Well, here we are in 2020. Joe Biden is an overwhelming favorite. 538 says Joe Biden is a 88% favorite to win, Economist 95%, the tipping point state is rated Lean D by all prognosticators. The median prediction on here is identical to 2016 + AZ. Everybody is in lockstep - Joe Biden will win overwhelmingly and there will be a blue wave that will sink Republicans in the Senate and allow them to expand their House majority.
I disagree. I think the presidential race is a toss-up, the Republicans will narrowly keep the Senate, and Democrats will retain a decreased majority in the House. I think we are likely in the midst of the biggest polling and forecasting failure in US electoral history - even if Joe Biden wins. I think the polls will end up worse - not better - than 2016. Certain pollsters - like Quinnipiac and Monmouth - should be shamed out of the business, but they won't be because they're propped up by liberal political prognosticators disguised as objective journalists to push a narrative that Democrats are winning by more than what will be revealed as reality on Tuesday (even Joe Biden's campaign manager said the public polls are "inflated").
I've never put my prediction out here that's this out of step with the groupthink on this website and the Election Twitter crowd. But it's what I think will happen based on other data mixed with my own assessment of the polls. I can be wrong - and if I am I will own it, but I think I will be closer to reality than most on here. I expect to be scolded, mocked, and insulted for posting this, because that's who you people are. It has become common place on Atlas for insecure election nerds to belittle others who don't view things the way they do and gang up with insults on those who dare to be vocal against the groupthink. It's why I've stopped posting for long periods of time this year, and that's why I'm done after 2020. I've never observed an environment so toxic to other people's predictions and opinions in the 7 years I've posted here, especially if you are in a tiny minority like myself.
What are those things that make me think Trump has a shot? Here are some of them:
1. Trump's historic party unity and turnout in the R primary (Norputh model)
2. Party registration trends in key battleground states towards Republicans (no, these are NOT just Republican-voting Democrats switching their registration - more on that after the election)
3. The undeniably strong economic comeback despite an artificial economic shutdown - GDP, unemployment, Housing, economic confidence, etc.
4. Social media and google search trends - yes, these actually are predictive.
5. The same polls that were accurate in 2016 and gave warning signs to Hillary - Trafalgar, Susquehanna, and another that shall not be named - are showing similar results to 2016 - albeit a little worse for Trump overall
6. Early vote data that suggests Democrats are underperforming their pace compared to the polls of where they "should" be at if they are to win overwhelmingly - and overwhelmingly Republican election day vote yet to come.
In addition, I've learned A LOT over the past few months about why the polls have overinflated Dem margins particularly in Midwest states over the past few months. I'll share those "secrets" with anybody who wants to know if they don't in a PM. All I can say is that it will continue happening as long as the parties keep trending in their current directions. And pollsters know that's it's become increasingly hard to get accurate polls, but I think many of them are content with trying to sell a narrative that Democrats are winning overwhelmingly, even if it's not true.
Nate Silver and the mafia that thinks like him, predictably, are ignoring these signs and slamming anything that goes against a blue wave - including a completely unprofessional war on Trafalgar. I hate to be the "in 2016" person but If you remember, Hillary was up by more in the battleground states in 2016 than the final averages, but Trafalgar and other right-leaning pollsters came in the last week and skewed the averages toward Trump. People ignored them and mocked them, called them garbage polls. That's happening again, the race is tightening and late deciders are disproportionately going to Trump. Maybe not in the polls you guys like, but I think it'll show itself on Tuesday.
Trump's approval is also at around 45-46%. I said months ago a 45-47% Trump approval would point to a toss-up election, and a 47%+ approval would mean advantage Trump. I expect Trump's approval without undecided would look like 46-47%, so my prediction below really isn't a stretch with the same standards I've had for months.
If Trump wins, I think you can expect any further trust in polling to collapse. In the event that does happen, one of my favorite things will be to watch the "experts" of the election - Nate Silver, Elliott Morris, Harry Enten, Rachel Bitecofer, David Wasserman, Larry Sabato, and others - meltdown and try and excuse their poor forecasting. Some of these people I don't think are bad people, but they are unbearably smug with their political prognosticating. Some of them in particular - Morris and Bitecofer especially, will be exposed as frauds.
