Trump rallies are killers.

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3722299

Careful regression analysis of the impact of 18 Trump rallies by Stanford University researchers concludes that approximately 30,000 cases of Covid-19 and approximately 700 deaths were the direct result of the large mask-free gatherings. These are cases and deaths that would not have occurred absent these rallies.

The tragic thing is that anyone who believes the very basic science could have – and should have – seen this tragic outcome as very likely. Sponsoring such gatherings amounts to criminal negligence if not manslaughter. Trump’s dismissal of this KNOWN science is why we are now going through the roof on new Covid-19 cases and are locked into avoidable deaths of a thousand or more per day for the foreseeable future.

40 thoughts on “Trump rallies are killers.

  1. At first glance, this not yet reviewed paper is in conflict with contact tracing which has not linked Trump rallies to more than a few cases.

    I am not expert in regression analysis, but I will pass this along to some folks who are. One glaring problem I can see is that they used the 10 weeks prior to the rallies as their placebo arm. In a period in which cases are rising already, comparing 10 weeks prior to 10 weeks after will guarantee a correlation.

    Of course, rushing this paper to the popular press, and it is cited A LOT, before critical analysis by other statisticians is suspect in itself.

    Like

  2. The methods of these Stanford researchers are bona fide statistical methods. But let’s say, for the sake of argument, that they are wrong by an order of magnitude. That leaves only 70 corpses in the wake of Trump’s science-denying rallies. Better?

    By the way, you seem not to have understood the description of what you call the “placebo arm.” It is NOT the period before the rallies, but similar counties matched by 24 criteria where no rally was held. One of those criteria was counties which were on similar Covid trajectories as the Trump counties before the rally. Thus, there was no built in bias toward increases that you claim.

    Liked by 1 person

    1. I heard back from Steve McIntyre. He hasn’t gone over the study yet but his initial suggestion is to remain skeptical as Europe is currently experiencing a greater resurgence along with larger localized outbreaks than in the US and obviously they are not connected to Trump rallies.

      Time of year might be the larger factor.

      Like

      1. I have no idea who Steve McIntyre is or why I should care what he thinks. However, if he is claiming that this is an artifact of the time of the year, then I immediately doubt he knows what he is talking about. It is the same time of the year for counties with Trump rallies and counties without. By the same token this finding has NOTHING to do with the fact that the virus is raging out of control here or in Europe. It is a finding based on DIFFERENCES between counties that were as alike as possible except for hosting a Trump rally. As bad as non-Trump counties may be, Trump rally counties are worse.

        Liked by 1 person

        1. Steve McIntyre is the guy who, along with Ross McKitrick, brought down Michael Mann’s Hockey Stick. He is a professional statistician.

          The comparison the study uses compares only to selected counties not all counties, and those Trump Rally counties are far from the worst.

          COVID districts

          If you are going to assert that correlation and causation are linked, then clearly voting for Democrats is many times more dangerous than going to Trump rallies. (No, I DO NOT make that assertion, it is only correlation)

          Like

          1. “If you are going to assert that correlation and causation are linked . . .”

            That is a nonsensical argument. It shows that as often as you make this argument you do not really understand it. Events and their effects ARE correlated. This is a study of events and the results that they CAUSED by KNOWN causal mechanisms.

            Liked by 1 person

          2. Nope.

            Correlation can involve causation, but it can just as easily be coincidence and even more likely, a common cause.

            If there is a correlation between A & B, it could mean A causes B, or B causes A, or that both A & B are caused by C, or it could mean nothing at all.

            Like

          3. “Correlation can involve causation, but it can just as easily be coincidence and even more likely, a common cause.”

            “Just as easily?” As I suspected you have no idea what you are talking about.

            If the CAUSE is known and its KNOWN effects are found, that correlation between the cause and the effects is NOT coincidence.

            In this study the CAUSE is large gatherings of maskless people chanting, cheering and shouting. The KNOWN effects of this CAUSE is the spread of the virus to other people. There is no valid reason to consider the increase in the spread of the virus found following these rallies to be a coincidence. Especially since the researchers were careful to exclude other differences that might explain what they found. Your denial is just another example of how you must not accept science or even common sense if you are going to support Trump’s “sociopath” (your word) behavior.

