COVID-19 and the Data Economy

A giant leap

At Deloitte, we are convinced that this pandemic has also fundamentally changed the way we value, manage and absorb data. COVID-19 has triggered numerous questions around collecting, sharing, using and interpreting data, with unprecedented levels of surveillance, privacy concerns, and misinformation at a global scale.

This Deloitte paper explores ways in which COVID-19 affects public perceptions and conversations concerning data value, data management, and data literacy while calling for:

• Increased deliberate investment from government and companies in the data literacy of citizens and the workforce
• The development of a clear data privacy framework, separating necessary tracking measures from improper use of personal data
• Action by the Belgian governments to complement the contact tracing centres with advice on contact apps
• A redesign of employee tracking data: direct access by employees to their own individual data can be used to boost productivity and well-being for remote workers rather than top-down monitoring practices
• The use of real-time data for data-driven economic scenarios, identifying trends to develop recovery scenarios
  

COVID-19 democratizes data interpretation skills

Throughout the outbreak, COVID-19 numbers have filled our newsfeeds. Models, statistics and projections have become critical as nations across the globe seek clarity during a period of uncertainty. There is an incredible amount of data. Making sense of the hourly updates and avoiding interpretation errors requires that people understand key data concepts.

No single news or historical event has made people aware of data interpretation and visualisation techniques more than COVID-19. There is no shortage of interesting and interactive graphs to be found online, as many organisations have put time and effort into quality COVID-19 data collection and digestible data visualisations. 

For the last months, much of the global population has been looking at data dashboards and visualisations of the daily COVID-19 numbers, trends and forecasts. Statistics communicated at press conferences or in the media have also led to a great deal of confusion, forcing policy makers to clarify and re-explain. Indeed, it has truly become a matter of life and death to question the data. Topics like sample sizes, exponential growth, curve flattening, and visualising the viral path have shaped everyday language. COVID-19 has surfaced the need for increased general data literacy and there is much to learn from this monumental time.

Exponential growth, the danger of averages, incomplete information and biased data

The growth rate of the virus has caught society by surprise, with the number of infections growing exponentially. The human brain is not wired to think in exponential terms. Growth is normally perceived in a more linear fashion.

One of the most cited statistics (for obvious reasons) is the average mortality rate from this virus, as the world is intensely interested in how dangerous COVID-19 is. Numerous mortality rate estimations are circulating around the globe, heavily varying by country. How can the same virus lead to such widely differing reported mortality rates? 

Strongly fluctuating numbers around COVID-19 infection rates have forced policy makers to explain the factors contributing to the difference we’re seeing, with perhaps the most important one coming down to simply how counting and testing influences the numbers. Hence, education about the value of random testing, which remains the most effective strategy to avoid selection bias and reduce the distortions in reported statistics, has been on the rise. Similarly, evolving mortality numbers have made the flaws of reporting averages increasingly apparent to people across the globe, helping people to understand how percentages are not evenly distributed and do not say much about individual risks.

Whether it’s about the need to adjust for population size, the complexity of making predictions in unprecedented events, or the influence of counting systems and testing on statistics (and a company or country’s reputation), people around the world are taking a crash course in basic data analysis and the pandemic has led to an acceleration in data literacy. 

COVID-19 pushes our thinking regarding the boundaries of data privacy

A number of governments have adopted digital tracking measures to track movements and trace contacts to monitor and mitigate the pandemic. Mobile applications using location data and disseminating information can often support these measures.

Tracking movements

Mobile phone operators are sharing mobile phone location data to analyse the mobility patterns of their users—including the impact of confinement measures on the intensity of contacts and the risks of contamination. Other companies have also started to share aggregate location data to help assess the effectiveness of self-isolation rules.

Clearly, movement data needs to be aggregated and anonymised, with several users being bundled together into one data package. This means that it is not possible to track individual citizens, merely the flows of movement. Some experts go so far as to say that aggregated data has no added value for monitoring the adherence to social distancing measures and that individual monitoring would be more effective in determining which individuals are disobeying lockdown orders. However, in addition to same technical hurdles, any kind of individual monitoring would probably be a disproportionate invasion of privacy.

