The coronavirus pandemic has dramatically changed the way governments everywhere, and at all levels, deliver services to citizens. There has been a widespread embrace of technology, new and old, even among the most reluctant. Medical services, education and work – for many of us – have moved online, and despite initial hiccups, seem to be serving us well. Technology has enabled us to stay socially close while remaining physically distant. We connect with our colleagues and friends on the Internet and critical public services are accessed through online portals. Coughs and colds are cybernetically diagnosed, and groceries are filled into virtual baskets instead of physical ones.
Technological advancements made a significant dent in helping the world manage the outbreak. They enabled scientists to track and identify the disease faster than ever, to share information and collaborate on research from all corners of the globe. Embedding sensing and data-gathering technologies in dense population centres (i.e., cities) would add even more value, allowing for decision-making based on accurate information that efficiently captures the incidence and spread of the disease. This intersection of data and technology has made a crucial impact by providing opportunities to make informed policy decisions, and has helped to keep the public abreast of the changing nature of the pandemic.
However, access to data doesn’t necessarily mean that it will be taken into consideration or that decision makers will know how to best analyze or use it for their purposes. This dynamic has played out around the world as the pandemic has evolved. While the emphasis in public policy in the last few years has largely been on increasing technological interventions to allow for greater data collection, there has been a marked capability gap in terms of how to generate and demonstrate value from said data to create effective public policy in many areas.
Part of the reason for this is a lack of training or understanding of best practices in government data use, but another is a challenge in human nature. The world is increasingly complex and interconnected; attempting to understand all of its interdependent parts and make decisions using data can be overwhelming, especially during a pandemic. As Vikram Mansharamani of Harvard University noted in his book Think for Yourself, we have understandably developed a tendency of outsourcing our critical thinking to technology or experts in the modern age because it is simply easier to do so than to individually make sense of the complex problems we are facing.
Technology and data have their role to play, but humans must not get lost in the mix. People, whether they realize it or not, have the unique ability to identify the blindspots in data that the tools they are using cannot. They have the ability to ensure public policy interventions are executed through the lens of a more holistic vision of society and the problems at hand. It is just a matter of knowing how to leverage the technology and data at their disposal and using their broad understanding of the world to complete the puzzle. A new decision-making framework is necessary to find the right balance between people, technology, and data to create positive outcomes in any future large-scale crisis, dispute, or consequential public policy arena. We call this nexus of human intervention, smart technology, and data people-driven smart policy (PDSP).
In order to understand PDSP in the context of pandemic response, it is necessary to break down its three component parts: smart technology, data, and people. Each of these play an integral role in the success of PDSP. This framework challenges the over-reliance on technology and data in public policy decision making, and puts people back in the driver’s seat.
Smart technologies in pandemic response
Just as some cultures dealt with the current pandemic more efficiently because they had already adopted healthy habits, such as mask-wearing, cities with good technology and data habits found that those capabilities can prove helpful in dealing with the current situation. We believe that there are five core smart technologies that can impact pandemic responses:
- The Internet of Things (IoT) encompasses the use of sensors and internet-connected devices to capture data in real time. These devices can be embedded in street infrastructure, buildings, cell-phone towers, and other public spaces. IoT is used to monitor air quality, traffic patterns, and even flood levels. Singapore and South Korea leveraged their experiences using traffic management technology and data to understand quarantine compliance and improve contact tracing.
- Mobile Communications Technology (MCT) broadly refers to the technology, apps, systems and networks built to transmit information through all digital channels. MCT enables city and state governments to improve emergency communications in disaster scenarios or expand WIFI access to underserved areas of the city. Barcelona’s app for seniors has been particularly useful in keeping isolated seniors connected during the crisis. It’s a striking example of a modest existing project that quickly became a lifeline for a vulnerable population.
