The Divided State of American Public Health

Modelling shows that a consistent approach to masks and lockdowns could have prevented two-thirds of U.S. deaths during the first months of the pandemic.

There are 50 states in the United States of America—and there have been more than 50 different approaches to suppressing and managing COVID-19.

Even as the pandemic radically reshaped American life, the federal government failed to provide strong and consistent guidance about how best to manage this public health crisis. Not surprisingly, individual states diverged in what they chose to do, and when they chose to do it.

On March 19, California became the first state to order a lockdown. On April 17, Connecticut mandated masks in public spaces. Other states, like South Dakota, chose to do neither.

The result has been a rolling series of localized COVID-19 outbreaks with the death toll exceeding hundreds of thousands. It is a tragedy of historic scope. But these varied policy approaches have also provided a trove of data that enables the examination of which public health policies have been most effective in the U.S. context.

New statistical modelling by McGill University’s Guillaume Roussellet, an assistant professor of finance, has shown that most COVID-19 fatalities in the U.S. could have been prevented if the policies enacted by states like California and Connecticut had been applied across the entire country.

“We observed the number of deaths per day, per state, and correlated those numbers with the different policies that have been implemented, and the date those policies came into effect. This has the potential to identify the different trajectories that these policies yielded,” explains Roussellet.

“Lockdowns reduced transmission by about 40%. Masks can reduce transmission by as much as 85% or even more if people would wear them correctly and regularly. But because not everyone will wear them—or wear them properly—their efficiency is reduced.”

In the article “Preventing COVID-19 Fatalities: State versus Federal Policies,” Roussellet and his co-authors, Jean-Paul Renne (Université de Lausanne) and Gustavo Schwenkler (Santa Clara University), found that masks had been similarly effective as lockdowns, reducing transmission by a little more than 40%. If either one of these policies had been implemented more widely, tens of thousands of lives could have been saved.

“If all states had implemented a stay-at-home order as early as California did, and if all states had implemented a mask mandate as early as Connecticut did, we found that about two-thirds of the lives lost would have been saved.”

The research also found that the states that acted to contain COVID-19 achieved some success.

“If individual policies had not been implemented at the state level, we would have seen between five and six times more fatalities by the end of September 2020,” says Roussellet.

“When states have been strict in their approach to tackling the spread of COVID-19, it has been somewhat effective. But the policies were not unified across the country, and some states decided not to implement them, which led them to faring worse.”

Travel bans achieve little, have unintended consequences

The research also found that inter-state travel bans achieved much less than either mask mandates or lockdowns. When people are prohibited from entering another state, it can even intensify the worst outbreaks. This is because of the exponential math of virus transmission. Preventing people from leaving the hardest-hit states or cities increases population density in the places that have the highest levels of transmission.
“The transmission rate of the virus is affected by social activities and the average way people interact with each other,” says Roussellet.
“The transmission rate of the virus is affected by social activities and the average way people interact with each other,” says Roussellet.

“Whether people are hugging or staying at a distance is going to make a difference. And when there is a huge disparity in these values across state lines, then it can make sense to allow people to go to a state where there is a lower transmission rate, because the growth of the virus is non-linear. A state where the transmission rate is twice as high as another state doesn’t make it twice as bad, it makes it much, much worse than that.”

Even in the hardest-hit communities, the overwhelming majority of people are not infected with COVID-19 at any given time, and preventing people from moving to areas with lower transmission rates can have the effect of making those existing outbreaks larger and more intense. Eventually, this spills over into neighbouring states.

“The main effect that a travel ban has is redistributing the fatalities across different state lines,” says Roussellet.

“When a state prevents people from entering and imposes a lockdown, it becomes a net exporter of infected people. Thus, any neighboring state that receives positive inflows of infections will see their cases rise over the long run. Our findings show that the two effects, both positive and negative, roughly compensate.”

Uncertain times make for difficult decisions

Roussellet’s analysis found that, in the U.S. context, masks and lockdowns have slowed virus transmission in approximately similar measure; but, given the severe economic impacts associated with lockdowns, he argues that mask mandates are the more desirable policy option.

“At the end of the day, policy makers have to make choices that involve some sort of cost-benefit analysis,” says Roussellet.

It is not our job as economists to say what they should do, or should not do.
“It is not our job as economists to say what they should do, or should not do. But it is our job to lay down the problems in the clearest possible way. Sometimes, it could feel to an outside observer that it is a bit too much, because we are talking about human lives. This can be very emotional, but at the end of the day, these are choices that policymakers will have to deal with—things like whether schools should be open or closed, and what sort of risk is involved. The pandemic has had an unequal impact across age groups and social groups, and policy decisions will have winners and losers, but policymakers still must make those decisions. They should have some idea of what the consequences will be.”