Skip to content

Many years ago, when he was a baby economist . . .

Jonathan Falk writes:

Many years ago, when I was a baby economist, a fight broke out in my firm between two economists. There was a question as to whether a particular change in the telecommunications laws had spurred productivity improvements or not. There a trend of x% per year in productivity improvements that had gone on for five years or so, and we were looking three years after the change, and the productivity improvements were still around 2% per year. Economist 1 said: “Nothing to see here. The productivity trend is unaffected.” Economist 2 said: “WTF? You think productivity changes come from nowhere? It’s the law change that allowed the last three years to happen.” The conversation got heated (WTF was the least of it) and Economist 2, and his whole group, were fired. It was really ugly.

I timidly approached the firm President and said: ‘Y’know, there is really no way to tell which one was right. There just isn’t enough data to tell, and so essentially it comes down to your priors about the causes of the trend.” He agreed, but said Economist 2 and his group had acted with disrespect toward Economist 1, and that’s why they were fired. He had no idea who was right.

The stuff you posted yesterday is a perfect example of that fight all over again. Looking at the estimated R panels by state, you could certainly not spot any actual impact of shelter-in-place orders. Everywhere R was falling before well before any stay-at-home mandate, often to 1 or below, as in NY, and the continued drops to 1 in places like MA don’t look any steeper than places before the orders. And just look at Florida and try and fit that in your model of governmental action.

At that point, all the commenters look at the same data and draw inferences based on nothing but their priors, as far as I can tell. What was the demonstration effect of Italy? Of Washington? Of NYC? Did schools closing before stay-at-home orders function as partial stay-at-home orders? Do people require stay-at-home orders at all? The answers to these questions will directly determine the effect of removing stay-at-home orders (really? you sure about that?) but even with multiple jurisdictions to help us tease out the effects, I suspect there just isn’t close to enough data to do so with any confidence. And that’s even assuming that the R-estimation model, which is absolutely necessary to move from uninterpretable disease and death curves to how-are-we-doing-against-the-disease curves, are perfect!

So sure, this is Stan multilevel modeling, and that’s great, but is there any way to tease out a stay-in-place order effect? And even if you could, do you have any confidence that the effect of relaxing it is symmetrically opposite? I love to answer questions with models, but this just looks too tough. Fortunately, everyone at the blog seems to be treating everyone else with respect, for some value of respect. So nobody has to get fired—yet.

I asked Falk if I could post this and he said, sure, but:

The 2% should be changed to x% to be parallel. (I think it was 2, but productivity changes are always around 2%.)

That’s news you can use!


  1. Adede says:

    I am a bit shocked by the part “Economist 2, and his whole group, were fired”. Well, I can understand firing Economist 2 for unprofessional conduct. But firing everyone who worked (for? with?) him seems a bit far. Unless, they were also flinging obscenities, it seems gratuitous to fire everyone who has the misfortune to be associated (possibly not by choice) with a bad actor.

    • Jonathan (another one) says:

      While neither confirming or denying that I am that particular Jonathan, it might have been, hypothetically speaking, a complicated situation in which a consulting firm had been purchased en masse and the fit between the two cultures was not perfect and this particular straw broke a particular camel’s back… hypothetically.

      • DickTheRapper says:

        Amazing! The two Jonathans describe an event similar to one I witnessed. Before my success in popular music, I worked in an economic consulting company.

        We had a dispute between two groups of economists about the effects of a change in communications policy. The debate was over the extent to which the savings that had accrued to the long-distance companies from the reductions in charges to them by the local operating companies for the origination and termination of long-distance calls had been passed on to consumers.

        One group—the one with local companies as clients—maintained that the long-distance companies had not reduced their prices and that therefore the savings could not have been passed on to consumers. The other group—the one with a long-distance company as a client—maintained that the savings had been passed on.

        The first group pointed to the BLS CPI for phone service, which had not gone down much. The second group pointed to (1) flaws in the BLS CPI for phone service, (2) the fact that our firm’s own long-distance bills had gone down far more than the CPI indicated, and (3) evidence that the long-distance company profits were steady or falling. (This was a few year before a number of the long-distance companies engaged in deceptive deals that fraudulently boosted their accounting profits. For those who are counting, Bernie Ebbers passed away about two months ago—shortly after his release from prison.)