- Artificial Intelligence (AI) is the practice of training computers and machines to imitate intelligent human behaviors. AI is used to help autonomous vehicles analyse video images to identify pedestrians and traffic lights, to support automated chatbots on government websites, and so on. AI is a fundamental part of advanced medical research, and can also be used by local public health agencies to accelerate epidemiological analyses by public health authorities, even to predict the evolution and spread of the outbreak.
- Blockchain may be best known as the underlying technology behind Bitcoin, a virtual currency. It also allows governments to conduct transactions securely, in a way that is auditable and verifiable. Blockchain has helped companies to exchange secure shipping information, and monitor food production and supply chains to ensure compliance with food safety regulations. In the context of a pandemic, however, it could be used to facilitate contact tracing while still protecting user privacy. In the emerging (and controversial, both from political and public health standpoints) conversation on COVID passports — digital “certificates” that show the bearer either does not currently carry the virus or has developed COVID antibodies and thus probable immunity — blockchain could ensure the system is trustworthy and secure (via encryption and decentralized storage), and the passports verifiable (via unique identifiers assigned to owners of the ledgers).
- Digital Communications, both public and private, play a significant role in disseminating information essential to coordinating a whole-of-society policy response. Government websites, listservs and email alerts are essential for communicating and interacting with citizens. Meanwhile, the algorithms of privately-owned social media platforms like Facebook and Twitter choose what information users see first, making them key actors in managing public communications. Social media monitoring can also help identify contagion hotspots and reinforce shelter-in-place and other communication goals that are critical during a pandemic.
Several countries had these capabilities in place before the pandemic and have used them to varying degrees of success. However, many others are now grappling with these core smart technologies, but without much experience using them for public purpose, the learning curve is steep.
Closing the gap in digital capabilities is now a new priority for both governments and the people they serve. A widespread understanding that a robust technology ecosystem can empower more resilient communities, is one of the important lessons of the pandemic. These technological capabilities can improve access to data and information sharing, which empower governments to take data-informed approaches in times of crisis, but how is data used in pandemic response?
The role of data in disease outbreaks
A major outbreak of Cholera swept through the Soho neighbourhood of London in the mid-1850s, killing 127 locals in the first few days and scaring a majority of the local population away. A physician named John Snow went door to door, interviewing local residents in the area about their symptoms. He plotted his results on a map and was able to conclude that the Broad Street water pump was the most likely source of the outbreak. Snow’s work earned him the title of the “Father of Epidemiology”. To this day, epidemiologists still use the same grassroots tactics and analytics techniques Snow employed, albeit more sophisticated.
Today, epidemiologists gather information more quickly using a mix of technological and human resources, with an aim towards collecting data in near real-time. However, these near real-time capabilities are not widespread; in fact, fewer than one in five countries globally received advanced scores on the 2019 Global Health Security Index for early detection and reporting capabilities. So, how can smart technologies enable or support real-time data collection and analytics for epidemiological response?
To answer this, it is important to first understand that epidemiological data collection uses a bottom up approach, starting at the local level in labs, and reporting up through the higher rungs of government. The early stages of this process are the most critical: the sooner investigators can get into the field to study emerging threats, the sooner mitigation measures can be taken. With technology such as IoT sensors maintaining a restless eye on the city’s movements, marrying big data with “shoe leather epidemiology” (the door-to-door method used by Snow) can go a long way in helping public health workers identify these invisible threats earlier.
We can already see the benefits of data analytics in tracking and stopping the spread of disease. Take San Diego’s response in 2017 to a Hepatitis A (Hep A) outbreak among its homeless population. Knowing that the disease spread through human fecal contact, healthcare professionals mapped the city’s public restrooms, instituted new guidelines for cleaning them, and deployed mobile hand-washing stations in areas considered to be “high-risk”, stopping the disease in its tracks.
New York City’s response to a Legionnaires’ outbreak illustrates how technology and data can detect epidemics before healthcare professionals are even aware of an emerging threat. A software program helped “identify a small cluster of Legionnaires’ disease before health care providers noticed an increase in cases,” triggering an early and effective public health intervention. NYC prevented a major outbreak thanks to having a robust technology ecosystem in place.
