The Failure of Expert Predictions and Models | The Coronavirus and Public Policy

hi I'm Alex Berenson I'm a former reporter for the New York Times and the author of the book tell your children the truth about marijuana mental illness and violence I've spoken to Hillsdale audiences about that in the past I'm here today to talk to you about the corona virus our modeling of it and what the failure of the most important models we've used to predict the course of the virus mean and what they should tell us about lock downs and our strategy going forward so let's talk a little bit about models and what they are and why they're important at the you know at the most basic level the corona virus is a is a novel illness it's a novel virus there are other kinds of corona viruses but when this one emerged in China in January to 2020 we didn't have very much idea about what it was how dangerous it might be medically and you know at a cellular level how it affected people but though the medical and biological issues are are sort of outside of my domain I'm more interested in the the submerged level the population level effects that we're trying to to figure out you know I'm not a molecular biologist I wouldn't pretend to know how the corona virus infects a cell on the you know on the most basic cellular level but I can tell you whether or not the predictions that are made about the transmissibility of the virus and the the ultimate population effects of the virus are what we're seeing in reality that's that doesn't require a PhD in molecular biology just requires the ability to compare different sets of numbers and track them in real time which is really what I've been doing for the last month or so so models matter why do they matter they matter because people are using them you know in in hospital systems at state levels at national levels at international levels to figure out how many people might become sick from the corona virus how many people might need ventilators or ICU beds from the corona virus how many people might die from the corona virus obviously if 10 percent of the population is going to die you know we need to take really drastic steps if 1 in 1,000 people is going perhaps the coronaviruses more like the flu and we can treat it more like the flu so we need to know where the epidemic is going to go in the days and weeks before it gets there or that's certainly what we're hoping to do and the models help us get there and the science of Epidemiology and infectious disease has been around a long time it's been it's been really developing for more than a century and a half so there's a you know there's a lot and epidemics come and go over time so there's a lot of history here there's a lot to draw on here so when people are talking about models what are they talking about they're talking about trying to predict the course of the epidemic in the coming weeks months you know ideally years although as you go further and further out it becomes harder and harder to predict how do you do that well you need some basic facts about the virus you need to know how transmissible it is in other words if I get sick how many people on average am I going to sick you know if if I'm only gonna sick in one person the virus the the epidemic is gonna spread slowly if I'm gonna sick in 20 people as actually can be the case in the case of something like measles the epidemic is gonna spread very rapidly we also want to know how quickly that transmission cycle takes place how many days am I likely to be infectious how many days am i likely to be infectious before I know I'm infectious before I'm showing symptoms that's a crucial variable just to give you an example of how big the uncertainties can be here if the are the transmissibility is - meaning one person infects two other people and the cycle is six days over the course of a month one infection might become 30 infections which you know that's that's obviously a big change but if the R is three one person is sickening three people or a little bit more than three yeah three point one to be precise and and the transmission cycle is four days over that same month one person can sicken or it's not that one person will sick in 10,000 people but one infection will become 10,000 infection so in our of three a transmission cycle of four one becomes 10,000 an R of two transmission cycles six days and R becomes a bit more than 30 over the course of month these are huge differences and on four end so that's one thing you need to know how fast the virus spreads we need to know where it goes we'd like to really know what what vectors are most likely to spread it in other words our children spreading this to adults or adults spreading this to children does this spread if we're outside just in talking to people or does it require pretty close contact you know is in-home transmission the main method of transmission is transmission on public transportation the main method these are these are more subtle questions but they're they're also crucially important in terms of Devon devising a societal response and we want to know can they are change how quickly you know what might drive our up or lower it does a there's a hard lock down lower it very fast as modern social distancing lower very fast those are those are you know again those are sort of second level questions before you can even get to those you need to know the basics about transmissibility and the reproductive cycle of the virus and again some of that depends on the biological characteristics of the virus which are you know which scientists in labs all over the world are trying to figure out but some of those depend on for lack of a better where the sociological characteristics of what we're doing are we pushing people into emergency rooms are there is there a Mardi Gras event where you know tens or hundreds of thousands of people are getting together for a parade all of those things are not actually so much about the virus as about us and so they can change very rapidly more rapidly actually than the than the fundamental characteristics of the virus so that so so transmissibility is one key number and then the other key number is