        The problem with the BLS statistics was simple. The carriers did not cut prices. They brought out new calling plans that were cheaper than the old ones. But the BLS figured that since these plans were cheaper than the old plans, which remained available, they were lower quality not price cuts. So the index wasn’t adjusted. Hausman describes many instances of similar flaws in the CPI. Search “Hausman BLS CPI,” and you will find several references. If customers changed plans, they got the lower prices.

        Arguably, the second group was intense in their criticism of the first group. I think part of the reason for that was that they thought that their position was so clearly correct (as subsequent events have proven), that the fact that management thought that there should be a “debate” over this issue was troubling.

        As in the cases mentioned by the Jonathans, the second group got fired. I’ve lost touch with everybody. Last I heard both groups were doing well and living happily ever after.

        Dick the Rapper

  2. jim says:

    Jonathan Falk: #1!!!!

    Exactly. People should be able to agree that there’s no data analysis approach to tease this out – not even close – and move on to more productive problems.

    Same with the education thing two days ago: the problem of whether or not a *general* college education has long-term wealth value is intractable; and it’s not even clear that it matters from a practical point of view, since we know there are career choices that have higher value.

  3. Tom Passin says:

    There are two features in the daily case counts that I think we can be fairly sure what the cause is. I refer to the US data as published by Johns Hopkins. There are two peaks towards the end of 2020. It helps to look at a smoothed curve overlaid over the spiky daily data.

    The first rise starts several days after Thanksgiving and the second starts several days after Christmas. The beginning of the change in the case counts is quite sudden, although to see this it helps to have looked at the data a great deal over time. I claim these peaks reflect the widespread travel and family congregations that have been so reported in the news media.

    To have a big change in the US data requires either a large change in one or two states with enough population and cases to affect the US totals, or smaller synchronized changes that are widespread across the country. Thanksgiving and Christmas qualify as the latter.

    The most straightforward way to understand how this worked is to suppose that the people who traveled and spent the holidays in close proximity to family and friends in effect increased R0 for the duration, and then after they got back home they went back to their pre-holiday habits. Thus R0 would have gone back to its previous value, and the daily case counts would have relaxed back to their previous trends over some period of time, presumably of the order of a few weeks.

    This is exactly the behavior we see in the data.

    Other features in the US data are less clearcut, and that’s in part because the actual information we want is the fractional rate of change of the daily data (which closely reflects the current R0 value) – which is extremely noisy – while the daily counts are an integrated version of that. Integration always blurs features so they are hard to distinguish.

    Only the two holiday features are definite and clear cut enough that (I believe) they can be ascribed to known events. To suppose that the effects of local changes such as lockdowns can be distinguished seems to me to be very doubtful because of the combined effects of the noisy day and insufficient numbers of cases. Only nation-wide trend changes have a hope of being reliably ascribed to specific causes, with the possible exception of large natural disasters.

    • When I looked at this in January, my impression was that all the kinks around Txg and Xmas were short term and due to reporting delays. Far fewer people are in the office filling out paperwork on thanksgiving, and the friday after, or on christmas eve or christmas day. If you look at a 7 day average it had a large decline starting right on the holiday for a few days, followed by a large rise that overshoots. this indicates reporting to me rather than any “real” effect.

      Here’s confirmed cases country wide:

      Deaths shows a similar structure, but deaths shouldn’t show such a structure so immediately after the holidays if it were from true infections at the holidays. hence, it’s reporting:

      • Tom Passin says:

        Yes, it’s true that one would expect to see data glitches at those times. Looking at the Johns Hopkins data, I can see that but it seems clear to me that there is more than just catch-up going on. For example, the daily count dives abruptly a day or two before Christmas then acts erratic for a few more days. However, the total area of these erratic spikes is small, much smaller than the area of the consistent increases of the following (roughly) ten days.

        • There was an enormous wave that started around Sept 10 or so, peaked around Jan 5 or so, and died out until about March 10, when the next wave started.

          During that time, we had two major multi-day holiday periods: Nov 24-Dec 2 and Dec 24 to Jan 4 or so. If you just run a smooth curve through all that data it works out just fine that lack of reporting during the holiday periods is made up for by overreporting shortly after. You can see this by comparing the detailed daily data and the biweekly data on the data explorer:

          Our World USA

  4. xyu says:

    reproductive number R is a bad indicator to monitor the epidemic. R ignores the overdispersion of contact structure which leads to superspreading events. R is also retrospective and time varying, and when most R is around 1, very insensitive to whatever changes in the environment.