For novel diseases such as COVID-19 for which there is limited data, existing datasets may not be so useful, but one can still use the information at hand to identify hidden threats. For example, artificial intelligence has been used by companies like BlueDot, a Canadian outbreak risk software startup, to scrape data from news reports, social media and public government documents globally to identify emerging disease outbreaks. BlueDot actually flagged the COVID-19 outbreak from a news report in China about a week before WHO or the CDC were able to catch it. Governments can leverage these capabilities as part of their technology ecosystem to monitor local conditions and take proactive actions to stop diseases from spreading.
The faster data can be collected and analyzed at the outset of a potential outbreak, the better we are able to respond to emerging pandemic diseases. Public health experts are doing everything they can to move to a more proactive, rather than reactive, approach to pandemic prevention and response, and big data and analytics are at the heart of those efforts.
People-driven smart policy: A holistic approach to data-driven decision making
Early on in the pandemic, Singapore was lauded as a champion in managing the coronavirus. Thanks to rigorous testing and using technology to bolster efforts in contact tracing , it was able to identify and strictly quarantine those that had the disease or were exposed to it, thereby mitigating its spread. But there was a gap: the government failed to extend these stringent standards to include migrant workers, many of whom live in cramped dormitories. COVID-19 flared up in that community. While Singapore had both smart technology and the data insights that go with it, its blind spot led the country’s number of infected to rise from 266 cases on March 17th to close to 50,000 today today. Migrants account for more than 90% of these infections.
In the US, a new study recently revealed major gaps in biosurveillance data in low-income zip codes where viral outbreaks often take the greatest toll. These are also the same places where COVID-19 has ravaged poor communities from the beginning, and has sustained throughout the pandemic thus far. Politics aside, even if the data and red flags raised by public health leaders in the US had been taken seriously and to the fullest extent, these major blindspots may very well have led to a loss of control of the virus in poor communities similar to the situation in Singapore.
While there is an inherent focus on technological solutions and data gathering in public policy these days, it is important to get an understanding of the inherent risks of this practice. Beyond the data blindspots highlighted above, there are several other real-world examples unrelated to the pandemic that demonstrate the dangers of leaning on data and technology too heavily.
For example, in the autonomous vehicle space, a safety backup driver in Arizona failed to prevent a self-driving vehicle from hitting a cyclist because they took their eyes off the road, relying too heavily on the vehicle’s sensors to identify pedestrians and other roadblocks. In policing, it is widely known that facial recognition technology performed poorly in accurately identifying people of color, yet police forces have relied on it to identify “criminals” for years, which has already led to wrongful arrests of innocent minorities. Artificial intelligence and algorithms have been known to help simultaneously reduce or exacerbate bias in human decision making, and which direction they go often depends on the training (or lack thereof) humans get to identify such biases and the principles of computer science that are used to build them. There are many instances where data and technology have helped to create better outcomes for governments and people, but we must not ignore these other examples that highlight the hidden dangers of leaning too heavily on them for decision making.
Clearly, using advanced technologies and data analytics is not a guaranteed solution to handle a pandemic, or any public sector problem, altogether. These tools, however valuable, can only go so far in understanding the complexities of the real world and the humans that inhabit it. Data and technology can inform policy, but inevitably, government officials, with their cultural and emotional intelligence, must use their knowledge to round out the insights that tech and data provide in order to devise best outcomes for citizens.
All this being said, we remain techno-optimists. We still believe strongly that technology and data, as illustrated earlier in this piece, add significant value to public sector policy making and encourage governments to embrace these tools. Of course, there is varied acceptance for monitoring through smart technologies globally, even in a pandemic. Their use is dependent on how much citizens trust their government, or are forced to comply. Cities who have successfully navigated these compromises by embracing smart technologies already have experience in determining what is possible and acceptable. Those that lack the experience have far more hurdles to jump through in terms of earning the trust of the citizenry.
