Hi, there. It’s Ross. On The Argument, we want to hear everyone’s perspective. And that includes you. Whether you’re a longtime listener or a new one, we’re asking you to fill out a survey about how you listen to this show and to others at nytimes.com/theargumentsurvey. We want to keep improving. And we can’t do that without hearing from you. Again, that’s nytimes.com/theargumentsurvey. And thank you in advance. Now, onto this week’s show.
I’m Michelle Goldberg.
I’m Ross Douthat. And this is The Argument. [MUSIC PLAYING] Today, what went wrong with polls in the 2020 election? And can we ever trust the polling industry again? And then Donald Trump still hasn’t conceded. What does that mean for the long 68 days until the inauguration?
Joe Biden is the President-elect of the United States, but not by the margin that lots of people expected and not with the coattails the Democrats had hoped for. So we have to ask, what were all those pre-election polls with wide 8, 9, and 10 point margins really capturing? What happened with polling in 2020? To answer our queries, or at least to start, we’ve brought on The Times’ polling expert extraordinaire, Nate Cohn, whose Twitter feed you may have been furiously refreshing, as I was all election week. Nate covers polling, elections and demographics for The Upshot at The Times. Nate, welcome to The Argument.
Thanks for having me.
It’s a pleasure, even if we’re going to give you a terribly hard time about the polls.
But the polls aren’t Nate’s — the polls aren’t either of the Nates’ fault, right? I mean, their model didn’t fail. It was the poll — the underlying polling failed. Or —
Yes, although The Times itself, we have our own polls.
OK, fine, fair enough.
So in some sense, we failed corporately. But so Nate, you wrote a piece saying that the predictions for this year’s election were even worse than 2016. So in spite of all the talk about how we were going to correct for 2016’s errors, we ended up doing even worse. And you’ve also offered several provisional theories about why. So why don’t we start with you running through some of those theories for us.
I think that, if I may just slightly revise your characterization of what I said yesterday —
Please, please do.
I think that the polling error this time, just in terms of the difference between the actual results and what the pre-election polls found, is almost exactly identical to the miss in 2016. But given that pollsters took a number of steps that we know improved the president’s standing in those surveys, the polling miss is, to my mind, a deeper and more significant miss than the one in 2016. And so I think we have to ask, why did the underlying data, the responses that we get to these surveys, get so much worse over the last four years that it canceled out whatever gains pollsters made by adjusting their samples in ways that led to a more conservative electorate? And I think there are four basic possibilities. One is that Trump supporters became less likely to respond to surveys, controlling for their demographic characteristics. The president attacked the media and the polling industry for years. The 2016 election result itself could have diminished trust in surveys. And all of this might have led to a decrease in the propensity for Republican-leaning voters to participate in the polls. A second possibility is sort of the flip side of that, that maybe progressives became more likely to respond to polls. We saw this huge uptick in political participation on the left over the last few years, the millions of dollars going to special elections and long shot candidates. Those same people who started donating for the first time may have also started taking polls for the first time. A third possibility is turnout. The turnout in this election was way up. I think that most polls found that that increase in turnout would be to the benefit of the Democrats and Joe Biden, because nonvoters in 2016 were disproportionately Democratic. I think it is quite possible that in the final account we’ll find that that was not true, that a disproportionate number of the new voters in this election were Republicans or Republican-leaning voters who came out to back the president’s re-election. And I don’t think that would — if that’s true, and unlike some of the other theories, that’ll be very easy to validate, that would not have been reflected by the pre-election polling this year. And I think the final possibility, which I think is the most interesting one in some important ways, is the possibility that coronavirus hurt the polls. This idea is primarily from a guy named David Shor who gained some notoriety this year because he was unceremoniously fired or whatever happened to him at Civis Analytics. And he believes that he has evidence that when the coronavirus hit, and Democrats took the threat from the coronavirus more seriously than Republicans, or at least they were likelier to stay home and social distance than Republicans and Republican-leaning voters and the supporters of the president, that that meant that their likelihood to respond to surveys increased. Because now they were at home with nothing to do while many Republican-leaning voters went along with their lives. And I do think there’s some elegance to that theory, because many of the polls that were conducted before coronavirus came very close to what the final results were in this election. We also know that there were studies throughout the election which found Joe Biden doing better in coronavirus hot spots. That absolutely did not prove to be true on election day. To take Wisconsin as the most obvious example, number one coronavirus hotspot of any of the battleground states, also the place where the polls were wrong by the most and also the rare state where Joe Biden was increasing in the polls heading into the election, which at least — it doesn’t prove anything. But all of what I just described is consistent with the possibility that there was some effect from coronavirus here. I also think it’s possible that, in the end, all of these various theories will work together in some combination of ways to add up to the polling miss. Or it’s possible that we’ll find that all of this is just theorizing with no serious evidence. And that we’ll be left without explanation. Or we’ll find that one is the complete explanation. I just don’t know at this point.