the fatality rate well the hospitalization rate but ultimately the fatality rate of the virus how many people who get this are going to get sick and how many of those people are going to die and obviously again a virus that kills and you know something like Ebola might kill half the people and in fact something like HIV before there were anti retrovirals would kill 100% of the people that infected you know with a you know with the with a very unusual exception or to everybody who got HIV ultimately died of of AIDS so some viruses are extremely Lethal other viruses like the flu much less the common cold are much less lethal and so that and the spectrum really can be anywhere from zero to a hundred percent and early on as you're as you're figuring out is you're trying to figure out these models you may not have great information about the fatality rate you're making some guesses and that was really very true in this case because a lot of the data was coming out of China and I would say you know people are pretty suspicious even now of some of the data that that the Chinese government has reported you know it's not entirely clear that they were counting cases accurately they kept changing the definitions of cases you know there are people who don't believe the fatality numbers that they reported the death numbers that they reported you know I don't really I try not to engage in conspiracy theory but you know China is in a an authoritarian state and I do think there are there are legitimate questions about the data that came out of China in the first month so the people who were who were who were charting these models who are creating these models especially in March as the as the epidemic spread you know outside of China to Europe and then to the United States we're doing so with somewhat limited information and they had to make guesses and when you're making these models I think there's a natural tendency probably to want not to err on the side of under estimating the danger he did so if you if you if you overestimate the danger maybe you know maybe you cause problems with the economy or other problems that you know these people who are making the models don't necessarily focus on but if you underestimate the data people die is I think you sort of the so the bias tends to be to overestimate the risk here and so again you need to we all should remember that as these models were being created when didn't have great data about either the transmissibility or the infection fatality rate and by the way the infection fatality rate can change too it can change for a number of reasons it can change if we get pharmaceutical interventions so you know people people have talked about antiviral drugs or you know or zinc and Zi packs and hcq and it's not clear by the way that any of those interventions actually work they need to be tested in placebo controlled trials where they're tested against basically nothing so we can see if they're better or worse than nothing because making guesses on the basis of anecdotes is not a good idea but over time you would expect and hope that physicians will come up with treatments for these you know for the corona virus that will reduce the fatality rate another another and there may be you know physical interventions like literally just having people lie on their stomachs it's called proning them that appears to actually help some people not become to not need to go on ventilators not become too oxygen deprived and so physicians learn and they learn quickly and but at the same time we need to make sure that they're not just going on anecdotes so it's so it's a very complicated process but the infection fatality rate can change and and it can change the other way you know I think there's increasing evidence and increasing debate in the medical community about the use of ventilators and whether whether if you overuse ventilators you might wind up hurting some people because of some specific characteristics of the way the corona virus affects the lungs and so ultimately if you use ventilators too aggressively you might even wind up killing some people who might otherwise have survived and that's you know that it's a very tricky balance you know I'm not a physician my wife is a physician and she actually treats coded patients and and I'm very respectful of the of the choices that physicians have to make in some of these cases with very limited information but the point is the infection fatality rate can change and so can the our okay so that's that's what the people who are creating the models those are the limitations they're facing and it's a little bit think of it a little bit like a hurricane prediction at some point if you have too many uncertainties the prediction is just useless so if if I'm predicting that a hurricane might hit Miami on a Tuesday in August and it might be a category three to category five that's a pretty you know I know and I know the date it's you know Tuesday August 21st I'm just making that data up but you know it might be Miami or it might be Fort Lauderdale or might be you know a little bit further north that's a good enough prediction that we can do something with that we can evacuate South Florida we can say to people we don't know that this is gonna be the worst thing ever but it could be and even if it's not it's gonna cause some problems so let's let's get you know let's get people in South Florida out of of Miami let's get them you know let's get them to someplace it's a little bit safer if my prediction is that there's gonna be a hurricane sometime in August that's gonna hit somewhere between Miami and New York City and it might be a category one or it might be a category three or it might be category five that's useless and it's especially useless if my policy prescription as a result of that prescription or that prediction is not just hey we need to evacuate the whole Atlantic seaboard but let's burn the whole seaboard down and to some extent that is the you know that is the choice that we've made with these lockdown so so at some point the prediction becomes too nebulous to be of any use no matter how well-intentioned it is and you know unfortunately