            Liked by 1 person

          4. Then where is the contact tracing? It should be readily available if there is any. Yet you don’t hear of any.

            Certainly risky behavior is an opportunity for infection, but you don’t know if that behavior occurred at the rally.

            First, there would have to be sick people at the rally. That would show up in contact tracing. So, where is it?

            Then you would have to exclude other opportunities for infection.

            In short, this is nothing more than correlation.

            In particular, look for the ‘C’

            Like

          5. Yes, where is the contact tracing? Just another of the many ways the Trump government has utterly failed to control the virus. There are very few places outside of New York where there is a serious – and expensive – effort at contact tracing.

            https://www.nationalgeographic.com/science/2020/09/contact-tracing-crisis-magnitude-hot-mess-america-fixes-coronavirus-cvd/#close

            So, really, the lack of data from contact tracing proves almost nothing. Besides that, many people who transmit the disease to others are not ill themselves.

            The finding of fact is that in communities that are as similar as possible those with a Trump rally have more disease than those that didn’t. It is possible that the methods used to find that fact are flawed but it is not apparent what that flaw might be.

            What is a possible separate cause of increased disease in communities where there were Trump rallies if it is not the rallies? Or what common cause would trigger both a Trump rally and an increase in disease in order to create this supposedly causeless correlation?

            Liked by 1 person

          6. Contact tracing is the responsibility of State Depts of Health, though Trump did offer to reimburse the costs. Some states did better than others.

            Like

          7. “Contact tracing is the responsibility of State Depts of Health”

            Yeah, I know. Trump punted on every aspect of controlling the virus. He believed that politically that was his best option. Pass the buck and point his finger at blue state Democrats. Too bad for him that he was so stupid as to believe that the virus preferred to spread in blue states only.

            Of course, the point here is that the lack of contact tracing data in an environment where little of that is done is not a reason to believe that the science of transmission was wrong when it came to Trump events.

            Liked by 1 person

          8. It wasn’t science.

            And where in the Constitution would you find authority for the President to bypass State Depts of Health? Obeying the Constitution is not “punting”

            Liked by 1 person

          9. That’s not how the Constitution works.

            The Federal govt is not empowered to do anything it chooses unless prohibited. It is prohibited from exercising those powers not enumerated.

            The States have to power to question people about exposure to disease, the Feds don’t

            Like

          10. “The States have to power to question people about exposure to disease, the Feds don’t”

            That is baloney. For example . . . https://www.cdc.gov/quarantine/aboutlawsregulationsquarantineisolation.html

            As someone who says he is willing to allow the “sociopath” Trump to run this country in exchange for his protecting the Constitution (Ha!), I would think you would have gone to the trouble to educate yourself on what it does and doesn’t do.

            But leaving aside such Constitutional “questions” there is nothing that prevented the federal government from taking the lead to encourage, organize and finance the NEEDED state and local efforts. That was not done because the significant money needed would have flowed to “failing Democrat controlled states and cities.”

            Liked by 1 person

          11. OK. where is the governor who can honestly say that he asked for COVID related help and didn’t gt it?

            Not bailouts for their past malfeasance, but help with COVID?

            Like

          12. Give me a break. The pandemic is a NATIONAL crisis. Our only hope for controlling the threat was/is a NATIONAL response with leadership by the President. Trump did NOT provide it. If you want to believe his punting of his responsibility was out of some kind of devotion to Constitutional authorities, knock yourself out. It is a laughable belief given Trump’s egregious disregard of those authorities in other areas – e.g., The Wall – but believe whatever floats your boat.

            As for the way you framed your question (“bailouts for their past malfeasance”), let me ask you – have you stopped beating your wife?

            Liked by 1 person

        2. Are you still pretending to not understand the study? Or are you really that confused? The study does not say that the Trump counties are the worst hotspots. It says that they are worse than very similar counties matched according to 24 different criteria. Duh!

