The Daily Signal - What’s the Real Data? A Look at the Relevant Statistics of the COVID-19 Pandemic

Episode Date: May 21, 2020

Would you believe that over half of the counties in the U.S. have had no COVID-19 deaths? This, as it turns out, happens to be true. Drew Gonshorowski, a research fellow in the Center for Data Analysis at The Heritage Foundation, and Kevin Dayaratna, a research fellow at The Heritage Foundation, join The Daily Signal Podcast to talk about the data behind the coronavirus pandemic. We also cover these stories: President Trump said he might keep funding from Michigan and Nevada over mail-in voting.  Secretary of State Mike Pompeo says he made the recommendation to Trump to fire Inspector General Steve Linick.  Senate Republicans have issued a subpoena to a company affiliated with the Ukrainian company Burisma Holdings, which Hunter Biden, the son of former vice president Joe Biden, sat on the board of.  Hosted on Acast. See acast.com/privacy for more information. Learn more about your ad choices. Visit megaphone.fm/adchoices

This is the Daily Signal podcast for Thursday, May 21st. I'm Virginia Allen. And I'm Richard Deltudis. Factor fiction, over half of U.S. counties have had no COVID-19 deaths. This, as it turns out, happens to be fact. Drew Gonscherowski, a research fellow in the Center for Data Analysis at the Heritage Foundation, and Kevin Dioratna, a research fellow at the Heritage Foundation, join me today on the Daily Signal podcast to talk about the data behind the coronavirus pandemic. Don't forget, if you're enjoying this podcast, please be sure to leave a review or a five-star rating on Apple Podcasts and encourage others to subscribe. Now onto our top news.

President Trump said he might keep coronavirus aid funding from Michigan and Nevada because they are pursuing mail in voting. On Wednesday, Trump tweeted, breaking, Michigan sends absentee ballots to 7.7 million people ahead of primaries and the general election. This was done illegally and without authorization by a rogue secretary of state. I will ask to hold up funding to Michigan if they want to go down this voter fraud path. A spokesperson for Jocelyn Benson, Michigan Secretary of State, said in a statement via M-Live. President Donald Trump's statement is false. The Bureau of Elections is mailing absent voter applications, not ballots. Applications are mailed nearly every election cycle by both major parties and countless advocacy and nonpartisan or

organizations. Just like them, we have full authority to mail applications to ensure voters know they have the right to vote safely by mail. Trump subsequently deleted the incorrect tweet and posted a new correct tweet. Trump also call out Nevada for their mail in primary election tweeting, State of Nevada thinks they can send out illegal vote by mail ballots, creating a great voter fraud scenario for the state and the U.S. They can't. If they do, I think I can hold a fund to the state. Sorry, but you must not cheat in elections. At Russ, vote 45 at U.S. Treasury. Secretary of State Mike Pompeo says he made the recommendation to President Trump to fire Inspector General Steve Lennox. President Trump said earlier this week that he had lost confidence in the Obama

appointed Inspector General and thus was terminating him. During a press conference of the Department of State on Wednesday, Pompeo was asked to provide more specific information as to why Lennox was fired. Pompeo responded, per the hill. So there's been lots of discussion about that I've read a number of reports. Let me say three things. First, the president has the unilateral right to choose who he wants to be his inspector general at every agency in the federal government.

They are presidentially confirmed positions, and those persons, just like all of us, serve at the pleasure of the president of the United States. In this case, I recommend it to the president. That Steve Linnick be terminated. Frankly, should have done it some time ago. Representative Elliott Ingle, Chairman of the House Foreign Affairs Committee, noted earlier this week that Winnick was in the process of investigating a 2019 emergency declaration that allowed America to bypass congressional approval to sell $8 billion worth of weapons to Saudi Arabia and the United Arab Emirates. to aid in their ongoing fight with Yemen.