what is obvious when I when I use the hurricane analogy was not obvious last month when we were talking about these models so the first model that really grabbed a lot of attention came out on March 16th and it came out from Imperial College in London and what it said was that two million people in the United States and 500,000 people in Britain would probably die from the corona virus if no steps were taken and that even if pretty aggressive mitigation as the authors called it was used the deaths would be at roughly half that level so more than in the u.s. a quarter million in Britain now the people who did this were led by a professor named Neil Ferguson who's a British professor who's a you know a very well respected epidemiologist I think what was forgotten at the time although it's become I don't say obvious but it's become noticed since is that professor Ferguson has a long history of making very aggressive predictions about death and from from epidemics he predicted in 2005 that as many as 200 million people might die from the bird flu that was that was around that year he predicted in 2001 that as more than a hundred thousand people in Britain might die from mad cow disease and he also predicted that it might lead to sheep and create mad sheep disease and that ultimately 150,000 people might die now obviously those estimates turned out not to be the case but nonetheless when the Imperial College model came out it was very it received tremendous attention in both the US and the UK and you know the New York Times oh she wrote a story a front-page story on March 17th talking about the impact that the model had had on policymakers in the US and the UK and and I'd been paying attention to the coronavirus pretty aggressively since since early or not not early since late January since you know that the numbers sort of came out of China and I and I was concerned about it and I would say that I was a little puzzled in late February because we didn't see the epidemics the secondary epidemics in other china mega cities that that were being predicted after the initial Wuhan epidemic we didn't see Beijing get overrun we didn't see Shanghai get overrun you know the epidemic seemed weirdly confined to Wuhan and or you know in Hebei Province and these other cities even after they began to you know reduce their lockdowns and quarantines they didn't show huge spikes so that puzzled me but then of course Italy northern especially the hospitals faced a terrible damage and that was you know that was very scary and then the the you know the paper came out on March 16th and I think I read it that night and what really struck me was the age stratification in the data meaning people over 70 and certainly people over 80 of account from the majority people over 70 come from the vast majority of coronavirus desk people over 80 account for the majority my best estimate is that worldwide more people over 100 have died than under 30 believe it or not and more people over 90 have died than under 50 so this is a this is an epidemic that is really really stratified by age your risks are much much higher if you're older which doesn't mean it's not serious and it doesn't mean that the deaths of people over 80 or 90 don't matter but it does it does suggest that the the virus itself is not so terribly dangerous if it isn't hurting younger people you know it's not the Spanish flu which you know which which which killed a lot of children a lot of young adults and a lot of older people so that you know so the Spanish flu analogy probably isn't a great analogy so so I looked at the at the Imperial College data and Imperial College was very explicit about this that to get numbers down to some reasonable level we were gonna have to suppress the they use the term suppression suppress Society for a very long time they said 18 months but if you actually read the paper it maybe the paper makes clear that 18 months is just essentially a number that they're using to say it might take 18 months to get to a vaccine and if we you know basically until we get to a vaccine suppression is gonna have to continue if we don't want these terrible deaths tools to mount a you know a million people in the US a quarter million in the UK and and so you know that that scared people even you know I think I think the top line number got a lot of attention understandably and people focus less on the age stratification and you know the the big gap in risk by age so so then what happened was I believe was March 23rd Neil Ferguson gave testimony to Parliament and he said well you know our new best estimate is twenty thousand deaths in Britain with it with a couple of weeks of lockdown and and that got some attention in the UK it didn't get any attention in the US wherewhere attention was becoming very very focused on what was happening in New York City and the death counts in you know in the hospitalizations in New York City and Governor Cuomo predicting that we might need as many as 140,000 beds here in New York State hospital beds as many as thirty to forty thousand ventilators really terrifying stories coming out in the New York Times about how you know a million people might need to be on ventilators which one Times reporter then said meant that a million ventilators might be needed which was an absurd statement in under any circumstance because a ventilator useless without the trained you know nurses and doctors and medical staff to run it and so there's no circumstance under which a million ventilators ever made any sense but but going back to Britain and to Professor Fergusson professor Fergusson who by this time had actually contacted the coronavirus himself been infected with the corona virus yet suddenly changed his estimate to 20,000 and he did this again it wasn't quiet it was it was it was before a parliamentary hearing in the UK but it didn't get very much attention and he said well we think the transmissibility has gone is higher than we thought so so what does that mean well if the transmissibility is higher that means a much larger group of people have gotten this over the course of the last couple of months than we realized and that's a good thing the more