Shadow1980 said:
Four weeks to go until Election Day. While things can still change quite a bit, looking at the polls as they stand now, things are definitely favoring Biden.
First off, I wanted to address approval ratings. Trump has been the single most consistently unpopular president since approval ratings have been a thing (at least since the 1950s). While some presidents have had lower lows, all of them had times where they were more popular than unpopular, their net approval rating being in the positive. But the running average for all polls on Trump's approval has had him underwater, and he has seen the most stable approval ratings of any president ever. His approval rating has consistently averaged in the lower half of the 40s, while his disapproval rating has consistently remained above 50%. Looking at state-level approval ratings from Civiqs, Trump's approval has been net negative in most competitive states, especially in the old "Blue Wall" states he unexpectedly and narrowly won in 2016 (Morning Consult also used to offer state-level approval, which they ceased doing so this past March, but they showed similar results, with Trump's net approval rating being poor in most swing states).
If there's one thing I've noticed about approval ratings, it's this: no president with net negative approval ratings just ahead of Election Day in their first term has won re-election. The last single-term president was Bush, Sr. His approval ratings were net negative ahead of Election Day 1992. Same for Carter in 1980. LBJ (only elected once; his first term was partial and doesn't count) was so unpopular he effectively got primaried out of office in 1968, withdrawing from the race after the New Hampshire primary and thereby failing to win the party's nomination for president despite being the incumbent. Ford and Truman may or may not be exceptions to this rule (Ford may have been net positive but still lost, and Truman may have been net negative but still won), but insufficient polling data exists to determine for sure; the last Gallup poll before both the 1976 and 1948 elections were in June of those years, far too early to give an accurate snapshot of their actual popularity in October/early November.
Granted, that we only have four Presidents since WW2 fail to win re-election (and only three with adequate polling data on their job approval) is a small sample size, but given their generally mediocre to poor approval ratings, it doesn't bode well for Trump.
Looking at the actual election polls themselves, the national polls themselves have consistently shown Biden with a very comfortable lead over Trump, with the polling averages consistently having Biden at around 50% since June and Trump at around 42-43% since early August. And polls over the past week since the debacle that was the debate show an ever so slight shift in Biden's favor.
But the election is of course not one big national election but 51 separate elections. So, let's look at the state-level polling.
Here's a map of all the competitive states, defining such states as ones where polling averages for this year have generally been within 5 percentage points and/or where the state has changed hands in the past three election cycles (i.e., 2008, 2012, & 2016):
So, let's do a rundown of each of them:
Indiana
Just to get this one out of the way, Obama's 2008 win in Indiana was clearly a fluke. Outside of that one election, it has been a solid red state. Trump is going to win this one. He won it comfortably in 2016, and polls show him way in the lead this year.
Arizona
This state appears to have been trending purple in recent years. While Trump won it in 2016, it was not a huge margin. He did worse there than any Republican has this century, and it was also one of the few states Hillary did better than Obama did in 2012. In the 2018 midterms they elected a Democratic Senator for the first time since 1988, and the special Senate election this year looks to favor the Democrat. Most polls have had Biden ahead, and he has maintained a lead over Trump in the polling averages, but not a huge one. Biden stands a good chance of picking up this state, but not by a huge margin. If Biden wins Arizona, it will be only the second time a Democrat has won the state in the past 50 years, and the first time one has done so since 1996.
Georgia
Like Arizona, my state of origin (I was born in Augusta) shows some signs of moving to swing state territory. Trump did worse in Georgia than any Republican has this century, while Hillary did marginally better than Obama did in 2012. The Senate races there appear to be competitive. And the polls haven't favored either Biden or Trump. Some show Biden ahead, some show Trump ahead, some show a tie. The polls show this state is a pure toss-up, but given Trump's approval ratings in Georgia are net negative, it's possible that Biden may stand a slightly better chance of winning than Trump.
Texas
While the Lone Star State has also shown signs of become increasingly purple, the fundamentals of this state still favor Trump. His most recent approval rating average is about at the break-even point, and most polls have shown Trump ahead by a few points, though in some he ties or is a point or two behind Biden. I'm going to rate Texas as "lean Trump," but it will almost certainly be the closest race there in a very long time. And as the demographics in the state continue to shift as more Latinos are added to the ranks of registered voters, Texas will likely be the state to watch in the coming decades, perhaps replacing Florida as the biggest prize to be fought over.