Senate Republicans have issued a subpoenaed to a company affiliated with the Ukrainian company Burisma Holdings, which Hunter Biden, the son of former Vice President Joe Biden, sat on the board of. The Homeland Security and Governmental Affairs Committee, which is chaired by Senate of Ron Johnson of Wisconsin, a Republican, voted to subpoena Blue Star Strategies, which is ties to the Ukrainian company Breazba Holdings. Hunter Biden resigned from the board of the president. Burisma Holdings in 2019. The subpoena, Per the Hill, is asking for records from January 1st, 2013 to the present of Blue Star Strategies related to work for or on behalf of Burisma

or individuals associated with Burisma. Johnson is also requesting an interview with Blue Star Strategies administration to review the subpoena. Proxy voting is allowed in the house for the next 45 days as of yesterday. Speaker of the House Nancy Pelosi of officially authorized proxy voting on Wednesday, which will allow one House member to vote on behalf of up to 10 other members of Congress. Proxy voting is being implemented as a way to keep representatives safe, who may feel uncomfortable returning to the chamber to cast votes during COVID-19. If a member chooses to vote through a proxy, he or she has to submit a letter with the names of those representatives who are allowed to vote for them and instructions for each vote. But House Minority Leader Kevin McCarthy warned during Wednesday's podcast interview with the Daily Signal that, quote, with a proxy vote instead of 435 representing districts across the nation that can be held accountable, 20 people control all of Congress because each member can call up to 10 proxies. So if Democrats have 20 people all holding 10 proxies, they can pass any bill.

The House members who serve as proxies are to state, quote, as the member designated by and then insert members' name, pursuant to House Resolution 965, I inform the House that, again, insert member's name, will vote yay, nay, or present per the hill. House Speaker Nancy Pelosi is once again calling out President Trump, saying he and those he works with have doggie do on their shoes. here is what Pelosi said Thursday, Mia da Hill. Yes. It's appropriate for the president to do that in the general. You're asking me about the appropriateness of the actions of this president of the United States so completely inappropriate in so many ways that it's almost a given. It's like a child that comes in with mud on their pants or something.

That's the way it is. They're outside playing. He comes in with it. He comes in with doggie-do on his shoes, and everybody who works with him has that on their shoes, too, for a very long time to come. Seattle wants Washington State to create a $100 million fund for undocumented workers who did not receive stimulus checks during COVID-19. On Monday, the Seattle City Council passed a resolution in a 9-0 vote that calls for the establishment of the fund but does not formally create it. Seattle Mayor Jenny Durkin plans to sign the resolution and said in a statement per the Seattle Times that, quote, looking out for the most vulnerable in our community is even more critical in times of crisis.

It is all the more important to ensure we are not pushing people further into the shadows. Heronhood applied for and was given $80 million in coronavirus recovery funding that they were not eligible to receive. Fox News reports. The Small Business Administration is overseeing the Paycheck Protection Program designed to shore up small businesses that have had to close or otherwise were affected by the coronavirus pandemic and allows businesses with fewer than 500 employees to be forgiven to the principle of a government loan if they don't lay off their employees.

The employees still have to pay the insurance. interest on the loans overseen by the Small Business Administration. The SBA is contacting all Planned Parenthood organizations that requested and received funds from the Paycheck Protection Program, letting them know that affiliates of larger organizations with more than 500 employees aren't eligible for PPP distributions. Per Fox News. The Department of Justice is standing up for religious freedom in California. Eric Dryband, head of the Justice Department's Civil Rights Division, sent a letter to Democratic California Governor Gavin Newsom on Tuesday, warning that his phase reopening plan was discriminatory towards religious gatherings.

California is in stage two of their phase reopening, which allows for restaurants and some businesses to open, but churches are currently not allowed to open until stage three. The letter referenced a former statement by Attorney General William Barr that, quote, government may not impose special restrictions on religious activity that do not also apply to similar non-religious activity. The letter, which has been posted on Twitter, says, simply put, there is no pandemic exception to the U.S. Constitution and its Bill of Rights. Now stay tuned for my conversation with Drew Gons Sharovsky and Kevin Dioratna on the COVID-19 data you're seeing on the news.

