2. Why measure digital platform employment and work?

The development of digital technologies and of new business models have contributed to the rise of online platforms and the emergence of different types of platform-mediated work, such as “crowd work”, “gig work”, and other forms of on-demand labour. Most of such work is performed by independent self-employed or own-account workers, in many cases only as a part-time job (e.g. on-call work).1

One positive aspect of digital platform employment is the increased efficiency of the matching process in the labour market, which may help to reduce frictional unemployment and skills mismatches. Another advantage, often cited by workers, is greater flexibility to choose when and where to work.

However, workers engaged in this type of work also report facing stressful situations more frequently than other workers, to perform more routine tasks and to have fewer learning opportunities (Pesole, 2018[1]). Broader concerns about digital platform employment relate to ensuring job and income security, access to social protection, career development, training, rights to collective bargaining and protection against algorithmic discrimination and opaque management practices. Some platform workers may face an increased risk of work accidents or job strain, including mental health problems. Platform employment and work also raise concerns for tax avoidance, distortions to product market competition, social dumping and race-to-the-bottom practices. Policy makers in many OECD countries have recognised these problems, and emphasised their determination to take steps to improve the working conditions of platform workers (von Der Layen, 2019, p. 10[2]).

This chapter provides the motivation for the statistical guidelines proposed in this volume:

  • First, it highlights the growing international demand coming from a range of different policy perspectives to measure digital platform employment, the number of workers involved, their individual and job characteristics, and their working conditions.

  • Second, it argues that better statistical guidelines would help improve the classification of platform workers, and therefore the enforcement of law (e.g. labour law, social security and taxation) with regard to digital platform workers. In some cases, new platform business models have spurred the growth of false or bogus self-employment, with implications for their employment conditions and protections. This misclassification needs to be addressed by correctly classifying platform workers within the framework of existing labour law.

  • Third, it examines how a lack of data also threatens to affect the development and implementation of existing and forthcoming policies in the areas of labour market regulations, social protection and social dialogue.

This chapter reviews existing evidence on digital platform workers by taking stock of a dedicated ad-hoc survey, namely COLLEEM, and explains the issue of digital platform workers classification. It then lists some key policy issues that are brought to the fore by the diffusion of platform work, and which call for better data by statistical offices and other data producers. Finally, it highlights major data gaps that will need to be filled in the future.

One of the major challenges of the platform economy is to gauge the size and characteristics of platform work. Definitions and measurement approaches directly affect the range of available estimates. Indeed, the size of platform work depends on the conceptual definition of what constitutes platform work and a digital labour platform, on the methodological approach used to collect the data, and the reference period used. Statistical offices around the world have launched initiatives to capture this new phenomenon, mobilising a range of sources (labour force surveys and special modules thereof, surveys about IT use by individuals, etc.). Unfortunately, lack of common definitions and classifications has led to a range of estimates that differ in terms of both the prevalence and intensity of the phenomenon, its pace of diffusion in different countries, the characteristics and working conditions of the workers involved.

This section presents findings from the JRC COLLEEM surveys, conducted over the years 2017 and 2018 among frequent internet users aged between 16 and 74 years old, which has the advantage of covering a broad range of European countries based on a common methodology. The JRC COLLEEM survey is an online panel survey covering 15 EU Member States2 plus United Kingdom, that gathered a total of 38,878 responses. A limitation of COLLEEM is that respondents were drawn from commercial online panels, and that participants to this survey might be more engaged in online work than the general population. Similarly, this survey may be less effective in detecting platform workers who supply labour on-location (Urzi Brancati, Pesole and Fernandez Macias, 2020[3]). Despite these limitations, COLLEEM is currently the most important source of information available providing comparative data on platform work in Europe.

The filtering question to identify platform workers in the COLLEEM survey follows the so-called “income approach” (see Chapter 3). That is, it asks respondents whether they have ever gained income from different online sources, excluding any labour supplied via capital platforms (e.g. AirBnb) or any additional income deriving from e-commerce, crowdfunding and similar activities.

The prevalence of platform work for those who ever provided labour via platform is around 11% of the working age population for the 16 countries covered, ranging from a maximum of 18% in Spain and to a minimum of about 6% in the Czech Republic.