    Basic R0 is a good indicator (and key characteristics of epidemic) to assess the potential of epidemic when nothing is done, but when something is done and the contact network is heterogeneous (and is likely reduced significant after intervention), there is nothing you can infer from small variations of R. So I think time varying, effective R is meaningless for monitoring epidemic. Recall after April 15,2020, most states have R below or around 1, and many states lifted mandates, and quickly, as we all know, a second wave came, and a third, and now a fourth.

    The actual number of new cases and the incidence (percent of new case/population) are the best indicators for policy decision. In traditional epidemiology, we plot the epidemic curve (i.e., the number of new cases over time) and summarize incidence (or cumulative incidence). This is what CDC original focused on, and they are correct in this regard. when there are new cases around, and the percent of people having been infected is low (or naturally immunized, a phrase invented in time of COVID), the epidemic will come back, and the resurge is predictable, no need math or R here.

    I think most people got carried away by R, a largely theoretical concept that has little use in practice. There are several papers suggesting R is not useful in a heterogeneous network. Even R0 is useless as all government will do something to control the epidemic.

    In April, 2020, the Wuhan city initiated a massive project to PCR test everybody in Wuhan (9 millions then), and they found 300 asymptomatic cases. This may seem ridiculous and extremely cost ineffective, but it worked. It cleared up all left over cases and there was no new cases in Wuhan afterwards. In this case, the government did not make such draconian decision based on R or whatever fancy indicators. Just plain old concept: “the number of new cases.”

    • Steve says:

      + 1 It still baffles me why in the U.S. the public has never been given the correct option. The issue should not be lockdown versus no lockdown, but test and quaratine versus let the virus spread. Lockdowns may have been necessary initially, but many parts of Asian/Australia have shown that contact tracing and quaratining the sick and exposed is effective and far less disruptive.

      • Clyde Schechter says:

        Absolutely right!

      • Jonathan (another one) says:

        I completely agree, but I believe the answer to that question is entirely political. Neither party has any stomach for *actual* quarantine, as opposed to voluntary-nobody’s-really-checking-and-my-family-can-hang-around quarantine. I heard a great podcast very early on with Lyman Stone who said that the best tool to stop pandemics is quarantine camps, and the laws enabling mandatory sequestered quarantine and the funding for those camps was actually already available. But he predicted (quite rightly) that Republicans would oppose them while Trump was downplaying the virus and Democrats would oppose them because (at the time) it was overwhelmingly urban Democrats who would be quarantined.

        Whenever I hear people say that South Korea did so much better than the US because of masks, or testing, or whatever, I remind them that they have mandatory sequestered quarantine. The Wuhan example is even better,

        • Steve says:

          You (Jonathan, definitely not the one Andrew is quoting), are quite right, but it is interesting to look at Australia and their success in implementing quarantines. I have lived there, and Australian politics and culture is not so different than ours. It include Murdoch’s propaganda machine and plenty of right wing don’t trust government or science types. Yet, they were able to implement real quarantines. I have followed it a little bit, but I am baffled by the difference. Was it just Trump or is there some other problem in our culture that made doing what was done in Asia/Australia/New Zealand so hard for us.

          • Michael Lew says:

            Differences like these?
            1. Single payer socialised health care.
            2. Payments to unemployed workers increased during the pandemic. (Falling back today, too soon, too bad.) Federal governmental subsidies for employees of companies affected by the lockdowns.
            3. Even the conservative federal government is mostly able to accept reality. (Unfortunately, there are a few Trumpians, and a variety of other nutters.)
            4. A relatively reliable and well trusted national broadcaster that much of the population accesses (to the disgust of Rupert!).
            5. State governments who are not enthralled by the false prophets of capitalism and Republicanism.
            6. Australia is one island and New Zealand is two.

          • Phil says:

            I don’t know the culture of anyplace else well enough to comment comparatively, but I can comment on the U.S. I don’t think our culture would allow it, not even close. Too many people across the political spectrum would object on civil liberties grounds, an issue closely related to distrust of government.

            I disagree with Dzhaughn, below, who says we wouldn’t be able to do it even if we wanted to — I don’t think the logistics would be substantially worse than in other countries — but without the will to do it, it wouldn’t happen.

            • Steve says:

              Probably true. But, when I came back from travelling abroad, I was under quarantine, and NYC knew it because I can back through a port of entry. I got a quarantine message every day from the NYC sheriff (yes NYC has a sheriff), threatening me if I broke quarantine. No one every visited me to actually make sure, but with testing that system could be used to quarantine. It would not be as good as in AU, but it would have been more effective and less disruptive than what actually did. And, if you said, “we are lifting the lockdowns and openning up because this quarantine system will work better” that could have been sold to Americans in April of last year. It baffles me that no one tried.

