Sex noise soundboards

Note: This essay was written about a month ago, for a school project. Some of the numbers and polling averages may be slightly outdated, but the point of the essay still stands.
Introduction
“DEWEY DEFEATS TRUMAN.”
That’s what the front-page headline of the Chicago Tribune said on November 3rd, 1948. It’s also what the polls had all been saying for months: that New York governor Thomas Dewey would defeat incumbent president Harry Truman and become the next president of the United States. And yet, he didn’t. Truman won reelection in a massive upset, defying the polls. Somehow, Truman had gone from trailing Dewey in polls by so much that cartoons like the following were created, to winning the election.
https://preview.redd.it/oqba22kugvzc1.png?width=800&format=png&auto=webp&s=92204f20feee6faea87f731a797760140c4a0814
Truman was a very unpopular president. His campaign was also plagued by third parties threatening to split his votes: Strom Thurmond on the right and Henry Wallace on the left. The way he was able to snatch victory from the jaws of defeat can’t be described as anything less than amazing.
Historians still debate over it, but the most popular theory is that Truman was able to win many voters who disapproved of him because he successfully painted Republicans as being worse than he was. This strategy was aided by Dewey’s weak campaign. Many voters didn’t like Truman, and when polled, wouldn’t say they would vote for him, but when the time came, they held their nose and pulled the lever for the president. The election was a lesson to not treat polls as gospel.
Today, the nation faces another presidential election. The Democratic candidate is incumbent president Joe Biden. He’s running for reelection despite concerns about his age and rumors he wouldn’t run again due to it. On the Republican side, former president Donald Trump is the nominee. He faced opposition in the primaries, most notably by former South Carolina governor Nikki Haley, but beat her and his other opponents without too much trouble. The election is the first presidential rematch since 1956. Several independent/third-party candidates are running too, the most notable being Robert F. Kennedy Jr. (RFK Jr. for short), nephew of JFK. He’s been polling very high for a third-party candidate, getting double digits in many polls.
Biden beat Trump in 2020, but opinion polls have been showing Trump leading Biden, often by large margins. As of April 1st, Trump leads Biden by 1.1% in the national polling average according to racetothewh.com, an election prediction/poll aggregation website. Biden won the popular vote by 4.5% in 2020, so this is a sizable swing right. Trump also leads Biden in all seven swing states. Below is a chart comparing the 2020 presidential election margin and the 2024 polling average in the seven swing states.
https://preview.redd.it/9wvdn2yzgvzc1.png?width=631&format=png&auto=webp&s=e9c69e14cedfecc11d866837b9533d3e39a30db0
It seems like Biden’s doomed. He needs to win at least some of the swing states to win the election, and right now he isn’t winning any of them. It looks like Trump is on track to becoming the second president ever to win a non-consecutive second term, after Grover Cleveland.
But there’s something else going on.
Biden’s bad polling situation seems simple on the surface. He’s incredibly unpopular, having an approval rating of 39.1% (net -16.3%) according to FiveThirtyEight. His bout of unpopularity seems to have started around the Afghanistan withdrawal, although when asking someone their reason for disliking Biden they’ll probably say something about his age or the economy instead. So, it makes sense that Biden would be polling badly. He’s an unpopular president, and people would rather have Trump.
But it isn’t that simple. Because looking deeper, there are some things that don’t make sense. Crosstabs of polls showing massive realignments not seen since the Civil Rights era. Other indicators of a president in trouble not showing up. Things that when put together, suggest Biden may not be in as much danger as the polls say.
When all the evidence is put together and analyzed, it’s clear that Biden is not doomed, not at all. Biden’s bad polling can be explained by two things. First, bad polling methodology underpolling his supporters. Second, people who are supporting third parties now, but will eventually return to Biden. These two things are both making Biden’s polling look bad, although which one has a stronger effect depends on the poll and the demographic group. Additionally, all the indicators other than the polls, like primary elections and special/off-year elections, don’t show Biden in too much trouble.