Also, if Trump wins, I'm expecting that political nerds in the media and on this forum will start making excuses why Biden lost, if they don't go all out and call it a rigged and stolen election. Lack of enthusiasm, lack of in person campaigning, and lack of Hispanic and African-American outreach will probably be some of the things they'll point to, despite brushing off or outright denying that any of these problems existed for months. Joe Biden was a bad candidate, perhaps, despite saying for months he was the best candidate they could've put up and way better than the "hated" Hillary Clinton who enjoyed large approval ratings for years before running for president. I expect the election runup narrative to be completely whitewashed and contradicted by the reasons they come up with for Biden's loss.
If Biden wins, they will gloat about how the polls were right, conventional wisdom made sense, and that they have been vindicated, even if someone like me is closer to reality than the 350+ Biden wins regularly predicted. They will continue to overpredicted Democrats in the 2022 midterms and beyond. I also think you can expect the Republican party doom to start to take over the pundit discourse much like it did post-2012, (even though Trump will likely do better with blacks and Hispanics in 2020!). Trump has obviously had many tactical and strategic errors throughout this campaign, if he loses it'll be moreso a reflection on that than some kind of nonsensical notion that Republicans just can't win elections anymore.
Here's the thing, I expect if anything my prediction to favor Republicans. I've been following this stuff now for 4 cycles (starting in 2014) and have overpredicted Democrats every time with the exception of the House in 2018 where I shorted Dems by about 5 seats. In 2018, I thought the same thing, that if anything my prediction was Republican-friendly. It turned out to not be Republican enough in key Senate and Governor races. Here's my
final 2020 prediction:
Trump/Pence - 289 (47.0%)Biden/Harris - 249 (51.5%)
Safe: 100% certainty
Likely: >90% certainty
Lean: 70-90% certainty
Tilt: 50-70% certainty
Safe Biden: 196
Likely Biden: 14
Lean Biden: 12
Tilt Biden: 27
Tilt Trump: 41
Lean Trump: 60
Likely Trump: 63
Safe Trump: 125
State (rounded to nearest 0.5%)AZ: Trump 49.5%, Biden 49.0%
CO: Biden 53.5%, Trump 44.0%
FL: Trump 50.5%, Biden 48.5%
GA: Trump 51.0%, Biden 48.0%
IA: Trump 52.5%, Biden 46.0%
ME: Biden 51.5%, Trump 45.5%
ME2: Trump 52.0%, Biden 45.0%
MI: Biden 50.0%, Trump 49.0%
MN: Biden 50.0%, Trump 48.5%
NE2: Biden 49.5%, Trump 49.0%
NV: Biden 50.5%, Trump 46.5%
NH: Biden 51.0%, Trump 47.5%
NM: Biden 53.0%, Trump 44.0%
NC: Trump 50.5%, Biden 48.0%
OH: Trump 53.0%, Biden 45.5%
PA: Trump 49.5%, Biden 49.5%
TX: Trump 51.0%, Biden 47.0%
WI: Trump 49.5%, Biden 49.0%
Nationally I'm predicting a larger polling error than 2016 - about 3 points (which is what happened in 2012, so it's not that out of the realm). I don't expect many states to have polling errors that benefit Trump, although some like Nevada, Arizona, and Texas will be the most likely to. I think the average polling error in the Midwest will be 6 points. An electoral college landslide is more likely to happen with Biden, simply because of how lopsided his states are safe for him compared to Trump's, and that helps explain his large national popular vote lead. But I do think the EC will be close (under 350 for either side). In the event that a Biden victory happens with under 300 EV's, I still expect a lot of harassment, even if my assessment was closer than the common 413-125 predictions. So I'm being pretty ballsy here.
