Biden is way ahead. Convince me I’m wrong
I want you to stress-test my reasoning to make sure it’s sound
I never really understood the saying “don’t miss the forest for the trees”. The shallow interpretation (don’t get caught up in minutiae that don’t matter) is not lost on me, but as someone who spends a lot of time pay attention to the minutiae of data, I’ve always thought that the only way to know you’re in the forest (the “wrong” place, according to the saying) is to explore it. That requires some nonzero amount of time spent inside the forest itself. Only once you’ve explored the area around you, taking stock of the variety of greenery and woodland animals will you notice that the trees enveloping you are more than you were looking for.
I find myself stuck in the forest today. It is a comfortable place for me. I do not believe someone can truly understand something without first getting lost in a maze of information. That goes double for election handicapping. But, as they saying implies, sometimes the forest is not the best place to be. We don’t need to know what species of moss grows on the tallest oak or the number of squirrels and rabbits sustaining the higher rungs of the food chain. We just need to know whether there are trees and how tall they are—whether a candidate is ahead, and how robust that lead might be.
So I invite you to argue with me today. Help me find my way out of the forest. What follows is my case for Joe Biden’s healthy lead in the election today. See what holes you can poke in it and leave your response in a comment below.
The way I see it, all the data indicate that Biden has a pretty strong lead in the presidential race. Just go look at my toy polling model (real model coming soon…) which aggregates state and national polls to project what might happen in the electoral college. It currently shows that Biden is up by roughly 5 points or more in all the “blue wall” midwestern states that doomed Hillary Clinton last time around. He also leads in Arizona and Florida and is right on the verge of being slightly favored in North Carolina. He’s only trailing by 2 points in Ohio, Iowa and Texas—margins small enough that even a normal-sized polling error could push Biden’s electoral vote haul up above Barack Obama’s 126-vote majority in 2012.
In addition to the state polls, national surveys currently suggest Biden will win the national popular vote by 8 percentage points—no doubt enough to win the electoral college. It would take a pretty large error to nudge that back down to Trump-friendly territory at this point. Partisan polarization makes such a shift unlikely, though it is certainly possible for Trump to claw back a few percentage points before November.
Then, there’s the economy. History tells us that presidents facing recessions perform poorly. And although I have serious doubts about how so-called “retrospective voting” holds up in an era of partisan hyper-polarization, I think our best guess is still that the economy will hurt Trump somewhat, even if that’s only by a point or two. (In contrast, the 1980 and 2008 recessions probably cost the incumbent parties a few points off the margins they would have won if they were presiding over healthy economies.
Presidential approval, too, suggests Trump will lose re-election. His rating currently stands at -11 percentage points according to FiveThirtyEight’s aggregate. That’s worse than most other incumbent presidents and puts Trump in the company of Jimmy Carter and George H W Bush, both of whom lost their re-election bids.
But maybe the polls are wrong. You can make a legitimate argument that the people some pollsters are interviewing aren’t demographically or politically representative of US voters. If some pollsters in key states fail to weight their polls by the education of their respondents, for example, they could repeat their mistakes in 2016 and underestimate Donald Trump’s margin by at least a few percentage points.
My rebuttal is that even if you dock Biden 3 or 4 points—large errors in the polls at this point in races during the polarized era, though not a total impossibility—he’s still ahead in enough states to win the electoral college.
This is not to say that Biden is a safe bet for November.
But as a way of stress-testing my mental and statistical models, I’m trying to talk myself out of seeing him as the strong favorite. Let’s see if you can help this exercise along.
I think Biden is a slight favorite. The distinction between slight and solid isn't that high, but here are some reasons to be a lot more bullish on Trump winning re-election:
1) Biden has a real problem with young voters- Maybe they just don't buy into his candidacy and he can't quite make up enough ground.
2) Trump's current problems with seniors will disappear once the campaign kicks off and if coronavirus problems go down
3) The decent possibility of a large recovery
4) Biden gets hammered and is viewed as corrupt and out of touch. The haters go toward Trump.
5) Trump can pull off one final electoral college v. popular vote split before 2024 makes his map untenable- Maybe he wins Arizona and Wisconsin by narrow margins while losing the popular vote by even more than he did in 2016
6) Too many unknowns to be very certain- Huge levels of partisanship. Social media and 24 hour cable news that's destroyed traditional campaigns. The possibility of a second wave. Biden and Trump both being old. A random supreme court vacancy. We haven't fully grasped how everyone reacts to events based on old models, and there are a ton of events we haven't modeled much.
All of those reasons--even some I'm not a giant believer in--give Trump at least a decent chance. I'd put Biden at around 65% to win.
Just my two euros' worth - I see one big problem here, Elliott. It is that your model and the pollsters assume that people will be allowed to vote in numbers that reflect the sentiments in the polls. It seems to me that we are reading lots of plans to prevent lots of people from voting. That fact alone throws a monkey wrench into the polling. As I have stated elsewhere, in the battleground states, and in Florida in particular, the key is a Republican governor and a lot of voters who will be prevented, again, from having their votes counted. Let us count the ways Florida voters have been undercounted in the past. Black voters blocked from getting to the polling places. Polling places closed. Bags of ballots dumped in the trash. The so-called "hanging chads" distraction. Secretary of State refusal to accept ballots. Etc. So that's one important way that you and the pollsters can be wrong - presuming that the number of people who will vote is not going to be reduced by policy obstacles. And that those policy obstacles will fall entirely on Democratic voters. As we know, there's where you vote and there's who counts the votes. In Florida, both have been interfered with in the past, and will be again.
Your analysis also doesn't take into account the numerous ways the GOP will work to prevent an election outcome that favors Democrats everywhere. We already know about the litigation filed to prevent absentee ballots. I do not doubt that these cases will wind up at the SCOTUS and that is not a happy possibility right now.
Cassandra speaks.