Naive realism and the pandemic

Ross Douthat writes,

it probably makes more sense to compare the U.S. death toll to similarly positioned and sized countries — meaning the biggest countries in Western Europe and our major neighbors in the Americas — than to compare us to a global average. And when you compare deaths as a share of population within that group of peer countries, the U.S. starts to look more mediocre and less uniquely catastrophic.

He is responding to the views of his newspaper’s writers and readers that the United States suffered many more COVID deaths because Donald Trump was President. I think that this view is very widespread and very wrong.

A lot of research suggests that non-pharmaceutical interventions made little or no difference in cross-regional and cross-country comparisons. Statistical comparisons aside, tell me what policies the President could have put in place that would have made a large difference. Show your work, keeping in mind how many deaths seemed to stem from New York subways and nursing homes.

Jeffrey Friedman introduced me to the term naive realism, which is an important concept with a misleading name. I would explain naive realism as follows.

A first-order naive realist believes that he knows enough to solve a problem if he were in charge.
A second-order naive realist admits that he does not know the solution, but he is sure that someone could solve the problem if that person were put in charge.

It seems to me that there are a lot of naive realists about the pandemic. To them, I would say the following:

1. Even now, there are huge gaps in our knowledge. It is not clear that after more than 6 months we know what the optimal policy should be or should have been.

2. Among experts, the most touted solution is massive testing along with “track and trace.” But in February or March, when this might have made a large difference, it was logistically and politically impossible. It was logistically impossible for many reasons, including the fact that the FDA was in the process of disapproving any test not issued by the CDC, and the CDC was in the process of issuing faulty tests. It was politically impossible because in early March, politicians on the left were positioning themselves against taking the virus seriously, arguing that to do so was anti-Asian and racist.

3. Health experts were against masks.

I am not saying that President Trump said the right things or gave good advice. But to me, the view that we would have had a different outcome under a different President seems difficult to support.

28 thoughts on “Naive realism and the pandemic

  1. “Statistical comparisons aside, tell me what policies the President could have put in place that would have made a large difference.”

    In the unlikely event that you are able to pass this challenge, please re-run your analysis, but this time without the benefit of hindsight.

    Hint: what was the *consensus* among the experts and politicians back in Jan/Feb/Mar?

    • Indeed, for the past 4 years, U.S. politicians have disagreed on Trump –the scapegoat for the Dems losing power in 2016. In February 2020, the disagreement had already led to high political tensions (remember the failed impeachment). Few “experts” expressed opinions before March 8, but they were not looking for a consensus. More importantly, the WHO empowered to track all epidemics and pandemics had not warned about COVID19. Then, for a few days after March 16, the U.S. press and social media found a focal point that would change the course of government interventions but it was a grotesque focal point: the famous Imperial College study. It took only 2 weeks to destroy the focal point, but many national governments had already recognized Covid-19 as a serious pandemic that required strong government responses (I live in Chile and I still complain about what the government did on March 16-20 –btw, the President, a Harvard Ph.D. Econ, was my colleague and friend before he changed to politics in 1990).

      In open societies, nobody should expect consensus on most issues. There was no consensus by March 8, and there has been no consensus since March 16 on the virus, the disease, and the pandemic in the U.S. and many other countries. Forget about WHO and focus on the U.S. public health bureaucracy. Albeit too early to pass judgment, the general impression is that most agencies at all levels of government failed to advance reasonable strategies at their level and contributed nothing to put together a national strategy. Why have they failed? When are we going to hear about what Dr. Fauci and his team have been doing for the past 35 years?

  2. It seems clear, to me at least, that culture, quality of a country’s bureaucracy, and local government decisions have each thus far been more important than the decisions of the head of government in explaining outcome. Does East Asia really have much much better prime ministers and presidents than Europe? I think you’re clearly right and anyone who think about it for a minute should realize it doesn’t make sense to put the majority of blame on the president, but it’s too politically appetizing to pass up.

    The actual argument by the NYT writer (I forget his name) Douthat is responding to was that we should attribute all US deviation from the global deaths per population to the president. That seems like reasoning almost no one would seriously entertain except in politics.