            • Dzhaughn says:

              My wild estimate was partly based on cultural considerations, so it is conflated with the question of what we are “willing” to do. Supposing the next pandemic disease kills 1 in 10 healthy people instead of 1 in 100 with a heavy skew toward the frail, sure, we will close the airports and the roads. But, that is roughly 10-100 times worse, so we are willing to accept 10-100 times the cost.

              But it is also just about complexity. Take NZ. It has a population about that of Oregon. Unlike Oregon, it has no land border, let alone open borders and a shared city with another state. It does not have to largely coordinate with a federal government in which it has about 10 about 600 legislative voices, and even less executive or judicial input.

              So I’m not amazed at what NZ can do as far as internal governance. And Oregon is so very much more manageable than, say, Ohio!

              • confused says:


                The US is rather unique in a lot of ways. There’s no other country of comparable development/wealth that is nearly so large (in population) or so internally complex.

                Also, we have a rather uniquely strong individualist streak, I think because the other Anglosphere countries e.g. Canada, Australia, NZ became independent much later and so never had a “Manifest Destiny” era as the US did.

                Also, the Cold War kind of set us up as a direct contrast to the Soviet Union; I think a lot of the externally-irrational quirks of US politics (like strong capitalism, small-government rhetoric, and evangelical Christianity being on the same political “side”) come from this (because the Soviet Union was communist, totalitarian, and officially atheist).

    • jrkrideau says:

      n April, 2020, the Wuhan city initiated a massive project to PCR test everybody in Wuhan (9 millions then), and they found 300 asymptomatic cases. This may seem ridiculous and extremely cost ineffective

      This strikes me as extremely cost effective. Wuhan, at last report, is functioning normally, the health services are not overloaded, and so on. A failure to do such testing and take other measures at the start of a pandemic is the extremely cost ineffective move. Compare New Zealand or Taiwan to the USA or France.

      • Dzhaughn says:

        You have to include the cost of the infrastructure of controlling the movement of population in your calculation.

        The cost of such control for New Zealand and Taiwan (for diseases largely contained outside those nations) is low. For China, the cost is hard to estimate because it was largely already paid; there was just an incremental bump for COVID.

        For the USA, the cost of introducing such controls would be extreme, I would wildly estimate at 10-100 times the total losses from COVID. It would take at least a generation to implement, even supposing there was popular acquiescence to the general idea.

        • Anonymous says:

          For China, the cost is hard to estimate because it was largely already paid; there was just an incremental bump for COVID.

          I am sorry, I just do not understand this.

          • elin says:

            They already had high levels of surveillance.

            • confused says:

              Yeah, the US just could not have done what was done there (or many other places). The time needed to set it up would have been far longer than the time available.

              (I also believe that that level of surveillance/control would not be acceptable even if it meant 0 COVID deaths, or even if COVID had been far worse than it actually is, but that’s a different argument.)

    • Julien says:

      Very good points. Re values that are thrown here and there (often without uncertainty) are basically just the slope of the epidemic curve, basically just removing the information about the amplitude of the incidence to only keep the slope. It’s actually *less* informative than just looking at the data! And it suffers from the same limitations, most of all differential under-ascertainment in time and across social groups.

    • Tom Passin says:

      Yes, indeed. It’s almost impossible to back out R from the data once anything starts to change. So it’s mostly useful to understand what kinds of things can happen given various conditions. For example, if lots of people start eating indoors in restaurants, there is going to be more transmission, and assuming or deriving an R value could be one way of expressing that.

  5. Tom Passin says:

    I wrote “… the combined effects of the noisy day”. Sorry, that was supposed to be “…noisy data”.

  6. Steve says:

    To the extent that Falk is analogizing the Ioannides controversy to a debate between economists over the effect of some regulation on the market, I couldn’t disagree more. There is a moral dimension in the Ioannides controversy. If Ioannides had said, “I have a different estimate of the fatality rate, which is much lower than others, but their is a lot of uncertainty and my estimate shouldn’t affect policy” that would have been fine. But, he was clearly advocating policy, and the choice here was assymetric. If we followed him, and he was wrong more would die and there would be more disruption to the economy. (Basically, that is what happened.) If we ignored his advice took the virus seriously and he was right, there would have been a temporary disruption to the economy, which could have been restarted quickly. In that situation, people in public health have a duty to advice caution. He did not. This is not a situation where there are costs and benefits on both sides. This was and is a situation in which there is exponential risk on one side and additive risks on the other. His actions from a leader in medicine were totally irresponsible.