Explaining Racial and Age Depolarization
First, context is needed for the rest of this essay to make sense. So, as was said earlier, 2024 polls are showing Biden doing much worse than his 2020 performance. That makes sense – Biden is less popular, so naturally fewer people want to vote for him. The strange part is what demographic groups Biden is slipping with. Instead of a mostly uniform shift, which would be expected, almost all of Biden’s losses seem to come among nonwhite voters – most significantly black and Hispanic voters. He’s also losing ground among young voters (usually defined as voters between the ages of 18 and 29). The Democratic Party traditionally does well with these groups, so this is of course concerning for Biden. Even more strange is that in some polls, Biden is actually making some inroads among the demographics that are historically the base of the Republican Party – those being white voters and seniors. Looking at the aggregation of crosstabs of polls during February, there are many abnormalities.
The aggregation shows Trump making massive gains among black and Hispanic voters (swings of R+28.4 and R+18.5 respectively) but making almost zero gains among white voters (R+0.1, but right under that there are slight blue swings with both college educated and non-college educated whites, likely a product of not all polls recording results for those groups). This is strange, to say the least. White people seem to be perfectly fine with Biden, while nonwhite people suddenly despise him. This phenomenon is called racial depolarization, or racedep for short.
Swings among different age groups are also odd. Trump is improving by 16.1 points among voters aged 18-29 but losing 1.8 points with seniors and 4 points with voters aged 50-64. Young voters are much more liberal than older voters. Every opinion poll and election result suggests this. Unless they’ve suddenly become much more conservative, them supporting Trump over Biden doesn’t make sense. Along with racedep, age depolarization ("agedep") is common in crosstabs of 2024 polls.
Those are not the only depolarizations supposedly going on, as can be seen in the tweet. Urban and suburban voters moving towards Trump while rural voters move towards Biden. Democrats moving towards Trump, Republicans moving towards Biden. Geographical and political polarization have been increasing in recent years, so this suggests a strange reverse of that trend. 2024 probably won’t be a large realignment, it’s more likely something is just wrong with the polls.
Explaining Primaries
Presidential primary season has been going on for a few months, after the Iowa caucus kicked it off in January. While Biden and Trump both won their primaries easily, how strong their performances were in different areas can reveal a lot about how certain groups are feeling about the candidacies of the two – like black, Hispanic, and young voters. But first, protest voting has to be explained.
When an incumbent president is running for reelection, they usually do not face much opposition in the primaries. Typically, only no-name minor candidates are the other people on the ballot besides the president. They do not have a chance at winning, but they do serve as a way for people who are upset with the president to express it. Sometimes, the “Uncommitted” option is also used to protest. Look back to 2012, when Obama was running for reelection. He swept the primaries, but his worst performances were in West Virginia, Kentucky, Oklahoma, and Arkansas, where he got under 60% of the vote.
The four states all had something in common: a lot of the registered Democrats were white conservatives who before 2008 voted Democratic, but switched to McCain because they didn’t like Obama’s dark vision for America. They voted against Obama in the primaries because they didn’t like him and didn’t want him to be the nominee. Those voters would then go on to vote Republican in the general election. The places that swung the hardest against Obama in 2008 were also the places where he did the worst in the 2012 primaries.
2004-2008 swing
2012 Oklahoma Democratic presidential primary
2012 Arkansas Democratic presidential primary
2012 Kentucky Democratic presidential primary
2012 West Virginia Democratic presidential primary
Now, those four states were already very red even before 2008, Obama was not going to win them and he did not need to win them. But if a candidate is doing badly in a potentially competitive state’s primary, they should heed the warning – or risk losing. Another good example of protest voting can be found in the 2016 Democratic primary. Hillary Clinton did very poorly in the Rust Belt states of Wisconsin, Michigan, and Pennsylvania – losing the first two to Sanders and coming close to losing the last. And where Sanders’s support was strongest was in rural areas – also the areas that swung the most towards Trump in the general election. Trump narrowly flipped all three of those states, winning him the presidency.
