The COVID‑19 pandemic has triggered the most severe recession in nearly a century and is causing enormous damage to people’s health, jobs and well-being. The extent of the crisis is still unfolding, but economies and labour markets have already been facing a double blow. On the supply side, many workers became ill, were quarantined, or were subject to lockdowns; many companies found themselves unable or even banned from operating. On the demand-side, consumers were unable or unwilling to consume, as job losses, lockdowns and fear of contracting the virus changed their consumption behaviour.

The damage to the economy will be significant. Even in the more optimistic scenario in which a second wave of infection is avoided, OECD projections (as of 10 June 2020) suggest a drop in global economic activity by 6% in 2020 and a rise in OECD unemployment to 9.2% from 5.4% in 2019. If a second wave of rapid diffusion of the virus does occur, triggering a return to lockdowns, global economic output is forecast to plummet 7.6% in 2020 and OECD unemployment to increase to 12.6% (more than double the rate prior to the outbreaks).

The evidence so far is that vulnerable workers are bearing the brunt of the crisis, particularly those who are self-employed or working in the platform economy (OECD, 2020[1]), in which apps or websites match customers to individuals offering services and mediate the transaction. Previous research by the AppJobs Institute showed that platform workers (i.e. those individuals offering services) were among the hardest hit by the crisis (Moulds, 2020[2]). In particular, platform workers face two major risks: i) some platform workers are key to ensuring the continuation of essential services during lockdowns and, while their activity has increased during this period, they have found themselves at substantial risk of exposure to the virus;1 and ii) some platform workers face income losses due to lower demand or inability to work due to illness, quarantine or caring duties.

Among platform workers, the incidence and impact of these risks differs according to the type of work being carried out (e.g. demand for delivery services through platforms has increased while demand for cleaning, babysitting and housesitting has decreased (AppJobs, 2020[3])), individual characteristics and family circumstances, measures taken by governments and by platforms, among other factors. There is also some evidence that the crisis has taken a higher toll on some types of workers. For example, women are over-represented in the most affected sectors and disproportionally hold precarious jobs (OECD, 2020[1]). Meanwhile, widespread childcare facility and school closures likely amplified their unpaid work burden at home.

The risks posed by the current crisis are compounded by the generally lower levels of access among platform workers to unemployment benefits, health insurance and sick leave, compared with individuals in standard (permanent, full-time, dependent) employment. While countries’ definitions and tests for employment status vary, many platform workers are considered self-employed, which means that they do not usually benefit from the same rights and protections as standard employees. At the same time, many platform workers do share characteristics with standard employees as well as some of their vulnerabilities to income loss. This has led the OECD to argue that rights and protections for this group of workers should be strengthened (OECD, 2019[4]).

During the COVID‑19 crisis, governments across the OECD have taken unprecedented steps to protect individuals performing non-standard work, including the self-employed (OECD, 2020[1]), by: providing cash payments to the entire population, providing payments specifically targeted to the self-employed (related to previous earnings or not), and expanding access (in many cases, temporarily) to sickness benefits and special paid care leave (OECD, 2020[5]), unemployment benefits and short-time work schemes to the self-employed. However, as evidenced by the data collection exercise carried out for this note, many platform companies themselves have also taken measures to protect the workers using their platforms.

To investigate these measures, a data collection exercise was carried out in June 2020 by the OECD in collaboration with AppJobs and the AppJobs Institute. AppJobs connects companies that run the apps or websites that mediate platform work (henceforth, “platforms”) with individuals seeking work. The AppJobs Institute was founded with the objective to use AppJobs’ data and market access to research, analyse and share insight about the platform economy with decision-makers and researchers.

Data were collected through three methods: i) a survey of platforms; ii) desk research to gather information publicised by platforms on their websites, in press releases etc.; and iii) a survey of workers using platforms through AppJobs. The survey of platforms and accompanying desk research were designed to focus on platforms in AppJobs’ contact list operating in an OECD member country and mediating services on location, hence most impacted by the pandemic (i.e. platforms mediating services delivered entirely online were excluded). The exercise captures information from 64 platforms mediating services as diverse as delivery (27 platforms), ride-hailing (8 platforms), babysitting (4 platforms), property rental (3 platforms), cleaning (3 platforms), car-sharing (2 platforms), removals and storage (2 platforms), gardening (2 platforms), pet-sitting (2 platforms) and a variety of other services (11 platforms). The list of platforms included in the data collection exercise is provided in Annex 1.A.