Can we talk about three for a minute, I mean, and why it is that you think that that one we can kind of definitively either validate or discount? I mean, that makes sense to me as someone who knows nothing about polling just in the fact that you saw so many of these Republicans running ahead of Donald Trump. So that there was this sort of red wave of Trump voters. And then you had a bunch of Republicans who also got whatever was the narrow never Trump vote, which is why you would see kind of a David Perdue or many of these Republican candidates running ahead of Trump’s margin, even as Trump’s margin was so much higher than the polls had led us to expect.
So we’ll be able to validate exactly who voted in this election, because most states — well, every state, in fact, has a database of people who are registered to vote. And they publish these voter registration files, as we call them. And on those files and they will indicate who voted in this election and who didn’t. And that will be used in a number of ways. One, in the states with party registration like a Pennsylvania or a Florida, we’ll be able to say, this was an electorate where registered Republicans outnumbered Democrats by two points in Florida. A second thing we’ll be able to do with that data is we’ll be able to append it back to the polls that were conducted before the election. Many surveys, including all of the surveys that were sponsored by The Times, are conducted off of these voter registration files where we literally pick people off the list. And we call people off this same list. So after the election, we’ll know which of our poll respondents actually voted. And we’ll be able to say, well, turned out that, while we talked to a lot of nonvoters in 2016, and while that group of nonvoters was Democratic, that the ones who showed up were Republicans. That’s not how it looked to us before the election, because the Democrats said they were going to vote, but they didn’t follow through. Or maybe it will prove to be the opposite, where a bunch of Republicans who weren’t sure that they were going to vote ultimately showed up. But either way, this is a question that we’ll be able to resolve with a lot of precision, and potentially soon.
I’m interested in number four for a minute. Because I remember on this show itself talking about some of the polling that The Times/Siena polls did during the Democratic primaries. And there was this moment when Elizabeth Warren was leading or close to leading in the Democratic primary polls. And we came out— I think it was a poll of the swing states, the Midwestern swing states, that compared how Biden performed to how Bernie performed and how Warren performed. And Biden clearly had an advantage. And it looked a lot like the advantage that he ended up with last Tuesday. And this was this weird moment in the Democratic primary where it almost felt like Times polls were sort of intervening in the race and signaling to Sanders supporters but maybe especially Warren supporters to freak out about their candidate’s electability or something. And then we all forgot about it, because the shape of the race changed so much. But under the COVID theory, this was an insanely stable race, right? Under that theory.
If theory number four is right, COVID changed the polls, not the race. And I do think it’s a very significant question. Because it doesn’t just affect what we think about the horse race. It has a lot of effect for the way we think about the arc of the Trump presidency and his handling of the coronavirus and what proportion of the American electorate is sympathetic to a more conservative approach to the issue. And if it turned out that large swaths of the electorate are not inclined to support social distancing to the extent that the polls suggested, that would be very important for understanding the capacity of our society to implement those measures going forward.
I suspect some of our listeners know about the challenges to polling these days in great detail. But others may not. Can you just tell us how hard it is to get someone to pick up the phone right now? Like, how many people do you have to call to get a polling respondent?
We usually call about 100 telephone numbers before we get a polling respondent. Now, granted, we work really hard to reach the people who are hardest to reach. So there are other pollsters who don’t go through that work. And they, as a consequence, have somewhat better luck in getting respondents. And they just collect more people — they have more people who are older as a result or something like that. A different way of thinking about it is that if you were an interviewer at one of our call centers that works for Times/Siena polling, you would go a whole hour dialing phone numbers. And you would complete about one interview during that time. And that interview might last 10 minutes and spend the next 50 minutes dialing numbers with no success.
And this is apart from all of the challenges of diminishing social trust, conservatives not wanting to talk to pollsters. The change in how people pick up the phone, right, is the fundamental challenge for the polling industry over the last 10 or 15 years. Or is that overstated, do you think?
Well, I think it’s certainly true from a cost standpoint, right. I mean, the more phone numbers you have to dial, the more hours you need to pay this poor interviewer for before you’ve collected one poll. So it’s had a huge effect on the cost of polling. There is less evidence that it’s had a huge effect on the accuracy of polls. We’re starting to get close to the point where I think maybe we can raise the possibility that it has had an effect on the accuracy of polls. To me, the polling error in 2020 betrays a pretty fundamental mismeasurement of white voters without a college degree. While in 2016, there just weren’t enough of those voters. In 2020, these voters were just mismeasured. I mean, they told a story about what happened for white working class voters in this election that wasn’t true. The polls said that they swung back towards Biden after swinging towards Trump in 2016. It just didn’t happen. And that’s a more fundamental failing than I think we’ve had in recent elections.
Can you talk about the social trust piece of this and sort of the ways that education weighting, I guess, was maybe supposed to correct for that, but in the end, didn’t correct for that?