2016 Wisconsin Democratic presidential primary
2016 Michigan Democratic presidential primary
2016 Pennsylvania Democratic presidential primary
2012-2016 swing
The polls said Clinton would easily win all three states, while the primaries said she would struggle in them – and the primaries were right.
The 2024 Primaries
Presidential primaries can give an idea of where a candidate might underperform in the general election, and 2024 primaries are no exception. If black, Hispanic, and young voters are upset with Biden, like the polls are suggesting, then they will protest vote against him. The first primary that will be examined is the South Carolina primary. South Carolina is 26% black according to the 2020 census, and that number is even higher among Democratic primary voters thanks to the racial polarization of the state – Biden won 90% of black South Carolinians in the 2020 election, while Trump won 73% of white South Carolinians.
https://preview.redd.it/x2t8cnl3hvzc1.png?width=338&format=png&auto=webp&s=2b5982c343da804a10a1221e623b2de84b2f1b86
South Carolina was also the first primary state (so Biden did not have momentum from winning contests at that point, nor was he the presumptive nominee), and the primary was open (meaning independents could vote), so the conditions for protest voting were as good as they could possibly be.
But despite all that, Biden got 96% of the vote.
If black people really are upset with Biden, they clearly don’t hate him enough to cast a protest vote against him. And looking at individual counties, there’s not even a correlation between the percent of black people and the percent of opposition vote. Biden got 97% of the vote in Allendale County (73% black, the blackest county in the state) and he got 95% in Pickens County (7% black, the least black county in the state). If anything, Biden did better in counties where there are more black people. And it’s not just South Carolina – in pretty much every state where black people make up a significant percentage of the Democratic electorate, Biden won by huge margins. He got 99% in Mississippi, 95% in Georgia, 90% in Alabama, and 86% in Louisiana. Biden came close to losing a few counties in Louisiana – but not the ones with lots of black people. The counties he did the worst in are heavily white. The same kind of people who gave Obama trouble in the 2012 primaries voted against Biden, too.
Evidently, black people aren’t protest voting against Biden. Young voters will be looked at next, using the Michigan primary. Just like South Carolina, Michigan has open primaries.
There was an organized campaign for the “Uncommitted” option in Michigan to protest Biden’s policy on Gaza and pressure him into calling for a ceasefire. The Uncommitted option did modestly well, getting 13% of the vote, slightly higher than it did twelve years ago when Obama was running for reelection. The Uncommitted campaign achieved their (unambitious) goal of 10,000 votes, getting slightly over 100,000. Biden got 81% of the vote, while Williamson and Phillips took the remaining 6%.
What’s interesting though, is where Uncommitted did the best. Its strongest performance was in Wayne County (which includes Detroit and a few other cities), where it got 17%. Wayne County is home to 140,000 Arab Americans who make up 7.8% of the county’s population, so the strong Uncommitted performance wasn’t surprising. The second strongest county for Uncommitted was Washtenaw County (also 17%), which doesn’t have many Arab Americans. What it does have, however, is the University of Michigan. With over 50,000 students enrolled, it’s one of the largest colleges in the country. Looking at a precinct map of the results for Washtenaw County, Uncommitted did well because UMich students were protest voting against Biden.
https://preview.redd.it/nov5qkx5hvzc1.png?width=629&format=png&auto=webp&s=cec905bdfdd4fa10be01d03a97a220925d4ffa6d
Ann Arbor, the city where UMich is located, had a very high percentage of Uncommitted votes. There’s no doubt about it, college students were voting Uncommitted to protest Biden’s handling of the war in Gaza.
Looking at college counties in other primaries, there was generally a trend of the Uncommitted option (or whatever name the state has for it) doing well. In Dane County, Wisconsin (University of Wisconsin), there was lots of protest voting against Biden. “Uninstructed” got 15% in Dane vs 8% statewide.
“None of these names” did well in Douglas County, Kansas (University of Kansas), getting 14.5% of the vote, compared to the statewide average of 10.3%.
And Uncommitted got a sizable 21% in New Haven, Connecticut (Yale University), compared to 11% statewide.