The broadest definition of ‘ever platform workers’ tells us about the extension of the phenomenon in each country. However, it brings little information on how many people routinely spend a significant amount of time providing services via platforms or make a living out of platform work. In order to get a measure of the prevalence of platform work that can be broadly compared to standard measure of paid work or employment, additional elements need to be taken into account: i) the amount of time spent on platforms; ii) the frequency of provision of labour services; and iii) the income gained. Figure 2.1 shows how the initial estimate for ‘ever platform workers’ change when frequency (i.e. labour services are provided via platform at least monthly), regularity (i.e. number of hours worked) and income earned (i.e. more than a quarter of the income of respondents) are taken into account.

About 11% of the working population in the 16 European countries covered by COLLEEM has provided labour services through platforms at least once; these are the ones defined as ever platform workers (Ever PW). When regularity of labour services provision is considered, the share for platform workers who have provided labour services at least monthly (Freq PW) is 8.6% of the working-age population. To identify employed person, the Labour Force Survey (LFS) asks respondents whether they worked for pay or profit at least one hour during the survey reference week. 3 In the context of COLLEEM, platform workers who worked at least 10 hours in the last month (>10hrs) are considered as potentially satisfying the LFS “one-hour criterion”, and hence could be considered as ‘employed’ in the LFS sense. In other words, 4% of the working-age population in the 16 European countries would be employed via platforms, and 2.5% of the working-age population made at least 25% of their income through digital platforms.

However, to better align the COLLEEM category of platform workers to the LFS one of people employed in main job, the criteria of a sufficient number of hours and income earned can be considered. Main platform workers can be defined as those who provide labour services via platforms at least monthly, who work on platforms at least 20 hours a week or get at least 50% of their income via platforms. These workers account for 1.4% of the working population in the 16 European considered.

The COLLEEM survey is perhaps not the best instrument to gauge the size of the platform economy, a question for which a Labour Force Survey (LFS) would constitute the main statistical vehicle as its sampling is more representative and accurate. However, the COLLEEM survey is very useful to describe the characteristics of digital platform workers and their tasks, which are not accessible via a LFS due to space constraints. Box 2.1 summarises the key findings.

The second wave of COLLEEM, run in 2018, also included a longitudinal component. 3,323 respondent who were identified as platform workers in the first wave were re-interviewed in 2018. By focusing on this sub-sample, (Urzi Brancati, Pesole and Fernandez Macias, 2020[3]) find that 41.4% of those identified as platform workers in 2017 remained in this category in 2018, as opposed to 58.6% who dropped out of it. Platforms mediating transportation services have a higher turnover rate than those who mediate professional online services or mediate microwork.

Platform workers are generally young. The average platform worker in the 16 European countries covered by COLLEEM is 34 years old. Software developers tend to be younger (32.5), and women providing such services are the youngest (30.8) among all platform workers. The average age of platform workers who provide services on location is 33, and about 34 for freelancers and micro-taskers. Women in all tasks are slightly younger compared to men in the same task.

About 4 in 10 platform workers are women. The gender gap is smaller when work-intensity in the platform is lower. Marginal platform workers are 55% men and 45% women; the gap increases with the intensity of the work provision, up to almost 70% men and 30% women for Main platform workers. Similarly, women tend to be more represented in specific tasks, in particular, freelance (43%) and micro-task (41%), while the share of women in software development is the lowest (24%) followed by transport and delivery (37%).

In 2018, COLLEEM also collected data on platform workers by country of birth. A much higher proportion of foreign-born workers provide labour services via digital platforms as compared to native workers. However, this pattern does not have a unique interpretation. On one hand, the large presence of foreign-born platform workers may suggest that work on digital labour platforms is not particularly attractive, since several studies have demonstrated how foreign-born workers tend to be employed in lower quality jobs while being overqualified (OECD, 2018[4])). On the other hand, the results may be an indication of the labour market opportunities that platforms provide to this population.

As shown on Figure 2.2, foreign-born workers are mostly concentrated in software development, a medium-high skilled task where the presence of migrants may suggest some outsourcing from companies to platforms to reduce labour costs, and similarly for firms in transport and delivery.

