Sweden’s Global Failure

Listen to this episode

S1: The following program has words that were once banned in Boston, as they say. No, not Bucky Dent or Manhattan clam chowder. Much worse words than that.

S2: It’s Thursday, May 7th, 2020, from Slate’s The Gist. I’m Mike Pesca. Terry Reid will be interviewed by Megyn Kelly.

S1: Does it show my age that when I heard this, I said, wait, Meghan Kelly even on TV? The answer is no. She’s on Instagram. Here’s the show my age part. My next reaction was, oh, so that’s not TV. Now, perhaps more millennial type person might say the same thing. Or maybe it’s even better. On the other hand, say there is a difference. Not that one medium in terms of bit rate or transmission. Speed beats the other, but actually in terms of journalistic standards. So during the entire timeline of the accusation, which I have been keeping an open but not uncritical mind on, one keeps one mind on something I’ve been waiting for the chance to see read interviewed by a real journalist backed by a real network with real standards and practices and accountability and reputational stakes. A Katie Helper podcast. Isn’t that crystal ball on Hill TV? Isn’t that Amy Goodman on Democracy Now? And she has done some great journalistic work over the years. She is not in the camp of decidedly journalists, decidedly non activist. All of these people are very happy to have someone who will damage the credibility of Joe Biden. I am not saying that’s what’s at play. But that is a common line between the outlets that Tara Reid has thus far appeared on. In an interview with Meghan Kelly. We don’t know so many things that we don’t know what considerations she has made. Kelly has made in regards Reid. We don’t know how much vetting her team is doing with Reid statements. We don’t know if questions or areas of questioning will be known to Reid beforehand. And throughout this time, I’ve been genuinely eager to learn more because I can’t emphasize this enough. I have in no way come to a conclusion about if the assault happened. I would not be terribly shocked either way. But I have wanted another shoe to drop. I need more information. This is not that shoe. Or maybe it’s a slipper. Or maybe it is the shoe, but it’s not dropping with a thud. It’s landing on plush carpeting. Or let’s put aside the analogy, shall we? Meghan Kelly on Instagram is not. Chris Wallace is not. Andrea Mitchell is not Anderson Cooper. There is no editorial oversight of Meghan Kelly’s Instagram channel. Kelly got a massive payout from NBC and might not need to reinvent herself or might be relishing the opportunity to reinvent herself as someone who I know pisses off Democrats. Who knows? Now, as far as the clips that have been released, they seem interesting. I’m not uninterested when Tara Reid says this.

S3: His campaign is taking this position that they want all women to be able to speak safely. I have not experienced that. If he’s watching this, what do you want to say to him? I want to say you and I were there. Joe Biden, please step forward and be held accountable. He should not be running on character for the president, the United States. You want him to withdraw? I wish he would, but he won’t. I wish he went. That’s how I feel emotionally. Do you want an apology? I think it’s a little late.

S4: There are lots of cuts to Meghan Kelly’s reaction to tarried statements. It’d be great TV if it were TV. It’s great. Instagram, I guess. I do think Reid will eventually have to do an interview with a more credible, established non ax to grind outlet. Or she might have no interest in talking to anyone like that. She might be gradually working her way up to that. If so, that will leave me hanging. We’ll all have unanswered questions because today in Vox, Laura McGann, an excellent journalist with lots of experience covering stories like these, wrote about her interactions with Reid. McGann interviewed Reid last year when she was talking about Biden touching her hair and neck in the course of that reporting began. Talked to a friend of Reid’s I’ll quote from that Vox article last year. Reid encouraged me to speak with a friend of hers who counseled her through her time in Biden’s office in 92 and 93. That friend was clear about what had happened and what had and quote is the friend talking on the scale of other things we heard. And I feel ashamed. But it wasn’t that bad. Biden never tried to kiss her directly. He never went for one of those touches. It was one of those. Sorry you took it that way. I know that is very hard to explain. The friend told me Mean McGann, she went on. What was creepy was that it was always in front of people. That account differs from Reid’s account of the sexual assault in the hallway. Okay, maybe the friend didn’t know about all of Reid’s experience. Only McGahn reveals the friend now tells her she did know about it writing. I asked the friend why then did she vote? So explicitly that Biden never tried to kiss her, touched her and appropriately, quote, it just organically rolled out that way. The friend said, reading, I had many conversations a year ago about what her degree of comfort was. She wanted to leave a layer there. And I did not want to betray that. It just wasn’t my place. Huh. That is complicated. And I’ll say it again. And this might get tedious hearing it because so many people interpret otherwise. But my talking through this outloud is not me disbelieving terror read. It may well could have happened and not in the.

