A Smaller Point About the Russian Vaccine

People are claiming that the incentive for each country to get a propaganda win by developing its vaccine is bad thing and that Russia is acting badly by introducing a vaccine that has not yet been adequately tested. The only argument I can see for that is the idea that if this vaccine flops that will make people less willing to take the next one, but unless the flop kills a lot of people that seems unlikely, especially since the fact that they are bringing it out early is being widely publicized.

The other, and I think more important, side of the argument is that if the vaccine flops that means Russia has wasted some money and if it has serious side effects, they will effects on Russians. If it succeeds, the rest of the world benefits, with Russia having provided a very large scale stage 3 trial at their expense.

The Russian Vaccine

Nobody appears to have any evidence that the vaccine is either dangerous or ineffective. What people are complaining about is a lack of evidence in the other direction, both because the Russians have not been publishing information on their research and because it is being brought out prior to the usual stage three testing. My conclusion, assuming as seems likely that Putin has or could hire competent people to develop a vaccine, is that it will probably work and probably not have serious side effects. 

My basis for that is what I have read about the other vaccine projects. It seems clear that everyone else is being very careful, unwilling to bring out a vaccine until they are virtually certain of both effectiveness and safety. A vaccine that has passed the stage two testing is probably safe and probably effective, but probably is not sufficient. 

Insisting that you have to be certain a drug is safe is good rhetoric but bad science, since safety is not an option. To see whyy, consider the case of  Diethylstilbestrol (DES), a fertility drug that turned out to produce medical complications in the daughters of some of the women who had taken it. The complications took a minimum of ten to twelve years to appear and the most serious, a form of cancer, appeared, according to varying estimates, in between one in 250 and one in 10,000 of the daughters. In order to be sure a drug did not have such an effect it would be necessary to give it to something upwards of a hundred thousand pregnant women then wait fifteen years or so and do extensive analysis of the medical problems of their children. Applying that standard to protecting against every long-term risk one could think of would make the production of new medical drugs very nearly impossible.

The current global death rate from Covid is about six thousand a day, so each month by which the introduction of a virus is delayed costs about 180,000 lives. That, plus other costs associated with the pandemic, should be set against the risks of introducing a vaccine earlier, when we can be less certain of its safety and efficacy. I do not know enough about the risks of vaccines that have passed the stage one and stage two tests, how serious and how likely they are, to judge whether the caution exhibited by everyone except the Russians would or would not pass a cost/benefit test, and I doubt that current policies are based on any such calculation. But, given how cautious everyone else is being, I doubt that the risk of serious side effects is as much as 25%. If I am correct, the odds are that Putin's decision will turn out to be correct ex post, whether or not it was correct ex ante.

Covid: The Invisible Elephant in the Room

Calculations of the requirement for herd immunity are based, as I understand it, on the implicit assumption that everyone is equally at risk, hence that the probability that one infected person will infect more than one other can be deduced from data on the early spread of the disease. That assumption is unlikely to be true, for both behavioral and biological reasons. Some people spend more time interacting at close range with those likely to be speaking loudly than others. And it seems likely that some can catch the disease more easily than others.

Suppose we drop that assumption. Suppose, for simplicity, that half the population consists of people vulnerable to the disease and half, for behavioral or biological reasons, invulnerable. Observing the early spread of the disease, we find that, on average, each infected person passes the disease on to two others. We conclude that we will only reach herd immunity when half the population have had the disease and become immune as a result.

But the relevant figure is not what fraction of the population has become immune but what fraction of the vulnerable population has. In my simple model, half the vulnerable population is only a quarter of the total population, so we reach herd immunity much earlier than the simple calculation implies. 

The real world distribution of vulnerability will, of course, be much more complicated than that, but the qualitative conclusion still holds. Over time, the people most vulnerable will be most likely to get the disease, so the average vulnerability of those who have not yet gotten it will decline, lowering the average number that each infected person passes the disease to. Hence herd immunity will come sooner than the simple calculation implies.

So far I have only considered differences in how easily individuals can get the disease. There will be similar differences, at least for behavioral reasons, to how easily individuals can transmit it. The two will tend to correlate — someone who spends a lot of time in loud conversation with lots of others will be more likely than average to get the disease and, if he gets it, more likely to pass it on. So, over time, the probability of transmission will fall as those most likely to transmit are selectively removed from the pool of potential transmitters.