There’s definitely some protest voting against Biden by young voters. But remember the reason most of them are unhappy with Biden in the first place: it’s because of Gaza. Trump is more pro-Israel then Biden, so it makes no sense for them to support him. That’s different from Haley voters, who are ideologically between Biden and Trump. Things may be more complicated than they seem, as will be discussed later, but first here’s the analysis of the third group Biden has been slipping with in polls: Hispanic voters. The Texas primary is a good place to judge how Hispanic voters are feeling about Biden. Texas has open primaries, like Michigan and South Carolina.
Biden did the worst in South and West Texas. One of the places he underperformed the most was the Rio Grande Valley (RGV). He got percentages in the 60s, 50s, and even 40s in many RGV counties, with his worst performance being in Zapata County, where he got a pathetic 40% of the vote.
The RGV is heavily Hispanic, so at first this seems like a validation of the polls showing Trump making massive gains among Hispanic voters – but it isn’t the only place in Texas where Hispanic people live. Biden performed very strongly in El Paso County, an 82% Hispanic county home to the city of the same name.
He also did well in places like Bexar County (San Antonio, 59% Hispanic), Dallas County (Dallas, 40% Hispanic), and Harris County (Houston, 43% Hispanic).
Looking at other states, it seems like Biden’s RGV performance was the exception, not the rule. He got 81% in Imperial County, California (86% Hispanic); and 83% in Santa Cruz County, Arizona (83% Hispanic).
Hispanic voters have been slowly trending towards Republicans over time, so Biden’s performances are even more impressive when that factor is taken into account. According to exit polls, Hispanic voters voted for Obama by 44 points, Clinton by 38 points, and 2020 Biden by 33 points. A lot of the people voting against Biden may be registered as Democrats but didn’t vote for him in 2020.
https://preview.redd.it/h35vewo8hvzc1.png?width=407&format=png&auto=webp&s=3c5b78394104a627ae1b8019db62aa1c3a4a1b70
https://preview.redd.it/jlo9nlhdhvzc1.png?width=377&format=png&auto=webp&s=726526e7da2a9c8690ab01e00a12e2e49265445d
https://preview.redd.it/l4tremrehvzc1.png?width=458&format=png&auto=webp&s=0744e5c12f7c0c4eb05ec84b59a070174b017b98
Overall, primaries don’t support the polls showing Trump making huge gains among black/Hispanic/young voters. There’s zero evidence black voters are upset with Biden. As for the other two groups, there are some signs of discontent, but not enough to warrant the double-digit swings polls are showing. Biden’s underperformances in college counties/Hispanic counties, when present at all, are usually less than 10 points worse than his statewide performance. And that’s assuming every single person protest voting will go for Trump. If all protest voters really do vote for the other party in the general election, say hello to Biden’s second term, because Nikki Haley regularly gets twice the number of votes in Republican primaries as Biden’s opposition does in Democratic primaries. Even after she dropped out.
Midterms, Off-Years, and Special Elections
At the same time Biden has been doing well in primaries, Democrats have been scoring wins in special/off-year elections. These elections are historically correlated with the popularity of the president, so they conflict with the polls showing Biden down. Look at elections during the last three presidencies to know what happens when a president is unpopular.
While Trump was in office, he was quite the unpopular president, and his party lost many elections because of it. Through 2017-2019, Republicans lost a net 8 governorships, going from 34 to 26; and a net 41 House seats, going from 241 to 200. The only chamber they managed to gain in was the Senate (thanks to a very favorable map and increased polarization causing many Democrats in red states to lose) – but not without losing a special election in Alabama, a deep red state that had voted for Trump over Clinton by almost 28 points.
This pattern continues to back when Obama was in office. From 2009-2011, when he was at the height of his unpopularity due to the state of the economy and Obamacare, Democrats lost big. They went from 28 governorships to just 20, 257 House seats to only 193, and 59 Senate seats to only 53. Like Republicans with Alabama during Trump’s presidency, Democrats managed to lose a Senate special election in a state considered safe for their party – Massachusetts, which had voted for Obama by 26 points in 2008.
