Too many policy discussions surrounding COVID-19 create a false dichotomy between those who prioritize the economy against those who prioritize reducing mortality. Data cateloging the spread of the virus, the severity and mortality rates, and time to vaccine are inconstant and contested from all sides of the policy debate. As a result, months into the pandemic, we (the public) don’t have a universally accepted model that predicts the end of the outbreak.
This uncertainty forces governing bodies to make a choice for when, what elements, and how to resume a functioning society. In the US the choice for how and when to reopen are left to the cities and states to determine. In some ways this make sense. Because the desease spreads unevenly and different regions will experience different severities of outbreak, multiple solutions might perform better than a one-size fits all approach. On the other hand, devoid of national guidelines, chaos reigns supreme.
Pundits, armchair politicians, and anyone with a computer have been quick to criticize places for opening too quickly or being too hesitant, but rarely offer a competing weighing mechanism or policy timeline.
I propose that we use the aggregate life-year change as the weighing mechanism to determine if a governing body should pursue a specific COVID-19 policy. The life year is a natural bridge between the economic and mortality focused sides of the debate. The idea behind the life year is to look at the loss of years from expected life that are tied to each COVID-19 death on one side and the number of anticipated year increases from a policy change on the other. There have been a number of studies on life year changes from educational attainment, child abuse, and income. Arguments for ending rest-in-place policies often cite educational losses, increase in child abuse, and decreases in incomes as reasons to reopen society.
Using the life-year scale will face two significant challenges. Many will be uncomfortable with discussions of death in years rather than lives and consider the scale dehumanizing. The second relates to quality of data, because mortality rates are still in question, critics will reject this scheme. To that end, I recommend using a very conservative mortality rate. Use a number that is almost certainly higher than truth across all age groups (e.g. 2% aggregate).
My recommendation is not a novel one. In a recent Marginal Revolution post, Tyler Cowen introduced a study that used life-year analysis (although he did not explicitly recommend its use). A group of researchers from South Africa explored the question of COVID-19 lockdown policies and expect that over 26,000 years will be lost to South Africa’s lockdown policies.
This analysis could help state and local leads make informed, data-driven, policy choices and provide structure for the national debate on COVID-19 policy.
Polly Wanna Math 