Iowa
This state swung hard for Trump in 2016, despite Obama winning by comfortable margins in both 2008 and 2012. Trump did better than any other Republican has this century, while Hillary did worse than any Democrat has in the same time span. However, the Democrats did make some gains in the 2018 midterms, winning a majority of the statewide vote and picking up two House seats in the process, though Iowans did narrowly re-elect their Republican governor. Most polls have shown Trump slightly ahead of Biden, but the running average is very close, and Trump's approval rating is net negative in the state. Iowa will be competitive, but I think Trump will edge out a narrow win.
Nebraska & Maine: NE-2 and ME-2
While there's no question that Trump will win the statewide vote in Nebraska and Biden will win the statewide vote in Maine, these two states are the only ones that split their electoral votes by congressional district (the statewide winner wins two electoral votes, plus one for each district they win). Obama picked up Nebraska's second district in 2008, and Trump won Maine's second district in 2016, the only time in recent electoral history where the overall statewide vote winner didn't also win every district. Polling data consistently shows Biden with a decent lead in NE-2, while ME-2 is harder to call, with some polls showing Trump in the lead while others show Biden in the lead. I'm going to rate NE-2 as "lean Biden" and ME-2 as "toss-up." If both states split their vote, they will be minor victories for the candidate that doesn't win the statewide vote, but every little bit helps.
Florida, Ohio, and North Carolina
These have been some of the most competitive states of the past several cycles, and it will probably be very close this time as well. Florida is the most important of these as it has more electoral votes than any other swing state, and is a must-win for Trump. However, most recent polls have shown Biden ahead by a few points, with the running average currently having Biden ahead by over 3 points. This is not insurmountable for Trump by any means, but if he doesn't make some gains in the polls before Election Day he's not likely to win it this time, and it is arguably a must-win for him. Right now, I'd rate Florida as "lean Biden," but only slightly. Except for the 1992 election, it has voted for the overall winner in every election since 1964, making it a "bellwether state." Also, it has only voted against the national popular vote winner twice since 1964, once in 2000 and again 2016, both elections notable for having a third-party candidate serve as a spoiler.
North Carolina is more competitive. While Biden has narrowly won most polls (usually only within the typical margin of error), there's still been some polls showing either a tie or a slight Trump lead. This one is currently too close to say how it will turn out.
Ohio meanwhile appears to be a pure toss-up. While the state did, like Iowa, swing hard for Trump in 2016, and a slim majority of Ohioans voted Republican in the midterms, so this does seem to indicate Ohio could favor Trump. However, Trump's approval rating in the state is a few points into the negative territory, and the polls have been all over the place, showing a tight race with no clear leader. Like Florida, Ohio has been notable for being a "bellwether state," voting with the overall winning candidate in all but one election since 1944 (Nixon beat JFK there in 1960). Will it maintain this trend, or will it fail to go for the winning candidate for the first time in 60 years? This one is also too close to call.
Michigan, Wisconsin, & Pennsylvania: The "Blue Wall"
These are arguably the most important ones. Trump won all three of these states by razor-thin margins, despite Michigan and Pennsylvania having been a reliably blue state since 1992 and Wisconsin having been as such since 1988. Granted, the margins of victor for Democratic candidates has sometimes been slim, but they've been consistently blue. This seemed to dispel the notion of the "Blue Wall" being unassailable. But did the Blue Wall really collapse, or was it left unguarded with the gate wide open?
Again, Trump only narrowly won those states, arguably in large part because of a sufficient number of progressive voters voting third-party (more on this later). While he did outperform Romney in all three of them, he didn't do any better than Bush did in 2004. And looking at both the results of the midterms and all of the polling data for the present election, it's looking like the Blue Wall's failure was only temporary. Biden's polling average has been at around 50% in all three states since at least the summer, with the overall lead being well outside the margin of error. State-level approval rating data also indicates that Trump is deeply unpopular in all three states.
If Biden wins all three of these states, he wins the election, even if he loses every other swing state. Trump needs to win at least one of them, and if he loses Florida, he needs to win all of three of them. This is not a likely outcome, and if things continue to remain favorable in the old "Blue Wall" for Biden, then his odds of winning go from "likely" to "nearly guaranteed."
