Posted on by Dr. Francis Collins
The alarming spread of coronavirus disease 2019 (COVID-19) last winter presented a profound threat to nations around the world. Many government leaders responded by shutting down all non-essential activities, implementing policies that public health officials were hopeful could slow the highly infectious SARS-CoV-2, the novel coronavirus that causes COVID-19.
But the shutdown has come at a heavy cost for the U.S. and global economies. It’s also taken a heavy personal toll on many of us, disrupting our daily routines—getting children off to school, commuting to the office or lab, getting together with friends and family, meeting face to face to plan projects, eating out, going to the gym—and causing lots of uncertainty and frustration.
As difficult as the shutdowns have been, new research shows that without these public health measures, things would have been much, much worse. According to a study published recently in Nature , the implementation of containment and mitigation strategies across the globe prevented or delayed about 530 million coronavirus infections across six countries—China, South Korea, Iran, Italy, France, and the United States. Take a moment to absorb that number—530 million. Right now, there are 8.8 million cases documented across the globe.
Estimates of the benefits of anti-contagion policies have drawn from epidemiological models that simulate the spread of COVID-19 in various ways, depending on assumptions built into each model. But models are sophisticated ways of guessing. Back when decisions about staying at home had to be made, no one knew for sure if, or how well, such approaches to limit physical contact would work. What’s more, the only real historical precedent was the 1918 Spanish flu pandemic in a very different, much-less interconnected world.
That made it essential to evaluate the pros and cons of these public health strategies within a society. As many people have rightfully asked: are the health benefits really worth the pain?
Recognizing a pressing need to answer this question, an international team of scientists dropped everything that they were doing to find out. Led by Solomon Hsiang, director of the University of California, Berkeley’s Global Policy Laboratory and Chancellor’s Professor at the Goldman School of Public Policy, a research group of 15 researchers from China, France, South Korea, New Zealand, Singapore, and the United States evaluated 1,717 policies implemented in all six countries between January 2020, when the virus began its global rise, and April 6, 2020.
The team relied on econometric methods that use statistics and math to uncover meaningful patterns hiding in mountains of data. As the name implies, these techniques are used routinely by economists to understand, in a before-and-after way, how certain events affect economic growth.
In this look-back study, scientists compare observations before and after an event they couldn’t control, such as a natural disaster or disease outbreak. In the case of COVID-19, these researchers compared public health datasets in multiple localities (e.g., states or cities) within each of the six countries before and several weeks after lockdowns. For each data sample from a given locality, the time period right before a policy deployment was the experimental “control” for the same locality several weeks after it received one or more shutdown policy “treatments.”
Hsiang and his colleagues measured the effects of all the different policies put into place at local, regional, and national levels. These included travel restrictions, business and school closures, shelter-in-place orders, and other actions that didn’t involve any type of medical treatment for COVID-19.
Because SARS-CoV-2 is a new virus, the researchers knew that early in the pandemic, everyone was susceptible, and the outbreak would grow exponentially. The scientists could then use a statistical method designed to estimate how the daily growth rate of infections changed over time within a location after different combinations of large-scale policies were put into place.
The result? Early in the pandemic, coronavirus infection rates grew 38 percent each day, on average, across the six countries: translating to a two-day doubling time. Applying all policies at once slowed the daily COVID-19 infection rate by 31 percentage points! Policies having the clearest benefit were business closures and lockdowns, whereas travel restrictions and bans on social gatherings had mixed results. Without more data, the analysis can’t specify why, but the way different countries enacted those policies might be one reason.
As we continue to try to understand and thwart this new virus and its damage to so many aspects of our personal and professional lives, these new findings add context, comfort, and guidance about the present circumstances. They tell us that individual sacrifices from staying home and canceled events contributed collectively to a huge, positive impact on the world.
Now, as various communities start cautiously to open up, we should continue to practice social distancing, mask wearing, and handwashing. This is not the time to say that the risk has passed. We are all tired of the virus and its consequences for our personal lives, but the virus doesn’t care. It’s still out there. Stay safe, everyone!
 The effect of large-scale anti-contagion policies on the COVID-19 pandemic. Hsiang S, Allen D, Annan-Phan S, et al. Nature. 2020 June 8 [published online ahead of print].
Global Policy Lab: Effect of Anti-Contagion Policies (University of California, Berkeley)
Video: How much have policies to slow COVID-19 worked? (UC Berkeley)
Hsiang Lab (UC Berkeley)
“Global Policy Lab Rallies for COVID-19 Research,” COVID-19 News, Goldman School of Public Policy, June 5, 2020.
Posted In: News
Tags: China, COVID-19, disease prevention, econometrics, epidemiological modeling, epidemiology, France, global health, global policy, Iran, Italy, New Zealand, novel coronavirus, pandemic, public health, SARS-CoV-2, shutdown policies, Singapore, social distancing, South Korea, virology