    • You are right. We often assume that one main responsibility of government is to protect citizens from natural disasters, but we cannot ignore that the structure of government differs greatly across countries. In the U.S. this structure involves divisions of government both vertically –implying the preservation of some autonomy by state and local governments– and horizontally to check –implying the separation of power at the federal and state levels. Comparison with other countries is quite complicated due to the several dimensions that define alternative structures of government (also, de facto powers of each government unit may differ from de jure powers, even in the U.S.). Some people think that in “the old days” was much easier for a President to coordinate all government units to protect from natural disasters but I don’t remember any disaster in the past 50 years which has not prompted conflicts between the President and some other units.

      Nobel Prize Winner Angus Deaton wrote a column blaming the current structure of government for the horrible response to Covid (but first he declared that “A malevolent, incompetent Trump administration bears much of the blame for America’s failure to control COVID-19.”; see https://www.project-syndicate.org/commentary/us-connecticut-compromise-1987-and-failed-covid-response-by-angus-deaton-2020-07?barrier=accesspaylog ).

      Finally, you are right about the column written by David Leonhardt (a submissive NYT columnist that claims some knowledge of economics) which prompted the Douthat’s column –one that is pathetic because he pretends to go beyond Leonhardt by introducing comparisons with other countries as if he were comparing two tomatoes.

  3. I agree that people tend to overly concentrate blame on authority figures, particularly those who are of the opposite party. That said, things that would have helped in the US include:

    1. Consistent messaging and coalition-building on the importance of social distancing and masking so that a) guidelines were followed more widely, b) they did not become partisan footballs
    2. A more complete and thoroughgoing travel ban, along with quarentining people returning from ex-US
    3. Once the CDC test blew up, we were in a hole. But I do not think the executive branch had a “this is priority 1, give me updates every 3 hours” response to the testing crisis. Would this have helped, and how much? Hard to know.

    All these actions might have been resisted. But in many cases, they were not attempted.

    Also, mobilizing support for policies is a political competency. Imagine a CEO who always explains failures with this excuse: “I have these great ideas, but the board vetos them.” Stipulate that it’s true — his ideas are great, and the board is wrong. This is still a failing CEO, because it’s the CEO’s job to persuade the board and build political support for his actions. The same applies to political leadership.

    • It’s difficult to have a serious discussion of what the Federal government could have done differently. Although there are many books on policy-making and regulation, you are not going to find anyone that reviews alternative types of regulation systematically. To simplify we can say that there three types: information, conditional authorization, prohibition. Since all regulations tried to prevent the bad effects of specific actions and omissions, it’s not surprising that often governments resort to all three types. Information is the less intrusive form of intervention and if properly executed it could alter people’s behavior at a reasonable cost. Prohibition is the most intrusive and the effects of someone’s action or omission on other people have to quite bad to justify a costly, effective enforcement unit. Conditional authorization can be seen as a way to enhance information (by increasing the cost of an action or omission) or as one to mitigate the high enforcement cost of prohibition.

      So to design a policy you better start from one of the two extremes. Let us say that Republicans start from information and Democrats from prohibition, it’s not surprising that we have observed long processes of designing and approving regulations (often accompanied by weak mechanisms of execution and control). Covid is quite different from all previous detonators of regulation because it was a shock, a sudden natural disaster, about which nothing or little was known. To make things worse, in the U.S. the structure of government imposes high costs of coordination and cooperation among the many units involved, much worse in times of high political tensions. So before finding the best intervention we should be concerned about the selection process and the political context.

      We can joke at Trump’s suggestion of taking medicine X (it’s just information to prevent infection or death) and at Biden’s suggestion to make masks obligatory (it means the prohibition of being outdoors without a mask). But we shouldn’t be surprised that none is taking seriously because they never will become official policy.

      Finally, the prize for the best meme goes to the mask: “Seeing how some people use their masks, I now understand how contraceptives failed.”

    • Consistent messaging and coalition-building on the importance of social distancing and masking so that …

      Which was out the window the moment public health officials told ordinary people that not only didn’t they need to wear masks, but that they shouldn’t and that masks might even be dangerous. Which those same public health officials knew was BS, a ‘noble’ lie intended to preserve masks for frontline workers — which were in short supply because of other government screwups. And those same officials lacked the not-very-great insight that people could make their own masks or use various other things (neck gaiters) as alternatives. Once the officials blew the trust of the public in that idiotic decision to lie to the public (and endanger them in doing so!), there was little hope of getting it back.

      • No, the public health officials were not lying about what they said with respect to masks at first. There was no noble lie – they were stating what health scientists had concluded well before Covid-19, that masks are not helpful in public and can give a false sense of security.