We have known for a very long time that people with a college degree and people who have a higher rate of turning out in elections are much likelier to take political polls than those who are not. And that makes sense. We know that college graduates are politically engaged. They subscribe to The New York Times. They donate to political campaigns. They’re listening to this podcast. The sort of people who want to do these sort of things may also be likelier to take a poll. In 2016, that was a problem for the political polling industry for the first time. Because Donald Trump did much better among people who did not have a college degree than those who did. And that didn’t used to be true. I mean, you can go back and look at the 2012 exit polls or 2008, whatever you want. Barack Obama and Mitt Romney both did about the same among voters with a college degree. So this dimension of non-response to political surveys, while true for a long time, did not bias political surveys. Now, in this election, we have an additional problem, which is that even though the polls properly represented the proportion of the electorate that didn’t have a college degree, the people who picked up the phone who did not have a college degree were less representative of that demographic group than they had been before. One of many theories for why that could be true is something like social trust. I have to say that I think the evidence on this is pretty mixed. But it is one possible explanation for why a certain kind of conservative voter who is distrustful of their neighbors and distrustful of society and distrustful of institutions would be less likely to take a telephone survey than a liberal who still has sort of idealistic West Wing-style visions of the way our government works or the way our society works.
Although, I mean, as maybe — I’m not quite a West Wing-style liberal. But somewhere in that spectrum. To me, it’s sort of psychologically opaque why if you are part of a social movement that spends a ton of time complaining about polls, saying that the polls are voter suppression because they don’t adequately represent your share of the populace, kind of in thrall to this demagogue who’s constantly saying the polls are wrong, the polls are underestimating us — why you wouldn’t take your opportunity to put your finger on them in your own direction?
So one, I think you can imagine that the distrust is so deep that you worry that your response is being manipulated. Or you think that you can’t trust the survey to even do a good job of doing justice to your side. That’s one possibility. I think another possibility is to sort of reframe it a bit. And it’s a little bit of a — to think of all the Trump supporters that have taken some version of — I don’t care about your feelings or that sort of attitude where it’s like, I don’t care about you. I think that if — to the extent that you think that the propensity to volunteer or to donate some of your time to a different cause is a relevant predictor of whether you want to take a survey, that it’s possible that the Trump presidency is sort of — we don’t want to spend too much time caring about other people sort of attitude, could have an effect on the kinds of people that are likely to respond to surveys. And then the final thing I would say is that I think that there are a lot of people, perhaps especially in rural America, who sort of would like to be left alone in a fundamental way and just don’t trust anything. One thing that I’ve taken note of this year, and I would like to — I may write this up myself at some point. But I would rather someone else do it, because I’m not an expert on this. But the Census Bureau publishes charts, maps that show in fine detail who has taken the census and who has not. And the census non-response rates in rural America and particularly in many of the sort of less educated outlying parts of rural America where the president has made his largest gains over the last eight years — the census non-response rates there are really high. And they’re much higher than they are elsewhere.
And did people take that as an early warning? I mean, if that was something that pollsters knew about?
I don’t think it’s something that pollsters knew about. I think that there are a handful of people who had been taking note of this during the year. I did take note of it. There was a question whether we should do anything about it. Census non-response rates are kind of challenging. I mean, a lot of these places are in population decline. It’s possible that a lot of these people are not responding to the survey’s because they’re not there anymore. And it’s not entirely obvious what you’re supposed to do with some of the information. But I do think that it was certainly a signal that’s consistent with the possibility. But this represents a broader withdraw from engaging with — whether it’s strangers or whether it’s the government or something else, that would be very difficult for pollsters to overcome if it’s true.
Nate, in the last five minutes we have you, I want to lift up to the — well, there’s sort of two questions. The macro level question is, will the country ever trust polls again? There’s also the more specific question of, will people in our profession trust polls again? Because, of course, we work with polls not just in presidential elections, but so much of what all of us as columnists and journalists do depends on public opinion surveys on a lot of other questions. But what do you think the best case scenario is for the polling industry sort of regaining public trust after not just sort of missing, but missing in the same direction in these incredibly high profile and fraught elections?
I don’t think that the polling industry will find it’s very easy to earn back the trust of either the media or the electorate. I do think that it’s important to keep in mind there are a range of possibilities for just how bad polling is. To take one example, we know polls are imprecise. They’ve never been perfect. Earlier in the cycle, I thought that polling was very illuminating on the Democratic primary — Joe Biden’s strengths and his resiliency despite all of the Twitter chat about his vulnerabilities on progressive issues, I think that was an important lesson for decision makers in the media and elected officials to fully internalize. I think it would have been very difficult to pull off without polling. Then there’s another possibility, which is that the polls are so inaccurate so as to be useless. I’m not sure whether I would go that far when I look at the balance of public polling. I do think that there’s a case that was true in this particular general election, that the polls were so off that you would have been just about as good to just look at the 2016 election results and assuming we were on track for a repeat. And that’s not quite true. I mean, Biden will do better than Hillary Clinton did and win the Midwestern states back. And then a final possibility is that the polls are so bad so as to be misleading. And I think there are cases of that too. I mean, Joe Biden’s campaign was in Ohio and Iowa before the election. He’s going to lose those states by eight points. He shouldn’t have been anywhere near there. There were Democrats who donated hundreds of millions of dollars in total to races that were not competitive in the Senate. That actively hurts Democrats —
In some of those cases they should have known.