And it goes even further back to Bush’s presidency. Backlash over the wars caused Republicans to lose 6 governorships from 2005-2007 (going from 28 seats to 22), 30 House seats (232 down to 202), and 6 Senate seats (55 to 49).
But despite Biden’s unpopularity and bad polling, Democrats have been doing well in elections despite precedent saying they shouldn’t be. The 2022 midterms, which were supposed to be a red wave, were anything but. Democrats flipped a net 2 governorships and 1 Senate seat, and only barely lost the House. The small majority Republicans won has been giving them trouble when trying to govern. Already, one Speaker was ousted and it’s possible a second might be too.
More recently, Democrats won the governorship in Kentucky and almost won it in Mississippi, both very red states. They flipped the Virginia state house and won a supreme court election in Pennsylvania by a large margin. Two months ago, they won a competitive special election for a House seat in New York by a decisive 8-point margin.
Interestingly, the normal pattern of an unpopular president’s party doing poorly manifested early in Biden’s term. After his approval rating crashed during the Afghanistan withdrawal, Democrats went on to lose the governorship (and state house) of Virginia, and almost lost the governorship of New Jersey. Both states voted for Biden by double digits in the 2020 election. Something changed between November 2021 and November 2022 to cause this shift. It might have been the Supreme Court’s decision to overturn Roe V. Wade and allow states to ban abortion. In several special elections right after the decision, Democrats overperformed massively. For example, Republicans won the special election for Nebraska’s 1st congressional district, which voted for Trump by 11 points in 2020, by only a 5-point margin. The election took place just four days after the Dobbs decision.
The Trump-backed candidates nominated in many Senate and governor elections could also be the ones to blame. Thanks to Trump’s endorsement, many extremist candidates won the primaries in key races. They often denied the results of the 2020 presidential election and had other problematic views. Most of them went on to lose the general election, sometimes by huge margins. Below is a table of all the results.
https://preview.redd.it/vx1ilmujhvzc1.png?width=633&format=png&auto=webp&s=2771b74c5d4257d66b4825078ada46216b0be9bd
Whatever the cause, Republicans flopped in 2022 and haven’t recovered since. And it doesn’t seem like Trump will be able to avoid the problems plaguing his party. His handpicked candidates were the ones that did terribly while other Republicans often did well; and the abortion issue isn’t just going away, not to mention Trump’s the one responsible for getting Roe overturned with his SCOTUS appointments.
Of course, there’s a counterargument: that Biden is somehow breaking historical precedent, and he’ll do badly while other Democrats do fine. That seems like a reasonable theory, until the fact that Biden vs Trump and the generic congressional ballot are polling exactly the same is considered. As of April 5th, at least.
https://preview.redd.it/l0ecq2slhvzc1.png?width=753&format=png&auto=webp&s=d8c231135e068129cc1f9c3e1a3b9b2ce41be3fb
Since work on this essay has started, Biden has experienced a little surge of support in the polls. It could just be noise, but it might be something else.
https://preview.redd.it/m14gsmjmhvzc1.png?width=1043&format=png&auto=webp&s=43bc8d8146b31f5a613a1e7a4adc4ca30a858750
Biden has also been polling as well as (or sometimes even better than) hypothetical Democratic candidates for president like VP Kamala Harris, California governor Gavin Newsom, and Michigan governor Gretchen Whitmer.
It could be argued Biden is only doing better because he has higher name recognition, and Democrats who don’t know the other three candidates are answering undecided. But Michelle Obama being extremely well-known didn’t stop her from trailing Trump by the exact same amount as Biden in a poll.
https://preview.redd.it/7h189dpnhvzc1.png?width=1200&format=png&auto=webp&s=42aa042e9462022d397bbe212c428e41f4d40c99
Democrats are doing much better in actual elections than in polls, and Biden’s polling the same as other Democrats. It stands to reason that Biden would also do better in an election than in polls.
