Thursday 29 October 2020
We are now less than one week away from the 2020 Presidential election and so far it has been a pretty uneventful and traditional campaign... nah just kidding, it's been every inch as much of a muck fight as one would have expected. The first term of Donald Trump is coming to a fittingly chaotic conclusion. The question now is whether he will in fact win a second term, or be replaced by the challenger Joe Biden. In this post I will attempt to shed some light on the likely outcome and make a few final predictions.
Introduction
To say that this has been an unusual election would be to put it mildly. Donald Trump is not a usual President and even as an incumbent he has seemed intent upon a "burn everything to the ground" approach to campaigning. He began the year as one of the most divisive and least popular incumbents in modern American history, freshly impeached for high crimes, amid a backdrop of near constant scandal and criminal investigation. This was before the pandemic even started.
What to expect? First let's begin with the fundamentals. An incumbent President usually does win re-election. In the last hundred years, only four out of eighteen Presidents have failed to win re-election on attempting to do so. Incumbency advantage is very real at the Presidential level. The economy is weak at the moment, but fortunately for Trump he seems to have been attributed only minimal blame for this. On top of this, the nature of the electoral college system appears to give Trump something of an advantage, to the point where he was able to win in 2016 despite losing the popular vote by millions of votes. In an ordinary year, you would expect a sitting President in this position to be a comfortable favourite for re-election, perhaps even as much as a 75% favourite. But Donald Trump has been an unusual President at the best of times. He was hugely unpopular even before the pandemic hit, and routinely trailed Biden in the polls. Then the pandemic did hit, Trump (rightly) took massive criticism for his response, and his polls sunk even lower to a point from which they have never really recovered.
That we haven't even mentioned the ongoing foreign election interference (mainly Russia but also allegedly China and Iran) speaks to how completely insane this election already is. That being said, I don't consider this to be a significant cause of uncertainty as to the election result, simply because any impact of these efforts, by their nature, one would expect to already be represented in the polling.
But to hell with that, we're going to predict anyway. Now let's give you the summary verdict that you came here for: The Ephemeric predicts Joe Biden to win the Presidency and the Democrats to hold the House of Representatives and retake the Senate.
Presidential Election Verdict: Joe Biden Elected President
Predicted Electoral Map: Joe Biden (D) - 313, Donald Trump (R) - 225.
The above map is based on data from a variety of aggregators, including Pollster, Fivethirtyeight, and analysts including Sabato and Cook Political, and shows the expected electoral map. The rest is pretty self explanatory: dark blue represents safe Democrat wins, light blue leans Democrat, grey is toss up. Meanwhile on the other side, light red to dark red represents lean to likely Republican.
House of Representatives Verdict: Democratic Majority
Current House Map: Democrats - 232, Republicans - 197.
Predicted House Map: Democrats - 240, Republicans - 195.
Approximate Net Change: Democrats gain 8-12 seats.
This is going to be the shortest part of this article. Very few analysts consider the Republicans to have any chance at all of taking back the majority in the House, and I concur.
Current Senate Map: Democrats - 47, Republicans - 53.
Predicted Senate Map: Democrats- 51, Republicans - 49.
Approximate Net Change: Democrats gain 4 seats.
Key states to watch: AZ, ME, CO, NC, IA, GA, KS, MT
If there has been one silver lining for the Republican party these last two years, it is that they managed to retain control of the Senate against the blue wave of 2018. This was not a surprise, even with a national situation that was highly favourable to Democrats that year. The Senate map in 2018 was one of the most lopsided in recent memory, very favourable to Republicans with Democrats defending several seats in deep red states. The 2020 map, in fairness, is also a decent Senate map for Republicans, though not nearly so much as in 2018. The combination of a more favourable Senate map plus strong national conditions make Democrats a slight favourite to win a majority of the Senate.
So there's a good six or so close races here, with Democrats only really needing to pick up one or two. All else being equal with such a number of close races, historically you would expect them to prevail in at least a few of them. Democrats will feel good about picking up the additional two seats they need.
Conclusion
So there it is. By no means a foregone conclusion, but the polling right now strongly suggests an election night that results in a Democratic trifecta, perhaps as much as a 7 in 10 likelihood. The unique conditions of 2020 and unprecedented threats being made against the democratic process are certainly setting a national mood of uncertainty. But for all the drama and theatricality, more often that not these things end up going more or less as expected. I understand the doubt, but ultimately expect the polls to be proven correct.