In some of those cases they should have known better. But there were polls from Quinnipiac University or whatever, a nonpartisan academic research center that had a tied race in South Carolina. And that’s bad. That is misleading.
Nate, we really, really appreciate you coming on, crossing the news opinion divide to share your insights.
Yeah, thank you so much.
And hopefully, the next time we have you on it will be the talk about how brilliantly New York Times polling performed in some future election cycle.
You know, I’m going to go off on a limb and say you won’t have me on if the polls do that well. So I’m not — [LAUGHTER] I enjoyed this so much that my incentives here are very conflicting going forward. [MUSIC PLAYING]
So remember that next time there’s a polling miss.
We’re in the business of that kind of incentive blurring.
Thanks for having me.
Here at The Argument we want to hear everyone’s perspective. And that includes you. Whether you’re a longtime listener or a new one, we’re asking you to fill out a survey about how you listen to this show and others at nytimes.com/theargumentsurvey. We want to keep improving. And we can’t do that without hearing from you. Again, that’s nytimes.com/theargumentsurvey. And thank you in advance for your help.
[MUSIC PLAYING] And we’re back. It’s pretty safe to say that by the time you hear this podcast, Donald Trump will still not have conceded the 2020 race. His team continues to file legal challenges over ballot counting, while his fellow Republicans seem, without necessarily endorsing all his claims, to be humoring him for the moment. So what happens now? Well, our next guest has the closest thing to a crystal ball that we could find. Rosa Brooks is the co-founder of the Transition Integrity Project, which last summer organized a series of war games that tried to play out what could happen after this election. And in September, she published a prescient piece in The Washington Post headlined “Trump Could Refuse to Concede,” in which she envisioned the blue shift of mail-in ballots, Trump’s Twitter allegations of fake votes, and the post-election lawsuits over alleged fraud, all of which has obviously come to pass. Rosa is also a professor at Georgetown Law School and formerly worked in the Departments of Defense and State. Rosa, welcome to The Argument.
Thank you, Ross. It’s great to be here.
So we are relying on you to tell us, since we’re living inside at least one version of a war game or simulation — in this version as it’s played out so far, Biden winning narrowly, seemingly. Trump refusing to concede. Legal challenges ongoing. What happens next?
I wish I had a crystal ball. I have something called a sarcastic eight ball. [LAUGHTER] But it’s not super helpful, because you shake it and it says things like, yeah, right, or, you wish, or, in your dreams, instead of giving any info.
Which sounds vaguely like what the president is saying to his aides when they approach him and suggest conceding the election.
No, I mean — and I should be clear, by the way. The Transition Integrity Project is not and was not an organization. It was just a series of simulation exercises that we did. I organized them together with Nils Gilman, a historian and about 100 different people participated over the series of four exercises we did. So the things that I talked about in that September Washington Post article you mentioned did not derive from my own fevered imagination, although my fevered imagination is capable of coming up with all kinds of apocalyptic scenarios. They derived from doing these gaming exercises where we took people who were experienced in both Democratic and Republican presidential campaigns, White House staff, Congress, state level office. And we put them on teams. We had a Trump team, a Biden team, a GOP elected officials team, and so on. And we gave them starting scenarios. We weren’t interested in the election. We figured other people were thinking about that. So all of our scenarios essentially started at about 4:00 in the morning on election night. They started with, OK, here’s how the electoral math looks. It’s 4:00 in the morning. Biden won these states clearly. Trump won these states clearly. Here’s what’s not known. And then we said to the teams, take turns, go. What are you going to do now? This kind of gaming is actually — if anything, the purpose of this kind of game is both to — it’s to push people to test their own assumptions. So if you’re sitting around — and when we started planning these games was a year ago now, more than a year ago. If you’re sitting around saying, oh, don’t be silly. Of course Trump will concede if he loses. That’s ridiculous. Because we have these magic guard rails of democracy that will keep us safe at the right moment. This kind of exercise can push people to go, oh, wait a minute. Maybe there’s nothing automatic about what happens. But they’re also a way to say, boy, how bad could things get? And identify ways to try to mitigate those risks and make sure that the sort of dark scenarios that these exercises did, in fact, get to don’t happen in real life. Now, unfortunately, [LAUGHS] either you could say that we turned out to be more prophetic than we intended to be or that our risk mitigation efforts were not as successful as we hoped they would be. Because we do seem to be at a moment that several of our scenario exercises got us to where, despite the fact that we have what kind of seems like a pretty definitive Biden win — it was really not that close. And yet, even so what we are seeing is Trump refusing to concede, a surprisingly high, and depressingly high, number of senior GOP officials saying, yeah, we don’t concede. We’re going to fight. Not only are we going to fight in court, but we’re going to try to persuade state legislatures to overturn the popular vote because we don’t trust it, and consider not certifying slates of electors for Biden. We’re seeing a massive right wing disinformation campaign, which seems to be having some real success saying things like, no, no, no, Biden didn’t win in Pennsylvania, and so forth. And I’ve got to say that this is one of those situations where the fact that our exercises ended up being prophetic does not give me the slightest satisfaction. In fact, it’s really, really scary and depressing.
