US 2020 Presidential Election prediction thread

[The Brain]

I really fail to see how the country would be better off if he got removed. Pence is in charge of the covid19 task force anyway--but assuming that he is so much better than Trump--you would still have 40% of the country fighting mad that their president was removed and Trump would likely be maximizing chaos from the sidelines (and probably running for president). It would be destabilizing long term.

He would not be running for president:

Judgment in Cases of Impeachment shall not extend further than to removal from Office, and disqualification to hold and enjoy any Office of honor, Trust or Profit under the United States: but the Party convicted shall nevertheless be liable and subject to Indictment, Trial, Judgment and Punishment, according to Law.

Why not? Because holding an office would be illegal?

[Eddie Teach]

He could run, he just wouldn't be on the ballot. 

[alfred russel]

You might be right politcally on Republicans not voting for impeachment, but for their votes Trump would be out of office and the president in the middle of a pandemic would not have Trump's psychological flaws. I think you hugely underestimate the impact Trump's indifference to actually being President has. I think it's very likely that you'd have far fewer deaths if Pence was in charge because I think he'd be trying to do the job. So on covid I think you would have more of a focused, coordinated policy that was not primarily focused on addressing the emotional mood of the leader. I don't think it's incompetence that is necessarily - I just don't think he cares about actually being President, for him it's a role on TV or on the phone to foreign leaders. So, in my view, even a very incompetent "normal" political leader would do better and make more of an effort than Trump.

And in terms of scale I believe the US has now, on a per capita basis, had more deaths than the UK and Italy. So I think it is safe to say it's had one of the worst experiences of this pandemic in the world.

It isn't really reasonable to hold republican legislators to task for not removing trump because a pandemic came later and he was the wrong person to lead america through it. I mean, if you were just to vote on competence you could have gotten him removed from office in his first year if legislators were able to anonymously vote just on those grounds.

Removing a democratically elected leader is a serious step that really needs something approaching a broad public consensus, and that just wasn't there.

In terms of deaths/capita, the US is near the top. Only a handful of developed countries are doing worse by that metric. However, you are still talking about profoundly disruptive restrictions on life that even in the US have cost 0.06% of the population. There are going to be a lot of long term effects from the disruptions, in addition to the short term effects on quality of life. There should be a balanced scorecard at some point looking beyond just a body count.

But more to the point--if you think that the US should have locked down more strenuously and Trump was the roadblock, it isn't at all clear to me that is the case. Trump talked the game of reopening, but as I've posted about, it was shocking that the federal stuff has been far more locked down than anything else in this state. He tweeted about "liberating" stuff, but his own government didn't comply. Some Republican governors, such as the one here, were much more effective at making sure things reopened. A President Pence could have more effectively implemented such a policy nationally.

OvB: point taken.

[Admiral Yi]

I really do wonder if there is a single person who could become President and unite the country again. Biden is more likely than any of the other Democratic candidates, but I almost think it's going to take a moderate Republican to win to do it, and no moderate Republican is going to win as things are.

We've painted ourselves into a corner, and I don't see a way out.

I was tempted to say just wait for the angry whites to die off, but then we'd still have the tension between the progressives and the moderates, and I'm sure with a common foe removed that tension would exacerbate.

[Sheilbh]

It isn't really reasonable to hold republican legislators to task for not removing trump because a pandemic came later and he was the wrong person to lead america through it. I mean, if you were just to vote on competence you could have gotten him removed from office in his first year if legislators were able to anonymously vote just on those grounds.

Removing a democratically elected leader is a serious step that really needs something approaching a broad public consensus, and that just wasn't there.

But it could have been a terrorist attack, or an economic crash. Presidents/executives deal with crisis - that's the job. There is no crisis that Trump is fit to deal with and that's why the personality of a leader matters. They're not just aggregates of policy opinion - character is important. As I say my view is Trump isn't interested in even trying to be President - he doesn't care, rather than just incompetence. So I think he is really unfit for dealing with any crisis, sadly the one he got was a pandemic. And this used to be a conservative opinion that character matters and that leaders should model the virtues and behaviour that they talk about.

The other point on removing a leader is fair but it's also an unknown. And again this goes to the role of leadership - are Congressional Republicans just meant to follow public opinion or attempt to shape. I think it's fair to say that anonymously briefing journalists about your concern doesn't have an effect, possibly because it's anonymous or easy for Trump to dismiss. But if they didn't think he was fit or that what he did was a crime, surely their role and what they owe the public is to vote that and try to shape the public consensus. It might have made no difference but maybe if there were a number of Republicans speaking about their concerns, on the record, on TV that might have shifted the consensus (or at least the middle of the consensus).

In terms of deaths/capita, the US is near the top. Only a handful of developed countries are doing worse by that metric. However, you are still talking about profoundly disruptive restrictions on life that even in the US have cost 0.06% of the population. There are going to be a lot of long term effects from the disruptions, in addition to the short term effects on quality of life. There should be a balanced scorecard at some point looking beyond just a body count.

Sure - it's just a metric you pointed to as relevant earlier in the pandemic. And I absolutely agree it's really important to strike a balance. I also don't think there's a clear answer of what works or what type of countries have done well in managing this (eg there's no correlation around state capacity, or pandemic preparedness ratings, or social spending, or austerity). I am far from someone who is obsessed with lockdowns. I think the key is probably actually widespread, rapid testing and clear messaging around self-isolation/quarantine until you get the all clear or for two weeks if positive. And again that goes to leadership and modelling the behaviour you want - Trump was apparently tested and even knew he was positive, but was apparently not self-isolating, still doing events and not informing people he was with.