Despite all the drama the polls have painted a pretty consistent and unambiguous picture. Biden leads, and pretty comfortably. His current national polling average of around 9-10% would be one of the largest margins of victory in the popular vote in modern history, far greater than Clinton's 3% popular vote victory in 2016 and greater even than Obama's landslide 7% victory in 2008. The state-level polling suggests a slightly closer race, with an implied national margin more in the order of 7-8%, but this is still pretty substantial by historical standards. The polls are pretty clear, Biden is ahead, and in any other election year this would be looking like a foregone conclusion. But from the lack of debates, to the global pandemic, to the attempted kidnappings and literal threats to overrule a democratic vote... this has not been a normal election.
So why isn't this a foregone conclusion despite Biden's commanding lead in the polls? First let's address the elephant in the room, the 2016 election. In 2016 the polls also pointed towards a Trump defeat and yet he won. If the polls were wrong in 2016, couldn't they be wrong again in 2020?
This idea that the polls are wrong or that they can't be trusted is something of a myth, albeit a persistent one. In 2018 the polls were pretty much spot on. Meanwhile the polling error in 2016 was nowhere near as significant as is commonly believed, and should not have come as a surprise given the high level of uncertainty being shown in the polls at the time. Hillary Clinton's final polling average in 2016 had her at a 4% national advantage and she won by 3%, pretty accurate indeed. The error at state-level was a little worse (WI, MI and PA in particular), but can be easily explained by the high level of uncertainty evident in the polls. While Hillary did lead in the state-level polling, she was still only polling in the mid 40s, with undecided voters numbering in double digits. That implies a high degree of uncertainty in the polls that people simply missed or ignored. Even with this uncertainty, the polling error was still mostly within or close to the margin of error. The reality is not so much that the polls were bad, but that people were simply misreading them and far more confident in Hillary's narrow polling leads than they should have been.
That is the big difference this year. To put it quite simply, the 2020 polls don't look like the 2016 polls. Biden is not polling in the mid 40s, he is polling at or above 50% nationally and in key states. Undecided voters are basically non-existent. Biden's margins are bigger than Hillary's, the polls show far fewer undecideds and thus less uncertainty. For Trump to win at the polls he needs a much bigger polling error than we saw in 2016.
This is not to say that Donald Trump does not have a chance. Just that the main uncertainty in the race comes not from the polls, as is commonly claimed, but from other factors. The biggest such factor, by far, is Covid-19.
The truth is, no one really knows what an election this year is going to look like. There has not been anything quite like this in recent memory, and we can only guess as to how, if at all, the pandemic will affect voting. We simply don't know whether or how the pandemic will affect turnout on election day, nor if such an impact would materially benefit one side over the other.
Nor, unfortunately, can we rule out a significant impact from various electoral shenanigans. The Trump regime's attempt to sabotage the postal service has been the subject of much criticism, and he has been completely open about his willingness to challenge the election results in the court if need be. Trump has stated on more than one occasion that it was important to have conservative justices on the Supreme Court to ensure he could swing a close election. Reportedly, there is at least one state government considering legislative manoeuvres to flat out overturn the vote if need be. Then there is good old fashioned voter suppression. This ranges from removing voters from the registers without notifying them, to reducing the number of ballot drop-off boxes in Democratic-leaning districts, to voter intimidation.
Perhaps most troubling, Trump's politicisation of mail-in voting has created a dynamic where the mail-in vote leans heavily democratic and the in-person vote is expected to lean heavily Republican. It is not beyond the realm of possibility that the President could hold a lead based on the election day in-person votes and then challenge in the courts to have states stop counting, or dismiss entirely, the mail-in vote.
That such nightmare scenarios can't be ruled out is a damning indictment of American democracy. But ultimately I doubt that anything of this nature will have enough of an impact to truly swing the election. Many mail-in votes are cast days or weeks before election day and for a lot of states will be counted even before the in-person votes. Nor is it particularly clear that voter suppression efforts will really work, or affect Democrats so disproportionately as to materially change the result (especially if such measures are targeted at in-person voters that, as we have already discussed, are expected to lean Republican). The fact is that every election has these theories and stories of purported efforts to "rig" the election, and they usually amount to very little. Trump is using these inflammatory claims to depress voter turnout, Democrats are using it to fire up their base. Barring something truly unprecedented, I expect the election to be determined by the vote, as always.
That we haven't even mentioned the ongoing foreign election interference (mainly Russia but also allegedly China and Iran) speaks to how completely insane this election already is. That being said, I don't consider this to be a significant cause of uncertainty as to the election result, simply because any impact of these efforts, by their nature, one would expect to already be represented in the polling.