So there was about a day and a half where it felt like, oh, we got so frightened for nothing. And actually everything is sort of playing out somewhat normally. And then some of these nightmare scenarios started unrolling. And so I guess my question is, when did you start to get really alarmed? And what are your fears about what happens next?
Well, I’m sort of characterologically always really alarmed. [LAUGHTER] I never get invited to parties, because I’m always a person who’s like, yeah, things are going to fall apart. So that said, even for a paranoid, apocalyptically-oriented person like me on Saturday when the networks started calling the election for Biden, I thought, oh my god. Thank goodness. We really dodged a bullet here. For a while there, I did. I did. I had a really kind of unfamiliar emotion [LAUGHS] of calm and relief. And I thought, wow. So we got about 36 hours in there, right, of thinking either we were paranoid nuts or we did a great job preventing this. And then the bad stuff started to happen. I mean, we all figured Trump wasn’t going to concede right away. There was zero chance of that. And in fact, I pretty much thought all along, Trump is never going to concede. It’s just not in his nature. But I did think, we’re going to see people like Mitch McConnell say — maybe not immediately, but in a day or two, say, congratulations, President-elect Biden. And go to Trump and say, Mr. President, I heart you, but you lost. And better luck in 2024. Or better luck with Trump TV or whatever your next venture is. And I’ll sit down with you as we start calling the moving trucks. And that did not happen. And that was ominous. First, the silence was initially ominous. And then things started getting even more ominous. We started seeing the GOP filing all of these lawsuits all over the place in an effort to stop the vote counting, get ballots invalidated, and so on. And we did the exercises — certainly the players playing team Trump and GOP officials did precisely that in all of our exercises. Unsurprisingly, these litigation efforts have been failing, failing, failing all over the place, because in court, unlike — even if you’re with a sympathetic Republican judge, in court, unlike on Facebook, you have to state a legally cognizable claim. And you have to make arguments. And you have to have evidence. And they just don’t. There’s no evidence of widespread voter fraud. But I think there’s — the purpose of this litigation is to create a narrative more than anything else, to sort of say to the base — and it’s connected to all these disinformation efforts to say, oh, well, where there’s smoke, there must be fire. They wouldn’t be litigating if there wasn’t some real issue. And people hear the litigation, and they miss the part about the cases getting dismissed for lack of evidence.
And how worried are you about the states refusing to certify the actual vote? Because that was part of your scenario also.
It was. I don’t know how worried to be. I mean, again, in my glass half full moments, I think, they’re never going to get — Biden won too many states. It’s not enough for them to get one state legislature to get wacky. And it would have to be getting wacky — it would have to be legislature saying, despite the absence of any evidence, we are going to attempt to subvert the popular will in this state and send an alternative slate of electors pledged to Trump. Could they manage that maybe in one state? Maybe. Could they manage it in three or four, which is what they’d have to do? I think it would be really tough to do. It’s also crazy, needless to say. Because as you were just discussing with Nate, one of the phenomena that we saw in this election was that Trump did worse than down ticket Republicans. So some of those very same ballots that voted for Joe Biden, that checked the Biden box at the top then checked the Republican House or Senate candidate box. So you’re really playing with fire if you’re trying to kind of invalidate those ballots. It also just doesn’t make sense, that if the Democrats are engineering a massive conspiracy full of fake ballots, then why don’t they sweep the House and the Senate too while they’re at it? What kind of conspiracy gives the presidency to Biden, but takes away a bunch of House seats and leaves him without a Senate majority? It doesn’t make sense. But it’s frightening.
The theory there is that — the conspiracy theory there is that it’s easy if you’re mass producing fake ballots to just have them vote for president and not fill in all the rest of the votes.
[LAUGHS] Well, maybe.
Just so we’re clear on what the —
What the conspiracy is.
What the conspiracy theory — well, look, so I mean, I am very aware of what all the conspiracy theories are. Because I’ve been following this in conservative social media and — including among people I’m friendly with over the last week or two. And I guess my main disagreement with how you guys are describing what’s happening is the idea that the legal motions being filed are there to further a deliberate disinformation campaign I think gets the causality wrong.
That may be right. That may be right.