  4. American culture has a strong bias toward freedom and diversity; people are accustomed to doing what they like with relatively little regard for the desires of authorities or for social norms (as compared to societies like China, Japan, or Singapore). The upside of this is that some of those people will do something profitable; the American entrepreneurial tradition is considerable because our economic system can leverage the successes and shut down the failures. But a pandemic exploits the downsides of our system; the virus can leverage diversity of hygiene practices the way the economy leverages diversity of business practices.

    • Jay, American culture is based on freedom and diversity. Americans raised on that culture are accustomed to doing what they like with strong support for the desires of authorities that comply with all social norms (as much as people expect everybody to comply with them). In the traditional cultures that you mention as an alternative to the American culture, people have been raised to be submissive to the privileged authorities that are not constrained by people’s social norms.

      Today the American culture is threatened by barbarians that reject it. Unfortunately, their plans are supported by D-politicians that fear losing again the presidential election. The government response to the pandemic has been very much conditioned by D-politicians’ struggle to regain power (this is why the response has not been coordinated by the many agencies at all levels of government and there has been little cooperation between the Federal and D-state governments). Indeed, people’s response to the pandemic has also been different across political jurisdictions, although there are strong similarities that may be explained by the American culture.

  5. I think he could have gotten on the mask train far earlier–I certainly did. I don’t think any politician would have though. I largely agree a Clinton presidency would have suffered from the same structural problems to this crisis.

    To use one of your ideas, Trump’s basic problem in this crisis was one of prestige. Anything he said or backed would be polarized almost immediately. Trump could have appointed someone with more prestige to manage the crisis–perhaps Bill Gates?

  6. I entirely agree. If your readers need to look at some of the evidence that non-pharmaceutical interventions made little or no difference, I cover some of it in a blog post – “The Mystery of the Lockdowns” – here: https://naimisha_forest.silvrback.com/the-mystery-of-the-lockdowns

    “In ‘How many lives would a more normal president have saved?’ New York Times columnist Ross Douthat wonders how many Americans died because of President Trump’s “abnormal” reluctance to embrace stricter lockdown measures. Douthat’s speculations never get close to the likely correct answer, though, which is zero: a ‘more normal’ president would have saved zero extra lives.

    That’s because recent studies find little evidence that differences in lockdown policies have had any effect on the pandemic.

    It’s a mystery why governments were and remain eager to plunge along a path with such meager benefits to offset daunting costs. The costs include not only the worst world recession in modern times but also the emergency powers states have seized to enforce the lockdowns and the resulting loss of civil liberties by citizens.

    The citizens’ loss has been the state’s gain. A new study finds that governments have declared emergencies mainly because of the new powers they stand to gain, not the severity of the pandemic.”

  7. One disadvantage that the U.S. has is an extremely high obesity rate. People who are obese or significantly overweight appear to have much worse outcomes with COVID-19.

    I’m sure that there are numerous other demographic differences that also could explain differences in outcomes.

  8. Here is word for word what I posted as an answer to the virus on Quora in early April.

    I would make no significant changes. We should keep business activity going where possible, isolate the at risk, and introduce substantially improved safety protocols on our regular activity such as work, exercise and shopping. If we had done something like this it would have saved tens of thousands of elderly and had reduced economic destruction.

    Start Quora answer:

    I believe we will come to regret our indiscriminate “lockdown” approach.

    I believe that a better approach is to offer near absolute lockdowns for the at-risk population, combined with reasonable, albeit extreme, “distancing” tactics for the young and healthy.

    The lockdown portion of the preferred strategy would offer near total isolation from the virus for the very old, and for those with serious pre-existing conditions such as diabetes, heart problems, immune deficiencies, extreme obesity and asthma. We could do this by creating quarantine zones in their houses, or by offering complementary shelters in hotels or government housing. Sanitized food and sundries would be delivered by the state or by loved ones to these at risk individuals, protecting them from exposure to the virus.

    The lockdown/quarantine would be completely voluntary but highly encouraged, as virtually everyone dying from the Coronavirus (studies indicate the number is somewhere between 95 and 99%) have at least one serious pre-existing condition, and many have two, three or more. By allowing those most at risk to have state-assisted lock down, the hospitals would be relieved of much of the influx of critical care which has so taxed their abilities.