people who've gotten it that means that the number of people walking around with no symptoms at all is higher than we think and that means that the virus is less dangerous than we think and now if if if 10 million people are infected and 10,000 ultimately die that's a 1 in 1,000 death rate that's like the flu if if if a hundred thousand people get it and 10,000 die that's a 10 percent death rate that's nothing like the flu that's much much worse and so and in that case we have to be terrified that ultimately you know those 10 million people will get it of those will die but what what but Professor Fergusson was saying was we think more people have gotten this and therefore it just doing the math it can't be as dangerous as we thought although he didn't explicitly come out and say that so I pointed this out in a couple of tweets the next day once I realized what he had said and and that's really that was about a month ago that's really when I started to become you know a little bit of people call me a contrarian or skeptic on this I think I'm just trying to point to data that you know that other that other outlets may not be pointing to but so so professor Ferguson then said well we didn't really change our model five hundred thousand two hundred fifty thousand twenty thousand all those numbers are in our model and that is true they're all in there but what he didn't say was that he fundamentally changed the assumptions around the model around the around the deaths around the twenty thousand death figure whereas before the assumptions in I'm talking about in the paper you can see it in the paper the assumptions are we need again 18 months of suppression to get to twenty thousand deaths even if we have just a few months of mitigation of somewhat severe societal measures but not a full lockdown we're gonna have a quarter million deaths all of a sudden he's saying well with just a couple of weeks of lockdown I think we're twenty thousand deaths and he also said explicitly most of those people will be will be elderly or quite sick or in and have less than a year to live they would have died anyway by the end of the year I think was his phrase I I can check that and so so you know he said twenty thousand deaths in the UK was the most likely estimate and again most of those people already already old you know already would have died quite soon anyway and so again I pointed this out he then responded not directly to me but you know he was responding clearly to the fact that I had pointed this out and said oh well we didn't really change anything and there were you know there were some media outlets that effectively accepted that narrative and nobody seemed to go back and say well professor Ferguson what about all these other times when you you know vastly overstated estimates for death counts so that so so the Imperial College model was the model that really grabbed the tension in the US and the UK and really pushed policymakers into a lockdown into lockdown mode I mean it certainly got the attention of the White House and it started this you know this sort of domino effect where first some states on the west coast you know lockdown and then New York said they weren't going to and then they did and then more and more states followed and you know there was increasing national pressure for the few states but you know by I would say the end of March the the last week of March the states that hadn't locked down yet we're feeling you know considerable media pressure and some public / political pressure to lock down and most of them ultimately did so so the Imperial College model you know was changed in a way that that that again again if you read the original paper it's clear that change that you know that that it was changed and I I don't I don't think that anybody who reads with a with a fair mind could could could dispute that but but actually the model that has become more important in the US and and is even enemies and is more provably wrong and more consequentially wrong in the last month is the University of Washington model okay so what is the University of Washington model well on on March 26th the University of Washington Institute for Health metrics and evaluation released a very detailed model state-by-state and actually they've extended it to many countries saying this is how many beds hospital beds we think states are going to need this is how many ventilator states are going to need this is how many deaths were gonna have and so and so this became a guide to state planning and what the March 26th model showed was that all states practically but especially New York State were in terrible trouble and and so and so in New York State I believe the the one of the or maybe the second iteration of the model what the prediction was it was going to need 76,000 hospital beds and you know in in in tens of thousands of ventilators an attic see wasn't as bad as the state's own projections which again were as many as a hundred forty thousand hospital beds and thirty thousand ventilators thirty to forty thousand ventilators okay and it and in many other states the predictions were quite you know they were extremely high ultimately I think they you know the the April fifth version of this predicted two hundred and sixty thousand hospital beds needed nationally which is a you know an enormous number okay so where did this come from here's what happened in New York City a'right you need to understand what happened to New York City from March about March 15th to March 25th hospitalizations increased dramatically they increased almost tenfold so you know in mid-march there were a hundred people showing up with kovat related illness in New York City hospitals are being hospitalized by March 25th that number was over a thousand okay and the key metric here is hospitalizations why because deaths are a lagging indicator and by the way right now in the United States because the death coding is so I would say aggressive and loose it's not even clear you know in some cases people who didn't even have the corona virus might be counted as a corona virus related death at the same time in places that are really badly and we know that