COVID Precaution Decision-Making:
15 Theses

Every decision about a COVID precaution is a balance.

We balance the expected benefits of the precaution (risk reduction, primarily) against its expected downsides (cost, hassle, unpleasantness, risk, etc.). And we balance the expected benefits of the precaution against the expected benefits of not taking that precaution (economic benefits, psychic benefits, educational benefits, etc.), and perhaps taking some other precaution instead.

We don’t have good data about many of the inputs to these calculations. And we don’t necessarily make the calculations explicitly or even consciously. But we necessarily are conducting some kind of balance.

It follows that COVID precautionary decisions are never dichotomous.

How much safer do we think Precaution X will make us, and how sure are we? What downsides do we anticipate, and how sure are we about those? What alternative benefits would we be forgoing, and how great do we expect them to be? What alternative precautions might have a better ratio of upsides to downsides? None of these is a dichotomous question; they are all matters of degree.

“X will make me safe” is almost never true, because there’s no perfect safety and no perfect precaution. “X will make me safer” is almost never dispositive, because it matters how much safer, with how much confidence, with what expected downsides and opportunity costs, compared to what alternative precautions.

It also follows that COVID precautionary decisions are never purely about COVID.

Some predictable outcomes of COVID precautions are unrelated to COVID itself – for example, their expected economic, psychic, and educational impacts. Most public health officials and experts are less attentive to the non-COVID outcomes of COVID precautions than they are to the COVID-related outcomes of those precautions. And most public health officials and experts are less attentive to the non-COVID outcomes of COVID precautions than the general public is.

Officials and experts mostly tell us what they know (or think they know, or want us to know) about the COVID-related outcomes, leaving us to look elsewhere for evidence about non-COVID outcomes.

It also follows that COVID precautionary decisions are never purely about data.

Decisions about COVID precautions are partly – often largely – trans-scientific; that is, they are about values as well as data. The data (or guesses at the data) tell us how much reduced risk to expect from a precaution, with how much confidence, with what expected downsides and opportunity costs, compared to what alternative precautions. But deciding how to balance these inputs requires value judgments.

People vary in how risk-tolerant or risk-averse they are – and they may be more or less risk-tolerant or risk-averse vis-à-vis one sort of risk than another (the risk of COVID, for example, versus the risk of job loss). People also vary in how they assess the acceptability of various downsides.

Some COVID precautions may be no-brainers, such that nobody who understands the available data could possibly accept Precaution X or reject Precaution Y. But most COVID precautionary decisions are debatable even among those who understand the available data. Often what conclusions to draw from the data is itself debatable – but that’s not what I’m focusing on here. Even if we somehow manage to agree on exactly what mix of upsides and downsides the data say we should expect from a precaution, we will still disagree on how to weight these upsides and downsides, and thus on whether the balance tips yea or nay.

It also follows that COVID precautionary decisions are contextual.

We don’t make three independent decisions about whether to take Precautions X, Y, and Z. We make sequential decisions that are affected by prior ones. Two diametrically opposed algorithms dominate this process.

• Risk budgeting/tradeoffs: Figure out how much risk to tolerate, and then allocate that risk. “I decided not to take Precaution X, so I’d better take Precaution Y” or “I took Precaution X, which earns me the right to eschew Precaution Y.” “I can let myself visit my vulnerable grandparent if I take an antigen test and I haven’t been in crowds without a mask for a few days.”
• Precedent / cognitive dissonance: Strive for consistency, viewing having taken (or not taken) Precaution X as a reason to take (or not take) Precaution Y. “I’ve been so careful all week, so it would be crazy for me to relax my guard now.” “I’ve been so carefree all week, so it would be pointless for me to become super-careful all of a sudden.”

Risk budgeting makes more sense insofar as the goal is balancing the benefits of a precaution against its downsides and against other benefits available if we eschew the precaution. But precedential thinking is psychologically compelling; we all prefer to avoid cognitive dissonance. The empirical evidence says precedential thinking is more common than tradeoff thinking in the realm of precautionary decisions, including COVID precautionary decisions. Most precautions against a specific hazard are positively correlated. Vaccinated people are likelier to wear masks; mask-wearers are likelier to avoid crowds. It is a challenge to find ways to encourage people to think in terms of risk budgeting.

Impacts on others are just additional factors to balance.