It's our priority at the Daily Signal to keep you informed during the coronavirus pandemic. Here's an important message from the White House Coronavirus Task Force. Hey, America. I'm United States Surgeon General Dr. Jerome Adams, and I just want to say thank you for following the president's coronavirus guidelines. Social distancing, mitigation, it's working. We know we're flattening the curve and saving lives, and it is all thanks to you. So keep at it and stay tuned for more updates from the coronavirus task force. I am joined today on the Daily Signal Podcast by Drew Gonscherowski,

a research fellow in the Center for Data Analysis at the Heritage Foundation, and Kevin Dioratna, a statistician, data scientist, and research fellow at the Heritage Foundation. Kevin and Drew, it's great to have you on the Daily Signal podcast. Thanks for having us, Rachel. Yeah, thanks for having us. So to start off with you, Kevin, you recently had a piece for the Daily Signal about the Imperial College COVID-19 model. Before we go into what was wrong with this model,

can you first off and explain to our listeners what this model was and how affected public policy? So the Imperial College model was an epidemiological model developed by people at a modeling team at Imperial College used to forecast the prevalence and mortality of COVID-19. They had been using in the past, actually, for other illnesses as well. They adapted for COVID-19.

And it was a very influential model, and it made it all the way to Downing Street in England and all the way to the White House over here, influencing policy decisions, including the recent lockdowns. Can we get into what was wrong with this model and why it was off? Well, firstly, so my colleague, Norbert Michelle, and I began on this type of research a couple of months ago,

given that there was all this modeling that was being used for COVID, and we looked at all sorts of statistical models in the Center for Data Analysis at Heritage. And so when I noticed that the government had been using or calling upon this model, I had specifically asked a number of members of the Imperial College team, including Dr. Ferguson himself, for their model to be able to play with it. And not surprisingly, I didn't get a response. A number of them just did not reply to my emails. I also asked a CDC for their model, and they did not either.

So then the question became, how did these people get their numbers? And not surprisingly, their model really had some serious issues with it, which came out in the subsequent weeks. Instead, in the meantime, we decided to take a publicly available epidemiological model, excuse me, and look at it ourselves, playing with the assumptions, testing robustness, and so forth. So yeah, you mentioned that the model had some serious issues. Could you detail maybe two or three of the biggest issues that had? Absolutely. So like I said, the model was not publicly available, so nobody could really see what was going on in this black box. There had been talk on the Internet, according to the journal Nature, that Microsoft had been collaborating with the Imperial College team to make these codes publicly available.

I heard about this in the middle of my own research using other models. and ultimately the model did become publicly available, but not the actual model that they used. It was some snippets of the model, some subset of the model. But the serious problems with this, which also carry over to the general model, are the first thing you got to check whenever you use one of these statistical models is, are the results reproducible? Can you actually get the same results they did? And in many cases, you at least can, but you would often check to see how the model defers under varying assumptions. But in this case, even if you make the same assumptions, you still get different answers, which is completely disturbing. And it's completely disturbing that this model actually made it all the way to policy.

So that is one issue. There are a number of other issues, including numerous bugs, poorly written code, and so forth. But the bottom line is this model fails the test, and not only fails a test, it fails a test miserably about reproducibility. Well, as you mentioned, Kevin, you and your colleague Robert Michel, I did decide to take a publicly available COVID-19 model and forecast the prevalence and mortality of the disease under variable plausible scenarios. What did you find in what you did in this situation? So it's a great question, Rachel. So Norbert and I looked at a number of assumptions.

Like any statistical model, these models are all grounded on assumptions. Specific assumptions we looked at are the mortality rates within the ICU. most of these statistical models assume that deaths associated with COVID occur in the hospital's ICU. So then the question is what is a mortality rate in the ICU? And that's also dependent on, firstly, some assumptions made by the user, but also realistically, when you think about it, who is actually being admitted to the hospital and are the elderly actually engaging in social distancing who are typically at most risk for serious illness associated with COVID? So we played with that assumption, varying levels of mortality within the ICU.

