According to a scientific paper just published, lockdowns in Europe and the US decreased COVID-19 mortality by a mere 0.2% on average, while the economic costs were enormous. This is no longer surprising. Most recent data actually tell us the same story: lockdowns were a failure. But of course, post diction with data is easier than prediction in a situation where there are no (or almost no) data. But then, why have so many countries chosen the lockdown option so forcefully in the absence of adequate data?
One possible answer points at two different but interconnected processes. The first refers to a dynamic of social conformism (or “informational cascade”) as in the famous Solomon Asch experiment: if individuals are under pressure to be right, rational conformism and system-wide error likely follow. Suppose a Prime Minister wants to figure out whether a policy (in our case, the lockdown) is good or bad. She gathers together a set of experts (say virologists) and asks them, in sequence, what they think. Now, suppose that each expert sincerely wants to give the right answer (there are no ideological incentives). Although none of the experts has perfect knowledge, each of them has insights into the situation: they know they are probably right, but they are not certain. They also know that this is true for all the other experts. The first expert answering the prime minister’s question has only her own knowledge to rely on, and gives her opinion. She may be right or wrong but gives her best guess. The second expert will take into account what he knows, and also the first expert’s answer. If the first expert gets it wrong, and the second expert agreed with the first expert, then the third expert will be influenced by the those who spoke out before he did, and will fall in line (quite rationally from his point of view), no matter what his private information would have suggested him to say to the prime minister. The same is true for all the others who follow. In particular, the statements by the last experts in the queue hardly carry new information. The experts’ answer will thus be unanimous, but may well be totally wrong.
The logic of information cascades can explain rational conformism, and also how groups of experts can rapidly change their minds. For example, imagine a situation in which after a set of experts gives the same reaction, one new expert comes into the picture, unaware of his colleagues’ opinion. He could thus offer new information to a rather uninformed environment; and the fresh information can be enough to reverse the self-propagating dynamic described earlier.
This is not the end of the story, though. Indeed, political incentives are also at work. Consider a political leader operating in a democratic context in the middle of the pandemic. She can follow two strategies: a) imitate what leaders in other countries are already doing (perhaps because some scientists support that option…); b) try a new road. Which of the two options will she choose? From a benefit-cost ratio, the answer is pretty obvious. She will be highly praised if she selects b), and b) turns out to be the correct choice. But she will pay a high cost if she goes for b), and b) is wrong: she and only she will be held responsible. The fact that the life or death of people in the short-term are at stake, a clearly important issue, magnifies the effects. On the other hand, she will not stand out as genius if she chooses instead a) and a) is correct. But she minimizes the loss of prestige if a) turns out to be wrong. After all, she was following “science”. The upshot is that a politician will have strong incentives to opt for the least risky strategy (“Così fan tutte“). Hence, the choice made by the first (reputed) democratic country is shared by all the others. This is what happened in the West, the where the Italian government made the first move and went for the lockdown.
Still, not all the countries decided to go the same way, or at least, not with the same intensity. How so? It is not necessarily because some countries cared about the freedom of their own citizens per se. Rather, it is because some democratic governments were particularly sensitive to the possible reaction of their citizens (at the voting booth). Some data seems to support this view.
The figure below compares the percentage of people from different democratic countries which were unwilling to sacrifice their own rights and freedoms in order to improve public health conditions during the COVID-19 pandemic; with the COVID-19 Stringency Index, based on nine indicators measuring governmental policies during the pandemic. The index goes from 0 to 100 (100=strictest) and includes school closures, workplace closures and travel bans. The percentages come from surveys carried out between March and June 2020, while the index refers to early 2021: the correlation is rather high and negative, as expected (r-Pearson: -.74).
The moral of the story is straightforward: You’d better consider the decision of the first-comers with a grain of salt. This also applies to the next crisis (climate change?).
