (note: I'm just pulling numbers and hypos. This could work in reverse as well). So if you balance your sample according to the vote totals in 2020, you'd end up with a skewed RV sample (though not necessarily as skewed in LV).

3. Finally, the electorate changes over time. Voters die, voters age in, voters move. This is a problem for all polling; the census bureau numbers, after all, will also be stale to some degree. But there are ways of researching demographic trends -- they might be inexact, but marketers have all sorts of sophisticated methods to determine, say, how many black people there are in GA in 2024 and where they live (and search history/social media plays a big role here). By contrast, the recalled vote can't really be researched and adjusted, or if there are ways of doing that, I don't know about them.

So what the Trump memo says is that the sample was skewed according to the #2 approach (recalled vote), because it was 7 points favorable to Kamala in an election that was actually tied. It's a reasonable point. The response would be something like the shy Trump voter effect I mentioned above, or some of the factors listed in point #3. No sample is perfect.

This is yet another reason not to focus on individual polls. The adequacy of the sample is usually only realized after the election is over.
 
Mulberry Heel said:
This is actually how Georgia used to hold "elections" for most statewide offices, including governor. They had a "county unit" system where each county in GA was given a certain number of votes, and small rural counties were deliberately given greatly exaggerated votes while cities like Atlanta were deliberately given a much lower number of county votes. So a GA county with 10,000 people received about the same number of votes for governor as Atlanta, which back then had 450,000 to 500,000 people. It was about as blatantly undemocratic as you can get, and was definitely designed to give all the power in state government to rural areas at the expense of urban ones. The US Supreme Court finally ruled this system as unconstitutional and overturned it sometime in the 1960s I believe.
Click to expand...
Today, the SCOTUS would rubber stamp that BS.
 
You must log in or register to reply here.
Back
Top