Prediction: Biden wins by a lot more than people are expecting [View all]

[Dr. Jack]

Let me say right off the bat, I of course don't have any secret insight or super special expertise on elections so take what I am about to say with a grain of salt.

With that said, I am going to predict that Biden will win by quite a bit more than people are expecting. I'm not basing this on claims like you see from Republicans that the polls are fake or "skewed" in some way. Instead I think the polls showing Biden hovering around 14 or 15 points nationally might prove to be the accurate ones and polls showing him around 8 or 9 are playing it too safe. Here is my reasoning:

1. Following 2016, many pollsters changed up their polling methodology to weigh heavier towards non college educated whites, or in other words Trump supporters. They did this of course to not underestimate Trump in places like Michigan, where they got it a bit wrong in 2016. I think however, that 2016s turnout models don't apply to 2020. 2016 was an anomaly and polls are somewhat overestimating the electorate being more trump friendly than it will be.

2. Nate Silver is suspicious that right wing leading pollsters like Trafalgar and Rasmussen are fucking with their numbers to force Trump's averages in the polls to be closer to where they want it to be. He doesn't think this is 100% malicious or that they are lying. Instead he seems to believe that these right wing pollsters want desperately for Trump to win and they are unintentionally finding evidence to prove their point. Also he thinks they are paranoid and that other non trump friendly pollsters are in the bag for Biden so they are releasing as many as their extremely generous polls to trump as often as they can to change the averages in Trump's favor. If right wing polls are indeed very inaccurate, which is looking pretty likely, and are way overestimating Trump then Biden's average lead could very well be understating democratic support.

3. Nate Silver also says he thinks his model may very well be too conservative right now. He said he doesn't entirely buy that Trump actually has a 12% chance of winning and that his model might be adding too much uncertainty to the projection which favors Trump.

4. The massive turnout this year will not be people turning out for Trump. Yes, Trump is going to get tens of millions of votes and Republican turnout will be high. However, experts think turnout will be between 6 million and 20 million votes higher in 2020 compared to 2016 (most likely about 12 million more). While I don't doubt some of those are Republicans, I think a significant majority are coming out for Biden. Between the pandemic, the economic collapse, the increased flirting Trump is doing with fascism and white supremacy, I believe a lot of people that sat on the sidelines in 2016 are now desperate for a "return to normalcy". I believe most of those extra 12 million votes are people that just want to be done with Trump and never have to deal with him again. And you have to ask yourself, what has Trump done to increase the size of his base or electoral coalition since 2016?

5. The massive amounts of money that Biden and the Dems are raising compared to Trump's campaign and the Republicans being nearly bankrupt shows that there isn't some hidden enthusiasm for Trump. Again, going back to point 4, if those extra 12 million voters this year were coming out in large part because of a "hidden Trump vote" then you would see him bringing in money like crazy and his rallies wouldn't be just 35 people all crammed together to make him look more popular than he is. Trump has his fanatical base, I don't deny that, but beyond that 35% of the country, no one else seems to be excited about the idea of a 2nd trump term.

6. Early voting turnout. Clearly the pandemic is affecting this quite a bit. Some people who would have turned out on election day during a normal year but I don't know if that entirely accounts for what appears to be a massive increase in turnout we are seeing in early voting amount democratic leaning demographics. Early voting among people under 30 for example is up some 1000% in places like Michigan and Pennsylvania. First, I highly doubt they are trump supporters. Second, if people were happy with Trump or fairly "meh" on the democratic candidate this year (like in 2016) why the high turnout among Dems and Biden voters? If people weren't all that motivated to get rid of trump or wanted to keep him around, why bother voting early in such massive numbers? Why would people wait in a 8 hour line a month before the election to cast a ballot? If people loved trump and hated Biden they would what's the rush to vote for Biden?

7. Congressional district level polls are showing massive Biden leads. I left this point to the end because it is the least certain but might prove to be a critical data point that people are missing. Nate Silver talked a bit about this on their Friday podcast. He said that polls on the Congressional district level, some of which ask about presidential candidate preference while polling house races (those polls are not included in state or national averages therefore aren't taken into account by much of anyone right now) are showing a very different story on the election. Specifically a massive Biden lead. Far larger than the national or state polls. Since these polls aren't used in averages then they aren't susceptible to Trafalgar and Rasmussen fucking around, like I mentioned in point 2. District level polling in presidential elections is somewhat new and experimental in the context of polling models like 538 but there is some evidence that they might actually be far more accurate than national or state polls. If that is the case Biden's lead might actually be closer to 14 or 15 points.


Again, don't take what I am saying as gospel but I from my personal prospective, I think Biden is going to end up winning by a lot more than most people are expecting.