S1: Oh, aliens could have abducted your cow. Way of happening. It really might have happened. But this is why we need established interviewers with clear track records of not trying to promote Bernie Sanders candidacy or not trying to promote Biden’s candidacy. Interviewing Tara Reid. Until then, we’ll have Meghan Kelly, who is to quote her Instagram bio. Happily married to Doug. No reason to think otherwise. Crazy in love with my children, Yates’, Yardley and Thatcher.

S4: Sure, why not? Journalist OK, but with our portfolio on the show today, let’s go to Sweden or let’s not. If you want to avoid Corona virus. But the Swedes are conducting an interesting experiment. Interesting because the results are clear, but they’re being treated like they’re quite unclear or even clear in the other direction. A smorgasbord of insight to follow. But first, Nina Everman, University of Tennessee professor who studies math and biology and vectors of contagion, is back. For every one chat I have with Nina Pfeiffer meant I cut it into two parts on the gist. So it’s an interview with an R, not of two point three and a Yelp rating of 4.5. Enjoy. Jennifer Afroman runs the Fetterman Lab, which studies how contagion’s spread contagion’s defined as specifically as immuno viruses and as generally as ideas. I began by asking Dr. Pfeifer men about this concern I have, which is these experiments that states like Georgia and Texas are undertaking. They’re opening up. And the conventional wisdom is we’ll get results and then we’ll get feedback from this experiment in about two weeks. That worries me because it seems more likely that it will take 20 to 40 days to really see what’s going on. And if we act on the data in only two weeks, might not the rest of us be led into some bad choices?

S5: You know what I’ve been worried about. I don’t know if you’ve thought about this or looked into it, but some states are opening, it seems to early. Different states are basing this on different criteria. Seemingly a lot of it is whatever the hell the governor wants to do. But, you know, there will be some feedback. And as I said on my show the other day, look, if I was a governor of Georgia, I wouldn’t open up to the extent that he is. But that’s going to tell us something. The data will be a little dirty. Georgia’s different from Tennessee, et cetera, etc.. I just worry that the lag on the feedback will be such that we might we might not have the ability to process it in time for the next state to make a decision. And I do think that other governors know this, that they think that, okay. They’ll open up and then we’ll see in two weeks what really happen. But I’ve been reading that it’s my take 20 to 40 days. And the perception that we’ll have good data in two weeks is perhaps more dangerous than not even going in with that perception at all.