AZ - I was previously bearish on Trump's chances there, but recently there's been some okay polls for Trump there and registration trends in Maricopa and early vote looks better than I expected, so I now think Trump will pull off a narrow victory, but I can still easily see how Biden can win it. Will definitely be closer than 2016 either way.
FL - I've become pretty convinced of a Trump victory here, even if it's extremely narrow. To say that Democrats have been underperforming their expectations with early vote is an understatement. Long live the Florida Dems incompetence. Trump will undoubtedly do better in Miami-Dade this time, offsetting some erosion in perhaps Hillsborough or Duval.
GA - Looks to be closer than 2016, but still on a bad night for Trump I still see him winning the state by a Kemp/Abrams margin.
MI - I expect no bigger polling average error than in the state of Michigan, where polling other than Trafalgar has consistently had Joe Biden either up double digits or 8-10. I expect Biden to narrowly win back the state, always a tough feat for Trump to hold, in part due to his ties to the auto industry, however little that matters, it could be decisive.
MN - Polling has actually been consistent in showing Biden up 3-6 points, but even Trafalgar has Trump down here, and the demographics of the state are simply friendlier to the Democrats than Wisconsin and Michigan. I do expect we'll see Dem erosion in the iron range, however.
NV - Previously thought it was gone for Trump - but good early voting by R's and bad Latino ground game by Dems has nudged it into play a little. And if we expect Trump to do better with Latinos and non-college whites - no better state where that would show up than Nevada.
NC - Early vote isn't good for Dems here, and with how Georgia and Florida will go, and with how wrong the polling has been here the past few cycles, I'm fairly convinced Trump is a small favorite.
PA - The Dems ran up the score with early voting here, but it looks like there'll be a large election day turnout - which of course will not be beneficial to Democrats. They do have the 3 day extended counting though. This is no doubt, the most important state in this election. If Trump wins it, he wins, if he loses it, his path is more difficult.
TX - Early voting is already all of 2016's vote right now, the profile looks identical to those who showed up in both 2016 and 2018. Even if Biden does as well as Beto with indies, he'll still lose. That's why I'm more confident in Trump's chances even though I think it'll still be a decently close race.
WI - oh boy, if this isn't a bold call I don't know what is. I think Wisconsin and Michigan will have some of the worst polling averages this cycle, competing with Ohio, Iowa, and Maine. Dems will run it up in Dane, no doubt, but I think they will have trouble with black voters in Milwaukee again. Biden may take back a county or two north of Madison but I think Kenosha Co will go to Trump and he'll improve further in the Northwoods and get back some third party WOW vote. Overall, a wash. Wisconsin's a state I can see Trump winning by 3 or Biden winning by 4, it'll really just depend on how those swingy rural independents go at the end.
One of the biggest stories coming out of election night will be the trends - the changing coalitions. Trump looks like he'll do better in Hispanic and black areas, particularly with men. Biden will do better with white women in places like Chester County PA, some of whom held their nose for Trump last time. Overall, I'm still predicting Trump doing worse overall than 2016, but I think some of the places he'll improve in will shock the Atlas echo chamber harder than even 2016. He'll do even better in places like Erie County, improve in heavily black urban areas by small amounts, but continue to do worse in white urban and suburban areas. Most people will just expect Trump do worse everywhere. My prediction is that it will not be a uniform swing, it'll be another coalition shift (but less so than 2016).
Overall, I think the electorate will look more R than it did in 2016 but independents will swing against Trump (this is common in incumbent elections). If Trump loses, it'll be because he lost too many independents in the rust belt. There's very little consensus among pollsters on how exactly they'll break. If Biden loses, I think it'll be because he either lost too many Obama/Hillary black voters in these key states or they didn't turn out.
So that's it. Either a Trump upset win or a narrow Biden win. Not confident in the winner. Before the pandemic, I viewed Trump as a narrow favorite. During the summer, I viewed him as a moderate underdog. Since August, I've viewed the race as a toss-up slightly leaning to Biden. Now I'm viewing it as a total toss-up but Trump winning as my final prediction.