          The authors are professional academics so it seem highly unlikely that they would deliberately publish a dishonest study or one with manifest errors. Why fight so hard against science? You would not go to a gathering of “sociopaths” (your word) refusing to wear masks. This study confirms how right you are to have that attitude.

          Your link to the fact that the death toll is concentrated in Democratic districts is not too nauseating but it is symptomatic of everything wrong with Trumpism. Biden intends to be the President of all Americans, not just those who voted for him. Yes, the virus death toll is highest in cities (so far, though a few rural upper mid-west states are catching up) and cities do tend to be represented in Congress by Democrats. Do you have a point to make or is this a rhetorical tongue sticking out?

          Liked by 1 person

  3. RE: “These are cases and deaths that would not have occurred absent these rallies.”

    Nope. They are cases and deaths that correlate with the rallies during the time interval studied. In the real world they might have occurred anyway, especially over a longer period of time.

    The purpose of the paper is to demonstrate a prediction model, not an epidemiological result (as implied by the post here). Whether or not the model is any good remains to be seen.

    Like

    1. As usual you have no idea what you are talking about. A word salad of meaninglessness.

      There is no “prediction model” in evidence. This is an after-the-fact analytical comparison between counties where Trump rallies were held versus very similar counties where they were not. Maybe the increased deaths in rally counties was just a coincidence but that is highly improbable when there is a causal explanation in plain sight that is FAR more probable.

      Science makes a prediction – when large numbers of maskless people from an infected population spend hours in close proximity with a lot of cheering, chanting and shouting conditions are ripe for the spread of the virus and new cases will be the result. This study confirms that prediction. In the “real world.”

      Liked by 1 person

      1. RE: “There is no ‘prediction model’ in evidence.”

        See page 3 of the paper.

        RE: “This study confirms that prediction.”

        Nope. It confirms a correlation, demonstrating an approach to disease forecasting, not a causation.

        Like

        1. On page 3 they authors make a clear distinction between what they are doing and what epidemiological prediction models do.

          If the demonstrated correlation between Trump rallies and measurable increases in sickness and death is NOT caused by those rallies, what alternative explanation can you offer? It could be pure randomness, of course, but the probability is vanishingly small.

          But, why are you quibbling? Do you not believe in science? Do you really think that large crowds of maskless, chanting, cheering and shouting people drawn from an infected population is safe? The finding is EXACTLY what the causal connection between pestilent exhalations and the spread of the disease would produce.

          Liked by 1 person

        2. RE: “Do you not believe in science?”

          I believe in science and I care about how scientific concepts are portrayed. In this case your statement, “These are cases and deaths that would not have occurred absent these rallies” is flat wrong. Your mistake is to confuse a correlation with a causation.

          I don’t doubt that the rallies had causitive effects, but there is no empirical data to support the claim you make. The paper itself doesn’t demonstrate the point you seem determined, falsely, to make.

          Like

          1. “Your mistake is to confuse a correlation with a causation.”
            “I don’t doubt that the rallies had causitive effects.”

            You don’t see the problem with these two statements I suppose?

            The whole study is based on “empirical data” which support the claims made – an extra 30,000 cases and extra 700 deaths resulting from Trump’s selfish and deadly decisions to organize these dangerous rallies.

            If the rallies are NOT the cause of the extra cases and deaths observed in the data, what would you propose? Remember that the comparisons are between counties that are alike based on 24 criteria.

            Liked by 1 person

          2. RE: “You don’t see the problem with these two statements I suppose?”

            No, I don’t.

            RE: “If the rallies are NOT the cause of the extra cases and deaths observed in the data, what would you propose?”

            I propose that the correlation is suggestive, but the cause is unknown.

            Like

          3. The cause is unknown?

            Uh, no, the study has found the cause – Trump rallies. If they were not the cause then there would not be statistically significant increases in those locations. Coincidence might make sense if the study was restricted to one or two communities, but with eighteen, coincidence becomes vanishingly unlikely.

            Liked by 1 person

  4. Sometimes it takes a “slap upside the head” for some people to actually wake up to the fact that there is a pandemic.

    Sad to say, those 700 deceased did not have to sacrifice their lives to go to a cult-like rally. Even sadder is that some of the victims didn’t go, but got infected by the attendees.