we're really badly impacted which basically means in New York City it's possible that there were some people who died who weren't counted even though they did die of coronavirus oh so the death counts are complicated hospitalizations are or a better metric for the most part because no good doctor is gonna hospitalized somebody you know for coronavirus or anything else who doesn't need to be hospitalized so here's here's here's the crucial part of this you become infected with corona virus it takes several days before you develop symptoms on average about five days some people develop more rapidly some people it takes longer then you get worse over time and over a period of days five days six days seven days ultimately you you get sick enough that you're in the hospital and from there you oftentimes you progress pretty rapidly you you know you might wind up in the ICU or on a ventilator or you might you know you might die or in some cases many cases actually you don't you know you you you feel better after a day or two in the hospital and you're discharged okay so but the main thing is there's a lag there's a lag between infection and symptoms symptoms and hospitalization in New York City the lockdowns I believe it was the state Governor Cuomo announced the lockdown on I want to say Friday March 21st and they were to take effect on March 23rd Sunday March 23rd what people were afraid of and what the models essentially predicted was that the lock downs occurred too late okay and that in the 10-day period before the lock downs took place there was going there had been a massive surge in hospitalizations and that surge was going to continue I'm sorry a massive search and infections in fact day after day infections increasing and those increased infections were going to lead to a massive wave of hospitalizations that wasn't going to peak on March 25th because that's too close to the lockdown but that was going to continue the increase increased day after day after day until five or six or seven thousand people a day were being hospitalized for this you know and and the hospitals were not gonna be able to cope with this and that that was what was terrifying to people that not that the lock downs were were we're going to work but that the lock downs weren't going to work because they taken place too late that is the core assumption at that you know even though it's not explicitly stated it is the core assumption in the University of Washington model certainly for New York State but what happened wasn't that at all what happened was that hospitalization stayed relatively high for several days really ten or more days following that March 25th date but they didn't continue to increase exponentially and so every day the models got a little bit more wrong so bye bye-bye you know early April the models were predicting or I shouldn't say the models there's one model it's the University of Washington model they're predicting that that as many as 60,000 people would need hospitalizations in New York State but in reality the number was more like 15,000 and and a gap continued to increase and so looking at that I sigh I realized wow there's something very you know there's something very seriously wrong with this prediction because this was released on March 26th and it's not trying to predict three months or a year what it is trying to predict those things in the future or five years in the future its predicting days in the future and it's completely wrong it's as if I said you know this hurricane again to come back to this I say on August 31st this hurricane is going to hit and it's gonna be a category 5 in Miami on September 3rd and the hurricane comes ashore on September 3rd but it's a category one so so the the the predictions were just way off and they again they were way off because they assumed this surge in infections had happened that didn't seem to have happened and we still don't quite know why that is but but one you know one good theory is that more people were infected earlier than we know so again that means that although there was a surge in infections in New York City in those days ultimately fewer people wound up in the hospital than we realized so so it was it was it was my prodding at the University of Washington model day after day that that again I would say increase my public profile a little bit and increased people's questioning about this because those those two models the Imperial College model and the University of Washington model led to the lockdown and are continuing to leads the lockdown and one of the one of the amazing things that happened was the University of Washington continued to update its model they would they updated it they've now updated I believe five times possibly six times in the last you know in the last month but but on April 5th they changed the predict projections and they changed them from from about 260,000 hospital beds needed to about a hundred and forty thousand hospital beds needed what was bizarre was even as they changed the model that day they didn't make it match what the actual numbers were in New York State that day so even the new model was wrong at the time it was issued and once I saw that I thought to myself this is really bad you know this is this is not this is not okay it's not okay that this incredibly important you know predictor that's driving public policy isn't even being corrected to what we know is actually happening in New York State and other states at this time now the models to be clearer have been more accurate about death counts now that's there's a couple reasons for that first of all again deaths are a lagging indicator and again and second of all there is this issue about how we're coding for deaths but the what we should care about more than anything else is hospitalizations because Hospital both because hospitalizations are a leading indicator and tell us how seriously the disease is you know for the general population and not just for a small group of elderly people who you know who are really at risk but also because if you remember a month ago when the lockdowns began the argument was we need to do this to save our healthcare system