
After the economic losses, unemployment and mass death of the COVID-19 pandemic, the question of how to effectively manage lockdowns has to be answered. Many experts believe another pandemic is inevitable. If that is true, we have to be ready with empirically tested policies that can maximize economic welfare while preserving life to the maximum extent possible.
We know, for example, that some diseases are highly contagious, while others are more lethal and that different diseases can affect population groups in different ways, with some more deadly for the elderly than for the young. We also know that the elderly, who generally have retired from work, are more able to engage in voluntary isolation than the young who must still work to earn a living.
What do all those things mean for the design of policies that aim to balance the economic and health effects of a pandemic? And how would the optimal policy in terms of lockdowns look like with diseases that have a less steep age gradient — i.e. that have a less disparate effect on the young and the old.
A New Economic Model of Pandemics
My colleagues and I sought to tackle these vital questions by developing an economic model of pandemics that features age differences and individual choice. We used information from the American Time Use Survey (ATUS) on time use among the young and old in the United States before the COVID-19 pandemic and Google mobility data from cellphones for their time use during that crisis to both develop and robustly test this model. We then sought to determine what kind of rational decisions people in different population groups would make with regard to social distancing, given their infection risk and the chance of a vaccine’s arrival.
Basic assumptions were made. Social distancing provides protection. But it comes at the cost of forgone earnings and diminished leisure enjoyment, which can only be partially ameliorated by teleworking and safe leisure activities. Initial disease symptoms also leave individuals and the government with incomplete information on infections, making testing capacity important. And the disease’s deadliness, the need to earn a living, and the probability of natural death are different for the working age population and for the elderly. This gives the government an opportunity to design effective policies that impose different lockdown restrictions according to age.
We first tested this model with real 2020 data from the COVID pandemic in the United States. During COVID, activities outside the home were reduced both by voluntary self-protective isolation and by government-imposed lockdowns. According to our model, this reduced outdoor activity led to a death toll that was 80% lower than in a purely epidemiological world where the pandemic raged on, but individuals failed to adjust their behavior.
Voluntary isolation was important in-and-of-itself. Imagine a world in which the government did not impose any lockdowns. In such a world, older individuals would shield themselves substantially. The young would also reduce work and outside leisure, but much less so due to their lower risk of dying and their need to earn a living (teleworking is a lower productivity activity than in-person work). The death toll in this no-lockdown world would have been 65% lower during COVID than in the epidemiological scenario in which individuals did not change their behavior.
The Benevolent Social Planner
Now, suppose a benevolent social planner chooses policies with the goal of maximizing the well-being of the different groups in society. Such a planner would consider the welfare of both the young and the old. According to our framework, the planner’s socially optimal lockdown would be stricter than what was implemented in the United States and would reduce deaths across both age groups. Increased restrictions would predominantly impact the young, while the old would gain more outdoor time due to a lower threat of infection. This asymmetry would be intentional: the young tend not to take strong precautions to limit the spread of the disease due to their low personal risk, leaving the old to bear an undue burden.
The framework we developed can also study lockdowns for different pandemics. We start by focusing on the Spanish flu pandemic of the 1910s. The optimal lockdown back then would have entailed milder reductions in social interaction, even though the overall death rate of this disease was higher. This is based on a combination of factors that were different 100 years ago: a younger population, the inability to telework, and a different virus, among others. Since young people had a higher chance of dying, they were more likely to voluntarily isolate themselves to limit their chance of death. The inability back then to telework would be a countervailing factor, as young people would have a harder time making a survival income from home. But mathematically taking these factors into account, the optimal lockdown would still be milder than under COVID.
The same framework applies to other diseases with different levels of infectiousness and deadliness. The age gradient is always a crucial factor: if the case fatality rate (CFR) is high among the young, a sizable and active group, fewer additional restrictions are necessary due to their greater voluntary precautions. Economic conditions are also critical. In scenarios where the older population is smaller or teleworking is easy, a less restrictive policy is optimal.
Testing During a Pandemic
We also explore several aspects of testing. On its own, testing does not eliminate a disease as infectious as COVID-19. It does, however, significantly alleviate the impact of the virus. For that reason, the optimal lockdown is influenced by the capacity of the testing regime. If that is robust, a less restrictive lockdown is warranted, reducing GDP losses and facilitating a quicker easing of restrictions. Where tests are costly and scarce, however, prioritizing them for the young is important, since this allows the targeted isolation of individuals that test positive who are otherwise likely to engage in more social interactions when not forced to isolate.
We hope our findings will make for more rational policymaking if and when another pandemic strikes. During COVID, lockdowns were highly controversial, given competing interests among different population groups and the inevitable health and economic sacrifices that different regimes entailed. A more empirical, evidence-based approach takes some of that controversy away, allowing for the design of policies that maximize welfare overall.
Leave a Reply