    • Jonathan (another one) says:

      While neither confirming or denying that I am that particular Jonathan, I believe the reference is not to yesterday, but to a post in May. If I were that particular Jonathan, I would steer clear of opining about Dr. Ioannides’ analysis.

  7. parkslope says:

    So ‘insufficient data’ on many known & unknown variables regarding the SARS-2 pandemic.

    That seems obvious now from an objective stance.
    But the emotional subjective viewpoint has long overwhelmingly dominated the general pandemic analysis and severe societal defense actions.

    Your economists anecdote illustrates the point of cautious objectivity versus subjectivity even among well educated professionals.
    The economist #2 group was silenced (fired) for “disrespecting” the more powerful #1 group. Power wins the endgame.

    Similar suppression of dissenting views is highly evident from the corporate media & government regarding their mainstream narrative of lockdowns, masks, testing, vaccines, etc, etc.

    • Andrew says:


      Yes, one of the problems of the mainstream narrative of lockdowns, masks, testing, vaccines, etc, etc., is that all these topics end up getting bundled in a way that can make no sense. We discussed that in another comment thread. I can understand how people can be anti-lockdown, but if you’re anti-lockdown I think you should be really really pro-mask: the point of lockdown is to reduce face-to-face interactions, so if you don’t have anything like a lockdown then makes it even more important to have masks when people are near each other indoors. But, with this bundling of attitudes and politicization of discussion, you get people who are anti-lockdown without being pro-mask, which doesn’t make sense at all. Or, I should say, it doesn’t make any sense from a public health standpoint but I guess it makes sense from some sort of political standpoint.

      • Well one of the participants in my informal group of infectious disease epidemiologists that vent once every two weeks, does argue lockdowns are not very effective when looked at over about year and includes peripheral health harms. They also insist that masks as worn by the public do not do much good. Exactly how to sort out their political views is not that clear. There does not seem to enough high quality data to “prove” them mistaken.

        • Rahul says:

          So I keep reading this objection about the utility of masks.

          Why is this so hard to settle via studies, any idea?

          Is it really true that we don’t have good data to support the use of masks or another of those science denier sort of memes.

          • Joshua says:

            People look at studies that don’t conclusively “prove” that masks work 100%, and conclude that “masks don’t work.”

            They also look at areas where spread has continued despite mask-wearing and conclude that means that “masks don’t work.” or they compare areas where there are mask mandates to areas where they don’t exist, and compare infection rates and reverse engineer to conclude the efficacy of mask-wearing without considering confounding variables.

            Probabilistic reasoning is hard, and understanding the compounding effect of individual marginal risk across a population is hard, and ideological “motivations” make it that much harder.

            • Martha (Smith) says:

              “Probabilistic reasoning is hard, and understanding the compounding effect of individual marginal risk across a population is hard, and ideological “motivations” make it that much harder.”

          • Mark S says:

            I found Scott Alexander’s post about masks from early on in the pandemic useful for covering a range of reasons why the mask/no-mask question is less straightforward than it seems (or seemed). This is before one gets to the effects of ideology. (Obligatory note that pointing to this post should not be interpreted as evidence that I think said effects are negligible or that it is incorrect to highlight them in current circumstances.)

            I apologize for being unable to figure out how to embed the URL using the comments interface. “Face Masks: Much More Than You Wanted to Know”,

          • Dzhaughn says:

            There is a big gap between proving “masks work” to “mask mandates are a cost-effective means of controlling a epidemic.”

            • confused says:

              Especially as I am not sure mandates have that much effect on “what people actually do” (especially in parts of the US where mandates will not in practice be enforced), and only the latter matters.

              Here in TX, businesses are largely continuing to require masks despite the state order going away, and basically everyone I know is continuing to operate at the same level of caution (ranging from zero, to extremely careful) as before.

              It is entirely consistent to believe both that masks *work* and that mask mandates *don’t*.

              (Not that I am saying I necessarily believe this — but it seems entirely possible.)

        • Steve says:

          Was there a “lock down” in this country that I was unaware of? Do you mean that time that New York and the Northeast sort of shut down and other parts of the country kept open?

