If you work long enough, you’re liable to have a crummy boss. Sometimes they give unclear directions, or lack empathy, or maybe micromanage every move you make in the workplace. More recently, sometimes they’re not even human. That is the case for millions of gig workers across the globe and growing number of employees in more traditional job categories. They are bossed by bots. Managed by machines. Supervised by software. The future of work that many envision is humans replaced by robots and overflowing unemployment offices. But it appears those robots could instead leapfrog us in the pecking order and land in the executive suite.
In theory, the arrangements should be perfect. The platforms used by ride hailing services or other on demand work companies are meant to provide the most efficient mechanism for matching consumers with work providers. But according to researchers at Colombia University and CUNY, the companies employing these “algorithmic management” systems often overstep their bounds by seeking to control the behavior of workers. This leads to increased anxiety for the workers subjected to such manipulation. That anxiety leads to burnout out or even rebellion tactics from the workers. Thus resulting in worse outcomes for the very consumers they set out to serve. Such behaviors have potential to negatively impact the business operations of the companies behind the algorithms. Instead of perpetuating the negative cycle, tech companies have the ability, or some would say imperative, to better implement algorithmic management practices.
Whether installing cable for a telecom conglomerate or selling memorabilia via an ecommerce platform, workers subjected to algorithmic management are struggling more and more with anxiety. Imbalanced power dynamics are nothing new to the workplace, but the tech behind algorithmic management has upped the ante. The feedback workers receive are constant and have immediate impact on their abilities to do their jobs. Too many negative reviews and a driver is given less assignments via the algorithm. Fail to lower prices enough and customers volley a sea of complaints that buries an online seller. In the traditional work setting, managers have the discretion to make accommodations for struggling workers. The laser focused algorithms lack such empathetic inputs and simply eliminate low performers or pressure them into leaving. If algorithmic management is to play a significant role in the future of work, practitioners must work to improve its implementation and create better work atmospheres for their workforce. Because when those workers reach their breaking points, they don’t always leave.
Workers fed up with algorithmic management find new ways to rage against the machine. Ride share drivers in the Washington DC area have implemented a plan to do just that. In a coordinated effort, all drivers shut off the app just as passengers are leaving their flights and searching for rides. The lack of availability causes a surge in prices. The coalition of drivers then log back into the app and capture the additional revenue from increased prices. This behavior is their response to years of pay cuts in spite of billion dollar revenues enjoyed by ride hailing companies. Elsewhere, food delivery app workers have chosen to outright protest their “employers.” Despite their “contractor” status and the advertised freedom to work their own hours, delivery app workers are nudged to take on certain jobs at certain times if they want access to the best paying gigs. As this leads to de facto working hours, workers feel entitled to the benefits that come along with traditional work arrangements like insurance and paid time off. In protest of the working conditions, some have chose not to work at all. As delivery companies compete for market dominance, they must also consider improved “employee” relations to mitigate disruptions of service. While many rely on algorithmic management practices to quickly scale up operations, there are approaches they can take to do so while maintaining employee friendly work environments.
One way to implement this in the right way is buy bringing on an Algorithm Bias Auditor. We introduced this role as part of our series on the Jobs of the Future. When most think about bias in algorithms, they consider gender or race/ethnicity discrimination. While those factors are certainly significant problems, bias can also be wielded against certain classes of workers within an organization. In the case of companies using algorithmic management, they must look to employ leaders that can assess how algorithms may be negatively impacting their own workforce. But the importance of ethics and humanity in the practice of algorithmic management extend beyond just one position. The entire platform must be designed with best practices for the well-being of workers in mind. The current “blackbox” approach to algorithmic management creates the most tension for workers. They don’t know when changes will come or why. And they often don’t understand exactly how they are evaluated or compensated. Practitioners must build trust within these systems first and foremost. With data collected at every level of interaction, the jobs can feel more like surveillance then work. Yet, all that data is seldom shared. Workers desire trustworthy employers and transparency about work processes.
Much of this can be accomplished through exercises that emphasize empathy and understanding. Gig economy workers often complain of feeling dehumanized by the work arrangements. Those programming the apps often think more of the technology undergirding them and the profits they make than the actual people that carry out the physical labor within the systems. The best way to alleviate this gap is to implement ride along programs or require all workers to take part in the labor intensive aspects of the company with regularity.
Algorithmic management helps companies to scale, but also presents a number of long term challenges. It leads to greater levels of anxiety experienced by workers that feel constantly surveilled. It also creates circumstances in which those workers feel the need to “game the system” with behaviors that may negatively impact customers, the company, or even workers themselves. At present, algorithmic management is largely relegated to gig economy work, but the practice has already been adopted by larger companies that employ maintenance workers. Soon enough, even those of us with white collar careers are likely to experience some form of algorithmic management. Before this approach becomes more widespread, leaders must work to craft it with more ethical and people-friendly components ensure a future of work that works for all.