Perceived COVID‐19 health and job risks faced by digital platform drivers and measures in place to protect them: A qualitative study

Abstract Introduction As they deliver food, packages, and people across cities, digital platform drivers (gig workers) are in a key position to become infected with COVID‐19 and transmit it to many others. The aim of this study is to identify perceived COVID‐19 exposure and job risks faced by workers and document the measures in place to protect their health, and how workers responded to these measures. Methods In 2020–2021, in‐depth interviews were conducted in Ontario, Canada, with 33 digital platform drivers and managers across nine platforms that delivered food, packages, or people. Interviews focused on perceived COVID‐19 risks and mitigation strategies. Audio recordings were transcribed verbatim and uploaded to NVivo software for coding by varied dual pairs of researchers. A Stakeholder Advisory Committee played an instrumental role in the study. Results As self‐employed workers were without the protection of employment and occupational health standards, platform workers absorbed most of the occupational risks related to COVID‐19. Despite safety measures (e.g., contactless delivery) and financial support for COVID‐19 illnesses introduced by platform companies, perceived COVID‐19 risks remained high because of platform‐related work pressures, including rating systems. We identify five key COVID‐19 related risks faced by the digital platform drivers. Conclusion We situate platform drivers within the broad context of precarious employment and recommend organizational‐ and government‐level interventions to prevent digital platform worker COVID‐19 risks and to assist workers ill with COVID‐19. Measures to protect the health of platform workers would benefit public health aims by reducing transmission by drivers to families, customers, and consequently, the greater population.