Conclusion
"But wait," you might ask. "What about the 2016 polls showing Hillary winning?" Yes, it is true that Trump did win despite the polls showing Hillary as the favorite. For what it's worth, the national polling data was quite accurate when gauging results of the national popular vote. As for the state-level data, there are several significant caveats.
1) Hillary wasn't exactly doing well. While she was in the lead in many swing states, her actual numbers weren't terribly impressive, including in the "Blue Wall," which is where I want to focus for now given their importance. She wasn't getting anywhere close to a majority in either Michigan, Wisconsin, or Pennsylvania, struggling to get to 50%, and her running average tending to sit in the mid 40s. That was worse numbers than any other Democratic candidate this century (incidentally, the final running average for Hillary was close to her actual vote share in those states). Meanwhile, Biden has, as mentioned, performed much better in the "Blue Wall" states, with a running average of around 50%. Overall, he's performing as well if not slightly better in those three states than Obama did in 2012, and Obama won those states by comfortable margins that year.
2) There was a significant contingent of undecided voters in 2016. In many swing states, upwards of 15% of poll respondents didn't pick either candidate. There is limited data to suggest a sudden surge for Trump in the final days, and exit polling data shows that people who made up their mind within the final month before the election favored Trump and were more likely to vote third-party. This time around, the number of undecided voters is much lower. Not only is Biden outperforming Hillary by quite a bit in terms of actual numbers (and not just lead), but there's a lot less wiggle room for undecided voters to swing the results. Furthermore, polling data suggests undecided voters are favoring Biden over Trump. That could change between now and Election Day—they are apparently undecided, after all—but Trump can't rely on last-minute decisions to vote for him to the degree he could last time.
3) Hillary was incredibly unpopular. Her favorability rating was among the worst of any major-party candidate in recent history. Trump's was actually worse, but we're talking about an election between two very unpopular candidates. Democratic enthusiasm was low, and a lot of progressive voters, including in key swing states, refused to vote for Hillary, either staying at home or voting third-party (speaking of which, see point #4 below). Biden meanwhile, though not exactly Mr. Popularity himself, has much better favorability ratings than Hillary, while Trump is still deeply unpopular. Democratic enthusiasm in general seems improved, with the total votes cast in the primaries being quite a bit more than they were in 2016, even in states that didn't switch from a caucus to a primary, and even in states that didn't have a primary until after everyone except Biden dropped out. This is likely because now that Trump-as-president is an actual reality and has been since January 2017, Democratic and Democratic-leaning voters are much more motivated because they really, really want to vote Trump out of office.
4) Third-party candidates over-performed in 2016. And progressives were more likely to vote for them. According to exit poll data, self-described progressive/liberal voters have always been more likely to vote third-party than conservatives. 2016 was no different, with polling data indicating that on average twice as many progressive voters voted third-party than did conservative voters. This by itself goes a long way to explaining Trump's razor-thin margins in several key states. If progressives voted third-party only at the same rate as conservatives, Hillary likely would have won, if only barely. This time around, third-party candidates are nowhere near as popular. Furthermore, polling data shows that third-party voters (as well as voters who sat out in 2016) favor Biden by a comfortable degree. This means the odds of the "spoiler effect" rearing its ugly head are greatly reduced, similar to how Nader served as the spoiler in 2000 but was a complete non-factor in 2004. As mentioned before, progressives are now almost certainly far less likely to want to risk a repeat of 2016, and are increasingly desperate to ensure Trump doesn't get a second term, especially once the prospect of losing the Supreme Court for a generation materialized.
Trump is incredibly vulnerable. We haven't seen a president be this unpopular in a very long time. Things have started to reach a critical mass this year, especially with the COVID-19 pandemic. A majority of voters think he has done a poor job with this crisis and are looking at how he's doubled down on downplaying the pandemic, especially after he has himself become infected yet refuses to take even the most basic precautions to avoid spreading the disease to his staff.
Odds are, barring some electoral shenanigans (e.g., trying to get mail-in ballots needlessly rejected), Biden is going to win this. If I had to put out a map right now showing the likely outcome, it would be this:
Well, that took a lot longer than I had hoped. That's all I'm going to write about on the issue for now. I may revisit things about a week out from the election if things have changed substantially.
|