S6: Yeah, yeah. So you’re completely correct. So we might start seeing some signal in 10 to 14 days, but it’s not going to be strong enough to know what what we’re looking at. And the three week to four weeks is really when we’re going to see the impact of partially because some of the states that are opening back up did have reasonably successful control for keeping the infection at the time at relatively low levels, not because they were doing anything so much better than other places, but just because the disease hadn’t gotten there yet. The population is sparser. It’s not as as centered. And in robust urban areas, testing wasn’t as good, but also routes of transmission weren’t as efficient. There wasn’t as much international or domestic travel to get to see those places. And so they were earlier in their outbreak when they shut down, which is great. But what that also means is that how much of the population needs to be infected by the gradual accrual of exponential growth. When you come out of shelter in place, it’s going to take longer. So, yeah, the two week mark is incredibly ambitious in terms of interpreting signal. And you’re right, if people are looking for it, it’s going to provide a sort of false reassurance that opening back up wasn’t that bad an idea. And it’s it’s the sort of thing where you don’t get to put the genie back in the bottle, opening up for three or four weeks and then shutting down is much, much, much, much, much worse and much harder to fix than even having stayed closed one more week and then opening it back up. So so there is a real danger in this sort of cascade effect of looking at early actors as indicators for the impact of their actions too early. And a large part of it really is just the social pressure of it’s hard to believe what feels to the average person, like a hypothetical instead of a scientific prediction based on math and science. And to the average person, it is true. Models are are thought experiments. And what I think we failed to communicate as scientists is that thought experiments are the basis of science. And when I thought experiments get good enough that they always help us make better choices and understand the dynamics of the real world. That’s when we turn them from. From hypothetical models into theory. And epidemiology is at the stage where for infectious diseases, this is theory. This is like saying if I release a bowling ball above your head, it’s gonna fall on your head because gravity, we don’t always understand gravity, but we know that it exists and we have theory to predict how things are going to fall on your head. We and we can use math to predict those trajectories. We use math to predict when comets are going to come back around and how the orbit of planets is going to work. And the general public doesn’t go, yeah, you know what? I’ll believe it when I see Jupiter again, they go, oh, OK. That’s when Jupiter is going to come. But right now, because we haven’t done the same level of explanation and we don’t have hundreds of years of astronomy to show to the general public about how this math works, they’re thinking of this math as a guess, and it’s not really a guess. It’s a tailored scientific theory based on different assumptions and those assumptions. Change, that means the different models make different predictions and those different predictions may be in conflict with each other. But that’s because they’re being asked slightly different questions. And so when we ask the same question, we get pretty good agreement from the different models. Maybe not to the exact number, but most of these models are not trying to predict an exact. They’re not trying to say seven. They’re trying to say seven versus seventy thousand.

S5: Right. Or, you know, a range between five and nine versus 70000.

S6: Yeah, yeah. And and it’s also the case. I’ve heard just personally a lot of pushback in the last few weeks of. Oh, but these models didn’t then predict, like early models didn’t predict what we’re seeing. And my answer is early models didn’t take into account the actions we’ve taken since they were to enable those actions. They were the warning signals that said what will happen if we don’t do anything? And then because of those models, it let us go. Oh, good. We should do something. There’s a model that says there’s a bowling ball over your head. You should step to the left. If you step to the left and the bowling ball then doesn’t hit your head. That model did what it was supposed to. Even though it predicted a bowling ball was going to hit your head and it didn’t. And that’s cause you step to the left.

S5: I do want to ask you about this idea that, you know, the models are the models are meant to be self-defeating. Right. Once you take into account once people start following the broad advice, well, then how do you know if the models ever work? If the design is to then look back and say, well, that was wrong. Well, how is it not just simply, you know, these models are just simply unfalsifiable? And you always have the escape clause of, well, you listen to what we were telling you.

S6: Sure. OK. So so there’s there’s three main ways that we actually deal with us. One is what’s called hind casting or bad casting, where what you do is is go back into the past and pretend you knew every action that was going to be taken, but not what it was going to do to the outcome. And so essentially, you have like precognitive ability for the interventions, but then you use the same model to build around those interventions and see if you can predict the outcome, knowing that each of those steps was gonna be taken. Right. You can’t do that in real time for a new disease because we don’t have anything to back cast. But this is how we validate the type of model itself. And we’ve done this a bunch with lots of different infectious diseases. And that’s how we know that the bones of the model are good. The skeleton is good. There is also the equivalent of proximate metrics. So even if we can never predict the actual future, if we always say try this and that thing always leads to decreases in new case incidents, or if you do this thing, it’s going to get worse. And we see that that when people do that, it does get worse. That’s not proof. That’s definitely not proof. But if we’ve got lots of different cases where the model makes predictions about the what ifs and those what ifs, even if we don’t have the counterfactual, if those what ifs are are correct more often than they’re not correct. That starts giving us a lot better idea that we’ve got a grip on the major players of what needs to be involved and what’s the other big way to figure out that the models have some some value. Awesome. So so this one is is the most remote, and that’s if the different types of models that we pick agree with each other. So there are a bunch of different types of models that we use. So I’m an applied mathematician. I use mathematical models. There are some really powerful statistical models. Those are very different from mathematical models. There are also simply some medical models of literally doctors doing different things in in sort of a scientific way of we’re just going to try these things and see what the outcome happens if if all three of these kinds of methods tend to start producing agreement. Then again, we start even though we are projecting into the future, we start having a little bit more faith that we’ve all captured something fundamental in a correct way. And again, it’s not that one is not great, but it is sort of a meeting of scientific minds in a way that gives us practical tools. So that’s not one to rely on by itself, but the combination of these three sets of tools of convergence in scientific inquiry, proximate metrics, giving us indicators that we were making kind of the right calls each time, even though it’s not directly the thing we’d really want to know. And back casting so that we can predict the future of the past in some sense. Those three together really do give us some good tools. And we’ve seen how these tools work over and over again to help us make decisions that have turned out to be really protective. And some of them are incredibly compelling.