In Belgium, the Data Against Corona Taskforce has been established to analyse data from telecom companies in order to assess the spread of the virus and identify risk areas. Agoria shared insights in March 2020, confirming that Belgians were largely remaining in their home postcode.

Telecommunications and other tech companies such as Google, Facebook and Uber have long compiled and shared aggregate location data, including for fighting pandemics. During the current COVID-19 crisis, a number of companies have started to share aggregate location data and related analytics that show where and when people are gathering.

Contact tracing

Many countries have already introduced mobile contact apps to proactively identify and warn people who have come in contact with verified infected patients. These initiatives have had some promising success. While contact monitoring is not new (GAFA or the Big Four, Google, Apple, Facebook and Amazon, have been doing it to a certain degree for years), the global public is now directly involved in the debate about the trade-off between privacy and public health. For instance, Apple and Google are working together on a contact tracing solution with the development of Bluetooth-based interoperable applications. As these functionalities would be integrated in the operating systems, this raises the question as to whether tracing capabilities would continue to be embedded and used beyond the COVID-19 crisis. 

A group of more than 100 civil society organisations published a Joint Civil Statement on the use of digital surveillance to monitor COVID-19 evolutions. The organisation asked that ‘the use of digital technologies to track and monitor individuals and populations is carried out strictly in line with human rights’, setting forth eight conditions to be met for digital surveillance (measures and technologies must be time-bound, proportionate, fit for purpose, lawful, avoid discrimination, properly safeguarded and include means for participation of relevant stakeholders).

In Europe, the European Commission published (non-binding) guidance to ensure that applications supporting the fight against COVID-19 comply with EU privacy and personal data protection regulation. For instance, Bluetooth communications between devices to determine proximity is recommended rather than geolocation data as this avoids the possibility of tracking. The Commission has also adopted a toolbox for the use of mobile applications for contact tracing and warning. The toolbox sets forth specifications to ensure that apps are installed voluntarily, approved by Health Authorities, interoperable, secure, and that data is shared with relevant epidemiological public bodies.

Currently, Belgium is considering introducing mobile contact apps. The hesitation can be largely attributed to beliefs that the current apps are not yet fit-for-specific-purpose leaving many open questions about due process, usage domains, and most importantly, individual privacy. Still, this could be a huge missed opportunity for enabling economic and societal recovery. Certain apps could function as a key complementary component to the existing call centres. For example, the Community of Madrid provides an app to enable citizens to carry out a self-assessment without ringing the city’s overburdened call centres.

The Belgian governments should evaluate and share insights on the relevancy of the different apps. These measures would be beneficial for a number of reasons:

• Increase awareness among citizens of the value of data and data privacy.
• Encourage app developers to step up their efforts in these domains.
• Encourage device producers to optimise their contact tracing technologies, and to specify limits to the applications of such insights.
• Set standards for solutions, hereby ensuring effectiveness and adequacy.  
  

COVID-19 raises fundamental questions about employee data rights, but opens promising paths to self-steering

Before the pandemic outbreak, employers were heavily investing in collecting and analysing data to improve and personalise the employee experience. When it comes to remote work in these unprecedented times, with management having significantly less face time with their teams, companies may want to adjust to a new normal and re-shift the focus of their data collection and analytical efforts from experience to productivity monitoring.

Indeed, the continued spread of COVID-19 is forcing millions of people to work from home, using the employers’ IT infrastructure and tools. As a large proportion of workers shift to this new reality, organisations become increasingly concerned about their workforce staying as productive as possible. 

As a result, some employers could start to analyse (digital) data traces to keep tabs of and monitor their employees. Many digital tools used in the workplace generate large amounts of data on the way employees work and can give employers unprecedented insights into who in their organisation may be floundering. 

