Previously, I’ve commented on how the data we are using to estimate the danger of this disease are extremely poor. Until very recently, for the purposes of estimating the danger, we have been testing the wrong thing (current shedding of the virus) the wrong way (mostly testing people who present themselves as sick). When you read that as of March 3, the WHO estimated the death rate of COVID-19 cases at 3.4%, you have to keep in mind they had non-random testing, testing only for current infection, and testing based almost entirely on sick people presenting themselves for care. The result is that there is severe selection bias which pushes the hospitalization and death estimates upward. The big question is by how much. None of us would be able to publish a paper in a third-rate econ or poli sci journal with such bad data; the editors would desk reject us. Nevertheless, governments around the world used such estimates to impose economic misery and dramatic restrictions on civil liberty on the masses.

On top of this, as economists and other math savvy people look into epidemiology, it’s becoming clear that the models they use are quite poor, because they have difficulty with endogeneity and with variance.

Shortly, I suspect my friend Phil Magness will go public with an article about how many of the epidemiological experts you see on TV and whose models are being used to create government policy have a long (20-30 year+) history of making dramatic and sometimes apocalyptic predictions about the dangers of past diseases, predictions which never came true, even though in the past governments did little to stop those diseases.

How does this bear on politics?

In Against Democracy and elsewhere, I’ve argued that competence is a precondition of political legitimacy and authority. The Competence Principle says:

It is presumed to be unjust, and to violate a citizen’s rights, to forcibly deprive a citizen of life, liberty, or property, or to significantly harm her life prospects, as a result of decisions made by an incompetent deliberative body, or as a result of decisions made in an incompetent way or in bad faith. Political decisions are presumed legitimate and authoritative only when produced by competent political bodies in a competent way and in good faith.

My main argument for this principle is by analogy to clear cases. I ask readers to imagine a capital murder trial. A defendant is accused of first degree murder. If found guilty, he will lose his property, his freedom, and possibly his life. Imagine the jury finds him guilty for any of the following reasons:

1. Ignorance: They simply ignore the facts of the case and flip a coin.
2. Stupidity/Lack of Understanding: The case requires sophisticated reasoning and analysis, which they lack the capacity to do.
3. Maleficence: They find him guilty because they hate people like him (e.g., suppose he’s white, rural working class Republican and they are average university professors).
4. Selfishness and Conflict of Interest: They find him guilty because they personally benefit from him going to jail or being executed. (E.g., suppose they own a rival business, or suppose they would get fame and fortune for being the jurors who put him away, regardless of whether he is actually guilty.)
5. Irrationality: They pay attention to the information, but process it in highly irrational ways, beset by a wide range of severe cognitive biases.
6. Conformity and authoritarianism: They find him guilty because they have a political bias to defer to state power, to do what is expected of them regardless of whether it’s right, or to be seen as doing something/anything during times of crisis.
7. Misinformation: The jurors decided properly in light of the information they had, but it later becomes clear the information was extremely poor, misleading, or false.
   

If we learned the jury found him guilty for any of this reasons, we would hold their decision is unjust. Moreover, it would be wrong to enforce their decision. The defendant could demand a retrial, and in many states, would be entitled to one.

I think this point generalizes to many political decisions beyond jury cases.. When a person or group makes a high-stakes decision, imposed involuntarily and through force upon others, a decision which can greatly alter people’s life prospects and deprive them of property, happiness, freedom, or life, that person or group must be competent in general, and must make that particular decision competently and in good faith. If they fail to do so, then their decision is presumed to lack authority (there is no obligation to obey it) and legitimacy (there is no moral permission to enforce it).

Now apply this to government actions on the basis of the COVID-19 disease.

As a philosophical matter, it’s easy to show that in principle, governments can restrict our freedom to stop the spread of disease. For instance, in The Journal of Medical Ethics, I have a paper arguing that governments can force us to accept vaccinations, not for paternalistic reasons, but to stop individuals from imposing unjustifiable risk of disease upon others. At his blog, anarchist libertarian powerhouse Michael Huemer says something similar:

Of course, what counts as unreasonable risk is open to debate. It’s going to have to do with the probability of harm, the total magnitude of the threatened harm, and how good one’s reasons are for imposing it (see previous post on meat & disease risk).

That’s the core of the libertarian justification for disease-prevention measures. Any individual who is at risk of carrying a communicable disease, such as Covid-19, is posing a risk of physical harm to others when he interacts with them. If the risk is ‘unreasonable’ (in light of the probability, magnitude, and reasons for imposing), then those under this threat would be justified in using coercion to protect themselves from the potential physical harm. Since individuals could justly do that, they can also delegate it to the state to do that (if you accept the state as legitimate in general).

The question of whether governments may in principle do what they are doing is not terribly difficult. But appealing to abstract principles is not enough to justify their actions. We need to know whether they made these particular decisions competently and in good faith, on the basis of good information. In the same way, it’s one thing to show in the abstract that states might have the right to punish criminals, but that doesn’t suffice to justify any particular jury decision. We still need to know whether the particular jury acted competently and in good faith, on the basis of good information.

This brings me to the upshot. Governments around the world appear to be relying on epidemiological models which suffer from serious endogeneity problems and which we know do not handle individual variance well, and which are constructed on the basis of the wrong data collected the wrong way. They thus appear to be deciding incompetently, on the basis of bad information. Whether they are acting in bad faith, I leave to you. (I would like to remind you, however, that we have plenty of evidence they often act in bad faith. For instance, bad faith is pervasive in the US criminal justice system.) Go ahead and remind yourself of your analysis of Bush’s decision to invade Iraq, or whether the Patriot Act and the surveillance regime it created is justifiable. While you’re at it, remind yourself of all of those papers published in political science showing that people have a bias toward authoritarianism during a perceived crisis. Surely, that bears on you now, no?

I very much doubt that there are “secret data” of the right sort collected the right way which all governments around the world are holding from us. Instead, they made dramatic decisions, decisions which have little effect on rich intellectuals like me, but which impose severe pain and suffering upon the poor. It’s looks to me like they are blatantly violating the Competence Principle and their decisions presumptively lack authority and legitimacy

The best argument against this position, I think, is something like this: We are in the midst of a possible humanitarian disaster, which could potentially kill millions or tens of millions. Leaders had to act fast on the basis of poor information. They saw what was happening in Italy and took extreme measures.

Maybe, but some rejoinders: First, governments could have collected better data earlier, before they shut the world down. Second, few governments are trying to collect good data now. It’s one thing to shut down in an abundance of caution, but they should subsequently do mass, randomized testing for antibodies so we can determine the real infection fatality rate. (That is, collect the right data the right way.) Why isn’t this being done en masse? Third, the argument that we are in the midst of a potential disaster and so had to act out of an abundance of precaution relied on things like the WHO estimates and other early models and estimates, all of which relied on the wrong kind of data (testing current viral shedding) collected the wrong way (mostly testing people who present themselves as sick). As I’ve been saying, none of you would get a paper published in a third-rate journal with that kind of data, and if I presented a paper using it, you would tear me apart. Fourth, whatever plausibility this argument may have, what about the contrary argument that the bigger the stakes, the better the information you must have?

Note well: I am not a “COVID-19 skeptic” or a conspiracy theorist. I don’t think there is a conspiracy; I just think there is mass government failure. I am not skeptical of the dangers of COVID-19; rather, I am uncertain how bad it is because the early work relied upon poor data and poor research methods.