In addition to the information collected from platforms, AppJobs members2 (i.e. individuals that have registered with AppJobs in order to find flexible work opportunities) in OECD member countries were invited to participate in the survey of platform workers, and 745 registrants chose to do so. The survey of platforms was designed to focus on platforms mediating services on location (rather than online), in order to capture not only the risk of potential income loss but also the risk of exposure to the virus while working. However, there was no such restriction in the worker survey, as many respondents use multiple platforms. When asked to classify the platforms that they used according to the type of service delivered, the most represented categories were: delivery (25%), freelance work (19%)3, cleaning (18%), driving (17%), online surveys (12%), teaching/tutoring (8%) and house sitting (8%).

The data collection exercise was not designed to deliver a representative sample. The desk research and survey of platforms are focused on platforms in AppJobs’ contact list, which is not necessarily representative of the full population of platforms. The worker survey may have also disproportionately attracted respondents who wanted to comment either positively or negatively on the topic, which could result in self-selection bias. Residents of Anglophone countries may be overrepresented due to conducting the survey in English only. For instance, 37% of respondents were based in the United States, 14% in the United Kingdom and 9% in Canada.

Figure 1 shows that platforms reported taking a variety of different measures in response to the COVID‑19 crisis to mitigate risks to health and income loss of workers using their platforms. 37 platforms (out of a total of 64 – or 58%) reported taking measures to promote social distancing and/or safe provision of services while 15 platforms (23%) reported taking measures to provide full or partial pay for sick or self-isolating workers. Two-thirds of platforms said they had taken at least one of the measures described and the average number of different types of measures taken by platform was 1.7 (out of 11 categories, as shown in Figure 1). All measures are discussed in more detail in the next sub-sections.

Platforms mediating delivery, freelance and online/remote services (services that experienced an increase in demand during the COVID‑19 crisis according to AppJobs (2020[3])) reported a higher number of different types of measures (2.4 types of measures on average) than platforms mediating cleaning, babysitting and housesitting services (services that experienced a decrease in demand) (0.5 types of measures on average).

The platforms experiencing greater demand during the crisis may have been offering greater support in efforts to attract new workers to their platforms. Platforms experiencing lower demand may have perceived less need to offer support due to a reduced level of activity. For instance, platforms mediating cleaning, babysitting and housesitting services (that experienced a decrease in demand) were half as likely to report taking measures to promote social distancing and/or safe provision of services compared to delivery, freelance and online/remote services (services that experienced an increase in demand).

By far, the most common measures reported by platforms were aimed at promoting social distancing and/or the safe provision of services, such as giving guidance on how to stay safe while working, encouraging contactless delivery, ensuring a safe working environment and refining their service offer. This is in line with other recent research on this topic (Fairwork, 2020[6]). Measures such as these were reported by 37 platforms, more than half of the 64 platforms included in the exercise. Many of these measures were aimed at promoting a safe environment for both clients and workers, and in most cases, it is difficult to distinguish between measures intended to protect one party or the other.

  • Two-thirds of delivery platforms reported taking measures to encourage contactless delivery and/or to remove the obligation to obtain a signature from the customer. US-based delivery platform GrubHub not only introduced contactless delivery, but also an option for drivers to stay in their vehicle while picking up an order.

  • Many platforms reported providing guidance to workers on how to stay safe during the pandemic. While many platforms referred workers to the World Health Organization (WHO) or national guidelines, some provided guidance specifically targeted to the type of service being offered. For example, delivery platform Quiqup published a step-by-step instructional video4 to demonstrate to workers in Dubai how to perform a contactless delivery.

  • Some platforms report that they have taken steps to prevent workers or customers who may present a risk from using the platform. For instance, Uber introduced an obligation for drivers and delivery workers to take a selfie before going online to prove that they were wearing a mask and an obligation for riders to confirm before every ride that they have taken specific hygiene measures like washing their hands. Delivery platform Roadie has been locking the accounts of sick or exposed workers (for a minimum of 14 days) until they can confirm that they no longer pose a health risk. Storage platform Clutter said that it would not service properties or individuals with confirmed cases of COVID‑19 or evidence of symptoms.

  • A handful of platforms reported refining the services offered in response to emerging risks. TaskRabbit paused IKEA assembly services in many US locations in accordance with governmental guidelines, while carpooling app Poparide said that its services should be used for essential travel only. While complying with government regulations or advice may seem a trivial measure, it is notable that the platform regulated the activities of the workers using the platform by suspending certain services to comply with government health and safety regulations – whereas without the mediation of a platform, a self-employed individual providing similar services would be responsible for this decision him or herself.