All of this tells us that herd immunity will come earlier than the simple calculation implies, but not how much earlier — that depends on the actual distribution of vulnerability, of probability of transmission, and the correlation between the two.

Has anyone done the research necessary to estimate these numbers and recalculated the requirement for herd immunity according?

A further problem with the simple calculation is that it ignores behavioral changes due to the pandemic itself. Presumably part of the point of lockdowns was to temporarily push the transmission rate below one, driving the virus to near extinction — far enough down so that it could be controlled by a test and trace approach. In most countries that imposed a lockdown that didn't happen, but even without a lockdown the existence of the pandemic changes behavior in ways that should reduce the transmission rate below its initial value, at least somewhat.

P.S. a commenter on another blog where I raised the post gave two links to relevant material. The first goes to the editors' blog of Science magazine. The most interesting bit is:

we were concerned that forces that want to downplay the severity of the pandemic as well as the need for social distancing would seize on the results to suggest that the situation was less urgent. We decided that the benefit of providing the model to the scientific community was worthwhile.

That implies that the editors believe that part of their job is filtering the scientific literature in order to bias the public perception in the direction they approve of, although in this case they decided not to. It follows that one cannot take the published scientific literature on any controversial issue as giving an unbiased picture of the actual science. That is disturbing, but not surprising.

The post contains a link to the paper, which appears to be simply a fancier version of my argument, without actual empirical data that could be used to figure out the size of the effect.

The second link goes to a paper which, judged by the abstract, does make an attempt at estimating real world numbers, and concludes that particularly hard hit areas, such as New York City, may already be close to herd immunity. I couldn't find the full text online.

But a commenter could.

Thoughts on Dating Apps

I have been thinking about the  problem that computer dating is designed to solve. I can see four different approaches, three that exist in both realspace and online, one that could exist online but I am not sure does.

1. Examine the characteristics of many people, looking for someone you believe would work as a partner.

Subscribers to OKCupid answer questions about themselves, if they wish many questions. They look at the answers of other subscribers to find someone who looks like a good fit. It is an approach that people routinely use in realspace, considering people they encounter of the appropriate age, sex, and marital status as possible partners — except in a society such as Saudi Arabia where contact between single men and women is very strictly limited, leaving it to a man's mother and sister to find him a suitable bride.

The great advantage of computer dating is that it lets you search a much larger number of potential partners than you would encounter in realspace. On the other hand, the information you are using is provided by the potential partner, designed in part to make him or her appear a good match, so may be less reliable than what you could learn by direct observation.

2. Search a set of people already selected for you to have characteristics that make success more likely.

OKCupid allows you to filter for people within a given geographical area, by age, gender and some other characteristics. Dating apps as a group carry this farther, specializing in Christians who want to marry Christians, Jews who want to marry Jews, individuals looking for casual sex, individuals interested in marriage and uninterested in casual sex, and  other categories of potential subscribers.

On the other hand, computer dating does not use the information implicit in who you associate with in real life, mostly people who have a good deal in common with you. A student at Harvard or Vasser is associating with other people who have passed the same filter, making a fit much more likely than with random pairing — this may be one of the major functions of institutions for higher education. Joining a political movement, becoming involved in a hobby, entering a profession, all put you in contact with a set of potential partners who share with you characteristics that might lead to a good fit.

3. Use an intermediary, someone looking at individuals and trying to match them up.

In the case of computer dating, the program can be the intermediary, showing you not a random collection of individuals but individuals that, in its judgement, fit what you are looking for and are looking for what you fit. The obvious way of doing this is based on the subscriber's account of his preferences for a partner but it also could be, perhaps sometimes is, based on someone else's theory of what characteristics go well together.

The realspace version in more traditional societies is the shadchen, the marriage broker. In a society such as modern-day America, it is done more informally. A colleague's wife suggests that there are a lot of nice girls at folk dancing — I go, despite not being a dancer, and meet the woman I have now been married to for almost forty years. I ask the wife of my married son, a perceptive woman who knows lots of people, if she can suggest anyone my single son or single daughter would get along with.

4. Use feedback on the results of attempted pairings to discover what characteristics predict successful ones.

The first approach depends on individuals knowing what sort of person will work for them as a partner. The third depends either on their knowing that or on someone else figuring it out for them. There might be a better way.

