Since the emergence of platform labour economy (with the largest platforms such as Amazon’s Mechanical Turk and Uber), research in this field has focused on two broad directions:
- Studies aiming to define, classify and estimate the current and predicted future size of the platform labour economy; and
- Studies on specific aspects of platform work, such as working conditions, motivations of platform workers, and platform workers’ legal labour market status.
In both areas, a plurality methodological approaches have been applied. However, each of them faces a number of shortcomings.
For example, to understand the prevalence of platform work, a number of surveys have been conducted in various countries. However, they suffer from several issues. First, robust surveys with probability samples are an extremely costly endeavour. Online panels have been used instead, but such a sampling approach cannot guarantee precision. Second, designing questionnaires is challenging too – platform work is a very diverse and constantly evolving phenomenon. Very few survey questions about it are relevant to all groups of remote and on-location, highly and low-skilled workers in different countries.
Although in-depth interviews with platform workers can allow the researchers to formulate better questions and reveal additional nuances, the data collected this way is even less generalisable.
Another approach to understand the prevalence of platform work could be asking platforms themselves. However, our attempts have illustrated an already well-known tendency of platforms to avoid sharing their data. For example, some platforms from both on-location and remote service sectors consider that information on the numbers of workers that they have is a commercial secret.
The automated data collection methods, such as web-scraping, then could present another solution. It has been employed, for example, by the Oxford Internet Institute researchers to measure the global remote platform work economy[1]. Data presented on worker profiles on a number of platforms can also be a good source of information on worker profiles, skills and earnings. However, this again faces a number of challenges. First, most platforms do not appreciate external parties crawling their websites, and implement various measures to block the scraping bots. Also, the types of data that can be scraped is restricted by the GDPR. Finally, while this approach is at least in theory feasible for studying remote platform work, it is not appropriate for on-location platforms providing taxi or delivery services. Their public interfaces just do not provide worker information.
Therefore, a combination of several approaches has proven to be the best way to go about studying platform workers. You can see the results of such an excercise in the Eastern Partnership countries in the upcoming reports on Online and Platform work in this region.
Please log in or sign up to comment.