  • Where the working environment is generally outside of the control of the workers, some platforms focused on measures to ensure a safe working environment. For example, TaskRabbit asked customers to clean doorknobs. Babysitting platform Sittercity gave advice to clients on how to discuss hygiene practices in the home with a new babysitter.

  • A small number of platforms reported enabling workers to have greater control over which tasks they accepted, allowing them to target less risky and/or more profitable tasks. For instance, Shipt allowed its shoppers to preview the shopping list for an order before accepting it, so that they could select the orders they felt more confident carrying out. Product availability estimates have also been added to the Shipt app in order to reduce the amount of time that shoppers spend standing in line.

A quarter of platforms in the study reported providing or reimbursing the cost of purchasing personal protective equipment (PPE), including masks and gloves, and/or hygiene products, including disinfectant wipes and hand sanitiser. In some cases, PPE was sent to workers’ homes or made available at pick-up points and, in others, platforms provided reimbursements for materials purchased. For instance, in the United Kingdom, Stuart reimbursed GBP 30 for protective equipment per 30‑day period for couriers who perform deliveries at least 20 days out of the period. DoorDash sent a kit comprising hand sanitiser, gloves and masks to workers’ homes. However, the survey of workers revealed that a number of them were dissatisfied with what was provided by platforms, as discussed below.

A handful of platforms reported partnering with telehealth companies to provide free or reduced-cost medical consultations or testing, primarily to workers in the United States. Delivery platform Doordash and its subsidiary Caviar offered USD 4 virtual medical consultations while ride-hailing platform Uber offered workers a free COVID‑19 risk assessment, which would refer workers at highest risk to a free virtual medical consultation.

In addition to measures taken to protect the health and safety of workers, some platforms, particularly those in the United States, reported providing temporary financial support for individuals who were either sick or forced to self-isolate. In the case of platforms that prohibited sick or exposed workers from accessing work through the platform temporarily (as discussed in the previous section), such financial support would be particularly important. In total, one in four platforms have provided such financial assistance, which mostly provided the equivalent of two weeks’ income (e.g. Uber, DoorDash, Postmates, Shipt, Caviar, Amazon Flex and Workforce Staffing) and sometimes involved eligibility criteria such as a minimum level of activity pre-crisis. For instance, the delivery platform Stuart paid two‑thirds of the worker’s average weekly earnings (based on during the 6‑week period between 10 February and 22 March), pro-rated for every day the worker was in isolation – up to a maximum of 14 days and capped at GBP 300 per period of 7 days. Only workers who had completed a minimum of 120 deliveries during the same 6‑week period were eligible. Italian freelance platform ProntoPro provided workers with a one‑off “convalescence allowance” of EUR 1 500 in case of hospitalisation and a one‑off payment of EUR 350 in the event of a mandatory quarantine. Delivery platform Doordash put in place a childcare compensation scheme, which offered a one‑off payment equivalent to 85% of two weeks’ average earnings for highly-rated and highly-active workers who could not work due to school closures. While several other platforms stated that they had provided financial support to workers, it was not clear what shape that support took or what the eligibility criteria were (e.g. Volt, Bolt, Instacart, GrubHub, Lyft).

Some platforms reported providing special support for workers facing financial hardship. Amazon set up a fund focused on supporting delivery drivers and temporary agency workers who face financial hardship as a result of a COVID‑19 diagnosis or quarantine (a wider range of qualifying events will be covered in the future). Applicants can apply for personal grants ranging from USD 400 to 5 000 per person. Other platforms set up relief funds during the crisis, including: GrubHub, Postmates and AirBnB – although the details of how these funds operate are not available.

Many ride-hailing and delivery platforms let vehicles to workers who do not own one. During the crisis, several platforms reported taking steps to remove charges/penalties for the early return of vehicles (Uber, Lyft), reduced rental rates (Uber, Lyft) or helped drivers in refinancing loans (Uber). Stuart (United Kingdom) provided free vehicle safety and roadworthiness tests, as well as discounted insurance policies.

During the crisis, several platforms reported making adjustments to their ratings systems to ensure that workers’ ratings were not unfairly affected by lockdown measures or periods of sickness. Last-minute cancellations due to illness did not affect workers’ metrics on Task Rabbit (including cancellations where the worker was worried the client might be sick) and several other platforms removed ratings below 5 stars during the period (e.g. Shipt, Instacart). The vacation rental platform VRBO (United States) gave individuals the equivalent of a 5‑star review if they provided clients with a 100% refund during the crisis. Canadian carpooling platform Poparide committed to waiving any cancellation penalties for any passengers or drivers who cancel due to illness.