I am imagining a dating app one of whose terms is that, after your first date with someone met online, you provide the app with feedback, report whether things went well or badly, with further reports after any later dates and a final report after you either give up on each other or decide on a long-term relationship. The program uses that information to figure out what characteristics, defined by the information provided to them by each subscriber — not "is X," which the program does not know, but "says he is X" — predict a successful match. The fact that I had a successful or unsuccessful date with someone with a given set of characteristics not only predicts what other people I will have successful dates with, it  predicts the same thing for other people who share my characteristics. Thus, over time, the program could provide experimental evidence on successful pairing based on self-reported evidence, making possible a better version of the previous approach.

Have any dating apps actually done this? If so, with what results?

I originally came up with the idea in the context of matching readers with books, inspired by a talk I heard on multi-dimensional voting theory, an approach to matching the positions of voters with the positions of politicians in order to predict electoral outcomes.

Wimps, Boors, and Philosophers

I am mining my accumulated blog posts to convert them into one or more books and have started the first chapter. It is based on exchanges with a group of libertarian political philosophers who describe themselves as Bleeding Heart Libertarians. It occurred to me that the argument I was making there was related to a different argument in a different post.

My central claim in the first group of posts was that, while the BHL were offering legitimate criticisms of arguments made by their fellow libertarians, they were unwilling to apply a similar standard to arguments made by academic philosophers. They wanted to add "social justice" to libertarianism but were unable or unwilling to give a clear explanation of what it was or on what basis it could be defended. They spoke respectfully of John Rawls but were not prepared to actually defend his central argument, which, I had long concluded, hinged on a claim about as defensible as 2+2=5, even less defensible than the libertarian arguments they had attacked.

My conclusion was that what they were offering was a version of libertarianism designed to appeal to their non-libertarian colleagues.

My other post came out of the controversy over the Ron Paul newsletters, some of which contained articles attacked as racist. While I may have missed something, I do not think any of them either asserted innate inferiority of blacks or hatred of blacks qua blacks. What they did was express a derogatory opinion of particular blacks — Watts rioters or muggers — in a gleeful fashion. They were thus likely both to appeal to racists and to offend liberals — more generally, people who accept current conventions of acceptable and unacceptable speech. My guess is that both effects were intentional.

I see the clash as between people who see non-PC speech as a  virtue and those who see it as a fault, between people who approve of offending liberal sensibilities and those who share enough of those sensibilities to prefer to respect them. The former group see the latter as wimps, the latter see the former as boors.The wimps have friends on the left they respect, so prefer to think of those on the left as reasonable people who are mistaken. The boors are more likely to have friends on the right, including some, such as religious fundamentalists or neo-confederates, whom the wimps disapprove of, so in that case the pattern reverses, with the wimps seeing those they disagree with as evil or stupid, the boors seeing them as reasonable people with, perhaps, some mistaken views.

Ron Paul, or whichever of his people wrote the relevant articles, identified with and was appealing to the boors and so offended the wimps. The Bleeding Heart Libertarians are professional academics, associate mostly with people well on the left, so speak respectfully of even bad arguments that such people respect and would like to revise libertarianism to make it more palatable to their left wing friends.

A Virtual Enchanted Ground Bardic Circle

As some of you know, I am a long time participant in the Society for Creative Anachronism, a group that does historical recreation, mostly medieval and renaissance but including everything before the seventeenth century. I normally spend two weeks of each year camped out with ten thousand other crazy people in a private campground about an hour north of Pittsburgh. One of my contributions to the event is hosting a bardic circle, a group of people sitting around a campfire in my encampment in the evening entertaining each other with poems, stories, songs, music, and on one occasion a play. The idea is to create the illusion of a group of period people entertaining each other, so all conversation is "in persona," done from the point of view of the invented period person you are playing. Not "this song is late Elizabethan" but "this is a song popular at Her Majesty's court." Pieces do not have to be actually period, although period pieces are appreciated.

Due to the present unpleasantness, Pennsic was cancelled this year, I think for the first time in the fifty some years it has been happening. I am accordingly holding my bardic circle online using Zoom, starting this Sunday, which would have been the first night of the second week, when I normally start the circle.