The Problem with the Polls
While primary and off-year elections suggest Biden isn’t doing badly, they still don’t explain the polls. One theory is that the black/Hispanic/young voters who don’t like Biden aren’t voting in any elections, that’s why Democrats are doing well. Perhaps the biggest proponent of this theory is Nate Cohn, the chief political analyst for the NYT.
This theoretical group of low-propensity Trump supporters who love answering polls but don’t vote in any elections sounds dubious, and that’s probably because it doesn’t exist at all. Biden’s bad polling is caused by two main things. The first is bad methodology, but before that is discussed, how polls work must be explained.
Polls work by contacting a certain number of voters, usually around a thousand, and asking them how they plan to vote. The 2024 options are usually Biden/Trump/Undecided/Other. Sometimes Other is changed to real third-party candidates, like RFK Jr. Polls also ask information on the voter, like their race, sex, age, and region. After data is collected, polls are weighted to reflect real demographics. For example, if a poll’s raw data has 40% of respondents living in urban areas while 60% live in rural areas, and the actual percentage of voters is 50% urban and 50% rural, then the responses of the urban voters are weighted higher. If that poll has urban areas voting 60D/40R and rural areas 40D/60R, then the raw data is 48D/52R while the weighted (and final) data is 50D/50R.
This seems like an effective way to avoid bias in polls, and account for lower response rates from certain groups. If rural voters are answering at a higher rate, just give them less weight. If Hispanic voters are answering at a lower rate, give them more weight. The thing is, voters don’t belong to just one group. A person can both live in a rural area and be Hispanic. And while groups (rural voters, Hispanic voters) are weighted, subgroups (rural Hispanic voters) are not.
Say, rural Hispanic voters are more Republican than urban and suburban Hispanic voters. Say, they’re answering polls at higher rates as well. Rural voters will be weighted lower in the poll, but that’s just all rural voters combined. Rural Hispanic voters are not weighted vs other Hispanics. That would lead to Hispanic voters in the poll being more Republican than they are in reality.
A typical poll has around a thousand respondents, and a margin of error of about ±3%. The sample sizes for different groups, however, are much smaller, which means a bigger margin of error. Let’s say Hispanic voters are 10% of the poll’s respondents, or a hundred in total. That’s a margin of error of ±8%, much larger than the ±3% for the poll as a whole. And if rural Hispanic voters are 20% of all Hispanic voters, that’s a margin of error of ±18%! Small inaccuracies in subgroups can cause a ripple effect that makes the whole poll wrong. Let’s do a simulation to show this effect off.

• True voting intention among all Hispanic voters is 63% Biden, 37% Trump (D+26).
• True voting intention among all non-rural Hispanic voters is 65% Biden, 35% Trump (D+30).
• True voting intention among rural Hispanic voters is 55% Biden, 45% Trump (D+10).
• 100 Hispanic voters answer the poll.
• Rural Hispanic voters make up 40% of the poll’s respondents (40 people), they make up 20% of the real Hispanic population. Since subgroups are not weighted, their influence on Hispanic voters in the poll is double what it should be.
• Due to the large margin of error of ±13%, rural Hispanic voters who answered the poll said they’d vote 45% Biden, 55% Trump. That’s 22 Trump voters and 18 Biden voters answering the poll. A proportional sample would have 22 Biden voters and 18 Trump voters. That’s just a 4-person difference.
• Non-rural Hispanic voters in the poll said they’d vote 65% Biden, 35% Trump (the true number).
• The average of Hispanic voters in the poll is 57% Biden, 43% Trump (D+14), a 12% swing from the true numbers.

And all that must happen for this problem to occur regularly is for Trump-voting rural Hispanics to answer polls at a slightly higher rate than Biden-voting rural Hispanics, and rural voters to answer polls at a higher rate than urban voters. And since polls collect responses from people who answer the polls first, the effect can happen easily.
You guessed it, this is happening in real life. And not just with Hispanic voters, but with everyone.
A pattern among 2024 polls is that rural voters are answering at a higher rate than urban/suburban voters. In one NYT/Siena poll (Trump+4), rural voters made up about 35% of the respondents, when they only made up 19% of the 2020 electorate.