Similar changes were made to some of the loyalty schemes for platform workers. For example, Uber reported that they protected the Uber Pro and Uber Eats Pro status of all their drivers and delivery people. Similarly, VRBO’s Premier Partners reported that that they kept their status (as long as travellers were given 100% credit or 50% refund for cancellations) and Lyft drivers could maintain their Rewards tier, even if they did not earn enough points during the period.

There are a number of other ways in which platforms reported trying to help workers financially during this difficult time. For example, Uber encouraged customers to tip while Instacart introduced bonuses for in-store teams, and several platforms have assisted workers in applying for government support schemes (e.g. Uber, Instacart, VRBO, BidVine and Lyft).

Finally, some platforms reported trying to assist workers to find alternative or additional work opportunities during the crisis. Uber developed a Work Hub designed to help workers find work (with either Uber or another company) and made it easier for people who have historically driven with Uber to find new ways to earn by delivering food with Uber Eats. Other platforms drew attention to the possibility to deliver services virtually rather than in person. AirBnB launched a new Online Experiences service, which allows individuals to host online events such as cooking classes.

A complementary worker survey was undertaken to ensure that the views of platform workers were represented, to complement and qualify the information reported by platforms. In particular, some external reporting suggests that some platform workers have encountered problems in accessing the aid provided by platforms (Fairwork, 2020[6]), including in obtaining promised PPE (Dickey, 2020[7]) and meeting strict and/or opaque eligibility criteria for sick pay (Brandom, 2020[8]). Other workers may be facing worsening working conditions due to other changes implemented by platforms during the crisis, such as pay decreases due to changes in the algorithm used by platforms (Hussain, 2020[9]). The worker survey cannot validate the measures reported by platforms, but nonetheless allows the views of workers to be represented.

Around 35% of respondents in the worker survey reported that their platform(s) had taken measures to assist them during the pandemic and just under half of this number (44%) said that they were satisfied with the measures taken. Figure 2 presents a breakdown of the prevalence of, and satisfaction with, measures taken by the type of platform, as reported by workers. Most workers were using more than one type of platform, and satisfaction and measures taken were reported at individual level, rather than platform level. The measures most frequently reported by respondents were full/partial sick pay (41%), measures to encourage social distancing/use of PPE (35%), communication with workers/associations (24%), flexibility in how to perform work (23%).5

Workers using platforms to provide services such as private car rental, delivery, driving and home/private sector rental were most likely to say (40‑50%) that their platform(s) had taken measures to assist them during the pandemic (although the number of respondents providing the rental services was small). Of those providing delivery and driving services, 45% and 40% respectively were satisfied with the measures taken, roughly matching the average across all types of platforms.

Workers using platforms to provide services such as pet-sitting and online surveys were least likely (9% and 12% respectively) to say that their platform(s) had taken measures to assist them during the pandemic, although those that did receive support reported higher levels of satisfaction (60% and 69% respectively) than the overall average. No clear distinction emerged in the prevalence of measures taken between services typically delivered on-location versus online.

Of those who rated the accessibility of the resources provided, more respondents said that it was easy/relatively easy (38%) to access these resources than those who said it was hard/relatively hard (30%).

When asked what measures respondents might like platforms to take to help workers during the COVID‑19 crisis, 44% of respondents said that they would like or needed platforms to take measures (additional to those already taken), while 26% said that there were no such measures. In order to prompt views on what types of measures were needed, respondents were given the opportunity to make an open comment to suggest how they would like platforms to help and 38% did so. Many of these workers expressed a desire for better employment opportunities or more work through the platform6. A number of them also complained about the PPE that was either provided or reimbursed. Some workers reported that PPE had been promised but not delivered, that PPE delivered had been of bad quality, or that PPE delivered or reimbursed was not sufficient in quantity for the period (e.g. just one bottle of hand sanitiser or one upfront payment). Other workers asked for:

  • Financial help, such as general financial support, hazard pay, days off, help with specific items (vehicle payments, food, rent and other basics), access to loans, charity or emergency funds;

  • Access to benefits, such as sick leave, benefits, furlough and other government assistance;

  • A safer work environment, including for clients to keep a distance and wear a mask;

  • Better treatment from platforms, such as evidence that they care about wellbeing of sick workers, a cessation of dismissals, better communication from platforms (including being informed when a co-worker tests positive), and general respect.