Presumably the main factor when making COVID precautionary decisions is the expected outcomes for the individual making the decision. But benefits/costs/risks to others are also often a factor in play, as are benefits/costs/risks to the society at large. I may decide to get vaccinated to protect my child or grandparent. I may decide to go to work while sick to get the job done and avoid overburdening colleagues.

Like all the other factors, these considerations of impact on others and on society are not dichotomous; they are not purely about COVID; they are largely trans-scientific, grounded in values and debatable even when the data are agreed upon; and they are contextual. They’re just additional factors to get loaded into the endless balancing process, where they may weigh heavily or lightly.

Experts and officials owe individuals the best available information.

To guide our COVID precautionary decisions, experts and officials owe us complete information about what they know and what they suspect (being careful to distinguish the two). That is, they owe us as much as they are able to tell us about the expected benefits/costs/risks of taking a specific precaution versus the expected benefits/costs/risks of not taking that precaution. They also owe us clarity about the extent of their uncertainty about the data they are providing. And when new data emerge, they owe us updated information that explicitly acknowledges what has changed.

Cherry-picking data to encourage certain precautions and discourage others is professional malfeasance. It remains malfeasance even if the encouraged precautions are wise and the discouraged ones are foolish. Experts and officials are not licensed to decide which information to give us based on their conclusions about which COVID precautions we should take. We are entitled to know what they know (or as much of it as they can tell us and we can master). Then we are free to decide for ourselves whether we reach the same conclusions they reached.

Experts and officials should not expect – much less demand – that we will always reach the same conclusions they reached.

Even as they provide us with all the data at their disposal, experts and officials should never forget that our COVID precautionary decisions are based on factors other than the data they are providing – data from other sources, perhaps; data from our own life experience; above all, the values that determine how much weight we choose to give each factor in the balancing process. As noted earlier, some COVID precautionary decisions may be no-brainers, such that nobody who understands the available data could possibly accept Precaution X or reject Precaution Y. But most COVID precautionary decisions are debatable even among those who understand the available data.

Experts and officials shouldn’t just recognize this. They should proclaim it and even interpret it: “People who think A and B are important priorities in their lives are understandably less likely to choose Precaution X than those for whom A and B are not so decisive. Precaution Y may be a more appropriate way for them to reduce their COVID risk.”

Experts and officials should recognize and acknowledge their own biases.

The vast majority of experts and officials with access to data relevant to precautionary decisions in a particular realm are likely to be highly atypical in their own precautionary decision-making in that realm. Thus, public health experts and officials are mostly people who value COVID precautions highly, who willingly accept the downsides and opportunity costs of those precautions in their own lives, and who believe that others should do likewise. Their expertise, in other words, comes with a bias – and they need to work hard to avoid misinterpreting that bias as part of their expertise.

They should acknowledge the bias matter-of-factly: “Of course I wear facemasks more than most people, and of course I wish everyone would. I’m a public health professional, so that’s my bias. That doesn’t mean everybody else should necessarily prioritize health over other priorities as much as I do.”

Experts and officials should accept the many factors that go into individual COVID precaution decision-making.

When people in authority are deciding which COVID precautions to encourage – I’m talking about “encourage” here, not “mandate” – they should do so in the context of all that has gone before on this list.

In particular, they should acknowledge:

• the uncertainties and gaps in the data they are sharing, and the fact that they have far more data on COVID-related impacts than on other impacts;
• the inevitable prioritization of health (especially health vis-à-vis COVID) over other values and priorities in the advice they are giving; and
• the importance of people assessing the data and the advice in the context of their own situations and their own values and priorities.

Experts and officials should try to help people think through their own COVID precaution choices.

COVID precaution decision-making is likelier to be wise and sustainable when people are considering their options in the context of their own priorities and values. To encourage this sort of thinking, contingent advice is most valuable: “If you’re A sort of person, you will probably prefer Precaution Y to Precaution X.” Also, it is useful to offer a “Plan B”: “We really hope people will choose Precaution X. But if X feels like it’s too unbalanced regarding safety versus living your life, we urge you to at least do Y.” “If you’re not planning to wear your mask throughout your whole flight, try to keep it on in the airport and at takeoff and landing, when the risk of infection is greatest.”

