The entirety of this post considers what massive turnout could mean. This is purely theoretical if we don’t actually get out our vote. We must maximize our turnout. If we do maximize our turnout and we end up seeing 154 million votes cast for president this time, then there are implications, in my view, to what that means for the election. However, the discussion is meaningless unless we get out our vote. I hope that this encourages us to get out our vote because that 154 million votes being cast has, in my mind, gigantic implications for the election and not just between Biden and Trump but also down ballot especially in the US Senate.
I was listening to a podcast by 538 and Nate said several things that should be extremely encouraging to us. The following is a very close paraphrase of what he said. The national polls might not have tightened that much, the state polls have more of an impact upon the model than the national polls and they show a slightly smaller lead than the national polls. The state polls suggest an 8 to 9 point lead nationally. There are more of the junk polls that do state polls, pollsters that they (538) are not crazy about like Trafalgar. They discuss herding which is when pollsters release results that they intend to converge to the mainstream of polling. Online / IVR polls (some that are less good have strong assumptions and massage their data because of a lack of a random sample) , quasi partisan polls (not following 538’s own strict definition) which are polls where it is clear who they favor and it is clear that they want to influence the narrative (by their write up) and they have extremely strong priors (by this Nate means prior assumptions) and the third type of herding is high prestige polling and you see this in their penultimate or last poll and they avoid publishing an outlier or weight it or manipulate it in order to get their results to be close to the consensus.
They discuss if house district polling is more reflective of the race because it avoids herding. Mostly they show very good results for Biden, an election in which Biden is leading by 12 or 13 points. This is true even if they are more partisan. National polls show a big lead, state polls show a more narrow lead, and district polls show a gargantuan lead . A true 5 point lead would show itself by presenting a Trump plus one poll or a Biden only plus one on the lower side and an 8 point lead at the top. Herding may be impacting the model. The very last polls may be less accurate due to herding.
538 has updated its turnout, revised upward because it looks like it could be huge. Their numbers, their projection is now that 154 million votes will be cast , the 88th percentile is between 144 million and 165 million votes — in 2016 the turnout was 137 million votes. Voting laws have, in general, been made more liberal and voting has been made easier. There are more ways to vote than ever Nate says. This doesn’t affect their model’s margin projection because they relied upon the pollsters. So, they are not adjusting the margin based upon the large turnout.
That to me means that they are very likely to be underestimating the Biden lead . Think about 154 million votes cast for president. Let’s say 4 or even 5 million votes are cast for third party or write in which is almost as large as 2016 and seems unlikely to me. We are still left with 149 to 150 million votes cast for the two major party candidates.
There, in my mind, is a definite limit to how many Trump voters there could be and it is much lower than how many Biden voters there could be. I have a hard time envisioning more than 65 million votes for Trump. I just don’t see that many Trump voters existing. This would mean that there would be 85 million Biden voters. At most, there are 68 million Trump voters. I think this is a hard cap. This would mean that there would be 82 million Biden voters.
So, let’s consider an 82 million Biden vs 68 million Trump election. This is an election that is not close. This is an election in which we likely win Georgia and Texas and we definitely win Florida and the election gets called by the networks on the night of 11/3 . This limits the damage Trump and Trumpsters can do in trying to cast doubt upon who won. That would be huge ! Next, think about what that would mean in the US Senate. In order to expand the Supreme Court (which to my mind is a non-negotiable — it must be done ) , we must have enough members in our caucus to be able to lose a few terrible conservative members’ votes. In this kind of election, I would say that it is likely that Jaime Harrison wins and MJ Hegar wins and Jon Ossoff wins in addition to Cal Cunningham and Theresa Greenfield and Sara Gideon and Mark Kelly and we end up with 53 or at the very outside 54 seats. A huge blow the doors out turnout especially benefits us in Texas and in Georgia and somewhat in North Carolina and maybe South Carolina as well (there won’t be a huge turnout there for Lindsay Graham — so this to my mind potentially benefits Jaime Harrison). That’s because we have latent votes and with suburban women voting D in huge numbers and an expanded and more diverse electorate which this would portend in these states. In very republican states, it may not necessarily benefit us. It also won’t help us that much if these votes are all in very blue states like California. Reverend Warnock will likely end up in a run off in January which could be tough, but definitely winnable due to the quality of the candidate. I think Dr. Gross and Dr. Bollier have outside chances of winning in a large turnout election because Alaska likes independents and it will be close and Kansas is a well-educated state considering how red it is. Governor Bullock will have to win on his own and I don’t know that a huge turnout helps him.
Now, 84 million votes cast for Biden gives him 53.2% vs 44.2 % for Trump with 68 million votes. Now, that, to my mind, is the worst case scenario . (notice that the projection seems to be based upon Texas and states other than California — if the increase were just in California, that would be a terrible disappointment but the early voting that is leading to this (including the fact that we are doing a better job at getting out our sporadic or infrequent voters out than they are) is based upon other states. Thus, it doesn’t seem to be the case that the huge increase in turnout is mostly or in large part from an increase in California).
On the other hand, 85 million for Biden vs 65 million for Trump would be 55.2% for Biden and 42.2% for Trump , a 13 point win. Now, to my mind, this is far more likely to be the outcome if we get a 154 million vote election. Trump only got a little over 62 million votes last time. The republican party has only shrunk since he won in 2016 and he is the reason why it shrank. He is vastly unpopular with broad swaths of the electorate. His handling of the novel coronavirus pandemic response is horrific and the numbers are only getting worse . We added nearly or more than 80,000 cases in one day yesterday alone, a record. Fewer Americans are employed now than were employed at the beginning of his term, the very first American president to accomplish that. Therefore, it seems clear that he is unlikely to increase his original turnout by much, if at all. Frankly, I have a hard time envisioning him increasing his turnout from 2016 at all just given the mismanagement of the novel coronavirus and its impact in deaths and on the economy. Nevertheless, this 65 million number increases his 2016 turnout. So, it is still somewhat on the conservative end.
Please remember that we won’t see a total of 154 million votes cast listed (if it does happen) for perhaps a month because absentee votes will trickle in. So, don’t say that this is wrong because we don’t see it on election night or even a few days after election day proper. It may take a week before we are close to this. Wise political commentators will not say that the turnout was much smaller than anticipated because the numbers are not immediately seen on election night. Anybody that says that should … well should be ignored.
Last point: Nate thinks that 538 accepts a lot of terrible junk republican pollsters but thinks that this in a way (I interpreted it) is better certainly for Biden because it is almost the worst possible outcome for Biden and underestimates Biden’s numbers. I kind of like that. Based upon state polling which they say has a lot of junk republican polling, the lead may look like it is only a little over eight points. However, it is probably at the low end of Biden’s likely outcomes.
Sunday, Oct 25, 2020 · 3:50:38 PM +00:00
·
Dem
Now at 59 million. MRA NY did some arithmetic. 137 million total votes in 2016 -47 million early votes in 2016= 90 million votes on election day in 2016 . So, Considering with 9 days left and not even a current count, we are at 59 million early votes, we should get to 75 million early votes easily by 11/3 and then another 80 million votes on election day (which would be a significant drop off) would get us to 155 million votes. So, we can get to 154 million or more votes total while having a significant drop off of votes on election day.