          Have your “infectious disease epidemiologists” do an experiment. Get them all in a pool, then order people in the shallow end to pee
          and the epidemiologists in the deep end to stop peeing. Wait awhile and measure both ends of the pool for urine concentration. Then, you can publish a study titled “Stopping People from Peeing in a Pool has No Effect on Urine Concentration.”

          Effects from mask wearing, social distancing, “lock downs” are not linear. Just as the virus spreads exponentially, we should expect all of the safety measures to greatly increase in effectiveness as we approach full compliance. Spoiler alert, that didn’t happen.

      • Joshua says:

        > But, with this bundling of attitudes and politicization of discussion, you get people who are anti-lockdown without being pro-mask, which doesn’t make sense at all.

        A lot of the “anti-lockdown” crowd think that “masks don’t work.” Some even think that they’re harmful. And as I’m sure you know, many think that wearing masks is an infringement on basic freedoms.

        So in that way it makes sense.

        • Kyle C says:

          “Think” is doing a lot of work there. It’s true they “insist” those things.

          • Joshua says:

            I don’t know how you’d get in between what they say they think and what they actually believe.

            • Kyle C says:

              I would inquire into the reasoning process. Having done this, my experience is you get “lol” or memes or insults because there is no basis of the statement, it’s a gesture.

              • confused says:

                Well, I think a lot of it is “emotional” rather than “reasoned”.

                But I don’t think it’s just “a gesture” — people wouldn’t take the risk just for a gesture if they really believed in the risk.

                I think it’s some combination of “not showing fear”/projecting an appearance of normalcy and not trusting the CDC etc./”showing that I don’t just follow whatever they say to do”.

                (And I think before a year ago, there was significant question at least in the Anglosphere — though apparently not in East Asia — whether *non-medical-grade* masks did anything useful. So to some people it seems like they ‘changed their story’.)

              • Joshua says:

                My experience is somewhat different. I get some of that, but I also get some “The Danish study shows masks don’t work.” or “The CDC originally said masks don’t work.” or, “People can still get infected of they wear a mask,” or “They don’t wear masks in Florida or other states doing better than New York.”

                Granted, these people aren’t likely representative. But I think there’s more of a mix than you suggest.

              • Joshua says:

                And yeah, there’s also lot of “Don’t be so afraid.” and “I can decide for myself and I’m not putting anyone else at risk

                They’ve been siloed in an echo chamber of misinformation

              • confused says:

                Well, “The CDC originally said masks don’t work” is I think part of what I’m talking about – if it’s interpreted less as “the data says they don’t work” and more as “the CDC changed their story, therefore there’s no reason to believe what they say now”.

                And I don’t think the two are really contradictory, either. If one has a prior idea that government advice on this sort of thing is questionable at best, one is much more likely to seize on any seemingly contradictory data (and think that its not being mentioned by the CDC etc. is further confirmation of their untrustworthiness).

                I don’t think that it’s all *that* unusual in the US to see public-health as annoying “nanny state” types.

                >>echo chamber of misinformation

                That’s certainly true… but IMO it’s not entirely a “cause” rather than an “effect”. In the case of COVID specifically, sure, those aspects of the media environment were pre-existing; but as for the social setup that led to its being so divisive…

                I feel like much of this arose to “fill a void”, with regards to the huge cultural divide between the rural/central US and the urban Northeast and West Coast (where most of the mainstream media is headquartered).

                That divide was easy to *exploit* – but I don’t believe that just changing the media environment would fix things (the “void” would still exist and be filled by something else).

              • confused says:

                Also, in April-May-early June it seemed very strange that the states that weren’t taking it as seriously were so lightly affected, especially after media claims that e.g. Texas and Florida would be hard hit “2 weeks after Spring Break”.

                I think seasonality was *way* underestimated as a factor up until the fall. There was so much of an attempt to say that it wouldn’t “just go away in summer” (which was indeed true) that the messaging felt like “it’s not seasonal” — which was *very* hard to reconcile with the fact that nearly all the seriously-hit places were cold-climate.

                By the time places like Texas and Florida started to see high death rates in July, the political divide on COVID was already well established.

  8. Yuling says:

    I once chatted with a selling side “analyst” about their quarterly-released “predictions”: the earning of this company would increase by 23.56-35.45% next year, and accordingly the share price should be $45.66-56.13. Out of curiosity, I asked what models they were using for such seemingly unlikely inferences and I recalled the answer was flawlessly computed by some particular function in excel.

Leave a Reply

Where can you find the best CBD products? CBD gummies made with vegan ingredients and CBD oils that are lab tested and 100% organic? Click here.