I think the way things work now in American politics — and this connects to the conversation we had in the last segment about collapsing trust in polling and every other institution, is that, yes, there are sort of cynical actors and grifters who fasten on to elections and sort of make what they know to be false claims. But there’s just a huge number of people who are invested in sort of narratives that emerge organically, that are conspiratorial but are also just sort of based on the fact that anyone can go and look at voting numbers on the internet and find what they think are anomalies and turn it into a Twitter thread that then gets picked up in six other places. And confirmation bias sets in. And I’ve spent the last week or so having arguments with — to an extent that I haven’t before about conspiracy theories, sort of batting down people saying, oh, but, Ross, didn’t you see that the number of voters in Wisconsin exceeded the number of registered voters. And isn’t that evidence of fraud? And then you have to say, well, but actually Wisconsin has same-day voter registration. So you’ve got the wrong denominator. But these are not people making things up. These are people who —
— thought Trump was going to win —
Right, but —
— were surprised by the outcome. And so that then — just to finish my brief here — you have a combination of a big chunk of the country that is primed or predisposed to believe the worst about its opponents joined to a president who — I always assumed Trump wasn’t going to concede, at least in the first week. That has not surprise me in the slightest. And so what I think you’re getting from Republican lawyers and from Mitch McConnell and from other people is not the worst case scenario of Republicans cooperating in a coup. But it’s the very predictable scenario of Republicans saying, OK, well, Trump has the right to make these legal challenges. And we’re going to do recounts. And so we’re going to wait to congratulate Joe Biden until those things have passed. And so I guess, predictably, I disagree with both of you about the scale of the alarm. I think liberals should still feel pretty comfortable about what Trump is doing. He’s ranting on Twitter. And his lawyers are filing — they’re filing lawsuits that are just —
You can never feel entirely comfortable about Twitter rants.
Not comfortable, but in the sense that this is still within the range that’s not Bill Barr sends in the National Guard to seize ballot boxes and these kind of things that were circulating in serious magazines before the election.
But Ross, you wrote on Twitter — and this is inexplicable to me —
Yes, let’s do it. Let’s do it.
— that you thought that —
— Senate Republicans were behaving — I don’t remember the word you used, but responsibly. That they were — rather than cynically cooperating with Trump, they were — well, I don’t remember the language you used. But you —
I said a minority of Republicans are cynically adopting Trump’s arguments. But most are trying to figure out how to do the right thing without ratifying his stab in the back narrative among voters who, right now, trust him more than them. So what I meant by that is that —
But they are ratifying his stab in the back narrative. I mean, that’s what they’re doing. They’re ratifying his stab in the back narrative. They are kind of reifying the most paranoid impulses of their base. The right thing to do would be to say that these theories are baseless. That we wish it had gone a different way, but actually, no, these states are not going to nominate an alternate slate of electors, which is what kind of Republican politicians are threatening to do. And also would be trying to step in at a time when the president is decapitating the Pentagon, filling the top ranks with fanatical lunatics. They would try to both mitigate that and speak to the huge part of the country that sees what’s going on and is absolutely terrified, and should not have to endure being terrorized in this way by their own government.
So I don’t totally disagree. And we can get into this a bit more in the sense that I think, and have always assumed, that having Trump in charge during a presidential transition, let alone one with a pandemic raging, would create various disasters that were independent of whether Trump actually tries to cling to power. But on the stab in the back narrative, you have — long before Trump, you had a dynamic in the Republican Party, dates back to the collapse of the Bush administration, where Republican voters don’t trust their leaders to a scale that has no parallel on the Democratic side. And it has produced a Republican Party that has been roiled by sort of insurgent waves that has helped make the party unmanageable and the country ungovernable. I don’t think you disagree with that.
No, I don’t disagree with that.
So the Tea Party wave led to this era of debt ceiling brinksmanship under Obama. And then the next wave was Donald Trump and gave us Donald Trump, right. So if you are Mitch McConnell, and you go out at a time when Trump still has the right to claim to ask for recounts. He still has room to file lawsuits. And you say, folks, this election is over. And Joe Biden is the next president. You are giving Trump — and not just Trump, but the Trump complex — you are giving them what they need to run a Trump version of the Tea Party in 2022 and to run either Trump himself or the Trumpiest figure imaginable— Don Junior, let’s say — in 2024, and say, we could have litigated this election. We could have had recounts. And Mitch McConnell shut it down because he’s a RINO, traitor, deep state, sell-out cuck. You have to do something in this moment to talk to the people who I know who are normal people who are reading stuff on the internet.
Well, they’re not that normal.
I mean, they’re more normal than any of us.
I think there’s a degree to which you are both right. I mean, I think in terms of the GOP leadership, we have some people who are like — for some people, this is calculated theater. And it’s calculated theater precisely because of what you just said, Ross. They’re thinking, OK, we know that the odds of actually overturning this election result are extraordinarily low. We’re losing in court, because we really don’t have evidence of voter fraud. We can do these recounts, but we know that historically recounts do not — maybe they change a couple hundred votes. But it’s not going to be close enough to change the outcome. We know that we’re not going to realistically persuade enough state legislatures to send alternative slates to the electoral college. But because our base trusts Trump more than they trust us, it’s really important that we demonstrate both to Trump, who’s like a giant toddler — look how hard we tried. We’re fighting for you all the way. And that we’re able to go back to our constituencies and say, we fought for you. We fought for you. We fought for you. Vote for me in 2022 or 2024. I absolutely anticipate —
Right, but I just want to quickly — I mean, I want to — I want you to finish what you were going to say. But I just want to quickly interject that that’s not them trying to do the right thing.