To be clear, none of this is to suggest that Trump does not have an ordinary electoral path to victory. He does, albeit a narrow one in the area of 20% (don't kid yourself, 20% chances can and do happen!). But increasingly it is looking as though his re-election chances depend on a combination of inherent pandemic uncertainty or anti-democratic shenanigans, and that is not a good position in which to be.
So this is the state of the race in a nutshell. Joe Biden leads and, barring something truly unprecedented, appears likely to win. But this is a year of unprecedenteds. Despite what anyone tells you, there remains enough uncertainty that no prediction can be called truly sound.
But to hell with that, we're going to predict anyway. Now let's give you the summary verdict that you came here for: The Ephemeric predicts Joe Biden to win the Presidency and the Democrats to hold the House of Representatives and retake the Senate.
Presidential Election Verdict: Joe Biden Elected President
Predicted Electoral Map: Joe Biden (D) - 313, Donald Trump (R) - 225.The above map is based on data from a variety of aggregators, including Pollster, Fivethirtyeight, and analysts including Sabato and Cook Political, and shows the expected electoral map. The rest is pretty self explanatory: dark blue represents safe Democrat wins, light blue leans Democrat, grey is toss up. Meanwhile on the other side, light red to dark red represents lean to likely Republican.
First thing's first. This is an unambiguously bad map for Donald Trump. He could win all of the toss up states above and still lose the election. He could win all the toss up states plus any one of the Democratic-leaning states and still lose the election. Those states won't even be enough and currently he trails in the polling average for all of them except Ohio. That sums up how difficult a position he is in heading into this final week.
A closer look at the polls shows Trump facing difficulties across the board. The suburban shift towards the Democrats that handed the party a blue wave in 2018 has continued through 2020. Women, a demographic that already leaned Democratic, have shifted even further to the left, including white women, a demographic that actually tilted towards Trump in 2016. Incredibly, Biden leads Trump among seniors, a demographic that almost always votes Republican (turns out seniors care a lot about Covid-19). The only demographic that bears any good news for Trump is among minorities, with whom he still does poorly but better relative to 2016.
You can see the impact of these trends on the map above, where Biden currently holds substantive leads in the key rust belt states and has even made in-roads into traditionally red states like Arizona (where Trump's animosity towards McCain has surely hurt the incumbent), Texas, and Georgia.
Looking at the toss up states, Trump could conceivably lose all of them but I don't think he will. His lead in Ohio has been small, but consistent, and the district level data looks promising for him. I also think he will win Florida. Biden has led in Florida for basically the entire year, but the state has a pretty clear track record of overestimating Democrats' in the polls, even in the 2018 midterms which were otherwise a blow-out for Democrats. Similarly while Biden has led consistently in North Carolina (another state that tends to overestimate Democrats), that lead does appear to be narrowing and I expect that this (admittedly minor) momentum could take Trump over the finish line there.
The opposite appears to be true of Georgia. Trump has led this state narrowly for most of the election cycle, but recent polls indicate a late surge for Biden. This is a state without a history of overrating Democrats that came extraordinarily close to going blue in both 2016 and 2018. If this momentum holds then there is a good chance that 2020 could be the year for Democrats in Georgia. The X-factor in this state, sadly, is corruption. In 2018, Georgia was widely criticised for voter suppression tactics, its lack of transparency, and suspiciously destroyed voting records in the wake of a razor thin vote. It could be enough to swing an otherwise close election.
But while these may be the closest and most interesting states at the moment, they are not the most important. Biden doesn't need to win any of these states. If he does, then it is almost certainly game over for Trump, but Biden could lose them all and still quite handily win the election.
No the key states, as in 2016, will be the rust belt states of Wisconsin, Michigan, and in particular, Pennsylvania. These are the must-win states that will determine who wins the election. If Biden wins them, he will almost certainly be President. If Trump wins them, he will almost certainly be President. The problem for Trump is that right now all of them show pretty decent Biden leads.
Of these, Wisconsin in particular looks rough for Trump. As one of the nation's worst-hit Covid states, Wisconsin was always going to be in an anti-incumbent (and anti-Trump in particular) mood. But the real damage was done in Kenosha over the summer where Trump was widely criticised for his handling of peaceful protests, which ultimately resulted in several dead at the hands of armed right wing militia. Wisconsin is a state that had been shifting right for several years now and not too long ago appeared to be Trump's most favourable in the region, but at this point it looks a safe bet to be carried by Biden.
