Article: Monday, 19 December 2022
Lots of people have switched to working on online labour platforms in the past couple of years, and they are likely to become more important as platforms become more advanced. They have an important role in developing and training AI, and there are several different types. Each kind suits a different type of worker – So can online labour platforms make themselves more effective at attracting and keeping the kind of workers they need? Assistant Professor Dr Olga Slivko of Rotterdam School of Management, Erasmus University (RSM) studied some of the factors influencing people using microtasking platforms to earn their living. The paper, Unemployment and Online Labor - Evidence from Microtasking will be published in MIS Quarterly.
Microtasks are small, well-defined jobs such as video screening, transcription, or picture matching, the kind of tasks that are useful for preparing training data for artificial intelligence (AI) and machine learning applications. They can be performed online and require only basic quantitative or reading skills. But during the period of the study – 2011-2015 – microtasking platforms didn’t see the same increase in participation alongside increasing unemployment rates compared to other online labour platforms.
Microtasking appeared to be most popular in areas where there was higher unemployment and more low-skilled workers, but still, the retention rate for workers was lower than might be expected. Dr Olga Slivko’s co researchers were Dr Ulrich Laitenberger of Institut Polytechnique de Paris, Steffen Viete of KfW, Dr Michael Kummer of University of East Anglia, Dr Kathrin Borchert of the University of Würzburg, and Prof. Matthias Hirth of the Technischen Universität Ilmenau.
They wanted to understand how unemployment makes people choose microtasking so platform managers could improve their infrastructure and address economic challenges more effectively. The ultimate aim is for microtasking firms to be able to provide more value to their AI clients.
The researchers analysed participation on microtasking platforms compared to unemployment rates from the period 2011 to 2015 using internal platform data from a large US-based online labour market (OLM) Microworkers.com, and compared it to unemployment data from the US Bureau of Labor Statistics. They found that higher unemployment causes more people to join the platform and complete tasks, but the rate of increase was only half that which had been documented for platforms requiring high-skilled freelancers.
The increase of participation on Microworkers happened predominantly in regions with a high proportion of white people between the ages of 45-64, particularly males, and in regions with a low proportion of college graduates and more low-skilled workers, illustrating that microtasking OLM platforms could become specialised at accommodating lower-skilled workers while leaving freelance online platforms to give opportunities for more highly skilled labour.
Dr Slivko points out the two main conclusions from the study:
“First, platform stakeholders can use our insights to assess the growth of their platforms in their own current economic conditions,” she said. “Our research points out the circumstances in which OLM for microtasks can successfully attract low-skill workers. The size of the effect compared to other online freelance platforms indicates that low-skilled workers may find the microtasking online labour opportunity less attractive, or it might be that they are simply less informed about this opportunity. From this, platform operators could develop strategies for better targeting this population segment to enhance the growth of their platforms.
“Second, low retention rates suggest that workers did not find microtasking attractive enough to keep doing it over a long period. Although microtasking offers great freedom, flexibility, and open access to work, platform managers might need to pay attention to their portfolio of tasks if they want to attract more educated part-time microworkers.”
The findings might also have implications for designing labour policies too. “Our findings suggest that OLM for microtasking can help in reducing regional mismatches between labour demand and supply – and wanting to avoid migration from regions that offer few job opportunities. Farsighted regulators could develop strategies for connecting online labour markets with more traditional forms of labour. OLM for microtasking could help in reducing interregional inequality and contribute to growth.”
Science Communication and Media Officer
Corporate Communications & PR Manager