Everyone is welcome, including people who only come to listen, but if you are going to be visible on Zoom you should be wearing something that looks plausibly period.

The schedule is:

Sunday, starting at noon Pacific time

https://zoom.us/j/93136961123

 

Monday-Wednesday starting at 3:00 P.M.

https://zoom.us/j/91931473332

 

Thursday starting at noon.

https://zoom.us/j/95542801469

 

Friday starting at 6 PM.

https://zoom.us/j/99099224114

A Density Query

One factor relevant to different rates of Covid deaths in different countries is population density. People often talk about herd immunity as if it required the same percentage of immune people everywhere, but that obviously is not true. In a society where the average individual only encounters someone else every three months, herd immunity starts at zero percent immune. The more people interact, the higher the fraction that must be immune in order that each infected person will, on average, pass the disease on to fewer than one more.

Frequency of interaction depends in part on population density, but not the density calculated by dividing population by area. Adding a million square miles with nobody living there to a country sharply reduces the average population density defined in that way but has no effect on the average density experienced by an individual. What the usual definition is measuring is  density per acre averaged over acres. What we want is the density averaged over people, giving heavier weight to acres that have more people on them. That difference is the reason that people in a country like the U.S., large parts of which are nearly empty, tend to greatly overestimate how crowded it is — almost all the places they observe are places with people in them.

Does anyone know of a source for the form of population density I want? It might turn out that some countries had a much higher density relative to others than the conventional figure implies.

More on Misleading Covid Reporting

I am seeing multiple news stories warning that Sweden is a horrible example of the bad result of not locking down. They report, correctly, that Sweden's deaths per capita figures is a little higher than that of the U.S., much higher than Denmark and Norway, although a little lower than the U.K.

What they do not mention is that the daily death rate has fallen by almost two orders of magnitude over the past three months, from a high of just over a hundred deaths per day to currently one or two deaths each day, making the Swedish per capita daily death rate less than a tenth the U.S. figure. By my calculations, if daily death rates stay the same in both countries, U.S. total deaths per capita will reach the Swedish figure in about two months.

Can anyone offer an explanation other than deliberate fraud for a news story that mentions the one figure but not the other? It is, after all, exactly the pattern we would expect if a strategy of getting to herd immunity had succeeded.

Source

P.S. I have now seen a different figure for Sweden, with daily deaths about ten times higher. I have not yet figured out why different sources are giving strikingly different numbers, but if the second number is correct, what I wrote in this post is wrong.

Pretending that Good News is Bad News

The headline of a news story today:

CDC study shows COVID-19 cases may be 10 times higher than reported

The story warns that this shows that states that are opening up are making a mistake, that the problem is worse than we thought. It does not seem to have occurred to the author that if there are ten times as many cases as we thought and the same number of deaths, that means that the disease is only a tenth as lethal as we thought it was, which is an argument in favor of opening up, not against. Nobody has been arguing that we should respond to each year's flu season with a lockdown.

The article also does not mention that this is the same result, in a somewhat stronger form, that was reported by people at Stanford quite a while back, on a similar basis, and for which they were severely criticized.

The explanation may be political bias. Trump has tended to downplay the seriousness of Covid, his opponents to do the opposite, and most reporters, probably including the author of this article, are opposed to Trump. But it may also be the bias that leads many people to believe that everyone else should be taking more precautions — against Covid, overweight, or anything else — than they are, hence to regard it as socially responsible to exaggerate any news that makes a threat look bigger, downplay or ignore news that makes it look smaller.

It is an issue I have discussed before, in other contexts, several times.

P.S. Looking at another source, it looks as though the claim is between two and thirteen times the number of known cases (for the U.S.). But I think previous mortality estimates already took account of some asymptomatic cases, so were using something larger than the known cases figure.

That said, the commentary continues to get precisely the wrong conclusion:

Jennifer Nuzzo, an epidemiologist at the John Hopkins Center for Health Security who was uninvolved in the study, told The Washington Post. “This study should put to bed any further argument that we should allow this virus to rip through our communities in order to achieve herd immunity.”

If the number of people who have had the virus is (say) twice the number we thought it was, then the deaths we have suffered so far have brought us twice as far towards herd immunity as we thought. Aiming at herd immunity may or may not make sense, but the new evidence makes the strategy look better than before, not worse, which is the opposite of Ms Nuzzo's claim.