S5: All right. Last question and then I’ll let you get about your work of thinking of solutions to save everyone’s lives. What are the circumstances by which this pandemic will play out that your profession broadly? You know. I don’t know how broadly scientists or at least epidemiologists modelers is. Valued and we say, wow, thank God for those guys and women. And they got it right. And what are the circumstances in which we. In my opinion, you know, missed the boat. And in general, people were like, well, why trust the scientists?

S6: Oh, I wish I had a good answer to that one. I think it’s so. Huh. So I think this is partially pessimism because I study the construction of reality by social convention. How people perceive events that they’ve lived through is strongly shaped by the post hoc narrative that everybody converges on afterwards. And so already I could see a narrative developing that said there. I mean, just a just among scientists right now, all of us are so deeply frustrated in the United States with where we are epidemiologically because all of us are looking back three months and going. If only we’d had better testing three months ago, we would not be in this situation right now. That already could be a narrative of the failure of the scientific community to have supported their existing enough tests and have rolled out those tests fast enough to cover the shortage more with more efficiency and rapidity. That’s that’s already a narrative that I could see building to failure. Or we just go, oh, well, well, what good is this? And part of the well, what good is this would be? Well, if you’ve never successfully communicated this to the people in charge, if you haven’t gotten their ear. If we have failed to lobby effectively, then then it is true. What what good is. Here’s how you keep yourself safe. If we’re just shouting that into the void, we’ve failed. And I’m equally guilty as as really as any one. I’m probably the most guilty because I usually love being an ivory tower academic. I love thinking my own beautiful thoughts and then publishing them and being like, well, when there’s a pandemic, you guys can call me and I’ll happily talk to you. But I do. I mean, I would also I spent 10 years consulting for the Department of Homeland Security on biosecurity and preparedness, among other things, and complex systems models. I was I was a researcher at one of their centers of excellence, and I loved doing that partially because there was this ready chain of communication where if something was going wrong, they could come to me and ask, or if ISIS something going wrong, I knew who to call and go, guys, fix this. So that center of excellence was a wonderful experience. But Homeland Security funds those in cycles. So it shut down as part of a natural process for them. Like, well, you had ten years and that was great. Now we’re going to work on some other stuff. I don’t know who to call anymore at Homeland Security. A bunch of the folks I would have called that I still had relationships with because of that center have now retired or moved on to positions in government elsewhere. On the turnover rate itself is something that I didn’t predict. The turnover at in the upper echelons of leadership in the last two years has itself been an incredibly problematic thing because those relationships went away and now literally nobody knows whom to call on in either direction. The people in leadership don’t know which researchers to call in which cases, and the researchers have no idea to whom to reach out. I think that’s easily a narrative of failure. So I think that those narratives of how did the science potentially fail are more about communication and the practice of science than they are about the science itself. And then the potential stories where we all get to be heroes are the ones where maybe we do come up with these off the shelf Cluj solutions of like, oh, crap. I mean, maybe it’s maybe it’s, you know, the next person thinking about how to build a better ventilator where people don’t have to be on it for long enough that they have really severe outcomes, even if they’re ventilated. Or maybe it’s you know, some people like me just looking at the news and being like, oh, God, restaurants should select us. Right. That’s right. It sounds it sounds silly, but if we have enough of those in parallel, not just to scientists, but that people thinking as scientists, people thinking, what are what are the engineering solutions based on the scientific understanding that we then get to enact that kind of save us from our own early failures of we could have. Yeah, we could have shut this down in the United States in February if only we’d had better testing and contact tracing. We didn’t. Oh, well, that means that that our wonderful, beautiful ivory tower epidemiological models for this are great tools for guiding benchmarks of how bad it can get and how good it can get. And we still need them. But we also need these really quick and dirty boots on the ground, like how do we prevent a worst outbreak in the next Smithfield meat processing plant? That’s it. I mean, I spent a lot of time last month working on models of jails because honestly, we had all of these models that said, OK, what about schools? What about assisted living facilities? What about nursing homes? What about hospitals without ever going, oh, my God. We have 2.5 million people who are basically stress and nutrition comp. Compromised, living in close quarters with poor access to medicine and hygiene. That’s going to be terrible for the whole country, not just the people in those facilities. These are things we’re quickly fixing. It might really help. And I could see that leading to like, yay, celebrate the epidemiologists as our saviors. Or I could easily see it, too. Wow, you guys are idiots. Why didn’t you know who to talk to about this when you realized it was gonna be a problem knowing America?