    We must protect those most at risk!

    The rest of the population would then go into extreme social distancing protocol. Those with symptoms would be expected to stay home, or go to the hospital for those rare cases which turn more serious. Masks would be mandatory outside of the home (unless exercising alone). Meetings would be strongly discouraged, and any large public gatherings or crowds would be expressly forbidden (church services, movies, concerts, dining halls, parties, etc). All mass transit will be expressly prohibited and closed.

    Retail establishments will enforce social distancing of customers, with mandatory hand sanitizing on entrance and exit along with automated doors. All employees will wear protective covering, and preferably separated with glass barriers at checkout, which will be cash-less with no exchange of “dirty” money. Ubers and taxis would be required to be cleaned and sanitized after every trip, with masks and preferably plastic shields between the driver and the passengers.

    Businesses would be required to test employees regularly as soon as tests are available. Anyone who could work at home remotely would be expected to do so. But businesses would still be open, and economic activity would not be suppressed except in the most dangerous situations (indoor restaurant, theaters, mass transit).

    Schools should stay closed until the governors believe the time is appropriate to cautiously re-open.

    Thus we would have a mixed strategy of near perfect isolation of those most at risk, combined with continued safe economic activity by the younger and healthier. Some of those working would still be exposed, though at a slow rate, allowing herd immunity to slowly build up in the greater population while protecting those likely to die.

    The governments should then focus their attention on incentivizing and fast-tracking cures and vaccines; facilitating tests, measurement, and statistical analysis; and supporting the manufacture and distribution of protective gear and hospital facilities. They should also enforce the above guidelines on social distancing and support and fund the voluntary quarantining of those most at risk.

    What the government should not do is molest people who are working or trying to work. They should also stop molesting people who are trying to exercise or walk outside in non-crowd conditions, such as the lifeguards who fined a lone paddle boarder who was exercising off the coast of Malibu and posed zero risk to anyone.

    I believe this mixed strategy would “flatten the curve,” reduce mortality rates, slowly and more safely build herd resistance, and minimize the economic catastrophe of our current bulldozer approach.

    No strategy is perfect, including the above. The point though is that the above strategy would be substantially better than what we are doing today, which is destroying global prosperity in the hope of delaying the spread of the virus until such a time as a vaccine is created.

    End April Quora answer.

    • “We must protect those most at risk!”

      No.

      In fact we shouldn’t “protect” anyone.

      Most high risk people got the message within the first month or two and learned how to protect themselves. They didn’t need assistance.

      The entire rest of this sounds like a hellish nightmare.

      The only justifiable case for lockdowns was to “flatten the curve” to preserve hospital space, and because aspects of virus transmission were unknown early on (were surfaces dangerous, who knew, turns out no). This was achieve very early on mostly through voluntary actions of individuals (and nursing homes eventually getting their acts together).

      Doing what you describe because of a slightly elevated death rate for the already old and sick is probably one of the worst moral tradeoff I’ve seen in my lifetime. It’s certainly the worst that I’ve been affected by on a day to day basis.

      The answer to the virus, at least by May when we had a reasonable idea of what it was simply “let them die.”

      https://www.youtube.com/watch?v=sfr2F8F1xms

      • When I wrote this they were still sending infected people to nursing homes and forcing “non-essential” businesses to close. My suggestion was the opposite. If you read my answer I was clearly against lock-downs. Did you read it?

      • @Rojellio very reasonable and probably far superior to what we got instead.

        @asdf I could be wrong and I’m too lazy to meander through the old comment threads, but I think you had much larger fear of the virus back in Mar/Apr/May despite some pretty good evidence to the contrary. Let me know if I’m misremembering. Also, “let them die” sounds fun and provocative, but I’m guessing that basically no one would support this.

        • @Rojello

          I don’t think schools should be closed. Parents should be able to assess the risk themselves, not self interested governors.

          It’s telling that private schools managed to re-open, school closures are primarily about people wanting to get paid without providing a service in return. If teachers or their unions refused to provide the service, their members should not have been paid the the funds remitted to parents for them to figure out how to care for their children. If we could only get at the money, maybe this blanket refusal to do their jobs would finally end public education. As Bryan Caplan says, “Not Even Babysitting!”

          I find the idea of extreme quarantine for people simply because they are old or have a medical condition hellish. From what I’ve witnessed of it, it doesn’t seem like any way to live. I’d take the virus risk, with that death is only a possibility, rather than a kind of living death that is a certainty.