    Our esteemed super spreader in chief probably just killed off hundreds of his most ardent supporters and infected tens of thousands more. And what is so puzzling is that these rallies are attended by the already committed. Politicizing the virus, like when he hassled Ingraham for wearing a “politically correct” mask the other day, is literally costing him votes. Unless of course, these folks were going to vote twice as recommended by the man himself. Then at least they have one in the ballot box while being lowered into the ground in another container.

    His concern for Americans is on par with a wild mushroom. Toadstool variety apparently.

    Liked by 2 people

    1. “Sad to say, those 700 deceased did not have to sacrifice their lives to go to a cult-like rally. ”

      Has contact tracing linked any of those cases to rallies?

      Consider that people who take such risks(and I wouldn’t, I haven’t been to a TLP meeting since March) are likely also taking risks elsewhere, such as going to Home Depot or a sports bar, or cchurch, or other gathering. Assuming that the rally was the actual point of contact is nothing more than an assumption.

      The sample is simply too small to make such an assumption based on statistics alone. Consider how many went to those rallies and did not get sick.

      Contact tracing is the only way to assign a source to such low probability events.

      Like

      1. Interesting and revealing perhaps.

        “Consider that people who take such risks(and I wouldn’t, I haven’t been to a TLP meeting since March) are likely also taking risks elsewhere, such as going to Home Depot or a sports bar, or cchurch, or other gathering.”

        That seems to take the position that Trump fans who attend rallies are also spreading the virus more than non-Trump folks because they are ignoring simple guidelines for one reason or another.

        Either way #LIBERATE has not done Americans any favors. Quite the contrary, it’s killing them.

        Liked by 2 people

        1. Or it could be that they have jobs that require them to go out in the public.

          It’s always attractive to think you know the answers because that gives you the false sense that if you do everything right you are safe, but we can only reduce risk, we can’t be safe until we are immune.

          Like

    2. RE: “Sad to say, those 700 deceased did not have to sacrifice their lives to go to a cult-like rally.”

      The paper is quite specific on this point. The 700 deaths include people who did not attend any rallies. That’s a key reason to be cautious in assuming that the rallies caused the deaths.

      Like

      1. You are quite right. Many of the 700 dead – in fact, most – did not attend a Trump rally. They were simply unfortunate enough to encounter people who did or the victims of the people who did.

        You understand, I hope, that the fact that so many who did not attend the rallies fell victim to the criminal negligence makes it even worse. Truly innocent bystanders unlike those who chose to ignore science.

        Liked by 1 person

      2. Attending the rallies was just a way of spreading the virus to those unfortunate souls who did not but were infected by those who did.

        On the other hand, we have this when the PA governor restricted his rallies according to state and CDC guidelines:

        Recounting how his campaign had to move the site of the rally to comply with Democratic Gov. Tom Wolf’s Covid-19 protocols, Trump said this:

        “I’ll remember it, Tom. I’m gonna remember it, Tom. ‘Hello, Mr. President, this is Governor Wolf, I need help, I need help.’ You know what? These people are bad.”

        (Quotes to show Trump words.)

        Now that is really sad.

        “I want my rally, waa, waa…” as he throws his binky across the platform.

        (My quotes before I get reprimanded)

        The question is interesting: Is a quarter million dead a win? Evidently. Perhaps a half million would really get a Nobel Prize in Medicine.

        Liked by 2 people

  5. “The sample is simply too small to make such an assumption based on statistics alone. Consider how many went to those rallies and did not get sick.

    Contact tracing is the only way to assign a source to such low probability events.”

    Your “critique” of this methodology is based on nothing but wishful thinking. The number who did not get sick has absolutely nothing to do with this analysis. What is “small” about the sample size? A lot of people live in 18 counties.

    Liked by 1 person

    1. RE: “A lot of people live in 18 counties.”

      How many? Let’s assume 100,000 residents per county for a total of 1.8 million residents. A statistical prediction of 700 deaths attributable to a single cause is almost meaningless.

      Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s