S5: I’m going to predict that lettuce in restaurants will be a bigger crowd pleaser than saving the jails. But still, I think they’re both very worthy.

S6: Having already gotten a huge number of e-mails from the concerned public telling me that I am wrong and bad about the saving the jail’s part, I believe pivot’s a lot of benefits. A modest Nina. Yeah. I’m gonna love doing both. I’m doing both. Because I think also what the people are missing about the jails is that it’s not the people in the jails. I mean, I personally also think that the people in jails shouldn’t necessarily die of Croner virus.

S5: Right? Right. But you could appeal to people even without any sympathy for the people in the jails. Just appeal to their own self-interest.

S6: Right. It’s literally the trolley problem of you write the philosophy problem of could you pull the lever to move the trolley from killing the one person, from killing three people to killing the one person? The answer on this is, if you do, are we really willing to kill more innocent people just to make sure we punish the people in the jails? And if the answer is yes, that’s not a society I’m necessarily on board with. But I will still quantify how many extra innocent people who have nothing to do with jail are going to get killed by. We just really wanted to punish the people in the jails. But yeah, I think the answer is yeah. Let us who’s gonna be much more popular?

S5: The Everman lab among its chief interests are mathematical modeling of infectious disease dynamics, like how individual behaviors and contact patterns cause different outcomes and the level of population sound applicable to the current moment. It is Nina Pfeiffer man gives the Heiferman lab its equanimity. I think I think that is the way to look at it. Nina, thank you so much.

S6: Thank you so much for talking with me. It’s always a pleasure.

S4: And now the spiel, Sweden is a country known for its beauty. It’s easily assembled furniture, its blondeness and its caring relationship for its elderly population. As detailed in the documentary Midsummer about an elderly person experiencing hardship and the entire village responds collectively. Cue the giant hammer to address the situation. In keeping with Thomas PIOs, Hammer and dance analogy. OK. I joke. Midsummer is a horror film, a folk horror film. But the folks in real Sweden have made some horrible choices. Maybe not because they’re unfeeling or stupid, but because they have a different take on how to combat Corona virus. They went with no real lockdown’s but advice advice to distance a bit. And if you say go to a nightclub, keep a few feet apart. Interviewed on the NPR podcast. The Planet Money Indicator about this, Sweden’s chief epidemiologist, Anders Techno was asked, Does it ever worry you that you’ve made the wrong decision?

S7: You know, I think everybody, least everybody who has any kind of a little bit to command over it about the decisions. Well, taking, though, everybody in every job is taking decisions that were taken before.

S8: Honor says that he and his colleagues, they themselves had been discussing this constantly at the public cafe that they all go to.

S7: Of course, they’re wondering if this is the right thing to do. I think we’re doing the best things we can figure out, giving the circumstances some of the tools, 100 days. And we’re probably going to discuss this for decades to come.

S8: If this was the proper thing to do or not Under says so far, the Swedish public has been very supportive of their approach. He gets a lot of e-mails from people encouraging him to keep going. And he says there are signs that the death rate in Sweden might peak next week. But if it does not, if the death rate does continue to grow, then the government would likely have to consider a full lockdown.