About 50% of platform workers possess at a least a higher secondary education, and about 40% have a bachelor’s degree or higher. This share is below 30% for foreign-born platform workers. The general level of education increases with the intensity of the work provision on platforms, while the differences across types of tasks are less remarkable. In general, freelance and software development present slightly higher shares of high-educated platform workers.

Platforms disrupt the traditional organisation of work by disaggregating jobs into the accomplishment of single tasks, and by introducing new managerial practices that facilitate the coordination of a multitude of workers to a multitude of tasks of different nature in a global market at almost real-time. As a result, platform workers often provide multiple type of services using different platforms. According to COLLEEM data, roughly 60% of frequent platform workers have provided more than one type of task. When a platform worker says that he or she has done more than one type of task, it means that they did for instance a micro-task and a delivery. The percentage of platform workers providing more than one type of task increases with the intensity of work via platform.

Place of living may also affect the type of services supplied by platform workers. Indeed, some services may not be available in areas that are more rural. Figure 2.3 shows a higher concentration of platform workers in big cities and metropolitan areas; this effect is stronger for on-location services, as online services can in theory be supplied from everywhere with adequate broadband coverage.

Another interesting information collected in COLLEEM is to which extent platform workers offer their services through more than one platform (‘multi-homing’). According to COLLEEM data, 40% of platforms workers4 use more than one platform, 52% use only one platform and the remaining 8% prefer not to answer. When looking at a breakdown by occupation, platform workers providing professional services tend to use more than one platform. This is the case particularly for translation services, as 40% declare to use more than one platform, and for interactive services (47%). By contrast, platform workers providing on-location and transport and delivery services are more engaged and locked-in with one platform, as only 31% and 36% respectively provide services in more than one platform.

The limitation of COLLEEM being a self-administered online survey becomes more stringent when analysing data on working conditions, in particular about the income earned via platforms. To gauge the working conditions of platform workers, two dimensions must be looked at, namely total hours worked and the income generated from them.

COLLEEM asks respondents about the total number of hours worked (both in online and offline labour markets) and the number of hours spent on platform. Figure 2.4 shows that most platform workers work around a total of 40 hours per week. (Urzi Brancati, Pesole and Fernandez Macias, 2020[3]) find that among main platform workers, 25% report to be employees, while among secondary platform workers this share increases up to 36%. These findings suggest that a high share of platform workers have additional regular jobs and use digital labour platforms as a secondary source of income. When considering hours worked via platform, Figure 2.4 shows a distribution skewed toward the left. Indeed, 44% of platform workers work less than 10 hours per week via platform, while above 60% spend less than 20 hours supplying labour via platform.

COLLEEM asks survey participants “how many hours do you work via online platforms?”. Answers to this question might underestimate the time spent waiting for an assignment or for searching online. Additionally, the survey asks on what basis platform workers get paid: about 69% of respondents report to get paid based on the task performed, which implies that many of them must undertake a substantial share of unpaid work in order to get paid work.

When considering the quality of the time spent on platforms, 54% of platform workers declare to work long hours, 68% claim to work at night and 74% to work during the weekend, with little variation among types of services. Generally, platform workers seem to concentrate most of their work in atypical or unsocial hours, implying worse working conditions and quality of job available on platform.

COLLEEM also elicits information about how much platform workers get paid for their last task, in an attempt to construct an indicator of payment per hour and payment per task for the different platforms and occupations. To minimise the effect of outliers, Urzi Brancati, Pesole and Fernandez Macias (2020[3]) restrict the sample to a pool of consistent respondents, and provide information on the average payments received for the 7 most popular platforms in the dataset. Then they calculate the median payments for the selected platforms in the COLLEEM dataset. These data are only indicative, due to both the small number of observations and the fact that the values reported are averaged across the 16 countries and for all types of services. Indeed, the price of a similar service may vary according to the country (e.g. a Uber ride in Paris or in Lisbon) and for those platforms offering several types of tasks, as such as Freelancer or PeoplePerHour, the variation across task-types could be substantial. The authors report that “the lowest median value in Freelancer is €5 per hour for micro-task jobs and €22 per hour for interactive services. There are no other official sources we could use to compare our findings with; however the Fair Crowd Work website collects information on the median payment in some selected platforms. The only two platforms that overlap are Clickworker and Upwork for which they report respectively a median value of €2.92 and €12.91”.