it is not we cannot have a situation where every hospital is going broke is overrun with kovat patients and can't do anything else and we are working our doctors and nurses and medical staff to death we can't have that happen that's and that's a totally reasonable way to look at it the problem is the models predicted a destruction of the healthcare system that hasn't happened at all not even in New York City in New York City got stretched and it probably you can argue guys stretch close to its limits but the overflow capacity that was brought in you know the the the hospital ships and the field hospitals have never been full so even by even New York City survived without being overrun and elsewhere this incredible paradox has happened and this is something I was reporting on a lot a couple of weeks ago I'd say at this point it's beyond dispute where hospitals all over the country are emptier than they were a month or six weeks ago and they are bleeding cash because they have postponed elective surgeries because people aren't coming into their emergency rooms because in some cases really necessary surgeries are being postponed and so our hospital system believe it or not is on the brink of collapse not from Kovac but because we reacted to these models by shutting down a ton of stuff that hospitals need to be able to survive financially and hospitals in the United States have already closed in the last month they've laid off employees and some have Believe It or Not closed so we are in the midst of the worst pandemic supposedly since the Spanish flu and our hospitals are closing for lack of business and once you realize that you realize that how important it is that these models have been wrong so so I would just I just urge everybody going forward to understand that you know there's been many predictions of doom made in the last month there's been the just wait two weeks predict there's in the model predictions there's been the wait 2 week predictions there's been the second wave which by the way is basically meaningless because the second wave could happen at any time it could happen next fall or next or in the spring of 2021 or you could happen in two years I mean you know at what point that we do we say okay there may ultimately be more deaths even significantly more deaths from this but we can't stay you know in our homes with no economy forever and and so so it really is the models more than anything else that have driven this and it's really important to understand how far from reality they've been and that's and that's what I've tried to do as much as anything else the last month anyway I thank you very much oh the one other point I would make is about testing so yeah that one thing that people have said is we need we need many many more tests well the truth is we don't necessarily need many many more tests to know if people are infected as long as they're not symptomatic or needing hospitalization you don't necessarily care if you have the flu or not if you're you know if you feel lousy for a couple of days but you you know you know it's the flu and you know it's gonna pass you probably just stay home for those couple of days the corona virus tests are necessary if corona virus is going to you know kill a lot of the door you know or seriously injure a lot of the people who get it and we have good interventions for them what is more important than knowing how many people are infected now what is much more important is knowing how many people have been infected and recovered and for that it you need a different kind of test which is called an antibody test which is a very simple test that just shows whether your body has produced an immune response to the virus and these are they're your they're relatively cheap they're relatively quick to do and they've been conducted now in a number of states and in a number of countries and almost universally they have shown that many more people have been infected and recovered from the corona virus than have active infections just yesterday New York City of respond number of New York State published a number for New York City showing the 21% of people in the city have been infected and recovered they have antibodies and by the way that number is probably low because it takes a little while to develop antibodies so over the course of a couple weeks the number tends to rise yeah there's no guarantee of that but it's likely but in many other jurisdictions there have been similar findings you know in a school in France a single high school in France 25 percent of students and teachers and and parents were infected in sweden and stockholm eleven percent in california it looks more like three or four percent in you know in geneva switzerland six percent and again that does not mean that there's not more you know virus out there that's gonna you know there's not gonna be more misery or you know some people might be infected and ultimately die because you have to get to a much higher level than that to have what's called herd immunity where the virus basically has infected everybody it can but those numbers suggest that the infection fatality rate is well below 1% and and probably below 1/2 percent and so and so that suggests again not that this is not real and serious but that it is less dangerous than we thought but one of the bizarre things that's happened is that the people the same people who made these models many of the same people you know based on what now looks like faulty data and questionable assumptions now that we're getting really hard data from the serology tests from these antibody tests they're pushing back and saying oh well the test those anybody test may not mean anything there may be false positives or or there may you know this means that so many people have it and don't even know it that it's more dangerous I mean so so that first question about false positive is a technical question but this idea that there's antibody tests that um that many people are infected and that's a bad thing and have recovered is a bad thing that's just nonsense there's that that's a that's that's the opposite of reality and why people won't acknowledge that I don't know