References

[3] AppJobs (2020), Coronavirus vs. Gig Economy - Many Left Jobless, https://www.appjobs.com/blog/coronavirus-effect-on-gig-economy (accessed on 20 July 2020).

[8] Brandom, R. (2020), “The human cost of Instacart’s grocery delivery”, The Verge, https://www.theverge.com/21267669/instacart-shoppers-sick-extended-pay-quarantine-leave-coronavirus (accessed on 23 July 2020).

[7] Dickey, M. (2020), “Gig workers say they are struggling to get personal protective equipment from companies”, TechCrunch, https://techcrunch.com/2020/04/16/gig-workers-personal-protective-equipment/ (accessed on 23 July 2020).

[6] Fairwork (2020), Fairwork Releases Report on Platform Responses to Covid-19, https://fair.work/fairwork-releases-report-on-platform-responses-to-covid-19/ (accessed on 15 September 2020).

[10] Folkhälsomyndigheten (2020), Förekomst av covid-19 i olika yrkesgrupper, http://www.folkhalsomyndigheten.se/publicerat-material/. (accessed on 21 July 2020).

[9] Hussain, S. (2020), “Target-owned Shipt sparks worker protest with pay change”, Los Angeles Times, https://www.latimes.com/business/technology/story/2020-07-20/shipt-shoppers-protest-pay-algorithm-change (accessed on 21 July 2020).

[2] Moulds, J. (2020), “Gig workers among the hardest hit by coronavirus pandemic”, The World Economic Forum, https://www.weforum.org/agenda/2020/04/gig-workers-hardest-hit-coronavirus-pandemic/ (accessed on 4 May 2020).

[1] OECD (2020), OECD Employment Outlook 2020: Worker Security and the COVID-19 Crisis, OECD Publishing, Paris, https://dx.doi.org/10.1787/1686c758-en.

[5] OECD (2020), “Paid sick leave to protect income, health and jobs through the COVID-19 crisis”, OECD Policy Responses to Coronavirus (COVID-19), OECD Publishing, Paris, http://www.oecd.org/coronavirus/policy-responses/paid-sick-leave-to-protect-income-health-and-jobs-through-the-covid-19-crisis-a9e1a154.

[4] OECD (2019), OECD Employment Outlook 2019: The Future of Work, OECD Publishing, Paris, https://doi.org/10.1787/9ee00155-en.

[11] Office for National Statistics (2020), Coronavirus (COVID-19) related deaths by occupation, England and Wales, https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/causesofdeath/bulletins/coronaviruscovid19relateddeathsbyoccupationenglandandwales/deathsregisteredbetween9marchand25may2020 (accessed on 21 July 2020).

Contact

Stefano SCARPETTA (✉ stefano.scarpetta@oecd.org)

Stijn BROECKE (✉ stijn.broecke@oecd.org)

Marguerita LANE (✉ marguerita.lane@oecd.org)

Notes

← 1. Emerging evidence (not specifically focused on platform work) suggests certain activities carry higher risk of exposure. For example, evidence from both Sweden and the United Kingdom suggest that road transport drivers (including taxi and van drivers) were among the occupations at highest risk of infection/death from COVID‑19. In Sweden, taxi drivers were found to be 4.8 times more likely to be infected by COVID‑19 than the general population (Folkhälsomyndigheten, 2020[10])). In the United Kingdom, road transport drivers were found to account for the largest proportion of deaths (35.2 deaths per 100 000 men) (Office for National Statistics, 2020[11]).

← 2. AppJobs has more than 1.7 million members, 1.4 million of whom are based in OECD member countries.

← 3. Freelance work can include, for example, design, programming, content writing, digital marketing, etc.

← 5. The findings are not directly comparable with information collected from platforms for multiple reasons, for instance: (i) platforms mediating only online services were excluded from the survey of platforms and accompanying desk research; (ii) the worker survey naturally gives greater weight to more popular platforms; (iii) platforms and workers may have described measures differently.

← 6. The survey did not attempt to assess whether comments such as this would have been made in the absence of the COVID‑19 crisis. As such, it is impossible to tell whether such comments are directly related to the COVID‑19 crisis or are more general comments on platform work.

Disclaimer

This paper is published under the responsibility of the Secretary-General of the OECD. The opinions expressed and the arguments employed herein do not necessarily reflect the official views of OECD member countries.

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area.

© OECD 2020

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at http://www.oecd.org/termsandconditions.