An Idea for Colleges in the Pandemic

The danger in realspace classes during the current pandemic is almost entirely to professors and staff, since students who catch Covid, unless they have some significant medical problem, have a very low chance of dying from it — as best I can tell under one in a thousand. For an adult over 65, which many professors nowadays are, the chance is much higher, perhaps two or three percent. The numbers are very uncertain because we do not know how many undiagnosed cases there are, but the relative risk is at least an order of magnitude greater, perhaps two orders of magnitude, for the older group.

That suggests a policy intermediate between realspace and VR classes — students free to interact in realspace, faculty in VR. Classes  work just as last year except that the professor has been replaced by a video screen. The class can see the professor on the screen, the professor can see the class via a video camera in the front of the classroom. It would be the same approach by which I once gave a talk in the country of Georgia from my office in San Jose. Judging by my experience, it works almost as well as a live talk.

Not all students or all teachers are the same. In an improved version of what I describe, students who are at serious risk could attend virtually, with the feed to the screen also feeding to their computer and with the option to ask questions of the professor over the internet. Teachers who are not at serious risk, such as graduate student teaching assistants, could if they wished attend in realspace, which would help for classes that involved more than verbal interaction.

The last year I taught at Santa Clara University, one of my classes consisted mostly of Saudi LLM students, including a women with whom I had some long conversations about how things worked in her country. A college is either for men or for women. A women's college not only has no male students, it has no males at all inside its walls. The only way a male professor can teach in it is over video.

If the Saudis can do it, surely we can.

One problem that remains is the risk of infected students passing the virus on to parents or, worse still, grandparents. To deal with that, isolate students for a week at the end of the school year, then test them to make sure they are not carrying the virus. Quarantine any who are. Students who have gotten the virus and recovered are free to go home for Christmas and spring break.

Andrew Sullivan on Leaving Print Media for Online Media

Andrew Sullivan has a very interesting column explaining why he is leaving New York Magazine. Unlike Bari Weiss, he has no complaints about how the editors of the magazine treated him. He is leaving because enough of the staff are intolerantly woke that someone with his views no longer fits in.

What has happened, I think, is relatively simple: A critical mass of the staff and management at New York Magazine and Vox Media no longer want to associate with me, and, in a time of ever tightening budgets, I’m a luxury item they don’t want to afford. And that’s entirely their prerogative. They seem to believe, and this is increasingly the orthodoxy in mainstream media, that any writer not actively committed to critical theory in questions of race, gender, sexual orientation, and gender identity is actively, physically harming co-workers merely by existing in the same virtual space. Actually attacking, and even mocking, critical theory’s ideas and methods, as I have done continually in this space, is therefore out of sync with the values of Vox Media. That, to the best of my understanding, is why I’m out of here.

What is interesting is not what he is going from but to. He concludes that print media, at least the parts he has been associated with, is becoming too ideologically narrow and intolerant to be useful. The reason that is not a terrible problem is that there is another alternative, online media, which can be much more open and diverse. His view of what his old blog was like is very much like mine of the Slate Star Codex commenting community.

But here’s what I do truly and deeply miss: writing freely without being in a defensive crouch; airing tough, smart dissent and engaging with readers in a substantive way that avoids Twitter madness; a truly free intellectual space where anything, yes anything, can be debated without personal abuse or questioning of motives; and where readers can force me to change my mind (or not) by sheer logic or personal testimony. I miss a readership that truly was eclectic — left, liberal, centrist, right, reactionary — and that loved to be challenged by me and by each other. I miss just the sheer fun that used to be a part of being a hack before all these dreadfully earnest, humor-free puritans took over the press: jokes, window views, silly videos, contests, puns, rickrolls, and so on.

So he is restarting his blog and shifting his efforts to it.

Jury Trial and the Problem of Punishing Government Misdeeds

The problem of government agents, in particular police, doing bad things is currently a live issue. It is one I have been concerned with for a long time — ever since Chicago police murdered two sleeping Black Panthers back when I was a graduate student in Chicago and were never charged with the crime. Looking back over old blog entries, I was reminded of a more recent and arguably equally outrageous case, the treatment of the FLDS, a fundamentalist Mormon sect that advocated and practiced polygamy, by the Texas child protection services. That was ended when a unanimous verdict of the state appeals court, followed by a unanimous holding of the Supreme Court, held that the acts were illegal, compelling the agency to return four hundred children to their parents.

