Work and Wages of Data Labelers in Kenya

Abstract

This research report is based on a survey conducted in cooperation with the Data Labelers Association in Kenya. The goal was to gain empirical insights to fuel a better informed debate in their mission for decent conditions for data workers. The survey was distributed across their membership, resulting in 57 completed surveys.

Key Insights

• Almost half of the data workers work for one platform, one third works for a BPO (Business Processing Outsourcing), whereas two in ten work for multiple BPOs or platforms.
• Only four in ten data workers say they have a contract with the platform/BPO.
• Labeling images is the most frequently mentioned task, followed by tagging photos, videos, or keywords.
• Almost half of workers work 40 hours per week, but the remaining group reports working between 4 to 70 hours per week.
• When asked about the definition of overtime, one-third say that overtime hours are paid as normal hours, one-fourth are paid with a premium and one-fifth are not paid at all. Half of the workers indicate that their platform/BPO does not have an overtime policy.
• Data workers' earnings are very volatile. Compared to the previous month, total earnings remained stable for only one out of seven workers.
• On average, female workers earn Ksh 21,849.72 and male workers earn Ksh 26,106.83 per month from their main platform/BPO.
• When asked whether their total monthly earnings are enough to cover living expenses, only one in five female workers and one in four male workers confirm this.
• Only one in twenty female workers and one in ten male workers are (highly) satisfied with their earnings.
• Medical insurance is the most frequent social security benefit available to data workers, one in five reports being covered by this insurance.
• Most workers report always working from home.
• Almost one in five workers is covered by a collective agreement (17% F, 17% M). However, almost all workers confirm that coverage is important.
• The survey occurred when many data workers were experiencing layoffs. Not surprisingly more than eight out of ten data workers worry about their job security, male and female workers equally, while six out of ten find their job mentally exhausting.