In another poll by Grinnell College (Trump+7), rural voters made up 27% of the respondents. Voters who said they lived in a “town” made up 17%, and it’s likely at least some of them would break for rural if they had to choose between urban/suburban/rural.
Looking at the 538 poll database, a clear pattern emerges. Polls that have Trump leading Biden have a proportion of rural voters that is way too high. Polls where Biden leads Trump have more normal numbers.
Rural voters tend to be more conservative and vote Republican, and sure enough, Republicans are answering at a higher rate then Democrats. (scroll to "Do you consider yourself a Democrat, a Republican, an independent or a member of another party?" for the NYT/Siena poll and the top of page 6 for the Grinnell College poll. Both show more Republicans answering the poll than Democrats.)
One pollster, Susquehanna Polling and Research, remarked that Trump supporters seem to have higher enthusiasm than Biden supporters, and so are answering polls at a higher rate.

The second reason why Trump may not be winning Pennsylvania has to do with who is answering polls. We suspect because Trump is the only candidate with “enthusiastic” voters, it’s Trump voters in particular who are disproportionately talking to pollsters. It’s the reverse of what happened in 2016, when the phenomenon of “shy” Trump voters meant that many pollsters undercounted Trump’s base of support. Many voters were afraid to admit they were Trumpers back then. Today, we suspect many pollsters are not adjusting their samples to account for this “non-response” bias, as it’s typically called. But SP&R is doing so.

Polls also say that Trump voters are more enthusiastic than Biden voters.

Republicans are slightly more enthusiastic ahead of November’s general election, edging out Democrats, according to a new survey.
In the poll, released Thursday by Gallup, 59 percent of Republicans said they are more enthusiastic about voting in the upcoming election than in previous years. Fifty-five percent of Democrats and Democrat-leaning independents said they felt the same.

Groups like seniors and white voters may not be swinging towards Trump because there isn’t an enthusiasm gap, unlike with black/Hispanic/young voters. According to a YouGov poll, groups that aren’t swinging towards Trump in the crosstab aggregate are also paying more attention to the election (and therefore are more enthusiastic, and answering more polls). When black/Hispanic/young voters start paying more attention, they’ll get enthusiastic and start answering polls, which should improve Biden’s polling.
https://preview.redd.it/0899t1ephvzc1.png?width=1074&format=png&auto=webp&s=4f9fe91a2d30381a9f08e7e1883b90679aefd6a0
And that rural Hispanic voter hypothetical was based on something real. Rural Hispanic voters were already more Republican than other Hispanics in the 2020 election; and Biden did badly in the rural RGV in the primaries while doing better in cities like El Paso. The difference may be even larger than it was four years ago, with rural Hispanics swinging against Biden while urban and suburban Hispanics don’t. Rural Hispanics make up a small percentage of Hispanic voters (scroll down to "Area type"), so this swing doesn’t mean much for Biden’s electoral prospects. It screws with the crosstabs of Hispanic voters, however.
As Biden’s voters become more enthusiastic and the gap closes, polls may start swinging towards him as more of his voters answer polls. There have already been signs of this happening, like that surge in support mentioned earlier. Perhaps it’s because of the recent ad blitz by Biden energizing his supporters?
Oversamples, and the True State of the Election
Biden voters are not answering polls as much as Trump voters, and this is creating big swings in crosstabs thanks to low sample sizes. Polls with bigger sample sizes would be much better. The margins of error would be much smaller and the crosstabs much more accurate. Unfortunately, it’s too expensive to make polls with huge sample sizes, but there’s still the next best thing – oversamples.
Oversamples are polls that poll only one specific group. While a normal poll polls everyone, an oversample might poll only black voters, for example. Because of the big sample sizes, oversamples are much better for determining the voting intentions of groups than just looking at the crosstabs of normal polls. Oversamples can also use more advanced methods of polling to reach people who may not respond otherwise.