| INTRODUCTION
Digital platform drivers, such as those working for Uber Eats, Amazon Flex, and Lyft, were busier than ever during the COVID-19 pandemic as the public attempted to avoid infection by ordering take-away food, shopping online, and taking ride-hails rather than public transportation. 1 Given airborne transmission of the virus and the nature of their work, this placed platform drivers in a position to become exposed to other people, infected with COVID-19, and to transmit it to others as they moved food, packages, and people from one location to another. 2 Although digital platform drivers operate between where people live (e.g., homes, care facilities) and wider communities, occupational and public health guidelines to protect these drivers from COVID-19 exposure and to mitigate their role in disease transmission have been limited. Given that people act based on their understandings, 3 we examined how the drivers understood and responded to perceived COVID-19 risks at work. The aim of our Canadian study was twofold: 1) to identify, from the perspective of platform drivers, the perceived COVID-19 exposure and job risks faced, and 2) to document the measures in place to protect the health of these workers and how workers responded to these measures.
The global digital platform economy continues to grow quickly. Mastercard and Kaiser 4 describe an annual 17% growth rate of digital platforms that allow freelancers to connect with individuals or businesses for short-term services or asset-sharing and an expected gross volume of $455B USD by 2023. Of this, 88% of revenue is based on transportation-based services such as ride-hail and food or package delivery. During the pandemic, e-commerce (i.e., requiring parcel delivery) made record gains; for instance, more than doubling in Canada over 2019-2020 5 and growing by 12% in Europe. 6 It is estimated that over 28M people in the EU work through digital platforms. 7 In Canada, the number of digital platform workers is unclear; however, an administrative database study of gig work (defined as unincorporated self-employed workers whose future business activity is uncertain, minor, or occasional) determined that gig workers comprised 8.2% of all workers in 2016. 8 Digital platform work is precarious, low-wage, and insecure. 9 Precarious work has grown in recent decades and is related to the transformation of employment conditions as a result of technological advances and the increased mobility of capital and workers. 10 It is generally characterized as involving contract renewal uncertainty, income inadequacy or volatility, and a lack of worker rights and protections and is associated with ill health. [11][12][13] While some precariously-employed workers have managers and workplaces, an added complexity facing platform workers is that they tend to be classified as self-employed. 14 This is despite the platforms having considerable and "employer-like" control over the activities of these workers, including standardization of services and pay rates. 15 The self-employment status of digital platform workers has been steadily contested in the law courts across international jurisdictions 16 and is the subject of a proposed European standard that, if passed, would provide a list of control criteria to determine whether the platform is an "employer." 7 Studies have identified ways in which digital platform gigs pose risk to workers' physical and mental health. Bartel et al. 17 found that ride-hail drivers, via encouraging nudges from their platform, drove for many hours without exiting the vehicle and experienced related back, leg, and foot pain. Due to limited bathroom access, drivers limited liquid intake to the point of dehydration. Their high amounts of sedentary behavior were often accompanied by unhealthy eating and weight gain. The study also found that mental health was affected by constant surveillance and the need to maintain adequate stores on platform rating systems. Similarly, others have noted that, while platforms promise flexibility and autonomy, the work actually involves unpredictability, work intensification, and financial hardship. 18,19 Nielsen et al. 20 found that safety systems for platform workers were weak or absent.
Platform workers can be categorized into two main groups: webbased (services performed remotely via the Internet) and locationbased (work in a specific geographical area). 21 In our study, we were concerned with platform workers' interactions with the public and so focused on location-based workers; specifically, those who delivered people, food, or packages. Because digital platform workers are mostly classified as self-employed, they are legally responsible for their own health and safety, as they are considered as controlling their work and working conditions. As such, it is up to them, rather than a workplace manager, to both identify workplace health risks and find ways to avoid them. Unlike regular employers, platform companies (which tend to refer to themselves as "tech companies") usually do not pay into social security systems, such as workers' compensation, employment insurance, and pension plans. As well, platform companies have mostly not been bound by labor laws that, for instance, require them to pay minimum wage, overtime, and provide holiday pay for workers and offer parental leave. 22 This lack of social security protection leaves platform workers vulnerable and downloads the risks of this study onto workers themselves. For instance, studies find that digital platform drivers often earn below the minimum wage, when costs such as gas, car maintenance, and waiting periods are factored in. 23,24 Studies of platform workers find that, although platforms emphasize the flexibility of workers to choose when they work, this characterization underplays the active role that digital platform companies play in standardizing services and monitoring workers. 15 Working times are mostly dictated by consumer demand and platform incentive payments and work is managed by the platform company to a significant degree. 21 In 2021, the European Commission 7 proposed a directive on improving working conditions in platform work. The directive proposes a list of control criteria to determine whether the platform is an "employer" with the aim of ensuring that people working through digital labor platforms are granted the legal employment status that corresponds to their actual work arrangements. The directive also calls for more transparency in algorithmic management. While the directive is subject to lobbying and change before final implementation, these initial proposals provide international policy leadership.
Interestingly, during the 2020-2021 waves of the COVID-19 pandemic, and despite platform companies' practices of classifying workers as independent contractors and making claims of being only a "technology company" or "digital marketplace," 25 some platform companies engaged in further employer-like roles by protecting worker health and safety through providing protective supplies to workers, such as hand sanitizer and face masks and offering some income compensation for workers when they were ill with COVID-19. 19,26 During the pandemic, governments in many countries changed their stance toward self-employed workers. 27 For the first time, they became eligible for emergency government income support benefits, such as the Coronavirus Aid, Relief, and Economic Security (CARE) Act in the USA 28 and, in Canada, the Canadian Emergency Response Benefit. 29 In Canada, in early 2022, legislation was passed in Ontario that established a so-called minimum wage for platform drivers. 30 Although the Ontario government appeared to be engaging with strategies that provide some worker protection to platform drivers, the new minimum wage has been criticized as it applies only to "active" times (i.e., passengers in the car, delivery underway) and so does not cover the whole work day, including waiting times. 31 Ultimately, the continued status of workers as self-employed means that they remain excluded from occupational health and safety standards and most employment standards.
COVID-19 has added a new layer of risk to the work of platform workers. Essentially, digital platform drivers are frontline, low-wage, and self-employed workers who, to date have had minimal occupational health guidance and support in the context of COVID-19. 32,33