Urzi Brancati, Pesole and Fernandez Macias (2020[3]) also report an estimate of pay per hour for the different type of tasks. According to their findings, the highest hourly paid tasks are software (EUR 23) and interactive services (EUR 16). Clerical, professional and sales tasks are all paid around EUR 14 per hour, on-location services and transport vary between EUR 9 and EUR 12, while the least paid are micro-tasks, at about EUR 6 per hour.

Finally, COLLEEM includes some questions on working organisation and the specific psycho-social conditions of platform work. As a result, half of platform workers find their work stressful, monotonous and unhealthy. Regarding working conditions, 80% of platform workers declare to be able to choose the pace of their work, although 70% declare to be under constant monitoring from the platform. Finally, two-thirds of platform workers consider ratings by clients as an important feature to escalate platform work, and declare to interact with peer platform workers.

Platform work blurs the line between dependent and self-employment5. Platform workers are typically classified in labour force statistics as own-account workers. However, this classification is often incorrect since, like employees, they often have limited control over their work (e.g., in some cases they cannot determine the prices of the services they provide, they are required to wear uniforms, they cannot choose the order of their tasks, etc.) and/or are dependent on their clients/employers in other ways (e.g. financially). Control may be exerted via technology-enabled monitoring, with the algorithm taking the place of a traditional manager, see Pesole (2019[5]) for a conceptual discussion on labour platform and algorithmic management.

The classification of platform workers as self-employed status has the effect of excluding them from rights, benefits and protections that are typically available to employees. One major concern is that some workers may be incorrectly classified as self-employed; and this allows platform operators to avoid regulation and taxation, giving them a competitive advantage over compliant firms in the ‘traditional’ economy and at the workers’ expense (Urzi Brancati, Pesole and Fernandez Macias, 2020[3]).

With new technologies blurring the lines between employment and self-employment, the capacity of labour inspectorates to monitor and detect breaches of labour regulations becomes crucial. For example, Spain has developed campaigns targeted at false self-employment in platform work, including a dedicated operative procedure providing specialised training to inspectors and implementing regional pilot programmes.

At the individual level, filing a complaint with a court against platforms can be daunting. It can be costly, complicated, and the outcome is often uncertain. In addition, workers may worry about retaliation whereby the platform operator removes access to the platform. Where these barriers exist, and where the consequences of abuse are minimal, platform operators may have little incentive to correctly classify workers. To address that issue, some governments have made it easier for platform workers to challenge their employment status, by placing the burden of proof on the employer rather than on the employee or by reducing court fees and by simplifying procedures.

Platforms are often described as intermediaries providing the infrastructure needed for the worker to find clients. However, this makes it hard to agree on who the employer is. Platform work typically involves a multiplicity of clients and tasks of a very short duration, even if these tasks are sometimes carried out on the premises of the client. Some authors have argued that the question of who is responsible for worker rights and protections should be analysed from the perspective of the key employer functions – from hiring workers to setting their rates of pay (Adams-Prassl and Risak, 2016[6]). The outcome of such an approach would be that employment law obligations are spread across multiple legal entities, rather than ascribed to a single employer.

There has been some discussion about whether the regulation of temporary agency work (TWA), which also feature multi-party (or triangular) employment relationships, could serve as a model for the regulation of platform work (OECD, 2019[7]). In TWA, the employment relationship is generally between the worker and the agency, and the latter is responsible for ensuring that labour law is complied with. It is not clear to what extent the TWA experience might be a useful example for regulating platform work (Lenaerts et al., 2018[8]) – although the TWA model seems to have been accepted by many platforms in Sweden (Söderqvist, 2018[9]) and several platforms worldwide have taken the initiative to treat their workers as employees (Cherry and Aloisi, 2017[10]).

Even when classified as self-employed, some platform workers share some characteristics of employees. They experience some form of dependence and/or subordination in their working relationship. Moreover, some platform workers may be facing a power imbalance vis-à-vis the platform operator, which could result in lower pay and worse working conditions than would be the case under market conditions. When classified as self-employed, platform workers do not generally benefit from the same legal protections as employees in terms of collective bargaining rights, social protection, and access to training. In all these fields, inadequate and uncertain evidence hampers effective policy responses.