S4: So that was almost a month ago. Yesterday, techno was on CNBC and other forums, armed with the statistics that he said would dictate his response, statistics that said he would be driven by and would react to accordingly. Well, when that interviewed the Planet Money, interviewed that I played, was taped. Sweden had never experienced a day with more than a hundred covered, 19 deaths. The day it was posted, 114 people died. The death toll roughly tripled after that interview for a country of 10 million to have recorded about 3000 deaths make Sweden one of the worst countries on earth for dealing with coronavirus 14th, an overall death, even though the population ranks eighty third in the world. Sweden has almost 30 deaths per thousand, which is higher than the US. It’s actually Sweden, twenty eight point eight eight. Let’s compare it to its neighbors. Denmark eight point seventy three per thousand. Finland 4.5 seven. Norway four point zero six. Its policies have led to the deadliest outcomes in Scandinavia by far by a factor of three to seven. It is abjectly speaking one of the worst outcomes of any country in the world, which would be sad if that’s all it was a strategy that did not pay off. That is at least being learned from. But that is not the lesson learned. Techno was on The Daily Show a couple days ago. He told Trevor Noah this.

S9: Everybody in Sweden who needs a hospital bed go to the hospital beds for corporate Nightengale for other diseases. So it kept on working an intensive care. Stephen had at least 20 percent of the beds free to any given time. So in that respect, the strategy has been successful. We managed to keep the level of spread on the level below the threshold of the health system, which I think is one very important part of any kind of strategy.

S4: Techno was also on CNBC today and said, quote, Of course, there is a huge regret over the fatalities, but we’re not really clear how that could have been avoided. Asked if Sweden could follow the same policy in any future outbreak. He answered, to a great extent, yes. And of course, Americans are taking that confidence or that answer and running with it if it suits their predispositions. So here is Clay Travis, who hosts a sports show called I’ll Kick the Coverage. But he also talks a lot about corona virus and freedom in the mainstream media as indulging in Corona virus fear porn.

S10: Sweden never shut down their schools. They never shut down bars and restaurants. They already are working rapidly towards herd immunity and the country, 26 percent of the country has already been exposed to the corona virus. And they did a lot better than many other European countries. And I think it’s gonna be a fascinating study in the years ahead to ask the question, was there any benefit at all to the lockdown?

S4: You know, I think it’s a fascinating and somewhat predictable study and motivated reasoning and an almost literal inability to assess clear facts when your prior positions want to create blindspots. Now, look, that policy might work for the Swedes, technical, by the way, as an actual scientist. And he will tell you that chasing herd immunity isn’t why Sweden underwent its experiment. Technol is pretty popular and most Swedes. All right. In a state of disinformation, they know there are a lot of deaths now and they hope that next year there will be a big drop off in the usual deaths.

S1: They see the idea being that the disease ravaged the elderly, nursing homes, people who were, to be blunt, going to die anyway. You don’t hear that sentiment in America. The Americans who promote the Swedish experiment do so by comparing the country to one of a few other European nations that are actually hit harder and also by downplaying the glaring statistical evidence in front of them. It’s not only America. It’s not even the worst of America. The New York Times had an article that was way too positive about the Swedish experiment. The Spectator, the British publication, once edited by Boris Johnson, ran a headline a few weeks ago that said The Swedish experiment looks like it’s paying off. It is not. I direct you to this headline from The Wall Street Journal a few days ago. Sweden has avoided a Corona virus lockdown. Its economy is hurting anyway. In Sweden, the experiment failed. On one front, maybe two, death in the economy. It didn’t save more Swedish lives, but they at least understand the facts and the implications of the facts. But in America, this Swedish experiment is a failure on all fronts. It failed to save Swedish lives. And those pointing to Sweden as a success are failing to see what really happened. I’m all for learning lessons when they occur, but that depends on an openness to facts and a basic mental capacity to interpret what those facts actually say.

S11: And that’s it for Today Show the gist, associate producer Margaret Kelly points to Sweden as an interesting experiment in having a reputation as stout, reliable and sturdy, but also making every piece of college dorm furniture that instantly cracks when you put a plate of spaghetti on it. Daniel Shrader, just producer, points to Mongolia as an excellent example of a country that is an anagram of mango oil. The gist when considering that clip of mid show ma, you need to thank me that podcasts are simply an audio medium. Desperate to brew. And thanks for listening.