Activity or productivity monitoring—a set of practices enabled by digital tools to monitor and track employee activity during work hours—may include URL-, app-, document- and time tracking as well as analysing emails and social media messages. Similarly, although being essential nowadays to effectively connect with others, tools that allow for instant communication, such as the messaging services of Skype and Microsoft Teams, offer monitoring features for employers. The videoconferencing app Zoom for instance, which has recently seen a surge in popularity, has a usage report that lets administrators look at the meetings, participants, and meeting minutes for each user.

It is clear how the exploding volume of highly sensitive data scooped up from individual employees is and will increasingly raise privacy concerns and adds a new dimension to the relationship between workers and their employers. The impact of employee monitoring and micromanaging can be fundamentally negative and actually do more harm than good. Sneaking around and looking at logs sends the message that a company does not trust its employees to take responsibility. It negatively affects the employer-employee relationship, thereby potentially triggering the opposite result than intended.

At the end of the day, employee productivity, performance, and extra-mile behaviour all boil down to the company culture you have established, the affective commitment employees have built towards the organisation, and the passion they have for the work they do. The real challenge for companies will be to navigate the needs of the business while empowering and trusting employees to work at a reasonable pace amid increased stress.

Self-steering through data

At the same time, the wealth of data collected by these apps and tools provide an excellent opportunity to empower the workforce. Giving employees access to their own data can be a great way to help them make changes in their lives, become more productive or disconnect from work. 

For instance, some apps build a timeline of users’ daily work activities by recording all active web and desktop activity. However, only the user can access the data and choose which data to share with the organisation. By helping people to build awareness around how they spend their time, these apps can trigger positive changes to often deeply established habits and unproductive routines (e.g. logging out of anything that doesn’t require their immediate attention, or structuring their day to become more present on one task at a time).

Therefore, giving employees access to their own data is a more effective way to increase productivity and makes the use of data beneficial to both the organisation and the employee. Moreover, the delineation of the target use of data greatly promotes transparency and accountability. 

COVID-19 strategies augmented through the use of real-time economic data and models

Few will disagree that COVID-19 is a ‘black swan’, an outlier event that catches the world unprepared. In such times, the limitations of historical economic data become painfully apparent. There is a significant lag on nearly all the types of data on which our economic models are built (employment data, consumer confidence, production data,…). Moreover, the current lockdown completely changes consumer behaviour, making obsolete the basket of products and services that is at the basis of inflation calculations.

In unprecedented times, what is needed more than data-driven forecasts are data-driven scenarios. Economists can look for new sources of data (both structured and unstructured) and incorporate more recent, real-time data to interpret the most recent economic trends activities and develop recovery scenarios.

Deloitte Canada has started working on such an initiative by constructing a COVID-19 real-time dashboard. The dashboard creatively combines publicly available data on pedestrians in metropolitan areas, commercial traffic activity, air quality indexes, and more to provide insights into where we currently stand in the crisis, and potential paths to recovery. Another such example is our State of the Consumer Dashboard.

Conclusions

We have reached four major conclusions in this paper:

1. People around the globe have learned to more critically interpret data and question its veracity and origin. It has forced governments, industries and companies to be more attentive and work hard to increase people’s confidence in data, thereby boosting data literacy at scale. An unprecedented—and overdue—acceleration of data awareness and skills.
2. For the first time, the debate between privacy and data access has turned into a debate about life and death. The stakes are high and very visible. Technology and contact apps are not the sole solution, but pretending they are not part of the solution is shortsighted. We call upon all societal stakeholders to use the COVID-19 crisis to advance Belgium’s position in the data space and deepen and widen the data debate in our country.
3. Data tracking can improve employee performance. But it has to be done by empowering the employees to analyse their own data, not by micromanaging log files.
4. Our world is highly connected. The current data lags in economic modelling are unacceptable to manage the risks of this connectedness. Real-time data and data-driven scenarios should become the future norm for economic modelling.

This global pandemic contains critical lessons about the way we value, manage and absorb data. Our governments, companies and citizens should use this crisis to accelerate the evolution toward a fair data economy.