so so so yeah so let me leave you with the testing and the fact that by all accounts there's now considerable evidence that the corona virus has spread much more quickly or and much more widely than we realized and and that's a good thing the more of us you know whether in the United States or anywhere else who have gotten this and recovered from it the better off we are so thank you very much I'm Alex Berenson and I appreciate having had the chance to talk to you hello and welcome to the second half of our program with Alex Berenson he just given us an excellent talk on the failure of expert predictions and models my name is John Jay Miller I'm director the Dow journalism program here at Hillsdale College in Michigan and Alex welcome back and thanks for agreeing to take a few questions John it's a pleasure I want to start with a big one which is this you've given us this great account of the problems of expertise and what experts have been saying we're seeing millions of people on the unemployment rolls you mentioned the problems hospitals are now facing have experts led us into making a gigantic public policy mistake well you know I I don't think that's exactly the right way to think about this you know the experts provide data and predictions but it's the job of political leaders you know executive legislative and it's the job of all of us to try to think about the costs and benefits here and again there there is a world in which the lockdowns make sense though you know that's a world in which a huge number of people are dying from this and we really have no choice but to you know have a crippled Society for a long time until we fix this that that is not the reality of what we're seeing here even the worst case projections don't fit that reality and so it's up to all of us to you know to show some critical judgment here and not just hide out and I can tell you I mean I don't know what your experience has been but so again my wife is a physician she does treat kovat patients and we've made a conscious decision that we're gonna try to keep our family you know functioning as normally as we can we're gonna go out with our kids we're gonna take them you know to the grocery store and stuff like that they're not going to be terrified because I truly believe that the mental health impact on children here far far outweighs any physical risk that they have but but most people we know aren't doing that at all most people are hunkered down cowering in their homes I know I can only imagine they're watching media outlets that are putting the worst possible spin on all this and they're not looking they're not exercising any critical judgment looking at any scientific papers looking at any government data for themselves and and I know that you know against a lot of this stuff is stuff that is beyond you know the the an average person's understanding certainly some of the technical stuff around the you know the biology of the virus but anybody can can look at a projection and look at what's actually happening that day and judge for himself whether those two numbers are the same and then ask what it means if they're not the same and and I just I feel like we've all let ourselves down what's a political leader supposed to do it someone's listening to public health officials listening to these studies they're saying all kinds of different things some of them are very scary and frightening you don't want to make a mistake you don't want to make a decision it's gonna lead to lots of deaths how is a political leader supposed to judge between different kinds of expert opinion and and land on the right course well I think this is a great question and here's the thing the situation a month ago was different than it is now things looked scarier a month ago than they did now we have more data and we know that outside of New York City and possibly a couple other hotspots there hasn't been a huge overrun so so when more data comes in if you care about data if you care about science you should adjust your response beyond that a political leader you know you're not just the the president of a nine year old in a nursing home you're the president of a six year old in a school who you know who may not have you know who may have parents who are abusive or neglectful and being seen by a teacher that day is the only adult who's gonna see that child who can make sure that he or she is being taken care of and not being abused you're the president of a small business owner who's invested his life in a you know in getting his dry cleaning business opening this week and all of a sudden it can't exist anymore it can't work anymore but he can't work as non-essential we have to make trades and and shutting the economy forever is not a fair trade or a good trade and there are things we could do that would mitigate the harm to the most vulnerable people here who again are overwhelmingly elderly and you know and even within that population the nursing home population is especially vulnerable and if we focused our attention on protecting those people if we had more temperature checks if we tried to make sure that staff were being tested regularly for the corona if we if we said you know what if there's any outbreak we're gonna make sure that all these people get the best care we can give them if we did that preferentially maybe we'd be able to actually reduce the number of deaths and and at the same time have a functioning society so we're the lockdown and shutdown orders where they were they a good idea based upon the information that that leaders had at that time maybe not the best information or the best information available but maybe not good information I mean I think that's a that's a point that we can argue and we'll argue about for years but to me you know it's more important right now April 24th what's the best thing to do going forward what's the best thing to do now that we have studies out of Germany and other European countries showing that social distancing without a heart lockdown actually reduces the infectivity of the virus in a way that prevents these searches from happening what's the best thing to do if there's increasing questions