Collective bargaining can give platform workers more say on their working conditions (while tailoring solutions to the sector or occupation), countering power imbalances in relationships between firms and workers (Lane, 2020[11]). However, the standard approach in antitrust enforcement has often been to consider all self-employed workers as undertakings, and therefore any collective agreement reached by platform workers as a cartel.6

Countries have followed different paths for granting collective bargaining rights to the self-employed. In some cases, regulators and enforcement authorities have taken a case-by-case approach to avoid a strictly procedural analysis of cases involving those workers with little or no bargaining power and exit options. In several countries (e.g. in France, Italy, Spain, etc.), independent unions of platform workers are already de facto negotiating working conditions for their members even if they are classified as self-employed, without any intervention from national antitrust authorities.7 Moreover, some independent unions have been created (e.g. in Italy and the United Kingdom), especially in the private hire or food delivery sectors. Unions’ engagement with platforms on behalf of non-standard workers has led in some cases to the signature of collective agreements as in Sweden and Denmark.

There are concerns about imbalances of power and evidence that some platform workers earn below the living wage (ILO, 2018[12]). For standard employees, legally binding minimum wage can help address in-work poverty, and many studies find that small increases in the minimum wage from a moderate level have no negative employment effects. In jurisdictions where a minimum wage exists, significant difficulties surround extending minimum wage legislation to platform workers. These include measuring what counts as work (i.e. should platform workers be paid for the time that they have an app open and/or the time they spend waiting/searching for tasks?) and how to deal with work carried out across national borders. In practice, some jurisdictions around the world have extended the right to a minimum wage to some categories of platform workers.8

Traditional concerns around working time revolve around the issue of excessive working hours. Labour legislation usually contains rules limiting working hours and requiring periods for rest and recuperation, including weekly rest and paid annual leave. Moreover, in the case of certain micro-task platforms, workers spend as much time searching for tasks as they do in performing them (Kingsley, Gray and Suri, 2015[13]). Some platforms have introduced their own working hour limits (e.g. Uber requiring drivers to rest for 6 hours after driving for 10 hours continuously in the United Kingdom) and workers have adopted their own informal practices such as daily routines and quota setting to manage their time (Lehdonvirta, 2018[14]). Data collected through platforms can help in monitoring working time. In practice, extending working time protections to platform workers faces several obstacles: many platform workers have several clients/employers at any particular point in time, implying that monitoring overall working time (and allocating responsibility) may be very difficult if not impossible.

When workers have employee status, employment protection legislation usually protects them against breaches of contract obligations on the part of employers, including remedies for unfair dismissal and wage theft. In the case of platform workers, wage and working conditions are often set unilaterally by the platform (or the intermediary) or by the requester (i.e. the individual or company who posts tasks), with no scope for individual workers to negotiate any of the conditions that must be accepted in order to begin or continue working.9 Similarly, the terms of conditions of digital intermediation services often establish that the platform can deactivate a worker’s account without providing a justification, sometimes even without previous warning – see e.g. (Kingsley, Gray and Suri, 2015[13]).10

Many platform activities are in the transport sector, where the risk of road accidents is high. Evidence suggests that the arrival of ridesharing is associated with an increase of 2-3% in the number of motor vehicle fatalities and fatal accidents as a result of increased congestion and road utilisation (Barrios et al., 2018[15]). There are also risks associated with online work – both physical and psychosocial, such as visual fatigue; musculoskeletal problems; work-related stress; chronic job and income insecurity; and isolation. Again, the question of employment status is critical here as Occupational Safety and Health regulation often applies to employees only.11

In some cases, platform operators have made commitments to improve working conditions, sometimes based on the understanding that their compliance with these commitments cannot be used to presume an employment relationship.12

Platform workers receive very little training and can be stuck in a low-productivity trap, or may be at risk of deskilling. Some governments have started to inform jobseekers of new opportunities in the platform economy, while at the same time ensuring the quality and sustainability of their work.13

As many platforms capture information about the payment exchanged for services (in addition to acting as an intermediary for these payments, in some cases), some countries have attempted to take advantage of this feature in order to tackle the underreporting of income for services carried out through platforms and to bring work traditionally performed informally into the formal economy.14

In emerging economies where informal employment has a high incidence, the platform economy may constitute an opportunity for many workers to formalise, since it can reduce the costs of formalisation and improve monitoring of economic activity through the digitalisation of transactions (Alonso Soto, 2019[16]). In practice, some platforms are already playing a role in facilitating access to social protection for their workers. For example, in Indonesia, GoJek offers help to its motorcycle taxi drivers to subscribe to the government health insurance program, while Grab Bike motorcycle taxi workers are automatically enrolled in the government's professional insurance programme.

