Daniel C. Elton, Ph.D.

A common refrain we hear from public intellectuals about vaccines prior to Phase III data is “we don’t know anything about the safety or efficacy of vaccine X”. This attitude is both false and misleading to the public, instilling uncertainty and fear about vaccines. To see why it is false, consider if a normal vaccine safety study was done, but by coincidence all of the vaccines were given in hospital rooms that were painted blue. Could we conclude on the basis of such a study whether the vaccine would be safe if administered in rooms painted red? Yes, we can, and we should. We can utilize two forms of reasoning to conclude that the vaccine is safe if given in red rooms, even though we have no data on the matter.

The first form of reasoning roughly approximates the way an ideal Bayesian statistical reasoner would function to compute what is called a “prior probability distribution”. Under this form of reasoning, we consider the millions of doses of similar vaccines (called the “reference class”) that have been administered. For instance, we might consider the vaccines developed for very similar coronaviruses like SARS and MERS.  We note that if the color of paint did affect the safety of those vaccines, this would have likely been detected over the course of prior studies and over the course of millions of doses given previously. Of course, there is a chance the correlation might have been missed. To figure out how big that is, we can go a level deeper and consider a reference class of “things people might notice or fail to notice in medical studies”. We can conclude that for prior vaccines, if such correlations existed they would generally be picked up. On the basis of this and the fact that no such correlation was ever discovered in the reference class of prior vaccines we can conclude that the probability of vaccines like the COVID-19 vaccine being dependent on the color of paint is very small. 

The second type of reasoning, which happens to be much more straightforward in this situation, is what the physicist David Deutsch calls “reasoning from our best explanation of the world”. According to the philosopher of science Karl Popper, we should reason using our explanatory theories of the world which have survived the most rounds of attempted falsification, and which have the highest degree of falsifiability (this rules out non-testable explanations like “vaccines work via invisible ghosts”). In more prosaic terms, this simply means reasoning using the best scientific theories which make predictions in the domain under consideration. We note that our best theories of vaccine function do not anywhere depend on the color of paint in the room. Instead they depend on things like T-cells, binding affinities of molecules, the concentrations of certain molecules in the body, etc. So, we decide that the vaccine is safe regardless of the color of paint in the room where it is administered. 

Both of these forms of reasoning are valid and both are foundational to science, rationality, and human progress. Both of these types of reasoning can be used to say that vaccines under development are likely to be safe and effective before any data comes in. It’s why a reporter who interviewed numerous top scientists reported that they all told him that “they expected the vaccines were safe and effective all along.” Yet instead of proudly sharing this important knowledge with the public, we rarely hear scientists say publicly that they expect the vaccines are safe and efficacious. Instead, they hedge, saying “we have to wait until the data comes in”. This is unethical both on Kantian grounds (they are lying) and on consequentialist grounds, because it leads to undue caution and the public being afraid of vaccines. 

Unfortunately, there is little incentive for scientists to tell the truth about what the likely risks and benefits are with new vaccines before full Phase III data is published. If, for instance, one or two people suffer severe side effects in a Phase III trial (which is rare, but has happened) a scientist who said they suspected it was “very safe” might receive harsh criticism for making a premature assessment. On the other hand, the same scientist will get no pushback for saying “we need to wait for data to make a judgement”. Indeed, they are likely to even be praised for exhibiting the virtues of “caution, prudence, and scientific skepticism”. Moreover, under no scenario should someone be allowed to get a vaccine until the full data comes in, even though it’s fine to allow people to sign up for studies where they have a 50-50 chance of getting the vaccine. Not very consistent, eh?

As US Transhumanist Party Chairman Gennady Stolyarov II has described in detail in an an earlier publication on this site, all of this is the result of a deeply flawed and deadly ethical principle called the precautionary principle, which unfortunately many people have fallen under the sway of. The principle originates in the environmentalist movement but is widely applied in medicine, and was instrumental in decisions such as the Bush administration’s ban on stem-cell research and decisions to ban life-saving GMO technologies such as golden rice. It has been formulated to varying degrees in several different ways. The United Nations World Charter for Nature (1982) issued one version of the principle, stating: 

Activities which are likely to pose a significant risk to nature shall be preceded by an exhaustive examination; their proponents shall demonstrate that expected benefits outweigh potential damage to nature, and where potential adverse effects are not fully understood, the activities should not proceed. 

The principle starts off OK but dives into serious error in the last line. The issue is that the precautionary principle only focuses on the potential adverse effects of proceeding and ignores the potential adverse effects of not proceeding, i.e., the effects of delay. As should now be clear in the case of the COVID-19 vaccines, not proceeding can sometimes be much more deadly than proceeding! There is often a high but unclear risk to not proceeding, and a low but unclear risk to not proceeding. (Picture two probability distributions, both wide (unclear) but one with a mean that is distinctly higher than the other). That’s where the precautionary principle throws expected utility theory (cost-benefit analysis) out and says we cannot proceed. The Nobel Prize-winning physicist Freeman Dyson stated the issue as follows: 

“The Precautionary Principle says that if some course of action carries even a remote chance of irreparable damage to the ecology, then you shouldn’t do it, no matter how great the possible advantages of the action may be. You are not allowed to balance costs against benefits when deciding what to do.” — Freeman Dyson, Report from the 2001 World Economic Forum

Imagine an alternative world in which our society and government was not under the sway of the precautionary principle. In this alternative world, scientists would give their truthful assessment of new vaccines to the public, stating that they are likely safe and effective, using one or both of the reasoning methods mentioned above. In such a world, given the clear potential harms of inaction, the public would be allowed to purchase new vaccines if they wanted, if the companies manufacturing them were comfortable doing so, and if they were fully informed prior to their decision that they were taking an unapproved product that carries potential risks but also potential benefits. Initially, only a few people would purchase the vaccines, perhaps on the basis of Phase I results. These would be folks like those who injected themselves with a DIY vaccine over the summer, and the tens of thousands who were willing to participate in clinical trials as early as last spring. Companies would be incentivized to survey those who took the vaccine and collect self-reported data on their outcomes, which is very cheap and easy to do. After a few months going by without any of those people keeling over and dying, and with very few (likely none) of those people getting hospitalized for COVID-19, more people would feel comfortable getting the vaccine. Things would quickly snowball, with more and more people becoming willing to get the vaccine. During this time the distribution system would have been stood up and become operational, with on-site stockpiles building up ahead of the FDA’s Emergency Use Authorization (currently, the FDA does not allow hospitals to stockpile unapproved vaccines ahead of their EUA). To present this case in its strongest form, in a future post I plan to estimate how many lives would have been saved, assuming many vaccines had become available to those who wanted them last March or August. However, I hope it’s easy to see that thousands of lives would have been saved in this alternative world.  

For more on the transhumanist alternative to the precautionary principle, the proactionary principle, see Max More’s excellent book chapter as well as the Wikipedia article and references therein.