By Bora Lee Reed

How does a team consisting of one professor and fourteen graduate students, postdocs, fellows, and staff, rally to conduct many months’ worth of analysis in just eight days?

On the morning of March 13, 2020, a few days before shelter-in-place orders went into effect for the San Francisco Bay Area, Professor Sol Hsiang and his wife were eating breakfast and listening to the news. As the director of a UC Berkeley lab that studies climate change’s impact on society, it hadn’t occurred to him to deploy the Global Policy Lab to understanding COVID-19. But when his wife suggested it, he was intrigued.

“Our team specializes in using econometrics to understand the effects of different policies around the world,” says Professor Hsiang. “For example, when hurricanes hit a country, what happens to the economy? That's a question we’ve asked in the past. We realized we could study the exponential growth of these infections in a similar way. Just as we’ve studied what happens to an economy when a hurricane hits it, we could study how an infection is impacted when a policy is applied to it.”

Professor Hsiang raised the idea at the lab’s all-team meeting. Everyone signed on immediately. 

[from to left to right: Jaecheol Lee, Emma Krasovich, Peiley Lau, Jeanette Tseng, Sol Hsiang, Luna Yue Huang, Ian Bolliger, Kendon Bell, Trinetta Chong, Esther Rolf, Tiffany Wu, Sébastien Annan-Phan, Hannah Druckenmiller, Daniel Allen, Andrew Hultgren]

“We were still in shock, trying to absorb news about the outbreak and adjusting to the sudden switch to quarantine life,” says Jeanette Tseng (MPP '16), a project manager in the lab. “This project gave us a sense of purpose.”

The team temporarily set aside current projects, which range from studying transboundary air pollution in Asia to estimating the economic impact of climate change on coastal regions to assembling 1.7 million historic aerial photographs to generate insights into the long-term relationships between society and the environment. 

They quickly determined that while epidemiological models were being deployed to shape policies like shelter-in-place, closures of restaurants and schools, and travel restrictions, little had been done to measure how well these policies were actually working. 

The team considered the places that had first been impacted by the outbreak, as well as the countries where team members spoke the language. With these criteria in mind, they decided to focus on China, South Korea, Italy, France, Iran, and the United States. 

Then they got to work.

“We realized the team had personal insights into countries where the outbreak had first occurred, and could look beyond the numbers to read government press releases and news articles,” says lab manager Tiffany Wu, a bilingual Mandarin speaker who helped collect health and city-level policy data for China. 

In an eight-day sprint, they gathered health and policy data for each country, ran the analysis, designed the figures, and wrote the paper. Despite sheltering-in-place in three countries (US, New Zealand, and Singapore) and across four time zones, they met daily via video conferencing to check on progress and to help one another. 

The result is the paper, “The Effect of Large-Scale Anti-Contagion Policies on the COVID-19 Pandemic” published in Nature.  [Read about the results of the paper]

“It was insane but rewarding,” says Andy Hultgren (MPP '16), an NSF Fellow from Agricultural and Resource Economics (ARE) who oversaw policy and health data collection for Iran. “With real-world data, we were able to underscore the point that health policies, while very costly, averted what would have been an overwhelming disaster.”

The team collated and constructed a unique policy dataset (now publicly available) from over 1700 different policies being deployed at the national, state, county, and city level across these six different countries. 

“Collecting and analyzing that amount of data usually takes several months,” says Jaecheol Lee, a Fulbright scholar with ARE who helped collect data from South Korea.

“We worked around the clock to publish meaningful and robust results to help policy makers,” says Sébastien Annan-Phan, a graduate student with ARE who worked on the analysis and contributed to French health and policy data collection. “It was an exhausting, unique, and amazing experience.”

The diversity of data sources presented challenges to the team, who had to categorize policies across many localities. 

“We had to take qualitative (and often messy) policy and code it in a standardized manner that would allow us to input into our main model,” says ARE PhD Candidate Peiley Lau, who headed up the policy data collection team. “We had to think very carefully about the many intricacies in the coding up of that data.”

The team also had to work in three different programming languages, making sure all the individual pieces could “talk” to one another. The process, notes Ian Bolliger, PhD Candidate in the Energy and Resources Group (ERG) and former researcher at the Institute for Health Metrics and Evaluation, had to carefully balance two often-competing values: speed and certainty. 

“It was going to benefit society to produce this research as soon as possible,” he says. “But we needed to ensure that the nuanced challenges of communicating across team members and programming languages did not lead to inconsistencies in results. ” 

As the project progressed, different sub-teams uncovered facets that shaped their overall approach and argument.

Emma Krasovich, who holds a MPH in Environmental and Global Health, conducted an epidemiology literature review that helped develop a more precise understanding of how the project’s empirical estimates might ultimately be useful to public health experts. 

“We had the sense that if we all put our minds to it and worked very hard, we could produce research that would make a difference right away,” she says. “In research, you don't often have the chance to work on something that may have an immediate impact. With this project, we all felt an urgency—we recognized we had the chance to do something special.”