          I’m against mandatory mask usage. I find it forced on people in nonsensical situations. There is no reason for people to be wearing masks in nearly any outdoor environment, and if you are in an outdoor environment where you feel the need to wear a mask you probably shouldn’t be there in the first place.

          Our family has been spending our summer outdoors going to parks, farms, outdoor dining, and gardens. We have seen some of our closest friends and family members from time to time. We even went to a modest Fourth of July party in someones back yard. While none of these events were overly crowded/packed in, we definitely did not maintain six feet of distance all of the time nor would that have been remotely reasonable thing to do. We haven’t worn masks whenever we are outside, nor during some of our briefer indoor private encounters. We’ve only ever used them in indoor public places that we can’t avoid (like the grocery store, or when closing on our house). We don’t make our three year old wear a mask, even though technically she’s old enough for all the mandates.

          We haven’t gotten sick Covid or otherwise. Nobody we know living a similar lifestyle has. Nobody we know living a riskier lifestyle then we have has gotten it. From what I can tell of the statistics the people getting Covid do not share our profile at all, despite our doing things on your list we aren’t supposed to.

          It’s pretty clear from the data that infection rates cap out at like 2X% of the population in whatever region, because it really isn’t that infectious if you avoid a few really high risk environments and activities (and most people in that 2X% can’t avoid those environments and activities because its their job or they are just single young people who don’t care). The death rate is wicked low at this point and something like the herd immunity is at a lower threshold than we thought.

          Time to stop telling kids to wear masks at the playground, for god’s sake. Let grandma hug her kids if they haven’t don’t anything high risk in the last week.

          @Hans Gruber

          Yes, I was more worried when it started. For one, I was really worried about surface transmission in the beginning. I was getting told that my Amazon packages might kill me. How could I send my kid to the playground if touching the same monkey bars meant death? Turns out surfaces shouldn’t have been a worry.

          I also didn’t know outdoors was dramatically safer or that you need a prolonged interaction with one or multiple infected people to get infected (passing by someone in the park isn’t really a risk).

          I also didn’t know that around 2X% of the population being infected would be enough to basically stop the virus growth in its tracks. Remember all those “predictions” back in March about infection rates.

          In March all I really knew is that infection growth rates seems exponential and the only country to deal with it at that point was China and they went total martial law lockdown.

          “but I’m guessing that basically no one would support this.”

          You may be right about public attitudes, but I don’t consider that a good thing. We shouldn’t be imposing this on society because a few tens of thousands old and sick people might die slightly later. If we are that kind of society that would make that choice, maybe we deserve to die.

  9. It is a heroic act of partisanship to see that the overall outcome couldn’t have been better and the buck doesn’t really stop at the top.

  10. Keeping in mind the deaths in New York nursing homes alone, a different president would have needed the media on side to take on Andrew Cuomo and reverse his order. Already that’s hard to imagine.

    In what scenario does the media not spend April and May writing love letters to Andrew Cuomo? In what alternative universe does the New York Times hold a politician on its own side to account?

    Assume that an alternative president was shouting from the rooftops that Cuomo’s nursing home policy was a disaster. It still wouldn’t matter if the media defended their man.

    There’s a xenophobic, hysterical, racist, authoritarian travel ban at New Zealand’s borders, but the media love that country’s Labour Party PM. So the media has its love affairs and that’s going to stay the same in every timeline.

    • Indeed, the great and the good were encouraging mass rioting while they were welding playgrounds shut.

      What’s the death rate in all those “irresponsible” Red States again?

      https://www.statista.com/statistics/1109011/coronavirus-covid19-death-rates-us-by-state/

      Even within states, how exactly is the fact that my governor won’t allow the slides at my local pool to be open impact the fact that the illegal Central Americans living in cramped apartments with too many people about a twenty minute ride away have slightly higher infection rates?

  11. I think the #1 policy that could have made a difference until basically early March was closing borders to all non-resident person, and quarantining everyone coming into the country, no matter from where. You’d have to keep it that way basically until now, but then you’d have very few cases and even fewer deaths.

    This is a policy that is entirely under the President’s control here in Brazil, but unsure about the US. Other than that, in hindsight, it would be better to actually let it rip and own it (which no one was ever interested in).

Comments are closed.