But more to the point--if you think that the US should have locked down more strenuously and Trump was the roadblock, it isn't at all clear to me that is the case. Trump talked the game of reopening, but as I've posted about, it was shocking that the federal stuff has been far more locked down than anything else in this state. He tweeted about "liberating" stuff, but his own government didn't comply. Some Republican governors, such as the one here, were much more effective at making sure things reopened. A President Pence could have more effectively implemented such a policy nationally.

It's not about lockdown at all - I'm not fixated on lockdown measures or masks for that matter. I think the stuff the federal government could have done better was issuing clear guidance and communications from the CDC, coordinating governors (without constant political sniping) on PPE, on lockdown measures and on testing, I think the CDC fucked up on testing to begin with (which isn't on Trump) but was slow to catch up. And my issue isn't Trump talking about "liberating" stuff or lockdown measures it's that I think his public messaging of wanting to move on reflecting an internal decision to move on. He got bored, he wasn't interested, this was taking up too much time on the networks and he wanted to go out and talk up other stuff. I think that had an impact and meant the federal government didn't stay as focused or engaged on this because they followed Trump's public statements.

My own view is that even an incompetent Pence administration would have tried to respond in a more focused, coordinated way and wouldn't have got bored, I suppose, because I think they would be more based in reality and not Trump's projection of reality.

[OttoVonBismarck]

Interesting development in terms of prediction stuff/polling; so not sure who has been following it but USC in conjunction with Dornsife has been running a static panel of 5000+ voters this year. They tried to select the panel to be demographically and geographically representative of the United States at large. The panel gives updates every day on who they prefer for President, and then they report the 14 day moving average of those every day. It's a static tracking poll, now the big caveat with polls like this is you're not taking a new sampling of voters every week, but using the same 5000 you started with (which is a big number), so any mistakes or failures you have in selecting those 5000 or modeling them onto the real population get magnified because these are your static voter set for the whole election.

The one thing polls like this are particularly useful for is showing trends over time. So maybe September 15 if it showed Biden +15, and then 15 days later, it shows Biden +10, we can't know for sure how well that equates to broader polling, but it shows us a useful trendline that voters are moving a bit a way from Biden. That's really the utility of a poll like this.

So anyway, one of the minor news stories in poll watcher land of late has been this poll had been very stable +10 Biden for a long time, and was showing a gradual drift towards Trump over the past 7-10 days or so. Some studious observers pointed out that this was not an actual trend, but an error in USC Dornsife's methodology. Some of these panelists eventually convert from "people who have said they are definitely voting" to "voters", as their states allow various forms of early voting. USC Dornsife quit asking those panelists questions after they had responded that they had cast their vote--their answer can't meaningfully change after that ballot is cast. But what they were accidentally doing is forgetting to carry forward these voters as supporters of whomever they had cast a ballot for in perpetuity, instead they cleaved their data entirely out of the moving average and the data set.

Biden has been getting a disproportionate share of early voting, and as more of those voters drop out of the panel, this means the tracking  number has indicated a move towards Trump. Now that they've corrected the error it shows no move towards Trump at all. Nate Cohn of the New York Times has just tweeted about it today: link

[DGuller]

  I know most things are obvious in hindsight, but it really should've been obvious even not in hindsight.  If you're doing statistics and are not constantly asking yourself about how your sample might be biased, you really shouldn't be doing statistics.

That said, the wrong number also seems useful, as long as it's properly labeled.  One of the concerns on the left about the election shenanigans is that those who will vote on election day would be disproportionately Trump voters, and election night partial results might set up a narrative that actual results may fail to upend.  That number that was calculated the wrong way can measure the extent to which this is a concern.  If it doesn't go down below 3 percent by election day, we should probably be safe.

[alfred russel]

I don't want to crap on a project that I know almost nothing about, but it seems that if you are asking a group of people routinely about their opinion on the presidential race, they are no longer going to be representative of the population at large.

[celedhring]

I don't want to crap on a project that I know almost nothing about, but it seems that if you are asking a group of people routinely about their opinion on the presidential race, they are no longer going to be representative of the population at large.

Yeah, these kind of panels seem to be particularly prone to the observer's paradox, imho.

[OttoVonBismarck]

Yep--that's also a big reason the 538 polling average weights the USC-Dornsife poll fairly lowly, while it's an interesting exercise to follow a static group for 5 months you also end up with a lot of weaknesses vs traditional polls which are getting a whole new sample of people every time they run their poll.

[Maladict]

Playing around with the map. MN, WI and MI should be a given for Biden.
He then doesn't actually need PA if he manages to get AZ which traditionally underpolls for the Dems.
And then he needs either NE2 or ME2 to avoid the tiebreak. Much too close for comfort, but still.

[Grey Fox]

I think Texas is in play a lot more than PA or OH.

[alfred russel]

New ABC poll out in Wisconsin: Biden +17.

[Habbaku]

I think Texas is in play a lot more than PA or OH.

Biden has consistently polled much higher in PA than in TX.

[Josquius]

With Texas I do wonder whether Republican complacency may come into things.
Pennsylvania knows its a battle ground and everyone will be out campaigning and voting. Down in Texas though...things may sneak in under the radar.

Yep--that's also a big reason the 538 polling average weights the USC-Dornsife poll fairly lowly, while it's an interesting exercise to follow a static group for 5 months you also end up with a lot of weaknesses vs traditional polls which are getting a whole new sample of people every time they run their poll.

I guess though it could be representative of people who pay attention to politics, who though a minority do exist. But ja, for the population at large its bad that they know exactly whats going on upfront.