What is clear from this research is that, in our quest to become tech and data driven, we have lost the critical human aspect in decision making. What we need is to find balance between technology, data and people so that we maximize the utility of all three parts of the equation to devise the best outcomes for citizens. What we need is a new framework that helps guide decision makers through the process of analyzing data so that we do not grant blind faith to the tools at our disposal.
This in essence is what we refer to as people-driven smart policy (PDSP), which we define as “a public policy approach that maximizes the utility of human social, cultural, emotional, and intellectual capacity, coupled with technological advances to fill in critical data gaps and drive inclusive, evidence-informed interventions.” In other words, PDSP prevents an overreliance on technology and data when making critical policy decisions and puts people back in the proverbial driver’s seat. Moving forward, we are conducting further research to help public sector leaders govern with data more effectively and are building a decision-making framework that helps policymakers through crises or moments of uncertainty.
The framework will be guided by a series of questions that help public sector leaders ensure they are not skipping critical steps or missing important data that can influence outcomes and to think critically about how to fill in critical information gaps. For example: For what populations do we lack information? What other agencies or entities might have the data we need? What are the limits of the data we have at our disposal? Which communities do we have data on and which do we not?
The idea is to ensure that the modern-day John Snows of the world are able to bring their critical thinking skills to the forefront. In Snow’s case, maps played a critical role in helping him identify the water pump that was the most likely source of the Cholera outbreak, but his ability to understand human consumption and behavior and his knowledge and theories of waterborne illnesses were critical in helping narrow his search. Humans have to fill the gap where technology fails and data is missing, or simply falls short of painting a complete picture of the world.
As we’ve seen in the case of COVID-19, having the most advanced technology and data analytics capabilities in the world does not result in the most effective public policy interventions. These tools in and of themselves will not bring down pandemic diseases; they are only as powerful as the people that choose to use them. Our hope is to be able to build a framework that puts informed human judgment at the centres instead – one that can adapt to any critical public policy decision where people’s lives or the wellbeing of the most vulnerable are at stake. We believe that if people-driven smart policy in its truest form was enacted, hundreds of thousands of lives would have been saved in our present crisis. We hope that by implementing PDSP, governments will be ready to save countless lives in the next one.
Rashi Khilnani is a public policy, strategy and communications consultant with a passion for adding public value through technological innovation and public-private partnerships. Rashi is a techno-optimist and has experience as an advisor to several startups in Canada and overseas. She is also a National Magazine Award nominated journalist, a former Media Fellow with the Asia Pacific Foundation of Canada, and a bursary recipient of the Shastri Indo-Canadian Institute. Rashi has a BA in Political Science and International Development Studies from McGill University and a MPA from the Harvard Kennedy School.
Matthew Leger is a policy analyst and researcher with experience in the fields of local and state government, emerging technologies, private & public sector digital transformation, workforce development, and homeland security. He currently serves as a Policy Research Analyst at the Ash Center for Democratic Governance and Innovation at Harvard Kennedy School and as the Director of Strategy at CONTRACE Public Health Corps in Washington DC. He earned his bachelor’s in public policy in 2016 and master’s degree in public administration in 2017, both from the Nelson A Rockefeller College of Public Affairs and Policy in Albany, NY.
Jean-François Barsoum is a researcher at IBM, where his focus centres on understanding and communicating the societal and environmental impacts of technology and smart cities. He is a director on the boards of the Climate Project Canada and the Canadian Water Network, a startup mentor at Techstars Montreal AI, the chair of the disruptive technology committee of the Montreal Chamber of Commerce, and a member of the committee overseeing the application of the Quebec Policy on Sustainable Mobility.
Edited by: Derrick Flakoll (in Canadian English)
Photo by: Christina @woctintechchat.com