When and where will infection rates from the coronavirus surge? How are people responding to quarantine measures? And how is the pandemic affecting people’s economic and social activity? The answers to these questions and the policy measures taken in response are critical to improving well-being and saving lives. The problem is that amidst the immensely complex and rapidly changing pandemic, governments most often lack the necessary data.
That, however, can change. Governments traditionally have used surveys and public administrative records to gather large amounts of information on their citizens. These sources, though useful, have shortcomings: they become available only with a considerable time lag, lack the required granularity, and provide only noisy measures of key variables. The silver lining to the pandemic’s arrival in the second decade of the 21st century is that we now have a broad array of new tools. These instruments, ranging from cell phones to social media and Google searches, can reveal critical and detailed information about what is happening to people’s health and how they are faring socially and economically virtually in real time.
Much of this big data are produced by private companies and have huge commercial value for them. So the trick for governments is to find a way to harness the data while protecting the companies’ commercial interests and, especially, citizen privacy. But these challenges can be tackled effectively, and there are numerous examples of collaborations between private companies, researchers, and governments showing the way forward.
Big Data and the Benefits of Close to Real Time Reporting
A free platform in the United States called the Opportunity Insights Economic Tracker, for example, provides up-to-date information from private companies and organizations on everything from consumer spending to small business activity and job postings. A project spearheaded by Alberto Cavallo and Roberto Rigobon at Harvard and the Massachusetts Institute of Technology (MIT) tracks inflation by collecting data from online sellers. And an initiative by MIT and the predictive analytics firm Endor, uses location data from cellphones to track the number of visitors from different neighborhoods to hospitals. This, in turn, allows to predict which neighborhoods are likely to have a surge in Covid-19 hospitalizations during the following two weeks — information key for the allocation of health resources.
Georeferenced data from cellphones can also reveal how citizens are responding to government policies and communication. A highly detailed, interactive portal created in part through our efforts uses cellphone data to reveal mobility in 22 countries of Latin America and the Caribbean, thus shedding light on the effectiveness of different lockdown measures. Cellphone data has been used to show how public statements from political leaders influences human mobility. And analysis of language on Twitter by different research teams from around the world has shed light both on public knowledge of the disease and attitudes towards public policies to combat it.
The Limitations of Big Data
Some adjustments are needed. Active users of cellphones, for example, are not representative of the entire population, and transactions collected by the banking system or online sellers are not representative of the entire economy. So methodological modifications have to be made to ensure accuracy. Some tweaking, through machine learning, is also required in interpreting text data from social media like Twitter where language is much different than that used in formal and academic writing.
Privacy protections issues are among the biggest concerns. Data from cellphones and social media accounts can reveal personal details about every aspect of people’s personal and professional lives, including where they work and live, who they socialize with, and their attitudes on sensitive topics. Several countries in Latin America have embarked on reforms that put their General Data Protection Regulations (GDPRs) in line with those of the European Union. But researchers and governments must make extra efforts. They must ensure in every instance that data is delivered and maintained in an anonymized fashion and that every step is taken to safeguard it. Protecting data is key not only to ensuring privacy but to protecting the commercial interests of companies, which collect data from cellphones, online purchases, and social media accounts to more precisely target their marketing efforts.
Accelerating Change Amidst the Coronavirus
These limitations aside, the era of Big Data promises vast improvements over the time-consuming, expensive and cumbersome ways of collecting information of the past. Those advances have only accelerated in the time of Covid-19 as researchers, businesses and increasingly governments collaborate to collect rapidly shifting data on all aspects of the pandemic. As they do and policymakers hone more informed and better responses, the benefits of expanding this kind of work will become self-evident.