There are three oversamples that are going to be examined here. The first is by Black PAC, and it’s an oversample of black voters.
https://preview.redd.it/epcr7xeqhvzc1.png?width=680&format=png&auto=webp&s=6938941ae9e6b345778035bfd45f7ceb81aa98ed
Trump gets a pathetic 8% of black voters, less than half of the polling aggregate showing him getting 18%. This, along with Biden’s strong primary performances, suggest that the bad polling for Biden among black voters is entirely due to bad polling methodology.
Next, Hispanic voters. An oversample of Hispanic voters by Univision shows Biden leading Trump 58-31 (27 points). Again, that’s completely different from the polling aggregate showing Biden winning them by only 6 points. It is a slight decrease from 2020, where he won them by 33 points; but like stated earlier, Hispanics have been trending right for a while, so Trump making small gains among them isn’t surprising.
And finally, young voters. Split Ticket, an election prediction and analysis website, polled young voters. They used live text interviews, rather than a normal method like calling landlines.
In the poll, Biden leads Trump 35-25, a 10 point lead. Biden is disapproved of by 68% of young voters, while Trump is disapproved of by 70%. Of the three oversamples, this is the only one that lines up closely with the crosstab aggregate (Biden+8). Biden won young voters by 24 points in 2020, so it looks Trump is making large gains among the group.
But it’s not that simple.
Biden and Trump have a similar total disapproval rating, but the number of respondents who strongly disapprove of Trump is 61%. For Biden, it’s just 44%. This means Trump likely has a lower ceiling of support with young voters than Biden does – it’s hard to get someone who hates you to vote for you.
Additionally, young voters who disapprove of both Biden and Trump overwhelmingly prefer Biden to Trump. RFK Jr. actually wins this group, but like all third party candidates, his support is declining as the election gets closer. The combined voteshare in polls for RFK Jr. and Cornel West (a left-wing independent candidate) has been steadily decreasing. 6 months ago, it was 17.9%. Today, it’s only 11.5%. This raises the question of who RFK Jr.’s supporters will break for when they realize he can’t win.
https://preview.redd.it/zt0t5ptzhvzc1.png?width=763&format=png&auto=webp&s=fd1f7c717e66e750c57e76eaa617966708ebd222
Based on the approval ratings of Biden and Trump, and the “double haters” who already have chosen sides, it seems like the vast majority of young RFK Jr. supporters will go for Biden. His lead among young voters will only increase as time goes on. Of course, it wouldn’t be a bad idea to run ads like these to speed up the process.
Split Ticket also conducted a poll using a more normal method, an opt-in web panel. This poll had Trump doing much better with young voters than in their live text poll. So yes, some commonly used polling methods don’t work correctly!
Conclusion
Biden has been polling badly lately. He’s been trailing Trump nationally as well as in swing states. Polls say key parts of the Democratic base, black/Hispanic/young voters, are abandoning Biden in huge numbers. But when looked at closely, it’s not so simple. Other signs for Biden are pretty good. He’s been doing pretty well in primaries, and Democrats have been doing well in special and off-year elections. Polls are underestimating Biden’s support due to bad methodology and Democrats not answering polls. Oversamples show Biden doing fine with black voters, and mostly fine with Hispanic voters. The only group he really needs to work on is young voters, by trying to decrease RFK Jr.’s support.
So, 2024 won’t be a red wave where Trump wins big. But current signs don’t suggest 2024 is going to be a blue wave either, just another extremely close election like 2016 and 2020 both were. But there’s reason to believe Biden might outperform his 2020 showing despite that.
The American public is not very engaged right now, as there’s still seven months until the election, so Trump’s latest ventures with the legal system aren’t on people’s minds. When people tune in more, he can only get hurt from it. There’s also the massive fundraising gap between the two, which Trump is scrambling to close.
Here’s a prediction for how the election will actually go (margins are 20+, 15-19.9, 10-14.9, 5-9.9, 1-4.9, <1).
https://preview.redd.it/ufw3oxa2ivzc1.png?width=810&format=png&auto=webp&s=55a5dcc6c246cb34381165d211b17181717ef196

theres no r34 flair😭