No, they’re being self-protective. Yeah, they’re being self-protective.
No, no, but see this is — I mean, this is, like, definitions of the right thing. But the most important thing for American politics going forward — not the only important thing. But one of the most important things is that we not have a replay of the Trump experience. Right? I mean, and you —
And I think we are getting set up for exactly that.
But you can’t — you have to make choices in this moment that have the possibility of a replay in mind. What McConnell and other Republican senators and some governors are doing, what seems to me — again, pending what happens in a week or 10 days — to be a reasonable attempt to not set up Trump for being the man who was betrayed by his party, which is what — and maybe I’m too worried about that for the future. But I think that that is —
I don’t know. I mean, when you look at people like Mike Pompeo saying, of course there’ll be a peaceful transfer of power to a second Trump administration. The appropriate responsible note of caution is to say, look, I would congratulate Joe Biden, but the president feels very strongly that he wants to see these recounts play out. It’s very likely that this won’t change anything. But you know, we’re going to wait and see what happens. And we’re going to hope for the best. And if, indeed, it plays out that Biden is the winner, as looks likely, we will congratulate him. But we want to — OK, fine. But this stuff about — Pompeo’s comments are just kind of shocking, frankly.
To pull us back to scenarios, I guess, what is the thing that in your view, Rosa, would take this from a situation of sort of alarm to actual crisis?
I think two things would really alarm me. One is if we started seeing a much more open and concerted effort to say to state legislatures, hey, the popular vote is fake. Overturn it. And if we started seeing state legislatures, more than one, respond to that. That would worry me very, very, very much. The other thing that would really scare me — one of the things that happened in a lot of our exercises is that in the scenarios in which Trump was refusing to acknowledge a Biden win, lots of protesters turned out in the streets basically saying, Trump, you got to go. You lost. And although Biden was saying peaceful, peaceful, peaceful, there was violence, we will say. And sometimes it was provocateurs. Sometimes it was armed counter protesters turning up. And in several of our scenarios, the Republicans in our scenarios, as the Biden team would say, well, we’re going to urge people to go out peacefully and demonstrate the popular will. Some of our Republican participants were like, that’s playing right into Trump’s hands. That’s exactly what he wants. Because he wants there to be violence. He wants there to be an excuse to use coercive force. The darkest and most paranoid vision that we could have here — and this is one that the firing of Secretary of Defense Esper plays right into, et cetera — is that Trump is prepping for a scenario in which he is able to use coercive force, including active duty military, to, quote unquote, “restore order.” Or one of the bases for invoking the Insurrection Act to use active duty military is — one of the legal bases is actually to, quote, “protect federal property.” But the claim would be that the places where these ballots are being counted or recounts being done are federal property. We have to send in troops to protect them. Esper’s head was clearly on the chopping block ever since he publicly said after Lafayette Square, invoking the Insurrection Act would be a very bad idea. Using active military troops in the United States would be a very bad idea. I’m opposed to that. The firing of Esper is ominous. But if we start seeing signs that that’s more than just a sort of Trump lashing out at people who are insufficiently obsequious, and that that’s actually attempting to clear the way for —
But wait, can I just say one thing? Because isn’t one sign of that — and I don’t think we know exactly what it means — is that it wasn’t just Esper, right?
They’ve decapitated a lot of the leadership of the Pentagon and put in these truly — these people who are truly nuts, right. The head of the policy shop at the Pentagon is this guy Anthony — is it Tata? Is that how you pronounce it?
Yeah, he’s truly nuts. Yes.
Right, this is somebody who tweeted — who’s called Barack Obama a terrorist leader, who has tweeted absolutely lurid fantasies about the execution of —
Tata is truly nuts.
— former CIA Chief John Brennan. Right? They’ve put all of these crazed loyalists at very high levels.
Well, not everybody — well, not all of them are nuts. I mean, I don’t think we have any reason to think that Chris Miller is nuts. I don’t think we — the guy who —
Who is, for listeners, now the acting Secretary of Defense.
But the biggest problem with these guys — and it’s Chris Miller is, I think, in his mid 50s. But he is a recently — retired five or six years ago. But going in the space of a few months from deputy assistant secretary level to the head of the NCTC to secretary of defense is — I don’t know that — I have no reason — in fact, I know people who know him, who say he’s a decent guy. But I think when you put people who are extremely inexperienced into those positions, you’re not doing it because you’re thinking to yourself, whoa, I cannot wait to rely on their sage advice. You’re doing it clearly because you’re thinking, I want people who I can push around.
But the reporting we have so far, which, of course, could be wrong— but suggests that the thing that Trump wants out of the national security establishment in the next two months is, one, a full withdrawal from Afghanistan, which Esper and others had resisted. And two, a bunch of declassification of Russia’s material.
Well, that wouldn’t be — that would be —
Which, in order to prove to Michelle —
Right, but that wouldn’t be at the military.
— once and for all —
But that’ wouldn’t be —
That would be Gina — that’s why Gina Haspel’s expected to be — the head of the CIA — to be one of the next to go. So I certainly think we’re seeing a purge across agencies.