about whether or not the panic that a lockdown can cause in the days before it's actually happening you know but when it's announced but before it happens might be driving people to emergency rooms might be causing them to trade the flu and coronavirus and other illnesses and actually gets them sicker and and you know increases mortality as may have happened in New York City last month I think I think it doesn't really matter you know it's it's for historians and you know and I guess investigative reporters ultimately to chase down how the decisions that were made last month were made I'm interested in making the right decision right now so Alex what's the right public policy solution right now we're having this conversation on Friday April 24th and today the governor of Michigan just extended the stay at home order in our state she loosened a few restrictions there are a few new ones but what would you recommend to the governor of Michigan really any governor if this person was to call you and say Alex Berenson what should I do in my state so there's a couple harder answers to that and there's a couple easy ones let me go with the harder ones first I think in most states there there's not a hospital system crisis which is almost every state we can very quickly look at reopening offices reopening retail and then you know in the next couple of weeks maybe at hospitality you know at bars and restaurants and then ultimately at you know at at big events like like concerts or sporting events there's very very good data showing that most of the transmission of this virus is intramural or in hospitals and nursing homes or to some extent on public transportation that all the other methods of transmission are secondary and so so opening offices opening retail opening construction letting people get back to work makes sense there are risks in it but that but it makes sense but here's here's two things that are very low risk that we should be doing right now right as in today we should be letting people go outside with no masks and you know if they want to wear a mask that's fine but outdoor transmission is not a major vector or even a minor vector for this virus there's very good data on this from many countries now this is not how it spreads and it's really bad for people's mental health especially with some are coming to deny them the chance to be outside and it's really especially bad for children and the second big thing we should do right now is figure out how to get schools reopened as quickly as possible because children and young adults are at very very low risk for this again the data is the same all over the world the number one risk factor in you know in becoming sick or dying from the coronaviruses age you know they're a couple they've been a couple of outlier cases where the media has focused on you know young children possibly dying in some of those cases it's not even clear whether coronavirus actually was a factor in the children's death but remember many children die every year from the flu you know depending on the year could be 50 or 100 or 150 we don't shut society down for that we don't shut schools down for that it is unfair to children to deny them the chance at schooling and in some cases it's worse than that many children in the United States 2,000 children in the United States approximately a year die from child abuse which is you know far more than will ever die from Corona and antis to lock these kids up in their homes where teachers who might be the only responsible adults in their lives can't see them where they're stuck with abusive or drug using parents all day and and where their whole family's under additional stress right now because that you know parents might be jobless they're worried about paying the rent is just beyond wrong so we really should be looking at focus at reopening schools Alex we're running out of time I got one more question for you and it draws from your earlier book your book from last year called tell your children which is about marijuana and mental health and violence and it was a subject of one of the most read and debated issues of in Primus which is the newsletter of Hillsdale College ever ever published you wrote that book last year and and when when when we talk about coronavirus of course we're talking about physical health sometimes we don't always talk about the mental health implications of this disease and of the lockdown the shutdown and so forth with these orders with staying at home with people out of work is there a greater threat now of drug abuse whether it's whether it's marijuana or opioids or something else III think there's some evidence of that at this point you know we I don't know that we have hard statistics on that and I'm reluctant to go past you know what the statistics say there's some evidence by the way of an increase in domestic violence already and again of child abuse in terms of drug use it's it's quite reasonable to assume that you know both prescription pharmaceuticals like benzodiazepines and you know and drugs of abuse like like cannabis or you know or alcohol or cocaine would be would be things that people would turn to right now to try to relieve their stress I'm not sure we have that data yet there's also considerable evidence that economic deprivation drives up drug use drives up suicides you know the opioid crisis is sometimes called the crisis a death of deaths of despair and it would be surprising if something similar didn't happen because of this you know perhaps because my wife is a psychiatrist so she's at you know she's a doctor a medical doctor but she's also focused on mental health I'm pretty I'm pretty attuned to this issue and it's something we really should be talking about much more Alex Berenson thanks so much for talking to us about expert opinion and models and their failures thanks very much Sean and thanks to all of you for watching this program from Hillsdale College

Channel: Hillsdale College

Views: 71,109

Keywords: hillsdale, politics, constitution, equality, liberty, freedom, free speech, lecture, learn, america

Id: mG2vdyfLv7U

Length: 50min 21sec (3021 seconds)

Published: Thu Apr 30 2020