Hearing from all voices during COVID-19

By Danielle Goldfarb

RIWI technology emerged to track public responses to pandemics across a broad set of the population in real-time. It can be used during COVID-19 to give us vital and otherwise unavailable near real-time information on infection prevalence, public responses to physical distancing directives, and economic and mental health impacts in all or any countries of the world.

Reliable, timely data is emerging as one of the critical factors in our collective ability to fight the COVID-19 pandemic. But in most countries, we lack reliable evidence on how many people have been infected, or continue to be infected. We also don’t have reliable and timely data on how citizens are actually responding to public policy directives during this time. Are citizens physically distancing themselves as recommended by public health authorities? 

Most typical methods of data collection on people’s opinions and conduct – such as traditional surveys and panels – rely on hearing from the most engaged voices. This is also true for news and social media. Vastly more people are excluded rather than included in these sources of information. 

But during an epidemic, hearing from the widest set of voices possible is not just a reflection of true democracy, it becomes vital to survival. As Ebola and other epidemics teach us, if we don’t hear from all sources, hidden data can unexpectedly lead to new outbreaks. We could miss pockets of communities where public health messaging has not penetrated or is not being heeded with sufficient intensity. We could fail to ensure the appropriate dose of economic relief at the appropriate time to populations that haven’t been heard – such as young people, for example, who don’t typically answer traditional surveys. This risks exacerbating a public health crisis when people continue to go out to work when they have symptoms. Many may not feel they have a choice because they are dependent on their paychecks. Alternatively, they may not be aware of the gravity of the situation. In addition to this behavior deepening a public health crisis, it can also induce a deep, long-lasting economic crisis. 

An additional data problem for the effectiveness of government and industry policies is the time lag inherent in the transmission of data. Confirmed COVID-19 cases reflect the state of infectivity days, sometimes weeks prior to their being reported officially. Official economic data are not reported until weeks and often months after they occur. 

In addition, some of the issues we are currently dealing with require answers to questions for which we have little or no reliable, recent data. These are questions such as: are all citizens actually following public health advice? How many people are – and aren’t – able to work from home? How many have paid sick leave? How many people aren’t going to be able to pay their bills? What is the impact on temporary or gig economy workers? How many gig/temporary economy workers are there in each country? (The current crisis brings into sharp relief the reason why traditional jobs survey data that ask, “do you have a job?” are inadequate.)

If public health officials could find out the truth in real time about who is and is not following physical distancing advice, they could target public health messaging more effectively, and, thus, contain the outbreak. If they could find out where incidence rates are growing sharply amongst the broadest set of the population, they could more effectively target resources to contain the outbreak. Economic policymakers could more effectively lessen the impact of the economic fallout if they had reliable, timely information on the current and future needs of the broadest possible set of the population. 

A key reason broad-based, timely sentiment data matters so much now is that public health requirements and economic requirements collide in this outbreak. Because we are all interconnected, what one person does or fails to do affects that person’s whole community.

If policy makers only hear from the most engaged citizens they will fail to meet the needs of those, for instance, who can’t afford not to work without appropriate economic relief, or those who do not appreciate the health risks of not maintaining physical distance from potentially infectious others. Policy makers simply will have no way to know who these people are. 

RIWI’s technology was created to try and solve these data gaps and problems – everywhere in the world. The technology had its origins in government-commissioned pandemic surveillance during the H1N1 outbreak, and has actively tracked the public’s response in real time among online populations during H7N9, Ebola, Zika. The technology is now actively used around the world to measure behaviors and opinions not just among engaged, highly vocal populations but also among those who are typically disengaged and who rarely respond to surveys. It has been used to measure labour market and consumer activity that is not otherwise captured by traditional data sources, which includes the global digital gig economy. And it has been used to predict headline economic data in China, the United States, and to forecast referendum and election results around the world, including the surprise victory of President Trump in 2016. 

During COVID-19, we have been tracking daily public sentiment in China and the US starting in January when the outbreak became public. As COVID-19 has expanded across the world, we’ve been measuring daily public sentiment in most countries that  have been hit by the outbreak. What is important to note is that over half of RIWI respondents have not taken any survey in the past month. We’ve been able to reach previously unheard voices. 

As of March 23, here are some key narratives we’re seeing emerge from our data, representative of the online population in each country under our evaluation. 

  1. While most people in Canada and the US are now practising physical distancing, including working from home, there are still non-trivial numbers who are only doing it some of the time or not at all. 
  2. The majority of Canadians, Americans, and Britons, South Asians, Nigerians, and Chinese express confidence in their public health officials. That confidence is growing in countries such as China, the US, and Canada as officials in each country are taking aggressive physical distancing measures. 
  3. Trust is essential in ensuring that people take proper protective actions. While there are some that still don’t appreciate the seriousness of this pandemic, only a very small percentage in each country does not feel any actions should be taken to prevent the virus.
  4. There is a massive shutdown of typical economic activity and no question there will be a significantly negative impact to the global economy. What is unclear is how severe, widespread, and long-lasting the impact will be. That depends on people’s levels of confidence and economic resilience. While those of us in Canada, the US, and Europe are now at the beginning of our physical distancing measures, China is at a very different stage and we are seeing relative economic resilience so far in RIWI China data, with levels of community, consumer, and business optimism nearing levels seen before the outbreak became public and despite very negative official data coming out of China during the shutdown period. We are tracking economic sentiment on a daily basis to determine how deep the economic fallout will be in other parts of the world.  

We continue to monitor broad-based sentiment on a daily basis on a range of critical public health and economic questions and across countries in all regions of the world — including sub-national jurisdictions.

One key gap is that no countries have reliable data on the prevalence of a virus in a representative random sample of the general population1. RIWI technology can do this and repeat it at regular intervals to estimate the incidence of new infections in every country of the world. We are seeking partners to undertake this critical work.

What else should we ask that would help decision makers during this global crisis? In addition to measuring COVID-19 prevalence and prevalence of physical distancing, there are a range of mental health and economic measures we could track on a daily basis anywhere in the world. Please reach out at with suggestions and comments of what we should ask here in North America or in any country around the world right now. We plan to share more detailed findings regularly as events evolve.

Watch RIWI CEO Neil Seeman’s upcoming virtual lecture hosted by University of Toronto’s Dalla Lana School of Public Health How is the public responding to COVID-19? on Thursday, March 26 at 12:00PM EST here.

  1. John P.A. Ionannidis, March 22, 2020, A fiasco in the making: We are making decisions without reliable data.