But I guess to me — again, I’ll have to — when we have the dire scenarios, I’ll have to eat my words. But to me, this seems like a repeat of the sort of pattern of the whole Trump presidency, where you have this —
Yes, right. You made me mad. You have to go.
Well, but where you have this anxiety among liberals about sort of his authoritarian power. But the actual danger is from total incompetence, having people who don’t know what they’re doing in high offices over a multi-month period. And I guess on that front, which is what I’m most worried about — I mean, Rosa, supposing that the worst case scenarios don’t come to pass, and in two weeks, Republicans are all calling Biden the President-elect. But the Trump White House just doesn’t cooperate with the transition. What do you see as the danger points there, where it’s clear that Biden is going to become president in January, but nothing is happening. No collaboration is happening to prepare for that.
Yeah, no, and I agree with you, Ross. I think that that darkest scenario is not super likely. I think it’s likely enough that it makes me very anxious in the sense that —
Right, but I mean it should be — the fact that it’s non-zero is extraordinarily alarming.
Yeah, that’s right. No, that’s right. And this is the metaphor that my friend Nils Gilman, who organized these exercises with me, always uses. If I said to you, there’s a 99% chance something won’t happen, you’d think, oh, good. Well, then let’s not worry about it. But on the other hand, if I said to you, there is a 1 in 100 chance that within the next 10 minutes a guy with a machete is going to walk into your house and start slashing away, you’d probably decide that, at a minimum, you’re going to lock your doors. And maybe you’re going to leave your house. But Ross, your question about the transition, what is supposed to happen, of course, is we get these so-called landing teams that will — appointees by Biden who will go into the agencies and just start kind of doing due diligence and saying, hey, what’s going on? What are the issues? Where are the vacancies? What’s working? What’s not? Can we see the memos? Where do things stand on x and y? And just kind of prepare so that when Biden is inaugurated and Biden appointees come in, they’re not sort of starting completely from scratch. They have some sense — there’s some sense of continuity. They’re also supposed to be getting classified intelligence briefings so they know what’s going on in the rest of the world. They’re supposed to be getting resources to pay staff, to have office space. And at the moment, they are not getting most of those things. And in fact, agency employees have been instructed not to cooperate with the transition teams. They’ve literally been instructed, you can’t talk to them. And so the best case scenario, it’s a mess. And it’s dangerous. And every US adversary thinks to themselves, hey, gee, this would be a really good moment to do something that in normal circumstances I know the United States would object to. But they’re too distracted and confused right now. So I think I can get away with it. That’s best case. Worst case is that our paranoia has been directed in the wrong place. It’s not a power grab by Trump domestically hoping to stay in the office. Worst case, these purges, et cetera, at the Defense Department, potential purges in other parts of the national security establishment, are designed to enable various forms of mischief involving Iran, involving the Middle East, involving potentially cutting deals that personally enrich Trump cronies, et cetera. And the goal is to get out of the way people who would say, oh, that’s a really bad idea.
I will say that my worries right now focus around the point at some point, 10 to 20 days from now, when I suspect liberal fears that Trump is going to pull off a coup will finally evaporate. But what will replace them is the question of transition cooperation. But on that note, contemplating wars and other catastrophes abroad, we want a recommendation for listeners who are going to live through the Trump to Biden transition and want to know what you’re enjoying in your life during this trying time.
Here’s my recommendation to everyone. Those listeners can’t see us, but I’m not sitting at a desk. I’m sitting in a recliner with my feet up. And in a world in which we all have to do endless Zooms and things like that now, I’ve always been somebody who wrote sitting on the sofa with my feet on the coffee table and the laptop on my lap, which is where laptops are supposed to be everybody. And I hated working at a desk. And then when the pandemic came along, and I had to do all of these Zooms, I started out thinking, oh god. Now I’ve got to be at a desk, because I’ve got to have my computer at the right height and all this kind of nonsense. And then at a certain point I thought, no, I don’t. So my recommendation to everybody is if you’ve got to read stuff on your computer, if you’ve got to watch TV, if you’ve got to do Zooms, you might as well recline.
Well, we appreciate you gracing us with that wisdom. Rosa Brooks, thanks so much for coming on.
Yeah, thank you so much.
My pleasure. [MUSIC PLAYING]
And that’s our show for the week. Thank you so much for listening. The Argument is, as always, a production of The New York Times Opinion section. Our team includes Alison Bruzek, Vishakha Darbha, Elisa Gutierrez, Phoebe Lett, Isaac Jones, Paula Szuchman, Kate Sinclair, and Kathy Tu. Special Thanks to Corey Schreppel. Next week, assuming that none of Rosa’s absolute worst case scenarios come to pass, we’ll be starting a series arguing about the first 100 days of the Biden administration. We’ll talk to politicians, analysts, and familiar faces and voices from Opinion about what Biden’s priorities might be and how both Democrats and Republicans might adapt to a new presidency. We’ll see you then.
Beautiful. Ah! OK.