A makeshift memorial for Li Wenliang, one of the doctors who warned the Chinese government about the coronavirus, February 2020
Stringer / Reuters

Political systems matter a great deal when it comes to disaster response. In general, democracies are better than autocracies at managing early warning systems: it is harder to suppress reports of an impending crisis in an open and transparent political system than in a closed system that incentivizes withholding unpleasant truths. In China, the central government sanctions local officials who fail to keep infectious disease rates below a prescribed level, thus encouraging them to underreport cases. That is what happened last December, when officials in Wuhan—the city where the novel coronavirus originated—kept quiet and punished the doctors who first reported the outbreak. No bureaucrat likes delivering bad news to superiors, but officials in democracies are more likely to do so because they know that if they don’t, the legislative branch, civil society, think tanks, and media could publicize their failure to report.  

It was this general logic that led the Nobel Prize–winning economist Amartya Sen to argue that no famine had ever taken place in a democracy. This claim has invited skepticism and critique, but Sen’s point that data collection and communications tend to be better in democracies was spot-on. Many of the worst famines in the twentieth century—such as the Ukrainian famine of 1932–33 and the Chinese famine of 1959–61—were well underway by the time the outside world became aware of the catastrophe. Then, in 1985, the United States Agency for International Development (USAID) developed a cure for this blind spot: the Famine Early Warning Systems Network, better known as FEWS, which has accurately reported most food crises in the past 40 years before they devolved into famines.

FEWS uses climate forecasting, nutrition surveys, satellite images, and data about the trade of staple grains and livestock to forecast food insecurity before it happens. It then produces maps of food-insecure areas as well as open-source analyses of brewing problems around the world—resources that make their way into the hands of thousands of development and humanitarian relief professionals who play a role in famine mitigation efforts. In his book Mass Starvation, Alex de Waal, a well-known scholar of famines, tracks famine deaths over the past 150 years. His analysis shows a noticeable decline since the mid-1980s, in part due to FEWS.

To resolve the ongoing problem of disease reporting in autocracies, U.S. policymakers could create an early warning system that applies the lessons of FEWS to pandemics. This Pandemic Early Warning System (PEWS) could serve as a sister mechanism to FEWS within USAID and operate using the agency’s existing emergency response mechanisms.  


Like its famine-detecting cousin, PEWS will rely on information gathered both from afar and on the ground. Several remote means of acquiring infectious disease data now exist, but they have yet to be systematized. Commercially available satellite photography can reveal the sudden appearance of mass graves or the increased use of crematoriums. Satellites can also track the size of crowds outside of health clinics and hospitals. A sudden influx of patients isn’t proof of an outbreak, but it does signal severe health distress and may justify an on-the-ground investigation.

Data from remote monitoring systems become even more powerful when analyzed in conjunction with online data. According to a new study by scientists at Harvard University, satellites detected a marked increase in hospital traffic in Wuhan beginning in August 2019, at the same time search engines recorded an uptick in searches within the Wuhan area for terms related to the specific and unique symptoms of COVID-19, the disease caused by the novel coronavirus. Taken together, the study’s authors argue, these two trends suggest that the coronavirus outbreak may have begun much earlier than previously thought.

Policymakers can track infectious diseases in animals before they make the leap to human beings.

Commercial transactions offer still more clues about the spread of infectious diseases. The Centers for Disease Control and Prevention already keeps tabs on seasonal influenza outbreaks by observing the daily sales of pharmaceuticals within the United States. During George W. Bush’s presidency, the U.S. Department of the Interior tested wild bird carcasses for influenza viruses along migratory routes. By cooperating with public health and wildlife authorities in allied countries, U.S. officials could expand these systems to include information about drug purchases worldwide and human-animal interactions in a variety of environments. In countries that refuse to collaborate, data can often still be collected or extrapolated from neighboring states.

Finally, policymakers can track infectious diseases in animals before they make the leap to human beings. From 2009 to 2019, USAID ran the PREDICT program, which was designed to identify the most dangerous animal pathogens that had the capacity to make the jump to human populations. (Seventy percent of new disease outbreaks over the past 50 years have been zoonotic in origin.) The PREDICT project managers at the University of California, Davis sequenced the genomes of 949 novel viruses and 217 known viruses. By restoring the PREDICT program and endowing it with substantial funds, the U.S. government could ensure that vaccines are developed before outbreaks of the most dangerous novel diseases occur. And by partnering with Gavi, the Vaccine Alliance (formerly the Global Alliance for Vaccines and Immunizations), the government could incentivize the production and dissemination of vaccines for diseases that PREDICT discovers.

Gathering all of this data under one roof would enable the United States to create a PEWS to function alongside the already successful FEWS. Such a system could furnish public health officials and policymakers with information in real time. It could also enable U.S. administrators to distribute a PEWS report—like its FEWS counterpart—around the world to civil society, the news media, legislatures, and international organizations, so that other states can take appropriate measures and, when necessary, pressure their neighbors for action.


But to be effective, PEWS will also need a rapid response mechanism that springs into action when other governments lack the capacity to respond or when international organizations fail to react. USAID already manages such a mechanism for other disasters. It consists of strategically located warehouses for vehicles, water treatment equipment (useful in the event of waterborne illnesses), shelter materials (for people displaced by outbreaks), food rations, personal protective equipment, and medical supplies, including vaccines and syringes. Using these tools, USAID responders would be able to conduct rapid assessments of a disease outbreak, implement mass immunization campaigns, distribute pharmaceuticals to local health officials, and inform community leaders and the public about how diseases spread and can be slowed by behavioral changes.

For more than three decades, USAID has deployed Disaster Assistance Response Teams (DARTs) around the world to quickly put boots on the ground during a disaster. (This includes previous disease outbreaks, such as the West African Ebola epidemic of 2014.) The Bureau for Humanitarian Assistance, to which these DART teams report, has a large discretionary budget—$5.7 billion, including food aid—and a staff of more than 450 easily deployable certified disaster operations officers, as well as the legal authority to spend money quickly. From the time the first DARTs were deployed in the summer of 1989, the teams have included refugee program officers from the U.S. State Department and scientists from the Centers for Disease Control and Prevention. USAID has responded to epidemics on a smaller scale in the past; it lacks only a specific mandate in law to be responsible for large-scale pandemic detection and response.

USAID already manages a rapid response mechanism for disasters.

A certain amount of redundancy between national governments and international organizations in pandemic early warning and response systems is essential. At the international level, the World Food Program has its own famine early warning system based on a different model of risk, just as the World Health Organization has the Global Outbreak Alert and Response Network, a disease early warning system, which failed to respond to both the 2014 Ebola outbreak in West Africa and the current COVID-19 outbreak. In the case of Ebola, the local WHO office believed that announcing an outbreak would damage the regional economy. In the case of COVID-19, we do not yet know why the local WHO office did not support a rapid response, but had there been other predictive systems warning of an outbreak, the WHO might have been under more pressure to act.

The coronavirus pandemic is not a once-in-a-century event. Infectious diseases with pandemic potential will become increasingly common in the years to come. The next major outbreak will likely emerge and spread in countries that lack robust public health infrastructure and effective disease surveillance systems; these countries may possess authoritarian governments that stymie effective early reporting. Without a PEWS system, the world could very well face yet another deadly pandemic in the not-so-distant future.

  • ANDREW S. NATSIOS is an Executive Professor at the Bush School of Government and Public Service at Texas A&M University and Director of the Scowcroft Institute of International Affairs, which has run a Pandemic and Biosecurity Policy Program since 2014. He is the former Administrator of the United States Agency for International Development.
  • More By Andrew S. Natsios