| METHODS
A qualitative approach was used to address our research question regarding the nature of occupational health risks faced by platform drivers during COVID-19 as core categories of analytic foci regarding platform driver COVID-19 risk are still relatively unknown and undeveloped. The goal of qualitative research is not to identify distributions and typical characteristics for statistical generalization but to gain insight into social processes and complexities of human interaction. 34 In this case, we were particularly interested in risk perceptions and decision-making rationales related to perceived COVID-19 risk management. While a recent mixed-method study shed light on health and safety risk perceptions of platform drivers, 35 their responses were limited to preset categories of anticipated response fields. By using in-depth interviews with open-ended questions, we were able to capture a wide range of risk perceptions of digital platform workers and managers in relation to key contexts of work pressures and platform occupational health interventions.
Our Ontario, Canada, based study used situational analysis, 36 which considers events and behaviors in organizational and policy contexts, to capture the contextualized work and health perceptions of digital platform drivers. For the purpose of our study, we focused on digital platforms that delivered food (restaurant meal delivery), packages (parcel delivery; also, pharmacy/grocery delivery by workers who also shopped for the items as a part of the service), and people (ride-hails), as these were very active during the pandemic. We use the term "drivers" in relation to all platform workers in this study, although some platform workers delivering food used a bicycle.
With our purposive sampling strategy, 37

| RESULTS
We conducted in-depth interviews with 30 digital platform drivers who worked across nine different platforms (see Table 1) and with three digital platform managers (we attempted to recruit 10 managers but only three responded to outreach). Interviews took place between September 2020 and February 2021. This was during Wave 2 of the pandemic in Canada and before the appearance of coronavirus variants of concern.
Most workers who participated (80%) were male; approximately half (52%) were members of racial minority groups; and the average age was 31 (range 21-54). Digital platform work was the primary source of employment income for 57% of workers in our sample, and the sole source for 27%. These rates are similar to those reported in a MACEACHEN ET AL. | 3 recent survey of New York City digital platform delivery drivers. 24 On average, the drivers in our study engaged in digital platform work for 23 hours per week and earned an average of $23,790 annually from this work. Many of our participants worked across multiple apps.
Publicly available and participant-supplied documentary data gathered included website information about the digital platform's COVID-19-related safety precautions and support to drivers. As well, participants provided screenshots from their smart phones that showed the digital platform's COVID-19 advice and supports.
Below, we detail our findings regarding the five key COVID-19 related risks faced by the digital platform drivers and how they were navigated. We first describe these worker's constant encounters with customers in the course of their work, including delivering to people in quarantine and isolation. Next, we consider safety measures, including contactless delivery, taken by the platform companies to reduce worker's exposure and ways that these had limited effectiveness. Third, we describe workers' challenge of managing customers refusing to wear masks. The fourth risk addresses time-related pressures built into platform's algorithms and enforced by the platform's rating systems, which workers described as prompting them to take health risks. Finally, we describe the loss of income risk to workers of self-reporting COVID-19 symptoms to their platforms. Table 2 lays out the risks by platform type.

| Constant customer encounters
All of the drivers in this study saw their jobs as entailing increased COVID-19 exposure because they were required to continually interact with members of the public, some of whom could be carrying the virus. At a time when public health messaging was asking people to physically distance by staying within their "social bubbles," ridehail drivers risked relatively high exposure as they were in close quarters with dozens of strangers in their car every day: You're always meeting strangers, new people coming in and out of your vehicle, so you're definitely more exposed that way. You're not just staying in your social bubble … I think just being in the car and having random people in and out is a risk by itself (Daniella, Uber) The food and package drivers in this study were aware that they had exposure to people with COVID-19 as those under isolation could not leave their homes and required home deliveries. A concern for drivers across all platforms was that they lacked information about the health status of their customers. In all, the platform drivers knew that they were potentially delivering to, or providing rides to, some people with the virus, and ride hail drivers felt especially exposed as they had customers in their cars. As they had only their own observations about which customers were ill, they were on heightened alert at all times.

| Weak platform safety measures
The digital platform companies did make some changes to their procedures to respond to risks of platform driver exposure to COVID-19. In particular, some ride-hail platforms required passengers (as well as drivers) to answer health-related screening questions.