Michigan is another state that Trump just about edged in 2016, but in which he has been polling poorly throughout this cycle. The most left-leaning of these three states, run by a popular Democratic Governor with whom Trump has inadvisedly insisted on waging a war of words, to the point where he almost appeared to have called for an armed rebellion, an accusation that can't have been helped by the recently foiled attempt by right wing militia (see the pattern here?) to kidnap the Governor.
But let's not mince words, Pennsylvania is the state. If you watch one state this election, it's this one. With 20 electoral votes, Pennsylvania is one of the most electorally valuable in the country. Trump has basically no plausible path to victory that doesn't involve winning this state. Conversely while Biden does have a plausible path to victory without it (by way of Arizona), he arguably becomes the underdog if Trump carries the state (also taking into account that if Trump overperforms in PA, he likely also overperforms in other demographically similar states). There is simply no state that is as important to this election as PA and there is a very good chance that this will be the single most determinative state.
To be quite frank, Pennsylvania also has a Republican controlled legislature and bizarre electoral laws that block the counting of any early or mail-in votes until after election night. So if there is any state where Trump could conceivably pull some shenanigans to block or overrule the popular vote, it's this one. To be clear, I still think this is incredibly unlikely to happen (or to work if it does happen), but tragically in today's America it can't be ruled out as a possibility.
Fortunately for Trump, Biden's lead in Pennsylvania has also been narrower than his lead in Michigan and Wisconsin. Despite being a sort of second home state for his opponent, Trump has managed to avoid making the sort of catastrophic blunder here that may have well have put WI and MI beyond his reach. To be clear: Joe Biden does lead the polls in this state and has done so consistently, but the race is close enough that it would only take a larger than normal polling error to put it into play, and the entire election along with it.
But just as this race could plausibly move in Trump's favour relative to the polls, it could equally do so in the opposite direction. Incredibly, Trump appears to be holding leads in deep red states like Montana, Kansas and (above all) Texas that are at least as narrow, if not more so, as the key rust belt states he needs to take from Biden. So yes, this is still a map where Trump could conceivably win, but it's also a map where Joe Biden could run the board and produce a truly historic landslide.
Whichever way you look at it, this is a bad map for Trump. Trump really has only the one plausible path to victory through Pennsylvania, but Biden has several potential options. More importantly, Biden leads in all the key states and either leads or is close enough in enough non-key states that you would expect him to pick up a few of those as well. That the closest battleground states appear to be the likes of Georgia, Iowa and Texas, rather than Michigan, Wisconsin or Pennsylvania is a very, very bad sign for Trump. He is playing defence at a time when he is already behind, and time is running out for him to turn it around.
House of Representatives Verdict: Democratic Majority
Current House Map: Democrats - 232, Republicans - 197.Predicted House Map: Democrats - 240, Republicans - 195.
Approximate Net Change: Democrats gain 8-12 seats.
This is going to be the shortest part of this article. Very few analysts consider the Republicans to have any chance at all of taking back the majority in the House, and I concur.
In 2018, Democrats rode a blue wave to a House majority with a historic 8-9% national margin. This result was achieved despite the fact that midterm elections are typically less favourable for Democrats and despite the fact that House is gerrymandered to the point where Democrats need a strong majority of the vote just to break even (see my 2018 post for a reminder of what "gerrymandering" means and why it is an antidemocratic stain on the country). In 2020, the national environment appears to be at least as favourable as 2018 (or at most 1% or so worse), plus the addition of the inherent benefits that the Democratic party typically sees in higher-turnout Presidential years. It is also material that progress has been made in numerous states to end the practice of gerrymandering.
All these factors suggest that Democrats should not only hold onto their majority, but are even poised to expand it further, even with all else remaining equal. My expectation is for a modest gain in the region of ten seats. No fewer than five, but no more than fifteen.
Senate Verdict: Democratic Majority
Current Senate Map: Democrats - 47, Republicans - 53.
Predicted Senate Map: Democrats- 51, Republicans - 49.
Approximate Net Change: Democrats gain 4 seats.
Key states to watch: AZ, ME, CO, NC, IA, GA, KS, MT
If there has been one silver lining for the Republican party these last two years, it is that they managed to retain control of the Senate against the blue wave of 2018. This was not a surprise, even with a national situation that was highly favourable to Democrats that year. The Senate map in 2018 was one of the most lopsided in recent memory, very favourable to Republicans with Democrats defending several seats in deep red states. The 2020 map, in fairness, is also a decent Senate map for Republicans, though not nearly so much as in 2018. The combination of a more favourable Senate map plus strong national conditions make Democrats a slight favourite to win a majority of the Senate.