Many types of platform work are, by nature, “cross-border”. The intermediation of digital platforms allows workers to be based anywhere in the world to carry out tasks and offer services to clients – private individuals or corporate entities – located in a different geography. This only applies to so-called “online” platform work (as opposed to “on-location” platform work, like ride-hailing and food-delivery), which ranges from micro-tasks such as image recognition to more complex services like software-development, graphic design, translations, social media and content editing.

In practice, around 80% of globally commissioned online work is done for clients based in OECD countries, whereas only 20% of the workers who carry out these tasks are based there. This lopsided picture poses significant challenges in terms of labour market fairness and efficiency. Given the considerable difference in the costs of living and working outside OECD countries, outsourcing online tasks may lead to unfair competition between workers based in different geographies. This may lead to a race-to-the-bottom in working conditions, to the divergence of standards in social policy provisions and to uncertainty as regards the jurisdiction of law and standards applicability across different geographies.

Since the emergence of online platforms, there have been several attempts to estimate the number of platform-mediated workers. Initial attempts made use of existing data sources, combined with strong assumptions, and led to varying estimates of the size of the platform economy. Since then, official statistical agencies have begun to introduce questions on platform workers into Labour Force Survey supplements and Internet Usage Surveys, with wide variation in results. These efforts highlight some of the difficulties in measuring platform-mediated workers (see Chapters 2 and 3).

First, a key problem of surveys is to ensure that respondents understand the meaning of platform-mediated work. The appropriate method depends on the research objectives, the resources available, and the trade-offs faced by researchers and statistical agencies. Many respondents report being platform-mediated workers due to the poor definitions presented to them, e.g. when asked whether ‘they make use of a computer or mobile app in their job’. Adjusting for obviously incorrect responses can considerably reduce the estimated number of platform-mediated workers.

Second, there are inconsistencies across countries in how platform-mediated workers are measured. For example, some surveys do not differentiate between capital and labour platforms. The vast majority of surveys use the last 12 months as reference period, whereas others use a single reference week in order to be consistent with labour force survey results.

The use of some big-data sources can allow researchers to improve their research question as new platforms enter the market. However, methodologies that rely on web-scraping may have problems of consistency over-time as platforms are added, or dropped, from the list of platforms that are monitored while other platforms may be created or cease to exist. In addition, use of administrative data is likely to be very limited due to differences in administrative systems across countries.

Finally, while the use of official surveys such as Labour Force Surveys is likely to provide accurate and robust estimates, problems of sample size reduces their suitability for gaining insights into the characteristics of platform-mediated workers. Most estimates of the number of platform-mediated workers are in the range of 0.5% to 2% of total employment. Even though sample sizes of labour force surveys are typically large, they will lack precision about specific characteristics of small groups in the population, such as gender or occupation.

Although the problem of sample size can be overcome through administrative data (e.g. social security or tax data), these have their own shortcomings that affect the measurement of platform work, e.g. workers may be below VAT reporting thresholds. Partnerships between government agencies and online platforms to improve tax collection have the potential to improve administrative data sources.

There are other emerging sources such as data compiled by platforms themselves or ‘platforms of platforms’ that can be used as stand-alone sources or to complement results from survey and administrative data sources. Although there are significant risks that obtained data is biased, this is typically the case of data on the demand for platform jobs and the supply of platform workers produced by new data compilers such as AppJobs Institutes (https://www.institute.appjobs.com/).15 For instance, recent studies have analysed trends of selected types of platform workers before and after the Covid-19 outbreak, and the gender and/or age distribution of employment in the gig and traditional economies by industry and cities. Other reports have focused on the challenges posed by the increased use digital platform employment for the ‘future of work’. Data on gig jobs and gig workers, including earnings and hours worked, are made available to researchers worldwide for studying the platform economy and the future of work. While there are several issues related to the nature of data compiled by platforms on digital platform employment and work and their representativeness, the availability of timely and granular data provide clear advantages for data analysis and policy developments on the future of work in a foreseeable context of growing incidence of platform employment.