| Mask-less customers
The ride-hail drivers felt particularly vulnerable to COVID-19 simply by going about the course of their work, as they had customers in their vehicle with them. However, many weighed the potential cost of this disease exposure against what they saw as the greater and more certain cost of loss of income if they did not engage in this study. One area where they regularly weighed risks was in relation to customers who refused or neglected to wear masks, which were required by digital platforms and also Ontario public health regulations. Some ride-hail drivers in our study described feeling able to provide a "friendly reminder" to customers to wear their masks and decline passengers who refused. A key feature of most of the digital platform work (except Amazon Flex) was the workers' overall ratings, which need to be maintained above a certain threshold to stay on the app. As customer ratings of the drivers were an important part of those overall rating scores, the drivers were motivated to behave in ways that pleased the customers, including tolerating risky customer behavior. For instance, one food delivery driver noted that, while she felt unsafe when customers without face masks got very close to talking to her, she was too afraid to ask them to wear a mask or move back because they might find it rude and take back their tips.
The structure of the platforms, with peer-ratings systems that invited customers to rate the drivers, was described by workers as creating an unequal environment that favored passengers. When weighing the cost of confronting a customer with the benefit of the customer wearing a mask, some platform workers felt that the costs outweighed the benefits.

| Time-related digital platform pressures
Time-related platform pressures were another COVID-19 exposure concern for food and package delivery drivers. That is, they had

| Risk of declaring COVID-19 symptoms
As self-employed workers, the drivers had to make their own decisions about what level of COVID-19 risk to accept and bear the consequences of lost income. As a ride-hail manager said, "It's really up to them [workers] how they control that risk" (Sonali, manager, Lyft). However, despite not being legally "employers," the platform companies did offer to provide some financial relief to their workers as COVID-19 increasingly became an occupational health and public health concern. For instance, Uber noted, "Any driver or the delivery person who is diagnosed with COVID-19 or is individually asked to self-isolate by a public health authority will receive financial assistance for up to 14 days while their account is on hold." 39 One ride-hail manager, describing the platform company's provision of funds to prevent workers from working while ill, was deliberate with her words, carefully communicating this was not a sick leave policy: We will pay them out a certain amount of money… We will compensate that driver … to incentivize them not to be on the road during…. I wouldn't call it a sick leave policy, but it's more of a benefit to encourage the driver to stay at home and take care of their wellbeing, so they're not coming out on the road (Trina, manager, Lyft) In addition to various forms of temporary COVID-19 sick pay for workers announced by platform companies, the Canadian government also temporarily provided emergency income support for which self-employed workers were eligible. The Canadian Emergency Response Benefit provided up to $500 per week for workers who stopped working for reasons related to COVID-19 and who had earned at least $5000 in the prior 12 months. 40 However, we found that most workers in our study did not take up either of these opportunities. The compensation offered by the platforms was less than what workers would typically earn, framed accordingly by one worker as "more so incentives to keep working" (Salma, Instacart). In all, despite platform companies and the Canadian government offering financial compensation for time off due to COVID-19, drivers largely did not take this up. Barriers to accessing this support largely centered on uncertainty about the application process, insufficient funds provided by the benefit, and workers' inability to wait without income for a benefit cheque to arrive.

| DISCUSSION
The aim of our study was to identify perceived COVID-19-related exposure and job risks faced by digital platform drivers and to document the measures in place to protect the health of these workers. By engaging in in-depth interviews with platform workers and managers, we were able to gain a grounded understanding of the COVID-19-related conditions of driver's work as well as their motivations and rationales for their actions related to COVID-19 risk-taking and protection. This study extends the limited literature on platform work and COVID-19 risks faced by digital platform drivers. 35 We also provide recommendations on occupational health interventions for this sector.