We also played with assumptions about the asymptomatic rate. As of now, in this country, Little is known about the asymptomatic prevalence of COVID-19. So we played with assumptions there. And we also played with assumptions about the basic reproductive number or the R-not value, which signifies basically how contagious the disease actually is. And like I said, we noticed that you can get vastly different projections of COVID-19, both in terms of spread and in terms of mortality, ranging from through August 1st, estimates of deaths, for example, as low as, say, 75,000, going all the way up to 1.1 million.

And so the latter assumption, those are definitely assumptions that people who want to cook the books will definitely try to emphasize by deliberately beefing up the model's assumptions to get those results. And yeah, I just think that in terms of these assumptions, very little is known about COVID. So it's very, very difficult. More and more, more than. information that's fortunately coming out, but it's very little as known about COVID. So it's important to update these models with the best assumptions and best information available. Well, in this, you also looked at the model's assumption about the virus's basic reproductive number and an average number of people who will get the virus from someone who is infected. Can you drill

down into what you found here specifically? Yeah, absolutely, Rachel. So we vary the basic reproductive number within what we believe were some very reasonable assumptions. The R&Ot value, as I mentioned in my blog, it was first popularized in the movie Contagion in the, in typical in movies, where it is the average number of people that a single infected individual will spread the disease onto. So we varied this basic reproductive number between one and a half and three, and we found that the deaths can range quite substantially. ranging from on the low end around 75,000 or so to all the way up to 1.1 million. And in terms of the spread of the disease, the spread of the disease in terms of the percent of the population infected

could range between, say, 15 percent or so of the population infected all the way up to well over 80 percent. So depending on the value of the Arnaut value that you're using. So with these results literally all across the map, these models, again, are clearly highly dependent on assumptions specified by the user. So, Kevin, given obviously the errors we discussed in the Imperial College model, as well as all of the information that's out there, the misinformation that's out there, do you think that the ordinary American is aware of how flawed these different models are and just generally when it comes to this, you know, this. staff we see on a dated basis. Do you think Americans are included into what really is going on? Well, we've tried to make that clear, Rachel, especially in my recent Daily Signal article that we had published on the Daily Signal website in the last couple of days where we expose the flaws with the Imperial College model. And it's important that we make the American public aware of

these things and that there's serious flaws in these models. More fundamentally, it's important that any model that lawmakers are using to make policy decisions be made publicly available. So the public can actually test the assumptions, play with the assumptions, and see how are these lawmakers actually getting these results? And how are they actually incorporating them into policy? Because if that is not the case, and for all intents and purposes, these policy makers can be using these black boxes to manipulate anything that they want. Well, Drew, you recently have a piece on the Dealey Slingland with a colleague hiding that 1% of all counties

representing 15% of the U.S. population are responsible for almost half of the country's COVID-19 cases and more than half of all the deaths. Can you unpack this for us? Right. So a lot of my work in the past couple months and also work with Norbert, Michelle, has been on effectively building out some basic information for readers and for the general public more broad. And so this sort of comes in two packages. It comes in tracking cases, which on the Heritage website, we have this great map that sort of gives you a bunch of basic information for your county about how many cases, the population density of your county, sort of basic information. So you can get an idea of how prevalent COVID is in your area. And one thing that we sort of discovered as we started working on this was that, as expected where people are more densely populated and more closely living,

it seems that the impact of this pandemic has occurred in sort of cities and largely or heavily populated areas. This does translate into some pretty surprising, not once you're familiar with it, but it does translate into some pretty surprising statistics as you've mentioned. Well, given your work on all this, you had noticed that, noted that 50% of all counties or 10% of the U.S. population have had zero COVID-19 deaths as in May 11th. Most people wouldn't know the stat looking at mainstream use coverage. Why would you say that this is the case? This is the case because in our country, people live,