A paucity of information about the prevalence of platform work and the characteristics of the individuals engaged in it hinders the development of adequate policy. While existing labour force surveys and household surveys provide valuable information on self-employment, fixed-term and part-time work, they have a hard time in identifying platform workers. While other types of statistical sources (such as the COLLEEM survey) as well as data directly provided by platforms (e.g. transaction records such as those submitted by platforms to tax authorities) can provide much needed information, official statistics have a very important role to play in providing the evidence needed to address the broad range of policy questions described above. The remainder of this Handbook focuses on concepts and definitions that statistical offices and other data producers could use to measure platform employment, while also reviewing existing data collection instruments and designing model questionnaires accordingly.


[6] Adams-Prassl, J. and M. Risak (2016), Uber, Taskrabbit, & Co: Platforms as Employers? Rethinking the Legal Analysis of Crowdwork.

[16] Alonso Soto, D. (2019), “Technology and the future of work in emerging economies: What is different?”, OECD Social, Employment and Migration Working Papers.

[15] Barrios et al. (2018), “The Cost of Convenience: Ridesharing and Traffic Fatalities”, https://doi.org/10.2139/ssrn.3271975.

[17] Bloomberg.com (2020), Uber Drivers, Pizza Delivery Workers Get Lift From EU’s Vestager.

[10] Cherry, M. and A. Aloisi (2017), “Dependent Contractors in the Gig Economy: A Comparative Approach”, American University Law Review, Vol. 66/3, http://digitalcommons.wcl.american.edu/aulrAvailableat:http://digitalcommons.wcl.american.edu/aulr/vol66/iss3/1 (accessed on 8 January 2019).

[12] ILO (2018), Digital labour platforms and the future of work: Towards decent work in the online world.

[13] Kingsley, S., M. Gray and S. Suri (2015), “Accounting for Market Frictions and Power Asymmetries in Online Labor Markets”, Policy and Internet, Vol. 7/4, pp. 383-400, https://doi.org/10.1002/poi3.111.

[11] Lane, M. (2020), Regulating platform work in the digital age, OECD Publishing, https://doi.org/10.1787/181f8a7f-en.

[14] Lehdonvirta, V. (2018), “Flexibility in the Gig Economy: Managing Time on Three Online Piecework Platforms”, New Technology, Work and Employment, Forthcoming, https://ssrn.com/abstract=3099419.

[8] Lenaerts et al. (2018), Online Talent Platforms, Labour Market Intermediaries and the Changing World of Work.

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

[4] OECD (2018), International Migration Outlook 2018, OECD Publishing, Paris, https://doi.org/10.1787/migr_outlook-2018-en.

[5] Pesole, A. (2019), How to quantify what is not seen? Two proposals for measuring platform work, European Commission.

[1] Pesole, A. (2018), Platform Workers in Europe, Publications Office of the European Union, Luxembourg, https://doi.org/10.2760/742789.

[9] Söderqvist, F. (2018), Sweden: will history lead the way in the age of robots and platforms? - Policy Network, Policy Network, https://policynetwork.org/opinions/essays/sweden-will-history-lead-way-age-robots-platforms/ (accessed on 12 March 2020).

[3] Urzi Brancati, M., A. Pesole and E. Fernandez Macias (2020), New evidence on platform workers in Europe, Publications Office of the European Union, Luxembourg, https://doi.org/10.2760/459278.

[2] von Der Layen, U. (2019), , https://ec.europa.eu/commission/sites/beta-political/files/political-guidelines-next-commission_en.pdf.


← 1. This chapter draws from the “Regulation of platform” policy note of the OECD Going Digital toolkit (Lane, 2020[11]).

← 2. The countries covered are Croatia, the Czech Republic, Finland, France, Germany, Hungary, Ireland, Italy, Lithuania, the Netherlands, Portugal, Spain, Sweden, Slovakia and Romania.