There are 35 Senate races this year. Democrats need to pick up a net of four for a majority, or three if they also win the Presidency (with the Vice President casting the tie-breaking vote). It just so happens that there are three clear pick up opportunities where I would consider the Democrats to be favourites to win: Arizona, Colorado, Maine.
First Colorado. This is, for me, a non-race. Colorado has been a solid blue state for the last few elections and it looks especially so in 2020. The Democratic candidate is John Hickenlooper, a huge name in Colorado and a popular former Governor. I'm not going to beat around the bush here, I don't see any shot for Cory Gardner. Given the national mood and where the state is at in 2020, Democrats should pick this one up comfortably and I really can't see any other outcome.
Arizona is a pretty interesting one. Historically a fairly deep red state, Arizona has been low-key shifting towards the Democrats over a few electoral cycles. 2018 proved to be the breaking point, with Democrats very narrowly winning a Senate race, the first statewide race the party has won in many years. But 2020 has been a completely different beast, with Democrats leading consistently and comfortably in both the Senate and Presidential races for basically the entire cycle. The Republican party's self destruction in this race has been baffling to watch, appointing the losing candidate of 2018 Martha McSally to this seat (and thus to become the presumptive party nominee). Losing candidates rarely do any better the second time around, especially when they've been holding the seat by appointment without having won it in the first place. She is an awful choice for a candidate and she has run an awful race. By contrast, her opponent is Mark Kelly, former astronaut and national hero without a political bone in his body, whose wife is a beloved former Congresswoman and mass-shooting victim. It seems incredible to say, Arizona is not a race in which the Democrats expected to be competing, but they're winning this race and at the moment they are doing so quite comfortably.
Then there's Maine. The Republican incumbent Susan Collins has been in big trouble since her vote for Supreme Court nominee Brett Kavanaugh. I don't think there's anyone in the world who would envy the tightrope she has had to walk as a Republican in a solidly blue state, but even under challenging political conditions she has done an exceedingly poor job of it. From her much derided, fawning speech announcing support for Kavanaugh, to her flip flops, she has (whether fairly or unfairly) become a poster-child for spineless sycophancy in politics, and it looks like her luck may have finally run out.
So that's three pick ups. But Democrats are almost certain to lose a seat in Alabama, where incumbent Doug Jones only eked out his unlikely victory by merit of being up against an accused pedophile (and even that was narrow). So they need at least one more, preferably two. But beyond these three, the map starts to look tougher for Democrats.
Cal Cunningham had been looking a tidy favourite to win in North Carolina, but recent polls have shown that race (and the state as a whole) narrowing in recent days in tandem with a minor sexting scandal that Cunningham has faced. He still leads in this race, but it's tight, and with the momentum arguably against him I don't think Democrats can rely on this race.
Elsewhere, Democrats appear to have taken something of a surprise polling lead in Iowa and trail narrowly in a Kansas race that has proven divisive on the Republican side. Montana, of all places, appears to be a pure toss up, with popular former governor Steve Bullock running neck and neck with the Republican incumbent. Then there's Georgia, a state that is not only crucial in the Presidential race, but also features two Senate races and polling that appears to be trending in the Democrats' favour in recent days. Both races in that state appear to be tied at the moment, with Democrats either slightly ahead or slightly behind depending on which poll you look at. The Georgia races represent good pick up opportunities, but it should be noted that due to Georgia's run-off system, if no candidate gets 50% of the vote on election day, the top two candidates head into a run-off at a later date, by which point the political environment may not be as favourable for Democrats.
These are the key races, but there are a few more that are still worth watching. First up is Michigan, one of the few states that Democrats are defending and a potential GOP target if they have a good election night. Currently, as in the Presidential race, Democrats look to have a consistent if modest polling lead, but it's worth watching. Keep an eye also on South Carolina, of all places, where Jamie Harrison has staged an unlikely, record breaking insurrection against prominent Republican Senator Lindsey Graham.
So there's a good six or so close races here, with Democrats only really needing to pick up one or two. All else being equal with such a number of close races, historically you would expect them to prevail in at least a few of them. Democrats will feel good about picking up the additional two seats they need.
Conclusion
So there it is. By no means a foregone conclusion, but the polling right now strongly suggests an election night that results in a Democratic trifecta, perhaps as much as a 7 in 10 likelihood. The unique conditions of 2020 and unprecedented threats being made against the democratic process are certainly setting a national mood of uncertainty. But for all the drama and theatricality, more often that not these things end up going more or less as expected. I understand the doubt, but ultimately expect the polls to be proven correct.