very differently, more or less across the entire country. And it's interesting. I grew up on a farm. I grew up in a very rural area. And you can go sort of days on end prior to all of this without seeing a whole lot of people. You'll go to work. You'll come home. You might go to the store. But for a lot of people living this way, they were already socially distancing. And it sort of makes sense when you live in a small or rural county that, you know, some counties in South Dakota have, you know, 1,300 people living in them. And it would be sort of an expectation that you would not have a lot of cases in that kind of county. You also had highlighted in a piece with Norbert Michelle that there are some counties in the United States that have had zero COVID-19 deaths. Can you

talk a little bit about that and how there are some points? places that haven't been affected. Right. So similarly, there are some places that have had really low cases and also, as a result, have had very, very few deaths. And this sort of translates more into when we're thinking forward about reopening and these questions about how do we start thinking about policy and how do we treat these counties differently than a county.

that's been really hit hard by COVID. Kevin, you are looking at these numbers. You're doing number crunching. How should Americans, when they see these news reports, what should they keep in mind about how statistics are reported as they see these news reports on TV? What would you tell them, as someone who crunches numbers all the time? To keep in mind.

So it's a great question, Rachel. So specifically regarding models and model projections, Any model is based on assumptions. And the real question is what assumptions are being made and are the, firstly, are those stated up front? Because if they're not, then it's very, very disturbing that suggesting that maybe there's, that lawmakers are trying to hide something. But the other question is, okay, so even if the assumptions are stated, what if you tweak these assumptions in other reasonable ways, how would the results change? And how are these potentially deferring results being used to guide policy? For example, regarding the Imperial College model, they've predicted that over 2.2 million people would die here in the U.S. by October.

One thing that would have been, you know, important to ask many people did was, okay, what assumptions do they make? And under other reasonable assumptions, what would be the associated forecasts? Well, before we conclude here, Drew you mentioned earlier about Heritage Foundation and the Interactive COVID-19 tracker that Heritage developed. Can you talk a little bit about how that was put together and how people can utilize it? So this tracker effectively updates, it's about three times a week currently. And it takes data from USAFax.org, I believe. and populates by county. So one of the main trends that we tracked

is just this rate of new cases, daily rate of new cases, over the past 14 days. So it sort of gives you an idea if a county has, let's say, one new case today and two new cases tomorrow, that would be sort of an accelerating trend, very simply explained.

But the map is sort of trying to tell you about that situation. So you can get an idea about whether or not your cases are sort of staying steady or accelerating or decelerating. One interesting thing that's happening now with that is a lot of states are starting to get increased capacity in terms of testing. And so some places might look like they have accelerating cases, but it really is just there. They have more tests than ever in some areas. I know in Michigan, that's my home state that over the weekend.

and they had more tests in sort of those three or four days than they had had for a while. And that resulted in sort of flagging a couple counties that you would still expect are doing better, but you sit there and you look at it and you go, oh, okay, this is because of increased testing. And I think that's ultimately what people will be able to use these sorts of trackers for is it gives them sort of some basic information

pretty readily available, pretty easy to understand that gives context to this. because a lot of this comes down to stories at a very local level, whether or not there's an outbreak at a nursing home or the prison population is being hit very hard by this. And sort of those things aren't necessarily brought to the forefront in a lot of the discussion we're having when it comes to policy. Well, Kevin and Drew, thank you so much for breaking down these numbers for us and joining us on the Daily Signal podcast. Thanks for having us. Yeah, thanks for having us.

And that will do it for today's episode. Thanks for listening to The Daily Signal Podcast. We do appreciate your patience as we record remotely during these weeks. Please be sure to subscribe on Apple Podcasts, Google Play, or Spotify. And please think of us a review or rating on Apple Podcasts and give us your feedback. Stay healthy and we will be back with you all tomorrow. The Daily Signal Podcast is brought to you by more than half a million members of the Heritage Foundation. It is executive produced by Kate Shrinco and Rachel Del Judas.

Sound design by Lauren Evans, Thalia Rampersad, Mark Geine, and John Pop. For more information, visitdailysignal.com.