← 3. According to the resolution adopted by the 19th International Conference of Labour Statisticians (ICLS), employment “comprises all persons of working age who during a specified brief period, such as one week or one day, were in the following categories: a) paid employment (whether at work or with a job but not at work); or b) self-employment (whether at work or with an enterprise but not at work)”, https://www.ilo.org/wcmsp5/groups/public/---dgreports/---stat/documents/normativeinstrument/wcms_230304.pdf.

← 4. Those data refer to platform workers who only perform one type of task, as those performing more than one type of tasks (i.e. delivery and micro-task) use more than one platform by default.

← 5. The problem, however, is not limited to the platform economy – many hairdressers, plumbers, and gardeners have faced similar challenges in the past. In some cases, the issue may be that these workers are falsely classified as self-employed in order to avoid regulation, or to access preferential tax treatment.

← 6. The recent plight of rideshare workers in Seattle illustrates this point. In order to help the wages and working conditions of drivers on ridesharing platforms, the city introduced an ordinance to allow the workers to collectively bargain with ridesharing companies. However, the law was challenged in court because it violated anti-trust laws.

← 7. The risk associated with this route is that it potentially creates uncertainty since it could be reversed without any legislative reform. However, this type of approach may receive further support from the European Commission in future. In March 2020, the European Commissioner for Competition said that she was examining whether the EU could help “people who work in a weak negotiating position” by giving “sort of European level guidance as to how to allow people to organize” without it being seen “as a cartel” (Bloomberg.com, 2020[17]).

← 8. Since January 2018, for example, New York City has imposed a minimum wage for Uber and Lyft drivers.

← 9. For example, on certain micro-task platforms, requesters can refuse completed tasks without providing a reason, in which case the worker receives no pay – see e.g. (Kingsley, Gray and Suri, 2015[13]).

← 10. In practice, there are precedents of social partners who have played a role in establishing simplified dispute resolution systems for the platform industry. For instance, three German platforms and the German Crowdsourcing Association drafted, in 2015, a Crowdsourcing Code of Conduct and established in 2017, in conjunction with five other platforms and IG Metall, an “Ombuds office” to enforce the Code of Conduct and resolve disputes between workers and signatory platforms – see (ILO, 2018[12]).

← 11. In France, the legislator has granted certain rights to platform workers through the August 2016 El Khomri law (or loi de travail) on labour, modernisation of social dialogue and securing of professional careers. Specifically, in cases where the platform determines the characteristics of the service provided and where the worker earns more than EUR 5 100 per year through the platform, the platform must provide reimbursement for insurance against occupational accident and illness.

← 12. In France, the 2019 Orientation des Mobilités law introduced the possibility for platform operators to draw up a social responsibility charter with a certain number of guarantees for workers.

← 13. For instance, the Finnish Public Employment Service has integrated a pilot called “New Forms of Work and Entrepreneurship”, into its digital job-market platform (Työmarkkinatori) to offer opportunities to jobseekers in terms of new forms of work and entrepreneurship, by linking them to invoicing companies and digital job mediation platforms. The Israeli Ministry of Labour and Social Affairs offers training in digital skills in order to allow workers to take advantage of opportunities in the platform economy. It operates a few small pilot programmes targeted at workers in new forms of work. One of these offers training to particular groups (people with disabilities, Arab women, ultra-Orthodox) on using online trading platforms and making a living on the global online market.

← 14. In order to tackle the underreporting of income, some countries have taken some policy initiatives. In Estonia, passenger transport platforms share information on the financial transactions between customers and drivers with tax authorities so that the tax authorities can prefill drivers’ tax forms. Since 2016, in Belgium, tax advantages (i.e. 10% income tax instead of 33%) have been granted to workers who earn under EUR 5000 annually through officially recognised platforms that withhold taxes at source and report earnings to tax authorities. As well as ensuring that taxes are paid, this preferential tax treatment was designed to incentivise side work in the platform economy.

← 15. AppJobs.com is a platform of platforms that has 1.3 million members in 45 countries. Data are available for any required frequencies (daily, weekly, monthly or annual) and individual characteristics (gender, age, etc.). They enable to monitor the demand for gig jobs and supply of gig workers by type of freelancing and gigs, industry, cities and region.

Metadata, Legal and Rights

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. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

© OECD/ILO/European Union 2023

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