| Interventions to prevent COVID-19 occupational exposure and job risk
Our in-depth interviews with Canadian digital platform drivers and managers led to the identification of five key COVID-19-related exposure and job risks that drivers faced in the course of their work.
Key exposure risk was the nature of ride-hail, package, and food delivery work, as it involved frequent contact with customers, some of whom would inevitably be virus carriers. Although the digital platform companies instituted COVID-19 risk-reduction measures for drivers, such as customer health statements, contactless delivery, and provision of masks and hand sanitizers, drivers and managers described many ways that these measures were ineffective or difficult to implement.
Given airborne transmission of the virus and public health mask mandates, mask-less customers also posed an exposure risk to drivers, but many felt that it was not worth refusing or contesting these customers because of the possibility of losing tips or getting a bad rating from the customer. Time-related platform pressures were a further risk encountered by the drivers. The need to rush to get packages and groceries delivered on time prompted drivers to breach public health guidelines requiring 2-metre physical distancing. Finally, we found that, despite digital platform companies and the Canadian government providing income support to COVID-19-positive workers, some drivers did not take this up because they could not afford the wait for benefits or risk the loss of income while denied access to the platform should they self-report and be deactivated.
There is emerging literature focused on health experiences and risks to digital platform workers during the COVID-19 pandemic, which echoes our findings. Blundell et al. 1 found that 78% of United Kingdom platform workers felt their health to be at risk while working but they continued to work due to economic insecurity.
Similarly, Beckman et al. 35 found that Seattle, USA, drivers were very concerned about COVID-19 exposure, and experiencing high stress.
Only one-third of the workers surveyed had been provided with free masks or hand sanitizer. An Uber survey of their American drivers found strong concerns among drivers about their physical and mental health during the COVID-19 pandemic. 41 The stress of COVID was widespread: in an international survey of digital platforms, Howson et al. 19 note the mental stress placed on workers due to the asymmetrical power of platforms and the lack of medical treatment available for COVID-infected workers in many jurisdictions.
We found that, as self-employed workers, platform drivers were generally expected to educate themselves about occupational risk and to avoid dangerous situations ("It's really up to them"). Such downloading of risks to workers is a key characteristic of precarious employment conditions. 10 The part-time and limited-term aspects of precarious employment also make it difficult for workers to form relationships to protect their health by collective action. Despite these barriers, platform workers have, in some cases, successfully unionized. 42,43 In the following parts of this discussion, we identify ways that digital platform companies and governments could make changes to address the risks identified in this study.  Recognition of platform drivers as "employees" rather than selfemployed would be an upstream preventive strategy that would make a positive contribution to government tax pools that, in turn, support all workers. That is, as employers, digital platform companies would be required to contribute to social security programs like employment insurance and pension plans. 22 Consequently, as employees, when platform workers become injured or ill, they would be able to access workers' compensation instead of needing to resort to basic social assistance payments. 19,54 Measures to protect the health of platform workers would have carry-on effects on public health. That is, COVID-19 exposure risk for workers translates into broader public health risks if drivers became infected with the virus and then pass it on to customers, families, and consequently, the wider population. As such, government measures to better protect the health of platform workers would also constitute important public health measures.

| Future work and conclusion
We recommend that future research on the occupational health of digital platform workers measure the extent of worker presenteeism, especially in relation to the transmission of contagious viruses. The effectiveness of occupational health interventions, such as income support benefits when ill, could also be measured among platform workers. What makes platform workers different from regular workers are pervasive and rather impersonal algorithmic digital platform pressures that nudge them toward accepting health risks, which creates a need to examine the effectiveness of occupational health interventions among this specific population. Future work could also examine the crossover effects of COVID-19 occupational health protections on public health outcomes. While platform work has been found to be gendered and racialized, 55,56 our study did not discern specific gender or race effects in relation to COVID-19 risks faced by drivers. It may be that low income created a common thread among participants, or that our qualitative study did not access a sufficient sample to explore these dimensions.
Strengths of our study include the qualitative approach that enabled us to identify a range of perceived COVID-19-related risks among digital platform workers and why the current strategies to mitigate risk appeared to be largely ineffective. Our findings provide information that can be used to develop future larger-scale studies regarding risk and to improve prevention strategies. Our study does not aim for statistical generalization; rather, our goal was to provide enough detail for analytic generalization, in which the reader is provided with sufficient information to decide whether the findings are relevant to other contexts or jurisdictions. 57 It is important to note that interventions to address perceived risks may not comprehensively address all relevant risks, including risks not directly perceived by workers.
In all, our study found that the nature of digital platform driving work was one of constant possible exposure to COVID-19 through work processes that created contact with customers and regular encounters with crowded environments. Despite digital platform company attempts to screen passenger health, institute contactless delivery, and provide financial compensation to workers when ill with COVID-19, we found that these had limited impact on the day-to-day COVID-19 exposure and job risks faced by digital platform drivers. It is important to consider that, as self-employed workers without the protection of employment and occupational health standards, platform workers are absorbing most of the occupational risks related to COVID-19. This arrangement also poses a societal risk because platform workers who take risks because of digital platform pressures can also spread the virus in the course of their work delivering food, packages, and people.