In the course of the events, the agency committed multiple acts that would, if done by a private person, be tortious or criminal. They extensively lied to the press about the facts, claiming that they had found a large number of minor women who were pregnant or mothers, a claim justified by ignoring documentary evidence of age and assigning ages to the women in their custody based on their own judgement. By the time the case got to the appeals court the total number was down from thirty-one to five, with no evidence that any of them were the result of anything illegal, Texas having had a minimum age of marriage, until a few years earlier, of fourteen, later sixteen. They held prisoner two adult women in late pregnancy on the claim that they were minors, one of whom had shown them her birth certificate and was in fact twenty-two. Only after their infants were born did the authorities admit they were both adults.

Insofar as there was any basis for the initial seizure, it was the belief that minor women might be forced into unwanted marriages. A minority of the Texas Supreme Court held that that would have justified the seizure of such women — there were 27 from 14-17 — but not the rest of the four hundred children seized. Insofar as any justification was offered for seizing boys it was to prevent them from being brought up to believe in polygamy. That makes the action an attempt to destroy a religion, which fits within the current broad definition of attempted genocide.

My feeling at the time, once it became clear how outrageous the behavior of the authorities had been, was that the victims should sue. Looking back at the case, my conclusion is that if they had sued, or if they had persuaded either Texas or Federal authorities to charge those responsible with crimes, they would have lost.

The reason is the jury system. The authorities lost the legal battle but mostly won the battle for public support. Most of the newspapers, with a few honorable exceptions, treated the claims of the authorities as true and ignored or downplayed later evidence that they were false. The perception in the minds of such of the public as read the papers and did not pay careful attention to the facts was that the FLDS was a cult that compelled underage women into the harems of older men, and ought to be suppressed. An honest judge, after carefully examining the evidence, might well have supported my conclusion — that the individuals responsible for the seizures ought to be in jail and the organization owed sizable damages to those they had imprisoned and slandered. The chance that every one of twelve jurors would have been willing to agree, to take the side of a generally despised minority against the authorities of the state of Texas, is close to zero.

The requirement of unanimous jury verdicts in the U.S. legal system is a good way of making it harder for the state to use the court system to misuse its power, to convict people of things they did not do or that were not crimes. But it makes it harder, even with an honest and competent judiciary, to use the court system to punish government actors for their misdeeds. In practice the most it can be expected to do is what the Texas court system, to its credit, eventually did, make them stop doing them.

A problem for current campaigns against unjustified killing by the police.

I discuss the issue of how to enforce law against law enforcers, in the context of a wide variety of legal systems, in one chapter of my most recent nonfiction book.

Covid 19 — A Question and a Comment

1. I have been wondering about the current evidence on mortality rates. Assuming, as I do, that the death figures are reasonably accurate — I calculated an upper bound for the number of people dying of something else while having Covid and being mislabeled in an earlier post — the uncertainty is the number infected. I gather that the high estimates of number infected, hence low estimates of the mortality rate, use the percentage of a random sample who show the relevant antibodies as evidence of how many have been infected.

But I have seen recent stories claiming that a large fraction of those who have been infected don't show the antibodies later. That is being discussed in the context of the question of whether having had it gives immunity, also relevant to whether a vaccine is possible. But it's also relevant to the death rate. If two-thirds of the recovered people fail to show antibodies, then a death rate calculated from the antibody data is three times as high as the real death rate.

Are there any competent webbed discussions of this issue? Anyone here done calculations?

2. I recently had an unpleasant thought about the politics of the response to Covid.

I think everyone agrees that the better the economic situation as of election day, the better the Republicans will do. It follows that Trump would weight economic costs of lockdowns and the like much more heavily, relative to health costs, especially health costs over the next couple of months, than his opponents.

According to which side you are on, you can describe that either as Trump killing people in order to make sure he wins the election or the Democrats trashing the economy in order to make sure he loses.

Of course, health costs also matter for voting, but my guess is that most people make their decision in terms of the current situation, not what it was several months earlier. It's even possible that voters judge by the amount of recent improvement as well as by conditions on election day, in which case bad conditions, economic or medical, a few months back, might actually benefit the incumbent: See how much better things have gotten.