As noted earlier (see #5), COVID precaution decision-making is best when we are thinking in terms of tradeoffs and risk budgeting. Menus of recommendations help encourage risk budgeting: “Here’s a list of things people do that carry some risk of COVID infection, and precautions people can take that reduce that risk. The list includes a point count for each activity and precaution. Start with the activities that are most important to you and the precautions that are least bothersome to you. If you add more activities, try to add more precautions too so your point total doesn’t go up. We hope you’ll aim for a point total of X or lower. If you can get below X, congratulations! Even if you’re above X, the lower your point count, the safer you are from COVID.”

Mandating COVID precautions is a balancing process too.

Just as individuals have to conduct a balancing process to choose which COVID precautions to take, authorities (governments, employers, etc.) have to conduct a balancing process to choose which COVID precautions to mandate. Part of this societal balancing process consists of considering the same factors individuals consider – but necessarily in a generic, “average” calculation that’s insensitive to individual differences. How much benefit can the typical individual expect to get from this mandate, and with what expected downsides and opportunity costs?

Special attention deserves to be paid to outliers. Immunocompromised people will benefit more than most people from a mask mandate. Religious people will suffer more than most people from closure of houses of worship. The resulting policy questions are daunting: Should everyone have to wear masks to help protect the immunocompromised, or should the immunocompromised wear respirators instead? Should everyone have to endure more transmission if houses of worship stay open, or should the religious do their worshiping at home instead? Note that these are predominantly values questions, not data questions; public health experts are needed to input some of the relevant data, but they are no more qualified than anyone else to conduct the balancing.

The balancing process for COVID mandates includes factors individuals don’t have to consider.

Obviously, the balancing process for mandates includes a paramount focus on some factors that are absent or de minimis in individual COVID precaution decision-making.

Among these:

• Effect on freedom. All COVID mandates reduce individual freedom, and individual freedom is highly valued virtually everywhere (though societies and subgroups within societies differ in how highly valued; some are more communitarian or more individualistic than others).
• Effect on quality of life. In the short term, mandated COVID precautions necessarily reduce quality of life to one extent or another. Their long-term effects are more complex. Some may alter the culture in ways that do permanent damage; others may evolve into communitarian norms that no longer need enforcement. And of course reduced mortality and morbidity, and reduced fear, are surely beneficial to quality of life.
• Effect on healthcare. Precautions that reduce COVID transmission can help keep hospitals from becoming overwhelmed, benefiting everyone who needs hospital care for any purpose. On the other hand, some COVID precautions deter people from seeking medical care they need, exacerbating health problems for decades to come.
• Effect on the economy. People who are told not to go to work, or who fear going to work, undermine the economy, as do people who cannot or choose not to shop. Business closures disrupt supply chains. On the other hand, people who come to work sick and sicken their coworkers also undermine the economy.
• Effect on the educational system. If schools are closed, children cannot learn, or at least do not learn nearly as much. They are also deprived of other benefits from school attendance: social interaction, exercise, nutrition. On the other hand, for some diseases, schools can be significant amplifiers of transmission. (The extent to which this is true of COVID is hotly debated.)
• Effect on social cohesion versus polarization. A national COVID precautionary policy can help unify a country against danger. A mandated COVID precaution that is widely disputed or despised, on the other hand, can polarize a country – with long-lasting negative effects that extend well beyond the danger that provoked the policy.
• Effect on compliance and respect for authority. A COVID mandate that is resentfully complied with nurtures outrage and rebellion. A mandate that is flouted nurtures disrespect for law, for authority, and for public health itself.

The case against COVID mandates is stronger than many experts and officials recognize.

Most (but not all) of the factors in #13 suggest that coercion has more downsides than upsides. It requires a stronger case to mandate a COVID precaution than to recommend one.

Most (but not all) of the downsides of coercion are the sorts of impacts that are outside the expertise of public health experts and officials. As a result, they are all too likely to underestimate these downsides when assessing the wisdom of a COVID mandate. To put the point more strongly: Public health professionals see clearly the upsides of COVID mandates, but may fail to see their downsides. They are arguably among the least qualified to weigh the pros and cons of coercion.

The case for COVID mandates should be made explicitly.

When people in authority mandate a COVID precaution, they owe the public an accounting of the balancing act they undertook before making that decision.

What did they see as the major upsides of the mandate they chose? What did they see as its major downsides? What alternative policies did they consider, and why did they choose this one instead? Under what circumstances will they rescind the mandate?